Io?z X A woi MAIM CIUMV FITD'if Indonesia The Transmigration Program in Perspective EcY .. ...l. .....t- .... - ..:!::..-:: ..... .-, . . . ..... . ::...: . ... . ... ... .. C' lA .1 ii C:::S.:; i i:1sS: Si . - i:.171:11:i i111± 7 iCL J- j ;: I ; ' ' -' :.A..i - .f'l ¼ :< - ' i.J..' i . i i} A WORLD BANK COUNTRY STUDY Indonesia The Transmigration Program in Perspective The World Bank Washington, D.C., U.S.A. Copyright (( 1988 The World Bank 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing July 1988 World Bank Country Studies are reports originally prepared for internal use as part of the continuing analysis by the Bank of the economnic and related conditions of its developing member countries and of its dialogues with the governments. Some of the reports are published informally with the least possible delay for the use of governments and the academic, business and financial, and development comnmunities. Thus, the typescript has not been prepared in accordance with the procedures appropriate to fonnal printed texts, and the World Bank accepts no responsibility for errors. Any maps that accompany the text have been prepared solely for the convenience of readers. The designations and presentation of material in them do not imply the expression of any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member countries concerning the legal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to Director, Publications Department at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncormmercial purposes, without asking a fee. Pennission to photocopy portions for classroom use is not required, though notification of such use having been made will be appreciated. The most recent World Bank publications are described in the catalog New Publications, a new edition of which is issued in the spring and fall of each year. The complete backlist of publications is shown in the annual Index of Publications, which contains an alphabetical title list and indexes of subjects, authors, and countries and regions; it is of value principally to libraries and institutional purchasers. The latest edition of each of these is available free of charge from the Publications Sales Unit, Department F, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66 avenue d'lena, 75116 Paris, France. Library of Congress Cataloging-in-Publication Data Indonesia : the transmigration program in perspective. p. cm. -- (A World Bank country study) ISBN 0-8213-1092-5 1. Rural'development projects--Indonesia. 2. Land use, Rural- -Indonesia. 3. Migration, Internal--Indonesia. 4. Human settlements--Indonesia. I. International Bank for Reconstruction and Development. II. Series. HN710.Z9C6225 1988 307.1'4'09598--dc19 88-20533 - iii - PREFACE The World Bank periodically prepares country sector reports on important economic sectors and development programs. These reports provide the basis for the Bank's dialogue with Government officials on the policies, investment priorities and financial requirements of important programs. Distribution of these documents is normally restricted to official representatives of member countries. However, where the issues are of general interest, and where the authorities of the country agree to waive normal restrictions on distribution, such documents can be made available to a wider audience. Such is the case with The Indonesian Transmigration Program in Perspective. Between 1980 and 1986 the Indonesian Transmigration Program supported the movement of more than two million people from the densely populated "inner islands" of Java, Bali, and Lombok to the less populated 'Couter islands" of Sumatra, Kalimantan, Sulawesi and to a lesser extent Irian Jaya. An equal number of people are believed to have moved on their own, attracted in part by family and friends settled with government support. The transmigration program has been the largest government-sponsored voluntary resettlement scheme in the world. Because of the size of the program and because a number of settlements were opened in forest areas, the program attracted international attention, particularly from environmental groups concerned with deforestation and conservation. In 1986 new settlement under the transmigration program came to a virtual halt as the government reassessed its priorities. This was partially in response to declining oil revenues. In addition, the lower than expected agricultural production during the initial development period convinced the Government of the need for additional investments in existing sites to increase incomes and create the conditions for sustainable economic growth and equitable development. This report was initiated in 1985 as a review of resettlement in the third five-year plan (1979-1984). By that time, data were available which permitted an assessment of migrant incomes and economic returns to settlement models. Evidence was also available on the role of the program in employment generation, demographic change and regional development. Environmental and social issues were also reviewed. Among the conclusions emerging from the report were that economic rates of return and initial settler incomes were low and that the program was complex and difficult to implement. Spontaneous migrants were not adequately incorporated into program planning and the scale of the program heightened concern about the environmental impact and potential social problems. Yet, the study also found that economic and social prospects for settlers were superior to those in Java, and transmigration appeared to have contributed significantly to employment generation and regional develop- ment. In light of this analysis, the review, which was written prior to recent resource constraints in Indonesia, recommended that (a) the rate of new settlement should be slowed and the major emphasis should be placed on the - iv - consolidation and improvement of existing sites; (b) new settlements should be based on more productive farming systems; (c) institutions responsible for planning and carrying out the program should be strengthened; (d) increased attention should be given to accommodating spontaneous migrants on productive holdings; and (e) higher priority should be given to social and environmental concerns. These recommendations were accepted by the Government. Different groups will have very different views on the merits and demerits of transmigration, but during the preparation of this report, it became evident that many misconceptions exist in public discussion of the transmigration program. The Government felt that the report provides at least partial answers to some questions and, for this reason, agreed in mid-1986 to release the document for general distribution. The World Bank would like to express its thanks to His Excellency, Mr. Martono, the former Minister of Transmigration,, who provided the full support of his institution for this review; to Dr. Sayuti Hasibuan, Deputy for Manpower and Population Affairs in BAPPENAS, who provided guidance to the review team; and to Messrs. Soedjino Hardjosoetowo and Djoko Hartono who have worked especially closely with Bank teams. Thanks are also due to Mr. Soegito Sastromidjojo, in the Central Bureau of Statistics, who carried out the income survey. It is not possible to thank individually all of the officials and individuals who provided information, help and support, and without whose assistance this report would not have been possible. A reenl esalse trnmirn vilg inJab Prvne Suata 4,~~ JIM&~~ ~ ~ ~ jM * - -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ....... Wor recnthe fesaldsished transmigrant settlemeint.mb CURRENCY EQUIVALENTS US$1.00 Rupiah (Rp) 1,650 /a Rp 1 million US$606 GOVERNMENT OF INDONESIA FISCAL YEAR April 1 - March 31 WEIGHTS AND MEASURES 1 kilometer (km) = 0.62 miles (mi)2 1 square kilometer (km2) = 100 ha = 0.39 mi 1 hectare (1 ha) = 2.47 acres (ac) 1 kilogram (kg) = 2.2 pounds (lb) 1 metric ton 9t) 2,206 pounds ACRONYMS, ABBREVIATIONS AND LOCAL TERMS AARD - Agency for Agricultural Research and Development Bangda - Directorate General of Regional Development BAPPEDAs - Provincial Planning Agencies BPS - Central Bureau of Statistics CRIA - Center for Research in Agriculture DGE - Directorate General of Estates DGFC - Directorate General of Food Crops FAO - Food and Agriculture Organization Gabah - Unmilled rice GOI - Government of Indonesia HH - Household IFAD - International Fund for Agricultural Development Inner Islands - Java, Madura, Bali and Lombok JMT - Junior Minister for Transmigration Kabupaten - District Kecamatan - Subdistrict IUCN - International Union for Conservation of Nature and Natural Resources MOA - Ministry of Agriculture MOF - Ministry of Forestry MOT - Ministry of Transmigration /a After September 12, 1986. All analysis in this document is based on an exchange rate of Rp 1,100, the prevailing exchange rate when the report written. NTB - Nusa Tenggara Barat NTT - Nusa Tenggara Timur ODA - Overseas Development Authority (UK) OSG - Office of the Secretary General (in MOT) Outer Islands - Sumatra, Kalimantan, Sulawesi, Irian Jaya and a number of small islands PIR - Domestically financed estate projects PMU - Project Management Unit, provides inputs and labor compensation to farmers PRPTE - Government-financed tree crop development program PTP - Publicly owned estate Repelita - Five-year development plan (Repelita I, 1969-74; Repelita II, 1974-79; Repelita III, 1979-84; and Repelita IV, 1984-89) SCDP - Smallholder Coconut Development Project Siap - Budgetary carry-over Sisipan - Fill-in transmigration program near already settled sites SRDP - Smallholder Rubber Development Project Susenas - National Socio-Economic Survey Swakarsa - Directorate for Self-initiated Movement TSP - Trisodium phosphate fertilizer UNDP - United Nations Development Programme WWF - World Wildlife Fund Table of Contents Page No. EXECUTIVE SUMMARY A. Background ............................................................. xv B. Evaluation of the Transmigration Program to Date ........... . xxii C. Factors Affecting Future Settlement .............. ........... xxxiv D. Summary of Major Recommendations ..................... ... xl MAIN REPORT I. OVERVIEW ................................................ .... 1 A. Background ........................................... .-. ............. I B. The Repelita III Transmigration Program ................. 8 C. The Repelita IV Transmigration Program ................. . 13 II. MICRANT WELFARE ................................... *........................ 17 A. A Comparison of Income in Sending and Receiving Areas ........................................................... 17 B. Factors Determining Transmigrant Incomes .......... ...... 22 C. The Sustainability of Settlements ............ ........... 28 D. Other Survey Findings .. ................................ . 34 E. Summary and Conclusions ...................... .. .... 40 III. ECONOMIC AND FINANCIAL ANALYSIS OF TRANSMIGRATION FARM MODELS ...... a...... o ........ o.................. ................... 41 A. Assumptions Used in the Economic and Financial Analysis ......................... o ........... o .... 41 B. Economic and Financial Analysis of Food Crop Farm Models ............................._....... ........ o.o.......... 47 C. Economic and Financial Analysis of Tree Crop Models ..... 59 D. Summary and Conclusions ...................... ................. . 64 IV. THE IMPACT OF TRANSMIGRATION ... oo ......... .. 67 A. Employment Generation... ............................ . 67 B. Demographic Impact o. .....- 73 C. Impact on the Regions ........ ....... o-oo ....... o..o..... . 80 Dr. Summary and Conclusions. ........................ 90 This report was written by Gloria Davis and Helen Garrison. Background papers were prepared by George Baldwin, Ray Byron, Mike Douglass, Geoffrey Fox, Anthony Whitten and Ken Woodward. Jean-Pierre Baudelaire, Fateh Chaudhri and Manuel Zenick also contributed to the work. -x- Page No. V. SOCIAL AND ENVIRONMENTAL CONSIDERATIONS .................... . 91 A. The Impact of Transmigration on Local People ............ 91 B. The Impact of Transmigration on the Environment . . 96 C. The Case of Irian Jaya . ................................. 103 D. Summary and Conclusions .................. 105 VI. AGRICULTURAL PROSPECTS .................................... . . 107 A. Upland Sites .. . . ........................................ 107 B. Swamp Reclamation Sites .............................. . . . 114 C. Tree Crop Development . . . . ............................... 119 D. Summary and Conclusions ......... . . . 128 VII. OTHER FACTORS AFFECTING THE SCALE OF THE FUTURE PROGRAM ..... 129 A. Land Availability . . ..................................... 129 B. Spontaneous Migration .................................. . 134 C. Financial Constraints . .................................. 143 VIII. INSTITUTIONAL CONSIDERATIONS .............................. . . 147 A. Institutional Development Within the MOT ................ 148 B. Interagency Coordination . . .............................. 151 C. The Changing Role of the MOT .... 155 D. Summary and Conclusions . . ............................... 157 IX. WORLD BANK INVOLVEMENT IN TRANSMIGRATION .................... 159 X. SUMMARY OF MAJOR RECOMMENDATIONS IN THE MAIN REPORT ........ . 169 TABLES IN THE MAIN REPORT 1.1 Basic Demographic and Land Use Data ........................... 2 1.2 Sponsored Transmigration, 1950-1984 .......................... 3 1.3 Repelita III Settlement, by Year and Island ............ ...... 11 1.4 Settlement From April 1, 1984 through June 30, 1987 by Type of Migrant Family and Island ........................... 15 2.1 Comparison of Transmigrant Incomes with Incomes in Sending and Receiving Provinces ................................... 19 2.2 Transmigrant Reports on Incomes in Transmigration Areas Compared to those Prior to Transmigration ................. 20 2.3 Distribution of Household Incomes among Transmigrants and Nontransmigrants .......................... o ............... 21 2.4 Income and Income Distribution, by Farm Model and Repelita .... 23 2.5 Percent of Community Energy Requirements Met from Reported Production, by Farm Model and Repelitap ........el a......... 24 2.6 Annual Production and Market Sales of Rice for Sponsored Transmigrants by Farm Model, 1 9 8 5 ......................... 25 2.7 Source of Income by Farm Model and Repelita, BPS Survey ....... 26 2.8 Monthly Household Incomes from the IFAD and Euroconsult BPS and MOT Surveys .......................... ............. 28 2.9 Comparison of BPS and MOT Income Surveys ...................... 29 - xi - TABLES (continued) Page No. 2.10 Average Monthly Family Income and Annual Rice Production by Year of Arrival, Sponsored Transmigrants, BPS Survey ... 30 2.11 Sponsored Transmigrants' Current Rice Production and Amount of Land In Production Relative to Two Years Previous ...... 32 2.12 Number of Families in Repelita II and III Transmigration Sites on Arrival and in February 1985 ..................... 33 2.13 Monthly Household Income by Type of Transmigrant .............. 35 2.14 Migrant Perceptions of Conditions in Transmigration Sites ..... 37 2.15 Per Capita Consumption in Transmigrant and Nontransmigrant Households ................................................ 39 3.1 Development Cost Per Family for Food Crop Models in Upland and Tidal Areas ........................................... 42 3.2 Main Characteristics of Observed Food Crop Models . . 50 3.3 Agricultural Production of Main Food Crops by Year on Site as a Proportion of Total Household Energy Requirement by Farm Model ................................................ 51 3.4 Economic Rates of Return Under Different Economic Assumptions and Household Income at Full Development .................. 52 3.5 Economic Rates of Return to Agriculture and Off-farm Work for Upland Models if Infrastructure Costs Increase Due to the Remoteness of the Sites ................................... 54 3.6 Main Characteristics and Prevalence of Tree Crop Models ....... 59 3.7 Financial Development Costs for Tree Crop Models Per Household ................................................. 61 3.8 Economic Rates of Return to NES Tree Crop Models .. . 62 3.9 Economic Rates of Return to PMU-based Tree Crop Models ........ 63 3.10 Economic Rates of Return of Tree Crop Farm Models with Alternative Tree Crop Commodity Prices .................... 64 Figure 3.1 Sources of Annual Income by Farm Model and Repelita ..... 48 4.1 Labor Inputs for Land Clearing by Farm Model .................. 68 4.2 Tree Crop Labor Requirements for Land Clearing, Development, and Maintenance and Harvesting ............................ 70 4.3 Household Annual Wages, On-farm and Nonagricultural Earnings, and Days Worked Off-farm by Food Crop Farm Model .......... 72 4.4 Sponsored Transmigrants as Percent of Receiving Province Populations ............................................... 75 4.5 Spontaneous Migration as Measured by Language Spoken at Home ...................................................... 77 4.6 Demographic Impact of Transmigration under Alternative Scenarios .. 79 4.7 Transmigration Development Budgets in Receiving Provinces ..... 82 4.8 Sponsored Transmigrant Land by Province, 1985 ................. 83 4.9 The Contribution of Repelita III Transmigration to Incremental Rice Production by Province ................... 85 4.10 Transmigration Road Construction in Relation to Local Infrastructure ............................................ 87 - xii - TABLES (Continued) Page No. 5.1 Forest Resources by Province Compared to Land Allocated for Transmigration in Repelita III ........................ 98 6.1 Reported Annual Production of Unmilled Rice in Upland Sites ..................................................... 107 6.2 Distribution of Agricultural Income in Upland Sites, 1985 ..... 113 6.3 Families Settled in Swamp Reclamation Sites in Repelita I, II, and III ............................................... 115 6.4 Transmigrants Settled on NES/PIR Schemes ...................... 120 6.5 Bank Projections for New Planting, Replanting and Rehabilitation of Tree Crops for Smallholders ............. 125 7.1 Estimated Area, Carrying Capacity and Forestry Status of Potential Sites in Central Kalimantan ..................... 130 7.2 Swamp Areas in Indonesia ...................................... 132 7.3 Breakdown of Self-Initiated (Swakarsa) Families Counted in Repelita III .............................................. 135 7.4 Budgets for the Transmigration Sector and Expenditures for Repelita III and Early Repelita IV ........................ 144 8.1 Possible Organization of Proposed Special Teams to Address Management and Policy Issues .............................. 154 9.1 Bank Support for Transmigration Projects ...................... 160 ANNEXES 1. Transmigration Settlement During Repelita III and Early Repelita IV Table 1: Repelita III Settlement by Year and Province Table 2: Repelita IV Settlement by Type of Migrant Family and Province through June 30, 1987 2. Transmigrant Income Table 1: Sites Covered in BPS Survey Table 2: Comparison of Income, Net Value of Production and Expenditure in Transmigrant Communities and in Sending and Receiving Provinces Table 3: Source of Income by Farm Model and Location 3. Financial and Economic Analysis of Transmigration Models 4. Impact of Transmigration on Provincial Populations Table 1: Population and Transmigrants in Receiving Provinces Table 2: 1970-85 Transmigrants as Percent of 1980 Provincial and Kabupaten Population Table 3: Population by Place of Birth, Place of Previous Residence, Place of Residence Five Years Ago and Place of Present Residence - xiii - ANNEXES (continued) 5. Classification of Forest Land, May 1984 6. Planned and Actual Plantings for Ongoing and New Smallholder Tree Crops Development Programs - Repelita III and IV 7. Description of Transmigration Surveys Reviewed 8. Methodology for Demographic Analysis Contained in Chapter IV 9. Summary TORs for Phase II and III Studies MAPS Sumatra IBRD 19537 Kalimantan and Java IBRD 19538 Sulawesi and Maluku Islands IBRD 19565 Irian Jaya IBRD 19566 EXECUTIVE SUMMARY A. Background Introduction 1. Transmigration is a multifaceted economic and social program, which has had a significant impact on employment, migrant welfare, and regional development in Indonesia. 2. In response to the inequitable distribution of land and labor, resettlement programs were begun in Indonesia in 1905. Since 1979, however, the scale of the transmigration program has significantly increased. In the second five-year plan (1974-1979) about 52,000 families were moved, while in the third five-year plan (1979-1984) about 366,000 families (1.5 million people) were resettled under the sponsored program. Of these, about 300,000 families were moved from the overcrowded islands of Java, Madura, Bali and Lombok, while 65,000 were resettled within the receiving provinces. Govern- ment also identified 170,000 families that moved without assistance, although the actual number is believed to be much larger. Of the sponsored migrants, 62% went to Sumatra, 19% to Kalimantan and 14% to Sulawesi. About 16,600 families, 5% of the total, were settled in Irian Jaya. 3. The fact that transmigrants move voluntarily to pioneer settlements involving considerable hardship, attests to the economic pressures on poor farmers in Java and Bali and to the opportunities they perceive in the outer islands. The fact that farmers settled under the sponsored program were, on the whole, moved in a safe and orderly way is commendable. However, the scale and rate of acceleration of the program in Repelita III have also given rise to a number of problems in need of attention. This report is intended to evaluate the transmigration program and make recommendations for future development. The Country 4. The Republic of Indonesia, with a population of about 168 million in 1986, is the fifth most populous nation in the world. Over 100 million people live on Java, an island with about 7% of the nation's land. With population densities in Java at 774 people/km2 (roughly the same as Bangladesh), agricultural holdings are small and incomes are low. In a 1984 survey by the Central Bureau of Statistics (BPS), average household incomes in rural Java were about US$730 and about 40% of the pyyulation fell below a poverty line estimated at US$540 per family per year.- Other islands have more land (although generally of lower fertility), less dense populations (about 33 people/km2) and a major portion of the country's nonagricultural resources, particularly oil and timber. For these reasons, households in the main trans- 1/ Based on an exchange rate of Rp 1100/US$1.00, the prevailing rate when this report was written. - xvi - migration receiving areas (Sumatra, Kalimantan and Sulawesi) are more pros- perous and incomes average about US$1,000/family/year. Key demographic data are provided in Table 1. 5. The significance of these demographic facts cannot be overstated. Nearly two thirds of Java's farm families have less than one half hectare of agricultural land, a subsistence size plot, and in spite of an active and successful family planning program which has reduced the growth rate to 1.8% p.a., population pressure has led to densely crowded conditions in lowland rice producing areas. Farmers have moved onto steep slopes and into forest reserves where cultivation and erosion have caused environmental degradation, siltation of reservoirs and canals, and downstream flooding. Urbanization in Java's major cities is occuring at a rate of 4% per year, more than twice the rate of population growth, and major investments in urban infrastructure and services are required to serve an estimated 1.6 million new urban residents each year. Table 1: BASIC DEMOGRAPHIC DATA Est. 1986 Population % Farm households with Province/Island population Area density Less than Over (million) ('000 km2) (People/km2) 0.5 ha 2.0 ha The Inner Islands Java/Madura 102.2 132 774 63% 4% Bali/West Nusa Tenggara 5.8 26 226 46% 10% Total 108.0 158 685 62% 5% The Outer Islands Sumatra 33.0 474 72 27% 23% Kalimantan 8.0 540 15 19% 39% Sulawesi /a 13.5 264 51 25% 40% Other eastern islands /b 4.1 60 68 - - Irian Jaya 1.4 422 3 - - Total 60.0 1,760 34 26% 30% GRAND TOTAL 168.0 1,918 88 - - /a Includes the Moluccas. 7T Other eastern islands which have not been transmigration receiving areas. Source: Statistical Yearbook of Indonesia, 1984. Land holding data are extracted from the 1983 Agricultural Census. - xvii - Intensively cultivated area in upland Java. Hills in the background show the trees remaining from a regreening program. A- ~~~~~~~~~~ I ~ ~ ~ ~ S~~~~~~~~~~~~~~~~ V~~ *, .. _ '.e'','. , '.f,d,,~~~~.. ...... ................_"z.b, Aerial view of recently established transmigrant community in Central Sulawesi. - xviii - Development Objectives and the Role of Transmigration 6. Employment creation and poverty alleviation are among Indonesia's key development objectives and they have been addressed through a broad spec- trum of programs. Major investments have been made in irrigation, fertilizer production, agricultural research and smallholder tree crop development, and the mechanization of agriculture has been discouraged on employment grounds. Government has invested in labor-intensive rural works programs, village improvement, education and health, and the Indonesian family planning program has made solid and impressive progress in promoting fertility declines by raising contraceptive prevalence. Government has encouraged the growth of small-scale industries, but the growth of export industries has lagged due to strong competition from other developing countries, import restrictions in developed countries, and a complex policy and regulatory environment. 7. Resettlement has also played a role in Indonesia's efforts to cope with population growth, reduce deforestation in the inner islands, create pro- ductive employment and stimulate regional development. Resettlement from Java began in 1905. By 1980 about one million people had been resettled, and about two million people were in the outer islands as the result of sponsored trans- migration and associated population growth. Government's interest in moving people has been reciprocated by a desire on the part of poor people to move. Transmigration is voluntary, and in recent years the number of people registering for the program has exceeded the number who could be moved. Farming households choose to resettle to meet their subsistence needs and to provide opportunities to their children. The Settlement Program 8. Definitions. Transmigration refers to the movement of people from overcrowded areas of the inner islands to less developed areas of the outer islands. Sponsored transmigrants receive government support and move to se- lected sites. Spontaneous transmigrants receive little or no government sup- port and move to areas of their own choosing, generally to be near families or friends. Government sponsorship is an important element in transmigration because traditional land tenure arrangements in the outer islands make it difficult for individual migrants to secure legal rights to land. In addi- tion, by selecting and supporting migrants, transmigration authorities have been able to target critical areas in Java and Bali for recruitment and ensure that even the poorest farmers can move if they wish to. 9. Resettlement under the official transmigration program has been based almost entirely on smallholder agriculture. (Parallel programs to transfer manpower to the outer islands as industrial or estate laborers have not been a part of the official transmigration program). However, different farming systems have been emphasized at different stages in the development of Indonesia's resettlement effort. (a) Irrigated Schemes. Under the Dutch, most resettlement occurred on new large-scale irrigation schemes in the outer islands. Between 1950 and 1970 government investment in new irrigation was limited, due to financial constraints, but most sponsored settlers were moved - xix - to areas intended eventually for irrigation. Some of these settle- ments have since received irrigation works. (b) Swamp Reclamation Schemes. Swamps were first opened for rice production and settlement in the 1930s. In these areas, in theory, tidal action pushes fresh water up the rivers and canals and onto low-lying land, making regular innundation of rice fields possible. In practice, most recently opened areas depend on bunded rainfed rice production. (c) Rainfed Schemes. In the second five-year plan (1974-79) an agricul- tural package (including planting materials, fertilizer, credit and extension) was developed to support rainfed agriculture on the less fertile soils of the outer islands. This package allowed settlement in slightly undulating areas unsuited to irrigation and permitted an expansion of areas suitable for settlement. To differentiate these schemes from irrigation and swamp sites in low-lying areas, they are typically called upland schemes, even though areas selected for food crops must have slopes under 8%. (d) Cash Crop Schemes. Under the Nucleus Estate and Smallholder (NES) program begun in 1978, estate managerial capacity has been harnessed for the establishment of cash crops for local smallholders and transmigrants. Most NES schemes have been based on tree crops (oil palm, rubber and coconut), although sugar and cotton have also been produced. Most settlers on these schemes also receive a small plot for subsistence food production. Recently, there has been a growing interest in using project management units (PMUs) to help new and existing transmigrants establish tree crops and other cash crops on their land. 10. The Settlement Process. Transmigrants are recruited in rural areas of Java, Madura, Bali and Lombok. They must be married, of good character and have previous farming experience. Typically, poor households encourage one married son or daughter to transmigrate, and if successful, other relatives follow. Sponsored migrants are moved by plane or bus and on arrival they re- ceive a small house on 0.25 ha of village land and 0.75 to 1.0 ha of cleared land outside the settled area. Additional land (averaging 1.0 to 2.0 ha) is reserved for future development. Public facilities including schools and clinics are located in the village center. Subsistence supplies are provided for one year while the land is being tilled and crops are being established. If crops fail, the provision of supplies can be extended. Settlements are expected to be self-sufficient at the end of five years. 11. In addition to the food crop package, most settlers on rainfed schemes are provided with planting materials for minor tree crops (coffee, cloves) and small livestock (chickens, goats). In the early years of settle- ment, farmers usually confine their agricultural activities to their cleared land (1.0 - 1.25 ha) since the reserve land is not allocated and labor is limited. Where reserve land is directly behind the houselot or cattle are introduced, more land is cultivated and farms are more diversified. In general, however, improving or expanding the area under production requires - xx - the introduction of financial and technical support at a later stage and this is called second-stage development. The Repelita III Transmigration Program (1979-84) 12. Program Formulation. Pressed by increasing environmental degrada- tion on Java and growing employment concerns, and encouraged by an improved financial situation, Government significantly expanded the transmigration program in Repelita III and targets for this period were set at 500,000 families. With these high targets, the central issues in the design of the Repelita III transmigration program concerned the farm model used, program scale, and investment levels. 13. Government officials recognized that it was technically possible to resettle smallholders on tree crop schemes that provided relatively high incomes and rates of return. However, they also believed that such programs would be costly, would slow the settlement process, and would create income disparities between beneficiaries and nonbeneficiaries. For these reasons, and consistent with Government's view that transmigration had to be carried out on a relatively large scale to realize program objectives, it was agreed that settlement would be based initially on schemes supported by rainfed food crops. This model was adopted because Indonesia was, at that time, the world's largest importer of rice, and because annual crops were the quickest to establish, they promoted early self-sufficiency, and they cost less than other models. Tree crops were to be introduced at a second-stage. In order to speed settlement, sites were selected and cleared on a plan-as-you-proceed basis, and to give a quick start to migrants, land clearing was mechanized and 1.0 to 1.25 ha of land per family were cleared. 14. To reach its targets, Government also made a number of changes in the implementation of the transmigration program. Most important, a number of major activities were taken from the Directorate General of Transmigration (DGT) and given to agencies normally responsible for each sector. The Ministry of Public Works undertook site selection and land preparation, the DGT carried out transmigrant selection and resettlement, agencies within the Ministry of Agriculture provided agricultural inputs, and the Ministries of Health and Education were responsible for activities in their sectors. Mechanisms were also established to promote interagency coordination. In response to the favorable resource position at the time, funding of the program was increased and ceased to be a major constraint. 15. Achievements. Although there were implementation problems arising from high targets, sponsored settlement was increased from 52,000 families in Repelita II to 366,000 families in Repelita III. Of these, about 22,000 were local families who elected to join transmigration settlements, and 42,000 were spontaneous migrant families who had settled in watershed and forestry areas in Lampung province and were resettled within the province (see Table 2). About 80% of all families were settled in rainfed sites based mainly on food crop production, 18% in swamp reclamation areas, and 2% on tree crop schemes. The total cost of the program was Rp 1.6 trillion (US$2.3 billion), of which about US$107 million (5%) was disbursed by the World Bank. - xxi - Table 2: SPONSORED TRANSMIGRATION PROGRAM, 1950-84 Five- Total Local Resettled Total year families families families families Total Year plan moved /a /b settled people 1950-54 21,037 0 1,280 22,317 87,000 1955-59 32,114 0 128 32,242 134,000 1960-64 26,456 0 0 26,456 111,000 1965-69 21,633 0 0 21,633 92,000 1969-74 (Repelita I) 39,436 0 75 39,511 240,000 1974-79 (Repelita II) 44,484 7,600 0 52,084 465,000 Subtotal 185,160 7,600 1,483 194,243 828,000 1979-84 (Repelita III) 301,279 22,284 42,414 365,977/c 1,492,000 Total 486,439 29,974 43,897 560,220 2,320,000 /a Indigenous families settled in transmigration sites. 7T Resettlement of sponsored or spontaneous migrants within the province. 7i In addition, 170,000 families were identified that moved spontaneously. Source: Official Summary of Repelita III Transmigration Program 1985, Ministry of Transmigration. 16. Program Development. In the course of Repelita III a number of adjustments were made in the implementation of the transmigration program: (a) site selection and evaluation were improved by the introduction of detailed planning and the involvement of the forestry agency and provincial governments in the selection process; (b) limitations on the food crop farm model were gradually acknowledged, and efforts were initiated to improve food crop production and to shift partly to tree crop schemes; (c) the importance of spontaneous migration was recognized and a Directorate was formed to promote unassisted migration, although no formal program was carried out in Repelita III; and (d) after five years of experimentation with multisectoral implemen- tation, a new Ministry of Transmigration (MOT) was formed in March 1983 in response to difficulties of coordination. 17. Serious implementation problems also persisted. While the location of sites improved, detailed plans were not always available, land clearing was often of poor quality, road construction and maintenance standards were low, - xxii - and the supervision of contractors uneven. In the course of Repelita III, land for settlement based mainly on food crop development became increasingly difficult to find, and land for large-scale settlements in Sumatra was virtually exhausted. The provision of agricultural supporting services, including input supply, extension and credit, was inadequate; and no progress was made on programs to introduce tree crops to existing settlements. Perhaps most important, institutional arrangements for coordination, planning, budgeting, and monitoring and evaluation remained weak. The Repelita IV Transmigration Program (1984-1989) 18. Concerned with the possibility of growing unemployment in Java, the Government called for the settlement of 750,000 families in Repelita IV. Of these, some 400,000 families were to be settled under fully sponsored programs and the rest were to move without assistance. Between April 1984, when Repelita IV began, and June 1986, about 154,000 families were moved on the sponsored program, and about 150,000 families were identified who had moved without financial assistance. 19. In January 1986 the Government made significant reductions in all development budgets in response to declining oil revenues. The overall devel- opment budget for the FY86/87 was cut by 22% and the budget for transmigration was initially cut by 56% of the previous year's budget. In May 1986, the MOT budget was further reduced to 38% of the FY85/86 figure. In response, MOT reduced its FY86/87 settlement targets from over 100,000 fully sponsored families to 36,000 families. To meet the annual target, it was assumed that the balance of movement would be from unassisted migrants. B. Evaluation of the Transmigration Program to Date Demographic Impact 20. Sponsored Migration. From 1950 to 1984, the Indonesian Government moved about one half million families to the outer islands. (The distribution of migrants by island is shown in Table 3.) Prior to Repelita III, the major islands received about the same proportion of transmigrants as their share of the total population. Approximately 65% of all transmigrants were sent to Sumatra, slightly more than Sumatra's 55% share of the outer island popula- tion, while Kalimantan and Sulawesi received approximately the same share of transmigrants as their share in the outer island population. Since 1980, Kalimantan and Irian Jaya have received a somewhat greater share of trans- migrants as the proportion of transmigrants moving to the more crowded provinces of Sumatra and Sulawesi has declined. Kalimantan increased its share from 14% in the 1970s to 22% in the 1980s, and Irian Jaya increased from 1% to 4% during the same period. Between 1950 and 1984 about 19,000 families settled in Irian Jaya, about 3% of the total moved. 21. Spontaneous Migration. There is no direct record of spontaneous movement, but the 1980 census indicated that there were about 3.6 million people born in Java and living in the outer islands. About 1.1 million people from the outer islands were living in Java. This suggests an out-migration to in-migration ratio of about 3 to 1. The census also indicated that 868,000 - xxiii - people had moved from Java to rural areas in the outer islands during the 1975-80 period, while the number of sponsored migrants was about 250,000, a ratio of 2.5 spontaneous migrants to 1 sponsored. Causality cannot be inferred directly from these figures, but there is a strong association between the provinces to which sponsored migrants have been sent in the past and those to which spontaneous migrants later move. In total, some 7.3 million people in the outer islands in 1980 spoke an inner island language (Javanese, Sundanese, Madurese or Balinese), including about 6.7 million in rural areas, or 16% of the rural outer islands' population. Of these, perhaps 30% were sponsored migrants and their descendents. Table 3: SPONSORED TRANSMIGRANT FAMILIES SETTLED SINCE 1950 Sumatra Kalimantan Sulawesi/a Irian Jaya Total 1950/54 20,400 1,400 500 - 22,300 1955/59 28,900 2,600 700 - 32,200 1960/64 21,000 4,500 1,000 - 26,500 1965/69 16,500 2,100 2,700 300 21,600 1970/74 22,000 6,000 11,400 100 39,500 1975/79 33,000 11,000 9,000 2,000 55,000 1980/84 227,100 70,600 51,700 16,600 366,000 Total 368,900 99,200 77,000 19,000 563,100 Percent 65% 18% 14% 3% 100% /a Includes the Moluccas and other small eastern islands. Source: MOT documents. 22. Impact on the Inner Islands. Although sponsored migration in Repelita III moved only 1.5% of the total inner island population and 15% of the incremental population, the cumulative effect of movement at these rates is significant. If sponsored transmigratio were maintained at the high Repelita III levels through the year 2000 - and were accompanied by sponta- neous migration at a rate of about 2 spontaneous migrants to 1 sponsored migrant, the population of Java in the year 2020 would be 21.5 million people (13%) less than without the program and transmigration would have absorbed 24% of the incremental labor force. If settlement levels are reduced to 200,000 families in Repelita IV and V, decline to 100,000 families in Repelita VI, and the previous assumptions were made about spontaneous migration, by the year 2020 the population of Java would be 9% less than without the program and the 2/ About 300,000 families moved in Repelita IV and V declining to about 200,000 families in Repelita VI. - xxiv - incremental labor force would be 19% less. Sponsored migration alone at lower levels would reduce the population by 3-4% and absorb 7-8% of the incremental labor force. Employment Generation 23. About 500,000-600,000 permanent jobs were created by the Repelita III transmigration program, at a roughly estimated cost of US$3,000 to US$4,000 per job in rainfed food crop schemes, and US$3,500 to US$4,500 per job in tree crop settlements. This is higher than the cost of employment creation in services, but considerably less than the cost of employment crea- tion in industry, estimated at US$10,000-20,000 per job. These figures do not take into account the employment generated indirectly by transmigration investment. Reliable data on the temporary work generated by the development of transmigration sites are not available, but it is clear that the employment impact of the program was very large. Employment generated in Repelita III by land clearing in transmigration sites is estimated at 18 million work days, or the equivalent of 63,000 years of full-time work. Assuming that local people or migrants worked on transmigration projects to supplement agricultural incomes and that they averaged no more than 100 days of work/year, 240,000 workers would have benefited from employment generated through land clearing alone. Of public investment programs, only labor-intensive rural works projects and tree crop development generated more temporary employment in Repelita III. Even more employment could have been generated by transmigra- tion if more labor-intensive methods of land clearing and road construction had been used. Migrant Welfare 24. About a dozen studies have been completed in the last five years which contain information on migrant yields and incomes. One of the most comprehensive is an income survey that was undertaken by the Central Bureau of Statistics (BPS) and the Ministry of Transmigration with financial assistance from the Bank. This survey was carried out in February 1985 and covered 2,200 households in 22 Repelita II and III transmigration sites. The survey covered 12 sites based on rainfed food crops, 7 tidal sites and 2 tree crop sites, and it included both successful areas and several on a list of critical sites. About 46,000 families lived in villages covered by the survey, roughly 10% of all migrants moved in Repelita II and III. For comparative purposes, BPS also analyzed a sample of rural households covered in the 1984 National Socio- Economic Survey (Susenas). This sample covered 17,000 rural households in Java and 2,800 rural households in receiving provinces. - xxv - Table 4: COMPARISON OF TRANSMIGRANT AND NONTRANSMIGRANT INCOMES Monthly % with monthly Annual Survey household incomes below Rp household date income 30,000 50,000 income (Rp) /a /b /c (US$) /a Transmigration sites 1985 58,300 20 50 636 Rural sending areas 1984 67,200 15 43 733 Rural receiving areas 1984 90,750 4 24 990 /a All 1984 values have been updated to early 1985 Rp or US$. 7T Rp 30,000 (US$27) is the monthly subsistence level for a family of five. 7c Rp 50,000 (US$45) is the family poverty line, estimated at Rp 10,000/ capita/month. Source: Sending and receiving areas, 1984 Susenas; transmigration areas, BPS Transmigration Income Survey, 1985. 25. Despite some problems in interpreting the BPS survey data (for example, possible understatement of home garden production used for home con- sumption and differences in the standard of living between the inner and outer islands), several points are clear. As Table 4 indicates, average migrant incomes in this sample are somewhat lower than the average income of families in rural Java, and significantly lower than the income of families in rural areas of receiving provinces. If corrected for an underreporting of subsis- tence production, incomes are about the same as Java, but lower than those in the outer islands. About 20% of households sampled in the BPS survey in transmigration sites were at or below subsistence levels in 1985, compared to 15% on Java and 4% in receiving provinces. Low incomes in transmigrant sites are not surprising since migrants were poor prior to movement, families are young, and some have been on site only a short time. Monthly incomes in BPS and MOT surveys for migrants who had been in upland communities for five years averaged Rp 75,000/household, which is slightly higher than the Java average, although incomes in older upland communities were low (see Table 5). - xxvi - Table 5: INCOME, INCOME DISTRIBUTION, AND RICE PRODUCTION BY REPELITA AND FARM MODEL Average Annual monthly Percent of household Arrival period household households below rice and Sample income Subsistence Poverty production farm model size (Rp) level /a line /b (kg)/c Repelita II Tidal 99 77,697 11 27 1,760 Upland 341 52,825 33 57 470 /e Upland with tree crops 100 66,595 /d 13 42 - 7 Subtotal 540 59,934 24 49 760 /g Repelita III Tidal 501 39,791 38 73 710 Upland 977 65,801 10 36 770 /e Upland with tree crops 120 63,923 /d 12 35 - 7 Subtotal 1,598 57,505 19 48 750 /g /a Subsistence level is defined as Rp 30,000/household/month. 7T Poverty level is defined as Rp 50,000/household/month. 78 Total production of unmilled rice averaged over all households. About 15% of all families do not produce rice, so yields are higher for those growing crops. /d These figures show little or no income from major tree crops, which are just coming into production if migrants were settled in Repelita II, or are being established if migrants arrived in Repelita III. /e Rice producing households produced 570 kg rice per family in Repelita II settlements, and 925 kg rice per family for Repelita III settlements. /f Not calculated. Many farmers work only for wages and sample sizes are small. 7g Average in upland and tidal sites only. Source: BPS Transmigration Income Survey, 1985. 26. The fact that transmigrant incomes are somewhat lower than average incomes in Java does not mean that individual families are no better off than they were before they moved. Virtually all migrants are drawn from the poorest part of the population in Java, but in transmigration areas 50% are above the poverty line. In addition, transmigrants have assets such as land, a house and possibly livestock, they generally have the means to meet subsis- tence requirements, and they have reduced expenses for basic needs such as housing. Most transmigrants confirm income improvement. About two thirds of all respondents in the BPS survey said their incomes were better than they were before they moved, 17% said they were about the same, and 16% said they were worse. - xxvii - 27. The BPS survey also indicates that incomes in transmigration sites vary significantly by farm model and period of settlement. As shown in Table 5, one third of families on older sites based mainly on food crops are below subsistence levels and 38% of those in new swamp reclamation areas are below this point, while the number is only 10-13% in other sites. If farmers in older upland sites have low incomes due to low soil fertility coupled with declining off-farm work, and if migrants on new upland sites follow this pat- tern, future incomes would be very low. Government argues that improved site selection and improved farming systems will help recent settlers to sustain or improve yields, but in order to evaluate this premise, better time series data are required. Since the poorest people in the survey are recent settlers in swamp reclamation areas, these data point to an urgent need to address produc- tion problems in recent tidal sites. 28. Several other important conclusions emerged from recent BPS and MOT studies: (a) Off-farm Work. A substantial proportion of income in migrant commu- nities is derived from off-farm work. The proportion of total fami- ly income from off-farm work ranges from 40% to 80% in recent stu- dies and is about 50% in the BPS survey. Several explanations are possible. Farmers may be neglecting agricultural production because of the wide availability of work generated in part by site develop- ment, or the potential for agricultural production may be limited either by poor services or poor soils or the absence of cash crops. In either case, these data indicate that many transmigrants currently find off-farm work more remunerative than full-time agriculture. (b) Income Levels Over Time. A breakdown of family incomes in upland areas indicates that incomes are highest for settlers who have been on-site for 5 or 6 years and lower for settlers who have been on site more than six years. However, these data are difficult to interpret as they are not from the same sites. It is therefore difficult to tell whether new sites, which are larger (permiting income differentiation) and better selected, can maintain their incomes over time or whether incomes will decline as wage work associated with site development decreases. The lack of conclusive data on this subject argues strongly for improved long-term monitor- ing of incomes and production on transmigration sites. (c) Sustainability. BPS data indicate that most communities grow over time, but that there is a turnover rate of about 10% of migrants leaving sites in the first five years. While this number is higher than official figures, it is not surprising or unusual in a program of this scale. There is no evidence of large scale abandonment in transmigration sites and those who leave are generally replaced by spontaneous migrants. (d) Expenditures. Data on migrant expenditures suggest that migrant families consume more rice, corn, and cassava, animal protein and vegetables than do other families in either the inner or outer islands. This seems plausible since most migrants have the ability - xxvi i i - to meet their subsistence needs from home production. On nonfood expenditures, the situation is reversed. Javanese spend nearly twice as much on fuel and 2.5 times as much for housing as transmi- grants since fuelwood is readily available and houses have been donated by the government. They also spend more for household items, durable goods, taxes, medical expenses and schooling, but migrants spend more on transport. If confirmed by further analysis, these data indicate that migrants, while poor, are on average meeting their subsistence food needs as well as rural Javanese or rural residents of the outer islands. Economic Analysis 29. Economic analysis is useful as one element in evaluating program performance and it is particularly important at a time of resource constraints as a means of choosing between future investment options. To evaluate the economic impact of transmigration, an analysis was carried out on observed production in upland and tidal sites, and on projected production in new tree crop sites. Rainfed schemes based mainly on food crops have a low rate of return to agricultural production using standard economic assumptions. This ranges from negative in the case of the low-input, low-output upland models to 4% in the case of upland sites with good market access. When benefits from off-farm income are included in the economic analysis, upland schemes based on food crops have rates of return ranging from 3% to 6% for the most commonly observed farming patterns, and up to 9% for sustained input models which can be replicated only in sites well suited to food crop cultivation. Although projected production is higher in tidal sites than upland sites, rates of return are about the same (3%-5%) because of the higher initial investment cost in swamp reclamation. 30. Economic rates of return based on observed production in food crop sites are lower than those projected in World Bank appraisal reports for transmigration projects. This is mainly because average realized crop yields have been lower than projected and world rice prices have dropped signifi- cantly since 1981 (an almost 50% decline in real terms from the 1976-81 average to the 1985-86 average). A model has been developed that illustrates the potential of food crop schemes if agricultural services are in place, marketing is improved and the use of inputs is not constrained. This model has a rate of return of about 11% and is based on yields similar to those in early Bank appraisal reports; however, production at the levels projected in this model has only been observed under research conditions. Production increasing investments (such as the introduction of cattle) in existing sites where previous investments are treated as sunk costs, have economic rates of return up to 25%. 31. Economic rates of return were also estimated for new settlement with tree crops under a range of assumptions regarding management standards, expected production levels and investment costs. Nucleus Estate and Small- holder (NES) schemes for rubber and oil palm have rates of return which are higher than the assumed opportunity cost of capital (10%). Project Management Units (PMUs) have not been used as a vehicle to provide tree crops to trans- migrants to date. However, if they were used to provide rubber to new - xxix - settlers, and if settlement costs were limited to NES levels, the economic rate of return to this model would also be acceptable. As second-stage development on already existing sites, PMU schemes have rates of return of 16-18%. Smallholder net income at full development is projected to be twice as high on tree crop schemes as on food crop schemes. 32. Therefore, from the point of view of settler welfare and economic returns, an argument can be made that future transmigration development should be directed at: (a) second-stage development on existing sites and (b) new settlement on NES/PIR schemes and PMU schemes. Use of these models would lead to higher settler incomes, higher economic rates of return than the food crop model, and would permit cost recovery in some cases through credit schemes. However, the use of these models would also entail higher initial investment costs and necessitate low levels of sponsored settlement because of limitations on financial and institutional capacity. Regional Development Considerations 33. Transmigration on the scale achieved in Repelita III had a signifi- cant impact on the regions to which migrants were sent. Sponsored migrants equaled slightly more than 40% of the total population in 2 out of 66 receiv- ing districts (kabupaten) and in 7 others they were more than 20%. These nine districts received about 47% of all those settled. In three provinces in southern Sumatra and one in Sulawesi, migrants cultivated between 10% and 20% of all agricultural land, and they farmed an average of 4% of the land in all receiving provinces. Repelita III transmigrants are estimated to produce about 240,000 tons of rice, about 2% of outer island production. Overall, migrants who moved since 1950 are estimated to produce about 2 million tons of unmilled rice annually, or about 33% of incremental rice production in the outer islands since 1950 and about 5% of Indonesia's total rice harvest. 34. The most significant impact of transmigration in the regions was on infrastructure. Access and main roads constructed during site development equaled 14% of all roads maintained by outer island districts in 1985. The ratio was highest in East Kalimantan where roads constructed under the Repelita III transmigration program were almost equivalent to the total provincial road network. In two other provinces they accounted for about half. In 9 of 66 receiving districts, transmigration doubled road infrastruc- ture. Many of these transmigration roads are poorly maintained and place a heavy burden on district public works offices when transferred to the prov- inces. Government has recently arranged for surveys and increased funds for maintenance, but institutional capacity is limited, and transfer procedures need to be further improved. 35. Transmigration has also led to an expansion of other services in the outer islands, particularly extension, agricultural input supply, education and health services. Consequently, careful planning and budgeting are necessary to ensure that adequate staff are employed during the settlement period, and effective handover of transmigration sites from MOT to the pro- vinces will entail further planning for the number of people to be transferred (extension workers, teachers, health clinic personnel), incremental budget provisions for their salaries, and arrangements for incorporating them into district or provincial services. - xxx - 36. Large-scale transmigration has had a significant impact on town and urban development. Towns in transmigration regions have experienced a boom in activity as a result of transmigration-related construction and the search for off-farm work by local people and by sponsored and spontaneous transmigrants. Many of these towns are growing at rates that will potentially double their populations in ten years. For smaller towns that have no formal urban status and depend mainly on central budgets, rapid increases in road traffic and accelerated growth may be difficult to accommodate under current planning arrangements. Cities may also find it difficult to keep pace with accelerated growth, and the capacity to deliver urban services is particularly low in large coastal cities and small frontier towns in the outer islands. There- fore, the growing demands for services in these areas must be taken into account in future development planning. 37. In 1985/86, the transmigration program provided one third to one half the total provincial development budget in eight of 17 receiving provinces. Only in four receiving provinces with very limited migration did the share of transmigration fall below 10%. While budget figures are not equivalent to expenditures, there can be little doubt that this spending has caused high levels of demand for labor in construction, secondary industries, and services, and has laid the groundwork for future regional development. However, if nonagricultural work should decline with proposed budget cuts, the potential exists for increased competition for employment and stagnating or falling wages in the regions. To avoid this, it is critical that reasonable levels of government expenditure be maintained in the outer islands in an effort to generate sustainable growth, particularly where the labor force has been greatly increased in Repelita III. The best way of doing this is to emphasize second-stage development and to support infrastructure maintenance. Social and Environmental Concerns 38. Social Issues. The most sensitive issues in transmigration center on social and environmental concerns. In general, past transmigration pro- grams have brought enough benefits to the regions in terms of labor, public investment, infrastructure and services to receive a generally positive recep- tion. Therefore, the main social concern has centered on the fairness with which land is acquired. In the early years of Repelita III, rapid land clear- ing without adequate planning led to conflicts between transmigrant interests and those of the local people, mostly over compensation for productive trees. As planning improved and the role of regional planning agencies in mediating land claims increased, these conflicts declined. In Repelita IV, the responsibility for land acquisition was returned to the provincial Governors, and in most cases where land is contested it is not pursued by MOT. For this reason, among others, there is now little scope for further large-scale land settlement in Sumatra without financial compensation for land. As land constraints grow, the possibility of providing such compensa- tion will have to be taken under consideration. 39. Concern also exists that isolated people who are unaware of their rights may be persuaded to relinquish land against their best interests. Government has introduced measures to take account of such people in the planning process, to determine their views on benefits and/or compensation, - xxxi - and to provide benefits including parallel development where desired. Local level mechanisms for discussing land claims are also being developed. However, successful implementation of these measures will require close super- vision by the Government. 40. To improve the social soundness of the transmigration program the Directorate for Social and Cultural Development within the Ministry of Transmigration should be strengthened. This Directorate, staffed with ade- quately trained technical specialists, could help resolve land disputes, develop parallel programs for local people and ensure fair treatment of isola- ted or ethnically distinct peoples. The monitoring unit in the Ministry of Population and the Environment (MPE) and the planning bodies of the local governments (Bappedas) should also assume an increased role in determining benefits for local people. The monitoring unit in MPE has been provided with funds for technical support to the transmigration program, but it has been slow to hire consultants, and further efforts to strengthen this unit are required. The Bappedas should be encouraged to give close attention to prob- lems of land alienation, particularly in areas where there are isolated or ethnically distinct peoples. 41. Environmental Issues. Transmigration has had a positive impact on the environment in Java by reducing demographic pressures on critical lands, thereby permitting reforestation and reducing erosion. But most environmental concern has been focussed on the outer islands where settlement has reduced land under forest cover and placed pressure on conservation areas and on wildlife habitats. 42. Indonesia's forests are a major economic resource as well as an important and valuable reservoir of tropical flora and fauna. The total area of forest in Indonesia is estimated at nearly 144 million ha of which 122 million ha, two thirds of the country, is under closed canopy forest. Of the 144 million ha forest area, three quarters is in production (commercial) forest, protected or conservation areas, and about one quarter is earmarked for conversion. Much of the conversion forest is already deforested, or along streams and roads where settlement and clearing has already occurred. Between 30-50% of the land cleared for sponsored migrants in Repelita III was fores- ted, and Table 6 summarizes the relationship between land allocated for sponsored transmigration during Repelita III and the proportion of conversion forest and forested land cleared on major islands. Less than 1% of the fores- ted area was used for sponsored settlement on any island and less than 5% of total conversion forest was cleared. Land opened by spontaneous migration would roughly double these figures. 43. To mitigate the adverse environmental impact of sponsored settle- ment, preference should be given to the development of grasslands wherever possible, mechanisms should be developed to provide adequate compensation to local people for underutilized, unforested land, and the capacity of the Ministry for Population and the Environment and the environmental studies centers in provincial universities to assess and monitor environmental impact should be strengthened. To reduce problems associated with increasing spon- taneous movement, it will be necessary to incorporate spontaneous migrants into planned settlements and to develop mechanisms for land purchase and land registration which permit spontaneous migrants to acquire suitable land. - xxx i- Critically eroded areas such as this in Central Java are major sending areas for transmigration. 44. In a country with Indonesia's population density and level of devel- opment, the pressure to place land suitable for agriculture into production is so large that it is unlikely that all forested land can be maintained under forest cover. However, with a vigorous program to protect those areas needed for production, protection and conservation purposes, about 57% of Indonesia's surface would remain under forest cover, about 7% less than the amount now under closed canopy forest. To realize these objectives, ecologically impor- tant areas would have to be identified and protected from encroachment, and buffer zones established between settled and protected areas. Such a program would require an increase in funds and increased attention to appropriate institutional development. Financial and technical support for a major effort to protect forest areas could be mobilized from bilateral donors, multilateral lenders and environmental groups, and should be encouraged by the Government. 45. Irian Jaya. Early government projections calling for an increase in settlement in Irian Jaya from about 16,000 families in Repelita III to 140,000 families in Repelita IV caused alarm among both social scientists and environ- mentalists. These targets were not realistic and have been reduced. The - xxxiii - Table 6: LAND ALLOCATED FOR SPONSORED MIGRATION IN REPELITA III COMPARED TO FORESTED LAND Forest as a Transmigration Transmigration % of provin- land as a % of land as a % of Province cial land all forest /a conversion forest /a Sumatra 54% 0.7% 4.3% Kalimantan 67% 0.2% 0.8% Sulawesi 57% 0.3% 2.1% Irian Jaya 70% 0.0% 1.2% /a Assumes 50% of cleared land on all islands was forested. Does not include spontaneous migration. Source: FAO/World Bank Cooperative Program, Forestry Project Working Paper, 1985. budget for FY86/87 calls for settlement of only 3,000 families in Irian Jaya and, on this basis, the Bank now estimates that only about 15,000 to 20,000 families will be settled in the province in Repelita IV. Nevertheless, since the people in Irian Jaya are racially and culturally distinct, the management of any settlement there will require special sensitivity. 46. The majority of Irianese live in the densely settled highlands and there are no plans for transmigration to those areas, but new settlement is contemplated on the north coast, on the western coast and in the south around Merauke. Steps have been taken to ensure that consultants involved in the site selection process take account of the local people, identify the areas needed by them to sustain their traditional livelihood, and assess their attitudes toward settlement, and toward culturally appropriate parallel bene- fits or compensation. However, the process of acculturation is a long one and for this reason, slow, well-planned settlement coupled with programs to bene- fit local people and respect their cultural identity must be encouraged. To avoid social disruption, the provincial government should be sensitized to the rights of isolated and unassimilated people, site selection procedures should be carefully supervised, and settlement in Irian Jaya should not be accele- rated beyond current low levels. - xxxiv - C. Factors Affecting Future Settlement Agricultural Prospects 47. Food Crop Farming Systems. In Repelita III about 290,000 families were settled in sites based on rainfed food crop agriculture, nearly 80% of all those settled. Rice production in these sites is low (500-1,000 kg/family) and the evidence suggests that farmers are producing mainly to meet the commu- nity's subsistence needs. The reasons for this are clear: (a) agroecological conditions in transmigration sites are not well suited to food crop production and overcoming soil limitations is both costly and time consuming; (b) market prospects for food crops in most sites are limited; and (c) research, extension, input supply and credit mechanisms remain weak. This is due in part to the problem of getting experienced staff in pioneer areas and to the inherent difficulties of coordina- tion in remote sites. 48. Steps should be taken to increase production and reduce marketing constraints: for example, semi-mechanized land clearing should be encouraged both to reduce soil damage and increase employment; more emphasis should be given to the incorporation of organic material into the soil; houselots and fields should be placed as close together as possible to reduce labor con- straints and provide protection from pests; and farming systems should be diversified (poultry, livestock and fish ponds should be encouraged). The fact that many of these steps have been recommended by the Ministry of Agriculture (MOA), but not yet adopted by MOT, points to the need for stronger coordination between the two. In general, however, the inherent limitations of the soils in transmigration sites, the high cost of producing food crops under adverse conditions, and the deteriorating market prospects for food crops such as rice, argue against the assumption that income growth in existing sites will come mainly from surplus food production. 49. Swamp Reclamation Sites. In Repelita III, about 67,000 families were settled in swamp reclamation sites, 18% of the total moved. Settlement in these areas has a number of attractive features: good soils in swampy areas have a high organic content and can produce relatively good yields even without fertilizers; the flat cultivated areas eliminate erosion problems and permit bunding to hold water on the land for wetland rice; the timber on swamp sites has less commercial value than that on upland sites; and since swampy areas are largely uninhabited before they are drained, they are relatively free of land claims. 50. There are also serious problems in swamp sites. Some areas have difficult soils which may retard or prevent the growth of crops, water manage- ment intended to prevent overdrainage and saline intrusion is a complex under- taking, and environmentalists are concerned about the unforeseen consequences of draining coastal swamps. Recently, farmers settled in tidally reclaimed areas have also experienced severe problems with rats and insect pests. - xxxv - Whereas farmers surveyed by BPS in tidally influenced areas settled in Repelita II were producing an average of 1,700 kg of unmilled rice annually, farmers in Repelita III sites reported an average of 700 kg of paddy per family and the lowest average incomes in the income survey. This is due to low production and limited opportunities for off-farm work. Potential trans- migrants are aware of these problems and there were some difficulties in recruiting settlers for tidal sites in Repelita IV. 51. Despite current problems in swamp reclamation areas, most technical specialists agree that tidal areas have good prospects for growing rice and for diversification into other crops. To realize that potential, however, a concerted effort is required to address the problems of these sites. Critical components are as follows: (a) improvements in agro-hydrological surveys; (b) judicious drainage and use of proper land development techniques; (c) review of soil properties after clearing in order to provide recom- mendations for on-farm management; (d) return to the linear settlement model in which fields adjoin the houselot; (e) provision of adequate agricultural support services based on appro- priate agricultural research; (f) development of an intensive pest management strategy; (g) increased attention to problems of potable water by MOT and to malaria control by the Department of Health; and (h) concentration of management in one entity to improve the interactive process of planning and implementation which is particularly impor- tant in swamp reclamation sites. Settlement in tidal areas is only justified if current problems leading to low production and low incomes can be overcome. Government officials have recognized this problem and targets were reduced in 1985, even before resource constraints affected the program as a whole. 52. Tree Crop Development. The introduction of cash crops such as sugar, cotton, cashews, spices, coffee and cocoa on appropriate soils is to be encouraged. However, some of these commodities face technical and marketing constraints and for this reason the best prospects for income increases in transmigration settlements are from the major tree crops (oil palm, rubber and coconut). Agricultural incomes on tree crop schemes are projected to be about US$1,500/family 10 years after settlement, compared to about US$700/family on upland schemes, and only tree crop projects offer scope for cost recovery. 53. Government has recognized the advantages of tree crops and taken a number of steps to promote tree crops for transmigrants, but only about 7,400 - xxxvi - transmigrant families were settled on NES/PIR projects in Repelita III (about 2% of the total moved). Low rates of settlement in Repelita III were due mainly to institutional and manpower constraints. Since then levels of settlement have been accelerated and a total of 32,000 families were settled on tree crop schemes as of June 1986, and an additional 56,000 ha of trees were ready to be transferred to transmigrants. 54. Measures needed to increase institutional capacity in the tree crop sector are set out in a recent Bank study on the prospects for Indonesia's major tree crops (Report 5318-IND, April 15, 1985). Other recommendations relevant to transmigration are as follows. (a) New tree crop development should, other things being equal, take place as close as possible to existing infrastructure, rather than in remote areas, and for financial and environmental reasons priority should be given to the development of grasslands. To obtain grasslands for tree crops, both local people and transmi- grants must be included in tree crop schemes. (b) Second-stage development, which involves the introduction of tree crops on existing sites, has the highest rates of return and the lowest financial costs of all transmigration models and should be strongly encouraged. (c) Opportunity also exists for expanding tree crop development in the private sector. To stimulate private investment, private sector investors should be relieved of the obligation to finance small- holder development from their own borrowing. 55. There are several mechanisms for establishing tree crops on trans- migration sites including project management units (PMUs), formed within the Directorate General of Estates (DGE) to assist farmers in establishing tree crops on their own land. PMUs provide planting material, fertilizer, exten- sion and incentive payments for major tasks (land clearing), and the cost of these inputs is covered through credit. These mechanisms show particular promise for assisting existing transmigrants to improve their production systems and Government is now considering a large-scale program to produce planting material for transmigrant communities. One problem encountered in planning these programs is that farmers have too little reserve land for 1.5 to 2.0 ha of tree crops, the minimum area needed to provide adequate incomes and repayment. To avoid this problem in the future, Government must ensure that transmigrants are allotted at least 3.0 ha on arrival, not 2.0 ha as currently planned. 56. An analysis of overall implementation capacity in the tree crop sec- tor indicates that there will be a serious trade-off between the quality and scale of the program. In order to maximize migrant welfare and stimulate eco- nomic development, Government should shift to more productive farming systems and this will limit future program scale. Recently, tree crop commodity prices have fallen to very low levels and this is a matter of serious concern. Bank commodity projections predict a return to prices which would make tree crops economically and financially attractive. However, should prices remain - xxxvii - low, Government will be faced with very difficult choices. Since few new sites are well suited for food crops, if tree crops are not an option, then tidally-influenced sites would have the best prospects for settlement. Any future settlement program relying mainly on tidal sites would necessarily be based on a mixture of food crops and tree crops because of soil conditions. Land Availability 57. Land availability is a serious constraint to future settlement. Sumatra may be divided into those provinces which are already fully settled (Aceh, North Sumatra, West Sumatra, Bengkulu, and Lampung) and those less fertile areas on the coastal lowlands (Riau, Jambi and South Sumatra) where limited prospects remain for large-scale settlement. Continued large-scale settlement in West Kalimantan is possible only if indigenous smallholders agree to make areas of low productivity rubber available or relinquish land under shifting cultivation. Settlement potential in Central and East Kalimantan is limited by forestry constraints. For example, three quarters of the area in Central Kalimantan that is suitable for agriculture is in production forest. In Irian Jaya the pace of settlement will be limited by social and environmental considerations. Land is not an immediate constraint in swamp reclamation areas, but continued settlement requires that problems mentioned earlier be overcome. 58. Province-by-province projections suggest that the maximum number of families which could be settled in large sites on reasonable agricultural land in the absence of financial constraints would be about 300,000 families in both Repelita IV and V, declining in Repelita VI. However, this "high scenario" could be achieved only if the tree crop program were accelerated, and if Government were prepared to reclassify some areas of production forest and to bear the cost of settlement in increasingly remote sites. These options are not likely to be possible and may not be desirable. 59. Uncertainties about land availability need to be resolved for rational planning to occur for any future settlement program. For this reason, ongoing Phase I evaluations of upland and tidal areas should be continued. These studies combined with natural resource inventories already under way should permit the completion of master plans on a province-by- province basis by the end of Repelita IV. This, in turn, could provide the basis for national master plan for transmigration for Repelita V. Preparation of this plan should be given priority by Government. The Prospects for Spontaneous Movement 60. The main constraint to spontaneous movement is the inability to find land free of claims and obtain legal rights to it. This is due in part to systems of land tenure, governed by adat, or traditional law. Under adat, the community may allocate the right to cultivate, but this does not necessarily guarantee security of tenure. Outsiders using community land may be subject to ongoing requests from the local people for compensation or even expropria- tion of the land if its value improves. For this reason the most important function of Government in the transmigration process is the provision of land which is free of alternative claims. - xxxviii - 61. A second constraint to obtaining land is the difficulty and high cost of land registration. It is possible to legally transfer land, but this takes cadastral surveys, legal checks and a long series of administrative procedures which are difficult and costly to arrange. The Directorate General of Agraria is also required to give priority in land registration to develop- ment projects, cadastral surveys in towns and urban areas, and land used for commercial purposes. Consequently, there is no land registration program at the present time which is affordable and responsive to the needs of rural smallholders. 62. A third constraint is the difficulty of finding capital for land purchase. There are many areas in the outer islands where land of moderate fertility is available at a "fair price," generally about US$40-60/ha. How- ever, migrants seldom have this amount of money and Government does not pay cash compensation for land, since it believes that land is a national resource to be managed for the common good. This means that better land is seldom relinquished for transmigration. A long-term MOT objective should be to pro- mote rural credit for land purchase through the banking system for those who wish to purchase land on their own. In the short run, consideration should be given to a fund within the MOT which could be used for land purchase by asso- ciations in which members mutually guarantee repayment. This has been done in a number of countries with reasonable success. 63. The amount and direction of spontaneous movement in Repelita IV will also depend on a number of other factors. First, it will depend on the state of the economy and the pattern of government investment. Second, it will depend on the rate of sponsored settlement. Since spontaneous migrants often follow relatives or friends, few inner island residents will move sponta- neously to pioneer areas, and the vast majority of spontaneous migrants will continue to go to already settled areas such as Sumatra. As much of the available land in Sumatra is already occupied, strong government encouragement to move in the absence of improved systems of land acquisition and strong measures for environmental protection could have negative consequences. Financial Constraints 64. During Repelita III, finances were not a limiting factor in transmi- gration development. Budgets for transmigration generally exceeded expendi- tures, and between 1981/82 and 1984/85 funds expended as a percent of funds budgeted, declined. Costs were lower than originally anticipated because Government moved only about two thirds of the families targeted on the sponsored program. Since fewer families were moved than originally intended, by the end of Repelita III the MOT had growing budget carryovers. 65. The situation in Repelita IV is quite different. In January 1986, austerity measures were announced which reduced the MOT budget from Rp 501 billion (US$452 million) in FY85/86 to Rp 323 billion (US$293 million) in FY86/87. Then in May 1986 it was announced that the portion going to MOT would be only Rp 208 billion, about 58% of FY 84/85 expenditures (the last year for which these data are available). - xxxix - 66. Under these austerity conditions, MOT has set the following priori- ties for financing: (a) fulfillment of MOT's commitment to existing migrants and maintenance and upgrading of existing sites; (b) completion of contracts already signed and completion of settlement in areas already cleared; (c) provision of funds for externally-assisted projects (these are expected to settle about 5,000 families in the next two years); and (d) institutional development including improvement in coordination, monitoring and evaluation, training and studies. Institutional Considerations 67. Coordination. Since the Repelita III transmigration program was carried out by several agencies, various mechanisms were established to promote interagency coordination. Among these was the appointment of a Junior Minister for Transmigration (JMT) who reported directly to the President. After the two main implementing agencies were combined in 1983 to form the Ministry of Transmigration, a new Presidential Decree was issued giving the MOT responsiblity for coordination. Nevertheless, interagency coordination has remained weak. 68. To improve policy-based planning and integrated project implemen- tation, coordinating bodies must be strengthened and there must be clear agreement between the various ministries on policy objectives and their means of execution. To facilitate interagency coordination the Secretariat in the Ministry of Transmigration (MOT) at the central level should be strengthened and similar Secretariats should be formed at the provincial and district levels to convene existing transmigration coordinating bodies and to provide support services to them. Mechanisms are also required to improve on-site coordination. 69. Planning and Budgeting. In the past, the Repelita III transmigra- tion program experienced serious problems in planning and budgeting. This was due in part to (a) the large number of agencies involved and the lack of authority or resources in the Office of the Junior Minister for Transmigration (JMT) to reconcile agency plans; (b) the difficulty of programming the long sequence of activities involved in settlement; and (c) a budgeting system that permitted expenditures over three years and thus created little need for careful planning. With the elimination of budget carryovers in FY86/87, improved planning and budgeting have become critical to successful implemen- tation of the transmigration program. Some positive steps in this direction have already been taken. The responsibility for transmigration planning and budgeting is now in the Office of the Secretary General in the MOT, and this office has adequate authority to carry out the task. However, because most of the staff involved are new and the task is extremely complex, the Office could benefit from technical support. - xi - 70. Monitoring and Evaluation. One of the weakest links in transmigra- tion to date has been monitoring and evaluation. In Repelita III the Bank and the United Nations Development Programme (UNDP) provided technical support to the Office of the JMT for monitoring project implementation, financial status and impact. This included an early warning system intended to detect and react to problems on site. Because of limitations in JMT staff and resources, this system was never effective and has been abandoned. Monitoring and evalu- ation functions are now in the Office of the Secretary General and steps to develop a responsive monitoring system are urgently required. The UWDP is providing assistance for special studies, and a consultancy under the Bank- supported Transmigration IV project also is designed to develop monitoring and information systems for the MOT. However, greater priority needs to be given by the MOT to the development of a simple monitoring system. 71. The Changing Role of the MOT. Twenty years from now, settlement following the current sequence of steps in the transmigration program will be limited. Instead, a broad spectrum of interagency programs and policy initia- tives will be required to foster labor migration and promote regional develop- ment. To move from the current highly logistical operation to one centering on policy formulation and facilitation will not be easy. To achieve these objectives, it will be important to identify the agency within MOT intended to provide guidance and policy direction, to give that agency appropriate autho- rity, and to strengthen it. At the moment, it appears that this role would best be filled by the Office of the Secretary General (OSG). If so, a con- certed effort is needed to define the function of the OSG, to attract the most skilled manpower to it, and to give it adequate authority to integrate the work of the other agencies. D. Summary of Major Recommendations 72. In order to sustain and improve migrant incomes and to capture the full economic benefits of manpower and infrastructure which are already in place, the main thrust of the transmigration program over the next few years should be on the maintenance and development of existing sites. Road maintenance will be a matter of high priority and attention should be given to upgrading existing production systems. To provide an appropriate administrative system for this development and to facilitate the transition between Ministry of Transmigration (MOT) authority and provincial control, the report recommends a bridging period under MOT authority for sites which are not self-sufficient at the end of five years. To ensure adequate financial and managerial resources for maintenance and upgrading, new settlement must necessarily be slowed. Government has recognized this point and deferred new planning and settlement. 73. With respect to the scale of future settlement and the types of productive systems used, the report recommends that: (a) fill-in programs which make use of existing infrastructure should be encouraged using sustainable production systems. Such programs can be carried out with lower costs if settlers are recruited among spontaneous migrants in receiving areas. However, the benefit package for farmers recruited in Java and moved to areas where they have no family support should not be significantly reduced; - xli - (b) future settlements based on food crops should be limited to areas where soils are good and marketing prospects are strong, or where plans for second-stage development are firmly in place; and a concerted effort should be made to address the production and management problems encountered in tidal sites which appear to have the best potential for food crop production; (c) new settlement with cash crops should be encouraged. This could be based on any commodity with reasonable marketing potential, but in most cases such settlement will be based on tree crops such as rubber, coconut and oil palm, which are well adapted to the soils of the outer islands, and for which mechanisms for establishment already exist; (d) new settlement with tree crops, Nucleus Estate and Smallholder schemes should be supported, and a major initiative is required to strengthen tree crop project management units (PMUs) and use them for tree crop establishment in conjunction with new settlement and second-stage development. Steps should also be taken to facilitate private sector investment in smallholder tree crop establishment; and (e) as the industrial (tree crop or cash crop) model involves cost recovery through credit, it is imperative that the size of the family unit can support loan repayment and family requirements. This will entail a unit of at least 2.5 ha in most cases, and the current limitation of 2 ha per family (including both houselot and food crop area) is unlikely to be sufficient in most cases and should be reconsidered. There will also be a need to re-examine and, if necessary, redesign site plans originally based on the standard settlement model where cash crops are to be introduced. 74. To enhance the quality of new settlement mechanisms must be developed witin the MOT to review management issues on a regular basis and to improve policy formulation and interagency coordination. Action plans should also be developed to overcome the most serious problems encountered in implementation. These should focus on settlement design (including selection of farm model, farm layout and incorporation of spontaneous migrants), physical development (land clearing, road construction and maintenance, and contractor supervision), the provision of agricultural supporting services, and program planning and coordination (planning, budgeting, monitoring and evaluation). Steps are also required to establish a system for assuring the orderly transfer of settlement to the provincial governments. 75. Increased attention is also needed to spontaneous migration. In the past decade, spontaneous movement is presumed to have occurred at rates one to two times as great as sponsored settlement. This rate of movement occurred in response to (a) employment opportunities generated by government investment and to (b) sponsored settlement in the outer islands. In FY86/87 Government hopes to facilitate the movement of substantial numbers of spontaneous migrants; however, it is not certain that Government can or should encourage - xlii - high levels of spontaneous movement if investment in the outer islands slows and employment opportunities decline, or if there are no new provisions for finding adequate land. A detailed review of spontaneous transmigration is warranted to more clearly define policies and realistic programs. 76. To improve the welfare of those who are already moving without assistance and to minimize environmental problems, steps are required to: (a) adapt the planning process to take account of future spontaneous migrants; and (b) develop procedures for land registration in the outer islands and provide credit for land purchase, particularly in those areas where sponsored migrants have already settled (Sumatra). These actions should be given priority by GOI. 77. Finally, to protect both the environment and the rights of local people, attention is needed to environmental and social concerns. On the environmental side, the use of secondary forest and grasslands should be maximized and full account should be taken of the analysis of forest resources carried out during site selection. This includes a delineation of buffer zones to protect ecologically important areas. In addition, priority should be given to identifying and gazetting environmentally important areas and protecting them from encroachment, and to developing clear policies on the preservation of production forest. The capacity for environmental assessment in regional universities should also be strengthened. On the social side, measures introduced in the recent Transmigration V project to take account of the needs and desires of ethnically distinct and unassimilated people should be rigorously implemented; and to ensure that local people, particularly those in Irian Jaya, have an opportunity to adjust to and benefit from development, settlement in such areas should be carried out on a gradual basis and not be rapidly accelerated. 78. Current price projections for oil are pessimistic from the producer point of view and suggest that the strategy proposed above will have to be maintained in the immediate future due to financial constraints. Even if the resource position improved, however, the general recommendations included in this report would not change significantly. Government should emphasize labor-intensive second-stage development to sustain incomes and encourage spontaneous movement to existing sites, and it should develop programs includ- ing increased access to credit to support land purchase, land registration and farm development, in order to facilitate the flow of labor to areas of economic opportunity. At the same time it should consolidate and improve new settlements in order to stimulate regional development. I. OVERVIEW A. Background General 1.01 Indonesia, the fifth most populous nation in the world, had an esti- mated population of 168 million in 1986. Of this number, over 100 million people lived on Java, an island with about 7% of the nation's land. Poyula- tion densities in Java averaged 774 people/km2 compared to 33 people/km in the outer islands, and Java population densities were about equal toothose in Bangladesh. 1.02 The significance of these facts cannot be overstated. Java has a total of about 20 million households of which about 17 million are in rural areas. About 63% of rural households have less than one-half hectare of land (a subsistence-size plot), while only 5% have 2 hectares or more. Small- holdings in the outer islands are somewhat larger: only a quarter of all farm families in the outer islands have less than one-half hectare, while 30% have 2 hectares or more. Soil fertility in the outer islands is also generally lower than in Java, although there are many areas in Java where soils are very poor and some areas in the outer islands, particularly in Sumatra and Irian Jaya, where soils are relatively good. 1.03 With 11.6 million farm families and 6.0 million ha of land under food crops, agricultural holdings on Java are small and incomes are low. In 1984, average incomes in rural areas of Java were about US$700/family/year and about 40% of the ruryl population fell below a poverty line estimated at US$540/family/year.1 The outer islands (islands other than Java, Bali, Madura and Lombok) have more land, less dense populations, and the major por- tion of Indonesia's natural resources, particularly oil and forest reserves. For these reascns households in many areas in the outer islands are somewhat more prosperous, and incomes in rural areas of the main transmigration receiving provinces (Sumatra, Kalimantan and Sulawesi) average about US$1,000/ family/year. Table 1.1 summarizes the main demographic features in sending and receiving provinces. 1.04 Faced with these disparities between land and population, Indonesia has sought to wed the underutilized labor of Java and the underutilized land of the outer islands through various programs of land settlement. These pro- grams began in 1905, and by 1930 about 100,000 people had moved to the outer 1/ All analysis in this report is based on an exchange rate of Rp 1,100/ US$1.00, the prevailing rate when this report was written. - 2 - Table 1.1: BASIC DEMOGRAPHIC AND LAND USE DATA Est. 1986 % rural HH % farm HH % farm HR F farm HH Province/Island population Area Population which are with less with less with more ('000) (km 2) (people/km2) farm HH than 0.10 ha than 0.50 ha than 2.0 ha Sending Areas Jakarta 8,164 590 13,837 0 00 0 West Java 31,876 46,300 688 63 18 67 5 Central Java 27,756 34,206 811 74 13 61 4 - Yogyakarta 2,913 3,169 919 85 16 61 5 East Java 31,639 47,922 660 69 14 62 4 Java 102,348 132,187 774 69 15 63 4 Bali 2,709 5,561 487 77 9 46 10 West Nusa Tenggara 3,108 20,177 154 67 11 46 11 Bail/NTB 5,817 25,738 226 72 10 46 10 Total 108,165 157,925 685 69 15 62 5 Receiving Areas Aceh 3,078 55,392 56 77 5 33 17 North Sumatra 9,668 70,787 137 77 8 39 13 West Sumatra 3,851 49,778 77 81 6 38 8 Riau 2,583 94,562 27 79 6 17 45 Jambi 1,822 44,924 41 81 5 17 44 South Sumatra 5,587 103,688 54 80 3 14 35 Bengkulu 986 21,168 47 88 2 12 35 Lampung 6,422 33,307 193 87 3 20 22 Sumatra 33,997 473,606 72 81 5 27 23 West Kalimantan 2,827 146,760 19 88 2 10 37 Central Kalimantan 1,159 152,600 8 79 2 7 51 South Kalimantan 2,328 37,660 62 80 6 37 33 East Kalimantan 1,690 202,440 8 71 6 17 49 Kalimantan 8,004 539,460 15 82 4 19 39 North Sulawesi 2,406 19,023 126 81 6 28 38 Central Sulawesi 1,604 69,726 23 85 4 13 49 South Sulawesi 6,666 72,781 92 79 7 31 36 Southeast Sulawesi 1,122 27,686 41 85 6 19 48 Moluccas 1,659 74,505 22 88 5 15 45 Sulawesi 13,457 263,721 51 82 6 25 40 Irian Jaya 1,364 421,981 3 85 * * * Other 3,613 62,750 58 * * * * Total 60,435 1,761,518 33 81 6 26 30 Legend: * = Data Uncertain RH = Households Note: Farm households are families that say their main work is in agriculture. Source: Statistical Yearbook of Indonesia, 1984, and the 1983 agricultural census. - 3 - islands on land settlement schemes and about 600,000 had been relocated as estate laborers. With the decline of plantation investment in the outer islands in the 1930s, transmigration became the major vehicle for movement. By 1980 about 1 million people had been resettled through sponsored transmi- gration programs, and about 2 million people were in the outer islands as a joint result of sponsored transmigration and associated population growth. In the third five-year development plan (Repelita III, 1979-84), 366,000 fami- lies, nearly 1.5 million people, were settled under the sponsored program, more than twice the number settled in the preceding 75 years (Table 1.2). During this period transmigration was the largest voluntary government- sponsored settlement program in the world. Table 1.2: SPONSORED TRANSMIGRATION, 1950-1984 /a Year Five- Total Local Families Total Total of year families families Resettled families people arrival plan moved /b /c settled settled (Repelita) 1950-54 - 21,037 0 1,280 22,317 87,000 1955-59 - 32,114 0 128 32,242 134,000 1960-64 - 26,456 0 0 26,456 111,000 1965-69 - 21,633 0 0 21,633 92,000 1969-74 (I) 39,436 0 75 39,511 176,000 1974-79 (II) 44,484 7,600 0 52,084 228,000 Subtotal 185,160 7,600 1,483 194,243 828,000 1979-84 (III) 301,279 22,284 42,414 365,977/d 1,492,000 Total 486,439 29,974 43,897 560,220 2,320,000 /a Settlement figures vary widely in government publications. This is due mainly to the practice of counting migrants by the year in which funds were budgeted for movement rather than in the actual year the migrants were moved. To the extent possible, this report indicates actual year of movement. /b Indigenous families settled in transmigration sites. 7i Resettlement of sponsored or spontaneous migrants within the province. 7E Government also identified about 170,000 families that moved sponta- neously. This figure is not included since it is not a complete enumera- tion of spontaneous migrant families, and comparable figures for earlier periods are not known. Source: Official Summary of Repelita III Transmigration Program, Ministry of Transmigration, 1985. -4- Development Objectives and the Role of Transmigration 1.05 Employment creation and poverty alleviation are key development objectives in Indonesia, and the country is widely regarded as one of the more successful developing countries in translating oil revenues into equity pro- grams. In agriculture, major investments have been made in irrigation, ferti- lizer production, agricultural research and smallholder tree crop development, while the mechanization of agriculture has been discouraged on employment grounds. In addition, the Government has invested in labor-intensive rural works programs, and in education, health and urban improvement, among others. Poorer regions have been given a larger share of development funds in relation to the size of their populations. The Government has also encouraged the development of small-scale industries, but the growth of export industries has lagged due to trade and commercial policies, weak commodity prices, and, to some extent, import restrictions in other countries. In 1987, however, significant steps were taken by Government to deregulate trade and stimulate export development. 1.06 To alleviate poverty and limit population growth, Indonesia also has an active family planning program. In 1970 the Government announced its intention to reduce the crude birth rate from 44 births per 1,000 population per year to 22 per 1,000 by the year 2000. By 1985 the birth rate was 33 per 1,000 and the growth rate had fallen to 1.8% p.a. in Java, and 2.3% p.a. for the country as a whole. Further declines will be harder to achieve on a voluntary basis, as poor families choose to have three or more children for many reasons, among them to provide economic security to the parents in their old age. 1.07 Despite these accomplishments, Indonesia has always found it diffi- cult to generate adequate employment for its growing population, and for eighty years transmigration has been seen as a partial answer to the Govern- ment's objectives of limiting population growth in Java, creating productive employment, and providing some of the poorest people in the country with suf- ficient land to meet their subsistence needs. In recent years, the pressures for resettlement have become increasingly intense as Landless farmers have moved onto ever more marginal land in Java and into poor areas in the cities. Those cultivating in upper watersheds and within legally protected forest areas have caused serious erosion, premature siltation of dams and irrigation canals, and widespread flooding, all with heavy economic and finan- cial costs to the country. The provision of urban services has also been costly, while failure to meet the expectations of the urban poor has poten- tially serious social and political consequences. 1.08 The Government's interest in moving people has been reciprocated by the desire of poor farmers to move in order to improve their own lives and the prospects for their children. Applicants for transmigration are usually land- less agricultural laborers with little likelihood of owning property or other assets. Their move involves the hardships of pioneer life, but offers opportunities not available in their area of origin. With the exception of small and unsuccessful efforts to relocate urban squatters in the late 1960s and early 1970s, and the infrequent use of transmigration to resettle people displaced by natural disasters, transmigration is voluntary. Transmigration officials target specific sending areas for recruitment (e.g., critical watersheds), but in recent years applicants from most areas have exceeded the numbers who could be moved. 1.09 The Government has also supported transmigration as a means of stim- ulating regional development. There are many potentially productive areas in the outer islands with population densities so low that regional development is not economica1ly feasible. In such areas, transmigration has been used to provide the critical mass to justify investment in infrastructure and services (such as agricultural extension and health clinics) and to provide labor and markets. Transmigration has also been encouraged as a means of reducing national food deficits by opening new lands for food crop production and by enabling new settlers to meet their subsistence needs. 1.10 Finally, transmigration has been seen by national leaders as a tool for national integration. Indonesia is a country with 13,000 islands and 300 languages linked initially by a common colonial history and prolonged struggle for independence. The ideal of cultural pluralism is recognized in the constitution and in Indonesia's motto "Unity in Diversity," and virtually all Indonesian ethnic groups preserve their mother tongue and many of their customs and traditions. However, Indonesian leaders have been both fortunate and skillful in forging a national identity based on a common national lan- guage, Indonesian (the native language of an area in Sumatra), a unified administrative system, and state philosophy which stresses equitable develop- ment. Migration of outer island residents, mainly to Javanese cities, and of inner island residents, mainly to rural areas in the outer islands, is seen by national leaders, from both Java and the other islands, as a means of promoting cultural contact and building nationaL unity. Transmigration Farm Models 1.11 Over the years, Indonesian resettlement efforts have had two major features in common. First, they have been government sponsored. Government support is necessary because many migrants are poor and cannot afford to move and establish themselves using only their own resources, and because tradi- tional land tenure arrangements in the outer islands make it difficult for migrants to secure legal rights to land without government support. Second, virtually all programs have been based on smallholder agriculture. In most schemes, migrants have been provided with free transportation to the site, land, housing, an agricultural package, and subsistence supplies for the first year, although the farming systems introduced have varied. 1.12 Dutch efforts to resettle Javanese as smallholders and most govern- ment transmigration efforts through the early 1970s were generally premised on irrigation. In early schemes, migrants received a 0.7 ha irrigated plot, an arrangement which placed considerable pressure on the land as children matured. In later projects, farmers were given 2.0 ha of land with irrigation potential but, due to institutional and financial limitations, irrigation infrastructure was deferred to a later stage. -6- Main drainage canal of tidal trans:igration community in South Sumatra. Several small businesses have developed along this thoroughfare. 1.13 A variation on irrigated farming was introduced in swampy areas of Kalimantan in the 1930s. In these areas, in theory, tidal action pushed fresh water back up river estuaries onto low lying land, making wet rice cultivation possible. In practice, most drained areas were used for rainfed rice. only small numbers were settled on such schemes in the colonial period and a major settlement effort announced by the Public Works Department in 1957 was ham- pered by lack of funds. Not until 1967-68 was the groundwork laid for the present program, which has since opened some 670,000 ha of tidally influenced land to migrant farmers. 1.14 Most of the transmigrants moved between 1950 and 1975 were settled on one or two hectares and produced food crops and minor tree crops (coffee and cloves) under rainfed conditions. Many farmers diversified their produc- tion systems over time and a few received irrigation or block-planted tree crops (rubber or coconuts) after several years. However, it was not until the second five-year plan (1974-79) that increased fertilizer availability and national rice shortages led to the development of an agricultural package intended to support rainfed food crop production on marginal soils. This -7- '41 ~ ~ 1 A Mm~~~~~~~4M Transmigrants in some schemes receive 1-2 hectares of rubber producing incomes of US$1,200-US$1,500 per year. These children are playing in a rbbe ara wichhasbeen in production about one year. package included 2.0 to 5.0 ha of land, of which 1.25 ha were cleared, and free planting material and fertilizer provided for three years. Most families settled in the third five-year plan received some variant of this model. 1.15 Plantation, or estate management, was first used to support trans- migration in the Nucleus Estate and Smallholder (NES) projects begun in 1978. Under the N,ES concept, funds were made available to expand existing estates and to increase their capacity to plant tree crops for smallholders including both local people and transmigrants. In NES projects, farmers generally received 2.0 ha of block-planted tree crops and 1.0 ha for food crops. Although costs are greater than those in standard projects, the cost of tree crop development is covered by loans to the smallholders and incomes are expected to be sufficiently high at project maturity to allow cost recovery. Recently, interest has grown in using project management units (PMUs) to help transmigrants develop tree crops on their own land. B. The Repelita III Transmigration Program Program Formulation 1.16 In planning Repelita III (1979-84), government officials saw trans- migration as an important tool for relieving population pressures, improving the welfare of the rural poor, protecting critical lands in the inner islands, promoting regional development in the outer islands, and increasing food pro- duction. To meet these objectives, the Government proposed to settle about 500,000 families during the plan period, on some 250 sites located mainly in Sumatra, Sulawesi, and Kalimantan. Of the settlements originally proposed, 18% were in swamp reclamation areas. 1.17 A central issue in designing the Repelita III program was the farm model and investment level to be used. Government officials recognized that it was technically possible to resettle smallholders on tree crop schemes that guaranteed relatively high incomes and high rates of return. However, they also believed that such programs would be costly in both financial and managerial terms, would slow the settlement process, and would create signifi- cant income disparities between beneficiaries and the local people. For these reasons, and consistent with the view that transmigration had to be carried out on a relatively large scale to realize program objectives, the Government decided that most settlement would be based on food crop agriculture. This model was adopted because annual crops could be established quickly, they promoted early self-sufficiency, cost less than other models, and provided no undue advantages to transmigrants in relation to the local people. Where appropriate, tree crops were expected to be introduced at a later stage, but no definite planning for this was undertaken. 1.18 Several other steps were also taken to facilitate large-scale set- tlement. First, organizational arrangements were changed. Prior to Repelita III, the Directorate General of Transmigration (DGT) in the Department of Manpower and Transmigration was responsible for all aspects of the transmigra- tion program. However, recognizing that institutional constraints limited the scale of resettlement, the Government undertook a major reorganization of the transmigration program prior to Repelita III. In March 1978, a Junior Minister for Transmigration was appointed to coordinate the activities of the various agencies intended to work on the program, and in August 1978 responsi- bility for project implementation was transferred from the DOT to the agencies normally responsible for each sector (such as Agriculture, Public Works, Health and Education). Overall, 7 Ministries and 53 Directorates General were to be involved, although only a few played major roles. 1.19 Second, funds were increased. The country's improved resource position allowed the Government to increase investment from about US$500 per family at the beginning of Repelita II (about US$1,000 in 1982 prices) to US$5,000 per family in Repelita III. Apart from inflation, the difference in cost was due to needed improvements in project design (the introduction of feasibility studies, topographic mapping, detailed village design), new compo- nents (link roads, hydrological surveys, agricultural research), and improved inputs (initial land clearing, the provision of cattle and fertilizer). The Government's indicative budget for Repelita III allocated about US$2.0 billion - 9 - for transmigration, nearly 6% of the projected development budget. 'With increased access to funds, financing ceased to be the major constraint to effective program implementation. 1.20 At the beginning of Repelita III (1979/80), the World Bank undertook a Transmigration Program Review (Report 3170-IND) to identify constraints to program development. This report projected only a moderate increase in the scale of transmigration and concentrated on ways to improve the program and eliminate institutional constraints. The review concluded that swampland set- tlements had good prospects, if costly infrastructure could be avoided, but that income prospects from rainfed food crops were limited in many areas and that a greater role should be given to mixed farming systems including tree crops and to second stage development among existing transmigrants. The report also recommended increased attention to site selection and evaluation, institution building and staff training, and efforts to improve the utiliza- tion of natural resources by developing grasslands and by utilizing timber from areas that were to be cleared. Achievements 1.21 In the first years of the Repelita III program, progress was slow due to a lack of prepared projects and inadequate interagency coordination. Later, however, significant improvements were made and the scale of movement increased dramatically (Table 1.3). In total, some 366,000 families were set- tled under the government-sponsored program, including about 22,000 local families and 42,000 spontaneous migrant families living in Lampung and reset- tled within the province. This is about twice the number projected by the World Bank. To meet the target of 500,000 families, an additional 170,000 families were identified who moved without government support, but the actual number of families moving spontaneously is believed to be much higher.2 The safe and orderly movement of nearly 1.5 million people on the sponsored pro- gram represents a commendable logistical achievement. That this many people were settled without a significant heightening of ethnic tension was also a tribute to both the Government and the Indonesian people. 1.22 Of the sponsored migrants, about 80% were settled on upland sites, 18% on sites in tidally influenced areas, and 2% on tree-crop schemes. Settlement by year and island is shown in Table 1.3; settlement by year and province is shown in Annex 1, Table 1. An evaluation of production, incomes and welfare in different sites is provided in Chapter II. The total cost of the Repelita III program was Rp 1.6 trillion (US$2.3 billion), of which approximately US$107 million (5%) was disbursed by the World Bank. 2/ Figures on unassisted movement are based mainly on information provided by sending provinces. It is uncertain whether all families who regis- tered to move from those provinces actually did move or whether most families who moved also registered. There is no doubt, however, that the number quoted understates the amount of spontaneous migration in the five-year period (see Chapter VII). - 10 - i.23 The World Bank's involvement in transmigration began in 1976 with a project (Ln 1318-IND) to settle 4,500 new families and to upgrade a community of 12,000 others. During Repelita III, the World Bank approved five additional projects, one each for large-scale settlement based mainly on food crop pro- duction, for settlement including tree crops, and for site selection and eval- uation, and two projects for settlement in reclaimed tidal swamps. In total, the Bank committed US$437.8 million to projects intended to settle about 57,000 families and benefit another 350,000 families through improved site selection and evaluation. A detailed description of Bank-assisted projects is provided in Chapter IX. 1.24 Other international and bilateral agencies also assisted the Repelita II and III transmigration program. Among the major contributors were the World Food Program (US$66.3 million) in Sumatra, Kalimantan and Sulawesi; the Asian Development Bank (US$34.3 million) in Southeast Sulawesi; the United States Agency for International Development (USAID) (US$20.8) in the Luwu Project in Sulawesi; the Federal Republic of Germany (DM 86.5 million) mostly in East Kalimantan; and the Netherlands (DC 26 million) in Bengkulu. The Food and Agriculture Organization and the Office of Project Execution of the United Nations Development Programme (UNDP/FAO, UNDP/OPE), France and the United Kingdom have also provided technical assistance to the Ministry of Transmigration. Program Development during Repelita III 1.25 Site Identification. The absence of suitable sites was a major con- straint to the start-up of the Repelita III transmigration program, and set- tlement on unsuitable sites was a major impediment to agricultural development under the plan. For these reasons, the World Bank agreed in 1981 to develop a program of site selection and evaluation. The Transmigration II project (Ln 1707/Cr 919-IND) was restructured to include settlement planning, and a second loan, Transmigration III (Ln 2248-IND), provided consultant support for land identification for 300,000 families. Under guidelines established for the project, sites were to be rejected if found to have unsuitable soils, steep slopes, or competing land claims from local people, forestry, wildlife reserves, or other development projects. Only about 20,000 families had been settled on sites identified under these procedures by mid-1986, so the overall impact of this program on transmigration had not yet been fully realized at the time of this sector review. During Repelita III studies were also ini- tiated in the Directorate General for Water Resources Development (DGWRD) to identify areas for swamp reclamation. - 11 - Table 1.3: REPELITA III SETTLEMENT, BY YEAR AND ISLAND 1979/80 1980/81 1981/82 1982/83 1983/84 Total Sponsored migrant families Sumatra 16,384 49,043 48,520 57,578 55,540 227,065 Kalimantan 565 11,976 17,378 17,488 23,207 70,614 SuLawesi /a 3,854 9,663 15,890 12,396 9,879 51,682 Irian Jaya 290 2,521 2,688 5,362 5,755 16,616 Subtotal 21,093 73,203 84,476 92,824 94,381 365,977 Spontaneous migrant 1,985 3,359 8,961 32,445 122,747 169,497 families Total 23,078 76,562 932437 125,269 217,128 535,474 /a Includes the Moluccas and other small eastern islands. Source: Figures provided by the Ministry of Transmigration. 1.26 One consequence of site selection on the scale undertaken in Repelita III was that it drew attention to land constraints and mobilized the planning efforts of other government agencies. Midway through Repelita III, the Directorate General of Fores7try completed maps intended to demarcate pro- duction, protection and conservation forests (see Chapter V), estates and individuals sought to register land for agricultural development, and environ- mental groups mobilized to identify areas of biotic importance. To reconcile these various interests, regional consultative assemblies were strengthened, and provincial governments took an increasing role in mediating between con- flicting land claims and development priorities. 1.27 These achievements were not without problems. The Directorate Gen- eral of Forestry's defense of its territories, whether forested or not, pre- vented migrants from settling on suitable unforested land in some provinces, while permitting settlement enclaves within forested areas in others. Funds provided to the Ministry of Population and Environment to monitor transmigra- tion were not used during the planning period, and regional governments often found themselves unable to resolve competing land claims. But although such factors caused delays and confusion, the involvement of these agencies repre- sented an important step forward in the process of reconciling development options, stimulating comprehensive area development planning, and alerting central government officials to emerging land shortages. - 12 - 1.28 Farm Model. The Repelita III transmigration program was based primarily on rainfed food crop agriculture. In support of this model, settle- ment was to be restricted to areas with less than 8% slope, 1.25 ha of land were cleared per household, and an agricultural package consisting of seed and fertilizer was provided to each transmigrant family for 3 years. Extension officers were also to be stationed in transmigration sites, and an expanded agricultural research program was initiated to improve farming systems on the acidic red-yellow podzolic soils common in the settlement areas. As part of this research and extension effort, trials were carried out on liming (to improve acid soils), soil conservation and weed control. 1.29 While recent studies indicate that most migrants have been able to meet their subsistence needs (see Chapter II), incomes from agriculture have been disappointing due to the low inherent fertility of the soils and the difficulties of providing timely inputs and reasonable extension in remote areas and of maintaining reasonable levels of fertilizer use once free supplies ceased. Many new tidal sites also had production problems. For these reasons, a midterm review of the Repelita III transmigration program carried out by transmigration consultants concluded that the farm model used was a major constraint to the successful implementation of the program. The review recommended allocating more money to agricultural development and shifting to more productive settlement models including irrigated food crops and tree crop development. 1.30 In recognition of these problems, a number of steps were taken during Repelita III to expand tree crop development for transmigration. Early in the plan a program was announced to settle 48,000 transmigrant families on NES schemes. Only about 8,000 families were settled on such schemes although considerably more trees had been planted than had been allocated to migrants by the end of Repelita III. Later, arrangements were made for sites identified under the transmigration program, but suitable for tree crops, to be used by the Directorate General of Estates (DGE) as tree crop schemes. However, little progress was made in developing needed programs for tree crop establishment on older sites. 1.31 Unassisted Migration. During Repelita I and II, all official migrants were, to the extent possible, intended to receive equal benefits, and a standard program was developed for this purpose. Partially assisted pro- grams were small and differed little from the regular program; but those moving outside the official transmigration stream received little or no assistance from the Government. In Repelita III, however, severaL steps were taken which were intended to facilitate unassisted movement, including the formation of a Directorate for Self-Initiated (Swakarsa) Movement, the completion of several studies on partially assisted settlement, and the inclusion of unassisted migrants in official targets. Although the Directorate remained weak and most proposed programs were not implemented, changes made in Repelita III signaled the Government's interest in promoting spontaneous movement to already existing sites (see also Chapter VII). 1.32 Organizational Arrangements. Organizational arrangements are described in detail in Chapter VIII. As might be expected, with the number of agencies involved in the Repelita III program, the integration of planning and - 13 - budgeting was complex, on-site coordination proved difficult to achieve, and reporting, monitoring and evaluation were complicated. Although improvements were made in the course of the five-year plan, the persistence of serious coordination problems throughout Repelita III led directly to the formation of a new Ministry of Transmigration in Repelita IV. 1.33 Emerging Issues. Prior to Repelita III the transmigration program had been modest (10-15,000 families per year) and individual settlements were small (200-500 families per site). Under these circumstances both the social and environmental impact were relatively limited. The seven-fold expansion of the program in Repelita III, coupled with the prospects of ever-increasing targets in subsequent five year plans, however, raised concerns about the rights of local people and the impact of the program on them and about the effect of the program on forests and wildlife. These issues, which are discussed in detail in Chapter V, emerged rather late in Repelita III as the scale of the program gradually became evident, but foreshadowed what were to become important concerns about the program scale and implementation in Repelita IV. C. The Repelita IV Transmigration Program Program Formulation 1.34 In planning for Repelita IV (1984-89), the Government recognized that adequate growth of both incomes and employment were key requirements for economic development and social stability. It also recognized that these objectives would be much harder to achieve in the late 1980s than in the previous decade when growth in both incomes and public expenditures was buoyed by rapidly increasing oil revenues. Under these circumstances, employment generation through labor-intensive investment was given high priority, and transmigration was seen as a major vehicle for job creation. Encouraged by the achievements of Repelita III, the Repelita IV plan proposed to settle 750,000 families in the outer islands. Of these, about 425,000 (60%) were to be settled under the fully sponsored pr ofram, including about 110,000 families to be settled on NES tree-crop schemes.- The cost of this program (excluding tree crop development) was projected at Rp 2.9 trillion (US$2.6 billion) in current terms. In recognition of problems encountered in the Repelita III program, a number of changes were made prior to Repelita IV. The first and most important was the formation of a new Ministry of Transmigration (MOT) responsible for site selection and land preparation (formerly in the Ministry of Public Works) as well as settlement and development (functions previously in the DGT). The formation of a Ministry devoted entirely to transmigration was intended both to affirm the importance of transmigration in the national development program and to improve coordination. Budgets for some functions such as the provision of education and health were given to the MOT, although the Ministries of Agriculture and Home Affairs retained control over cheir own funds. 3/ Different figures were issued at various times prior to Repelita IV and numbers varied widely. These figures were set as the program was about to begin. - 14 - 1.35 Organizational changes were also made as a part of the Government's general effort to decentralize planning and implementation. The most impor- tant of these was to shift the procurement of goods and services from central to provincial agencies and from provincial to district levels. Other changes, begun in Repelita III, included increasing the emphasis on tree crops and un- assisted movement. The gradual acceptance of tree crop models was reinforced by surplus rice production in Indonesia as a whole, and by continuing low incomes in both upland and tidal food crop sites. Proposals were developed for an accelerated oil palm development program covering 300,000 ha, and for utilizing private estates to establish tree crops for both local people and transmigrants. The interest in spontaneous migration also increased, neces- sitated in part by financial constraints. 1.36 In addition, Government emphasized a number of new elements in the Repelita IV program: (a) Farm models were be diversified. In addition to integrated farming and tree crops, models were to be developed for livestock, fisheries, industry and mining. (b) Increased participation of indigenous people in settlement projects was to be encouraged. During Repelita III the number of local people in transmigrant settlements was fixed at about 10%, but in Repelita IV this was to depend on the needs and interests of the local population. (c) Increased attention was to be given to the development of existing sites. Second-stage development was to be encouraged and carried out in cooperation with the private sector wherever possible. 1.37 To assist the Government in preparation of the Repelita IV program, early in the plan period the World Bank appraised a second project primarily for site selection and evaluation (Transmigration V, Ln 2578-IND). This proj- ect was intended to identify sites for an additional 200,000-300,000 families and to improve procedures for taking environmental and social factors into account in site selection (see Chapter V). In addition, it provided technical support for supervision of land clearing contractors and for institutional development. The Start-up of the Repelita IV Program 1.38 Between April 1984, when Repelita IV began, and June 30, 1987, about 179,000 families were settled under the fully sponsored program, 17,000 families were settled on partially assisted programs, and an additional 281,000 families were identified that had moved spontaneously (see Table 1.4 and Annex 1, Table 2). It is important to re-emphasize that registration pro- cedures intended to count unassisted migrants in receiving provinces had not yet gone into effect, and that the total number moving may have been larger. Nevertheless, these 477,000 "official" families roughly equalled the Govern- ment's target of about 150,000 families/year including both sponsored and spontaneous settlers. - 15 - Table 1.4: SETTLEMENT FROM APRIL 1, 1984 THROUGH JUNE 30, 1987 BY TYPE OF MIGRANT FAMILY AND ISLAND Partially Island Sponsored assisted Spontaneous /a Total Sumatra 99,100 8,600 183,800 291,500 Kalimantan 44,900 5,800 73,500 124,200 Sulawesi /b 25,500 2,400 18,700 46,600 Irian Jaya 9,400 50 5,100 14,550 Total 178,900 16,850 281,100 476,850 /a Partial enumeration based on government records. 7% Includes the Moluccas and other small eastern islands. Source: Figures provided by the Ministry of Transmigration. 1.39 In the first three years of the plan period, however, the transmi- gration program encountered several major implementation problems. First, new organizational arrangements, new people in key positions in the MOT, plus the relocation of staff from the Ministry of Public Works to MOT caused predict- able adjustment problems. Second, the decentralization of procurement, which entailed the introduction of new procedures at the provincial level, caused delays in the tendering of land clearing and settlement contracts. Third, production problems in tidal sites combined with financial constraints in the Ministry of Public Works led to a slowdown in the development of swamp recla- mation areas. Finally, with the large number of settlements under MOT author- ity, funds were needed in increasing amounts for upgrading activities in existing sites. It is noteworthy that the availability of upland sites was not a major constraint to the start-up of Repelita IV. 1.40 It was also clear that the very large scale of transmigration had heightened the visibility of the program and polarized views on its merits, particularly outside of Indonesia. Some criticism directed toward the program was based on faulty information about the Government's intentions and methods of implementation, and the Government made an effort to address these through a series of meetings coupled with visits both to successful and less successful sites. Other criticism was directed at the Government from single issue interest groups, which were often unaware of the difficult trade-offs involved. Finally, still different concerns were registered by Indonesians and others, who were sympathetic to government objectives, but concerned about high targets, implementation problems and the social and environmental impact of the program. - 16 - Program Revisions as a Result of Declining Oil Revenues 1.41 In January 1986, oil prices fell and declining revenues caused Government to cut the proposed budget. The overall development budget was cut from Rp 10.7 trillion in FY85/86 to Rp 8.3 trillion in FY86/87, a reduction of 22% in nominal terms; and the transmigration sector budget was reduced from Rp 578 billion in FY85186 to Rp 325 billion in FY86/87, a reduction of 44% in budgeted funds, and a decrease of about 10% when compared to FY84/85 expenditures. In response, the MOT reduced its FY86187 targets for settlement on sites already under preparation from 100,000 families to 36,000 fully sponsored families (26,000 families on food-crop schemes and 10,000 on tree- crop schemes). An additional 20,000 families were targeted to receive tree crops under partially assisted programs. 1.42 In May 1986 the MOT budget was further reduced to about Rp 208 bil- lion or about 58% of FY 84/85 expenditures (the last year for which data are available). Under these circumstances priority was given to the maintenance and upgrading of existing sites, to the completion of projects already under- way, and to institutional development. All land clearing in new areas was deferred, except on externally assisted projects expected to settle no more than 3,000-5,000 families/year. With these changes, the MOT began to empha- size improvement in the quality of settlements in order to improve migrant welfare and to attract spontaneous migrants. 1.43 These revisions have significant implications for the transmigration program. Since the upgrading of existing sites is being given highest priority and no land clearing is being carried out on new sites, sponsored settlement is expected to drop in Repelita IV. The total number of sponsored and partially assisted settlers is not expected to significantly exceed the number settled by June 1987 (195,750 families). This would result in settle- ment levels of about 1.0 million people in Repelita IV, or about two thirds the number settled on sponsored programs in Repelita III. 1.44 To adjust to these new conditions, the Government and the Bank have reviewed the Transmigration V project and reduced funds allocated for new site selection and settlement planning. Technical assistance for the supervi- sion of land clearing has been redirected to road maintenance. Given resource constraints, the remainder of Repelita IV is expected to be a period of (a) upgrading and consolidation on existing sites; (b) institutional develop- ment and training; (c) planning for Repelita V with more attention to second stage development, regional development and environmental considerations; and (d) policy formulation intended to promote spontaneous movement in Repelita IV and beyond. - 17 - II. MIGRANT WELFARE A. A Comparison of Income in Sending and Receiving Areas The Data 2.1 In the past five years a number of transmigratio7 studies have provided information on migrant incomes and expenditures.1 Of these, the most comprehensive was a Bank-assisted income survey carried out by the Ministry of Transmigration and the Central Bureau of Statistics (BPS) in February 1985. This chapter draws heavily on the results of the BPS survey because the raw data were available to Bank staff for analysis, and the BPS survey, which was drawn from the National Social-Economic Survey (Susenas), permited comparisons between transmigrant and nontransmigrant households. 2.2 The BPS transmigration survey was carried out in 110 villages in about 22 locations in seven major receiving provinces. In each village, 20 households were randomly selected for interviews and, in total, 2,200 households in Repelita II and III transmigration sites were covered. About 46,000 families lived in survey villages, slightly over 10% of all families moved in Repelita II and III. 2.3 The sampling framework adopted by BPS reflected the distribution of migrants by farm model, period of settlement (Repelita), and class of migrant (sponsored/spontaneous) as of December 31, 1982, when survey preparation began. Upon completion, the survey covered 12 upland and 7 tidal sites, plus two nucleus estate projects, and it included both successful and unsuccessful sites. A list of sites, dates of settlement, population and sample size is provided in Annex 2, Table 1. In order to compare the incomes of transmigrant and nontransmigrant households, BPS extracted a sample of rural households in sending and receiving provinces from the 1984 Susenas. This sample included 17,241 households in rural Java and 2,831 households in rural areas in the seven receiving provinces covered by the transmigration study. 1/ Small-scale surveys: Marshal Khan, "An Inquiry into the Agricultural Economy in Singkut," UNDP/FAO Project INS/78/012, December 1980; Robert Stuart, "Employment in Singkut, 1983," Euroconsult, 1983; Barbara Chapman, "Diet and Production Survey of Sitiung, West Sumatra," Tropsoils Project, December 1984; Levine, et al., "Transmigrants and Tolakis in Kecamantan Tinanggea, Southeast Sulawesi," December 1982; and, Levang and Marten, "Sembamban I: Agro-economic Survey of a Transmigra- tion Center on South Kalimantan," Indonesia-Orstom Transmigration Project, Jakarta 1984. Large-scale surveys: BPS Transmigration Income Survey, 1985; Smallholder Cattle Development Project Socioeconomic Impact Survey, 1985; and MOT survey by the Directorate for Economic Development. These studies are described in Annex 7. - 18 - 2.4 As with any large-scale socioeconomic survey, there are some prob- lems in interpreting BPS data. For example, income from agriculture is based on the farmers' estimates of the value of annual crop production (both sold and consumed); yields and incomes are then divided by twelve to get the aver- age monthly household income from agriculture. This is deemed necessary because of the seasonality of agricultural work. In contrast, household incomes from other productive activities such as handicrafts, construction, trade and transport are recorded for three months and divided by three to get monthly income (again assuming some seasonality) while wages, pensions and rent are recorded for only one month. This may have the effect of slightly understating agricultural income which must be recalled over a Longer period of time. 2.5 A second problem is that only a limited number of responses are possible on income from food crops, and respondents on the average report less than three crops. This means that only major commodities are fully recorded and incomes from home gardens are underreported. Based on BPS expenditure data, the margin of error appears to be about Rp 1,000-2,000/capita/month or about Rp 5,000-10,000/household/month.2/ These problems are not serious when comparing populations within the BPS survey, but they lead to difficulties when comparing the data to other transmigrant surveys that record agricultural incomes in more detail. Despite this, the BPS income survey, when comple- mented by other surveys and micro level studies, provides the best information collected to date on migrant incomes, patterns of employment and welfare con- siderations in Repelita II and III settlements. Incomes in Sending and Receiving Areas 2.6 Table 2.1 summarizes the results of the 1984 Susenas in rural areas of sending and receiving provinces and of the 1985 transmigration income sur- vey. This sample shows average household incomes in 1984 of Rp 62,000/month in sending provinces and Rp 83,000/month in receiving provinces. If 1984 figures are updated to 1985 prices to make these figures comparable to the transmigration survey, average monthly household incomes are Rp 59,000 (US$53) in transmigration sites, Rp 67,000 (US$61) in sending areas, and Rp 91,000 (US$82) in receiving areas. (All dollar figures assume the exchange rate of Rp 1,100/US$1.00 in effect at the time this report was prepared.) 2.7 As noted, there are some problems in interpreting these data. For example, the fact that the value of home garden production is understated means that the incomes of subsistence food producers (such as transmigrants) are also understated if compared to rural residents depending on wage incomes to cover food expenses. In addition, many migrants may have been on site for only two to three years and, therefore, may not have reached stable income levels; and migrant households in general may be slightly smaller and younger than other households and have lower incomes because of this. On the other 2/ A 1983 survey in transmigration sites in Lampung estimated home garden production at Rp 51,750 annually. If updated to 1985 prices and divided by 12 this would also be about Rp 5,200/household/month. - 19 - hand, the purchasing power of households in the inner islands and the outer islands differs. In many more remote outer island areas, prices are higher than they are in Java and a higher income is required for a comparable standard of living. Table 2.1: COMPARISON OF TRANSMIGRANT INCOMES WITH INCOMES IN SENDING AND RECEIVING PROVINCES Average Average Average household expenditures household income income as a % in 1985 prices Sample (Rp/month) of income Rp/month US$/year Rural Sending Areas - 1984 West Java 3,452 67,702 90 73,660 804 Central Java 3,527 53,537 85 58,248 635 D.K. Yogyakarta 1,275 66,970 83 72,863 795 East Java 4,155 58,764 84 63,935 697 Bali 1,461 77,047 78 83,827 915 NTB 1,506 56,208 83 61,154 667 Total 15,376 61,738 85 67,171 733 Rural Receiving Areas - 1984 Riau 380 104,951 84 114,187 1,246 South Sumatra 883 79,305 95 85,284 941 Central Kalimantan 271 76,894 91 83,661 912 South Kalimantan 229 75,091 95 81,699 891 East Kalimantan 393 89,865 86 97,773 1,066 Central Sulawesi 195 58,407 79 63,547 693 Southeast Sulawesi 480 86,346 81 93,944 1,025 Total 2,831 83,396 88 90,735 990 Transmigrants in Receiving Provinces - 1985 Riau 300 65,513 95 65,513 715 South Sumatra 981 52,888 87 52,888 577 Central Kalimantan 320 39,131 97 39,131 427 South Kalimantan 239 55,131 97 55,131 601 East Kalimantan 100 107,893 63 107,893 1,177 Central Sulawesi 100 60,231 78 60,231 657 Southeast Sulawesi 160 89,062 57 89,062 971 Total 2,200 58,343 84 58,343 636 Source: Sending and receiving areas, 1984 Susenas; transmigration areas, BPS Transmigration Income Survey, 1985. - 20 - 2.8 If we were to assume that transmigrant incomes understate the value of agricultural production by Rp 10,000/month, transmigrants would have aver- age incomes that are about the same as those in sending areas, but are still less than incomes in the outer islands. The relative success of rural families in the outer islands is due almost entirely to higher incomes from tree crops (Rp 16,000 in the outer islands vs. Rp 2,985 for those in sending areas and Rp 1,660 for transmigrants), as food crop incomes cover the same range in all groups (see Annex 2, Table 2). 2.9 The fact that adjusted average transmigrant incomes are about the same as average incomes in sending areas does not mean that individual trans- migrants are no better off than they were in Java. The World Bank's poverty index for Indonesia is about Rp 10,000/capita/month or Rp 50,000/house- hold/month. Fully 40% of rural Javanese households in the BPS survey fell below this figure in 1984. Virtually all transmigrants are drawn from these lower income households, but in transmigrant communities 50% of all families are above the poverty level. This is a significant achievement. Transmi- grants also have assets that are not valued in the income survey. They own land, a house, and possibly livestock; they have the means to meet subsistence needs and they have reduced expenses for items such as housing. 2.10 These considerations are reflected in the judgement which migrants make of their own situations. As Table 2.2 indicates, when asked whether they are better off or worse off than before transmigration, about two thirds of all migrants say they are better off than they were, while 23% of Repelita II migrants and 14% of Repelita III migrants feel they are worse off. This rela- tively high proportion of "dissatisfied customers," in comparison with earlier studies, reflects the inclusion in this study of migrants in critical sites. Table 2.2: TRANSMIGRANT REPORTS ON INCOMES IN TRANSMIGRATION AREAS COMPARED TO THOSE PRIOR TO TRANSMIGRATION Repelita II Repelita III Total Percent reporting that their incomes are: Better 66 67 67 Worse 23 14 16 Just as good 10 17 15 Just as bad 1 2 2 (Sample) (540) (1,598) (2,138) Source: BPS Transmigration Income Survey, 1985. Three villages (60 people) with migrants arriving in 1984 are excluded from this sample since they have no income from agriculture yet. - 21 - Income Distribution 2.11 Data on income distribution are shown in Table 2.3. Five income benchmarks are given. At the lower end of the spectrum, Rp 20,000/month/ household generally implies serious deprivation; Rp 30,000/month/household is regarded by Government as the minimum subsistence level, defined as 240 kg Table 2.3: DISTRIBUTION OF HOUSEHOLD INCOMES AMONG TRANSMIGRANTS AND NONTRANSMIGRANTS Monthly household incomes (% households) Sample Under Under Under Over Over size Rp 20,000 Rp 30,000 Rp 50,000 Rp 75,000 Rp 100,000 Sending Areas - 1984 West Java 3,452 4 10 37 31 16 Central Java 3,527 8 23 59 18 9 D.K. Yogyakarta 1,275 5 14 44 29 16 East Java 4,155 6 19 53 22 11 Bali 1,461 1 6 30 40 21 NTB 1,506 8 24 55 21 10 Total 15,376 5 15 43 25 13 Receiving Areas - 1984 Riau 380 0 2 9 72 46 South Sumatra 883 2 5 26 42 23 Central Kalimantan 271 1 1 24 47 20 South Kalimantan 229 2 6 29 40 24 East Kalimantan 393 2 6 22 51 31 Central Sulawesi 195 2 5 44 19 8 Southeast Sulawesi 480 1 5 24 52 31 Total 2,831 1 4 24 48 27 Transmigrants - 1985 Riau 300 4 10 34 39 16 South Sumatra 981 9 25 59 19 10 Central Kalimantan 320 11 32 68 10 3 South Kalimantan 239 1 11 46 26 10 East Kalimantan 100 1 4 12 71 49 Central Sulawesi 100 0 17 50 19 10 Southeast Sulawesi 160 2 2 21 34 13 Total 2,200 6 20 50 26 13 Source: BPS Transmigration Income Survey, 1985. - 22 - milled rice equivalent per capita per year,3/ and Rp 50,000/month/household is about equal to the poverty level in the inner islands and in the provinces to which a majority of transmigrants are sent. At the upper end of the spectrum, Rp 75,000/month/household is a figure at which food consumption begins to decline as a percentage of total household expenditures, implying that food needs have been met. These benchmarks have been developed and used in conjunction with the Susenas data in the past, but are subject to the interpretation problems mentioned earlier. 2.12 In the sample from the 1984 Susenas, about 5% of rural households in sending areas and 1% of rural households in receiving areas fell below Rp 20,000/month, while about 6% of transmigrant incomes were below this level in 1985. Similarly, in 1984 about 15% of households in sending areas and 4% of households in receiving areas fell below Rp 30,000/month compared to 20% of migrant households. Using Rp 50,000/household/month as a cutoff point for the poverty line, 40% of inner island households, 24% of households in receiving provinces, and 50% of migrant households fell below this amount. About 25% of inner island households and 26% of transmigrant households exceeded Rp 75,000/month, compared to 48% of outer island households. About 13% of inner island and transmigrant households make more than Rp 100,000/month compared to 27% of outer island households. 2.13 In summary, both the level and distribution of transmigrant incomes closely resemble those in sending areas. According to these data, perhaps 20% of migrant households are near subsistence levels, but fully 50% are above the poverty line. This is a significant accomplishment. Migrant incomes do, however, vary considerably between receiving provinces. Those in East Kalimantan and Southeast Sulawesi appear to be doing as well or better than their local counterparts, while those in sites sampled in South Kalimantan, South Sumatra and Central Sulawesi are not doing as well. It is noteworthy that surveyed migrants in Central Kalimantan are very poor. More detailed analysis by province is not possible, as the numbers in most provinces are small. Further analysis therefore focuses on the factors affecting migrant success within the total migrant population. B. Factors Determining Transmigrant Incomes Farm Model and Length of Time on Site 2.14 Perhaps the most important findings in the BPS survey are related to income differences by farm model and length of time on site. As Table 2.4 indicates, the highest average monthly household incomes in transmigrant com- munities are in earlier (Repelita II) swamp reclamation (tidal) sites while the lowest average incomes are in recent (Repelita III) tidal sites. Recent upland sites based on rainfed food crop agriculture are doing somewhat better than older upland sites in this sample, and those with tree crops have 3/ Milled rice retailed at about Rp 300/kg in 1984/85 when the survey was done. Therefore the calculation is 240 kg/person x 300 Rp/kg x 5 people = Rp 360,000/12. - 23 - intermediate income levels and no apparent change in total income between Repelita II and Repelita III sites. The sample size for migrants with tree crops is too small to provide statistically significant results. Table 2.4: INCOME AND INCOME DISTRIBUTION, BY FARM MODEL AND REPELITA /a Family Income per capita per month (% sample) Sample income Under Rp 6,200 Rp 10,000 Over size (Rp/mo) Rp 6,200/b to 10,000/c to 20,000 Rp 20,000 Repelita II Tidal 99 77,697 11 16 54 19 Upland 341 52,825 33 24 30 13 Upland/trees /d 100 66,595 13 29 35 23 Subtotal 540 59,934 25 24 35 16 Repelita III Tidal 501 39,791 38 35 21 6 Upland 977 65,801 10 26 43 21 Upland/trees /d 120 63,923 12 23 42 23 Subtotal 1,598 57,505 19 29 36 16 /a Total sample of 2,138 excludes three villages of migrants who arrived mainly in 1984 (without time to establish comparable agricultural production before the survey). /b The Government's minimum subsistence level per year, or the consumer value of 240 kg milled rice per year. /c The World Bank's projection of the absolute poverty level. 7d Tree crops not yet in production. Source: BPS Transmigration Income Survey, 1985. 2.15 These findings are also reflected in the proportion of migrants in different income brackets. For purposes of this analysis, it was possible to divide total family income by the number of people in each household to calcu- late a monthly per capita income figure. The findings are striking. As Table 2.4 indicates, more than one third of settlers in the poorest type of sites (Repelita II upland and Repelita III tidal sites) have incomes below the subsistence level, in contrast to about 10% of settlers in other types of sites. Similarly, nearly 60% of settlers in Repelita II upland sites and nearly three quarters of those in Repelita III tidal sites are below the absolute poverty level, while the number in other groups is only about one third. These figures have been reviewed with officials in the Ministries of Transmigration and Agriculture who confirm that older rainfed sites have lower - 24 - incomes than more recent ones and that a number of Repelita III tidal sites have serious production problems. Tidal sites are also heavily represented on the MOT's list of least successful sites. About 40% of sites on this list are swamp reclamation sites, even though these sites were only 18% of all those settled in Repelita III. 2.16 In order to determine whether these communities are covering their nutritional requirements, Table 2.5 shows food crop production from major sub- sistence crops averaged over all households in the six samples. These data show significant surplus production from Repelita II tidal sites, sufficient production in Repelita III upland sites to cover the communities' nutritional requirements, and production below the levels required for community self- sufficiency in Repelita II upland sites and Repelita III tidal sites. Sites based on tree crops do not meet their subsistence needs from food production, but from wage work that is a part of the development program. Table 2.5: PERCENT OF COMMUNITY ENERGY REQUIREMENTS MET FROM REPORTED PRODUCTION, BY FARM MODEL AND REPELITA /a Average % community energy Sample production (kg/family) requirements from size Rice /b Corn Cassava own production Repelita II Sites Tidal 99 1,484 473 437 185 Upland 341 592 86 925 88 Upland with tree crops /c 100 603 38 1,076 89 Repelita III Sites Tidal 501 740 150 357 87 Upland 977 788 249 753 116 Upland with tree crops /c 120 481 118 1,233 95 /a Total sample of 2,138 excludes 3 villages of migrants arriving mainly in 1984. /b Unmilled rice. 78 Tree crops not yet in production. Source: BPS Transmigration Income Survey, 1985. 2.17 These data are also supported by farmers' reports on the amount of rice sold. Table 2.6 shows annual rice production, sales, and value of sales. On average, farmers in Repelita II tidal sites are the only sponsored transmigrants in the income survey who sell a significant (US$106) amount of rice. All other farmers, on average, sell less than US$25 of rice per year. - 25 - This indicates that while rice production is undoubtedly important for home consumption, it is not a major source of cash income for transmigrants, other than those in early tidal sites. These data also suggest that a major determinant of migrant welfare is the type of site to which the migrant is moved. Questions of sustainability will be reviewed in Section C. Table 2.6: ANNUAL PRODUCTION AND MARKET SALES OF RICE FOR SPONSORED TRANSMIGRANTS BY FARM MODEL, 1985 /a Sample Rice /b Amount of Reported value size produced rice sold /b of sales ----- (kg/family) ----- (Rp'000) Repelita II Tidal 79 1,620 790 117 Upland 291 600 110 29 Upland with trees 80 530 20 6 Repelita III Tidal 406 730 140 20 Upland 816 840 200 27 Upland with trees 95 540 90 10 Overall Average 1,767 780 180 28 /a Sponsored transmigrants who arrived in 1984 are omitted from this table because they would have just begun agricultural production. /b Unmilled rice. Source: BPS Transmigration Income Survey, 1985. Sources of Household Income 2.18 BPS Survey. BPS data on sources of migrant income are summarized in Table 2.7 by Repelita and farm model. These data indicate several important points: (a) Incomes from food crops and agricultural production are generally low. Less than one quarter of total income is from food crops, except in older tidal sites where the amount is one third. Less than half of income in any farm model is from agriculture. - 26 - Table 2.7: SOURCE OF INCOME BY FARM MODEL AND REPELITA, BPS SURVEY (Average Monthly Household Income) Govt. Sample Food Total agri- assist- size crops culture /a Wages Other ance Total Values (Rp) Repelita II Tidal 99 24,787 32,859 15,803 28,467 567 77,696 Upland 341 10,557 15,176 16,337 17,278 4,032 52,823 Upland with trees 100 7,781 32,761 9,034 20,385 4,415 66,595 Subtotal 540 12,652 21,674 14,887 19,905 3,468 59,934 Repelita III Tidal 501 9,954 16,214 4,619 13,898 5,060 39,791 Upland 977 15,321 23,294 14,807 17,651 10,049 65,801 Upland with trees 120 9,636 17,195 25,008 14,962 6,758 63,923 Subtotal 1,598 13,212 20,616 13,012 16,844 7,033 57,505 Percentages (%) Repelita II Tidal 99 32 42 20 37 1 100 Upland 341 20 29 31 33 8 100 Upland with trees 100 12 49 13 31 7 100 Subtotal 540 21 36 25 33 6 100 Repelita III Tidal 501 25 41 12 35 13 100 Upland 977 23 35 23 27 15 100 Upland with trees 120 15 27 39 23 11 100 Subtotal 1,598 23 36 23 29 12 100 /a Includes food crops. Source: BPS Transmigration Income Survey, 1985. (b) The two farm models with the lowest incomes, i.e., older upland and new tidal sites, both have average incomes from food crops of about Rp 10,000/month. However, farmers in recently settled tidal areas also have very low off-farm incryes, and about one third of the wage work in other food-based sites.- 4/ The reason for this is unclear. Access is a difficult problem in tidal sites, but Repelita II sites which are also in remote areas report wages which are three times as high. - 27 - (c) The farm models with the highest total agricultural incomes, Repelita II tidal and tree crops, also have the highest incomes under "other," a category which includes handicrafts, small enter- prises, trade, transport and construction. This may indicate that these communities are turning surplus agricultural production into other forms of employment, particularly in small industries and services. 2.19 The proportion of total income from off-farm work varies signifi- cantly between sites in the BPS survey (see Annex 2, Table 3) and ranges from one-third in Sei Rateh (a tidal site) and Teluk Dalam (an upland site) to more than two-thirds in 8 of the 22 locations. The relatively low incomes from agriculture in relation to previous expectations, and the high incomes from nonagricultural work, have significant implications for an understanding of the transmigration process. 2.20 Other Survey Findings. Since there is some concern that the BPS survey may underreport agricultural income, BPS data were compared with a number of other studies. For example, an IFAD-assisted project to provide cattle to transmigrants also carried out a study in February 1985, the same month as the BPS study. It differs from the BPS survey in several respects. All smallholders who were interviewed were, in theory, full-time farmers eligible to receive cattle under the livestock distribution program. Of the 1,200 families interviewed, 600 had received cattle and 600 had not. Since cattle in this project are distributed first to leading farmers, the 600 recipients are probably representative of the best farmers in transmigrant communities. A second difference is that the study covered only gross cash income and did not estimate the value of subsistence production. Even though nonagriculturalists were excluded from this study, the survey found that 42% of cash income was from off-farm work, with the percentage ranging from 30% to 75% over 21 IFAD sites and adjacent areas (Table 2.8). 2.21 Several other studies have shown equally high incomes from off-farm work. For example, SCET, the firm that supervised initial development of the Bank-assisted site at Baturaja, noted that two-thirds of family income was from off-farm activities in the initial years, particularly from wage work associated with tree crop establishment. Similar studies carried out by Euroconsult in six Repelita II settlement sites in Singkut showed 75% of cash income from off-farm work. The same survey in an adjacent World Bank-assisted site in Kubang Ujo showed 81% of cash income from off-farm work. MOT surveys, described more fully in Section C, also report more than 40% of total family income from wage work. 2.22 Even though the BPS, IFAD and Euroconsult surveys are not entirely comparable, due mainly to the difference in handling production consumed at home, the figures for off-farm work are consistently high (Table 2.8). Sev- eral possibilities exist to explain rhis. First, farmers may be attracted to the certainty of wage work generated by site development. Second, farmers may be experiencing declining yields due to poor soils or a lack of inputs, forcing them to find off-farm work. It is also possible that recent settlers, benefiting from improved site selection and services, will find agricultural production more remunerative and will not seek off-farm work so readily in the future. - 28 - Table 2.8: MONTHLY HOUSEHOLD INCOMES FROM THE IFAD AND EUROCONSULT BPS AND MOT SURVEYS Sample On-farm Off-farm Total Off-farm size income income income income as ----------- (Rp) ----------- % of total Gross Cash Income /a IFAD 1,200 18,000 13,000 31,000 42 Euroconsult lb 180 8,000 25,000 33,000 75 Net Income /c BPS-Repelita II /d 341 33,000 20,000 53,000 38 BPS-Repelita III 7d 977 23,400 43,000 66,000 65 MOT-Development stage 389 30,000 24,000 54,000 44 MOT-Stabilization stage 413 41,000 34,000 74,000 46 /a Excludes the value of agricultural production consumed by the household. 7r Survey conducted in April 1983 in Singkut transmigration site, Jambi Province. Off-farm income includes pensions. Singkut was settled from 1974 to 1979. /c Includes the value of agricultural production consumed by the household. Excludes government subsidies and pensions. /d Sponsored migrants in upland sites only. 2.23 Government officials tend to believe that transmigrants neglect their farming because of the wide availability of off-farm work generated by transmigration development, and should this source of work disappear. transmi- grants would return to agriculture. Other observers believe that agricultural production is limited and that if off-farm work declines, incomes will suffer. These important issues are considered again in Chapters V and VI. Whatever the case, these data suggest that many transmigrants currently find off-farm work more remunerative than full-time agriculture. The data provide a striking contrast to the widely held picture of transmigrants as subsistence agriculturists largely isolated from the regional economy. C. The Sustainability of Settlements Yields and Production over Time 2.24 MOT and BPS. Although many transmigration communities have persist- ed in the outer islands for several decades, a major question asked of trans- migration is whether agricultural production can be sustained on marginal soils and, if so, at what level. To address this question, among others, the MOT Directorate of Economic Development carried out a major income survey in 1985 using MOT staff. This study included 36 upland locations and 5 tidal sites, and covered about 1,500 transmigrant families, all settled in Repelita III. For purposes of MOT analysis, families were divided into those in the - 29 - consolidation phase (1 to 1-1/2 years on site), the development period (1-1/2 to 2 years) and the stabilization period (3 to 5 years). The results bear a striking resemblance to those of the BPS survey. As Table 2.9 shows, average monthly incomes in upland sites are about the same in both studies and in- crease for Repelita III migrants the longer they have been on site. The MOT survey also supports the BPS survey in showing relatively low incomes from food crops (about one-third), and from agriculture in general (about one- half), and significantly lower incomes in recent tidal sites when compared with recent upland sites. MOT has concluded from the evidence of increasing food crop income in the five year period, that migrants are meeting their subsistence needs and increasing their production over time. Table 2.9: COMPARISON OF BPS AND MOT INCOME SURVEYS Total income Food crop income MOT stage/BPS year Sample size (Household income, Rp/month) Upland Sites MOT BPS MOT BPS MOT BPS Consolidation/1983 396 194 28,500 20,700 8,600 12,800 Development/1982 389 175 53,800 54,300 19,300 9,600 Stabilization/1980 413 204 74,100 75,900 25,000 24,800 Tidal Sites Consolidation/1983 98 9 12,500 28,100 3,700 3,000 Development/1982 91 133 37,000 33,500 11,700 10,700 Stabilization/1980 91 109 48,200 32,000 22,300 9,900 2.25 However while incomes from food crops in upland sites in the BPS survey are highest for settlers who have been on site for five to six years, they are lower for those who have been on site nine to ten years (Table 2.10). Farmers from this period were not covered in the MOT survey. Rice production in the BPS survey also shows a decline from 800-900 kg p.a. in the first five years to 300-400 kg p.a. in years 8 and 9. These data are difficult to inter- pret since they are not time series data, but a cross-section of communities settled in different years. Thus, while it is likely that older sites have lower agricultural incomes due to declining soil fertility, it may also be possible that the improved conditions in Repelita III settlements and the large scale of such settlements (which permit the growth of markets and income diversification) will allow income growth after year 5. Neither assertion can be definitely confirmed from the data at this time. - 30 - Table 2.10: AVERAGE MONTHLY FAMILY INCOME AND ANNUAL RICE PRODUCTION BY YEAR OF ARRIVAL, SPONSORED TRANSMIGRANTS, BPS SURVEY Upland sponsored Tidal sponsored Unmilled Unmilled Year of Sample Total Food crop rice Sample Total Food crop rice arrival size income income production size income income production _---- (Rp) ------ (kg) ----- (Rp) ------ (kg) 1974 - - - - 15 110,562 12,528 1,272 1975 66 44,617 9,871 337 15 61,333 26,146 2,767 1976 141 46,887 5,014 506 24 59,539 23,500 628 1977 22 59,796 7,456 2,190 9 39,574 24,293 2,500 1978 61 56,986 7,096 320 14 58,682 27,748 1,221 1979 71 64,895 20,634 1,224 14 64,361 37,977 1,964 1980 204 75,886 24,781 887 109 31,937 9,941 763 1981 286 52,833 13,380 912 159 37,645 9,817 748 1982 175 54,270 9,620 638 133 33,456 10,704 778 1983 194 20,726 12,806 931 9 28,111 3,032 323 Source: BPS Transmigration Income Survey, 1985. 2.26 Migrant Reports. One problem with the previous information is that it is not chronological, that is, it does not document changes on a single site over time. However, some information is available from the BPS survey on changes in production and area under production for the same farmer. Settlers were asked if their present rice production had declined, improved or remained about the same relative to two years ago and if their area under cultivation had increased, decreased or remained the same (Table 2.11). About 40% of settlers overall reported that their rice production was lower than two years previous, and about the same percent reported increased production. Consis- tent with other results presented in this chapter, settlers on older tidal sites most frequently reported production increases (62% reported an in- crease), while 40% of Repelita III tidal farmers experienced declining produc- tion. Over half (56%) of older upland farmers reported declining production while only 31% showed improvement. About half of all older upland farmers also reported they had less land under cultivation than they had two years previously. 2.27 In contrast with upland areas, where older sites on average have lower levels of production, reasonable production has been maintained in some tidal areas for 20 years or more without fertilizers. Recently, however, as evidenced in Tables 2.10 and 2.11, new sites have experienced serious agricultural problems. These are most likely related to site selection (i.e., location on deep peats and acid soils), inappropriate farm layout, and inadequate pest control. Technical specialists remain optimistic that these difficulties can be overcome, and these findings and their implications are discussed in Chapter V. - 31 - 2.28 Thus, the data do not tell a uniform story. This may be due to the fact that farmers have good and bad years, and that some sites are improving while others deteriorate. Settlers who take land out of production may be Transmigrant family with neighibours in front of a new house constructed with wages earned from estate development. doing so because yields are declining and returns are low, or because off-farm work is increasing, or a combination of both. The lack of conclusive evidence on these issue argues strongly for long-term monitoring of yields and production on representative upland and tidal sites. - 32 - Table 2.11: SPONSORED TRANSMIGRANTS' CURRENT RICE PRODUCTION AND AMOUNT OF LAND IN PRODUCTION RELATIVE TO TWO YEARS PREVIOUS /a Sample % % % Farm model size Increase Decrease No change Change in Production Relative to Two Years Previous Repelita II Tidal 78 62 10 28 Upland 277 31 56 13 Upland with trees 78 44 35 21 Repelita III Tidal 396 36 40 24 Upland 606 44 35 21 Upland with trees 76 34 33 33 Overall 1,511 40 39 21 Change in Land Under Production Relative to Two Years Previous Repelita II Tidal 79 52 14 34 Upland 291 37 47 16 Upland with trees 80 42 39 19 Repelita III Tidal 397 43 10 47 Upland 622 39 10 51 Upland with trees 79 53 13 34 Overall 1,548 41 19 40 /a Sponsored transmigrants who arrived after 1982 were omitted from the table because two years previous to the survey they would still have been in their area of origin. This cross-tabulation on yields also omits 37 respondents who did not answer the survey question on change in production. Source: BPS Transmigration Income Survey, 1985. Return Migration 2.29 One additional piece of evidence on the sustainability of settlements comes from information on the number of families who leave transmigration sites. Government estimates that return migration is about 1-2% during the five-year period when settlements are under the authority of the MOT, but these figures have been regarded with some skepticism since higher return rates are typical of large-scale voluntary programs. - 33 - Table 2.12: NUMBER OF FAMILIES IN REPELITA II AND III TRANSMIGRATION SITES ON ARRIVAL (A) AND IN FEBRUARY 1985 (B) Sponsored families All families /a Settled 1985 B/A Settled 1985 B/A (A) (B) x 100 (A) (B) x 100 Repelita II Sites Upland 5,483 5,676 109 6,143 6,680 109 Upland with tree crops 1,676 1,703 102 1,702 2,524 148 Tidal 1,255 1,210 96 1,390 1,453 105 Subtotal 8,414 8,589 102 9,235 10,657 115 Repelita III Sites Upland 18,265 16,912 93 21,736 21,839 100 Upland with tree crops 2,926 2,952 101 3,434 3,631 106 Tidal 10,770 10,408 97 11,761 11,962 102 Subtotal 31,961 30,272 95 36,931 37,432 101 Total 40,375 38,861 96 46,166 48,089 104 /a Includes sponsored, retired military and local families. Source: BPS Transmigration Income Survey, 1985. 2.30 BPS data throw some light on this subject. In addition to income data collected by the enumerators, information was collected for each village on the number of transmigrant families on arrival and the number residing in the community at the time of the survey.5/ As the last column in Table 2.12 indicates, transmigration communities on average maintain their populations and increase their number of households over time. Repelita II sites included in the survey increased an average of 15% and Repelita III sites also show stability or growth in total households. It is evident from the table's first two columns, however, that some turnover of households has occurred. Of the 93 villages where return migration occurred, 23 (26%) show a reduction in the number of sponsored migrants. 2.31 Unfortunately the manner in which the data were collected obscures the actual turnover rate among households. When the sons and daughters of transmigrants grow up and marry, the number of households increases. Some enumerators recorded the increase from the division of sponsored migrant 5/ The census was conducted in 110 villages and data on return migration were collected for 93 villages covering over 40,000 households. - 34 - families under "sponsored migrants," while others included this under the category "other." The records look very different. For example, if a commu- nity began with 500 families, 50 left and 70 new families were formed, the first type of record would show a net increase of 20 sponsored families with no record of leavers, while the second type of record would show a loss of 50 households or 10%. To assess the proportion of households leaving transmigra- tion sites, a list was prepared of only those Repelita III villages where the enumeraor clearly followed the second procedure and the number of leavers was clear.6 Of the Repelita III villages 62% were in this category. These villages show an average turnover rate of 13%, 15% for upland areas, 12% for tree crop areas (consisting entirely of return migration in one government- financed estate in Batulicin) and 9% in tidal areas. Neither turnover by farm model nor by village is closely related to the number of families below the poverty line. 2.32 These figures are not unexpectedly high, nor, of themselves, a cause for alarm. Not all families attracted to transmigration areas are well suited to the transmigration life and the fact that families leave and are replaced attests both to the voluntarism of the program and the pressure on spontaneous migrants to find land. On the other hand, the fact that some few sites lose substantial numbers of settlers (20%) suggests that these sites are doing less well than others and such information could be used to identify potentially serious problems in the field. It is noteworthy that even in poorer areas in the survey there have been few cases of significant abandonment of transmigra- tion sites. D. Other Survey Findings Incomes and Household Characteristics 2.33 Table 2.13 summarizes BPS data on income broken down by the charac- teristics of migrant households. As the table shows, those few farmers who have retired from the military and elected to settle in transmigration sites have significantly higher incomes than other settlers. This appears to be due entirely to pensions (which average Rp 40,000-60,000/month) rather than to differences in agriculture or household production. Sponsored and spontaneous migrants do about equally well, although spontaneous farmers do not generally obtain the benefits given to sponsored families. 2.34 Of some significance is the fact that local settlers appear to do slightly better in transmigrant communities than transmigrants. This may be due to two factors. First, source of income. Sumatran settlers show higher incomes from trade, while local Kalimantan and Sulawesi settlers have higher government subsidies and pensions. Second, selection. There is some evidence that local families who are not doing well leave transmigration sites and return to their nearby villages. This would tend to bias the sample toward those who have better incomes. 6/ Repelita II villages were excluded as most had been handed over to the provinces and recent data are incomplete. - 35 - Table 2.13: MONTHLY HOUSEHOLD INCOME BY TYPE OF TRANSMIGRANT (Rp/month) Sample Total Agricul- Wage income Type of migrant size income ture and pensions Other /a Type of Movement Sponsored 1,800 54,441 18,876 22,370 13,195 Spontaneous 152 56,741 23,533 27,113 16,095 Retired military 57 117,706 16,325 92,056 9,325 Local resident 186 74,045 20,581 33,270 20,194 Area of Origin Transmigrants Java 1,856 55,757 18,490 24,319 12,988 Bali 133 60,788 18,004 23,924 18,860 Nontransmigrants Sumatra 114 71 $24 18,252 43,522 9,450 Kalimantan 71 77,977 21,924 27,885 28,168 Sulawesi 24 82,741 23,909 24,412 34,420 Level of Education None 435 51,357 19,141 19,529 12,687 Less than primary 798 52,649 17,561 23,167 11,921 Primary 759 61,191 19,866 27,061 14,264 Junior high school 134 81,467 14,759 39,801 26,907 High school 71 76,954 19,356 42,189 15,409 Age of Household Head <-20 21 52,046 12,009 25,766 14,271 21-25 198 52,791 16,832 22,654 13,305 26-30 397 55,044 16,847 22,519 15,678 31-35 394 56,608 18,445 26,232 11,931 35 + 1,189 60,179 19,610 26,543 14,026 /a Includes home production other than agriculture (handicrafts, etc.), transport, trade and other nonwage income. Source: BPS Transmigration Income Survey, 1985. 2.35 Income also appears to be slightly correlated with education, although the data indicate this is due to increased wages and off-farm earn- ings, rather than increased agricultural production. Since village officials, teachers and the like have slightly higher incomes from wages and also have more education, this can be easily explained. Income is also very slightly related to the age of the household head, but in this case the difference is attributed to increased agricultural production. A breakdown by agricultural income shows small but continuous increases in all categories of agricultural - 36 - production including food crops, tree crops and livestock over time. This may be a function of merging upland with tidal populations, as upland sites alone do not show this effect. Increases in agricultural income with the age of the household head are also slight when measured on a per capita basis. A regression analysis run on these variables shows that only the association between retired military and total income is statistically significant. Migrant Perception of Welfare by Site 2.36 The BPS survey included several qualitative questions intended to assess migrants' perceptions of their situation in transmigration areas. These data, summarized in Table 2.14, indicate that over 60% of all migrants in all categories think their incomes have improved over what they were prior to migration. Less than 15% believe their incomes are worse, except in Repelita II upland sites where 31% feel their incomes have deteriorated. About 50% of all migrants believe their incomes have improved over the past two years, but 22% report that they have declined and about the same number report incomes which are the same. On transport, the migrants provide their most negative assessment, with more than two thirds saying that transportation is worse than in the sending area and only 17% saying it is better. About 40% say their health is better and 49% say it is just as good as in Java, while only 11% say it is worse. This reflects improved access to health facilities in transmigration areas and relatively low health standards in the sending areas. Migrants in tidal areas do not say they are less healthy than at home, despite the reputation of tidal areas for having serious health problems. Those that say they have worse health are most frequent in Repelita II upland communities where incomes are very low. 2.37 Among those sites which have both low incomes and negative reports, the most conspicuous are those villages in Pematang Panggang settled in Repe- lita II. About 46% of farmers in these settlement areas say their incomes are worse than they were before, and 40% say incomes are declining. Fully 34% say their health is worse, more than twice as many as in any other site. Migrants in more recently settled areas of Pematang Panggang are not so negative. Only 8% say they are worse off than before, although 33% say their incomes are declining. They are also well within the range of overalL responses on health, with more than 50% feeling that their health is just as good as it was before they moved. 2.38 Several features of Pematang Panggang explain this situation. The community was settled during 1974-84 and has the misfortune to be almost entirely isolated during the rainy season. Road maintenance is poor and a proposed highway which would have penetrated the area has not been construc- ted. Under these circumstances farmers in the survey show both low returns to agriculture and limited off-farm work (see Annex 2, Table 3). Farmers in new communities appear to have higher returns to agriculture at present, but whether production can be sustained has yet to be demonstrated. The role of infrastructure in regional development is discussed further in Chapter IV. - 37 - Table 2.14 MIGRAT PERCEPTIONS OF CONDITIONS IN TRANMN IGRATION SITRS (Percent of Respondents) Income no. and Income nov and Transport before trans.igration two years ago here and In Java Health here and In Java Jost Just Just Juat Just lost lost That Site Retter Worse as good as had Better Worse as good as bad letter Worse as good as bad Retter Worse as good as bad Repelita IT Sites Tidal Sei Rateh 85 5 10 0 72 3 26 0 1 92 S 0 15 10 74 n Upang Delta 88 13 0 0 13 5 80 3 25 18 58 0 53 0 4R 0 Tanah Grogot 80 5 15 0 65 ln 25 0 58 0 42 n 37 5 53 5 Subtotal 85 8 7 0 46 5 47 1 22 44 34 0 35 5 9q 1 Upland Pematang Panggane 46 46 7 1 47 40 10 1 16 70 2 1 73 34 42 n Ba.bulu Darat/Sepaku 65 8 28 0 48 15 20 18 40 43 18 0 90 5 43 3 Malonas 88 9 2 1 R7 2 I1 1 21 79 0 0 67 9 24 0 Subtotal 60 31 8 1 59 26 11 4 2n 74 3 21 39 23 17 1 Upland w/Treeo Baturaja 69 13 18 0 64 22 14 0 43 42 15 0 44 7 4R 0 Subtotal 69 13 18 0 64 22 14 0 43 42 15 0 44 7 48 0 - Repelita 1I Total 66 23 in 1 57 21 19 3 25 63 11 1 39 17 43 1 Repelita III Sites Tidal Air S.gtha- 84 4 13 0 66 16 17 0 15 so s 0 54 7 39 n Terusan Tengah 37 35 24 4 25 51 21 2 1Q 67 12 3 58 A 33 0 Pangkoh 53 9 23 15 27 28 24 22 5 90 40 5 17 A 71 3 Saka Lagon/Sf.gga1 Muhur 70 8 13 10 48 i1 in 33 9 99 0 0 95 10 39 0 Subtotal 64 13 18 5 45 27 20 8 13 71 1S 2 46 8 46 I Tpland Belis 55 21 22 2 45 24 28 3 33 36 74 6 3R 6 56 1 Teluk Ruantan 60 15 25 0 52 12 13 3 20 62 18 0 42 7 52 n Pematatg Pangga.q 85 8 7 1 93 31 14 1 7 RQ 2 2 33 9 95 I S-oggai W,aras 51 14 16 0 31 20 49 0 I 41 42 14 19 10 71 0 Batullicin 76 11 12 1 54 13 74 8 17 67 17 4 4n 9 51 0 ltu-ai/Hanlalipan 65 7 28 0 31 16 48 4 7 8R 5 n I5 7 78 n Teluk Dalat 85 1n 5 0 55 35 10 0 9n 40 10 n 90 0 20 n Lahunbuti 53 30 15 2 67 8 24 8 11 72 13 4 49 11 44 0 S,btotal 68 15 16 1 92 22 23 4 15 69 13 4 17 8 S5 n Upland ./Trees Pat-raja 74 8 1J Q 68 11 19 1 29 69 6 0 50 13 38 n Tepong Tandung 65 15 20 0 85 10 5 0 9 95 0 0 40 II 50 0 Bat ,lirln 65 0 35 0 31 13 56 0 0 90 5 5 39 15 90 0 S.btotal 71 8 22 0 66 11 22 2 25 69 6 n 46 13 42 n Repeitta III Total 67 14 17 2 91 23 22 5 15 69 13 3 40 0 SI 0 Total 67 16 15 2 53 22 21 5 17 68 13 3 40 11 40 0 - 38 - Analysis of Household Consumption 2.39 In addition to income data, the transmigration income survey included one section on household consumption based on previous Susenas ques- tionnaires. This section covered household expenditures for food for the past week and for nonfood items for the past month. The initial data tabulation is set out in Table 2.15 and shows a serious data problem. Monthly expenditures exceed total family income by about 6% for upland and 25% for tidal areas, by 25% in Java and 5% in the outer islands, and are not consistent with the ana- lysis of expenditures presented by BPS (see Table 2.1). These discrepancies may be due to errors in the pricing of some products. If the error introduced in this analysis is systematic across all samples, however, these preliminary findings are extremely significant. 2.40 Briefly, the data suggest that transmigrants on a per capita basis consume more rice, corn and cassava than do families in either the inner or outer islands. The ranges are well within plausible limits. Dryland migrants consume more corn and cassava than tidal migrants which is consistent with production patterns. (Soils in tidal areas are generally too acidic for corn and too wet for cassava.) Migrants also consume significantly more fish and slightly more meat and eggs than nonmigrants, but this may be inflated by the inclusion of salt fish in the subsistence package. Migrants, particularly in the uplands, are also found to consume more vegetables and fruits than do other respondents, which is plausible since all transmigrants can grow vege- tables and fruits in their garden lots. It is noteworthy that the per capita cost attributed to food is about equal for transmigrants and Javanese and about three quarters of that for residents in the outer islands. Since this does not reflect an increase in quantity, it may well be a measure of the higher cost of living in nontransmigrant areas in the outer islands. 2.41 On nonfood expenditures the situation is reversed. Javanese spend nearly twice as much on fuel and 2.5 times as much for housing as transmi- grants. They also spend more for household items, durable goods, taxes, medical expenses and schooling than do transmigrants, but migrants spend more on transport. The fact that medical and schooling expenses are lower in transmigration communities may reflect government subsidies through the trans- migration program. That migrants in tidal areas spend less on housing, cloth- ing, household items (such as furniture and kitchen utensils), but almost the same on food, may be attributed to their lower overall incomes and their need to meet their subsistence requirements before other expenditures can be made. Residents in receiving areas spend more than migrants or Javanese on housing, clothing, household goods and on relative luxury goods such as tobacco, alcohol and cosmetics, reflecting both the higher cost of living and higher disposable incomes in the outer islands. 2.42 These data indicate that migrants, while poor, are on average meeting their subsistence needs as well as or better than either rural Javanese or rural residents of the outer islands. And, because they have lower expenses for housing and fuel, they may even be better off than their Javanese counterparts. - 39 - Table 2.15: PER CAPITA CONSUMPTION IN TRANSMIGRANT AND NONTRANSMIGRANT HOUSEHOLDS Transmigration Non-transmigration Item Upland Tidal Outer Java areas areas Islands Weekly Food Consumption Starch (kg/capita) Rice 2.35 2.55 2.20 2.21 Fresh corn 0.30 0.08 0.07 0.07 Dry corn 0.06 0.03 0.01 0.01 Ground corn 0.22 0.14 0.05 0.30 Cassava 1.10 0.60 0.54 0.32 Ground cassava 0.54 0.40 0.07 0.08 Sweet potato 0.11 0.10 0.07 0.15 Other starches 0.34 0.41 0.27 0.03 Protein (kg/capita) Fish 1.54 1.34 0.80 0.54 Meat 0.13 0.04 0.06 0.03 Eggs 0.74 0.47 0.58 0.56 Vegetables and Fruits (kg/capita) Vegetables 1.95 1.70 0.37 0.40 Beans 1.32 0.80 0.06 0.12 Fruits 1.48 1.05 0.45 0.26 Monthly Nonfood expenditures (Rp/capita) Energy/fuel 696 738 1,002 1,464 Housing 355 239 1,328 908 Cloth & clothing 1,208 931 2,995 1,483 Transport 429 322 257 231 Medical expenses 189 120 377 399 Schooling 198 141 225 244 Household items 786 389 1,011 685 Durable goods 311 186 400 626 Taxes and insurance 27 40 105 123 Optional Expenditures Ceremonial expenses 572 350 419 299 Tobacco and alcohol 841 853 1,248 716 Recreation 62 55 16 24 Cosmetics 143 104 337 190 Total Expenditures (Rp/capita) Food (week) 2,220 2,036 2,875 2,110 Nonfood (month) 5,295 3,859 8,742 9,696 Total Month 14,818 12,593 21,075 18,748 Total Year 66,681 56,668 94,837 84,366 (sample size) (1,550) (640) (2,755) (6,490) Source: BPS Transmigration Income Survey, 1985. - 40 - E. Summary and Conclusions 2.43 Although there are problems in interpreting the BPS income survey, several points are clear. First, migrants appear to have incomes which more closely resemble those of rural popuLations in Java than the outer islands. This is due in part to higher incomes in the outer islands from tree crop production, as all categories of respondents have relatively low incomes from food crop agriculture and considerable off-farm work. Second, transmigrants are neither as well off as advocates of transmigration would wish nor as poor as detractors claim. About 20% of migrant families fall at or below subsis- tence levels and about 6% are classified as severely deprived, but 50% have incomes above the poverty level (Rp 10,000/capita/month), and about 13% have incomes over Rp 20,000/capita/month, twice the poverty line. These data are thought to be conservative as the BPS survey appears to understate the value of food produced for home consumption. 2.44 Perhaps the most important finding of the survey is that incomes and the number below the poverty level vary by farm model and period of settle- ment. About one third of farmers in Repelita II upland sites and 38% of farm- ers in Repelita III tidal sites have incomes below minimum subsistence levels, while the proportion in other types of sites is 10-13%. If farmers on older upland settlements have low incomes because of declining soil fertility cou- pled with declining off-farm work, and if Repelita III settlements follow the same pattern, this would be a matter for serious concern. Government argues that improved site selection and farming systems will enable recent settlers in food crop sites to sustain or improve yields and incomes, but this remains to be confirmed and must be closely monitored. Further analysis is provided in Chapter VI. The income data also point to the urgent need for attention to the problems of recently settled tidal sites. 2.45 A second significant finding is that a very large proportion of transmigrant incomes is from nonagricultural work. This figure varies from 30% to 80% in different studies cited in this report but it constitutes an important part of transmigrant incomes in all cases. This contradicts the picture of transmigrants as full-time farmers relying primarily on their own production and it leads to several important questions about future develop- ment in transmigration sites. 2.46 MOT and BPS surveys show increasing agricultural production in the first five years on upland sites. Older Repelita II sites have lower agricultural incomes in BPS surveys, but it is too early to tell whether Repelita III sites, which are larger and better selected, will face the same declines. Even in older areas, we do not find large-scale abandonment of sites and farmers have persisted for many years on subsistence agriculture and wage work. 2.47 Finally, the data on migrant expenditures suggest that transmigrants consume as much food or more than their counterparts in Java and in the outer islands, and by implication, they are equally well nourished even if their total incomes are lower than Javanese averages. They also have assets such as houses, cattle and land which make their situations more secure. For these reasons two thirds of migrants in the BPS survey say their lives have improved. - 41 - III. ECONOMIC AND FINANCIAL ANALYSIS OF TRANSMIGRATION FARM MODELS 3.1 Economic analysis of transmigration cannot capture the unquantified costs and benefits to the transmigration program, however, it is useful as one element in evaluating the overall program and as a method in choosing between alternative transmigration investments. This is particularly important in the context of current resource constraints. Also, as the Bank had previously calculated expected economic rates of return to transmigration investments, it is useful to compare what is now known with prior expectations. 3.2 Representative transmigration farm models are evaluated in this chapter to assess the financial and economic returns to the Repelita III transmigration program and to compare returns to future settlement under different approaches. The emphasis in the first section is on the food crop model, because this model characterizes almost all transmigrant sites developed during Repelita II and III. In the second section, tree crop models are examined as full development options (tree crops are planted when the settlers arrive) and as second-stage development (as further investment in existing sites). Assumptions in the models are fully described, and sensitivity analysis of the major assumptions is included. This was judged necessary because of the uncertainty surrounding potential agricultural production and off-farm work in transmigration sites. Section D presents the summary and conclusions of this analysis. A. Assumptions Used in the Economic and Financial Analysis Quantified Costs 3.3 Settlement. In order to derive representative costs, the most recently approved budget estimates were used for settlement on upland and tidal sites in South Sumatra.- Average costs for South Sumatra upland sites are estimated at Rp 5,840,000 or US$5,300 (mid-1986 prices), and for tidal sites are estimated at Rp 8,227,000 or US$7,154 due to the additional costs for drainage (see Table 3.1). Unless otherwise stated, all costs are presented in mid-1986 prices and an exchange rate of Rp 1,100 = US$1 is used throughout. 1/ These budget estimates (1984/85) were revised slightly for the analysis. Project administration was included at 15% instead of 10%. Site selection, believed to have increased in real terms since 1984/85, was taken from the MOT 1986/87 budget requests. All figures are shown in mid-1986 rupiah. Drainage costs for tidal sites were provided by the Ministry of Public Works. These figures exclude the costs of schools and health centers which would also be provided in the area of origin. - 42 - Table 3.1 DEVELOPMENT COST PER FAMILY FOR FOOD CROP MODELS IN UPLAND AND TIDAL AREAS (1986 prices) Upland Food Crop Model Tidal Food Crop Model (Rp '000) % of total (Rp '000) % of total Site selection 513 9% 513 6% Site preparation Land clearing 733 13% 706 9% Road construction 805 14% 740 9% Housing 635 11% 762 9% Public facilities 248 4% 239 3% Drainage N.A. 0% 1,728 21% Subtotal 2,421 41% 4,175 51% Selection, recruitment, and subsistence Selection 33 1% 33 0% Relocation 1,119 19% 1,119 14% Subsistence supplies 639 11% 959 12% Other services: health, education, MOT services 50 1% 50 1% Subtotal 1,840 31% 2,161 27% Agricultural development Food crops - input package 277 5% 277 3% Tree crops, livestock, industry and cooperatives 30 1% 28 0% Subtotal 306 6% 305 3% Subtotal All Activities 5,080 87% 7,154 87% Project administration (15%) 762 13% 1,073 13% Total All Activities 5,841 100% 8,227 100% Total in US$ Relatively Accessible Site US$5,310 US$7,154 More Remote (25% more) US$6,638 N.A. /a Very Remote (50% more) US$7,966 N.A. /a Tidal sites, given their water access, are assumed to vary less in cost. - 43 - 3.4 South Sumatra costs are representative of the most accessible sites. More remote sites, such as those in Kalimantan and Irian Jaya, cost significantly more to develop. For the purposes of the financial and economic analysis, upland sites are broken down into three categories: relatively accessible (e.g., areas of Sumatra), more remote (East, Central and West Kalimantan, Central Sulawesi, and the Moluccas), and very remote (Irian Jaya). Even within provinces, however, the accessibility of sites varies considerably. Future settlement will necessarily be based on increasingly remote sites, and for analytical purposes, more remote sites are assumed to cost 25% more than the relatively accessible sites, and very remote sites are assumed to cost 50% more. This applies only to upland models; tidal models, given their water access, are more uniform in cost. 3.5 Settlement with local people on upland sites is assumed to cost about US$1,200 less than the basic model (the costs of upland settlement, less selection and relocation costs). In the economic analysis, schools and health centers are omitted from project costs, because they would be provided in Java, and the remaining investment costs are shadow-priced using the general conversion factor of 0.8 for Indonesia. Some transmigration infrastructure, such as roads and schools, benefits local people and spontaneous migrants as well as transmigrants. Since these benefits cannot be quantified, investment costs which benefii/local people and spontaneous migrants have been excluded from the analysis.- 3.6 Labor. The opportunity cost to transmigrant labor is the shadow- priced income foregone in Java, which is estimated at Rp 220,000 per year in mid-1986 rupiah.3 In the farm modRls, the cost of the household's labor is included at this fixed amount, and if labor requirements exceed household labor, the additional labor is costed at the daily wage rate on the outer islands (Rp 1,200/day for 1986), shadow priced. The financial cost of hired labor is the daily wage rate. The models cover a 30-year period, and labor costs are estimated to increase (in real terms) at the same rate as projected 2/ This analysis assumes that 20% of people within the site are local people, traders and other types of migrants (see Table 2.13). Therefore, about 20% of the cost of site selection, land clearing for village facilities and the village facilities; and 50% of the cost of access roads (which also benefit people outside the settlement) are deducted. This is about 10% of total investment costs. 3/ Survey data show that the majority of transmigrants were landless or nearly landless in Java, with very low incomes. Their subsistence income in Java is estimated at Rp 270,000 per year (300 man-days at a wage of Rp 900/day) and this is shadow priced with a conversion factor of 0.82, to give an opportunity cost of labor at Rp 220,000 per year. The methodology used is outlined in Ghanem and Walton, "Indonesia: The Use of Shadow Prices," World Bank working paper, August 1, 1984. - 44 - per capita increases in gross domestic income.4- When second-stage develop- ment is considered, the relevant opportunity cost of labor is the shadow wage rate in the outer islands. Given that labor is a major input into the farm model, the sensitivity of the economic analysis to assumptions regarding the economic cost of labor is tested in para. 3.35. 3.7 The transmigrant family is typically young at arrival, and total family labor supply for both agricultural and off-farm activities is assumed to be about 350 days/year, the first year on site. The family labor supply is assumed to grow at 4% p.a., as the children age and other relatives join the household, until it stabilizes at about 520 days/year, 11 years after arrival. 3.8 Other Inputs. Planting material is assumed to have the same econo- mic and financial costs. Urea, TSP (triple superphosphate), and pesticides have divergent economic and financial prices, and these are presented in Annex 3, Table 1. The opportunity costs of the soil and forest resources made available to transmigrants have not been quantified. Quantified Benefits 3.9 Agricultural Benefits. Benefits in the food crop models are from a combination of rice, corn, cassava, small livestock (chickens, goats, etc.) and fruits and vegetables. The value of houselot production and small live- stock used mainly for home consumption is valued at about Rp 110,000 in year five (para. 2.06). Peanuts are grown in all upland models and benefits from the sale of cattle offspring are included in some models. Economic and financial prices 'verge for rice, corn, and peanuts, and are detailed in Annex 3, Table 1.- Import parity prices are used for corn and peanuts. Through 1983 Indonesia was a rice importer, but it has been highly successful in expanding rice production over the last decade and in 1985 did not require imports. For the next several years, Indonesia will probably be a marginal importer/exporter. The economic price for rice from 1985 onwards is therefore measured as an average of the import and export parity price. Prior to 1985, it is taken as the import parity price. The world price for rice has fallen significantly in real terms since the latter half of the 1970s. This, in combination with the fact that Indonesia is now a marginal importer/exporter, means that the economic benefits from rice production in the transmigration food crop model are significantly lower than those projected at the beginning 4/ Annual growth in per capita gross domestic income is projected as follows: 1982-86 0%, 1987 1.3%, 1988 1.1%, 1989 2.8%, 1990 2.5%, 1991-95 3.5%, 1996-98 3.7%. All parameters from 1999 to the end of the 30-year period are constant. 5/ Commodity prices are based upon January 1986 Bank price projections. The impact on the analysis of the August 1986 Bank price projections, which became available shortly before this document was finalized, is evaluated in para. 3.34. - 45 - of Repelita III.6/ The impact on the economic analysis from alternative economic values for rice is discussed in para. 3.33. 3.10 Off-farm Work. Farm models used in appraisal reports for transmi- gration projects have thus far only analyzed agricultural income. It is evident from recent income surveys, however, that a substantial proportion of total household income is derived from off-farm work. Many farmers seem to follow the strategy of producing food crops to meet their own subsistence needs, while pursuing whatever off-farm employment opportunities exist, since these usually offer a higher return to labor than cultivation of additional food crops that must be marketed. 3.11 As a benefit to transmigration, the portion of off-farm employment that is not part of transmigration development (such as the construction of roads and houses for new sites) and is permaIent is included in the farm model as an additional benefit. It is assumed that no additional costs are associa- ted with the creation of that employment. Income from transmigration site development is not included in the analysis because the final benefit to this work (agricultural production) is already incorporated. The models that incorporate off-farm employment assume that households meet their agricultural labor requirements; and then, if excess labor exists, household members work off-farm to a maximum of about 190 days/year (i.e., the model assumes there is only a finite amount of off-farm work available). The estimate used for the maximum amount of off-farm employment is derived from recent income survey, and is included at 20% less than the average observed from these surveys.- The amount earned from 190 days of off-farm labor is Rp 220,000/year in 1986 (approximately the same as income foregone in Java), and grows in real terms thereafter in line with projected growth in per capita gross domestic income. In the three most common farm models, low-input tidal with severe pest problems, low-input tidal with good pest control, and low-input upland, the household is presumed to engage in the maximum amount of off-farm employment after the first few years on site. Because of the variation in employment opportunities between sites and the possible temporary nature of some types of employment, caution should be used in assessing the viability of models which rely heavily on off-farm employment to raise incomes above subsistence levels. 6/ The price for 5% broken white milled Thai rice, FOB Bangkok, averaged US$438/t during the period 1977-81. During 1982-86 the price declined in real terms, and is projected at US$205/t in 1986. The price is projected to increase in the long run, to US$362/t in 1990 and US$343 in 1995, still about 20% lower in real terms than 1977-81 prices (figures all in 1986 dollars). 7/ It is assumed here that 80% of off-farm employment in the income survey is permanent and 20% is of a temporary nature. Therefore, only 80% of observed levels of earnings are included in the farm models. If more off-farm work were temporary, that is, generated by the settlement process itself, the rate of return which includes off-farm work would be lower. - 46 - Transmigrants pit-sawing timber from their own site for cash income. Skilled transmigrants also make use of opportunities to make furniture. - 47 - 3.12 In the analysis below, three rates of return are presented: the returns to agricultural activities alone, the returns incorporating off-farm employment, and the returns when spontaneous and local people are included in the analysis. This permits comparison with earlier forms of analysis. The last rate of return represents the Bank's best judgement on the economic rate -of return to the model analyzed. B. Economic and Financial Analysis of Food Crop Farm Models General 3.13 Most farm models that follow are based in early years on survey data from existing upland and tidal areas. Detailed budgets for all of the farm models are presented in Annex 3, Tables 3 through 11. For reference, Figure 3.1 shows the breakdown of average income from agriculture and off-farm work for Repelita II and III upland and tidal sites from the BPS Transmigra- tion Income Survey of 2,200 transmigrant households carried out in February 1985. This information was supplemented by a 1985 socioeconomic survey of 1,200 farmers which was carried out in connection with the IFAD-assisted Smallholder Cattle Development Project (IFAD Loan 35-ID). In addition, 14 small-scale surveys of upland sites and seven small-scale surveys of tidal sites were reviewed for information on production, areas planted, and off-farm income. (These surveys are listed in footnote 1, Chapter II.) These surveys were generally consistent with BPS findings. 3.14 Several upland models are discussed in the following analysis. Of these, the low-input/low-output upland model is the most common and represents expected levels of income and production if no further investment were made in transmigration sites. The upland model with cow is typical of models in which some additional investment has been made, and yields and incomes under this model are found in some older settlements. The upland diversified model is typical of the small number of farmers with good market access who raise cash crops (chilies, vegetables and fruits) and who purchase part of their subsis- tence food from the proceeds of these crops. The upland model with sustained input reflects the yields and incomes of the best transmigrants in the best sites and it indicates the potential of the food crop farm model in areas which are well suited to food crop production. Two models are also considered which are based on yields observed with pilot farmers under research condi- tions in which input use (fertilizer, pesticides and labor) is not constrained by cost or risk factors and soils are sufficiently good to benefit from high fertilizer levels. Rates of return to these models are not presented in the tables, but are discussed in the text. 3.15 Although considerable diversity exists across sites in terms of agricultural production, in the most commonly observed models average input use and yields are below those projected in early Bank Staff Appraisal Reports. While yields are generally sufficient to meet the household's subsistence needs (Chapter II), for upland food crop sites, yields of rice are on average less than half of that planned; yields of both cassava and corn are below forecast levels; and peanuts are planted on less than 20% of the proj- ected area. Overall, production of rice is about half that forecast; produc- tion of secondary food crops is about 15% to 25% of forecast levels (Annex 3, - 49 - FIGURE 3 1: SOURCES OF ANNUAL INCOME A1 BY FARM MDDEL AND REPELITA (constant 1985 Rp '000) 900 - On-Farm * 600 - . Agricultural Income /2 50 .-.o 400 30 Non- Q 300 - - _ - _ _ 0 - ;- f-; ~~~~~~'griCUltUrAL 200 - ___ Earnings /3 Wage 100 -4 Average, Average, Average, Average, Repelita II Repelita III Repelita II Repelita III Upland Farmers Upland Farmers Tidal Farmers Tidal Farmers N - 341 N - 977 N - 99 N - 501 /1 Excludes transfer payments. 7T Income from food crops, tree crops, livestock, includes vAlue of own consumption. 73 Excludes wage income and transfer payments. /4 Agricultural or non-agricultural wage income. Source: BPS Income Survey of Transmigration Areas, 1985. - 49 - Table 17 presents yield assumptions from Bank-financed transmigration projects for rice). Rice yields on tidal sites are also below the forecast level of 2.6 tons/ha at full production, although areas planted are roughly in line with appraisal estimates. Table 3.2 summarizes the main characteristics of the models under consideration, and Table 3.3 summarizes the agricultural production and the proportion of a young household's energy requirements met by production under the main food crop models. Detailed Features of Observed Upland Farm Models 3.16 - Upland Low-Input (Annex 3, Table 3). This model reflects the as- sumptions that no additional investment is made in existing transmigration sites and that yields and incomes remain at level now observed in many Repelita II and III upland transmigration sites.- The model includes a 0.25 ha houselot on which mixed subsistence crops (mainly fruits and vege- tables) are grown and a 1.0 ha field under food crops. Reserve land is not in production. Rice yields are highest in the initial years, declining after free inputs stop to about 0.75 ton/ha in year 4. This pattern of inter- cropping results in an annual gross household production of about 750 kg unmilled rice, 100 kg corn, 1,800 kg cassava, and 20 kg of peanuts at full production. Even though production levels are low, this is greater than the subsistence needs (in energy terms) of a young family of two adults and three children (see Table 3.3). Under this model, moderate levels of fertilizers are used as long as they are part of the 3-year subsistence package, dropping thereafter to 40 kg each of urea and TSP annually. Total labor required is between 160 to 200 days per year, freeing considerable family labor for off- farm work, where such work is available. 3.17 This farm model has low financial and economic returns to agricul- tural production alone. Annual net household income at full production from agriculture is about Rp 290,000/year (mid-1986 prices). After meeting subsis- tence requirements, however, the household earns about Rp 250,000/year from off-farm work (year 10 on site). With this additional income, annual net household income rises to about Rp 540,000 in year 10 on site. The economic rate of return when benefits from off-farm income are included is positive, but low, at 3% (Table 3.4). This analysis assumes that employment opportuni- ties are limited to about 190 days per household per year. If labor opportu- nities are not limited and grow in relation to the size of the family and if this employment is permanent (i.e. not generated by transmigration develop- ment) the estimated economic rate of return would be about 5% and net family income would be about Rp 725,000 in year 10. 8/ This model includes agricultural incomes and production patterns similar to those of older upland sites surveyed in the 1985 BPS Transmigration Income Survey. - 50 - Table 3.2: MAIN CHARACTERISTICS OF OBSERVED FOOD CROP MODELS /a Model Main Characteristics Prevalence Upland Low-Input 1.0 ha under food crop cultiva- Most Repelita II and tion. Household grows rice, corn many Repelita III upland and cassava to meet own consumption sites, without addi- needs. Labor available for off-farm tional investment. employment. Low-Input Similar to low-input model with the Increasingly prevalent with cow introduction of a cow. Cattle in Sumatra and Kaliman- provide manure for fertilizer, draft tan. Representative of power to speed up land preparation sites with additional and cash income from the sale of investment in agricul- offspring. ture. Diversified Farmers intensively intercrop 0.5 ha Observed on a limited food crops and 0.5 ha cash crops basis in transmigration including mixed fruit trees, sites with good market bananas, coffee and cloves. Small access. amounts of labor available for off- farm employment. Sustained Farmers intensively cultivate rice, Observed among the best Input corn, cassava, and peanuts on 1.0 ha farmers in sites with with a marketed surplus. Recom- good market access. mended levels of fertilizer are used and some family labor is available for off-farm work. Tidal Low-Input Food crop areas have severe pest Observed in tidal sites with pest problems. Insufficient rice, corn settled during Repelita problems and cassava grown to meet own con- III and believed due to sumption needs. Household is de- design problems which pendent upon off-farm employment. can be addressed. Low-Input Low-input cultivation of basic food Frequent, especially in with pest crops. Subsistence needs met. sites settled in control Labor available for off-farm employ- Repelita II. ment. Improved Higher-input food crop model. Rarely observed, but Sizeable surplus to market after often recommended on meeting subsistence needs. Less technical grounds. labor available for off-farm work. /a All models assume an 0.25 ha houselot on which vegetables and fruits are grown for home consumption. - 51 - Table 3.3: AGRICULTURAL PRODUCTION OF MAIN FOOD CROPS BY YEAR ON SITE AS AT=PORTION OF TOTAL HOUSEHOLD ENERGY REQUIREMENT BY FARM MODEL /a Production (kg) Years from arrival: 2 10 Upland Models Low Input R'Ti-ce 540 750 Corn 90 100 Cassava 1,500 1,800 Peanuts 15 20 Milled rice equiv., calories (kg) 910 1,151 Proportion of total energy needs 1.07 1.35 Low Input with Cow Ricce 600 860 Corn 105 140 Cassava 1,500 1,500 Peanuts 15 24 Milled rice equiv., calories (kg) 964 1,240 Proportion of total energy needs 1.13 1.46 Diversified (Cash Crops Also Raised) /b Rice 400 400 Corn 70 70 Cassava 1,200 1,200 Peanuts 60 90 Milled rice equiv., calories (kg) 744 768 Proportion of total energy needs 0.87 0.90 Sustained Input Rice 500 1,300 Corn 300 1,000 Cassava 1,800 2,600 Peanuts 500 800 Milled rice equiv., calories (kg) 1,581 3,291 Proportion of total energy needs 1.86 3.87 Tidal Models Severe Pests Rice 540 450 Corn 75 80 Cassava 900 1,200 Milled rice equiv., calories (kg) 700 738 Proportion of total energy needs 0.82 0.87 Low Input Rice 630 2,000 Corn 140 160 Cassava 1,200 1,200 Milled rice equiv., calories (kg) 915 1,825 Proportion of total energy needs 1.08 2.15 Improved Ri=ce 1,225 3,840 Corn 200 450 Cassava 1,000 1,500 Milled rice equiv., calories (kg) 1,301 3,405 Proportion of total energy needs 1.53 4.00 /a Transmigrant families are assumed to be made up of two adults and three young children. Their energy requirements (calories/day) are calculated by taking daily nutrition requirements for their age group and comparing total requirements with total calories per year provided by agricultural production of rice, corn, cassava and peanuts. Energy needed by a trans- migrant family/day (calories): male aged 20-39, 2,530 calories, lactating woman aged 20-39, 2,350 calories, child aged 4-6, 1,450 calories, child aged 1-3, 1,160 calories, child aged 6-12 months, 900 calories per day. Total calories per family per day equals 8,390. Total calories per family per year equals 3,062,350, equivalent to 851 kg rice, or milled rice equivalent. Energy in terms of milled Conversion: Energy cal/kg) rice equivalent Unmilled rice 2,34U 0.650 Corn 3,610 1.003 Cassava 1,095 0.304 Peanuts (unshelled) 2,938 0.816 Milled rice 3,600 1.000 /b Above list does not take account of cash crop production. - 52 - Table 3.4: ECONOMIC RATES OF RETURN UNDER DIFFERENT ECONOMIC ASSUMPTIONS AND HOUSEHOLD INCOME AT FULL DEVELOPMENT /a Annual net Economic rates of return household income Agricul- Spontaneous year 10 on site /b tural Agriculture and locals Agricul- Agriculture production and off- taken into ture and off-' alone farm work account /c alone farm work (1) (2) (3) ------ (Rp'000) ------ Upland Food Crop Models Low input /d 2% 3% 291 537 Low input with cow 7 5% 6% 506 751 Diversified 2% 5% 6% 656 783 Sustained input 4% 8% 9% 506 982 Tidal Food Crop Models Severe pests /d /d /d 228 474 Low input 7 3% 4% 529 775 Improved 3% 4% 5% 870 930 /a Most accessible model (costs for Sumatra). 7% Excludes transfer payments such as pensions not generated by the model. 7i See para. 3.05. Off-farm work also included. 7-7 Number is negative. 3.18 Upland Low-Input with Cattle (Annex 3, Table 6). This analysis is based on the low-input model assuming 1.25 ha in production, with the addition of a cow. Government programs in Sumatra, Sulawesi and Kalimantan are provid- ing cows to existing transmigrants with the agreement that they repay the cow by returning two offspring within a given time period. Benefits from the introduction of a cow into the farming system are that: (a) cows are used to plow, thereby reducing labor spent on land preparation; (b) manure is used to fertilize the houselot and/or field crop; and (c) cash sales from offspring supplement farm income. Yields are increased slightly over the low-input model and are 1,200 kg/ha of rice, 700 kg/ha corn, 6,000 kg/ha cassava, and 300 kg/ha peanuts at full production. Inorganic fertilizer is used on the field crops as long as free inputs are supplied (three years) and thereafter a small amount is applied to the houselot garden. Labor savings from the cow (in terms of reduced time for land preparation) of about 30 days/year are offset by 20 additional man-days/year for animal care, including forage production. 3.19 Agricultural income (Rp 510,000) in this model is about 75% higher than in the low-input model, and rises to about Rp 750,000 with off-farm work at full development. This model assumes that the farmer will pursue off-farm work instead of opening up reserve land because of higher returns. In spite - 53 - of improvements in agricultural production, rates of return to agriculture alone in this model remain low (-1% to 0%). Once off-farm income is included, the economic rate of return is about 6%. The fact that economic rates of return remain low under conditions which are relatively attractive to the farmer is a reflection of the high initial cost involved in resettlement. If a cow, or other agricultural investment, is provided to existing transmigrants as second-stage development and all transmigration investment costs are considered sunk, the incremental rate of return would be about 25%. 3.20 Upland Diversified (Annex 3, Table 4). This upland model is based on farming syst ems observed on the fields of progressive farmers with good market access.9 This model has production on a 0.25 ha houselot and an adja- cent 1.0 ha field. In this model, farmers intensively intercrop 0.5 ha with food crops for subsistence, and peanuts are planted on 0.15 ha following the rice harvest. Under this regime, farmers produce 400 kg of rice, 70 kg of corn, 1,200 kg of cassava and 90 kg of peanuts in addition to vegetables and root crops grown in the houselot. On the remaining 0.5 ha, the family plants mixed fruit trees, mainly citrus, but also banana, coffee and cloves, main- taining reasonable levels of fertilizer and pesticides in this area (200 kg each of urea and TSP and 6 liters pesticide). 3.21 This model assumes that crops can be marketed and will produce a net annual income of about US$550, bringing household income above the poverty level. The model has an economic rate of return of about 6% to agriculture and off-farm work. In order to achieve a 10% rate of return, sales of vege- table$ _and fruits would have to increase from about US$550 to US$930 per year.l°l'This model requires about 400 man-days/year of labor which is roughly equivalent to available family labor for the first several years of settlement. By year 10 on site, the family labor supply has increased and approximately Rp 120,000 is added to household income from off-farm work, raising household income to about Rp 780,000/year (US$710), above the poverty level. 3.22 Upland Sustained Input (Annex 3, Table 5). Yields and production under this model are observed among the better farmers in transmigration sites with reasonable market access and regular input supply (extension, fertilizer and credit). Rates of return generated by this model would, therefore, be possible if interventions could be made which would help all farmers, on 9/ Adapted from Bevan, "Review of Rainfed Production on Upland Soils," Working Paper, FAO/World Bank Cooperative Programme Investment Centre, 1985. 10/ This high level of sales of fruits and vegetables (at least US$930 per household per year) does occur occasionally, for example, in transmigra- tion sites near Samarinda in East Kalimantan, where farmers market chilies and citrus in the provincial capital. - 54 - average, to achieve these yields. Farmers under this model cultivate the standard 1.0 ha field lot and 0.25 ha houselot; however, inputs are higher than in the low-input model, averaging 200 kg urea, 100 kg TSP, and seven liters of pesticide. With this input use, farmers produce about one ton of rice in year 5 (1.3 ton/ha, but intercropped) and 1.3 tons in year 10 (1.7 ton/ha intercropped). Households also obtain 900 kg of corn, 2,400 kg of cassava and 800 kg of peanuts in year 5, increasing slightly thereafter. This is more than sufficient to meet the families' subsistence needs, and the surplus is marketed. With one hectare under production, labor is available for off-farm work. 3.23 The rates of return to this model are still modest at 4% for agri- culture alone, but for agriculture and off-farm work rise to 9% (7% and 6% in more remote sites). Household incomes are higher, averaging about Rp 670,000 in year 5 and Rp 980,000 in year 10. These incomes place farmers well above the poverty level. However, these yields and incomes are not expected on a large-scale because of lack of market access, and technical and institutional limitations described in Chapter VI. Table 3.5: ECONOMIC RATES OF RETURN TO AGRICULTURE AND OFF FARM WORK FOR UPLAND MODELS IF INFRASTRUCTURE COSTS INCREASE DUE TO THE REMOTENESS OF THE SITES /a Farm model Base case /b More remote Very remote Upland low input 3% 1% 0% Upland with cattle 6% 4% 3% Upland diversified 6% 4% 3% Upland sustained input 9% 7% 6% /a More remote sites cost 25% more to develop than the most accessible areas. Very remote sites (Irian Jaya) are assumed to cost 50% more to develop than the most accessible sites. In Repelita IV and V an increasing proportion of sites will be more remote. /b Takes local people and spontaneous migrants into account. Research-Based Upland Models 3.24 High-Input, Labor-Intensive (Annex 3, Table 7). This model is not observed except in research stations, but it is widely described in the lite- rature and is included for comparative purposes. The model is based on the heavy use of fertilizer and other inputs as indicated by the results of research work, and requires high standards of management and good access to markets. The model analyzed adapts research recommendations made by the - 55 - Central Research Institute for Agriculture (CRIA).14/ In this model, rice, corn, and cassava are intensively intercropped on 1 ha followed by peanuts and cowpeas on 1 ha. The household has a large surplus of these crops after meet- ing its own consumption needs and markets the rest. High-input levels are maintained, averaging about 270 kg each of urea and TSP per year, and high levels of pesticides are used, averaging 35 liters annually at full produc- tion. Labor requirements (600 man-days/year) exceed family labor supply, requiring hired labor and precluding off-farm work. Under these circumstances yields are projected at 1,500 kg/ha of rice, 1,200 kg/ha corn and 10,000 kg/ha cassava intercropped. Peanuts yield 600 kg/ha. 3.25 The net financial return from agriculture to the farmer under this model would be about Rp 540,000/year at full development, because of the high demand for labor and inputs. However, since no labor is available for off- farm work, household incomes in this model are no higher than in the low-input model when off-farm work is included. If the risks (climatic, pests) to the farmer's capital are taken into account, this helps explain why farmers opt for the low-input model. This model has the lowest economic rate of return of all models which include off-farm work and rice yields would have to increase from 1,500 kg/ha to 5,500 kg/ha in order to raise the rate of return to 10%. 3.26 High-Input with Cattle (Annex 3, Table 8). In order to explore the rate of return to food crops under optimal conditions, one model was developed that assumes the cultivation of 2.0 ha of food crop land (using draft power) and rice yields averaging 2.0 tons/ha in year 5 and 2.5 tons in year 12. By year 10 the household would produce a total of 3.8 tons of rice, 3.3 tons of corn, 6.4 tons of cassava and 2.0 tons of peanuts and have an average annual income of about US$1,500. To obtain these yields farmers would have to use 250 kg/ha urea, 200 kg/ha TSP, 100 kg/ha KCI and also apply lime, assumed to be available from a nearby quarry. Family labor is fully used in cultivation and gathering animal feed and no off-farm work is assumed. 3.27 The rates of return to this model are high in relation to other upland models at 11% for the base case (9% for more remote and 8% for very remote areas). This model is, however, not observed in farmers' fields at the present time for reasons discussed in detail in Chapter VI. Should experience on existing farmers' fields demonstrate that household production could be maintained at these levels, the conclusions about economic rates of return generated by past experience would have to be reassessed. Until then, this model is regarded mainly as an indication of what would have to be accom- plished to make the food crop model economically attractive in areas with good market access. 11/ Adapted from Bevan, "Review of Rainfed Production on Upland Soils," Working Paper, FAO/World Bank Cooperative Programme Investment Centre, 1985. - 56 - Tidal Farm Models 3.28 All tidal models assume that the farmer is given 2.0 ha, a 0.25 ha houselot and a 1.75 field divided either into one or two plots. This is typi- cal of the majority of Repelita II and III sites. (Bank-assisted projects provide 2.25 ha.) Inputs other than pesticides are not used in the basic tidal model as soil fertility is assumed to be sufficient to produce reason- able rice yields in the initial years.1- 3.29 Tidal Low-Input With Pest Problems (Annex 3, Table 9). This model represents a number of tidal sites settled during Repelita III that have severe pest problems and low resulting yields; it indicates production and incomes if no further investments are made and if yields remain at current levels. This is not expected to be the case. Most observers believe that yield and pest problems are due in part to the nucleated village layout and the presence of uncleared land within the settlement area in Repelita III tidal sitTe3 and that measures can be introduced to control these problems.- With the linear layout used in other models, the field is surrounded by other cleared areas, pests have fewer places to hide, and farmers can scare away larger predators. Some settlements in nucleated sites with severe pest problems report average yields of 500 kg rice/ha/year or less. In this model, rice yields are higher, at 750 kg/ha, but only 0.6 ha of rice is planted. Corn and cassava are grown on 0.3 ha each, intercropped, and yields are low averaging 400 kg/ha of corn and 6,000 kg/ha of cassava. Fruit, vegetables, taro and the like are grown on the houselot. 3.30 Food crop production under these conditions would be below household subsistence requirements, and rates of return are negative. (Remote sites are not included as costs for tidal settlement are virtually the same in Sumatra and Kalimantan and have not been estimated for Irian Jaya.) With this crop- ping pattern, only 87% of the total energy requirements of the household would be met from home production or rice, corn and cassava (Table 3.3). Household income from agriculture alone, about Rp 230,000 at full development is 25% lower than the low-input upland model. Labor requirements for on-farm are only 160 days/year, so additional income can be earned from off-farm work as 12/ At the end of Repelita III, Government began to distribute a small fertilizer package to tidal sites consisting of 100 kg urea and 50 kg TSP. A majority of sites, however, were developed without fertilizer input. 13/ In nucleated sites, houses are grouped and the 1.0 ha field plot is separated from the houselot. Repelita III tidal sites were developed using a nucleated layout instead of a linear layout for several reasons. The nucleated design has the advantages of more concentrated land clearing blocks, greater economy of road networks, greater proximity to social services, and economy for future utility service. The main advantage of the linear design is proximity of the farmer to his farmlands. The farmer can open up more land and cultivate it more intensively with less destruction by predators. - 57 - long as such income opportunities exist. In the model, household income rises to about Rp 475,000/year when off-farm work is included, but at this income level the household is still below the poverty level. 3.31 Tidal Low-Input with Reasonable Pest Control (Annex 3, Table 10). This model indicates the potential of well-designed tidal settlement without further investment. It is observed in many older tidal sites and some sites settled in Repelita III. In many Repelita II sites farmers were given up to 1.75 ha in one plot adjacent or near to the houselot. With this farm layout, settlers have a strong incentive to clear and cultivate their entire 1.75 ha in order to control pests. Under optimal conditions, about 1.6 ha are under rice production by year 7 and corn and cassava are raised on 0.2 ha each. Yields, based on intercropping, are estimated at 1,250 kg/ha for bunded rainfed rice, 800 kg/ha corn, and 6,000 kg/ha cassava, which still reflect moderate problems of pest control. Rice1production is estimated at two tons rice/household which provides a surplus.- Corn production is estimated at 160 kg and cassava at 1,200 kg, sufficient for home consumption. Peanuts are not grown in any of the tidal models. The rate of return to agriculture alone is still negative and with off-farm work is 4%, reflecting high investment costs in relation to incremental production. (Rice yields would have to rise from 1.25 t/ha to an average of 4.2 t/ha in order to increase the rate of return to agriculture to 10%.). However, financial returns to the family are about Rp 530,000 annually from agriculture at full production and labor requirements, estimated at 300 days per year, permit limited off-farm work. When off-farm work is included in the model, family income rises to Rp 775,000 per year, well above the poverty level. 3.32 Tidal with Improved Food Crops (Annex 3, Table 11). The improved farm model is seldom observed in practice in tidal areas. It assumes 2.0 ha in production or 1.0 ha double cropped with HYV rice and assumes that farmers maintain fertilizer use at rates of 160 kg urea and 80 kg TSP/ha. It also assumes relatively high applications of pesticide (7 liters/ha) throughout the production period. Rice is planted on 1.6 ha and yields rise to 2,400 kg/ha by year 7. Corn and cassava are planted on 0.3 and 0.1 ha respectively with yields of 1,500 kg/ha and 15,000 kg/ha. Rates of return to agriculture alone in this model, at 3%, are significantly improved over those on the low-input model. Labor inputs are significantly higher than in the previous models, an extra 150 man-days per year at full development, reducing opportunities for off-farm work, but financial returns to the farmer from agricultural produc- tion are considerably improved, at Rp 870,000/year at full development (about Rp 350,000 higher than the previous model). If off-farm earnings are included, average incomes in both models are similar at full development (at about Rp 930,000 per year), and rates of return are about 5%. However, even with higher incomes many farmers are expected to choose the low-input model with pest control because it entails less risk. 14/ This model is similar to production reported by Repelita II tidal farmers in the BPS 1985 Income Survey, who were selling the most rice of any group in the Survey (an average of 0.8 tons of rice are reported to be sold annually by Repelita II tidal farmers; the average for the sample overall was 0.2 tons of rice marketed annually). - 58 - The Impact of Alternative Assumptions on the Economic Analysis 3.33 Alternative Food Crop Prices. The economic rice price used in the models is an average of import and export parity price from 1985 onwards, instead of the import parity price for rice. It could easily be argued that because of the significance of subsistence consumption of food items, economic and financial farmgate prices understate the value to the household of food produced and consumed by the household. If the food were not produced on the farm, the household would need to get supplies from another source or buy food at retail prices. In order to capture this benefit in the models, economic home consumption prices were assessed on the basis of the cost of supplying food to consumers in the sending areas (i.e., the without project case). In pricing rice and corn this way, the home consumption value is about 30% higher than if measured at farmgate prices. When the models were rerun with these higher prices for home consumption quantities, however, this was found to have vitually no impact on the rate of return. Comparison was also made using the farmgate price derived from import parity for rice instead of an average of import and export parity. This also had little impact on the rate of return. 3.34 As this report was being finalized, the August 1986 Bank commodity price projections became available. For rice, corn and peanuts, these more recent projections are lower than the January 1986 projections used for the analysis. The economic farmgate prices based on August projections are about 25% lower for rice, 5% lower for corn, and about 10% lower for peanuts over the period of analysis for the farm budgets. The economic farmgate price for urea is about 8% lower, and the price for TSP is roughly the same. The rates of return fall by 1-2% across all food crop models when these latest price projections are used. 3.35 Alternative Assumptions for the Economic Cost of Labor. One of the goals of transmigration is to move poor, generally landless underemployed households from Java to the outer islands where they can become more produc- tive. The opportunity cost of labor in the economic model presented here is the shadow priced subsistence income of poor rural households in Java (see para. 3.06), estimated at Rp 220,000 per year for 1986. Since labor is a major input to the farm models, the sensitivity of results of the analysis to the assumption of the opportunity cost of labor was tested for the most commonly observed model, the low-input food crop model. If labor is not shadow priced, but simply equated to subsistence income foregone in Java, the rate of return remains negative for agriculture alone. If the shadow-priced subsistence income is halved, then the rate of return remains negative for agriculture benefits, and rises from 2% to 5% for agriculture and off-farm benefits. At the most extreme position, with the opportunity cost of labor in Java valued at zero, the rate of return from agriculture would rise from nega- tive to 3%, and the rate of return from agriculture and off-farm work would be 8%. The models presented assume that labor costs as well as off-farm income increase in real terms with projected growth in per capita gross domestic income, but the results of the analysis are not sensitive to this assumption, i.e., if no growth in labor costs and off-farm income is assumed, the results would be virtually the same. - 59 - Economic Return on the Repelita III Settlement Program 3.36 About 366,000 households were settled during Repelita III by the MOT. Only about 2% of the total were settled with the tree crop model (prin- cipally rubber). Based on the previous economic analysis by farm model, the rate of return to the Repelita III transmigration program based on food crop settlement on tidal and upland areas (98% of the families settled) was between 3X and 6% (depending on the frequency of the most representative models) if benefits from off-farm work are included. Nonquantifiable benefits such as regional development are not included in this rate of return. C. Economic and Financial Analysis of Tree Crop Models Tree Crop Models 3.37 Tree crops are better suited agronomically to much of the low fertility soils of Sumatra and Kalimantan than food crops. The establishment of tree crops has been promoted by the Government through externally financed Nucleus Estate and Smallholder (NES) Projects, and domestically financed projects of the same type (PIR). They are also developed by projects in which smallholder tree crops are established by Project Management Units (PMUs) under the Directorate General of Estates. Examples of PMU projects include the Bank-assisted Smallholder Rubber Development Project (SRDP, Cr 984-IND) and Smallholder Coconut Development Project (SCDP, Ln 1898-IND). The salient features of these projects are summarized in Table 3.6. Table 3.6: MAIN CHARACTERISTICS AND PREVALENCE OF TREE CROP MODELS Model Main Characteristics Prevalence NES and PIR Government-owned estates (PTPs) open About 6,100 transmigrant type projects new land, block plant tree crops, households were settled and 80X of the area planted is later with rubber and about distributed in 2 ha blocks to local 1,300 with oil palm families and transmigrants. The during Repelita III. cost of tree crop establishment and settler housing is covered by credit. SRDP and SCDP Externally assisted projects in Not currently used for type projects which Project Management Units transmigrants. (PMUs) help farmers plant or replant uneconomic rubber or coconut small- holdings on their own land. 3.38 Financial and economic analyses of these schemes were described in the World Bank report "Indonesia--The Major Tree Crops: A Sector Review" - 60 - (April 1985). The analysis that follows differs from that in the earlier report in three ways. First, commodity prices have been uyated in accordance with the Bank's January 1986 commodity price projections.- Second, the models have been adapted to include costs for relocation and settlement. Third, the assumptions in the analysis regarding the opportunity cost of labor and benefits from off-farm work have been changed from those used in the Sector Review in order to make the analysis as comparable as possible with the food crop models presented in the previous section of this chapter. Table 3.7 presents the development costs of the various tree crop models. Adaptation of Costs of Models for Settlement with Transmigrants 3.39 PMU-based Models. The PMU approach has not been used to settle transmigrants thus far, so that the model included here is hypothetical. This model is presented only for rubber, but the PMU approach could also be used for coconuts. Settlers are provided with 2 ha of tree crops and a 0.5 ha houselot/garden area. Since this model has not been used to settle transmi- grants, two versions of possible settlement costs are presented (Table 3.7). The first uses MOT settlement costs similar to the upland food crop model, except the land clearing and food crop input package costs are halved because the food crop area is reduced. This results in overall financial development costs per household that are somewhat higher than NES schemes. Theoretically costs could be limited to NES levels, and the second cost version is similar to NES schemes. The first version settles households with 2 ha of rubber for about US$9,200, the second for about US$7,900. 3.40 One PMU model is also presented for second-stage development of existing transmigrants' reserve land (1 ha) in which 0.5 ha of the already cleared food crop area is planted with tree crops along with the 1 ha reserve land. This model was analyzed because the majority of transmigrants settled on upland sites during Repelita III do not have access to more than 2 ha of land. As a minimum of 1.5 ha is required for a viable tree crop holding, this leaves a 0.5 ha houselot/garden area. In financial terms, this second-stage development is estimated to cost US$3,046 per household. Land clearing is only necessary for the 1 ha of reserve land. Fertilizer, materials and equip- ment, labor, and overheads are reduced to reflect 1.5 ha of tree crops. 3.41 NES Schemes. For NES schemes, which include some social infra- structure (e.g., housing) in the basic tree crop investment, MOT's additional investment costs for selecting and moving the transmigrant household and pro- viding some services is estimated at US$1,360 per household (1986 prices). Costs and Benefits of Tree Crop Models in the Economic Analysis 3.42 Costs. Materials and equipment, agricultural machinery, housing, roads and buildings in the economic analysis are shadow priced with the stan- dard conversion factor of 0.8 for Indonesia. Economic prices of fertilizer, 15/ The impact of the recently available August 1986 commodity price projections is presented in para. 3.46. - 61 - Table 3.7: FINANCIAT. DEVELOPMENT COSTS FOR TREE CROP MODELS PER HOtJSEHOLD /a (Rp'000, 1986 prices) PMIJ planted rubber, NES (2 ha tree crops) full development, 2 ha tree crops Oil Coco- Cost Alternative Cost Alternative Rubber palm nut One /b Two /b Time period of initial development (6 yrs) (4 yrs) (6 yrs) (6 yrs) (6 yrs) Tree Crop Development Costs Labor 1,860 1,675 1,929 1,89f) 1,84() Fertilizer/agrochemicals 1,046 1,176 849 1,144 1 ,144 Aaterials/equipment 827 660 937 729 729 Agricultural machinery 154 133 133 71 71 Overheads /c 560 527 557 564 564 Infrastructure /d Housing 1,162 1,162 1,162 - 1,162 Roads 942 942 848 251 942 Buildings 157 126 157 63 157 Other Cash payment to settler (NES only) 73 73 73 - 73 Otiher 452 452 534 193 452 Subtotal, Tree Cron Dev. (Rp -OOO) 7,235 6,926 7,179 4,905 7,184 (US$) 6,577 6,297 6,527 4,459 6,531 Additional Settlement Costs for Transmigration NES Projects with Trans,igrants Ancillarv services 5( 50 50 50 Settlement and transport 1,252 1,252 1,252 1,252 Project administrationi (15%) 195 195 195 195 Subtotal, Additional Settlement Costs for NES Transmigranits (Rp-Ono) 1,497 1,497 1,497 1,497 (U155) 1,361 1,361 1,361 1,361 Ministry of Transmigration Providing Settlement Infrastructure Total settlement costs for food crop model, with land clearing and food crop inpuit package halved (to reflect reduction in food crop area from I ha to 0.5 ha) (Rp '000) 5,262 (US$) 4,784 Total, Tree Crop and Tranasmigration Settlement Costs (RpV000) 8,763 8,214 8,659 10,167 8,681 (1JS$) 7,966 7,467 7,872 9,243 7,892 /a NES and PMtJ schemes assumed to pay settlers a daily wage rate of Rp 1,750 per day. Labor requirements during the establishment period are a total of 1,063 days for NES rubber, 957 days for NES oil palm, 1,102 days for NFS coconut, and 1,080 days for PH13 rubber. /b Cost alternative one assumes Ministry of Tranismigration provides settler infrastructure at costs similar to those for upland food cron settlement, except land clearing and food crop inpuit packages are halved. Cost alternative two assumes settlement infrastructure is provided at levels similar to NES rubber schemes. 6oth models are hypothetical because transmigrants, to date, have not been settled with PMU-provided tree crops. /c 15Y1 of the direct financial costs of field crop establishment (labor, fertilizer/agr)chemicals) and materials and equipment. Ld Fuill cost of infrastructure. Does not take costs or benefits to spontatneous migrants and local people into account. - 62 - as in the food crop models, are some 75% higher than domestic subsidized prices. For full development of tree crops with transmigrants, the opportu- nity cost of household labor is the subsistence income foregone in Java, shadow priced as in the food crop model (para. 3.06). For second-stage devel- opment (tree crops planted for existing settlers), the opportunity cost of labor is the income foregone from other activities when labor is shifted from off-farm work to tree crops. 3.43 Benefits. Economic and financial price projections for rubber, oil palm, and coconuts are presented in Annex 3, Table 2. Production and input assumptions presented in Annex 3, Tables 12 to 16, are taken from the Tree Crop Sector Review. Because of better management, NES rubber schemes have higher yields than most PMU schemes. PMU-planted rubber for existing transmi- grants (on 1.5 ha) is assumed to have yields 5% lower than PMU-planted rubber for new transmigrants, simply because the farm layout is not as orderly. All models include net benefits from a 0.5 ha houselot/garden area. These net benefits are derived from the upland low input food crop model. In all models, household labor supply is projected along with demand for agricultural activities. If excess labor is available, rates of return are also calculated for models including off-farm work, after labor requirements for tree crops and the garden are met using the same assumptions about off-farm work as in the food crop models (para. 3.12). Detailed assumptions about off-farm work under each model are presented in Annex 3, Tables 12 to 16. Since there are fewer spontaneous migrants in tree crops sites than in upland sites, neither costs nor benefits attributed to spontaneous migrants and locals are taken into account. 3.44 Rates of Return to NES Schemes. NES rubber and oil palm schemes are economically attractive as full transmigration development schemes, with rates of return of 13% each (Table 3.8 and Annex 3, Tables 12-14). NES coconut schemes do not look as attractive. If tree crop development costs increase by 25%, rates of return fall from 13% to 11% for NES rubber schemes, from 9% to 7% for NES coconut schemes, and from 13% to 11% for NES oil palm schemes. With benefits from off-farm work, rates of return are slightly higher, 15% for NES rubber and oil palm and 11% for coconut. Table 3.8: ECONOMIC RATES OF RETURN TO NES TREE CROP MODELS: NEW SETTLEMENT WITH 2 HECTARES OF TREE CROPS PER HOUSEHOLD (1986 constant Rp '000) Rate of Rate of return return on on tree crops Farm model tree crops NPV (10%) and off-farm work NPV (10%) NES Rubber 13% 2,418 15% 4,086 NES Oil Palm 13% 2,067 15% 3,585 NES Coconut (Hybrid) 9% -875 11% 856 Source: Annex 3, Tables 12, 13 and 14. - 63 - Table 3.9: ECONOMIC RATES OF RETURN TO PMU-BASED TREE CROP MODELS: NEW SETTLEMENT AND SECOND-STAGE DEVELOPMENT WITH TREE CROP AND FOOD CROP BENEFITS (1986 constant Rp'000) Tree Crops Tree Crops and Off-Farm Work Economic rate Economic rate of return NPV (10%) of return NPV (10%) Full Development 2 ha rubber Cost alternative 1 /a 9% -671 11% 419 Cost alternative 2 7a 11% 773 13% 2,275 Second-Stage Development 1.5 ha rubber 16% 2,214 15%/b 1,795/b /a Cost alternative 1 assumes the Ministry of Transmigration provides settler infrastructure at costs similar to those for upland food crop settlement, except land clearing for food crops and the input package are halved. Cost alternative 2 assumes settlement infrastructure is provided at levels similar to NES rubber schemes. /b If off-farm work is available and transmigrants are given tree crops as second-stage development, benefits lost from shifting their labor into tree crops slightly reduces the rate of return. Source: Annex 3, Tables 15 and 16. 3.45 Rates of Return to PMU-based Schemes. As second-stage development, the PMU-based rubber model has a rate of return of 16% (Table 3.9 and Annex 3, Table 16). At full development, the return is 9% under Cost Alternative One, and 11% under Cost Alternative Two (Annex 3, Table 15). 3.46 Impact of Changing Commodity Prices. Long-term price projections for tropical tree crop commodities are low. Prices for tropical tree crops declined in 1985 and are projected to continue to decline in real terms. However, Indonesia has an advantage in labor availability and domestic demand (for coconut and oil palm). Because of the sensitivity of results to price projections, rates of return were recalculated with prices 25% higher and 25% lower than the World Bank's January 1986 price projections for tree crop development. Results are tabulated in Table 3.10. NES rubber and oil palm are marginally economically attractive as full development schemes with a 25% fall in projected prices. The PMU second-stage development model continues to have a return above 10% even if prices fall by 25% in real terms. As this report was being finalized, the August 1986 Bank commodity price projections became available. Palm oil and rubber economic farmgate prices are about 15% lower in these more recent projections than the January 1986 price projections for the time period that the economic analysis covers. Copra prices are about - 64 - 30% lower than the January 1986 prices. Rates of return using these prices are presented in Table 3.10. Table 3.10: ECONOMIC RATES OF RETURN OF TREE CROP FARM MODELS WITH ALTERNATIVE TREE CROP COMMODITY PRICES /a Commodity Commodity August 1986 prices prices commodity price Base case 25% higher 25% lower projections (%) (%) (%) (%) New Settlement (2 ha tree crops, 0.5 ha garden/houselot) NES coconuts 9 13 5 5 NES rubber 13 16 10 12 NES oil palm 13 16 9 10 PMU planted rubber Cost alternative 1 9 11 6 8 Cost alternative 2 11 14 7 10 Second-stage Development (1.5 ha rubber, 0.5 ha garden) PMU planted rubber 16 20 12 14 /a Excluding impact of off-farm work. Based on January 1986 Bank commodity price projections unless otherwise indicated. D. Summary and Conclusions 3.47 Most transmigration settlements to date have been developed with food crops. Seven observed food crop models and two research-based models have been described in this chapter, covering upland and tidal sites. In all seven observed food crop models, economic rates of return were estimated based upon agriculture alone and upon agriculture and off-farm income. In all models, rates of return from agriculture were low and varied from negative for the most commonly occurring low-input models, to positive (4%), for upland sites with good market access and sustained use of inputs. Most transmigrants seek off-farm work to supplement agricultural production, and off-farm work makes an important contribution to settler incomes. When benefits from off- farm income are included in the economic analysis, rates of return to agriculture and off-farm income average 3-6% for the most commonly observed low input/low output upland and tidal models. If the production achieved by the best farmers with good market access were earned by all farmers on average the rates of return to agriculture and off-farm income would average 9%. This is not expected for reasons outlined in Chapter VI. 3.48 Many research models for food crop production in transmigration sites rely on a high-input strategy to keep yields above a subsistence level over time. However, in practice most settlers adopt a low-input/low-output food crop model, since input supply is irregular, credit is scarce, markets - 65 - are inaccessible, and crop failures may occur every few years. The economic analysis also indicates that labor-intensive, high-input agricultural produc- tion is not financially attractive when compared to off-farm work. 3.49 Economic rates of return estimated here for models based mainly on food crops are lower than those projected in World Bank appraisals of transmi- gration projects. This is mainly because average realized crop yields have been lower than projected, and rice prices have decreased significantly since 1981 (a drop of almost 50% in real terms from the 1976-81 average to the 1985- 86 average). 3.50 An economic analysis was also undertaken for several tree crop farm models. Tree crops formed only a small part of the transmigration program through Repelita III, but were increasingly important in Repelita IV. The Government's financial cost to settle a transmigrant household with tree crops varies from US$7,470 to US$9,240 (mid-1986 prices) depending upon the commodity and institution involved. This is 40% to 80% higher than the upland food crop model. Tree crops can also be added as second stage development on already existing sites, at a cost of about US$3,050 per household for PMU- planted rubber on 1.5 ha. With this additional investment, tree crop models have higher economic rates of return and higher settler incomes than the food crop models. 3.51 From an economic point of view, NES rubber and NES oil palm schemes have the highest rate of return of the new settlement options at 13% each. As second-stage development to already existing sites, PMU-planted rubber has a higher rate of return of 16%. 3.52 The net incomes of tree crop smaliholders at full development are projected to be about twice as high as those of food crop transmigrants, even when off-farm income is included in food crop settlers' income. Smallholders with tree crops also have the potential to supplement income with some off- farm work. From the point of view of settler welfare and economic returns, a strong argument can be made that transmigration development be based on: (a) second-stage development of existing sites with tree crops, livestock, fish ponds or other forms of development; or (b) new settlement on NES- or PMU- based schemes. This leads to higher settler incomes and higher economic rates of return than the food crop model, but necessarily entails a much reduced program compared to the scale of Repelita III, because of limitations in institutional capacity in the estate sector. The rate of return to second stage development is the highest, and remains economically attractive even with commodity price declines of 25% in real terms. The implications of this analysis are reflected in the changes in program direction adopted by MOT. 3.53 The economic models discussed in this chapter do not include other complex costs and benefits that are difficult to quantify, but are nonetheless essential to consider. Some of these costs include environmental degradation in the receiving province, and possible depressed wage rates in the outer islands. Benefits include broader regional development from roads, temporary and permanent employment generation for the local population as well as trans- migrants, and alleviation of population pressure (and its environmental impact) in the sending areas. These issues are discussed further in Chapters IV and V. - 67 - IV. THE IMPACT OF TRANSMIGRATION A. Employment Generation 4.1 Without migration, the labor force of Java and Bali would increase by about one million people per year over the next few years. Therefore, finding adequate employment for new job seekers is among the most important challenges facing the GOI. In the past decade, the Government has generated employment by expanding areas under production and intensifying production on cultivated land, by supporting large-scale, labor-intensive rural works programs and by encouraging resettlement. With resource constraints, the growth of temporary employment through rural works programs is expected to slow while the need for employment creation will be increasingly critical. For this reason, it is important to understand the amount, type and cost of employment generated by transmigration during Repelita III. The following section provides a methodology for assessing the amount of employment resulting from transmigration in Repelita III, and it evaluates the role of the program in employment generation. Definitions 4.2 Transmigration settlement generates both temporary and permanent employment. Temporary employment is produced in the development phase of a transmigration site, but is not sustained over time. Most of this employment is from land clearing and the construction of roads and houses. Permanent employment continues after the development phase and includes sustainable jobs in agriculture and services in the transmigration sites. Indirect permanent jobs are also generated in transportation, marketing, social services, and the supply of inputs to transmigrants. Local residents benefit from the transmi- gration program through indirect employment, particularly in trade and transport; and spontaneous migrants who move to an area for the opportunities created by transmigration investment also benefit from indirect employment. Assumptions Used in Calculating Temporary Employment 4.3 The development phase of transmigration creates a large demand for labor to clear land, construct houses, buildings and roads, build drainage systems, and plant tree crops. Labor inputs for land clearing, which is one of the major employment activities, are shown in Table 4.1 and described below. (a) Tidal Sites. In tidal sites, land clearing is generally done with chain saws and manual labor, since the soil cannot support heavy machinery. Construction of the main drainage works is capital- intensive, while the construction of tertiary canals by the transmi- grants is labor-intensive. In South Sumatra, where many transmigra- tion tidal sites are located, contractors often mobilize labor from West Java for land clearing and construction. - 68 - Wage laborers involved in clean up following mechanical land clearing. Table 4.1: LABOR INPUTS FOR LAND CLEARING BY FARM MODEL Farm model Man-days/ha ha/household Man-years/household /a Tidal sites 80 1.5 0.42 Upland food crops Mechanical 30 to 35 /b 1.5 0.16 to 0.18 Semi-mechanical 60 to 70 7r 1.5 0.32 to 0.37 Tree crops 65 /c 2.0 0.45 /a Assuming one man-year equals 286 man-days. 7r Varies depending upon the type of vegetation prior to clearing (secondary forest, grassland, etc.). /c Average of labor required for secondary forest, bush/grassland and grassland. - 69 - (b) Upland Food Crop Sites. Semi-mechanical methods (using chain saws and hand labor) were generally used to clear upland transmigration sites until the beginning of Repelita III when the program shifted mainly to the use of mechanical methods. However, both agronomic and employment considerations argue strongly for semi-mechanical land clearing. Aside from preservation of the topsoil, the major advantage to semi-mechanical methods is employment generation. While mechanical methods use 30 to 35 man-days/ha, semi-mechanical methods use about twice this much, or about 250 more man-years per site (assuming 1,500 households per site). (c) Tree Crops. Labor demand from the development phase of tree crop schemes for transmigrants is the heaviest of all the farm models discussed. Not only must land be cleared for both food crops and tree crops, but roads and houses must be constructed, a cover crop planted, and tree crops established and maintained. Land clearing is generally done using semi-mechanized methods, because the area does not need to be completely cleared and semi-mechanical clearing is cheaper. Assuming that the nonquantified labor inputs for roads, houses and buildings are roughly comparable across farm models, tree crop investment generates the most labor during the development period. Tidal sites are next, with semi- mechanized land clearing in upland food crop sites close behind. Mechanical land clearing on upland sites has considerably less employment impact. Assumptions Used in Calculating Permanent Employment 4.4 Permanent Agricultural Employment. Permanent employment consists of both direct and indirect employment generated by transmigration. Labor requirements for food crop agriculture are estimated in Chapter III (see Annex 3, Tables 3 to 11). In the most commonly occurring Repelita III development model, the low-input, upland model, farm labor requirements sta- bilize at 160 days/year, the amount necessary to meet subsistence needs. Thereafter, the household is assumed to pursue more remunerative off-farm employment. Assuming full employment is 286 days of work per person per year, then 0.6 agricultural jobs are generated from settling one household under the upland model. Tidal sites follow roughly the same pattern, although, on aver- age, settlers spend most of their time on agriculture on these sites. As Table 4.2 indicates, of the three tree crop commodities, rubber requires the largest labor input (0.94 man-years per household per year), followed by oil palm (0.64 man-years) and hybrid coconut (0.58 man-years). Thus oil palm and coconut employ the settler at about the same level as food crop models, roughly 0.6 man-years/year. - 70 - Table 4.2: TREE CROP LABOR REQUIREMENTS FOR LAND CLEARING, DEVELOPMENT, AND MAINTENANCE AND HARVESTING /a Commodity Rubber Oil palm Hybrid coconuts Development phase Ic Development period 6 man-years 4 man-years 6 man-years Land clearing /b 65 man-days/ha 65 man-days/ha 65 man-days/ha Tree crop establishment 479 man-days/ha 426 man-days/ha 447 man-days/ha Total, man-years per household /c 3.8 man-years 3.4 man-years 3.6 man-years Maintenance and 0.94 man-years 0.64 man-years 0.58 man-years harvesting /c per household per household per household /a Excludes settlement infrastructure. 7b Average for secondary forest, bush/grassland and grassland. 7i Two hectares per household, 286 man-days per man-year. Source: "Indonesia - The Major Tree Crops: A Sector Review," April 15, 1985. 4.5 Permanent Off-Farm Employment. It is more difficult to calculate how much off-farm work is permanent and not related to the settlement process. As in rural Java, productive activities in transmigration sites are highly diversified. In rural areas of Indonesia, in general, nonagricultural economic activities made up at least 33% of employment in 1980- Among the most important sources of employment generation in rural areas in Java are timber milling, plywood industries, rice mills, brick and tile makin cottage industries, food processing, weaving and the production of clothing.- Survey data indicate that these types of activities are also important sources of employment for transmigrants. Some of this employment is related to felling, cutting, sawing and selling timber from uncleared lands; and if a site is part of an ongoing transmigration development, settlers may also do construction work on transmigration houses, roads and other buildings. Small-scale surveys also report many other activities, with most generated within the settlement itself: carpentry, forging, brick and roof tile production, food processing 1/ BPS, 1980 Census. This may underrecord nonagricultural employment as many of those recorded as employed in agriculture also engaged in non- agricultural activities. 2/ Gavin W. Jones, "Links between Urbanization and Sectoral Shifts in Employment in Java," Bulletin of Indonesian Economic Studies, Vol. XX, No. 3, December 1984. - 71 - (tempe and tahu), wickerwork, saw mills, retail shops, market activities and small businesses. Although one would expect the service sector to expand as migrant welfare increases or with the length of time on site, the 1985 BPS Transmigration Income Survey does not show a correlation between agricultural success and the amount of off-farm work, nor does there appear to be a significant relationship between wages and length of time on site. This may be due to the interaction of a number of complex factors which hide such associations. 4.6 Figure 3.1 in Chapter III shows the average earnings overall and by farm model from the three sources: wages, nonagricultural activities, and on- farm agriculture, and the implied days worked for wage income. These data are summarized in Table 4.3. The data are shown only for tidal and upland food crop sites; tree crops were omitted because of their smaller sample sizes. For the sample overall, settlers work about 130 days per year (0.46 man-years) for wage income, assuming they earn Rp 1,200 per day. This ranged from 46 days of wage labor per year for Repelita III tidal settlers, to 163 days of wage labor per year for Repelita II upland settlers. One might hypothesize that wage opportunities would decline over time if highly dependent on ongoing transmigration development; however, Repelita II and Repelita III upland sites do not differ significantly in the amount of wage income. Repelita II tidal sites have higher wage incomes than Repelita III tidal sites, but this may be because a high proportion (30% of respondents) were located in remote Central Kalimantan (see also Chapter II). The Amount and Cost of Employment Generated by Transmigration 4.7 If household members work, on average, 130 days per year (0.46 man- years) for wage income and if they also earn almost as much from other more informal nonagricultural activities, off-farm work would provide households with between 0.7 to 1 man-year of employment per household per year. Adding this to the earlier rough estimate of 0.6 man-years of work in agriculture, transmigration is assumed to generate about 1.3 to 1.6 jobs per household from both on-farm and off-farm activities. This would mean that the transmigration program generated an estimated 500,000 to 600,000 full-employment, permanent jobs during Repelita III. This figure excludes employment generated indirectly by transmigration for local residents and spontaneous migrants. Data on temporary employment suggest that about 18 million days or 63,000 years of full-time employment were generated by land clearing alone. 4.8 At a cost of US$5,310 31 per household for the upland food crop model, transmigration would generate employment at about US$3,300 to US$4,100 per permanent job. For tidal sites, with settlement costs of US$7,150 per household, this would be about US$4,500 to US$5,500 per job created. Less information is available about the off-farm work of tree crop settlers, but assuming they have roughly the same off-farm employment, then it would cost about US$3,600 to US$4,200 to create a job in tree crops. (These estimates 3/ 1986 US dollars, base case. More remote site development is estimated to cost 25% to 50% more. - 72 - are so rough that no significant differences in job creation costs between models can be inferred.) This analysis does not take into account permanent work found by spontaneous migrants who move at a lower cost. The cost of transmigration-related job creation is considerably lower than the cost of job creation in the industrial sector (estimated at US$10,000-US$20,000/job), but more than the cost of job creation in service industries. Table 4.3: HOUSEHOLD ANNUAL WAGES, ON-FARM AND NONAGRICULTURAL EARNINGS, AND DAYS WORKED OFF-FARM BY FOOD CROP FARM MODEL (Rp'000 1985) Nonagri- Sample cultural On-farm Total size Wages/a earnings Total income income /b (1) - (2) (1)+(2) (3) (1)+(2)+{T) ----------------------- Rp'000 ----------------------- Income Tidal Sites Rep. II 99 190 200 390 460 849 Rep. III 501 55 124 179 228 407 Average 600 78 136 214 266 480 Upland sites Rep. II 341 196 98 294 225 519 Rep. III 977 178 141 319 319 638 Average 1,318 182 130 312 295 607 Average /c 1,918 156 128 284 295 579 -- Days Worked Off-Farm -- Tmplied Labor Days Id Tidal Sites Rep. II 158 167 325 Rep. III 46 103 149 Average 65 114 178 Upland sites Rep. II 163 82 245 Rep. III 148 117 266 Average 152 108 260 Average, Total 130 107 237 /a All wages from both agricultural and nonagricultural activities. 7b Excludes transfer payments such as money sent from relatives, pensions and government assistance. /c Excludes 60 respondents settled just prior to the survey date. 7T Days of labor implied at a daily wage rate of Rp 1,200. Source: BPS Transmigration Income Survey, 1985. - 73 - 4.9 Interpretation of these figures must be qualified for several reasons: (a) we cannot be absolutely certain how much off-farm employment is dependent upon short-term transmigration program development expenditures (e.g., construction of settlers' houses) and might not therefore be sustained; (b) both the quality and quantity of work are important. Low value- added jobs contribute little to productivity and income levels, while better jobs with better incomes indirectly generate growth and further employment; (c) to calculate net employment generated, some account must be made of the jobs foregone when a family is moved from Java. In spite of these caveats, there can be no doubt that transmigration was a major vehicle for employment generation in Repelita III; and no Government investment program other than tree crop development is believed to have generated more self-sustaining, full-time employment in the third five-year plan. 4.10 Future Prospects. In most receiving provinces, government expenditures associated with transmigration have been greatly reduced in Repelita IV. In many of these provinces, sources of economic growth and labor absorption exogenous to the transmigration program are also sluggish, given the current weak commodity prices for exports from the outer islands. If nonagricultural work also declines because of reductions in transmigration expenditures, migrants will go farther afield searching for work. Under these circumstances the potential exists for increased competition for employment and downward pressure on wages in the receiving provinces. To avoid this scenario it is critical that reasonable levels of government expenditure be maintained in areas where the labor force has been greatly increased from transmigration in Repelita III. The best ways of doing this are to support infrastructure maintenance, to invest in second-stage development in existing transmigration areas, and to use labor-intensive forms of development. B. Demographic Impact The Impact of Sponsored Migration 4.11 The distribution of transmigrants relative to the distribution of people in the outer islands is shown in Annex 4, Table 1. It is noteworthy that over Repelita I and II (1969-79), the major islands received about the same proportion of transmigrants as their share of the total population. Approximately 60% of all transmigrants were sent to Sumatra, slightly more than Sumatra's 55% share of the outer island population, and Kalimantan and Sulawesi received approximately the same share of transmigrants as their' share in the outer island population. This is apparently related more to accessi- bility, carrying capacity, and the availability of infrastructure and services than to conscious design. Since 1980, Kalimantan and Irian Jaya have received a somewhat greater share of transmigrants than before. Kalimantan, which - 74 - accounted for 14% of all transmigrants in the 1970s, increased its share to 22% during the Repelita III and early Repelita IV period, while Irian Jaya increased its share from 1% in the 1970s to 4% of transmigrants in the 1980s. 4.12 Table 4.4 shows the number of sponsored transmigrants in relation to provincial populations as of 1980 and 1985. Although the number of transmi- grants more than doubled during Repelita III in 9 of 18 receiving provinces, the share of sponsored transmigrants in provincial populations was no higher than 12% in any province, and in most cases it was under 8%. The highest proportions of sponsored migrants were found in southern Sumatra, in the sparsely populated province of Central Kalimantan, and the smaller provinces of Central and Southeast Sulawesi. The more densely populated provinces of North Sumatra and South Sulawesi show low proportions of recent migrants, but they were major receiving areas in the prewar period. The eastern islands have not been a prime target for transmigration. By 1985, about 20,000 trans- migrant families had been settled in Irian Jaya, a province of about 1.3 million, and about 200 families had been settled in East Timor. 4.13 At the district (kabupaten) level the effect of transmigration is more apparent. Annex 4, Table 2 shows the number of sponsored transmigrants settling in outer island kabupaten from approximately 1970 to mid-1985, and at this level the impact is striking. Two kabupaten in Lampung and South Sumatra each absorbed more than 200,000 sponsored transmigrants, and together they accounted for one quarter of all the transmigrants moved in Repelita III. In each of seven other kabupaten, more than 45,000 transmigrants arrived within the five-year period. These 9 out of 66 receiving kabupaten absorbed 53% of all transmigrants. Almost three-quarters of all transmigrants were accommo- dated in 20 kabupaten. Sponsored transmigrants equaled slightly more than 40% of the total population in two kabupaten in 1985 (one each in South Sumatra and South Kalimantan) and in seven others they equaled more than 20% of the population. With such high concentrations of migrants, the impact of transmi- gration on people, production and administrative services is significant in these districts. The Impact of Spontaneous Movement 4.14 It is difficult to estimate the rate of spontaneous migration in Indonesia since these migrants did not register in the past, but the 1980 cen- sus contained two types of questions which touched upon the scale of sponta- neous movement: the first was on previous residence, the second on mother tongue. The data on prior residence (Annex 4, Table 3) indicate that in 1980 there were about 3.6 million people born in Java but living in the outer islands, while there were about 1.1 million people living in Java, but born elsewhere. This suggests a lifetime out-migration rate from Java over three times as great as the rate of in-migration. In the five years preceding 1980, about 1.1 million people moved out of Java and about 420,000 moved in, for an out-migration ratio of 2.6 to 1. Of those moving out of Java, 868,000 (80%) moved to rural areas, while the number of transmigrants in this period was less than 250,000. This means that for each family moved by the transmigra- tion program from 1975 to 1979, at Least 2.5 others moved to rural areas in the outer islands in other ways. Causality cannot be inferred directly from these data, but there is a strong association between the provinces to which - 75 - sponsored migrants have been sent in the past and those to which spontaneous migrants later move. This is particularly true of rural migrants who find it difficult to identify and secure land without the help of family and friends who have previously moved (see Chapter VII). Table 4.4: SPONSORED TRANSMICRANTS AS PERCENT OF RECEIVING PROVINCE POPULATIONS Sponsored transmigrants Sponsored as x of transmigrants Number of sponsored provincial as % of popu- Province transmigrants ('000) po2ulation/a lation increase 1971-80/b 1980-85/c 1980 1985/d 1971-80 1980-85 Aceh 9.6 61.0 0 2 2 6 N. Sumatra 1.8 37.1 0 0 0 3 W. Sumatra 34.8 23.2 1 2 6 6 Riau 29.3 177.9 1 8 6 52 Jambi 96.0 107.8 7 12 22 35 Bengkulu 41.7 61.1 5 11 17 34 S. Sumatra 141.3 379.0 3 10 12 48 Lampung 133.3 188.2 3 5 7 13 Sumatra 487.8 1,035.3 2 5 7 21 W. Kalimantan 23.7 131.9 1 6 5 46 C. Kalimantan 9.4 109.8 1 11 4 65 S. Kalimantan 41.0 91.4 2 6 11 41 E. Kalimantan 29.6 55.5 2 5 6 14 Kalimantan 103.7 388.6 2 6 7 37 N. Sulawesi 11.2 18.8 1 1 3 8 C. Sulawesi 51.5 75.5 4 8 14 29 S. Sulawesi 36.5 25.0 1 1 4 5 S.E. Sulawesi 37.9 92.1 4 12 17 62 Sulawesi 137.1 211.4 1 3 7 19 Maluku 4.3 35.1 0 2 1 17 Irian Jaya 10.6 75.6 1 6 4 48 Other/e n.a. 7.9 n.a. n.a. n.a. 2 Total 743.5 1,753.9 1 4 7 25 /a Source for data on provincial population -- BPS, Statistik Indonesia 1984, Table 3.1.2; data on households from 1983 Agricultural Census. /b P. Gardner, Provincial Population Projections (Jakarta: UNCHS/COI NUDS Project, 1985), Table 4. /c Transmigration Department, Daftar Proyek Transmigrasi Yang Dibina Tahun 1985/86 (to August 1985). /d 1980 figures include migrants moved from 1971-80 and 1985 figures include migrants moved from 1971-85. /e Nusa Tenggara Barat (NTB), Nusa Tenggara Timur (NTT) and East Timor. - 76 - 4.15 Data on place of birth do not convey the full impact of transmigra- tion, however, as the children of migrants born in the outer islands cannot be distinguished from the local population in such records. For this reason the 1980 census asked what language was used in the home. These data, summarized in Table 4.5, indicate that there were 7.3 million people in the outer islands in 1980 who spoke an inner island language (Javanese, Sundanese, Madurese or Balinese). Of these, about 5.2 million (70%) lived in Lampung and North Suma- tra, areas receiving transmigrants and plantation labor prior to 1940. Although the data on early movement and return migration are not precise, it is unlikely that more than 2.0 million of the 6.7 million migrants to rural areas are the descendants of transmigrants and plantation laborers. The other 4.7 million people are spontaneous migrants, their descendants, or those moving in the civil service or for other work. This again suggests that total spontaneous movement has occurred at a rate of about 2.4 to 1 over the past 50 years. 4.16 If we eliminate Lampung and North Sumatra from the analysis, to exclude the "pull" influence of those moved in the colonial period, transmi- grants moved to other provinces between 1950 and 1979 plus their descendants would have numbered about one million in the 1980 census, had no one returned to Java. The 1980 census, however, records nearly 2.2 million inner island language speakers in these other provinces, of whom 1.9 million are in rural areas. This suggests a "pulling power" of at least one family for each family moved in the last 30 years. This number is significantly lower than the figures which include Lampung and North Sumatra since (a) Lampung is the most accessible area to Java and has attracted the vast majority of spontaneous migrants; and (b) the majority of sponsored migrants to other areas have moved in the last decade and have had less time to attract others. Table 4.5: SPONT4NEOUS MIGRATION AS MEASUREI) BY LANGUAGE SPOKEN AT IHOME /a Percent of inner Pro jected island language Percent population Toner loner speakers in outer of rural Percent in Individuals in 1980 from island island Ratio of islands due to population rural areas moved as sponsored language language Rural rural sponsored from speaking sponsored migration speakers speakers excess spontaneous: migration /b sponsored inner island migrants daring 1950-78 enumerated enumerated (spontaneous sponsored Rural only Total Total provincial migration language fron from natural in 1980 in 1980 migrants) migrants (M) (X) population, 1980 (2) (%) Province 1950-78 increase census, total census, rural (4)-(2) (5)/(1) (2)/(4) (2)/(3) Rural only Total (2)/(9) (4)/(9) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) Aceb 9,600 10,900 175,300 156,200 145,3300 15.0 7 6 2,377,000 2,610,500 0 7 N. Sunatra 15,600 26,400 1,767,700 1,668,600 1,642,100 104.6 2 1 6,223,500 8,350,900 0 27 Riau 16,500 22,770( 189,500 148,200 125,400 7.6 15 12 1,575,600 3,406,100 1 9 W. Sumatra 35 900 48,200 56,100 52,700 4,400 0.1 92 86 2,973,000 2,163,800 2 2 iambi 70,700 82,700 255,300 236,200 153,400 2.2 35 32 1,261,600 1,444,400 7 19 Bengkulu 30,600 37,20n 134,900 132,200 94,900 3.1 28 28 695,500 767,900 5 19 S. Sumatra 262,400 430,100 635,000 585,700 155,600 0.6 73 68 3,360,700 4,627,700 13 17 Lampung 222,701) 350,400 3,400,89n 3,163,100 2,812,600 12.6 11 10 4,047,300 4,624,200 9 78 Subtotal Sumatra 664,500 1,009,000 6,615,000 6,143,10() 5,134,100 7.7 16 15 22,514,500 27,995,900 4 27 W. Kalimantan 32,6(10 43,200 197,600 153,40(1 110,2(( 3.4 28 22 2,067,900 2,484,900 2 7 C. Kalimantan 14,50n 20,000 62,90n 35,90() 15,90o 1.1 56 32 855,900 954,100 2 4 S. Kalimantan 50,100 70,200 115,700 97,300 27,000 0.5 72 61 1,622,300 2,063,200 4 6 E. Kalimantan 41,200 59,300 126,20( 62,200 2,900 0.1 95 47 729,300 1,214,600 8 9 Subtotal Kalimantan 138,600 192,800 502,50(1 348,900 156,100 1.1 55 38 5,275,500 6,716,900 4 7 N. Sulavesi 17,700 25,500 31,0()0 29,900 4,300 0.2 85 82 1,760,200 2,114,800 1 2 C. Sulawesi 42,200 54,900 53,600 51,100 (3,800) -0.1 107 102 4,963,400 6,059,500 1 1 S. Sulaaesi 55,40n 68,300 71,600 70,90(0 2,500 0.0 96 95 1,169,000 1,284,500 6 6 S.E. Sulnvesi 31,600 39,400 46,00n 45,11)0 5,600 0.2 97 86 853,500 941,400 5 5 Subtotal Sulawesi /c 147,100 188,200 202,200 197,000 8,700 0.1 96 93 8,746,300 10,400,300 2 2 East Nusa Tenggara 2,100 2,400 3,2(( 7(1() (1,600) -0.8 324 74 2,531,500 2,736,900 0 0 Maluku 4,200 7,1(00 16,300 15,900 8,700 2.1 45 44 1,255,500 1,408,400 1 1 Irian Jaya 4,800 6,600 4,400 800 (5,800) -1.2 799 150 869,900 1,107,200 1 0 Total 961,400 1,406,300 7,343,700 6,706,60( 5,300,30)) 5.5 21 19 41,193,400 50,365,900 3 16 Encluding N. Sumatra, Lampung, & Irian Jaya /d 718,100 1,022,700 2,17(0,70() 1,874,10() 351,30) 1.2 55 47 30,052,600 36,283,400 3 6 /a Inner island language speakers include those speaking Ja-nese, Sundanese, Madureso and Balinese. /b Including offspring of sponsored migrants. 7E In Sulawesi, many spontaneous migrants have registered as snonsored. 7i EIcLuding North Sumatra and Lampuog because of large migrant communities settled there prior to World War El. Excluding Irian Jaya because of apparently sizeable census undercounts. Source: 1980 Indonesian census snd MrOT records. - 78 - 4.17 Thus, while the proportion of migrants in Indonesia is small (over 95% of all Indonesians were on the island of their birth in 1980), the mobile people in the population have had a large impact on the provinces to which they move. Overall, in 1980, inner island speakers made up 27% of the rural population of Sumatra, 7% of Kalimantan and 2% of Sulawesi. In rural areas, about 78% of the people in Lampung spoke an inner island language, 27% in North Sumatra, 19% in Jambi and Bengkulu, and 17% in South Sumatra. Demographic Impact of the Program through the Year 2020 4.18 Sponsored transmigration in Repelita III represented an estimated movement of about 2% of the combined population of the provinces of Java and Bali and about 15% of the incremental population increase between 1980-85. However, these figures understate the impact of the transmigration program, since they cover a very short period and do not take into account the cumula- tive impact of moving young couples with the highest fertility levels. They also exclude the effect of spontaneous migrants who follow sponsored settlers. 4.19 In order to assess the longer-range demographic impact of transmi- gration, Bank staff developed two scenarios based on alte47ative assumptions about the level of sponsored migration between 1980-2000 ._ Both scenarios were developed prior to the cessation of land clearing for new settlement in 1986. The high scenario projected Repelita III levels of movement (300,000 families) in each of the third, fourth and fifth five-year plans, declining to about 240,000 families in Repelita VI. This would entail the movement of about 3.3 million people through the year 2000. These figures represent the maximum level that could be achieved under the sponsored program if land availability issues were resolved, implementation capacity in the tree crop subsector were significantly improved (see Chapters VI and VII) and financial resources were available. The intermediate scenario assumes that settlement will fall to 200,000 families in Repelita IV and V and taper off to 100,000 families in Repelita VI, ending thereafter. This would entail the sponsored movement of about 1.9 million people through the year 2000. In both scenarios population projections were run on sponsored settlement alone and with two levels of spontaneous movement: the first would result in 88 spontaneous migrants over 20 years for every 100 moved on the sponsored program; the second would result in 181 spontaneous migrants for each 100 moved. This is thought to represent a reasonable range of spontaneous movement where sponsored settlement levels are high. 4.20 The results of this exercise are summarized in Table 4.6. In the absence of any redistribution, the population of the inner islands is expected to reach intermediate 170 million by the year 2020 compared to 107 million today. Under the intermediate scenario for sponsored migration, and with high but historically observed levels of spontaneous migration, the population of Java would be reduced by 15 million (9%) of what it would otherwise be in the 4/ Assumptions of projected fertility and mortality rates for the inner and outer islands, labor force participation rates, and spontaneous migration are detailed in Annex 8. - 79 - year 2020, and transmigration would absorb 19% of the incremental labor force of Java. With the lower ratio of spontaneous migration, transmigration would still absorb 12% of the incremental labor force and the population of Java would be reduced by 10 million. Table 4.6: DEMOGRAPHIC IMPACT OF TRANSMIGRATION UNDER ALTERNATIVE SCENARIOS ('000 people) Population Difference in 2020 as a Annual with and proportion of population Change in without population growth rate incremental Population migration without 1980-2020 labor force year 2020 year 2020 migration --------- (%) --------- No Migration Inner islands 169,136 - - 1.4 Outer islands 106,753 - - 1.9 High Levels of Spon- sored Settlement /a Inner Islands Sponsored only 161,526 -7,610 0.96 1.3 -8 Low spontaneous 155,361 -13,775 0.92 1.2 -16 High spontaneous 147,594 -21,542 0.87 1.1 -24 Outer Islands Sponsored only 115,006 8,253 1.08 2.1 10 Low spontaneous 121,453 14,700 1.14 2.2 22 High spontaneous 129,455 22,702 1.21 2.4 34 Intermediate Levels of Sponsored Settlement /7b Inner Islands Sponsored only 163,857 -5,279 0.97 1.3 -7 Low spontaneous 159,566 -9,570 0.94 1.3 -12 High spontaneous 154,032 -15,104 0.91 1.2 -19 Outer Islands Sponsored only 112,530 5,777 1.05 2.0 10 Low spontaneous 117,043 10,290 1.10 2.1 18 High spontaneous 122,764 16,011 1.15 2.4 27 /a Assumes sponsored movement of 300,000 families in Repelita IV and V, falling to 240,000 families in Repelita VI. /b Assumes sponsored transmigration levels of 200,000 families in Repelita IV and V, falling to 100,000 in Repelita VI. Note: Low spontaneous means that each mover attracts 0.5 people; high spontaneous, each mover attracts 0.75 people within ten years. Assumptions of projected fertility and mortality rates for the inner and outer islands, labor force participation rates, and spontaneous migration are detailed in Annex 8. - 80 - 4.21 These figures highlight two important points. First, intermediate rates of sponsored and spontaneous migration sustained through the year 2000 (with a reduction in Repelita VI) would have a significant impact on popula- tion growth and labor absorption in Java. A transmigration program at these level would have a flattening effect on the growth rate of Java, reducing it from an average of 1.7% to 1.4% p.a. over the period 1980 to 2000. Second, this level of movement would also have a significant impact on the outer islands, increasing the growth rate from 2.3% to 2.8% per year. By extension, the impact of transmigration on regional development to date is critical to an evaluation of the future program. If change brought about by transmigration can be managed to create employment and benefit the outer island peoples without jeopardizing their institutions and resources, this would be a strong argument for program support. C. Impact on the Regions 4.22 The acceleration of the transmigration program during Repelita III (1979/80-1983/84) and the very large numbers of people moved to the outer islands during this period changed the view that transmigration could only make a marginal contribution to regional development. During Repelita III the estimated movement of between 1.5 to 2 million people not only added substan- tially to local populations in receiving areas, but also brought unprecedented increases in land clearing, road construction, public services, and financial resources. In addition, recent studies showing that many transmigrants are regularly involved in agricultural wage work and nonagricultural employment outside transmigration sites indicate that transmigrants are potentially filling, and perhaps generating, local demands for production and employment in receiving provinces. This section discusses the impact of transmigration on regional resources--people, land, production, infrastructure, and institutions--in order to evaluate the nonquantified costs and benefits of transmigration, and to assess the role of transmigration in regional develop- ment. The impact of transmigration on local people and the environment is described in Chapter V. Transmigration Expenditures in the Receiving Provinces 4.23 Since Repelita III, transmigration has accounted for a large share of the development budgets of many receiving provinces. Table 4.7 shows the distribution of the transmigration development budget across provinces. The share of transmigration in provincial development budgets was large throughout Repelita III. As the table shows, the 1985/86 transmigration budget in 14 of 18 receiving provinces exceeded the total transmigration budget for the respective provinces during 1979 to 1982. While some increase is due to depreciation of the rupiah, the transmigration budget also increased in real terms through FY85/86, and represented an increasing share of the total development budget in most outer island provinces. 4.24 Although budgets are not equivalent to expenditures, the contribution of the transmigration program to the provincial budgets is striking. In 8 of the 17 provinces shown in Table 4.7, the FY85/86 transmi- gration budget provided from one third to almost one half of the total - 81 - provincial development budget.5/ In only four provinces, all with very low numbers of transmigrants, did the proportions fall below 10%. In five outer island provinces (South Sumatra, Bengkulu, Lampung, South Kalimantan and Maluku), transmigration contributed 20% to 30% of the FY85/86 development budget, in four (Riau, Jambi, Central Sulawesi and Southeast Sulawesi) it contributed 30% to 40% and in four other provinces with relatively small populations (West, Central and East Kalimantan and Irian Jaya) it contributed 40% to 50%. 4.25 It is important to recognize that these expenditures are incremental to other provincial expenditures, that is, in addition to what would otherwise be allocated on a per capita or area basis. For this reason, such expendi- tures generate considerable employment, both for local people and transmi- grants, and are undoubtedly a leading source of growth in many provinces. Everywhere, small towns near transmigration sites appear to be growing at very high rates, attracting both local people and spontaneous migrants to off-farm work. In fact, much of the increase in public resource expenditures in the outer islands can be justified only because of the exceptionally rapid popula- tion increases stemming in part from transmigration. At the same time, as noted in the previous section, concern exists that a rapid curtailing of expenditures in these provinces, either because transmigration moves to other provinces or because of continuing resource constraints, could make these provinces vulnerable to increased unemployment. Land Development 4.26 Table 4.8 shows the distribution of land made available for spon- sored transmigrants in the receiving provinces according to official MOT sta- tistics. Consistent with settlement patterns in general, Sumatra accounts for almost 64% of household and farm land allocaoed under the transmigration program, Kalimantan 20%, and Irian Jaya 2%.6 Of more interest is the relationship between transmigrant land and total land in use in the provinces. In three provinces in Sumatra (South Sumatra, Bengkulu and Lampung) and in Southeast Sulawesi, the amount of land allocated to sponsored transmigrants is over 10% of provincial land used for agriculture. In most cases, however, transmigrants make up a larger proportion of farm households than of agricultural landholdings, presumably due to the extensive agricul- tural practices used by local smallholders in the outer islands. These figures do not take into account land obtained for agricultural purposes by spontaneous migrants. 5/ The development budget includes all line agency expenditures financed through the central government, except for routine expenditures and special presidential programs (INPRES). 6/ The discrepancy between land available f 3r sponsored transmigration (8,287 km ) and that allocated (6,901 km ) and in use (4,910 km ) is apparently due to the practice of allocating 1.25 ha and retaining much of the land for future development. - 82 - Table 4.7: TRANSMIGRATION DEVELOPMENT BUDGETS IN RECEIVING PROVINCES, 1979-82 AND 1985/86 /a Distribution Transmigration as % Transmigration among receiving total development Province development budget provinces /d budget in province 1979-82/b 1985/86/c 1979-82 1985/86 1979-82/b 1985/86/c --- (Rp billion) --- ----- (%) ---(%) ------… Aceh 15,193 16,017 3.3 2.6 11.0 15.0 N. Sumatra 6,659 17,460 1.4 2.8 3.0 8.5 W. Sumatra 5,147 9,016 1.1 1.5 3.0 6.6 Riau 62,956 49,397 13.6 8.1 24.0 35.8 Jambi 28,367 30,275 6.1 4.9 29.0 37.5 S. Sumatra 114,486 60,966 24.7 9.9 44.0 29.8 Bengkulu 16,505 14,472 3.6 2.4 20.0 24.1 Lampung 10,316 30,068 2.2 4.9 9.0 25.7 W. Kalimantan 29,098 61,781 6.3 10.1 26.0 44.4 C. Kalimantan 22,765 49,084 4.9 8.0 33.0 48.0 S. Kalimantan 39,030 26,695 8.4 4.4 27.0 24.1 E. Kalimantan 26,297 60,046 5.7 9.8 29.0 42.9 N. Sulawesi 7,265 7,519 1.6 1.2 8.0 8.9 C. Sulawesi 20,622 27,111 4.4 4.4 25.0 33.9 S. Sulawesi 6,678 15,582 1.4 2.5 4.0 9.7 S.E. Sulawesi 24,755 26,280 5.3 4.3 37.0 35.7 Maluku 11,634 23,804 2.5 3.9 17.0 25.4 Irian Jaya 16,668 87,699 3.6 14.3 27.0 45.7 Total 464,441 613,272 100.0 100.0 20.0 16.5 /a Recently high carry-overs in transmigration budgets mean that budget figures are higher than actual expenditures (see Chapter VII). /b Source: UNDP/OPE, Transmigration Program Second Phase Evaluation (Jakarta, 1982), Table 3.9. Figures are totals for three budget years and do not apparently include carry-overs (siap) from previous years. /c Source: Ministry of Finance, Laporan 1985/86 dari SDPD Bandung. Figures include siap from 1982/83 to 1984/85. /d Figures do not include amounts allocated to Jakarta or sending provinces. - 83 - Table 4.8: SPONSORED TRANSMIGRANT LAND BY PROVINCE, 1985 Transmigrant Allocated (2) land (km2)/b as a % of Sponsored Provincial Avail- Allo- Total provincial transmigrants Province agricultural able cated in use agricultural as a % of farm land (km2)/a (1) (2) (3) land households Aceh 10,965 269 173 106 1.6 3.4 N. Sumatra 13,725 148 127 98 0.9 0.7 W. Sumatra 7,713 116 90 90 1.2 1.5 Riau 7,101 801 679 312 9.6 14.1 Jambi 6,626 525 376 330 5.7 11.3 S. Sumatra 10,622 1,740 1,711 1,681 16.1 10.1 Bengkulu 1,527 272 230 168 15.0 15.3 Lampung 7,774 1,018 1,001 563 12.9 7.3 Sumatra 66,053 4,889 4,387 3,348 6.6 6.3 W. Kalimantan 6,831 586 250 109 3.7 7.7 C. Kalimantan 6,479 514 479 265 7.4 17.0 S. Kalimantan 6,223 465 373 253 6.0 7.3 E. Kalimantan 4,841 303 256 197 5.3 13.1 Kalimantan 24,374 1,868 1,358 824 5.6 9.6 N. Sulawesi 3,549 80 66 48 1.9 1.4 C. Sulawesi 63,001 380 300 201 0.5 9.1 S. Sulawesi 15,308 120 115 87 0.7 0.7 S.E. Sulawesi 3,363 387 364 277 10.8 10.8 Sulawesi 85,221 967 845 613 1.0 3.3 Maluku n.a. 175 149 15 n.a. 4.6 Nusa Tenggara Barat 3,717 45 25 23 0.7 0.4 Irian Jaya n.a. 337 131 85 n.a. 0.3 E. Timor n.a. 6 6 2 n.a. 10.3 Total 179,365 8,287 6,901 4,910 3.8 5.8 /a Includes land used for house compounds, gardens, shifting cultivation, and sawah in 1981; excludes pastures, uncultivated swamp, ponds, forested land or unutilized land. BPS, Statistik Indonesia 1984. lb Land allocated to sponsored migrants moved in Repelita III. Transmigrant land includes only house and field land. Source: Dir. Jen Pengerahan dan Pembinaan, Buku Data Usaha Tani 1985. - 84 - The Contribution of Transmigration to Regional Agricultural Production 4.27 The contribution of transmigration to regional agricultural produc- tion is difficult to assess for several reasons. First, it may take several years for migrants to establish themselves and for production, particularly of tree crops, to stabilize. Second, many transmigrant agricultural systems are upgraded over the years, through the addition of irrigation or tree crops, but there is no way to distinguish the contribution of transmigrants from that of the local population in provincial statistics. Finally, provincial data them- selves are rather unreliable, particularly for secondary food crops and tree crops. For this last reason, the following analysis will focus on rice production. 4.28 Table 4.9, which summarizes the recent impact of the Repelita III transmigration program on regional rice production, assumes (a) that 80% of all sponsored migrants produce rice with an average yield of 700 kg/family, (b) that one spontaneous family has already settled in each province (other than the Moluccas and Irian Jaya) for each sponsored family, and (c) that spontaneous families produce rice at the same level as sponsored. Using these assumptions, which are believed to approximate the actual situation, the incremental increase in rice production in the outer islands from Repelita III transmigration is estimated at about 240,000 tons per year, about 10% of the increase recorded in the outer islands over the past four years and about 3% of total incremental rice production in Indonesia. These figures, while not insignificant, do not point to Repelita III transmigration as a major factor in recent gains in rice production. This is not surprising in view of the relatively low productivity of food crop sites. 4.29 Transmigration has, however, made a significant contribution to incremental rice production in some provinces. In five provinces, recent transmigrants contributed an estimated 20-30% of incremental rice produc- tion: Bengkulu (19%), Jambi (23%), West Kalimantan (25%), South Sumatra (27%) and Riau (30%). In two provinces, their contribution is above 40% (South Kalimantan (45%) and Southeast Sulawesi (49%)); while in two sparsely populated provinces lacking traditional irrigated rice, it constitutes a major portion of the increase: (Central Kalimantan (81%) and Central Sulawesi (96%)). In East Kalimantan and Irian Jaya, the very small numbers involved make analysis difficult, but it appears likely that transmigrants may account for the entire incremental production in these provinces. It should be under- stood that due to the highly unreliable nature of the figures used, this analysis is meant only as a rough approximation of the contribution of recent transmigrants to rice production. 4.30 When all transmigrants are considered, the impact of the transmigra- tion program on rice production is more impressive. For example, in Lampung where 70% of all residents are of inner island descent, it is reasonable to assume that at least 70% of all incremental production in the last five years Table 4.9: THE CONTRIBUTION OF REPELITA III TRANSMIGRATION TO INCREMENTAL RICE PRODUCTION BY PROVINCE, 1980-1984 1984 Transmigrant Incremental production /a 1980 1984 Incremental Transmigrant families Sponsored Spons + Spont Provincial Provincial Provincial contribution Province 1979/80- (A) (B) production production production --- (%) 1982/83 ------------------------------('000) …-------------------------- (A) (B) Aceh 9,625 5.4 10.8 666 876 210 3 5 N. Sumatra 3,534 2.0 4.0 1,595 1,952 357 0 1 W. Sumatra 3,993 2.2 4.5 1,051 1,311 260 1 2 Riau 19,554 11.0 21.9 244 317 73 15 30 Jambi 11,361 6.4 12.7 389 444 55 12 23 S. Sumatra 66,616 37.3 74.6 872 1,147 275 14 27 Bengkulu 8,917 5.0 10.0 174 228 54 9 19 Lampung 6,316 3.5 7.1 686 1,030 334 1 2 Sumatra 129,916 72.8 145.5 5,678 7,305 1,627 4 9 W. Kalimantan 13,297 7.4 14.9 547 606 59 13 25 C. Kalimantan 10,197 5.7 11.4 206 220 14 41 81 1 S. Kalimantan 15,114 8.5 16.9 735 773 38 22 45 X E. Kalimantan 6,759 3.8 7.6 138 144 6 63 -/b n Kalimantan 45,367 25.4 50.8 1,626 1,743 117 22 43 N. Sulawesi 2,991 1.7 3.3 200 291 91 2 4 C. Sulawesi 11,905 6.7 13.3 228 242 14 48 96 S. Sulawesi 2,763 1.5 3.1 1,799 2,406 607 0 0 SE. Sulawesi 13,880 7.8 15.5 59 91 32 24 49 Sulawesi 31,540 17.7 35.3 2,286 3,030 744 - - Maluku 5,712 3.2 - 25 22 -3 - - Irian Jaya 9,775 5.5 - 3 4.4 1.4 -/b - Maluku & I.J. 15,487 8.7 _ 28 24.4 - -7 Total Outer Islands 222,310 125 240 9,618 12,102 2,485 5 10 Total Inner Islands - - - 20,032 25,392 5,360 - - Total Indonesia 222,310 125 240 29,650 37,494 7,845 2 3 /a Assumptions: (A) Assumes 80% of all families produce rice at 700 kg/family or 560'kg/family average over all transmigrants. (B) Assumes ratio of one spontaneous family/sponsored family and production at above rates. /b Error. - 86 - (estimated at 334,000 tons) is related to transmigration.7/ Assuming that inner island language speakers in the outer islands in 1980 produced the same proportion of rice as their share in the local population, transmigrants would be responsible for the production of 1.8 million tons of unmilled rice (gabah) or 15% of outer island production. By the same reasoning, transmigrants would be re ponsible for the production of some 2 million tons of unmilled rice in 1985,_y about 17% of outer island production and 5% of Indonesia's total rice production. Other Bank studies of the contribution of transmigrants to rice production in the outer islands have also estimated about 2 million tons. This is a small but significant contribution to the increase of about 20 mil- lion tons in the country's annual production over the last 20 years. Infrastructure 4.31 One of the most important factors promoting regional development and contributing to the integration of transmigrant activities and the regional economy is the construction and maintenance of roads. An extensive, reliable transportation network is essential to the marketing of agricultural commodi- ties, and to finding off-farm employment and maintaining household incomes. Data are not available for the total number and length of roads built under the auspices of transmigration since Repelita I, but since a majority of migrants were sent to the outer islands in Repelita III, the figures in Table 4.10 are highly indicative of the overall distribution of roads. As the table shows, between 1981 and 1985 about 4,200 km of new access roads, 7,700 km of new main roads and 14,000 km of village roads were constructed under the transmigration program, increasing provincial totals by an average of 20%. This is a significant achievement, and it lays the groundwork for economic development in many areas where it would not have been possible before. 4.32 Table 4.10 also shows the impact of transmigration roads on the road stocks in specific provinces. Because most roads constructed under the trans- migration program will eventually become the responsibility of the district (kabupaten) governments, the ratio between transmigration roads and district roads is most important, and indicates the additional burden that will have to be assumed by the districts. Comparing the length of access and main trans- migration roads to the length of district roads indicates that the ratio is highest in East Kalimantan where transmigration roads constructed between 1981/82 and 1984/85 were almost equivalent in length to the local district road network. Lampung and West Kalimantan had about half as many transmigra- tion roads as local roads, and in Riau, Jambi, South Sumatra, Central Kalimantan, and Irian Jaya transmigration roads account for 21-28% of local roads. 7/ While transmigrants have lower yields than local people due to their location on more marginal soils, virtually all transmigrants grow rice while many local people, particularly in Sumatra, concentrate on cash crops. 8/ About 1,800,000 tons (15% of total outer island production) plus 240,000 tons from Repelita III migrants. Table 4.10: TRANSMIGRATION ROAD CONSTRUCTION TN RELATION TO LOCAL INFRASTRUCTURE % Total Provincial Road Network Repelita III Total Repelita Repelita III Transmigration transmi- National/ Kabu- Transmiigration roads /a III Trans- roads as % of local roads gration Province Total provincial paten Access Main Village migration roads (4+5) as (4+5+6) as (4+5+6) as roads per (1) (2) (3) (4) (5) (6) (4+5) (4+5+6) a % of (3) a % of (3) a % of (1) province ----------------------------------- (ki) ---------------------------------- Aceh 9,982 2,906 7,076 118 218 404 336 740 5 10 7 3 North Sumatra 15,132 5,188 9,944 121 223 414 344 758 3 8 5 3 West Sumatra 8,632 2,772 5,860 52 96 178 148 326 3 6 4 1 Riau 7,262 2,344 4,918 364 670 1,240 1,034 2,274 21 46 31 9 Jambi 4,580 2,047 2,533 233 430 796 663 1,459 26 58 32 6 South Sumatra 9,692 3,891 5,801 568 1,051 1,946 1,619 3,565 28 61 37 14 Bengkulu 3,527 1,031 2,496 108 200 370 308 678 12 27 19 3 Lampung 4,596 2,007 2,589 480 884 1,637 1,364 3,001 53 116 65 11 Sumatra 63,403 22,186 41,217 2,044 3,772 6,985 5,816 12,801 14 31 20 49 West Kalimantan 4,182 1,812 2,370 398 737 1,365 1,135 2,500 48 105 60 10 Central Kalimantan 4,355 722 3,633 353 653 1,210 1,006 2,216 28 61 51 8 South Kalimantan 4,119 1,177 2,942 150 277 514 427 941 15 32 23 4 1 East Kalimantan 3,338 2,550 788 269 497 921 766 1,687 97 214 51 6 00 Kalimantan 15,994 6,261 9,733 1,170 2,164 4,010 3,334 7,344 34 75 46 28 1 North Sulawesi 5,627 1,645 3,982 55 101 188 156 344 4 9 6 1 Central Sulawesi 6,309 2,800 3,509 186 344 637 530 1,167 15 33 18 4 South Sulawesi 17,807 3,394 14,413 70 130 240 200 440 1 3 2 2 Southeast Sulawesi 5,012 1,305 3,707 162 299 554 461 1,015 12 27 20 4 Sulawesi 34,755 9,144 25,611 473 874 1,619 1,347 2,966 5 12 9 11 Eastern Islands 10,293 4,465 5,828 129 238 441 367 808 6 14 8 3 Irian Jaya 5,194 643 4,551 361 668 1,227 1,029 2,256 23 50 43 9 E.I and Irian Jaya 15,487 5,108 10,379 490 906 1,668 1,396 3,064 13 30 20 12 Total 129,639 42,699 86,940 4,177 7,716 14,282 11,893 26,175 14 30 20 100 /a Access road at 8.8 m/household; main road at 16.2 m/household; village road at 30 m/household. Source: Directorate PLP, 1985. - 88 - 4.33 The road network established due to the Repelita III program is expected to at least double the infrastructure stocks of the two districts where new transmigrants are equivalent to 40% of the 1980 population and in the seven other districts where they number 20% or more of the 1980 popula- tion. The large proportion of transmigration roads in some districts and provinces indicates the need for close coordination between transmigration and provincial public works offices in the design, maintenance and transfer of these roads, and this is emerging as a major issue in Repelita IV. 4.34 The problem has several dimensions. First, transmigration roads were initially designed to low standards and constructed by land clearing con- tractors who in many cases had little experience in road construction or main- tenance. Second, funds were not provided for maintenance while roads were under MOT authority during Repelita III. Thus, when sites are to be transferred to the provinces, roads are already in very poor condition and their rehabilitation overwhelms the limited financial and technical resources of the district public works offices. To compound the problem, district public works offices, which are generally weak, are particularly weak in those relatively underpopulated districts where transmigration occurs. This argues for an ancillary program to develop public works services in those transmi- gration receiving districts where the maintenance task will be most demanding and the capacity most limited. 4.35 Another element of this problem is related to the absorption of roads into the provincial and national road networks. National and provincial public works offices have a numbering system for roads under their authority, and funds for maintenance are allocated according to the length and condition of numbered roads, However, there is no automatic system for incorporating transmigration roads into the programs of the relevant public works offices and a lag of several years in incorporating them is not uncommon. The Ministry of Public Works has recently begun an inventory of all transmigration roads constructed in Repelita III, but arrangements have yet to be made for ongoing review and registration prior to transfer. 4.36 A separate problem is that incorporation of some new settlements into regional markets will require an extension of the national road network. For example, in South Sumatra where nearly 400,000 people were settled in the early 1980s, major investments in new access routes and/or a realignment of existing roads will be required to realize the economic poten- tial of new sites, but new construction wpll be difficult to undertake because of the high capital investment required.9- This suggests that transmigration 9/ The proposed road from Tanjung Karang in Lampung to Palembang is a case in point. This road would pass through an area where some 350,000 transmigrants are settled, provide access to sites such as Pematang Panggang which are almost entirely isolated, and provide a direct link between Java and Palembang (the major city in Sumatra). Both the economic impact and traffic potential of this road are unquestionable, but construction has not yet been initiated, mainly because of resource constraints. - 89 - planning must take a longer perspective and that the full costs of integrating transmigrant communities into the regions must be taken into account in the planning process. More specifically, a mechanism is now urgently required for identifying major road investments needed to support the regional development programs initiated by transmigration and for scheduling new investments. This will be difficult to support in light of budget constraints. The Impact of Transmigration on Regional Services 4.37 Roads are only one example of the expansion of rural infrastructure and institutions in response to transmigration. Other services which have been greatly expanded include extension, agricultural supply, education and health. During Repelita III, each of the line agencies responsible for these services held their own budgets, and planning and coordination were difficult (see Chapter VIII). In Repelita IV, the budgets for some activities were given to the MOT. Under these arrangements, the coordination of implementa- tion during the development period is expected to improve, but problems in the transfer of such services to the provinces are expected to increase. There- fore, effective hand-over must entail early planning for the number of people to be transferred (extension workers, teachers, health clinic personnel), adequate budget provisions for their salaries, and arrangements for incor- porating them into district or provincial services. Problems will also be exacerbated as the number of Repelita III sites to be transferred expands. 4.38 Large-scale transmigrant activities also have a major impact on urban development in their vicinity. Towns, even small ones, in transmigra- tion regions often experience a boom in activity either as a spin-off from construction, the relocation of civil servants and their consumption expendi- tures, or the search for "informal sector" work by the transmigrants them- selves. Many of these towns are growing at rates which, if they continue, will double their populations every ten years. For smaller towns which have no formal urban status and depend exclusively on central government budgetary allocations, rapid increases in road traffic and accelerated growth are diffi- cult to accommodate under current planning arrangements and will continue to be so unless special measures are taken which recognize the impact of transmi- gration. Indonesian cities also find it difficult to keep pace with accel- erated growth, particularly where the capacity is already low as in the large coastal cities and the small frontier towns in the outer islands.40/ 4.39 The point to be made is a simple one. Transmigration on the scale achieved in Repelita III is placing large demands on regional services and budgets in Repelita IV. These demands are so great that special arrangements including revised planning and budgeting procedures at the central level and accelerated institutional development at the local level are required. The Government has recognized this point, and is now using its limited resources 10/ In the outer islands, only about 30% of the larger urban centers have access to clean water supplies, less than 30% have access to sanitation facilities, and only 25% have any kind of formal garbage collection service. - 90 - for maintenance of existing sites, while cutting new development expenditures. D. Summary and Conclusions 4.40 This chapter illustrates the impact which the transmigration program has had on employment generation, demographic trends in the inner and outer islands and on regional development. Data analysis indicates that the transmigration program generated an estimated 500,000 to 600,000 permanent jobs in Repelita III. This is more permanent, self-sustaining employment than in any other government investment sector, other than tree crop development. The cost of creating a permanent job in transmigration is estimated at about US$4,000, which is higher than the cost of jobs in service sectors, but lower than the cost of creating employment in industry (estimated at US$10,000- $20,000 per job). This cost does not take into account benefits derived from temporary employment and spontaneous transmigration. 4.41 At the settlement levels achieved in Repelita III, transmigration absorbed about 12-15% of the incremental labor force in Java and Bali. If transmigration were sustained at half the levels achieved in Repelita III through the year 2000, and accompanied by spontaneous migration at historically observed levels, it would by the year 2020 have reduced the population of Java by 10% and absorbed about 20% of the incremental labor force. Even with lower rates of spontaneous transmigration, the program would reduce the population by 6% and absorb about 12% of new laborers. 4.42 The data also show that sponsored transmigrants settled during the Repelita III period were equal to as much as 40% of the 1980 population of some kabupaten; that farm land opened for transmigrants was equal to 10% or more of the estimated total area of agricultural land in four provinces; and that rice production by Repelita III migrants, while small in total, was significant in some receiving provinces. Road construction resulting from transmigration was estimated to have almost doubled the length of roads in one province and to have increased the stock of roads by as much as 50% in seven provinces and by 20% overall in receiving provinces. Other services which have been greatly expanded in response to transmigration are health, education and extension. 4.43 These data indicate the very significant impact which transmigration has had on both the inner and the outer islands. They also suggest that the expansion of population, regional infrastructure, and services on this scale require adequate arrangements for the transfer of roads and administrative functions to the provinces and the reallocation of funds from the center to the periphery to increase needed services. Investments should also be channeled to past receiving provinces to maintain employment levels in these areas. This may be difficult in the face of increasing resource constraints. - 91 - V. SOCIAL AND ENVIRONMENTAL CONSIDERATIONS A. The Impact of Transmigration on Local People People of the Outer Islands 5.01 There were about 60 million people in the outer islands in 1986; slightly over half (34 million) lived in Sumatra, about 12 million lived in Sulawesi, 8 million in Kalimantan and 6 million in the remaining eastern islands (see Table 1.1). Of these, about 1.5 million are regarded as isolated peoples, groups which for geographic, historic, or cultural reasons are not fully integrated into the national administrative system. Transmigrants have, in the past, had only limited contact with such people, who are generally located in remote or inhospitable areas with limited agricultural potential. * 5.02 Outer island people belong to a very large number of ethnic groups whose traditions, social organization and agricultural practices range from the well developed kingdoms of densely settled, wet rice agriculturalists; through settled groups mainly dependent on estate crops (rubber and coconut); to smaller, more dispersed groups, depending partly on shifting cultivation for their subsistence. While most "isolated" people live in areas with margi- nal soils, and therefore practice some form of shifting cultivation, not all do; and isolated people in the highlands of Irian Jaya, for example, have complex sedentary farming systems and high population densities. 5.03 In the past, transmigration programs have brought enough benefits to the regions in terms of infrastructure, employment, markets and labor to be generally welcome, and the provinces themselves have sponsored or assisted resettlement from Java and Bali, in addition to that supported by the central government. As noted in the income survey, outer island residents have higher incomes than transmigrants, on average, and with few exceptions they have more political power in the receiving areas. Therefore, while interpersonal and even interethnic conflicts do arise, they have been relatively rare, and no large-scale movement of people in modern times has been associated with less communal tension. In spite of this record, there are several areas where the social soundness of the program must be carefully monitored, for example, in land acquisition, in the provision of benefits to transmigrants and to local people in the development area, and in the treatment of ethnically distinct peoples. The following section discusses these points and makes recommenda- tions for improving the social soundness of the transmigration program. Land Acquisition for Transmigration 5.04 Basic Agrarian Law. Among the most frequent questions about the transmigration program are those centering on the procedures for land acquisi- tion. The Basic Agrarian Law of 1960 reflects a view of the nation set out in the constitution and reiterated in the state philosophy (Pancasila). This philosophy holds that the state has the right to use land resources to benefit the people as a whole. The law recognizes and protects customary land rights, unless they conflict with national interests, and it supports land registra- tion both to provide security of tenure and to indicate land without claims that can be reallocated for development purposes. - 92 - 5.05 The agrarian law recognizes traditional (adat) law. In most areas of Indonesia where adat law is followed, the community recognizes the right of the smallhoider to cultivate in a given area and to be compensated if trees or plants are damaged. Ordinarily, however, the individual does not own the land itself or have rights to it other than those accorded by community consent. Alienation of land must also occur with community consent and under these circumstances problems related to dividing compensation and competing land claims are common (below). Critics of transmigration have also voiced concern that agrarian law does not recognize the rights of shifting cultivators to land which has been cleared in the past but is not currently under production. While this is true of both national and traditional law, provisions in the basic agrarian law mandate Government to act in the best interests of the people, and local governments can, and frequently do, prevent the expropriation of land claimed by local inhabitants, even if it is not under permanent cultivation. 5.06 Compensation. The Government does not pay cash compensation for land relinquished for transmigration purposes, although it does provide cash compensation for tree crops in areas to be used for settlement. There are a number of reasons for this. The Government argues that compensation for land would both raise the price to local smallholders and make development programs less replicable; it feels that local people should be prepared to relinquish unutilized land for programs that will contribute to local and national deve- lopment; and it acknowledges that it would be difficult to ensure fair distri- bution of cash compensation for land where individual land claims are not clear. This policy has been possible in the past because adequate land has been made available on a voluntary basis. There is now little scope for addi- tional large-scale settlement in Sumatra without mechanisms for land purchase, and as land increases in value elsewhere (due partly to settlement) this compensation policy will have to be re-examined (see Chapter VII). 5.07 Land Disputes. Disputes over land and compensation take several common forms. In Sumatra, where adat law generally recognizes the right of the group to all land not under the authority of other local groups (i.e., whether cultivated or not), land cannot be acquired without the consent of the community, and most disputes are over compensation for productive trees or crops. The Government has standards for payment (depending on the age and productivity of trees) which are generally regarded to be fair, but disputes may arise over the number and type of trees, and there has been a tendency, of late, for local people to plant crops in areas being cleared, specifically for the purpose of obtaining cash compensation. Local groups also argue among themselves about the appropriate distribution of compensation. It is important to note that such disputes occur even where transmigration is - 93 - welcome.l/ Land disputes frequently delay settlement because provinces mediate land claims and pay the compensation, and the relative priority given to the issue by the MOT and the provinces may differ. 5.08 In Kalimantan and Sulawesi disputes are more likely to result from the competing claims of groups moving onto "state land." In West Kalimantan, for example, the Iban, local people who have been moving inland for many years, are aggressive in opening new territory which brings them into conflict with other local peoples and transmigrants. Because of the ambiguous status of land in West Kalimantan the most serious conflicts have arisen in this province. In Central and Southeast Sulawesi, several well publicized cases have involved confrontations between transmigrants and other immigrants, particularly the Buginese who are not part of the transmigration program. These conflicts have been over commercial transactions and over competing claims to land that the province allocated for settlement. Disputes of this sort are to be expected given the scale of the transmigration program and it is important to emphasize that in spite of such incidents, disputes over land have seldom generated permanent antipathy between ethnic groups, due in part to the efforts of provincial governments to resolve these problems equitably. The Provision of Benefits to Local People 5.09 In the past, transmigrants were small in number and were settled on marginal land, mainly with annual crops. They had few advantages in relation to the local population and because of their tenacity they generally earned local respect. Under Repelita III and the first two years of Repelita IV, however, transmigration occurred on a larger scale and benefits to the transmigrants were increased. Since the program shifted its emphasis from settlement to second-stage development, local people are increasingly sensitive to the benefits being provided to transmigrants and they are seeking parallel benefits. Several options are available. 5.10 Settlement within Transmigration Sites. In Repelita III, about 10% of the places in transmigration sites were allocated to local people. This was increased in Repelita IV. Local people and transmigrants in such sites tend to work separately, due to their different languages and cultivation practices. Some local people do well in transmigration settlements, but many maintain their residences in nearby villages, live off the site and return only to farm. Others sell their lots to spontaneous migrants and return to the more familiar life of long-established villages. It is not surprising that some local people, who have joined settlements to obtain land, prefer to return to their own villages: productivity is often low in transmigration sites, services are not well developed, and transmigrants have different 1/ In the Bank-assisted Transmigration II project (Ln 1707-IND), an area in Jambi Province was identified which was suitable for the settlement of 30,000 families. Problems over land classification and compensation ultimately reduced this number to about 10,000. Local people were not opposed to settlement, but wanted higher levels of compensation than the Government was prepared to provide. - 94 - languages and traditions. For these reasons, the benefits to be provided to local people through transmigration cannot be limited exclusively to settlement within transmigration sites. 5.11 Parallel Development. Local people outside of transmigration sites have benefited from new settlement through improved access and provision of services, and provinces have made an effort to upgrade facilities for local people in the vicinity of transmigration sites. In the future, however, if new settlement were to occur in more remote areas, more systematic attention will have to be given to parallel development for local people. This could involve the provision, as needed, of infrastructure, facilities, wells, and supplementary agricultural inputs (seedlings, livestock and extension). The Ministry of Transmigration could play a useful role in identifying parallel programs, by either paying for them or mobilizing provincial authorities to do so. 5.12 Second-Stage Development. In Repelita IV, the Ministry of Transmigration redirected its efforts from settlement to second-round investments in transmigration sites to improve infrastructure, introduce major tree crops and regularize land registration. There would be no equity issue if this assistance were directed only at critical sites, or if local people in the vicinity were already well-off. However, it is unlikely that major second-round investments can be made in transmigration areas without arousing resentment unless parallel benefits are provided to local people. To address this problem, it is necessary to provide many aspects of second stage investment to both local people and transmigrants, and to do this, investment plans must be prepared that take into account the needs of major receiving districts. This can be done by preparing plans for second stage development on a regional basis. The Treatment of Ethnically Distinct People 5.13 The People. Most outer island people are sufficiently well integra- ted into the political process to protect their own interests and to accept or reject transmigration depending on the benefits to them. But concern exists that less assimilated people may relinquish land against their best interests, or that very isolated people may be unable to assess the situation properly or assert their views. 5.14 About 1.5 million of the 60 million people in the outer islands are isolated people not fully integrated into the national administrative system. While small in number and ethnically diverse, isolated people are present in many outer island provinces. For example, there are a small number of nomadic, forest-dwelling people called the Kubu in the more remote areas of Jambi and Riau (Sumatra) who are so invisible that their exact numbers and habitat are not known. In the past these people retreated with the forest as local development progressed. To protect these people from both indigenous development and transmigration, and to prevent expropriation of their lands, the Governor of Jambi announced plans in late 1985 to protect a forest area where the Kubu can follow their traditional way of life. A study on the area to be delineated for this purpose is underway. To date, contact with and treatment of such people has been largely a provincial matter. - 95 - 5.15 Concern has also been expressed about the rights of the Dayak people in Kalimantan, although the term "Dayak" refers to many groups in Kalimantan which vary greatly in their degree of isolation and assimilation. These groups range from the aggressive and politically astute Iban of West Kalimantan, discussed earlier (para. 5.08), to other inland peoples who are more reclusive and retiring. Recognition of these cultural differences in the design of appropriate benefits or protection is important. It is noteworthy, however, that most Sumatran and Kalimantan people also have many common cultural practices, and when assimilated they are not distinguishable from other Indonesian nationals. This is not true in Irian Jaya, where the people are racially and culturally distinctive, and where management of cultural differences will require particular sensitivity. Because of the special features of the Irian situation, this issue will be discussed separately in Section C. 5.16 Policies. Bank policy on the incorporation of ethnically distinct people in development projects is being used by the Government as the basis for its own policies in transmigration areas. This policy indicates that development projects that use occupied lands should only proceed where adequate safeguards are provided for the local people. It also notes that nondisruptive acculturation is necessarily a slow process, and that projects affecting previously isolated people must provide the time and conditions for such acculturation. Projects affecting isolated or distinct people are required to include: (a) the recognition, demarcation and protection of areas containing the resources required to sustain their traditional means of live- lihood; (b) appropriate social services, including protection against new diseases and maintenance of health; (c) maintenance, to the extent desired by the local people, of their cultural integrity; and (d) a forum for the participation of local people in decisions affecting them and provision for adjudication and redress of grievances. 5.17 Guided by this policy and with the cooperation of the Government of Indonesia, the Bank's Transmigration V project (Ln 2578-IND) included a number of specific steps to ensure that the rights of local people were taken into account in site selection. Consultants undertaking site screening and evaluation were required to provide information on the number of local people in proposed project areas, describe their culture and subsistence base, and identify the area needed by them to pursue their traditional way of life; they also assessed the attitudes of local people toward compensation, resettlement, and the provision of infrastructure and services outside the site. To carry out these tasks, consulting firms employed appropriately qualified social scientists, and the project advisory group engaged an anthropologist/ sociologist on a full-time basis to oversee consultant work. Inclusion of these provisions in the project has had a positive effect in raising the level of awareness in the implementing agencies of social considerations. Improving the Social Soundness of Transmigration Programs 5.18 At the present time new settlement has been slowed and social issues are in abeyance. Were new large-scale settlement to be resumed, however, key institutions should be strengthened to deal with social issues. - 96 - (a) The Ministry of Transmigration is currently receiving technical assistance through the Transmigration V project to evaluate social policies, to resolve land disputes, to determine appropriate parallel benefits or to monitor treatment of isolated or ethnically distinct people. As these functions would increase in importance in any future program, the Ministry of Transmigration should also strengthen the existing Directorate of Social and Cultural Development to participate in these functions. The responsible Directorates should be staffed with appropriate technical specialists. (b) The Ministry of Home Affairs (MOHA) has a direct relationship with the Governors and Bappedas which are, in turn, responsible for protecting the rights of local people through the Directorate General of AGRARIA. It would therefore be appropriate for MOHA to organize seminars to sensitize provincial staff to the special problems of less assimilated people and to the Government's own policies in relation to them. MOHA and the Governors are also in the best position to ensure that the concerns of local people are adequately registered and addressed, and close supervision of these activities in the implementation period is required by Bappedas. (c) The Ministry of Population and Environment also has a role to play in monitoring the development of appropriate social policies and it could profit from additional technical support from either local or foreign anthropologists. There are also demographic programs in a number of local universities which could provide social assessment of the more traditional groups, working either for the Ministry of Transmigration or the Ministry of Population and Environment. Given the number of groups with responsibilities in the area, it would be helpful if an interagency coordinating committee, reporting to the Minister of Transmigration, were formed to deal with social issues (see Chapter VIII). B. The Impact of Transmigration on the Environment 5.19 Transmigration has as one of its major objectives the alleviation of population pressure on critical lands in Java and Bali. As noted in Chapter I, in many upper watersheds in the inner islands, food crops are cultivated on dangerously steep slopes and there has been encroachment on legally protected and production forests. Under such conditions, erosion has increased, there has been rapid siltation of dams which has necessitated the construction of new reservoirs, and downstream flooding is acute. To address these problems, the Government must relocate people from critical areas, and transmigration, with the promise of land ownership, offers an acceptable alternative to many farmers. The Government has also used transmigration as one option in the relocation of people displaced by natural disasters and development projects, people who might otherwise move onto increasingly marginal land. In this respect, transmigration has contributed to environmental protection in the inner islands. - 97 - 5.20 Most attention from environmentalists, however, is focused not on environmental degradation in Java, but on the outer islands and on the impact of transmigration, particularly spontaneous migration, on forest reserves, on areas such as tidal swamps, where the long-run consequences of drainage are uncertain, and on Indonesia's wildlife and biotic resources. The data on these subjects are poor and difficult to analyze, but some relevant facts are known. The following section will focus on the impact of sponsored and spon- taneous transmigration on forested land and on the role of shifting cultivation by local farmers and transmigrants. The purpose of this section is both to reduce misunderstanding about the transmigration program and offer concrete suggestions on ways in which environmental concerns can be better addressed. The Impact of Transmigration on Forested Land 5.21 Indonesia contains about 10% of the world's tropical rain forest and perhaps half of that remaining in Asia. It is one of the most biologically diverse areas in the world and has more than 500 species of mammals, 1,500 species of birds (17% of the world's av fauna) and a botanical diversity which includes over 10,000 species of trees.2 For this reason, Indonesia's forests and wildlife are a matter of international interest and Indonesia's steward- ship of them a matter of utmost importance. Because a portion of new settle- ment occurs on forested land, transmigration is an important topic among those interested in conservation of the world's natural resources. 5.22 The Conversion of Forested Land. The total area within forest boundaries in Indonesia is estimated at nearly 144 million ha or about 75% of the nation's land. Of this, 122 million ha is closed canopy forest. Areas classified as protection forests and nature reserves cannot be used for sponsored transmigration and are eliminated from consideration in the site screening process. Production forest is that area required for the sustained production of timber. It can be reclassified if the land is suitable for agriculture and if equivalent land is returned to production status. In practice this is difficult because of the strong economic interests of forest concessionaires. Conversion forest is land that can be converted to agriculture or other uses. Most of this land is already deforested or in areas along roads or rivers where exploitation by smallholders is inevitable. About 25% of all of Indonesia's land is in protection or conservation forest, 32% is production forest, and 16% of the area within forest boundaries is earmarked for conversion. Annex 5 shows the classified forest area by province as of May 1984. 5.23 Table 5.1 compares forest reserves by province with the land allocated for transmigration in Repelita III. The Ministry of Transmigration estimates that an average of 30% of land cleared for the sponsored transmigra- tion program was in heavy, logged-over forest. The remaining land cleared was about equally divided between secondary regrowth and grasslands, although the 2/ From a 1986 project proposal prepared by the International Union for Conservation of Nature and Natural Resources (IUCN). Table 5.1: FOREST RESOURCES BY PROVINCE COMPARED TO LAND ALLOCATED FOR TRANSMIGRATION IN REPELITA III (ha '000) Allocated as Allocated as Forest as a % of total a % of conver- Province Defined Conversion a % of Transmigration land /a forest land sion forest area forest forest prov. land Available Allocated 30% 50% 307 50% Aceh 5,539 3,282 192 59 27 17 0.1 0.2 2.7 4.5 N. Sumatra 7,168 3,526 253 49 15 13 0.1 0.2 1.5 2.5 W. Sumatra 4,230 2,942 437 70 12 9 0.1 0.1 0.6 1.0 Riau 9,456 6,546 1,754 69 80 58 0.2 0.3 1.0 1.7 Jambi 5,100 2,614 1,013 51 53 38 0.3 0.5 1.1 1.9 S. Sumatra 10,278 4,028 1,186 39 174 171 1.0 1.6 4.3 7.2 Bengkulu 1,978 992 193 50 22 23 0.6 1.0 3.6 6.0 Lampung 3,200 1,244 0 39 102 100 2.4 4.0 - - Subtotal 46,949 25,174 5,028 54 489 429 0.4 0.7 2.6 4.3 W. Kalimantan 14,600 7,695 1,508 52 57 25 0.1 0.1 0.5 0.8 C. Kalimantan 15,300 10,997 3,000 72 51 48 0.1 0.2 0.5 0.8 1 S. Kalimantan 15,114 2,029 284 55 47 37 0.5 0.8 3.9 6.6 E. Kalimantan 3,700 15,951 3,500 75 30 26 0.0 0.1 0.2 0.4 X Subtotal 54,824 36,672 8,292 67 187 136 0.1 0.2 0.5 0.8_ N. Sulawesi 2,751 1,583 699 58 8 7 0.1 0.1 0.3 0.5 C. Sulawesi 6,803 4,165 335 61 38 30 0.2 0.3 2.7 4.5 S.E. Sulawesi 3,814 2,190 699 57 39 36 0.2 0.5 0.5 0.8 S. Sulawesi 6,292 3,351 259 53 12 12 0.1 0.2 4.2 7.0 Subtotal 19,660 11,289 1,992 57 97 85 0.2 0.3 1.3 2.1 Maluku 8,572 5,096 436 59 18 15 0.1 0.1 1.0 1.7 Irian Jaya 41,066 28,816 11,775 70 34 13 0.0 0.0 0.0 0.1 Subtotal 49,638 33,912 12,211 68 51 28 0.0 0.0 0.1 0.1 Total 171,071 107,047 27,523 63 824 677 0.2 0.3 0.7 1.2 /a Land allocated for transmigration shown in Table 5.2. Source: FAO/World Bank Cooperative Program, Indonesia Forestry Project, Working Paper I, October 1985. - 99 - type of land cover varied by province. The table shows the estimated amount of forest clearing if 30% or 50% of the total transmigration area had been in forest land, both in relation to total forest area and conversion forest. Assuming that 50% (rather than 30%) of all cleared land was forested, land allocated to sponsored migrants exceeded 1% of the total forest area in only three provinces in southern Sumatra. Land made available to transmigrants, but not cleared, would increase these numbers slightly; and it is estimated that land claimed by spontaneous migrants would double the amount of land brought into production by sponsored migrants in Repelita III and would therefore double the percentages shown. 5.24 Table 5.1 also shows that transmigration has brought differing degrees of pressure to forest lands in different areas. Not surprisingly, South Sumatra province with high levels of settlement had a greater portion of conversion forest opened and for this reason, among others, there will be no further sponsored transmigration to that area. With the exception of densely populated South Kalimantan, however, forested land cleared for sponsored migration on the island of Kalimantan comprised less than 1% of the total forested area and perhaps 1-2% of the land intended for conversion. Irian Jaya, the largest province in Indonesia, shows the least impact. 5.25 Despite the fact that only a small proportion of Indonesia's total forested area was utilized in Repelita III, settlement in the five-year period reduced logged-over forest areas in Indonesia by at least 200,000 ha and reduced secondary regrowth by an equal amount, and it also put a smaller uncleared area under the command of sponsored migrants. These numbers suggest a loss of forest cover of perhaps 500,000 ha to the official program in the third five-year plan. Of this perhaps half was secondary regrowth which, while less rich in terms of diversity, nevertheless has environmental import- ance. If spontaneous migrants occupied an equal area, up to one million hectares of forested land would have been brought into agricultural production in the five-year plan. This is an important loss of tropical rain forest. 5.26 Ideally, no forest areas would be used for development. However, past experience in developed countries and current patterns in developing countries suggests that it will be virtually impossible to prevent the conversion of some forested land to agricultural purposes. For this reason, the effort to protect forests must concentrate on the demarcation and protection of land unsuited to agriculture. To do this it will be necessary to prepare master plans which clearly delimit areas for conservation and timber production; and indicate those areas with agricultural potential. 5.27 Land resource inventories carried out in the 1970s noted that there were 30 million hectares of degraded grassland in Indonesia and the question arises whether this land should not be used for transmigration, rather than forest. In Repelita III grasslands were used wherever possible as they are cheaper and quicker to clear, and very large settlements in Riau, for example, were reclaimed from grassy areas. However, Indonesia is so humid that secondary forest usually reasserts itself within a very few years unless persistent human activity, such as burning for pasture or short fallow culti- vation hinders regrowth, or unless the land is seriously degraded and unsuited to trees, as in the shallow soils of the highlands of Sulawesi. In both cases - 100 - land is rejected for transmigration either because of local land claims or soil infertility. Since most grassy areas suitable for reclamation and available for settlement were occupied in Repelita III, less of this land will be available in the future. 5.28 Transmigration and Shifting Cultivation. Shifting cultivation is a method of agriculture used by indigenous people on low fertility soils. When the natural vegetation is burned, nutrients are available in the soil for cultivated crops for a year or two, and when exhausted, new land is opened and the old land is left to regenerate. It is important to recognize that shift- ing cultivation is ecologically sound when population densities are low and the fallow period is long enough to permit forest regeneration. With recent advances in public health and population growth, however, shifting cultivation has led to extensive forest clearance, a short fallow period, and land degradation in many areas. 5.29 Recently, critics of the transmigration program have linked destruc- tion caused by shifting cultivation to the "nomadic practices" of spontaneous transmigrants. This is incorrect. Javanese transmigrants, whether they have been settled under sponsored programs or moved spontaneously, rarely engage in shifting cultivation on a large scale because they have limited access to land. They do encroach upon forested areas, occupying one to two ha per family, but they do not ordinarily put land under production on a temporary basis. 5.30 Most degraded land has been made so by indigenous cultivators. For example, the critical lands of Central Sulawesi and the northern reaches of South Sulawesi, mentioned above, are areas where indigenous cultivators have moved from old lakebeds of reasonable fertility up the sides of mountains with very shallow soils. The resulting grasslands are not suitable for cultivation by either local people or transmigrants. Similarly the degraded areas of Sumatra (the Toba plateau, Padang Bolak) are the result of intensive indigenous cultivation. In East Kalimantan where farmers are blamed for recent forest fires, almost all shifting cultivation is done by local people or Buginese immigrants who have no links to the transmigration program. 5.31 Analysis of maps and figures available for Central Kalimantan shows that about 3.4 million ha or 22% of the province have been cleared, 750,000 ha are degraded and abandoned, and only 300,000 ha are under permanent cultiva- tion. With approximately 260,000 farm families in the province in 1983, this suggests a ratio of 13 ha per farm family. Since there were fewer than 10,000 transmigrant families in Central Kalimantan in 1983, and most had been re- cently settled, this forest clearing cannot be attributed to transmigrants. Similarly, although land use maps are not yet completed for West Kalimantan, it is estimated that up to one-third of the province (4.8 million ha) has been cleared by a farming population of about 660,000 families. This is not an argument for complacency. However, it does suggest that the problems of forest degradation have been caused in large measure by local people and will not be solved by a halt to transmigration. To resolve the problem, local production systems must be upgraded (through the introduction of more productive tree crops, for example) as this will both reduce shifting cultivation and make some land available for others. - 101 - 16 Local farmers practicing shifting cultivation on the shallow soils of highland Central Sulawesi have left permanent grasslands behind them. Other Areas of Environmental Concern 5.32 Swamps. Questions have been raised about the environmental implications of drainage in tidally influenced areas. In particular, concern is expressed about the acidification of rivers in areas where peats are drained, and about the loss of mangroves or wildlife habitats. Environmental impact studies of both Sumatra and Kalimantan, undertaken by Indonesian universities, have not shown irreversible damage to streams or rivers, but concern over the long-term impact remains. In 1984 a long-term monitoring study of the effects of drainage was begun under the second Bank-assisted Swamp Reclamation Project (Ln 2431-IND), but this study focuses primarily on the Bank-assisted settlement area and additional work could profitably be done on other sites. Stronger linkages should be developed between universities and the Ministry of Population and Environment. 5.33 Wildlife. The Indonesian Government is committed to the conserva- tion of forests and wildlife species and has demarcated areas for their con- servation. Areas demarcated for conservation purposes cover 10 million ha and - 102 - 300 sites, and include 17 national parks, and 70 nature and game reserves of which 4 are biosphere reserves. These areas are underfunded and inadequately managed, and an international effort to mobilize funds and technical expertise to improve wildland management is urgently required. The Government is committed to the preservation of endangered species and has relocated some animals, including elephants, from existing transmigration sites in South Sumatra. However, human encroachment on the natural habitats of wild animals will continue to be a problem and mechanisms must be developed to monitor and protect critical areas, and to provide the human resources to carry out these tasks. 5.34 Spontaneous Migration. Transmigration settlements are envisaged as growth poles, but the implications of this, particularly in terms of the space required, have not yet been incorporated into planning and this has had nega- tive environmental consequences. In Lampung, which has received the most unsponsored transmigrants, for example, the integrity of forest and nature reserves is threatened, and in the past few years the Government has under- taken a program to resettle 45,000 families within the province. To help avoid such problems in the future, spontaneous migrants must be incorporated into the planning process and land must be reserved for them. It is also important that the Government develop a mechanism to facilitate land registration and develop credit mechanisms for land purchase. These matters are discussed further in Chapter VII. Improving the Environmental Soundness of Transmigration 5.35 Institutional Development. Three specific agencies have a strong role in improving the environmental soundness of transmigration and protecting wildlife and the environment: the Ministry of Population and Environment; the Regional Environmental Studies Centers, which are located within selected universities; and the Directorate of Conservation within the Ministry of Forestry and the Environmental Directorate within the MOT. 5.36 The Ministry of Population and Environment is in a good position to monitor the transmigration program from an environmental point of view and to help develop policies and guidelines for those carrying out site selection, environmental assessment studies, and land development. Technical support was provided to this agency under the third Bank-assisted transmigration project, and consideration should be given in the future to increasing specialist support to the agency, particularly in wildlife management, swampland assessment, and other technical specialities. 5.37 The Regional Environmental Studies Centers in provincial universities are in a good position to carry out environmental assessment studies and are already doing so, particularly in swamp areas. They could also play a key role in ensuring that buffer zones are established near transmigration sites and that wildlife areas are protected. These centers are currently being assisted by the Bank and UNDP, but they need to be further strengthened, and linkages with the transmigration program should be formalized. Resources for both domestic and external training should also be increased, and support should be given to these universities to train officials and MOT staff in environmental matters. - 103 - 5.38 The Directorate of Conservation in the Ministry of Forestry (MOF) has the most important role to play in environmental protection as the bulk of all lands which should be gazetted and protected are forestry lands. Conside- rable work has already been done by the Ministry, by universities, and by organizations such as the World Wildlife Fund, to identify areas of biotic importance and a National Conservation plan was prepared in 1982. Therefore, consideration should now be given to a major program to consolidate this work, determine priority areas for protection and begin the very large tasks of translating this plan into reality. 5.39 An Environmental Directorate has recently been formed in the Directorate General of Settlement Preparation within the MOT. Because of its location, this Directorate can potentially play an important role in improving the environmental soundness of site selection and settlement development and it can serve a useful role in facilitating the work of interagency teams developing policies for environmental protection (Chapter VIII). 5.40 Any effort to mobilize and support an environmental protection program would benefit from close coordination between Indonesian line agencies, particularly the Ministry of Forestry, which has the mandate to carry out such programs; the international environmental organizations (World Wildlife Fund, IUCN, among others) that have the technical expertise to assist in the definition and delineation of priority areas to be protected; and bilateral and multilateral agencies, which have the financial resources to assist in implementation. One possibility would be for the Ministry of Forestry or the Ministry of Population and Environment to form a consortium of environmental organizations and donor agencies interested in environmental protection and to coordinate their technical and financial inputs. It is likely that grants and concessional aid would be made available for a systematic effort in this direction. C. The Case of Irian Jaya Special Features 5.41 Irian Jaya, which is about the size of Spain or California, is the largest province in Indonesia and the least densely populated with about 3 people/km2. In general, soils in many parts of Irian Jaya are somewhat better than those in areas such as Central Kalimantan. Of the population of 1.2 million, some 800,000-900,000 are people of Papuan descent. Non-Papuan peoples include long-time residents from eastern Indonesia, located primarily on the islands and coastal areas in the west; more recent migrants from surrounding areas; civil servants and transmigrants. About 19,000 transmigrant families were settled in Irian Jaya through Repelita III. Initial targets for Repelita IV projected some 140,000 families moving to Irian Jaya within the overall 750,000-family target, and the prospect of movement on this scale raised alarm among environmental and social scientists. Actual movement in Repelita IV is expected to be only about 7-8% of targets. With the reduction of the land clearing program in February 1986, total settlement of sponsored and partially-assisted transmigrant families in Irian Jaya had reached about 10,000 families by June 30, 1987. - 104 - 5.42 The Government is committed to settlement in Irian Jaya to (a) help realize the demographic and employment objectives of transmigration; (b) provide the people and economic infrastructure, such as roads, which are crucial to regional development; and (c) promote national integration. The speed of settlement in the area, however, is constrained by the absence of infrastructure and services and the high cost of establishing them. In addition, the social and environmental considerations noted below argue for slow and systematic development. Social Issues 5.43 Perhaps 0.5 million Irianese live in the fertile highlands of the Baliem Valley and Paniai Lakes area where they pursue their traditional way of life with minimum exposure to the modern cash economy. Their main contact with the outside world is through an active missionary network and more recently, through government administration and the educational system. Outside the highlands area, the environment is less hospitable and lowland people are more sparsely settled, often basing their subsistence on shifting cultivation, hunting and gathering, or sago harvesting. Because the highlands are fully occupied, there are no plans for transmigration settlement there. Most sites were planned for coastal areas in the north and west, and on the south coast around Merauke, areas where populations have remained low in relation to carrying capacity due, in part, to climatic factors such as drought and flooding and to endemic malaria. These locations were chosen for settlement as they were felt to have the best agricultural potential of those areas not densely settled. 5.44 The most sensitive issue in such settlement is the fairness with which the rights of the local people will be taken into account. As noted earlier, concern for this issue figured prominently in the design of the most recent Bank-assisted project, Transmigration V. Up to 50% of all those to be settled in transmigration sites in Irian Jaya can be native Irianese, and many of these are likely to be migrants from the highlands and small island areas where economic opportunities are limited. The Government has also recognized the need for culturally appropriate "parallel development," i.e., comparable services provided to local people who do not want to settle in transmigration sites. However, as Bank policy statements acknowledge, the process of acculturation and growth of awareness about one's rights and interests is a long one and for this reason, slow, well-planned transmigration coupled with sound programs to benefit the local people and respect their cultural identity is required. 5.45 Regional development studies financed by UNDP and executed by the Bank are focusing on programs to assist the indigenous people of Irian Jaya, and should be very useful in defining culturally appropriate development. Steps to improve the social soundness of the transmigration program will also contribute to the quality of settlement in Irian Jaya, and a Directorate in the Ministry of Transmigration to develop and monitor programs for indigenous people, could, if appropriately staffed, develop policies to ensure that the rights of Irianese were protected and that benefits provided are consistent with the culture and level of economic development of the people. Since traditional law in Irian Jaya has not been codified, consideration should be given by the MOHA to providing support for such work. - 105 - Environmental Concerns 5.46 Irian Jaya has the largest undisturbed lowland rain fore) in Southeast Asia and is unequalled in biotic richness and diversity.- In keeping with the importance of the area, the Government together with the Food and Agriculture Organization, the World Wildlife Fund and others, has prepared a series of proposals for a system of parks and nature reserves in the province. The Government is committed to conservation of these areas and has rejected transmigration sites which infringe on existing or proposed reserves. However, forest concessions, proposed roads and the prospect of large-scale settlement will have a significant impact on the environment unless a major effort is undertaken to complete the gazetting of environmental protection areas, parks and wildlife protection areas. Plans are also needed to keep roads at a distance from protected areas and to create buffer zones between transmigration sites and protected areas. Support for these activi- ties from environmental organizations and bilateral and multilateral agencies would be relatively easy to obtain. Support should also be considered for the Environmental Study Center of Cenderawasih University, Jayapura, to monitor the environmental impact of transmigration in Irian Jaya. 5.47 With respect to transmigration planning in Irian Jaya, the World Wildlife Fund specifically recommends: (a) integration of environmental planners into presettlement planning teams and review of settlement plans by the State Ministry for Population and Environment (MPE); (b) adequate supervision of land clearing contractors to ensure adherence to site clearing instructions; (c) elimination of wasted forest resources by close coordination of land clearing with logging interests; and (d) development of an extension service through the Bureau of Environ- ment in the MPE, to create a public awareness and appreciation of conservation values, including adequate information to the settlers themselves. These recommendations should be adopted and efforts should be expanded to ensure the preservation of important wilderness areas. D. Summary and Conclusions 5.48 Impact on Local Peoples. Past transmigration programs have received a generally positive reception, and no program of this magnitude has been carried out with less communal tension. Attitudes could change, however, if the scale of the program increased, or if migrants were sent in large numbers 3/ For details see the 1985 World Wildlife Fund study "Conservation and Development in Irian Jaya: A Strategy for Natural Resource Utilization." - 106 - to more remote areas, or among people who are culturally different. Parti- cular sensitivity is required in the management of development in Irian Jaya. To improve the social soundness of the transmigration program, the appropriate Directorate within the Ministry of Transmigration should be strengthened to deal with policies for compensation, resolution of disputes, the development of parallel programs for local people, and the protection of ethnically distinct groups. A number of steps have been taken by the Government to assess the needs of local people in transmigration areas and to ensure that they benefit from the program, but successful implementation will require (a) involvement which is voluntary and gradual and provides the time for necessary adaptation; and (b) careful monitoring and strengthening of the institutions responsible for this task. 5.49 Impact on Forested Land. Migrants on average receive and cultivate less than 2.0 ha of land per family. Since 1980, the Government has moved about 500,000 families, and it could possibly move 500,000 more by the end of the century, depending upon the scale of the future program. This means that, at an upper estimate, some 2.0 million ha could potentially be put into production by sponsored migrants. An equal area may be put into production by spontaneous migrants. If 50% of this land is forested, this would mean the loss of 2.0 million ha of forested land by the year 2000. This is a very large area, although it is less than the area under shifting cultivation in Central Kalimantan alone. It is equivalent to about 10% of conversion forest and 2% of the total forested area of the outer islands. The loss of this forested area must, therefore, be weighed against other economic and employment considerations. 5.50 It is unlikely at Indonesia's level of development, that areas suitable for agriculture can be maintained permanently under forest cover. Strong programs are therefore needed to protect areas unsuited to cultivation, to protect wildlife and biotic reserves, and to bring spontaneous migrants into the planning process to reduce their encroachment on protected land. Programs are required to upgrade indigenous agriculture, thereby reducing shifting cultivation and freeing land for population growth and new settlement. In the future, mechanisms must also be developed for land purchase and registration that permit migrants to acquire land already under cultivation. - 107 - VI. AGRICULTURAL PROSPECTS A. Upland Sites 6.1 In Repelita III about 290,000 transmigrant families were settled in upland sites based on rainfed food crop agriculture, nearly 80% of all those moved. The upland food crop model permitted the rapid establishment of fami- lies and a large increase in the scale of the program; it enabled transmi- grants to meet their own subsistence needs and quickly reduce their dependence on the Government; and it was less costly than settlement based on swamp reclamation or tree crop development. To move beyond subsistence, transmi- grants were expected to gradually put more land into production, use improved farming and conservation techniques, and increase their yields. Officials expected that the sale of surplus production would increase farmer incomes and reduce national and regional food deficits, particularly in rice. Recent studies indicate, however, that rice production is modest (Table 6.1), and incomes and economic rates of return are low. Therefore, in order to evaluate the prospects for this model, it is necessary to understand the factors which limit production in upland areas and to ask whether these limitations can be overcome. Table 6.1: REPORTED ANNUAL PRODUCTION OF UNMILLED RICE IN UPLAND SITES Average Average Families family Settlement family growing production years Sample size production rice for growers (kg) (Z) (kg) BPS Survey Repelita II 1975-79 419 470 83% 567 Repelita III 1980-82 824 824 87% 947 Repelita III 1982-84 746 746 83% 898 IFAD Survey 1974-84 1,200 760 86% 883 Source: Farmer reports, BPS Transmigration Income Survey 1985 and IFAD Transmigration Survey, 1985. Factors Affecting Food Crop Production in Upland Sites 6.2 Agro-ecological Conditions. Most transmigration sites are located in upland areas with red-yellow podzolic soils of low inherent fertility. These soils are generally characterized by low pH and aluminum toxicity, which restricts the growth of certain species, and by high erodibility. High levels of exchangeable aluminum, which increase with depth, cause shallow rooting of - 108 - plants, and limited organic matter reduces the ability of the soils to retain water and nutrients. Typically, yields on such soils are moderate to good the first year after clearing and burning, and they improve in the second and third year as organic materials (roots, stumps) decay. Yields decline there- after as nutrients are leached from the soil and weeds become harder to control. Most experts agree that land clearing with chain saws and manual labor is less destructive than mechanical clearing, but regardless of how land is cleared, fertility eventually declines to low levels. 6.3 These problems can be addressed, although not entirely overcome, by liming to increase the pH and reduce aluminum toxicity; by applying chemical fertilizers, particularly phosphates and urea; by incorporating organic mater- ial into the soil on a regular basis to add nutrients and improve the water retention capacity; and by introducing erosion control measures such as bund- ing and bench terracing. However, many of these measures involv,e complex technologies and are either costly or time-consuming to carry out. For example, bench terracing is estimated to require 200-250 man-days of labor per hectare; and liming, at two tons of lime per hectare, may cost up to US$80- $120, the equivalent of 80-120 days of off-farm work. The distribution of lime also entails considerable extra effort on the part of the farmer, particularly if his field is not adjacent to a road. Under these circum- stances, such practices are seldom adopted by the farmers unless labor or capital is subsidized, or unless they can be certain of adequate returns. 6.4 Climatic factors work against the certainty of adequate returns. Although most sites in western Indonesia have at least 1,500-2,500 mm of rainfall and a seven- to nine-month growing season, rainfall is highly localized and periods of drought occur even in the wet season. Rainfall at the beginning and end of the rainy season is particularly uncertain. Where soils have high organic content and good water retention properties, dry spells are not a particular problem, but on red-yellow podzolic soils where water retention is poor and crops tend to be shallow rooted, rainless periods at critical points in the growing season can cause serious damage to crops and weaken their natural defenses against disease and insect pests. Agro- ecological studies conclude that the soil and climatic properties of most settlement sites make them well suited to deep-rooting, acid-tolerant tree crops such as rubber, suited to certain tuberous crops such as cassava, only marginally suitable for the production of rice and corn, and unsuited to plants preferring less acidic soils such as soybean. 6.5 One approach to overcoming agroclimatic constraints is through agricultural research intended to develop cropping systems, farm management practices and cultivars suited to outer island conditions. Such research is being carried out by the Agency for Agricultural Research and Development at 20 sites in Indonesia, of which about half are in transmigration areas. This research has already led to the introduction of the rice/corn/cassava system of intercropping used by most transmigrants. To date, however, research has been carried out without adequate attention to the socioeconomic and risk- aversion factors which influence farmers' judgements, and the analysis presented in Chapter III indicates that some proposed high-input/high-output models have lower financial returns than traditional farmer practices. The Agency for Agricultural Research and Development is aware of this problem and modifying its work accordingly. - 109 - Transmigrant site located in grassland area. The alang-alang (imperata cylindrica) in the foreground is easy to clear but difficult to eradicate completely. 6.6 Marketing Constraints. Some transmigration sites are located near major markets, and transmigrants there have earned sizable incomes from the sale of chilies, vegetables, citrus and the like. However, most transmigra- tion sites, by definition, are located in sparsely populated areas. Given the long distances to major markets and the poor quality of roads, transport costs are high and discourage trade. In the past, limited access has meant that perishable goods such as vegetables and fruits were produced almost entirely for home consumption; and nonperishables such as surplus rice and peanuts could be sold to middlemen in the markets, though at low farmgate prices reflecting high transport costs. 6.7 To address marketing constraints, the Government has two mechanisms for intervention. The first and most important is purchases by BULOG (a government body used mainly for the purchase and redistribution of rice). In theory, BULOG purchases all surplus rice meeting specific quality standards from village cooperatives at a specified floor price. In practice, with the rice surpluses of the past year and severe storage problems, BULOG is pur- chasing only rice of the best quality. Under these circumstances BULOG - 110 - purchases have slowed, rice surpluses have gone onto the market, and the price has declined. It would be possible for BULOG to give preference to rice purchases in transmigration sites, but this would be costly to the economy since transport costs are high and it is far more expensive to grow rice in transmigration areas than in either Java or the better areas of the outer islands. 6.8 The second method of intervention is through the establishment of agroprocessing industries in or near transmigration sites. There has been some success in using government estates (PTPs) and private estates either to grow cash crops using transmigrant labor or to purchase cash crops grown by migrants. For example, migrants are growing sugar for PTPs in transmigration areas in Sumatra and Kalimantan, and private enterprises purchase tobacco and cotton from migrants in Sulawesi. Second-stage developments of this type should be strongly encouraged. The Government has also discussed the pos- sibility of linking transmigration settlements to major industrial develop- ments such as the aluminum factory at Asahan, Aceh, or the proposed pulp and paper mill in Sesayap, East Kalimantan. Such projects could circumvent mar- keting problems and should also be encouraged, but the number of settlements possible on this basis is small. 6.9 Where transmigration sites are located near markets or have good infrastructure, private sector investment typically enters the surrounding areas without government intervention. For exampl.e, outside of Way Abung, an older settlement area in Lampung (settled 1965-74), three mills using cassava and cassava by-products were established when roads were improved, and several small estates (500 ha or less) were developed for cassava, peanut and coconut, taking advantage of transmigrant labor. Government regulations intended to protect transmigrants and support the development of cooperatives have discouraged the establishment of small enterprises in transmigration sites in the past and these regulations should be changed. In general, however, private investment in most transmigration areas is likely to be slow since development costs are high in transmigration areas and the returns to invest- ment are low. 6.10 Other Constraints. There are a number of other constraints to production and those related to the remoteness of transmigration sites are particularly difficult to address. (a) Extension. Experienced extension workers are not eager to live under pioneer conditions in transmigration sites, so staff in those areas are generally recent graduates with little field experience. Their numbers are small, transport is limited, and they lack the skills to evaluate diverse conditions and to introduce the complex technologies required to manage upland soils. (b) Input supply. The supply of inputs is also erratic in remote areas. Steps have been taken to improve input supply but the support packages of seed and fertilizer are often delayed and seed is frequently of inferior quality. - ll - Cc) Credit. Production credit is only available to a limited degree in specific soyabean and maize production programs, and there are virtually no credit programs for other production systems or for opening and developing reserve land. Specific recommendations to address these problems were given in the first Transmigration Sector Review (April 1981), and some improvements have been made since then in the delivery of services. On the whole, however, most of the problems noted then still persist and cannot be easily overcome. There- fore, while improvements should be strongly encouraged, it is unlikely that the quality of extension and input supply will be significantly different in the near future. 6.11 Design Considerations. While agronomic conditions and constraints on marketing and services are difficult to address, production constraints in transmigration sites arising from factors related to project design are easier to control. Chief among these are the following: (a) Land Clearing. Soil scientists, economist, and environmentalists generally favor manual or semimechanical 1 land clearing methods over the mechanical methods normally employed in the clearance phase of transmigration. While it is possible for experienced operators to use large machines in ways that minimize soil compaction or the loss of topsoil and soil nutrients, care is not always exercised. Therefore, given the difficulty of supervising and managing mechanical land clearing, manual or semimechanical clearance is a desirable alternative. Such methods would also generate more employment. (b) Farm Layout. Before 1980, most sites were laid out with fields located directly behind the houselot or relatively nearby (the linear settlement model). In Repelita III, however, houselots were generally grouped into villages of 350-500 families and fields were located outside the village area (the nucleated settlement model). Nucleated villages promote interaction, improve access to schools and health centers, reduce the required road infrastructure and facilitate the introduction of water and electricity when the commu- nity is developed. But they have disadvantages for agricultural production: it takes more time to get to the fields, the use of intermittent family labor is constrained, the use of inputs is reduced, pest protection is poor, and farm diversification is limited. For these reasons, food crop production on linear settle- ments is higher than on nucleated settlements. If agricultural production is to be maximized, serious consideration should be given to returning to the seminucleated model. 1/ Semimechanical land clearing involves the use of chainsaws and sometimes winches to remove large logs. - 112 - (c) Land Allocation. During the 1950-80 period, farmers were ordinarily given 2.0 ha of land, a sufficient holding if irrigation was intro- duced. With the adoption of the upland farm model in late Repelita II, holdings were first expanded to 5.0 ha and then reduced to 3.5 ha when it became clear (i) that family labor was insufficient to cultivate 5.0 ha, and (ii) that holdings of this size would reduce the number of families which could be settled. Since most families cultivate no more than 1.0-1.25 ha, the Government has recently decided to reduce holdings to 2.0 ha. Unless the farm size is carefully considered, however, this decision could potentially prevent the introduction of tree crops at a later date. This would, in turn, limit production and income on family farms. For this reason, settlements should include at least 2.5-3.0 ha/family, and should be sufficient to generate target annual incomes of US$1,500 per family. (d) Road Maintenance. One of the most severe constraints to input sup- ply and marketing is the quality of roads in transmigration sites. Difficult road conditions raise transport costs, restrict the use of inputs, and hamper marketing. The Government has recognized this problem and begun to allocate funds for road rehabilitation. Improvements are still required, however, in design standards, contractor supervision and routine maintenance. Farmer Responses to Upland Conditions 6.12 Farmer response to the risks inherent in rainfed food crop produc- tion is predictable and relatively well documented. In general, farmers avoid large investments of either labor or capital in agricultural production. They do not use recommended applications of fertilizer and pesticides, nor will they undertake costly erosion control even if yields can be increased through such measures. Instead, they concentrate on the production of food for sub- sistence purposes and for sale within the community, and they diversify their production systems away from food crops over time. In the BPS survey, Repelita III migrants obtained about 30% of their total incomes from food crops and older Repelita II migrants, 20%, while Repelita II migrants obtained three times as much from tree crops as those settled in the early years of Repelita III. Livestock, poultry, eggs, fish and vegetables, raised mainly for subsistence, produced 18% of farm income (see Table 6.2). 6.13 Government officials in the Directorate General of Food Crops (DGFC) are aware of the constraints to production and have made a number of sugges- tions to improve farmer cropping practices. Chief among these is a recommen- dation that rice should be produced mainly on wet ricefields (including tidal swamps), and that upland areas should be converted to diversified "conserva- tion farming." The model proposed by DGFC and endorsed by the Agency for Agricultural Research and Development (AARD) would involve a combined field and houselot of 0.75 ha on which a mixture of food crops, tree crops and small livestock would be grown. Advantages to the combined houselot/field plot are that labor can be used more effectively, plant protection can be intensified, soil fertility can be improved by the application of manure and other organic wastes, and postharvest protection is better. Small livestock including - 113 - chickens, ducks, rabbits and goats would be an important part of this farming system, contributing both to home consumption and cash incomes. In addition to the area provided to meet subsistence needs, 1.5-2.0 ha per family would be set aside for cash crop development. This land would be allocated when tree crops or cash crops were established and then only to those farmers prepared to undertake the credit and labor obligations entailed in tree crop develop- ment. A shift to this model in upland sites should be strongly encouraged. Table 6.2: DISTRIBUTION OF AGRICULTURAL INCOME IN UPLAND SITES, 1985 (Annual Income - Rp'000) Repelita II Repelita III Repelita III % of 1974-79 1979-82 1982-84 Total Total (Sample size) (419) (292) (788) (1,499) Food crops 112 210 164 158 27 Tree crops 51 17 6 21 4 Other agriculture 105 108 107 107 18 Subtotal 268 335 277 286 49 Nonagricultural /a 308 320 287 299 51 Total 576 655 564 585 100 /a Excluding pensions and transfers. Source: BPS Transmigration Income Survey, 1985. Summary 6.14 There is a broad consensus on steps which should be taken to support agricultural production within the existing program. First, those constraints that are a function of design problems should be changed. Manual or semi- mechanized land clearing should be encouraged, both to reduce disturbance to the soil and to generate employment. Houselots and food crop areas should be combined, land allocation should ensure that adequate land remains available for second-stage development, road maintenance standards should be reviewed and funding for maintenance increased. Second, research, extension, input supply and credit arrangements should be strengthened. Finally, upland farm models should assume that farmers will produce food crops largely for subsis- tence purposes and on-farm diversification should be encouraged. The fact that many of these recommendations have been promoted by the Ministry of Agriculture (MOA), but not adopted by the Ministry of Transmigration, points to the need for a stronger role for MOA in setting agricultural standards and for closer coordination between the two (see Chapter VIII). 6.15 Some older transmigration areas, such as those in Lampung, are now the heart of more densely populated regions and since marketing prospects in - 114 - these areas have improved and the prospects of regular input supply are better, experimentation intended to increase food crop production beyond household consumption levels should be focused on existing sites in these areas. In general, however, the inherent limitations in food crop cultivation in remote transmigration sites, the high cost of producing food crops under adverse conditions, and the deteriorating market prospects for commodities such as rice, argue against the assumption that income growth in most upland sites will come from surplus food production. B. Swamp Reclamation Sites 6.16 In Repelita III, about 67,000 families were settled in swamp recla- mation sites, 18% of the total moved (see Table 6.3). Settlement in areas of swamp reclamation have a number of attractive features. Good soils in swampy areas have a high organic content and can produce relatively good yields even without fertilizers. For example, tidally influenced areas in South Kalimantan have annually produced two tons of rice/ha for over 40 years and farmers in the older (Repelita II) swamp reclamation sites reported rice yields of 1.7 tons/family and the highest incomes of all food crop cultivators in the BPS transmigration income survey. Areas cultivated are flat, which eliminates erosion problems, and they lend themselves to bunding that holds water on the land and benefits wetland rice. The timber on swamp sites has less commercial value than that on upland sites and since swampy areas are largely uninhabited before they are drained, they are relatively free of land claims. 6.17 There are also serious problems in swamp sites. Some areas have difficult soils which may retard or even prevent the growth of certain crops, water management intended to prevent overdrainage and saline intrusion is a complex and inexact science, particularly under Indonesian conditions, and concerns exist about the possible environmental consequences of the large scale drainage of coastal swamps. Recently settled farmers in tidally reclaimed areas have also experienced very severe problems with rats and insect pests, and farmers in Repelita III swamp sites report the lowest average incomes in the BPS income survey. These swamp sites have the most people below subsistence production, the highest rates of abandonment, and very low incomes from off-farm work. In recognition of the difficulties faced by the settlers in tidal areas, subsistence supplies have been extended from 12 to 18 months by MOT, Repelita IV targets for tidal areas were reduced even before budget constraints slowed the new settlement program, and funds have been diverted to canal rehabilitation and the upgrading of existing sites. Factors Affecting Swamp Development 6.18 Physical Features. Lands used for swamp reclamation are general y located in tidally influenced areas near the coasts of the major islands.2 These areas are very flat and have poorly formed drainage networks. For this 2/ Freshwater inland swamps also have settlement potential but they are less extensive and have not been developed on a significant scale to date. - 115 - reason, low-lying areas are often inundated by surface runoff during heavy rains, by river flood waters in the rainy season or by tidal action which raises the level of the river and backs seawater or freshwater over the land. Such water is trapped when floods or tides recede, and the resulting swamps are difficult to drain on a small scale and have high levels of malaria. For these reasons, most tidally influenced areas are inhospitable to people and have very sparse human populations. Table 6.3: FAMILIES SETTLED IN SWAMP RECLAMATION SITES IN REPELITA I, II AND III Repelita I Repelita II Repelita III Total 1969-73 1974-79 1979-84 Sumatra Riau 355 700 12,865 13,920 Jambi 2,585 500 4,845 7,930 South Sumatra 2,100 4,800 34,140 41,040 Subtotal 5,040 6,000 51,850 62,890 Kalimantan West 500 2,600 2,600 5,700 South 1,535 750 1,690 3,975 Central 1,675 250 10,875 12,800 Subtotal 3,710 3,600 15,165 22,475 Total 8,750 9,600 67,015 85,365 6.19 There are two main types of soils in tidal areas: peat soils and unripe clays and each requires a different system of soil and water manage- ment. Some unripe clays, which are stable when submerged, contain pyrites, which oxidize when drained causing a high level of acidity in the soils. This lowers the pH, increases aluminum toxicity and may be detrimental to plants. Such acidification can generally be prevented or controlled by maintaining a high water table or by running fresh water over the land to leach out the toxic elements. To date, very conservative cultivation practices, adopted to avoid exposing potentially acid soils, combined with high rainfall have kept acidity within manageable limits in most swamp reclamation areas. 6.20 Deep peats are a more serious impediment to agriculture, and due to micronutrient deficiencies, among other problems, rice cannot ordinarily be grown in peat soils over one meter thick. Even in shallower areas, irrever- sibly dried peats can cause total crop failure and some transmigrants have experienced problems with peat fires. Soil salinity may also have an adverse effect on crops and in many areas measures must be taken to contain or prevent saltwater intrusion. - 116 - 6.21 Agricultural Development. Because of the technical skills required for large-scale drainage works, the identification and initial development of tidal lands is the responsibility of a Directorate within the Ministry of Public Works (MPW). As the first step in opening tidally influenced lands, this Directorate digs a large navigation canal which permits access by barges of up to 50 tons and allows water to flow freely out to the river or sea. Thereafter, primary and secondary canals are constructed. Tertiary canals are established by the farmers in their fields. In some areas the construction of drains has lowered the water table and caused oxidation of acid sulphate soils. To avoid this, simple hydraulic structures have been recommended to maintain the water table, prevent oxidation of the soils, and prevent salt- water intrusion and flooding, but water structures which have been constructed are poorly maintained by the farmers and may be totally neglected. More com- plicated gates and polders are not economically justified. 6.22 Until the beginning of Repelita III, MPW also undertook both research and extension in swamp reclamation areas. These functions were transferred to the Ministry of Agriculture in 1978, but systematic research in swamplands by AARD only began in 1984. Farming practices developed by MPW involve very shallow cultivation to avoid exposing acid sulphate soils. The use of plows and draft animals has been discouraged, even though areas of nonacid soils might potentially profit from puddling by water buffalo or cattle. Where soils permit and labor is available, farmers may elevate the soil to create alternate high and low areas in the fields to be used for rice intercropped with vegetables such as eggplants, fruit trees and coconut. Produce is transported by foot or bicycle to the primary canal where it is rafted to local market centers. Since most sites in Sumatra and many in Kalimantan are over 10 hours by boat to major markets, transport costs and consumer prices are higher, on average, in tidal areas than in upland sites. Experience to Date 6.23 General. The First World Bank Transmigration Sector Review conclu- ded that the income of migrants with two hectares of wet rice justified swamp reclamation, if deep peats and acid sulphate soils could be avoided and if drainage canals and other infrastructure could be provided at relatively low cost. This conclusion was based on studies which showed that many families in Upang Delta annually produced an average of two to three tons of rice/family, well beyond their subsistence needs. The recent BPS income survey indicates that Upang settlers are still doing well and produced 1.7 tons of rice/family in 1984. Repelita II tidal settlements generally produced 1.7 tons, but Repelita III tidal sites produced only an average of 0.6 tons of paddy/family, about equal to older upland sites. Within sites, yields are also very erratic and many recent settlers produce less than their subsistence needs. 6.24 Some hypotheses are possible for the lower yields in Repelita III areas. First, Repelita II sites may have been selected from better areas. Although most sites have been identified under plan-as-you-proceed procedures, nearly seven times as many people were settled in swamp areas in Repelita III as Repelita II and this may mean that some recent sites are less favorably located. Government officials acknowledge that high Repelita III targets led to rapid settlement prior to thorough soil studies and that some families - 117 - settled on deep peats or areas subject to flooding have subsequently had to be moved. Master plan studies currently under way should help to identify land suitable for settlement, but the wide variation within sites and the need to link information on soil type to on-farm management, suggests that there is also a need for careful land evaluation after clearing, and for farm level studies of settled areas on which recommendations about farm practices can be based. 6.25 Two other factors that may be affecting yields are inappropriate farm layout, and partial clearing of farmland. The shift from linear to nucleated settlements appears to have had a more important impact on tidal sites than upland sites. The BPS data show annual incomes from food crops of Rp 177,000 in linear swamp settlements and Rp 93,600 in nucleated settle- ments. Most of the nucleated settlements are newer and also have other problems. However, the data suggest that a return to linear settlements should be considered. Some Repelita III sites, including Bank-assisted sites, also left uncleared land within the agricultural area on the assumption that land clearing could best be carried out by the migrants themselves. It is now evident that this practice has contributed to pest problems. 6.26 Several other factors affect productivity in all swamp sites: (a) Transport. A major problem in swamp areas has been maintenance of the navigation and primary canals to permit passage of boats into and out of sites. MPW has recognized this problem and recently diverted funds from new settlement to canal maintenance. (b) Market Limitations. In the early years of swamp settlement when there was a strong demand for rice and prices were relatively high, middlemen actively sought rice for purchase in tidal sites. Recently, with high transport costs and falling prices, this trade has declined. While potentially affecting all transmigration sites, this phenomenon is likely to be particularly serious in tidal sites that are not highly diversified and that appear to have limited off- farm work. (c) Health. Health problems are more serious in tidal than in upland sites for two reasons: first, the widespread incidence of malaria; and second, problems of potable water supply, caused by salt water intrusion which makes canal water undrinkable in many areas in the dry season. Rainwater collectors have been distributed, but are not sufficient for family needs and many families buy water from traders. To reduce these problems, specific efforts need to be made in tidal areas by MOT and the Ministry of Health. (d) Management. Establishing settlements in swamp reclamation areas requires the close coordination of the Ministry of Public Works and the Ministry of Transmigration. This has been difficult in the past because of very different institutional arrangements in the two agencies. Therefore, to improve implementation it is important that one agency have an overview of the entire operation. Consideration might also be given to bringing land development, settlement and - 118 - agricultural development under a single management entity in swamp development projects. 6.27 Bank-assisted projects. A July 1985 review of the first Bank- assisted swamp reclamation project in Karang Agung, South Sumatra (Loan 1958-IND) makes specific recommendations for addressing the problems of tidal sites. To reduce severe rat and pig problems, the review recommends that the farm layout be changed, that a protective perimeter (a moat or fence) be established around the area, and that all land within this perimeter be completely cleared, that better coordination be exercised in planting and harvesting and that extension officials be trained to coordinate intensive campaigns for pest control. The report also recommends an adjustment of farm layout to include all cultivated areas in one plot near the house and it suggests that this is the single most important factor in stimulating land reclamation, increasing labor availability, and improving farm management and pest control. 6.28 With respect to soil and water management, the report recommends improved mapping of the area once the land is opened and closer attention to soil differences. Recommendations are also made to improve farm practices including early on-farm diversification based on poultry and small livestock, aquaculture and coconuts; improved training and extension for farmers; esta- blishment of seed farms; and improved marketing infrastructure, including adequate markets, docks and transport facilities. Some of these measures have been put into place and yields in 1985/86 are expected to improve. Conclusions and Recommendations 6.29 Despite the serious problems now experienced in swamp reclamation areas, most technical specialists agree that tidal areas have good prospects for growing rice and for diversification into other crops. To realize that potential, however, a concerted effort is required to systematically address the problems of these sites. To do this, Government should take the necessary steps to: (a) improve past practices of site selection and evaluation, taking into account distances from markets and the availability of off-farm employment as well as technical parameters; (b) promote judicious drainage and proper land development; (c) establish a mechanism for reviewing soil properties after clearing and settlement in order to provide recommendations for on-farm management; (d) review the linear settlement model and decide on optimal farm layout; (e) provide adequate cleared land to facilitate quick startup and discourage pests; (f) improve agricultural support services; - 119 - (g) develop an intensive pest management strategy, including agreement on the type of physical barriers, land clearing practices and organizational arrangements best suited to pest control; (h) increase attention to problems of potable water by (MOT) and of malaria control (Department of Health); (i) pay close attention to on-site transportation, particularly from fields to secondary canals; and (j) improve management and agency coordination. 6.30 Large-scale new settlement in tidal areas does not appear justified until current problems leading to low production and low incomes can be over- come. Government officials have recognized this problem and they have redirected funds to upgrading and second-stage development. There is also concern about the viability of communities based primarily on rice production, and every effort should be made to promote on-farm diversification including, where possible, the introduction of perennial crops. On the other hand, the ready availability of tidally influenced lands and the good yields in successful sites, argue strongly that priority should be given to resolution of problems in this farm model. C. Tree Crop Development 6.31 Most tree crop development in transmigration sites is relatively new, so experience is limited, but barring a prolonged downturn in world commodity prices the outlook for settlers on such schemes is promising. Agricultural incomes on tree crop schemes are projected to be about US$1,500/family in year 10, Nucleus Estate and Smallholder (NES) rubber and oil palm and Project Management Unit (PMU)-based rubber tree crop projects have an internal rate of return which exceeds an opportunity cost of capital of 10%, and tree crop projects with relatively high incomes offer scope for cost recovery. Not only are tree crops critical for proposed pioneer settle- ments in remote areas where both markets for food crops and off-farm work are limited, but demand for tree crops is also very large in existing transmigra- tion sites. 6.32 The Government recognizes the advantages of tree crops and in recent years has taken a number of steps to promote the development of tree crops for transmigrants (Chapter I). However, the impact of tree crop development on the transmigration program to date has been small. Only about 7,400 transmi- grant families were settled on NES/PIR projects in Repelita III (about 2% of the total moved) and an additional 3,800 families were settled in 1984/85, for a total of about 11,200 transmigrant families settled under the NES/PIR program through 1984/85 (Table 6.4). During this same period about 20,500 local families were settled on NES/PIR schemes. Preliminary figures suggest that an additional 20,670 transmigrant families were settled in NES/PIR projects in 1985/86 for a total of about 32,000 families in such schemes in mid 1986. - 120 - Table 6.4: TRANSMIGRANTS SETTLED ON NES/PIR SCHEMES (Households) Actual Projected /a Scheme 1980-1984 1984/85 1985/86 1986/87 NES Rubber /b 891 2,446 6,250 5,900 NES Oil Palm - - 6,605 4,600 NES Coconut - - 3,749 3,377 Subtotal 891 2,446 16,604 13,877 PIR Khusus Rubber /c 5,207 614 7,925 6,708 PIR Khusus Oil Palm 1,294 796 8,268 4,600 Subtotal 6,501 1,410 16,193 11,308 Total 7,392 3,856 32,797/d 25,185 /a Includes both local people and transmigrants (NES). 7b Nucleus Estate and Smallholder Projects are externally assisted. 7 Perkebunan Inti Rakyat (PIR) Projects are domestically financed. / Of these, 20,700 families were settled. Tree Crop Establishment 6.33 The main constraint to expanded tree crop development, whether for local people or transmigrants, is institutional capacity. Tree crops in Indonesia are established in several ways. (a) Spontaneous Planting. Indonesia has about 6 million ha of rubber, oil palm and coconut, of which about half was planted by small- holders using unselected planting material and family labor. Yields and returns to labor on such stands are low, and in many areas rubber tapping and copra production are neglected at current commodity prices. (b) Project Management Units. To improve the quality of smallholder planting, about 870 Project Management Units (PMUs) have been formed by the Directorate General of Estates. About half of these PMUs are for rubber and coconut development, and the other half are for minor tree crops (coffee, cloves), or cash crops such as sugar or cotton. PMUs provide planting materials, inputs and extension to local farm- ers as well as compensation for all or part of the cost of farmer labor. Funds for this purpose are provided as a credit to the farmer to be repaid when trees are mature. There have been no PMU programs directed specifically at transmigrants to date, although local smallholders may include former migrants. (c) Public Sector Estates. Tree crops are also established by publicly owned estates (PTPs) under the NES/PIR programs. In these programs, - 121 - estates establish a nucleus of estate-owned trees and then plant additional areas for smallholders. The PTPs construct roads, houses and community facilities and they block-plant tree crops using settler labor. When mature, trees are divided into 2 ha plots and distributed to the farmers who obtain credit for both the cost of tree crop establishment and housing. (d) Private Estates. There are about 850 private estate companies with rubber as their main crop and 18 are planted to oil palm. Most rubber plantations are small and need rehabilitation. Only 3% of the 3 million ha under coconut is managed by private estates. Private estates have not yet been used for smallholder development. During Repelita III about 120,000 ha of tree crops (22%) were planted by the PTPs and about 430,000 ha (78%) were planted by PMUs. The area planted under each program is shown in Annex 6. Oil palm accounted for 6% of the area plan- ted, rubber for 34%, and coconut for 60%. 6.34 In 1984, the World Bank undertook a study of the prospects for Indonesia's major tree crops (rubber, oil palm and coconut) during Repelita IV. This Sector Review (Report 5318-IND, published April 15, 1985) endorsed an official (PTP and PMU) planting program of about a million ha for Repelita IV including up to 770,000 ha for smallholders. The report concluded that Indonesian rubber would remain competitive on the world market and that oil palm and coconut production should be increased, both to meet the domestic demand for edible oils and to earn foreign exchange. The cost of the proposed program was estimated at US$2.9 billion (1984) for Repelita IV and to finance a program of this scale, the report recommended: (a) that the provision of funds for smallholder credit be shifted from the development budget to funds mobilized by the banking system; (b) that the liquidity of the estates be improved so they could prefinance their own development from borrowing; (c) that ways be sought to decrease the cost of tree crop establishment; and (d) that the private sector be involved in estate crop development. 6.35 Implementation constraints were also reviewed in detail in the Sector Review. The report concluded that the rapid buildup of the NES programs in Repelita III and, more recently, the increasing demands for ever larger PIR programs had overstretched the management, financial and technical capacity of the PTPs and resulted in a downturn in the quality of plantings, a reduction in investments for maintenance and replanting on existing estates, and a deterioration in the financial condition of the PTPs. The rapid expan- sion of the PMU program without commensurate manpower development had also resulted in poor quality plantings, financial problems and target short- falls. An action plan to address these issues is being developed by the Government, but implementation will understandably be slow, and any recommen- dations intended to expand tree crop establishment for transmigration must be carefully formulated to avoid exacerbating problems in the sector. Tree Crops for Transmigrants 6.36 Existing Programs. Experience gained during Repelita III provides a number of useful observations on ongoing PTP and PMU programs and their - 122 - complementary roles in tree crop development. The use of public sector estates is preferred when new infrastructure must be developed and large areas are to be block-planted, but PTPs are not well suited to the development of scattered fields or to the replanting of existing smallholdings. PMUs, on the other hand, are better suited to the establishment of tree crops on existing smallholder lots, and they could potentially be used in conjunction with settlement by the MOT to develop tree crops in new sites. 6.37 The choice of commodity also affects the management system required. Establishment of oil palm is best done by PTPs or private estates because the development of associated processing facilities and regular collection requires managerial skills and financial capacity that cannot easily be provided by smallholders or by PMUs. Estates have no technical pro- blems establishing rubber, but rubber is less costly in financial terms when established by PMUs since they pay incentive payments to the farmer-owner which are less than the full value of his labor. The better PMUs are tech- nically able to produce planting material and to assist with rubber estab- lishment, and PMUs are particularly well suited to the establishment of coco- nut which is the least technically demanding smallholder crop. These commod- ity distinctions are already evident from data for Repelita III. During that five-year period, all smallholder oil palm was established by PTPs, rubber establishment was split almost evenly between PTPs and PMUs, and virtually all smallholder coconut was established by PMUs. 6.38 New Programs. Since early 1984, three programs have been proposed to increase the number of transmigrant families which can be settled with tree crops. (a) Accelerated PIR. Under this program, five of the most capable PTPs are to get funds from the development budget to accelerate the establishment of oil palm for transmigrants. The MOT would be responsible for selection, settlement and social development, and the PTPs for housing, infrastructure and tree crop establishment (as under current NES/PIR projects). Settlers would receive the initial settlement package on grant and obtain credit for tree crop estab- lishment. Under the initial proposal, estates were to plant 500,000 ha in 5 years. This target has since been reduced to 300,000 ha, but given existing demands on the estates, the Bank estimated in late 1985 that only 65,000-100,000 ha could be planted in the proposed time frame. This number may be further reduced if commodity prices remain low, a factor which would reduce the PTP's incentive for new investment and make financing more difficult. (b) PMU Programs. The Government has recently shown an interest in using PMUs to establish tree crops on new sites and a sizeable nursery program is currently being considered to produce planting material for new and existing sites. If linked with new settle- ments, the DGE would have an input into settlement design and the MOT would be responsible for settlement planning, land development for the houselot and food crop areas, the selection and transfer of migrants, and for their initial support. In the second year after arrival, PMUs would assist the farmers in opening and planting one - 123 - ~, Transmigrant farmers harvesting oil palm fruits planted for them by a Government owned estate. hectare of tree crops and in the third year they would help plant a second hectare. Introduction of such a model for rubber and coco- nuts and possibly other crops, would significantly improve the income prospects of new settlers. (c) Private Sector Tree Crop Development. A Ministerial Decree was issued in June 1985 which provided a framework for private sector investment in tree crops. Each investor is required by the Government to develop 3 ha of smallholder land for each 2 ha of nucleus estate and to prefinance smallholder development costs until the trees are transferred to the participating farmer, at which time the investment is to be repaid by a state bank. The principal concerns of the investors are export taxes, which have been relatively volatile in the last several years; regulations which require the sale of palm oil at a controlled domestic price; high interest rates for borrowed funds; and the possibility that conversion of the smallholder area and repayment of funds pre- financed by the estate will be delayed. - 124 - 6.39 To improve the financial returns to private sector investment, the Bank has recommended an exemption from export taxes for new private sector investors (windfall profits can be captured through an existing income tax); freedom to sell at international prices; relaxation of the planting ratio of smallholders/estates; and relief from obligation to prefinance smallholder development. 6.40 The DGE target calls for the private sector to plant 360,000 ha of oil palm and settle 180,000 transmigrant families from 1984/85 to 1988/89. The Bank estimates that the maximum likely to be planted is about 100,000 ha over the next five years and even this number is unlikely to be achieved if palm oil prices remain at the low levels of mid-1986. 6.41 Second-Stage Development. To date very little estate or PMU capacity has been focused on the upgrading of transmigration sites. One exception is the Bank-assisted NES III project (Loan 1751-IND) which uses PTP VI to plant rubber in Rimbobujang, an ongoing transmigration settlement in Jambi Province. As noted, the Government and Bank have also developed a proposal to establish nurseries for rubber and coconut which could be used for tree crop establishment, under PMU direction, in existing transmigration sites, and this should receive strong support. 6.42 One problem encountered in planning for second-stage development is land fragmentation and land shortages in transmigration sites. At least 1.5- 2.0 ha of tree crops/family are needed to improve family incomes, permit cost recovery, and, in the case of rubber, to discourage overtapping. But in many sites, reserve land has not been allocated, and in some places it has been settled by spontaneous migrants and local people. In some areas, reserve land is too far away to be developed and maintained, while in others only 1.25 ha or at most 2.0 ha are available for both food crops and tree crops. This points to the need for ensuring that future land allocation is sufficient for second-stage development. On existing holdings of 2.0 ha, the DGE has proposed that the reserve land of 0.75 ha be opened and cleared in year one and that 0.75 ha of cleared food crop land be planted to tree crops and inter- cropped with rice three years later. In year six, when intercropping ceases, the first hectare of rubber would be tapped and food crop production would be limited to a 0.5-ha garden lot. Under these circumstances, a portion of family food would be purchased. The use of this model should be supported. The Proposed Tree Crop Program in Support of Transmigration 6.43 Table 6.5 gives the Bank projections on the maximum number of hectares which could potentially be planted for smallholders over the five- year period 1986/87 to 1990/91, a period which covers the remainder of Repelita IV and the first years of Repelita V. The table indicates that in the absence of financial constraints, an estimated 250,000 ha of major tree crops could be planted under ongoing NES/PIR programs during Repelita IV and an additional 64,000 ha could potentially be established under the two proposed programs (accelerated oil palm and smallholder tree crop development by private estates). About 250,000 ha of tree crops could also be established by PMUs during the remainder of Repelita IV, if adequate funds are available. Table 6.5: BANK PROJECTITNS FOR NEW PLANTING, REPLANTING AND REHABILITATION OF TREE CROPS FOR SMALLHOLDERS /a (ha planted) Repelita IV Repelita V Program Bank projections Total Families/ Planting program 1984/85 1985/86 1986/87 1987/88 1988/89 Total ha Trans families /b 1989/90 1990/91 year Ongoing Estate Projects A. NES Projects NES - Oil palm 13,200 9,200 10,000 10,000 10(,0o 52,400 60 15,720 15,000 15,000 4,500 NES - Rubber 12,500 11,800 9,000 7,000 6,000 46,300 50 11,550 6,000 6,000 1,500 NES/Trans - Coconut 7,500 6,800 6,000 6,000 - 26,300 100 13,150 - - - Subtotal NES 33,200 27,800 25,000 23,0no 16,000 125,000 New 40,420 40,000 45,000 6,000 B. PIR Projects PIR Khusus - Oil palm 16,500 9,200 10,000 10,000 10,000 55,700 80 22,280 15,000 15,000 6,000 PIR Khusus - Rubber 15,850 13,400 6,000 8,000 11,000 54,250 80 21,700 15,000 18,000 7,200 PIR Khusus - Rubber/ oil palm 5,250 1,300 3,000 3,000 3,00(1 15,500 0 - 3,000 3,000 - Subtotal PIR 37,600 23,900 19,000 21,000 24,000 125,500 New 43,980 33,000 36,000 13,200 Total A + B 70,800 51,700 44,000 44,000 40,000 250,500 New 84,400 73,000 81,000 19,200 Proposed Estate Projects Ut PIR Akselerasi - Oil palm - - 10,000 12,000 12,000 34,000 80 13,600 15,000 15,000 6,000 PIR Swasta - Oil palm - - - 10,000 20,000 30,000 80 12,000 40,000 50,000 20,000 Subtotal New PIR _ _ 10,000 22,000 32,0(0 64,000 New 25,600 128,000 146,000 26,200 Total A + B + C 70,800 51,700 54,00( 66,000 72,000 314,500 New 110,000 128,000 146,000 45,200 Ongoing PMU Projects SRDP - Rubber 4,772 5,815 7,000 9,000 12,000 38,587 50 9,650 18,000 25,000 6,250 SCDP - Coconut 4,763 11,240 8,000 3,000 8,000 35,003 50 8,75( 10,000 15,000 3,750 PMU - Rubber - 23,293 7,000 6,000 6,0()0 42,293 50 10,550 5,000 5,000 1,250 PMU - Coconut - 31,277 30,000 35,000 40,000 136,277 50 34,050 40,000 45,000 11,250 Subtotal PMU 9,535 71,625 52,000 53,000 66,000 252,160 Old 63,000 73,000 90,00(1 22,500 Total A + B + C + 0 80,335 123,325 106,000 119,000 138,000 566,660 - 173,000 201,0no 236,000 67,700 /a Do not reflect recent budget constraints which may cause a downward revision of targets. lb At 2 ha/family. - 126 - 6.44 Standards already set for ongoing NES/PIR programs would permit the settlement of transmigrants on 50-60% of the area planted for smallholders, and about 80% of land to be established under new PIR programs is earmarked for transmigrants. Thus, Bank projections initially suggested that about 75,000-85,000 transmigrant families could be settled under ongoing NES/PIR programs in Repelita IV and an additional 25,000 families could potentially be settled if the accelerated oil palm and private estate programs moved forward quickly. In total, therefore, the Bank estimated that Government could settle some 75,000-100,000 families with tree crops in Repelita IV and that the number settled could potentially rise to 40,000-50,000 families annually by mid-Repelita V. These figures are now thought to be high in view of budget constraints. 6.45 These figures also illustrate the trade-off in transmigration between the quality of settlements and the number of families that can be settled under the sponsored program. The Government originally proposed to settle 750,000 migrants in Repelita IV of whom about 400,000 would be moved with full sponsorship; of these about 100,000 families were to be settled on tree crop sites. This implies that all other families were to be settled with food crops. To improve the quality of the program by relying principally on new settlement with tree crops, the Government would have to significantly curtail the targets for sponsored settlement. Issues and Recommendations 6.46 New Settlement vs. Second-Stage Development. If all future tree crop development by estates and PMUs were used for new settlement, higher targets could be met. However, this is neither feasible nor desirable since (a) there is a strong need for replanting and new planting among indigenous smallholders who are the backbone of the traditional tree crop sector; (b) providing tree crops to local people has long been seen as a way of shifting from extensive to intensive agriculture and freeing land for new settlement; (c) there is an urgent need to establish tree crops for existing migrants in order to promote growth and diversification; and (d) second-stage development is generally more economically attractive than new settlement since previous expenditures are treated as sunk costs. For these reasons, proposed PMU programs to expand new planting or replanting on existing farms, including those of both local smallholders and transmigrants, are appropriate. 6.47 Program Scale and Managerial Emphasis. The projected planting program of about 570,000 ha (Table 6.5) is only one half the government target and only 75% of that initially endorsed in the Tree Crops Sector Report. However, most observers feel that pressure to exceed the figures shown would result in poor planting and undue managerial and financial strain and should not be encouraged. The capacity of the private sector has not been tested, and with low commodity prices the figures provided are thought to be optimistic for Repelita IV. Therefore, the main opportunity for expanding rubber and coconuts is through the PMU program and this depends on the speed with which PMUs can be rationalized and staffed with adequately trained personnel, operating under adequate supervision from the DGE. This particularly applies to the government-financed tree-crop program (PRPTE) which has been poorly administered in the past. - 127 - 6.48 There are a number of advantages to PMU operations. First, exter- nally assisted programs have demonstrated that plantings of reasonable quality can be established even if the farmer is reimbursed for less than the full value of his labor. Incentive payments reduce government financial outlays, limit farmer indebtedness, and often increase his commitment to the program and to the resource he has obtained. In addition, since PMUs are generally run by agricultural high school graduates, overheads are low when compared to those of the PTPs, and PMU programs can potentially be expanded more rapidly. The greatest scope for expansion in the tree crop program appears to be in the establishment of coconuts by PMUs and secondarily in the expansion of rubber areas by PMUs or estates. 6.49 Management and Coordination. In March 1986, a Ministerial Decree was issued which gives the budget and authority for tree crop development to the Directorate General of Estates (DGE), establishes new arrangements for financing both credit and noncredit components and provides incentives for development to public sector estates. This decree lays the groundwork for future estate crop development for transmigrants and includes a coordination team headed by the Junior Minister for Tree Crops and including the heads of the two Directorate Generals from the MOT, among others. Full support from the MOT will be necessary to ensure a strong and growing role for tree crops in transmigration schemes. (These issues are reviewed in Chapter VIII.) Priorities 6.50 Given that there is a sound rationale for increasing tree crop production in Indonesia, and that management and financial capacity are con- strained and likely to continue to be so under the current austerity condi- tions, attention must be given to priorities in tree crop development. New development should, other things being equal, take place as close as possible to existing infrastructure rather than in remote areas, and make use of exist- ing grasslands (which are cheaper to clear) rather than forest. To obtain the right to use such land, local people must have access to some of the benefits of tree crop schemes and, for this reason, the distinction between estate crop projects for locals and for transmigrants should be abolished, -so that each scheme incorporates both types of settlers. 6.51 There is also scope for associating new settlements with PMU development of tree crops but the benefits of this must be weighed against the economic advantages of providing tree crops to existing farmers, whether transmigrants or iocal families. As noted in Chapter III, second-stage development with tree crops has the highest rate of return of the models considered and the lowest financial cost. Such development would also have the effect of increasing regional production and spontaneous migration. Caution should be exercised, however, in assuming that estate and/or PMU capacity to plant tree crops can be rapidly expanded. This must be a slow and systematic process done in the context of improving implementation capacity of the tree crop sector as a whole. - 128 - D. Summary and Conclusions 6.52 A major finding of this study is that food crop production in upland sites appears to be carried out mainly to meet the subsistence needs of the transmigrant community. Furthermore, the report concludes that this is due to agroeconomic conditions, marketing and manpower constraints which are diffi- cult to address. To improve income prospects in existing sites, priority should be given to improving agricultural supporting services (research, extension, input supply and credit) and to second-stage development, including adequate maintenance of existing roads. 6.53 Tidal sites have better prospects for food crop production than upland sites, but incomes in ReDelita III tidal sites are very low, due appar- ently to inadequate site selection, inappropriate farm layout, and a relative scarcity of off-farm work. For this reason, the report endorses the Government's reduction in targets for tidal areas and it recommends a shift in farm layout and a concerted effort at pest control to see whether current problems can be overcome. If so, and if settlement is to continue in tidal areas, more attention must be given to site evaluation, soil surveys and site planning. Site selection should be placed in a regional development framework which takes account of markets and employment prospects, management should be better integrated, and a major effort is needed to address problems of water supply and malaria control. 6.54 Tree crop schemes entail less risk to farmers, and provide higher incomes and better economic returns. The report therefore supports Government efforts to mobilize the estate sector for new settlement, and it recommends a strong effort to expand PMU capacity to plant for existing smallholders. Since coconuts are one of the easiest commodities for PMUs to establish, there is a need to expand PMU programs to establish coconuts for migrants in coastal upland and tidal sites. The declining trend in commodity prices for tree crops is a matter of considerable concern. However, even with lower price projections, economic returns are equal to or above 10% for rubber and palm oil. To protect farmers from the effects of periodic low prices, food crop cultivation for subsistence should be encouraged. 6.55 The foregoing analysis suggests, however, that there is a trade-off between the number and the quality of settlements. To maximize migrant welfare by shifting to cash crop or tree crop development in the future, the Government would have to limit the amount of new settlement. If Government chooses to continue large scale settlement and does not develop a mechanism for second-stage development, settlements would be likely to remain at subsistence levels and contribute only marginally to regional growth. Should market prospects for tree crops deteriorate, the Government will be faced with very difficult choices. - 129 - VII. OTHER FACTORS AFFECTING THE SCALE OF THE FUTURE PROGRAM: LAND AVAILABILITY, FINANCIAL CONSIDERATIONS, AND SPONTANEOUS MIGRATION A. Land Availability Prospects for Settlement on Upland Sites 1/ 7.01 Sumatra. Sumatra can be divided into those provinces where the possibilities for large-scale settlement were almost fully exhausted by the end of Repelita III (Aceh, North Sumatra, West Sumatra, Bengkulu and Lampung) and those where new large-scale settlements will be virtually completed in the course of Repelita IV (Riau, Jambi, and South Sumatra). Thereafter, new settlement is expected to be largely through tree crop schemes, fill-in programs assisted by the MOT, and spontaneous movement. 7.02 The number of families which can be settled on tree crop schemes in Sumatra will depend partly on institutional capacity (Chapter VI) and partly on the ability to persuade local smallholders to upgrade and consolidate existing holdings, thus making way for partial settlement by transmigrants. NES schemes which benefit local smallholders and also include transmigrants have been generally successful in obtaining land, but tree crop schemes intended solely for transmigrants may be less attractive to local people and less helpful as a mechanism for identifying land. Similarly, fill-in programs, which involve the settlement of transmigrants on empty lands within or adjacent to transmigration sites, are likely to be small unless a mechanism for compensation is developed for local smallholders. 7.03 Although there are limitations on further sponsored settlement, Sumatran provinces are expected to receive the bulk of spontaneous migrants in the foreseeable future. Successful settlement of these migrants and protec- tion of land unsuited to agriculture will depend on the development of mechanisms for (a) land registration, (b) land purchase, and (c) demarcation and protection of areas unsuitable for settlement. These matters should receive high priority from the GOI and are covered in more detail in subsequent sections of this report. 7.04 Kalimantan. Estimating the settlement potential of Kalimantan is more complex. The areas suitable for agriculture in the small province of South Kalimantan are almost fully utilized and continued large-scale settle- ment in West Kalimantan is possible only if indigenous smallholders agree to consolidate areas of "wild" rubber or to relinquish land under shifting culti- vation. They would be more likely to do this if given financial compensation or the opportunity to obtain high-yielding rubber through estate crop 1/ Most of the information in this section is drawn from a study to assess land resources, under way with assistance from the British Overseas Development Authority (ODA). - 130 - programs. East Kalimantan and Central Kalimantan have considerable potential for tree crop development, but most land is in production forest, infrastruc- ture is poorly developed and routine services are limited. Therefore, given the long period required to develop infrastructure and services, and the high cost of doing so, potential future settlement in Kalimantan would not be expected to exceed past levels of 18,000-20,000 families per year and these figures could be reached only if some land in production forest were reclassified. 7.05 Forestry constraints are well illustrated in Central Kalimantan. Because of its large land areas, this province was believed to have the greatest potential for settlement with tree crops. For this reason it was the first area reviewed in detail by the Land Resources Development team from ODA. Over half the province is unsuitable for any type of agriculture due to infertile sand terraces, deep peat, steep-sided hills or mountains. The area suitable for agriculture is almost entirely of low fertility soil and most (80%) would require high technology to develop (such as high-input tree crops). In addition, virtually all land is forested. Only 9% of Central Kalimantan is outside forest boundaries, 28% is in conversion forest, 40% in production forest, and the remainder in protected areas. Table 7.1 summarizes the settlement potential of Central Kalimantan within forest class. Table 7.1: ESTIMATED AREA, CARRYING CAPACITY AND FORESTRY STATUS OF POTENTIAL SITES IN CENTRAL KALIMANTAN Settlement Potential /a Total Recommended No. of (estimated families) families land use sites (A) (B) (A + B) Upland food crops /b 23 4,500 26,700 31,200 Wetland food crops 7b 7 8,990 2,610 11,600 Nucleus estate projects 32 34,850 77,830 112,680 Commercial estates (no food crop component) 19 7,930 68,890 76,820 Total 81 56,270 176,030 232,300 /a (A) If settlement is confined to conversion forest or nonforest areas. (B) If production forest suitable for agriculture is reclassified. /b The introduction of tree crops is also recommended in these models. Source: Review of Phase lB Results, Central Kalimantan (RePPProT), June 1985. 7.06 Several points are evident from Table 7.1. First, only about 43,000 families could be settled on food crop land (20% of the total). Of that number, only 13,500 families could be settled if there were no settlement in - 131 - areas now classified as production forest. All told, about 50,000-60,000 families could be settled on land which is in conversion or unforested areas, while about four times that many could be settled if forest boundaries were changed. To do this, mechanisms for the review and reclassification of production forest would be required. 7.07 However, there are many reasons why the conversion of forests should be weighed carefully. Forests have commercial value and development in fores- ted areas is often not economically attractive if foregone uses (i.e., commer- cial forestry) are taken as project costs. Land clearing for settlement also reduces the area under forest cover and may lead to erosion and negative effects on the environment and wildlife. These trade-offs call for some guidelines for development. Since forests have ecological and commercial value, they should not be cut or converted unless the proposed development is economic and sustainable. Evidence from Chapter II suggests that food crop farm models are not economically attractive on soils such as those in Central Kalimantan and should not be contemplated in remote upland areas. Tree crop development, on the other hand, has better economic prospects and is less destructive to the environment. This does not mean that all land suitable for tree crops can or should be developed. Most inner Kalimantan areas will be expensive to develop and relatively unattractive to established estates. Therefore, in light of budgetary constraints, projected low commodity prices and marketing difficulties, remote areas in Kalimantan would have low priority for tree crop development. 7.08 Eastern Islands. The eastern islands have very limited prospects for new settlement. Sulawesi could absorb no more than 20,000 additional sponsored families and as the remaining available area is degraded uplands and ultra-basic rock unsuitable for agriculture, spontaneous movement into such areas would be limited. The Moluccas have little land available for settle- ment and are unlikely to attract large numbers of spontaneous transmigrants given their rather limited economic possibilities and the number of migrants from the eastern islands competing for jobs in those areas. 7.09 Irian Jaya. As noted earlier, Irian Jaya is the largest province in Indonesia and one of the least densely settled, averaging three people/km2. The indigenous population is located mainly in the fertile highlands, and some areas, particularly in the south around Merauke, are only sparsely settled due, in part, to malaria and to the limitations of traditional production systems. Although the Merauke area has considerable land suitable for settle- ment, there are physical, financial and social constraints associated with development in the area. 7.10 Physical development near Merauke will necessarily be slow because there is little existing infrastructure, and road construction is difficult due to the absence of suitable foundation material. There is insufficient local transport on rivers to permit large-scale development at present and the local labor supply is limited. For these reasons, development of the area is likely to be costly and marketing difficult. Therefore, future settlement could only be contemplated if planned in the context of an overall regional development program based on crops with good market prospects. - 132 - 7.11 To improve planning, the Bank is financing the preparation of a regional development plan in the Merauke area. This plan includes provisions for the development of economic activities, the demarcation of lands for local people and the provision of services to them. Concentrated development in an area such as Merauke is believed to be preferable to widespread scattered settlements, both from the point of view of planning and monitoring and of minimizing adverse effects on indigenous populations. However, financial and social constraints are likely to slow development in Merauke. 7.12 In general, the most important reasons for proceeding slowly in Irian Jaya are social and environmental. As noted in Chapter V, most Irianese follow traditional subsistence patterns and have limited contact with outsiders. To provide them with the benefits of development and to prevent demoralization and social tensions, change must be gradual and priority given to the needs of the local people. This is not consistent with rapid, large- scale settlement. The environment in Irian Jaya is also an important resource, and care must be given to the development of a comprehensive plan for demarcating and protecting valuable wilderness areas. Bank projections in light of such considerations place any future settlement for the province at less than 5,000 families/year. If such settlement were carefully managed and monitored, these settlement levels could be achieved with a minimum disruption to the indigenous population. This would, however, depend on strong resolve on the part of Government. Tidally Influenced Areas 7.13 As Table 7.2 indicates, a substantial portion of Indonesia's total land area is in coastal swamps. Since these areas are often malarial and generally cannot be cultivated unless drained, they are virtually unpopulated and for this reason attractive for settlement if malaria can be brought under control. The Repelita IV target originally called for the settlement of 75,000 families in tidal areas, but in 1985 this was reduced to about 37,000 families due to financial constraints and production problems. This level is now expected to be considerably lower. There were no plans to open swamplands in Irian Jaya in Repelita IV. Table 7.2: SWAMP AREAS IN INDONESIA ('000 ha) Swamps as % Island Swamp area Total area of total area Sumatra 13,211 47,360 28 Kalimantan 12,764 53,946 24 Irian Jaya 12,780 42,195 30 Sulawesi 469 18,204 3 Total 39,224 161,705 24 - 133 - 7.14 Since 1981, consultants have been carrying out a nationwide inventory of coastal and near-coastal swamplands. This study has surveyed some 25 million ha of swampland in Sumatra, Kalimantan and Irian Jaya. Of this, 5.5 million ha were classified as having settlement potential. The largest areas are in the coastal lowlands of southern Sumatra (1.3 million ha), West and Central Kalimantan (1.2 million ha), and Irian Jaya (2.8 million ha). This indicates that sufficient land would be available in the coastal plains to potentially settle 10,000 families/year through Repelita V and VI, if pest and production problems could be overcome, and if better microplanning were done to eliminate unsuitable soils within sites. Without these measures, future large-scale development should not be encouraged. 7.15 Mangrove areas are protected in Indonesia, but environmentalists are concerned about large-scale drainage in swamps in the absence of detailed stu- dies of the effects on soils, water, fish life and marine habitats, particu- larly from the acidification of water in drained sites. Wetlands are also important as wildlife habitat. As noted in Chapter V, environmental impact studies are included in both the First and Second Bank-assisted Swamp Reclama- tion Projects (Loans 1958 and 2431-IND), but more could be done to formalize institutional arrangements and provide adequate funding for studies and for wildlife management in tidally influenced areas, and to establish plans for development which indicate crucial areas for conservation and protection. Summary 7.16 In summary, land availability is a serious constraint to future settlement. Only a limited amount of land remaining in upland areas is suit- able for food crop production, but a rather large amount of outer island land is suitable for tree crops. In Sumatra and West Kalimantan the main limita- tion on tree crop development is the presence of low productivity "wild rubber" and upgrading and intensifying local smallholder agriculture would be a prerequisite to further sponsored settlement. In Central and East Kalimantan the introduction of tree crop schemes is limited by economic and forestry constraints, while in Irian Jaya the pace of future settlement is limited by the absence of existing infrastructure and by environmental and social considerations. Land per se is not an immediate constraint in swamp reclamation areas, but settlement in these areas should not be considered until production problems are overcome. 7.17 Province-by-province projections made prior to reductions in the settlement program in 1986 suggested that the maximum level for sponsored settlement, consistent with land availability and institutional capacity, was about 300,000 families in Repelita IV and V, declining in Repelita VI. Settlement levels are now expected to be about 200,000 families in Repelita IV. Future high levels of settlement could be achieved only if agreement were reached on forest classification and if the Government were in a position to bear the cost of settlement in increasingly remote sites. 7.18 The natural resource inventories already under way should permit the completion of master plans on a province-by-province basis by the end of Repelita IV and completion of this work should be given priority by the GOI. - 134 - Such plans would lead to a more realistic assessment of settlement potential for Repelita V and permit planning consistent with both economic development and environmental protection. A knowledge of the area under forest cover is also critical to policy formulation and further settlement planning. B. Spontaneous Migration 7.19 Promoting spontaneous migration is a major objective of the Govern- ment, and has gained increasing attention as financial resources have become more limited. Recognizing the importance of large-scale unassisted migration, a Directorate for Self-initiated (Swakarsa) Movement was set up in the former Directorate General of Transmigration in 1979 and its tasks were set out in a Ministerial Decree issued in 1982. Several partially assisted programs were envisioned, including the identification and settlement of spontaneous migrants already in transmigration sites. Although some spontaneous families were settled in existing areas, no program of partially assisted movement or of support for spontaneous migration was developed in Repelita III. However, the Repelita III target of 500,000 families was eventually achieved by including 169,500 "spontaneous" families in the total moved. Of these, about 30,000 families were moved with partial assistance and the remaining 140,000 moved without assistance, but were identified in sending or receiving areas (Table 7.3). It is important to emphasize that these families were identified to round out official targets, and the data collected from sending provinces is not a good indicator of the number of people who actually move. 7.20 Of the target of 750,000 families for Repelita IV, the Government initially intended that 315,000 would be settled in upland and tidal programs, 105,000 in tree crop programs and the remainder, 330,000 families, would be settled on various partially assisted or unassisted programs. To ensure that unassisted movers would be counted, Government also required that first-time migrants register before they move. The Process of Spontaneous Movement 7.21 In the past, the Government has been limited in its ability to identify and generate the policies needed to stimulate spontaneous migration. To understand what policies are needed, it is necessary to understand both the process of spontaneous movement and the constraints to settlement. Spontane- ous migrants moving outside of government-assisted programs must ensure that they have adequate funds for transport and initial subsistence and that they have reasonable certainty of a cheap place to stay and employment in the destination area. Very poor people cannot afford to move on their own, even to areas with good employment prospects. Indeed, the very poorest people seldom, if ever, move without outside help, and virtually all spontaneous migration occurs through pre-existing chains of social support. When labor shortages arise in transmigration sites, earlier settlers alert friends and relatives, encourage their immigration and assist in their initial support. It follows that settlers in pioneer areas need the most assistance from Government (because they do not have family and friends to help them), and that if employment is generated in the receiving area, subsequent migrants will require less and less assistance. - 135 - Table 7.3: BREAKDOWN OF SELF-INITIATED (SWAKARSA) FAMILIES COUNTED IN REPELITA III With Assistance Moved with assistance from provincial governments 2,675 Other 27,875 Subtotal 30,550 Without Assistance Relatives in projects 46,132 Those obtaining identification forms in sending areas 33,293 Settled in projects 4,877 PIR projects /a 4,264 Unexplained 50,381 Subtotal 138,947 Total 169,497 /a These families obtain considerable assistance and should be regarded as sponsored settlers. Source: Joan Harjono from MOT statements on Swakarsa transmigration issued in 1984. 7.22 This suggests that the benefit package should be diversified by region, remoteness, and stage in the settlement process. For example, the demand for land is so strong in Sumatra that virtually all land identified under the fill-in (sisipan) program could be settled if infrastructure and access were provided; but settlement in the more remote areas of Kalimantan and Irian Jaya would require substantial government investment and support. Conversely, given a much reduced sponsored program, relatively little spon- taneous movement is expected to occur in Kalimantan and Irian Jaya. Constraints to Spontaneous Movement 7.23 Once spontaneous migrants are in the receiving area, the major factor limiting settlement is the inability to find appropriate land and obtain legal rights to it. This problem, in turn, is due to (a) traditional systems of land tenure in the outer islands, (b) difficulties in land registration, (c) problems in providing compensation to land owners, and (d) Government policies which have encouraged complete occupation of available sites to meet settlement targets. These points are expanded briefly below. - 36 - Local people in a spontaneous settlement along the trans-Sumatra highway. (a) Land Tenure. Land rights in rural areas of the outer islands are generally governed by customary law (adat). Under customary law, authority over land may reside in extended families or local territorial groups, and individual smallholders are seldom in a position to transfer the ownership of their land. If a spontaneous migrant wants land, he does what is called ganti-rugi -- compensates for loss, either to the individual or group. Compensation may be for the loss of the right to use the land or the loss of productive trees, but it is not generally for the land itself. Therefore, if the land improves in value, as it does with increased immigration or the planting of perennial crops, migrants may be requested to pay additional money or to return the land. They may even be subject to counter claims that the person who received the initial payments was not the person who had the right to do so. For this reason, many people are unwilling or unable to move without government assistance in finding land. - 137 - (b) Land Registration. It is possible to legally transfer land title in the outer islands, but this entails official surveys and complex registration procedures which are costly and difficult to arrange. While the Directorate General of Agraria has staff in all outlying districts, they must give priority in land registration to development projects, to cadastral surveys in urban areas and towns and to registration for commercial enterprises, so the registration of smallholder land is difficult. Furthermore, the cost of cadastral surveys and land registration is currently about US$70, well beyond the capacity of most transmigrants to pay. (c) The Cost of Land. Migrants are also limited in their ability to purchase land. There are many areas in the outer islands where land of moderate fertility is available at a "fair price", often about US$40-60/ha. However, migrants seldom have this amount of money and Government does not pay cash compensation for land (See Chapter V). Therefore, since land cannot be alienated without local consent, an impasse results in which the better land is seldom relinquished for transmigration, and spontaneous migrants settle on steep slopes or in protected forest areas, even though there are large areas of better, underutilized land available at a reasonable price. (d) Planning. Finally, spontaneous migrants are not taken into account in the planning process. Under the pressure of high targets, most areas which were identified in Repelita III were settled with fully sponsored migrants, and little thought was given to the amount of land needed to incorporate spontaneous migrants or to provide for family expansion. To overcome these problems it is crucial that Government develop new approaches to facilitate spontaneous movement in the long-term, and develop specific medium- and short-term programs to promote the settlement of past spontaneous migrants and incorporate future ones into the settlement plan. Successful programs to facilitate spontaneous movement will require actions to address these constraints. The Long-Term Program 7.24 Once agricultural land suitable for large-scale sponsored settlement in the traditional receiving provinces of Sumatra, Sulawesi and Kalimantan is used, most resettlement occurring in these islands would be by spontaneous migrants seeking underutilized land or off-farm employment. In many areas this situation already exists. The Government must therefore give increased attention to those policies which facilitate spontaneous movement, but reduce overcrowding on old sites, competition with local residents, and encroachment on forested areas. The three main components of such a program are (a) improvements in the mechanisms for land identification and land registration, (b) provision of capital for land purchase and land registration, and (c) the development of strong measures for environmental protection. - 138 - 7.25 Land Registration. The best way to facilitate land transfer in the immediate future is to increase the incentives for district and sub-district officials to identify and register landholdings suitable for small numbers of migrants. To establish a program along these lines, the MOT might allocate a lump sum of perhaps US$150-200 for settling a family on 2 ha of land. Up to half this amount would be allocated to the district level Agraria office for land surveys and registration. The remainder would be paid to the district (kabupaten) or subdistrict (kecamatan) treasury as an incentive to identify land and facilitate registration. Land registration criteria would eliminate steep slopes and protected areas and ensure reasonable access. Although it would undoubtedly be a slow process to develop the required procedures, such a program could reduce the cost of settlement per family to 3-4% of current costs. 7.26 Credit Mechanisms for Land Purchase. In the past, many communities willingly relinquished underutilized land to the Government for transmigration in order to obtain associated infrastructure, markets and services. In most areas of Sumatra even marginal land now has a financial value and people with claims to such land are generally unwilling to part with it without remunera- tion. At the moment, however, Indonesia has no form of rural credit for the purchase of land by individuals or groups outside of government programs. Both the Government and banks have been wary of credit for land purchase, fearing that borrowers might be unwilling or unable to repay, or that the availability of credit would drive up land prices. However, the development of credit or the provision of funds for land purchase is a necessary aspect of rational land allocation, and such mechanisms must now be developed. 7.27 As a temporary measure, while more general credit mechanisms are being developed, the MOT should consider providing a line of funds for land purchase and land registration. Initially, these funds could be made avail- able to NGOs or other migrant associations which could guarantee repayment. In the long run, however, rational, economic land use will depend on the development of credit mechanisms that enable individual migrants to buy land and improve its productivity. Complete cost recovery would reduce the cost of the program to the cost of administration alone. Even if such funds were not repaid, however, this program could potentially reduce the cost of settlement to US$200-500/family, about 4% to 10% of the current cost. 7.28 Environmental Protection. Wherever spontaneous migrants go, they try to settle on land that does not have local land claims. These areas are often either poorly suited to agriculture (e.g. steep slopes) or in protected forest. Therefore, any program to encourage spontaneous migration must also be accompanied by strong measures for environmental protection. This means that areas of production and conservation forest must be clearly identified and demarcated so migrants are aware of the status of the land. Local officials should also be made aware of land status and there should be sanc- tions against officials who permit settlement in protected areas as well as against those who encroach on such land. Since spontaneous movement is already occurring at high rates in Sumatra, priority should be given to land demarcation on that island, and master plans for environmental protection should be developed by the Bappedas and ratified by MOT and MOF in the receiving provinces. - 139 - I .I .I6 ~r". , ,. e.. N Spontaneous transmigrants developing sawah (wet ricefields) from forested land. Medium-Term Program (Repelita V) 7.29 Due to financial and implementation constraints, any future settlement in the remainder of Repelita IV is likely to be slow. This means that the opportunity exists to plan a Repelita V transmigration program which takes into account the locations to be settled, the mix of settlement models (upland food crops, tree crops, and swamps), and levels of support ranging from fully sponsored through partially assisted and unassisted migrants. The main elements to be considered when deciding on the level of assistance to be provided to migrants in the Repelita V program are (a) location; (b) area of recruitment; and (c) phasing of settlement. 7.30 Fully assisted. If sites are new and/or remote, or if those being settled are poor families recruited in Java or Bali (i.e. families without relatives or friends in the receiving areas), the full settlement package must be provided. There should be no exceptions to this rule. There is no "icushion" in the current transmigration package which can or should be left out if farmers have been recruited in Java and have no alternative means of - 140 - support. Saving money by reducing the subsistence or agricultural packages, for example, could have serious negative consequences with no major saving in cost. Therefore, in planning the Repelita V program, it should be assumed that farmers sent from Java or Bali to new and remote locations should have full support. 7.31 Partially assisted. Fill-in (sisipan) sites (small areas between large-scale sites) and the later phases of successful new sites can be settled using partial packages if migrants are recruited among spontaneous migrants and second generation families already in the receiving areas. Because such households already have family and friends in the area, because they know the area and the difficulties inherent in settlement, and because they are, in the case of spontaneous migrants, self-selected, they can make reasonable deci- sions about whether they have the resources and ability to cope with new settlement in the absence of full support. 7.32 Recruitment in the receiving areas implies a shift in emphasis from counting migrants "moved" to counting those families "settled". The Government is concerned that such a shift may lessen the demographic impact of transmigration in critical areas, but this need not be the case. Intensive recruitment in critical areas such as upper watersheds could continue under the fully sponsored program, and spontaneous migrants from these areas would be expected to follow. 7.33 Several options are available for partially assisted schemes: (a) Homesteading. The basic component of any partially assisted package is land. For the last decade or so a number of writers have proposed an Indonesian homesteading policy such as that used to settle the outlying areas of the United States and the Philippines. Under homesteading, the Government would identify and obtain the land to be settled, divide it into plots and possibly provide minimum infrastructure. These plots would then be given to farmers registering on site. Such a program would work best in Sumatra where the infrastructure is already available and where many migrants would have relatives in the vicinity. It is unlikely to be successful in the more remote areas earmarked for transmigration. The cost of such a package would be on the order of $500/family, e.g., the amount needed for site selection and evaluation. (b) Other Land Allocation Schemes. The second most important element is access. In successful settlement areas, where infrastructure exists or is being put into place, the provision of individual plots of land along village roads would be more than sufficient to induce settlement by spontaneous migrants already in the area. To encou- rage resettlement in slightly less developed areas, 0.25 ha house- lots along the road could be cleared as roads were constructed. To promote agricultural production, foodcrop lots should be allocated directly behind the houses in order to facilitate clearing and development by family labor. Land titling should be done after the land is cleared by the farmer. Adequate reserve land should also be provided for second stage development. The cost of this package is estimated at about US$1,000/family. - 141 - (c) Land Allocation with Other Benefits. Most partially assisted migrants would also profit from start-up funds for housing and agricultural supplies. Ideally this would be in the form of a cash grant of at least US$500 which could be used by the migrants for needed goods and services. The Government is wary of cash grants since it is concerned that migrants may register for the cash alone or that funds may be diverted. Recognizing these problems, goods could be provided in kind. In order of priority these are: (a) subsistence supplies; (b) agricultural inputs; and (c) building supplies. These elements are worth about $1,000. Construction of a house, rather than the provision of building materials, would add about $500 to this package. The cost of this partially assisted package including building supplies would be roughly US$2,000, and with a constructed house it would cost about US$2,500, about half the current cost. (d) Unregistered Migration. Any mandatory registration, particularly if accompanied by checks in the area of origin or the receiving area, will potentially discourage circulatory migration and free move- ment. For this reason mandatory registration is to be discour- aged. Instead MOT should introduce positive incentives for people to check in with transmigration offices and register in both Java and the outer islands. The incentive for check-in in Java might be to obtain discounts on transportation, maps and information. The incentives in the receiving area might be in the form of information on benefits which can be obtained in terms of land, extension, or other forms of support. The number of unassisted migrants should be counted in the receiving provinces. 7.34 Toward a Staged Program of Settlement. The Government has experimented in the past with programs where a portion of the site was allocated to sponsored migrants, a portion to local people, and a part was reserved for spontaneous migrants, the so-called tri-partial approach. In the light of budget constraints, a new program of this type is required on a phased basis. In the first stage, pioneer settlers (who can be recruited in the sending or receiving provinces) should receive the full-benefit package. In the second stage, partially assisted migrants could be recruited in receiving areas. Land would also be reserved for third-stage settlement by spontaneous migrants and the second generation. Local people could be settled at any point in this process. Short-term Measures 7.35 As of FY86/87, the Government had cleared land for 60,000 families and constructed houses for 54,000 families, but due to budgetary constraints it had funds for settling only 26,000 families on the fully sponsored upland program and for settling 10,000 fully sponsored families on tree crop schemes. Under these circumstances, there has been growing interest in partially assisted programs which could potentially accelerate settlement while reducing costs. It is important that the government recognize, however, that the urgency of budget constraints does not alter the guidelines listed above: - 142 - (a) the first settlers in pioneer sites, settlers in remote sites where off farm work is not widely available, and settlers recruited in Java and moved to areas where they do not have family or friends should be fully supported; (b) if the package is to be reduced, settlers should be recruited in the receiving areas; (c) the better the area, and the more existing and spontaneous migrants in the vicinity, the more the package can be reduced. Within these guidelines, partially assisted programs are appropriate. The programs cited above would also go a long way toward relieving the pressure for land by spontaneous migrants in existing sites and a special effort should be made to ensure that surplus land within existing settlements (that beyond 3.5 ha/family) is allocated to spontaneous migrants before the community is transferred to the province. Institutional Mechanisms for Program Development 7.36 There is a Directorate of Self-Initiated (Swakarsa) movement in the Directorate General of Mobilization and Development; however, this Directorate is relatively new and its staff are not sufficiently senior to recommend major policy changes. For this reason, program development has been slow and has focused on the components of the settlement package (i.e., eliminating the cost of transport, land clearing or houses), rather than on supporting the ongoing process of spontaneous settlement. 7.37 Recently, the office of the Secretary General in the MOT has tackled several important program and policy issues by forming special teams drawn from select senior officials and the staff of appropriate Directorates. This model could also be used to design key policies to support spontaneous migration and develop short-, medium- and long-term programs for partially assisted settlement. These policies and programs could be developed in the office of the SEKGEN and then discussed by the Minister with the other Ministries invol"ed. The Swakarsa Directorate would function as the Secretariat for this purpose. The Prospects for Spontaneous Movement 7.38 About 1.1 million people moved from the inner to the outer islands during 1975-80, of whom 878,000 moved to rural areas. At least 600,000 of these people were not s gnsored transmigrants. This is the equivalent of about 150,000 families - of spontaneous movers or perhaps 30,000 families per year. With the level of investment and amount of development activity in the outer islands in Repelita III, it is reasonable to assume that this level increased significantly, at least doubling to 50,000-75,000 families/year. 2/ This is an estimate since spontaneous migrants often move individually. - 143 - Government targets for FY85/86 called for over 100,000 families to move on partial or unassisted programs, and the question is whether this level of movement can be achieved or should be attempted. 7.39 The amount and direction of spontaneous movement in Repelita IV will depend on a number of factors. First, the state of the economy and the pat- tern of government investment. The relatively large amount of spontaneous movement from the inner to the outer islands in the last ten years has been in response to labor opportunities created there. Some of this employment has been generated in the process of creating transmigration sites, some as a result of the need for secondary services by migrants, some in different sec- tors such as mining and manufacturing. If public investment in the outer islands slackens and economic development slows, the number of spontaneous migrants will lessen. 7.40 Second, since spontaneous migrants often follow relatives or friends, it follows that spontaneous migrants will not be the first settlers in pioneer areas. In the absence of a strong sponsored program, only a few inner island residents would be expected to move spontaneously to Irian Jaya. More would move to Kalimantan, where there are already sponsored settlers and where employment opportunities in the timber industry are attractive, and the vast majority of spontaneous migrants will continue to go to Sumatra. Since much of the available land in Sumatra is already occupied, strong government encouragement to move there in the absence of improved land acquisition systems and strong environmental protection measures could have negative consequences. Chief among these would be a flood of unskilled labor onto provincial markets which would drive down wages and increase underemploy- ment. These factors should be carefully monitored by the MOT. 7.41 In summary, it is not certain that the Government can achieve high levels of spontaneous movement to the outer islands if investment there slackens, and efforts to sustain high levels of movement in the absence of employment creation could have an adverse effect on the regional economies. For this reason it is important that the MOT use targets only as guidelines and concentrate on the key policies required to facilitate spontaneous move- ment. Among these, establishment of a responsive system of land registration, establishment of a short-term (MOT-supported) and longer-range (bank- supported) credit system for land purchase, and development of strong environ- mental protection measures are urgently required. Other programs such as homesteading and partially-assisted settlement packages will help circumvent immediate budgetary constraints, but they cannot circumvent long-term problems of land availability and economic land use in the outer islands. C. Financial Constraints Repelita III 7.42 During Repelita III, finances were not a limiting factor in transmi- gration development. As Table 7.4 shows, in four of the last six years for which data are available, budgets for transmigration exceeded expenditures, and between 1981/82 and 1984/85 funds expended as a percent of funds budgeted actually declined. The main reason for decreasing expenditures was that - 144 - implementation constraints led to the decision taken toward the end of Repe- lita III to include spontaneous migrants within the official 500,000 family transmigration target. Since budgets were compiled on the assumption that 500,000 families would be settled under the sponsored program and only 366,000 families were settled in this way, by the end of Repelita III, the MOT had growing budget carry-overs from preceding years. For example, in October 1985, actual expenditures from the 1984/85 budget (ending 6 months earlier) totaled only about 20% of the FY84/85 budget. The bulk of MOT expenditures in 1984/85 was from previous years' budgets. Table 7.4: BUDGETS FOR THE TRANSMIGRATION SECTOR AND EXPENDITURES FOR REPELITA III AND EARLY REPELITA IV (Rp billion) Repelita III Repelita IV 79/80 80/81 81182 82/83 83/84 84/85 85/86 Transmigration budget /a 138 332 370 475 460 515 546 Actual expenditures /a 143 300 376 387 399 360 - Expenditures as a % of budget 104 90 101 81 87 70 - Total development expenditures 2,490 3,865 4,603 5,091 7,419 7,462 - Transmigration expendi- tures as a % of total expenditures 6 8 8 8 5 5 - /a This table shows budget approved by BAPPENAS for the sector rather than the MOT and actual expenditures within the FY as recorded by the Ministry of Finance, not expenditures from the FY budget. Source: Ministry of Finance and MOT. Repelita IV 7.43 During Repelita IV, the budgetary situation is dramatically different. With the fall in oil prices austerity measures were announced in January 1986, which reduced the MOT budget from Rp 501 billion (US$452 mil- lion) in FY85/86 to Rp 323 billion (US$293 million) in FY86/87, a reduction of about 35%. (Since preceding budgets had overestimated MOT expenditures, this did not necessarily imply a 35% cut in the transmigration expenditures.) Then in April 1986 with the prolonged recession in oil prices, the MOT budget was further reduced to Rp 208 billion (US$180 million). This amounted to 42% of the FY85/86 budget and 58% of FY84/85 expenditures. 7.44 Under the circumstances, in mid 1986 the MOT reviewed and revised its priorities. Under the FY86/87 program the following items are listed in descending order of emphasis. - 145 - (a) Site Development. Highest priority is given to the fulfillment of MOT's commitment to existing migrants through the provision of social and economic elements of the standard transmigration package (subsistence supplies, agricultural inputs). Second, is the upgrad- ing of production systems for those farmers who are not yet self- sufficient, for which a budget for 35,000 families is provided. Third, is the upgrading of infrastructure, including the rehabilita- tion of 3,500 km of roads and 3,700 bridges. The Government places priority on these matters, both to meet its commitment to those who have already moved, and to improve the quality of life in existing settlements in order to attract spontaneous migrants. (b) Committed Projects. Three types of commitments are recognized. First, priority is given to the completion of contracts that the MOT has already signed. Although most Repelita IV contracts with provincial firms are single year contracts, several multi-year contracts are still underway for planning and site preparation. Second, the MOT recognizes the need to complete investments in housing and settlement for sites which have already been cleared, in order to take advantage of the financial investments already made. Third, MOT has given priority to the provision of counterpart funds for externally assisted projects which augment the resources available to the Ministry. These projects, assisted by IBRD and ADB, are projected to settle only about 4,000 families in FY86/87. Counterpart funds are also allocated for the continuation of studies and the provision of services (e.g., World Food Programme) involving external assistance. (c) Institutional Development. Finally, the Government has given priority to institutional development, including the coordination of ongoing services, monitoring and evaluation, training, efficiency improvement and studies. This is an appropriate response to difficult resource conditions. Implications of Budget Reductions 7.45 These priorities have significant consequences for the transmigra- tion program. First, and most important, no new land clearing will be begun in FY86/87 except for 3hat which is intended to make areas already cleared, ready for settlement,3 and for the relatively small area under externally assisted projects. This does not mean that transmigration will halt. As of March 31, 1986, 53,700 houses had been completed by MOT, and the DGE held a 3/ Typically houselots are cleared and houses are constructed before field areas are cleared. This means that some land clearing will have to be carried out to permit settlement in areas where houselots have been cleared and houses established. - 146 - budget for 26,000 houses which would permit settlement in areas where 56,700 ha of tree crops had been established for transmigrants. 7.46 A second implication of current planning is that programs will have to be developed for partially assisted settlers. About 52,000 houses will be available to migrants on upland and tidal schemes, but sufficient funds are available to move only 26,000 families on fully sponsored programs. Similarly the DGE may have houses for nearly 34,000 families, but the MOT budget includes funds for moving only 10,000 families. MOT officials explain that this discrepancy is intentional, and the residual houses are to be filled by partially assisted migrants who do not get other support. As noted in the previous section, for this strategy to be successful, partially assisted migrants should be given preference for the best (most accessible sites). It is also urgent that programs to settle partially assisted migrants be initiated immediately as houses cannot remain unoccupied in Indonesia for more than a few months without severe deterioration. 7.47 Finally, the budget situation has important consequences for planning. The Government has already identified areas that are potentially suitable for 400,000 families on the outer islands. Of these, land for about 170,000 families is in Kalimantan where local land use and forest constraints may limit actual settlement, and sites for 140,000 families are located in Irian Jaya, where settlement prospects are likely to be limited by costs and by social and environmental considerations (Chapter V). This large backlog of potential sites, combined with a downturn in settlement levels, means that planning must also be slowed, and to this end the second phase of studies proposed under the Bank-assisted Transmigration V project has been cancelled. The Future Transmigration Program 7.48 Oil revenues are not expected to return to previously high levels in the immediate future. For this reason, the Government will face trade-offs in the future between second-stage development or new settlement, and between high-input programs (cash crops) on a smaller scale or lower-input projects (annual crops) on a larger scale. The Government can contain costs in three main ways: (a) by reducing the sponsored program and emphasizing fill-in (partially assisted) and spontaneous programs; (b) by reducing the number of people settled in more remote areas; and (c) by reducing targets. Tree crop development reduces total program cost on a longer term basis because 50-75% of costs can be recovered. Since they cost more initially, however, tree crops do not address the immediate problem of financial constraints. Cost savings cannot be realized by reducing unit costs or limiting upgrading programs without serious negative consequences for the migrants, and cost recovery on food crop schemes does not seem feasible given the low incomes of migrants. - 147 - VIII. INSTITUTIONAL CONSIDERATIONS Overview 8.01 During Repelita III, the responsibility for transmigration was distributed among the line agencies normally in charge of each sector. Chief among these were the Directorate General of Transmigration (DGT) in the Ministry of Manpower and Transmigration, the Directorate of City and Regional Planning in the Ministry of Public Works, the Directorate General of Agraria in the Ministry of Home Affairs, and the various Directorates General in the Ministry of Agriculture (MOA). Coordination between agencies was provided by a Junior Minister for Transmigration (JMT). Under the Indonesian system, Junior Ministers provide coordination across Ministries and provide information and advice to the parliament and the cabinet on the progress of interagency programs. 8.02 These arrangements were not satisfactory. The Office of the JMT had insufficient authority to coordinate planning and budgeting, interagency objectives differed, and the coordination of implementation was poor. Because of these problems, a new Ministry of Transmigration (MOT) was formed in 1983. The former Directorate General of Transmigration became the Directorate General for Mobilization and Development (DGMD), and the Public Works Directorates responsible for site selection and evaluation and for land prepa- ration became the Directorate General for Settlement Preparation (DGSP) within the MOT. The budgets of some agencies such as Health and Education were also incorporated into the MOT budget, while other agencies such as Agraria and the Directorates in the Ministry of Agriculture retained control over their own funds. Under the new Ministry, the staff and functions of the Inspector General and the Secretary General were expanded in order to improve financial and administrative control. 8.03 While these steps have laid the groundwork for future development, problems remain. Planning, budgeting, monitoring and evaluation, which are the chief management tools of the Ministry, remain weak; and with the combina- tion of the two most important agencies in the program, coordination with other agencies has lessened. In addition, due to the pressure of high targets during Repelita III and the first two years of Repelita IV, the agency focused on the very difficult tasks of site selection, site preparation, and the selection and transfer of migrants under the sponsored program, while policy and institutional issues took a back seat to program expansion. 8.04 This chapter will look at the most important institutional matters affecting transmigration implementation: planning and budgeting, monitoring and evaluation, manpower development, and structural arrangements within the MOT. It will also discuss problems of interagency coordination and the MOT's future role. It concludes by indicating needed shifts in direction and empha- sis over the next 20 years. - 148 - A. Institutional Development Within the MOT Planning 8.05 Problems of interagency coordination during Repelita III had a very serious impact on planning. In the initial years of the program, targets were established with only limited knowledge about land availability, and budgets were prepared based on targeted figures for each province and district (kabupaten). The Ministry of Public Works (MPW) agencies fared better than others because the agency was relatively centralized, tendering occurred at a national level, and there was some flexibility in the use of funds, so that if sites could not be opened in one district, contractors could be deployed to another. MPW also had the advantage of doing the initial work on each site. . Agencies later in the sequence, such as the DGT, had more problems with planning and budgeting and those decentralized agencies last on site, such as the Directorate General of Food Crops, had a difficult task. 8.06 During Repelita III, planning improved. Better site selection and evaluation allowed more accurate estimates of the number of families to be settled on a given site; annual meetings of provincial officials and line agencies provided a forum for obtaining a provincial consensus on the targets and areas to be developed; and midway into the period the Governor's offices began providing a "green light" for settlement after the necessary first steps had been completed. Planning and implementation were also increasingly decentralized and the provincial planning agencies (Bappedas) played a growing role in reconciling development objectives and facilitating implementation. However, many management problems persisted and the recent reorientation of the Ministry from new settlement to second stage development and changes in budgeting procedures that eliminate budget carry-overs, are placing increased demands on the planning system. 8.07 In Repelita IV, the MOT's Office of the Secretary General (OSG) is responsible for planning, budgeting, monitoring and evaluation. This office sits outside of the Ministry's two Directorates General and should be in a good structural position to undertake these tasks. However, integrated planning functions are new to the office, and it is not yet adequately staffed. Some planning and monitoring functions are also being carried out in other agencies (Directorates) within the MOT and the relative roles of the OSG and the Directorates are still being clarified. High priority should, therefore, be given to strengthening the Planning Bureau in the OSG and to reducing duplication of functions in the MOT. Budgeting 8.08 Budgeting is a subsection of planning. If program targets are unrealistic, if information on the status of implementation is not available, or if unit costs are not updated from the field, adequate budget preparation is impossible. In part, because of these problems, MOT carry-overs in September 1985 amounted to more than twice the budget allocation for FY86/87. Under the new budget procedures, carry-overs of funds from one year to the next are not permitted and budget preparation will have to be improved. Elements of the process should include: - 149 - (a) the preparation of realistic program targets at the provincial level; (b) development of realistic implementation schedules for contracted works and schedules based on actual experience to date; (c) careful review of actual unit costs on a province-by-province basis; (d) close monitoring of implementation progress so the amount of work already done (and to be done) is known; (e) computer processing of financial information; (f) annual updating of key parameters (implementation time, unit costs); and (g) manpower development commensurate with the task at hand. Improvements in bookkeeping and accounting procedures are also required, and the Bureau of Finance in the MOT is in need of technical support. Monitoring and Evaluation 8.09 One of the weakest points in transmigration has been the monitoring and evaluation of project implementation (quantity, quality and cost) and project impact. In Repelita III, the Bank and the United Nations Development Programme (UNDP) provided technical support to the JMT to monitor project implementation, finance, and impact, and to develop an early-warning system intended to detect and react to problems on site. Because of limitations on JMT staff and resources, however, this system was never effective and has been abandoned. New systems must now be developed to meet these functions. 8.10 In Repelita IV, the Planning Bureau in the OSG has been given the task of overall monitoring and it is attempting to establish a comprehensive monitoring system. However, the ability of the Planning Bureau in the OSG to develop an adequate monitoring system has been hampered by the fact that the office is trying to gather too much material. Data sheets are too long, they include information not readily available to MOT staff, and are therefore difficult to complete, and they are collected monthly, thus taking a great deal of time. For these reasons, collection rates are low (recently 35%) and declining. To reduce this problem, monitoring undertaken by the Planning Bureau should be greatly simplified and confined to key indicators of physical and financial progress. Questionnaires should not contain information unavailable to the site manager, or, if this information is critical, mechanisms should be introduced for contractors and other agencies to provide such information to the site manager. Information should be collected no more than once every three months. 8.11 At the same time that monitoring in the OSG is simplified and given priority, monitoring systems at the provincial level and within the Directo- rates General should also be redefined and developed. - 150 - (a) Local level. Since procurement, tendering and contractor supervision have been decentralized, these tasks must also be monitored at the provincial and district levels. Information on migrant welfare and specific site problems must be provided on a regular basis to the provincial MOT head (Kakanwil) so that action can be undertaken to resolve problems before they become serious. Reporting systems already exist, but technical support from the OSG to improve local level monitoring systems should be considered. (b) Directorates General. Technical data on the progress and quality of implementation should flow from the provinces to the Directorates General. In addition, since each Directorate General is required to develop guidelines for implementation, they should be able to carry out special studies that would assist in policy formulation, for example, studies on road standards, the condition of potable water, fertility levels and child nutrition in transmigration sites. Certain data collected from the provinces by the Directorates General should be forwarded to the Planning Bureau, and the roles of these agencies in data collection should be clarified. Manpower Development 8.12 The Ministry of Transmigration could profit from both a simplification of structure and consolidation of existing staff numbers. However, future functions will require a clearer policy orientation and good high-level staff. These staff should be recruited from the best officials in provincial offices, so that they have field experience and so that career opportunities are developed within the agency. Improved training is also required. 8.13 In order to impove the quality of manpower in the Ministry of Transmigration a number of initiatives have been taken. Technical support has been provided under the Bank-assisted Transmigration IV project to assist the training unit within the OS to assess training needs, assign priorities and develop the curriculum and instructors required for the task. Training mate- rials are being developed and diploma courses have been established at Gadjah Mada University to train transmigration officials in the theory and practice of the program. About 600 staff of MOT are also to be trained under the United States Agency for International Development assisted management training program, open to all agencies, and about 60 staff will be sent overseas by MOT to earn advanced degrees. These efforts deserve continued support. Other Steps Needed to Improve Quality Control and Program Management 8.14 The Role of the Directorates General in Project Implementation, As noted, setting standards and drawing up contracts, procuring goods and services and supervising works were largely in the hands of the national offices of the various line agencies in Repelita III, particularly for larger contracts. In order to decentralize authority and responsibility and to - 151 - distribute the benefits of government investment to the regions, these func- tions were transferred to the provinces in Repelita IV. This action has placed the central Directorates and Directorates General in a weak position, and contractor supervision and quality control have suffered. To address this problem it is urgent that institutional arrangements be reviewed with an eye to improving quality control. To do this the central Directorates General should be given a stronger role in setting standards, in strengthening the capacity of provincial staff to do adequate supervision, and in monitoring the quality of such work. The Inspector General working through the central Directorates also has a strong role in reviewing the quality of works under- taken at the provincial and local level and can withhold funds where quality is inadequate. 8.15 Management Review. At the present time, senior MOT officials are aware of the institutional problems facing the Ministry of Transmigration, but they are often unclear about mechanisms for addressing them. Most high level decisions must be made by the Minister and his advisory staff, but they do not have adequate time to consider policy issues systematically. To address this constraint, there should be a formal review of management issues on a regular basis. This should be carried out by a team of senior officials, including the Director Generals and Directors, provided with secretariat services by the Office of the Secretary General. Such a team should evaluate core management functions in the MOT (planning and budgeting, monitoring and evaluation) and make recommendations for improvement. It should also review on-site manage- ment, arrangements for settlement transfer and contractor performance. Alternative arrangements are possible, but the need for high level staff in the MOT to give collective attention to management and policy issues on a regular basis is clear. B. Interagency Coordination Organizational Arrangements 8.16 Repelita III. With the large number of agencies involved in Repelita III, a number of mechanisms were put into place to achieve interagency coordination. Chief among these were the following: Junior Minister Special Minister for Coordination between Ministries BAKOPTRANS Interministerial Council SATDAL TRANS Committee of Director Generals SATBIN TRANS I Committee of Provincial Level Staff SATBIN TRANS II Committee of District Level Staff In addition, regional coordinators (KORWILs) were to be appointed by the Junior Minister to coordinate activities at the district level, and site coor- dinators (KORLAPs) were to be appointed to coordinate on-site interaction. 8.17 Under these arrangements the Office of the Junior Minister held a pivotal role, and for this reason Bank financial support was provided to the JMT to provide management assistance and to set up a planning, budgeting and monitoring and evaluation system. As noted, this assistance was not well - 152 - utilized as the Junior Minister did not have the authority, permanent staff or budget to carry out the task. Under these circumstances, coordination at the center remained weak, KORWILs and KORLAPs were not appointed, and overall monitoring was poor. Inadequacies in this system led directly to the forma- tion of the Ministry of Transmigration in 1983. 8.18 Repelita IV. Under Presidential Decree 59, 1984, the responsibility for coordination of the transmigration program was given to the Minister for Transmigration assisted by (a) a Secretariat in the Office of the Secretary General; (b) a supervisory team in the Inspectorate General; and (c) a technical team consisting of the various Director Generals. There is no longer any policy making body such as the BAKOPTRANS, although agencies can be called together by the MOT. The Governor and Bupati (district head) are responsible for coordination in the province and district, respectively, and they are assisted by committees consisting of representatives of the various implementing agencies. These arrangements are spelled out in further Ministerial Decrees. In spite of the provisions for coordination, however, there are now a number of areas where improved coordination and integration of purpose are urgently required. Key Areas for Coordination 8.19 Agricultural Development. Several problems hamper coordination be- tween the MOA and MOT. Staff in the MOA feel they have only limited influence over the choice of farm model, selection of sites, and adoption of technical parameters. For example, since the early 1980s, MOA officials, supported by the AARD, have opposed the nucleated farm model and advocated complex farming systems rather than food crop monoculture, but these ideas have yet to be adopted. At the same time, these staff are poorly equipped to be assertive in policy matters. The Agricultural Technical Team is chaired by a single senior official who operates with limited staff, and most team members are concerned with activities other than transmigration. As a result, policy issues are not clearly formulated and brought to the attention of MOT. Finally, in matters of dispute, no agency has clear authority. In order to formulate the accelerated oil palm program, for example, senior officials outside MOT and MOA had to be involved to reconcile conflicting views. The current situation is not satisfactory. The Agricultural Technical Team should be strengthened, and a mechanism should be developed for reviewing agricultural policy in the MOT, and for integrating the views of MOA and MOT. 8.20 Forestry and the Environment. There are several agencies concerned with forestry matters and environmental protection. The first is the Ministry of Forestry (MOF) which is responsible for identifying, demarcating and pro- tecting forest areas, among other tasks. One problem in reconciling forestry and transmigration objectives is that there is no clear agreement on the areas within the Ministry of Forestry's jurisdiction. The MOF is basing its classification of protection, production and conversion forests on inaccurate 1:500,000-scale maps and has not shown an interest in using the land system classification maps at 1:250,000 prepared by MOT and BAKOSURTANAL (the national mapping agency). To resolve these issues: (a) MOF should utilize the most recent land cover mapping information to improve the accuracy of forest delineation; (b) the criteria for preservation of forest and for - 153 - conversion should be more clearly formulated; and (c) MOT and MOF should develop mechanisms for deciding on land availability early in the planning process. In matters of environmental protection, close coordination is also required with the MPE. The Directorate of Environment in the Ministry of Transmigration also has an important role to play in identifying important issues and agencies in resolving them. 8.21 Settlement Transfer. Coordination with the provinces is also needed to facilitate settlement transfer. Transmigration settlement areas are under MOT authority until a reasonable level of self-sufficiency is achieved, a period averaging about five years. The settlement is then transferred to the province which assumes the responsibility for public administration, mainte- nance of roads and public facilities, health, education and other routine services. Prior to transfer, these services are paid for by the MOT. The two main problems in settlement transfer are in road maintenance and the transfer of staff. (a) Roads. Roads are often so deteriorated on transfer that their improvement overwhelms the limited capacity of the district public works offices, and for this reason upgrading is often deferred. To address this problem, mechanisms are needed for road maintenance while sites are under MOT authority, and the provinces require ample forewarning of infrastructure to be transferred, of its status, and of the upgrading work required (including financial implications). (b) Transfer of staff and officials. Many of the staff in schools, clinics and the extension service in transmigration sites are on honoraria rather than in structural positions, and they expect to become civil servants when the community is transferred to provincial authority. With budget constraints, this may be difficult, and long-range planning by MOT and the provinces for the transfer of these staff is imperative. 8.22 Spontaneous Migration. Chapter VII argues for a major initiative by the MOT to develop systems of land registration and mechanisms for land pur- chase in the outer islands in order to facilitate and regulate spontaneous mi- gration. The agency responsible for surveys and registration is the Directo- rate General of Agraria in the Ministry of Home Affairs. Arrangements for land purchase would involve the Directorate General of Regional Development (Bangda), the Ministry of Finance and the development banks. The agency responsible for initiating these activities is the Directorate of Swakarsa (Self-Initiated) Settlement; however, staff in this Directorate are not sufficiently senior, or experienced, to convene and guide the agencies which would have to be involved. 8.23 Social Affairs. Improved coordination is required between regional officials and the MOT in the evaluation of policies for compensation, the settlement of disputes, the identification and provision of culturally- appropriate benefits to local people, and the treatment of ethnically distinct peoples. Policy formulation and coordination on these matters could poten- tially be handled within MOT by the Directorate for Social and Cultural Development or by a special Directorate formed for this purpose. - 154 - Steps Needed to Improve Coordination 8.24 National Level. It is necessary to further develop the resources in the Planning Bureau of the OSG to support the interagency coordination committees established under Presidential Decree 59. To do this, special teams might be formed, as required, to review and recommend policies on matters requiring interagency coordination. These teams could be provided with secretariat services by Directorates within the MOT. Possible organiza- tion arrangements are shown in Table 8.1. Formation of such interagency teams or the establishment of alternative methods of promoting interagency policy formulation and coordination are critical to further program development. Table 8.1: POSSIBLE ORGANIZATION OF PROPOSED SPECIAL TEAMS TO ADDRESS MANAGEMENT AND POLICY ISSUES Issue Key Members Secretariat Management Team Senior MOT officials Office of the Secretary General Policy Development teams Agricultural Development OS, DGSP,/a DGMD /b, Directorate for Agricultural Technical Team Economic Development Forestry and Environment MOF, MPE, DGSP, DGMD Planning Bureau (DGSP) or the Environment Directorate (MOT) Social Affairs Bangda, DGA, Health, Directorate of Social Education, DGMD Cultural Development Spontaneous Migration Bangda, DGA, OSG, DGMD, Directorate of and Land Registration DGSP Swakarsa Migration Settlement Transfer Bangda, DGMD, DGSP Directorate of Social Cultural Development /a DGSP: The Directorate General of Settlement Preparation in the MOT. 7i DGMD: The Directorate General of Mobilization and Development in the MOT. 8.25 Provincial/District Level Coordination. Transmigration coordination committees already exist at the provincial, district and subdistrict levels to coordinate the day-to-day activities of transmigration. The work of these groups has improved significantly in the last five years, and could be further facilitated by the provision of secretariats with adequate budgets within the transmigration offices at the provincial and district level. As with the central secretariats, regional secretariats would establish agendas and issue proceedings. - 155 - 8.26 On-site Coordination. Although mechanisms for coordination between provincial agencies have been improved, on-site coordination remains weak, particularly where site development entails the coordination of a number of agencies as in swampland development. This is partly because the site manager for MOT has no control over non-MOT staff, and decision makers affecting the settlement are located variously in the subdistrict, district or provincial capitals depending on the agency in question, thereby making coordination difficult to achieve. One solution to on-site coordination problems is to form a Project Management Unit (PMU) at each settlement with control over its own budget and the ability to purchase needed services. This model worked successfully in the first Bank-assisted transmigration project and in two other externally-assisted projects, the Luwu Project assisted by the United States Agency for International Development (South Sulawesi) and the Southeast Sulawesi Project assisted by the Asian Development Bank. A second possibility is to give the MOT project head authority over all other agencies at the site level. Other options may be possible, but it is clear thaL action must be taken to improve on-site coordination. C. The Changing Role of the MOT 8.27 Preceding sections of this report have suggested that prospects are limited for the food crop farm model, and that there are serious land and bud- get constraints. The report, therefore, recommends a shift to new settlement with tree crops, an emphasis on second-stage development (including but not necessarily limited to tree crop development), and steps to facilitate and improve spontaneous migration. These recommendations have significant insti- tutional implications. New Settlement with Tree Crops 8.28 In the past the bulk of the work of the MOT and the funds expended by it were devoted to activities involved in settlement on food crop schemes: land identification and site preparation, selection, transport and initial support of settlers. In the NES/PIR tree crop development programs, the estates take the initiative in identifying and clearing land, and esta- blishing houses and physical facilities. When the social and administrative infrastructure is ready, the estates through the Governor indicate that settlers are required. MOT then selects and transports migrants, and settles them, while the estate employs the migrants as wage laborers in tree crop development until trees are mature. Although these arrangements have worked relatively well with only 30,000 families involved, they have the disadvant- age, from the point of view of the MOT, that they do not permit adequate forward planning and they reduce MOT's control over a major part of its future program. 8.29 In March 1986 a Ministerial Decree was issued that provides funds for tree crop development by public sector estates from the transmigration development budget, but gives the Directorate General of Estates (DGE) the authority for tree crop development. This decree also establishes a coord- inating committee chaired by the Junior Minister for Tree Crops and including representatives from MOT. There are a number of advantages in having the DGE take the lead in planning and implementing tree crop development, and the pre- - 156 - sent policy of financing smallholder development through the DGE in close cooperation with the MOT is welcome. Nevertheless, close coordination between the DGE and MOT is imperative and the MOT has a strong role to play in land identification, settlement, and in monitoring migrant welfare and ensuring that migrants are treated equitably between schemes. This should be super- vised by the Directorate of Economic Development. Second-stage Development 8.30 Ordinarily, transmigration projects are under the authority of the MOT until they are self-sufficient, at which time they are transferred to the provinces and the buildings, staff and services are taken over by the appro- priate provincial agencies. In the coming years the MOT intends to give increased emphasis to road rehabilitation, the provision of cash crops and tree crops, and other upgrading activities on sites under its authority. However, if these activities take longer than the period the schemes are under MOT authority, problems of budgetary responsibility arise. This is one reason that second-stage development has been carried out mainly after the settle- ments have been transferred to the provinces and then by sectoral agencies in the course of ongoing development work. For example, in the past, PMU pro- grams have not established tree crops for transmigrants until after the set- tlement is transferred to the province, which means that the time from the date of settler arrival to tree crop maturity is about 15 years. This is unsatisfactory. 8.31 There are several possibilities for addressing this problem: (i) the MOT could begin the work while the site is under its authority and the line agencies would have the budget after settlement transfer. Although this appears to be the simplest, line agencies are reluctant to be committed in this way to activities which may have priority to the MOT but not to the agency involved. Alternatively, (ii) the MOT could hold the budget for second-stage development even after the community is transferred to the pro- vince; (iii) communities could remain under MOTs authority until all needed work is finished, or (iv) MOT could enlist sectoral agencies to begin work using their own budgets before the transfer of sites. This option could potentially involve a bridging period between the time when sites are under MOT control and fully transferred to the province. 8.32 The second option could be taken under review, but the third option would probably prolong dependency and limit Bappeda involvement, and should not be encouraged except on critical sites. There are strong arguments to be made favoring the last option. MOT staff are on site, and after 2-3 years they are in a good position to describe the status of development (current incomes, land availability, etc.) and recommend upgrading and second-stage development. MOT could collate these data at the provincial level and negotiate with sectoral agencies on the location and type of activities needed for second-stage development. Funds earmarked for second-stage development could then be allocated to the line agencies and the MOT could ensure that technical Ministries give priority to second-stage development in planning and budgeting. This approach may be difficult in light of budget constraints, but decisions are required before large-scale second-stage development is begun. - 157 - Spontaneous Migration 8.33 To date, the MOT has had only a modest role in promoting spontaneous migration because it has viewed the opportunities for intervention in terms of the sponsored settlement process--registration, selection, transport, settle- ment, and agricultural development--and has asked the question, how many of these elements should be provided to partially-assisted migrants, and at what cost. As this report has emphasized, however, in the future, partially-assis- ted programs should be tailored to the level of development of the settlement and surrounding region (i.e., the more developed, the less support). In the longer term, the management of spontaneous migration will require the identi- fication of constraints and the formulation of policies to overcome them. This cannot be done by a small Directorate alone. Instead, this report recommends that an interagency team be formed to discuss and develop needed policies and programs to promote less costly settlement. D. Summary and Conclusions 8.34 Planning, Budgeting, Monitoring and Evaluation. Program planning (including budgeting) and monitoring and evaluation have been weak links in the transmigration program to date, and these tasks are likely to become more important in the future as innovative programs are initiated and resources are constrained. To improve performance in these areas, a high-level team should be formed in the MOT to review management functions, determine problem areas and develop mechanisms to resolve them. Technical support should also be provided to strengthen the OSC. Institutional structures should be simplified, manpower requirements defined and training improved. To improve quality control, the role of the central Directorates should be defined in relation to the provinces and mechanisms should be developed in the MOT for regular management reviews of line agency functions. 8.35 Coordination. Despite the adjustments made over the last few years, poor interagency coordination persists. In the past, the coordinating func- tion was carried out by formal committees and supported to some extent by the JMT and advisory staff. These functions must now find a new home. Recently, considerable authority has been given to the Office of the Secretary General (OSG) in the MOT and the ability of this agency to serve a coordinating func- tion would be strengthened by the establishment of a strong secretariat which convenes regular meetings for this purpose. In addition, some interagency mechanism must be developed to deal with policy formulation and implementation problems in agriculture, forestry and the environment, spontaneous migration, settlement transfer, and social affairs. 8.36 The Future Role of MOT. Twenty years from now, the possibilities for settlement following the current sequence of steps in the transmigration program will be limited. Instead, a broad spectrum of interagency programs and policy initiatives will be required to foster labor migration. To move from the current highly logistical operation to one centering on policy formulation and facilitation will not be easy. To achieve these objectives, it will be important to identify the agency intended to provide guidance and policy direction, to give it appropriate authority and make the maximum effort to strengthen it. At the moment, it appears that this position would best be - 138 - filled by the Office of the Secretary General (OSG). If so, a concerted effort is needed to define the function of the OSG, to attract the most skilled manpower to it, and to give it adequate authority to integrate the work of the other agencies. Technical direction and supervision would be left in the hands of the various Directorates General. - 159 - IX. WORLD BANK INVOLVEMENT IN TRANSMIGRATION Overview 9.1 Since 1976, the World Bank has been involved in the evolution of Indonesia's transmigration program through five Transmigration projects, two Swamp Reclamation projects, and through Nucleus Estate and Smallholder (NES) projects which provide tree crops and other cash crops (sugar) to both trans- migrants and local people (see Table 9.1). Two projects have been appraised that provide cattle in transmigration areas where livestock are in short supply, one receiving financial assistance from the International Fund for Agricultural Development (IFAD), the other from IFAD and the World Bank. Projects for agricultural research and extension have included components to strengthen agricultural supporting services in transmigration areas. In support of this lending program the Bank has carried out two Transmigration Sector Reviews, one in 1980, the second (this report) in 1986. 9.2 Bank commitments for upland transmigration and swamp reclamation total about US$560 million or 20% of Bank lending for Indonesian agriculture and 7% of the Bank's lending to Indonesia in the past decade. Bank-assisted projects, when completed, will settle about 47,000 families in new sites, and upgrade existing settlements including about 16,000 families. IFAD and Bank- assisted projects will provide cattle directly to about 120,000 families and benefit about twice this many through the redistribution of offspring. An additional 400,000 families will benefit through improved site selection and evaluation. In addition, the Bank has committed about US$680 million for Nucleus Estate and Smallholder projects which will settle about 95,000 fami- lies of whom perhaps 25% will be transmigrants. Bank disbursements amounted to about 5% of the cost of the Repelita III program, and are estimated to be about 10% of expenditures in Repelita IV. 9.3 Bank-supported projects have introduced a number of innovative features into the transmigration program including improved site screening and evaluation, support for and evaluation of the major farm models used, assis- tance for program management, and the development of procedures for utilizing timber from transmigration sites. They have also experienced a number of problems. Some problems are common to all projects in Indonesia (land alloca- tion, coordination and quality control), but others are due to the difficul- ties inherent in programming a complex sequence of events and to high program targets which strained institutional capacity. On balance, however, the World Bank believes that the transmigration program has both merited and profited from Bank support. Projects in Support of the Upland Transmigration Program 9.4 Transmigration I. The first Bank-assisted transmigration project (Ln 1318-IND) was approved in 1976 and completed in 1983. It consisted of two pilot schemes in southern Sumatra: (a) the rehabilitation of an existing settlement of 12,000 families in Way Abung, in Lampung province; and (b) the Table 9.1: BANK SUPPORT FOR TRANSMIGRATION PROJECTS Total Actual or project Bank Effective estimated Project Ln/Cr number cost assistance date closing date Components -- (US$ million) -- Transmigration Projects Transmigration I Ln 1318-IND 56.8/a 30.0 11/30/76 04/13/83 Settlement of 4,500 families in Baturaja; rehabilitation of communities with 12,000 families in Way Abung. Transmigration II Ln 1707/ 242.0 90.0/ 10/04/79 12/86 Settlement of 20,000 new families in Cr 919-IND 67.0 Jambi and South Sumatra provinces; upgrading at sites with 4,000 families in Jambi; site selection and evaluation for 50,000 families. Transmigration III Ln 2248-IND 185.8 101.0/b 07/01/83 06/87 Site selection and evaluation for 300,000 families and settlement of 2,000 additional families in Baturaja. Transmigration IV/b Ln 2288-IND 61.2 53.5 09/06/83 06/89 Settlement and tree crop establishment for 6,000 families in East Kalimantan. Transmigration V/c Ln 2578-IND 145.5 97.0 04/29/86 12/90 Site selection and evaluation for 300,000 families; institutional development. Swamps I Ln 1958-IND 44.6 22.0 07/09/81 12/87 Reclamation of 9,000 ha of swampland in Karang Agung and settlement of 3,200 families. Swamps II Ln 2431-IND 108.3 64.8/b 09/05/84 06/91 Reclamation of 30,000 ha of swampland in South Sumatra and settlement of 11,500 families. Smallholder Cattle IFAD, 40.0 /d 09/01/81 02/87 Distribution of 45,000 head of cattle to Development' Ln 35-10 transmigrant and local families in existing settlement areas of Sumatra. Smallholder Cattle Ln 2628-IND 66.4 32.0 04/17/86 03/92 Distribution of cattle to about 76,500 Development- famililes in Sumatra, of whom about 90% are expected to be transmigrants. /a Cost as originally estimated at appraisal. At completion, the project cost was US$57.3 million. 715 Loan amount excludes front-end fee. Total project cost and loan amounts here are revised estimates. 7-c This loan was reduced to about 100,000 families in light of budget constraints. 7r Project was financed by GOI .and an IFAD loan for US$21.8 million, but apnraised a.nd administered by the Bank. - 161 - settlement of 4,500 families near Baturaja, in South Sumatra province. It also provided funds for research on cropping systems and for long-term monitoring and evaluation. 9.5 New settlers in Baturaja were provided with 5.0 ha farms which included 1.25 ha cleared for food crops and 1.0 ha of rubber trees established for the smallholders by a public sector estate (PTP X). These benefits were provided to the settlers as a grant. To promote food crop development, farmers were given cattle for draft power, and free agricultural inputs (seed, fertilizer and pesticides) for three years. Transmigrants also received a house and one year of subsistence supplies. The project was implemented by a Project Management Unit (PMU) with considerable financial autonomy, and both domestic and foreign technical assistance was provided to support the PMU. 9.6 A Project Performance Audit Report (No. 5157, June 25, 1984) concluded that the project was successful. Farmers in a number of locations produced marketable food surpluses, and wage employment generated by ongoing settlement and by tree crop development provided reasonable cash incomes. These incomes sources combined with future rubber revenues are expected to produce annual household incomes at full development of about US$1,500/family, exceeding the target income for rural areas of about US$1,000/family. Spon- taneous transmigrants now comprise about 27% of all settlers. Rehabilitation in Way Abung also generated considerable employment and project activities combined with other developments in the area (the establishment of a sugar plantation and the introduction of irrigation) have raised farmer incomes and stimulated regional development. 9.7 A number of lessons were learned in the implementation of Transmigration I. Yields were lower than projected and reserve land was not put into food crop production on the large scale envisioned in the appraisal report; lowered expectations were then incorporated into later projects. Farmers using draft animals were able to open up to one-half hectare more of land than those who did not, and this led to the development of two subsequent cattle distribution projects. The monitoring and evaluation component, undertaken by a university team, started late and did not provide timely information to project management, so later support was provided to develop monitoring capacity in the line agency. Finally, the amount of technical support and the absence of cost recovery led to questions of replicability, and subsequent Bank-assisted projects used institutional arrangements which were less management-intensive and included mechanisms for cost recovery for tree crop establishment and livestock distribution. 9.8 Transmigration II. The second Bank-assisted transmigration project (Ln 1707/Cr 919-IND), approved in 1979, was intended to settle 30,000 families along the new Trans-Sumatra highway in Jambi province and to improve Singkut, an older settlement of 4,000 transmigrant families. The project was intended as a model for the Repelita III (1979-84) transmigration program. Under Transmigration II, technical assistance was also provided to the Junior Minister for Transmigration to aid in coordination of the overall program. - 162 - Cattle distribution programs are used to increase land in production and improve incomes. 9.9 The settlement component of the Transmigration II project encoun- tered several start-up problems, due in part to a shortage of suitable land. Detailed mapping in the original project area revealed sufficient flat land for food crops for only about 10,000 of the proposed 30,000 families. The project was, therefore, reformulated to settle about 20,000 families, 10,000 in Ja'mbi and 10,000 in South Sumatra, and the balance of funds was realLocated to site screening and evaluation for the program as a whole. The Transmigra- tion II site screening and evaluation component, known as the Site Feasibility Studies and Engineering (SFSE 1980), was carried out by consultants under the supervision of the Directorate of City and Regional Planning in the Ministry of Public Works assisted by an expatriate Technical Advisory Group. This component was successfully executed and set new, improved standards for site screening and evaluation. 9.10 Settlement commenced in October 1981 and was completed by the end of 1986. The settlement component experienced a number of implementation and coordination problems. Second-stage development in Singkut started slowly due - 163 - to problems of interagency coordination, plans were not made for the introduction of tree crops into new settlement sites, and the components to promote spontaneous migration were not implemented. The technical team provided to assist in the coordination of implementation was not well utilized since the office of the Junior Minister (JMT) did not have adequate authority and staff to carry out its task. However, managerial problems highlighted, in part, by project consultants and JMT staff, led eventually to the formation of a new Ministry in Repelita IV. 9.11 Transmigration Sector Review. In 1981 the Bank reviewed the Goverment's upland and swamp transmigration programs (Transmigration Program Review, Report No. 3170-IND). The report described the history of the trans- migration program, acknowledged the role that the program could potentially play in meeting the Government's development objectives, and concentrated on steps needed to improve project implementation. A major conclusion of the report was that yields and income prospects in lowland swamp reclamation areas were promising, but that a greater role should be given to mixed farming systems including tree crops in upland schemes. The report also recommended that increased attention be given to site selection and evaluation, institution building and staff training, and that a major effort be made to promote better resource utilization through the development of grasslands and the utilization of timber from forested areas which were cleared. The report projected an increase in the size of the transmigration program in Repelita III to about 175,000-200,000 families over five years, about half the number actually moved, and it indicated that the Bank was prepared to assist the Government in the implementation of the program as long as reasonable standards of settle- ment were attained. 9.12 Transmigration III. Given the difficulty of finding sites suitable for large-scale settlement, and given the considerable risk to migrants settled on unsuitable sites, in 1983 the Bank agreed to finance a project pro- viding technical support for the Government's entire program of site selection and evaluation, so long as future upland settlement was confined to sites selected under improved procedures. The Transmigration III project (Ln 2248- IND) was intended to identify sites for up to 300,000 families to be settled during the latter part of Repelita III and early Repelita IV. Under the SFSE 1983 planning program, better reconnaissance was initiated to exclude unsuit- able sites early in the planning process, expanded terms of reference were drawn up which permitted investigation of a larger number of farming models and required the financial evaluation of typical settlements, increased attention was given to determining the amount and value of timber on the site in order to decide whether the site should be accepted, and the Technical Advisory Group was strengthened. 9.13 In addition to site selection and evaluation, Transmigration III also financed: (a) the settlement of 2,000 families adjacent to the first Bank-assisted project in Baturaja, South Sumatra, and additional services and facilities in the existing settlement; (b) staff training in the Directorate General of Transmigration; (c) agricultural research in support of transmigra- tion under the Agency for Agricultural Research and Development; and (d) tech- nical assistance to the Ministry of Population and Environment (MPE) to monitor and improve the environmental soundness of transmigration projects. - 164 - 9.14 Most components are finished and the project is to be completed in 1988. In total, sites have been identified for about 250,000 families. As of June 1986 only about 20,000 families had been settled on sites selected under the SFSE 1983 program (many more were in progress), so an evaluation of improvement on such sites was not possible. Nevertheless, GOI officials and Bank staff believe that standards introduced under the SFSE program will improve migrant welfare and resource utilization. The Baturaja extension was also completed successfully. 9.15 Transmigration IV. This project (Ln 2288-IND), which will settle 6,000 families in East Kalimantan, was identified in 1979 and approved by the Bank in late 1983. This long gestation period was due to the introduction of several innovative features in line with recommendations in the Transmigration Sector Review. The project established procedures to permit the recovery of commercially valuable timber in areas to be cleared, and it developed institu- tional arrangements for the introduction of 2.0 ha of tree crops early in the settlement period. The project also included components to improve the coordination of Bank-assisted projects and to facilitate implementation of the GOI program. 9.16 The settlement model established under Transmigration IV influenced the design of GOI's transmigration tree crop development program adopted in early 1986; but due to delays in contracting, caused by falling timber prices and a lower amount of recoverable timber than bidders anticipated, land clear- ing and settlement were delayed. Contractors started clearing in November 1984, but three of four contracts were cancelled and replaced by new contrac- tors in mid-1986. Although the slow start-up was disappointing, timber in the site cleared by a logging concessionaire has been utilized, the remaining sites have been logged, and procedures established under the project to ensure timber utilization have been applied to other areas. By June 1986 over 600 families had been settled. 9.17 Transmigration V. Under Transmigration V (Ln 2578-IND), approved in 1985, six firms were to be engaged in the first phase of studies and eight in a second phase to identify land for a total of about 300,000 families. Out- line terms of reference are provided in Annex 9. This project also included technical assistance to improve mapping, to help set standards for land clear- ing, to supervise contractor performance, and to undertake a regional develop- ment study of southern Irian Jaya. The project was viewed as a means of improving mapping, containing program targets for Repelita IV (set at 750,000 families), promoting a shift to models other than food crops, and improving the environmental and social aspects of the transmigration program. 9.18 The first round of six site selection studies got underway in late 1986, but with the reduction in the settlement program resulting from the budget constraints of 1986, the second round of eight studies was cancelled. The reformulated project would help identify sites for about 100,000 new settler families and additional studies would be carried out to determine appropriate second stage development on existing sites. Consultants initially hired to supervise land clearing would be used to review standards for roads and bridges and to help plan and implement programs to rehabilitate existing physical infrastructure. Preliminary work on the Irian Jaya regional develop- - 165 - ment study has begun, but early indications suggest that the financial costs and social implications of settlement will dictate very slow development. Other Projects Supporting Transmigration 9.19 Swamp Reclamation I and II. The first Swamp Reclamation Project (Ln 1958-IND), approved in 1980, was intended to reclaim some 9,000 ha in Karang Agung, South Sumatra, settle 3,200 families, and demonstrate improved methods of implementation, land reclamation, land clearing and agricultural development. Funds were also included for the preparation of master plans identifying priority areas for swamp reclamation. Problems were experienced in the coordination of infrastructure development and settlement, which were carried out by different agencies, and in 1985 this project encountered production problems, due mainly to pests. Action plans intended to alleviate these problems have been put into effect and harvests improved significantly in 1986. 9.20 Based on the findings of the first Transmigration Sector Review (1981), which found the best rice yields in settlement areas in swamp develop- ment sites, the Bank decided in 1982 to expand the pilot scheme in Karang Agung, and in May 1984 the Swamp Reclamation II Project (Ln 2431-IND) was approved. This project was designed to reclaim 30,000 ha of tidal swamps and settle 11,500 transmigrant families in an extension of the First Swamp Reclamation Project. It also supported institution building for the Direc- torate of Swamp Reclamation, lowland agricultural research, and studies for future swamp reclamation and rehabilitation projects. Infrastructure development in this project has just begun. 9.21 Smallholder Cattle Development. The World Bank also prepared and appraised two projects which provide cattle to local smallholders in transmi- gration areas where animals for draft purposes are in short supply. The first of these projects (IFAD-Loan 35), was financed by the Government of Indonesia and the International Fund for Agricultural Development (IFAD) and adminis- tered by the Bank. It was intended to procure cattle in Java and South Sulawesi and distribute them in Sumatra. By the end of the project in 1987 about 51,500 head of cattle had been distributed (about 46,000 head to transmigrants and 5,500 to local smallholders). Purchase and transfer were accomplished without major losses and the average area cultivated by the 60% of farmers using their animals for draft purposes increased by about half. These receiving cattle are required to return two heifers to be redistributed to other smallholders, and the original investment is, therefore, expected to benefit over 100,000 families. 9.22 The Second Smallholder Cattle Development Project (Ln 2628-IND), approved in 1985, is intended to procure 84,000 head of cattle from Nusa Tenggara Timur and Sulawesi for distribution to about 76,500 transmigrant and local families in Sumatra. This project is jointly financed by the Bank, IFAD and the GOI. The return and redistribution of heifers under this program is expected to benefit over 180,000 families. - 166 - 9.23 Nucleus Estate and Smaliholder Projects. Since 1978, the Bank has also assisted, a major program of tree-crop and cash-crop development through its Nucleus Estate and Smallholder projects. In these projects, existing public sector estates are provided with technical and financial support to establish their own (nucleus) holdings and to develop adjacent areas for smallholders who may be indigenous people, existing transmigrants, or new settlers. When completed, these projects will settle about 95,000 families with 2.0 ha of tree crops and additional food crop land. About 10,000 families will be drawn from existing transmigration settlements and another 10,000-20,000 may be relocated from Java depending on the number of local people who elect to join the settlements. Most NES projects have been based on the major tree crops, rubber and oil palm, but one project (Ln 2344-IND) has been used to develop sugarcane on the holdings of local people and transmigrants. Lessons Learned 9.24 Program Scale. Bank assistance to the transmigration program began in 1976 when settlement levels averaged about 10,000 families/year. These numbers were judged to be small in relation to the overpopulation and increas- ing environmental degradation on Java, and to the resources available in the outer islands, so decisions about program scale were related largely to implementation capacity and to the cost and benefits of specific settle- ments. The Transmigration Sector Review, issued in 1981, predicted an increase in settlement to about 40,000 families a year and concentrated mainly on ways to improve project implementation. The settlement of 366,000 families in Repelita III was nearly twice the number predicted by the Bank. In addition, the 1981 census indicated that the number of migrants to the outer islands doubled in the decade between 1971 and 1981 and the evidence suggests that the rate of movement increased in Repelita III. This level of sponsored movement, coupled with evidence of growing spontaneous migration, has changed the framework for program evaluation. Transmigration is now a program which significantly affects regional development and natural resource utilization, and decisions about the future program must be made with reference to appropriate program scale. 9.25 Farm Model. The sustainability of the food crop farm model was an issue in the design of both the first and second Bank-assisted transmigration projects and was a major topic in the Bank's first Transmigration Sector Review. This Review indicated that there were many communities which had persisted for years on poor soils in rainfed areas, but that incomes in such communities were low. It, therefore, argued that the rising expectations of rural settlers and the location of sites on increasingly marginal land provided a strong argument for expanding the role of tree crops in the transmigration program. Consistent with arguments in the Sector Review, terms-of-reference for consultants doing site selection and evaluation under the Bank-assisted Transmigration III project were expanded to include an analysis of several productive models, and Transmigration V established detailed procedures for planning tree-crop and other types of sites. Transmigration IV, located in East Kalimantan, included 2.0 ha of tree crops/family. - 167 - 9.26 Spontaneous and Partially Assisted Movement. Transmigration II included a number of components to promote spontaneous movement. These compo- nents were not carried out by the DGT/MOT. The reasons are diverse. The Government was preoccupied with the larger questions of land identification, project implementation, and settlement for sponsored migrants; innovative components in the Bank projects proved difficult to implement in a large program with uniform standards throughout; and the Directorate responsible for developing programs for partially assisted and spontaneous migrants did not have sufficient authority to carry out this task. However, the current scale of spontaneous movement, the potential for environmental problems created by this level of movement, the increasing difficulty in finding suitable land for sponsored settlement, and the Government's financial constraints, mean that policies for facilitating and regulating spontaneous migration are now urgently required. The mix of fully sponsored, partially-assisted and spontaneous migrants and the policies needed to support them should be developed for Repelita V. 9.27 Environmental and Social Issues. Although environmental and social issues have been discussed in past Bank reports and figured prominently in the design of Transmigration V, with the increased scale of transmigration in Repelita III and IV environmental and social issues have become increasingly important, and a major program to address these issues is now required. Steps needed to promote environmental protection and programs for the local population are discussed in detail in Chapter V. Action on these measures will be difficult, since the required activities (such as the development of environmental master plans) are not fully under MOT control. Nevertheless, it is clear that the MOT must take an increasingly strong role in mobilizing and coordinating programs that affect transmigration and that it must be instrumental in the development of adequate social and environment safeguards in transmigration areas. Because of the importance of these issues, in preparation for Repelita V steps should be taken to develop environmental masterplans to support the Environmental Directorate in MOT, and to form or strengthen a Directorate in the MOT dealing with local people. 9.28 Project Management. The first Bank-assisted project was implemented by a Project Management Unit (PMU) and was the most efficiently implemented of the Bank-assisted projects. The Second and Fourth Transmigration Projects and the two Swamp Reclamation Projects have been carried out under the same insti- tutional arrangements as the overall program and have experienced problems due, in part, to difficulties in coordinating the chain of events prior to and after settlement. Steps needed to improve overall coordination must entail realistic targets, a strengthening of MOT management, and establishment of appropriate coordinating structure at the site, district and provincial levels. To improve on-site coordination it will be necessary to return either to PMUs (in which implementation authority, including financial control, is under the MOT site manager), which might be possible in a program of reduced scale, or to develop alternative arrangements to strengthen the role of the MOT site manager. Such steps will also be necessary to obtain the flexibility necessary to develop diverse farming and settlement models. 9.29 Bank-assistance to the Sector. The first Bank-assisted projects in the upland and swamp reclamation programs were site-specific projects which - 168 - included all aspects of land development and settlement. Because of the large number of components in such projects, and the Bank's emphasis on innovative features intended to improve project implementation, these projects have been difficult to implement. And even though the demonstration components (e.g. procedures for timber disposal) have played an important role in program evolution, the consensus is that any future projects should be simplified. The site selection and evaluation components, and the cattle distribution projects, which either assisted the entire Government effort or focused on a single subsector, have been more successful. This has been due to a clear definition of objectives and the means to achieve them, the introduction of uniform procedures, and the provision of technical support in the sector to be developed. Possible subsectors for program-wide support if settlement is resumed would be land registration, land preparation and/or road construction, tree crop development, site selection and detailed planning in swamp areas, the upgrading of health (wells, sanitation, clinics, family planning) and social services on sites already selected under Bank-assisted projects, and projects for environmental protection. - 169 - X. SUMMARY OF MAJOR RECOMMENDATIONS IN THE MAIN REPORT Second Stage Development 10.1 At a time of resource constraints, there is a strong argument for further investment in existing sites in order to sustain and improve migrant incomes and to capture the economic benefits of the manpower and infrastruc- ture which are already in place. In addition, the Government should make an effort to sustain reasonable levels of investment in those provinces and districts which received large numbers of migrants in Repelita III, in order to prevent increasing competition for work, falling wages and possible second round flows of labor either back to Java or to outer island cities. 10.2 The most important form of second-stage development is likely to be in tree crop establishment in existing sites. To establish tree crops for transmigrants, the PMU programs must be strengthened and expanded, and the MOT should conduct an inventory of sites to determine which ones require tree crops and whether the sites have sufficient land for tree crop development. To support such development, steps are needed to mobilize private sector investment in tree crop establishment in transmigration sites. 10.3 Maintenance of settlement infrastructure must also be given high priority. In transmigration settlements, road maintenance is particularly critical to economic growth. The Government has taken steps to identify and register transmigration roads, and assistance in absorbing and upgrading transmigration roads is urgently required in the districts that received large numbers of transmigrants. 10.4 There are a number of other improvements that can be made in trans- migration sites. For example, the introduction of livestock, fishponds and minor tree crops, the acceleration of land titling, and the improvement of research and extension should be encouraged. However, this report cautions that steps to ameliorate natural conditions, through liming and bench terrac- ing for example, may be too costly to be replicable on a large-scale. The report also suggests that integrated, multi-component programs will be diffi- cult to manage and it encourages an approach in which the MOT, in conjunction with the line agencies, identifies the needs of existing sites and contracts with the line agencies for these services. New Settlement 10.5 Program Consolidation. Fill-in programs which make use of existing infrastructure should have priority over new settlement in pioneer areas. Such programs, which involve the settlement of migrants within or adjacent to existing sites, will have the highest rate of return and can be carried out under partially-assisted programs if settlers are recruited among spontaneous migrants already in the area. Programs of consolidation will depend heavily on regional authorities for land identification and implementation, and are expected to contribute substantially to regional development. - 170 - 10.6 Choice of Farm Model. In the ongoing second-stage development program, and in any future new settlement, the following principles should guide the choice of farm model. (a) Annual Crops. Settlements based mainly on annual crops may be possible in limited areas with good soils, in swamp reclamation areas (d, below) and in fill-in areas where ample opportunities for wage work already exist. (b) Tree Crops. Tree crop programs, such as NES/PIR projects, have good prospects as they provide higher incomes and rates of return than projects based on annual crops in marginal areas and they have the potential for cost recovery. PMU programs can be tapped for new settlement, but priority for PMU schemes should be given to older sites. (c) Other Cash Crops. Cash crops such as sugar (Sumatra and Kalimantan) and cotton (SE Sulawesi) may have potential either in first or second stage development. However, most cash crops (fruits, vege- tables, soyabeans) will be more attractive in densely settled areas with better market prospects. (d) Swamp Reclamation. Settlement in areas of swamp reclamation has a number of attractive features. However, to realize the potential of swamp sites, steps are needed to improve agro-hydrological surveys, drainage, and land development techniques; to improve coordination and agricultural supporting services; to develop an intensive pest management strategy; and to address problems of potable water and malaria control. Partially-Assisted and Spontaneous Migration 10.7 In the near- and medium-term (Repelita IV and V) two principles must be borne in mind: first, pioneer settlement should be fully sponsored, as settlers in new areas will not have family, friends or well-developed institutions to support them. Second, partially-assisted migrants should be recruited among spontaneous migrants and second generation families already in the receiving areas. In the longer term, however, (Repelita VI) the success of transmigration program will depend on the ability of MOT to facilitate the settlement of individual families moving on their own. To promote this: (a) incentives should be developed to encourage local-level officials to identify, register, and transfer underutilized land to transmigrants and to arrange for financial compensation as required; (b) credit mechanisms should be developed to permit poor smallholders to purchase and register land; and (c) emphasis should be given to parallel programs for environmental protection in areas where spontaneous migrants are expected to settle. - 171 - To avoid the adverse effects of spontaneous movement, it is important that the MOT use targets only as guidelines and concentrate on the key policies required to facilitate settlement. Environmental and Social Factors in Transmigration 10.8 Environmental Soundness. A major effort is needed to identify, demarcate and gazette areas needed for conservation purposes. Improved maps are being prepared by Bakosurtonal, and additional specialist assistance could facilitate the completion of the land resource suitability maps needed to reconcile development priorities, forest boundaries, and environmental master plans. Even when maps are available, however, considerable work will still be required to assess areas of biotic importance and to demarcate areas for conservation. To facilitate this task, a consortium should be considered, consisting of donor countries and environmental groups willing to provide financial and technical support. Government officials and consortium members could decide on the most efficient ways of studying and gazetting important lands and determining the priority to be given to specific areas for protec- tion. Since wetlands appear to have good potential for settlement, these areas should also receive attention to ensure an adequate strategy for their development. 10.9 Social Soundness. Further steps need to be taken to provide bene- fits to local people in the vicinity of transmigration settlements and to link second-stage development to regional development. To improve the MOT's abili- ty to understand and assist local people in the vicinity of transmigration settlements, and to aid MOT in developing appropriate procedures for the provision of compensation and parallel benefits, either the Directorate for Social and Cultural Development should be strengthened or a new Directorate within the MOT should be developed to deal with matters relating to local people. To improve social assessment, the demographic departments of local universities might be used to undertake assessment studies and impact monitor- ing; and provincial and district officials in receiving provinces should be sensitized to the special issues concerning isolated people. Institutional Development 10.10 Management. The MOT is aware of implementation problems, but has very few staff in positions to address management issues. To address this problem, senior MOT officials should meet on a regular basis to review the core management functions (planning, budgeting, monitoring and evaluation) and make recommendations for their improvement. To improve the role of the central Directorates should be strengthened and contractor supervision should be improved. 10.11 Interagency Coordination. To improve interagency coordination: (a) The office of the Secretary General in the MOT should convene an interagency forum on a regular basis to identify implementation problems and evaluate progress. A secretariat should be formed to support the work of these meetings and issue regular reports. - 172 - (b) Consideration should be given to the formation of interagency policy development teams for agricultural development, forestry and the environment, social affairs, and spontaneous migration, and these teams should report regularly to the interagency forum or to the Minister. (c) Working secretariats should be formed within the provincial and district transmigration offices to convene interagency meetings and prepare materials. Increased budgets should be allocated for this purpose by Bappenas. (d) The MOT project head (pimpro) on each site should be given the authority to convene and coordinate other agency heads (pimpros), regardless of their location or status. 10.12 Planning and Budgeting. Under new budget procedures which allocate funds for only one year, much better planning must occur. Recently a special team has been formed to reconcile program and budget functions. High priority must be given to institutionalizing the functions of this team in the Office of the Secretary General, MOT. 10.13 Monitoring and Evaluation. There is an urgent need to strengthen monitoring and evaluation: (a) Office of the Secretary General (MOT). The monitoring system in the OSG should be greatly simplified. Only major indicators of physical and financial progress should be aggregated at this level to provide needed information for programming and budgeting. (b) Directorates General. Detailed monitoring of project implementation and impact should be carried out by the Directorates General and this material, where necessary, should be forwarded to the OSG. (c) Provinces. With the decentralization of procurement and implementa- tion responsibility, there should be increased emphasis on develop- ing appropriate monitoring and reporting capacity within the provinces, and clear decisions on that information to be channeled to the Directorates and the OSG. (d) Other Agencies. Improved monitoring and evaluation are also needed in the agricultural sector and could be carried out by a strengthened Agricultural Technical Team or by AARD. The capacity to undertake environmental monitoring should be developed under MPE. Technical support for monitoring and evaluation should be used to build an overall system operating at all these levels, and adequate budget should be provided for this purpose. The Future Transmigration Program 10.14 Due to financial constraints, implementation problems, and evidence of low farm incomes, Government is now concentrating on maintenance of settlement - 173 - infrastructure and second-stage development, and this policy is likely to continue through Repelita IV. During this period, emphasis should also be given to planning and preparation for Repelita V and to institutional development. 10.15 Forward planning for Repelita V should entail the preparation of base maps indicating those areas suitable for settlement and those with forest claims or other contingencies; the production of draft master plans indicating areas for upgrading and new settlement; the mix of farm models; and areas where sponsored, partially assisted and spontaneous migrants are to be settled. A major effort should also be given to the development of parallel master plans for environmental protection and the development of action plans to address implementation constraints. 10.16 Whether or not sponsored migration is resumed on a large scale, spontan- eous migration is expected to continue, and programs to channel spontaneous migrants onto good agricultural land are urgently required. The number of spontaneous migrants who will seek land in Repelita V is difficult to predict and will depend on economic factors, on the attractiveness of past settle- ments, and on the ability of the Government to establish policies which support land acquisition and land registration by individual families. - 175- ANNEX I Table 1 Repelita IIT Settlement by Year and Province (Families) Total Province 1979/80 1980/81 1981/82 1982/83 1983/84 Repelita Ill Sumatra Aceh 0 2,580 3,724 4,390 77 10,771 North Sumatra 0 855 1,986 1,086 4,139 8,066 West Sumatra 1,000 2,000 397 1,040 3,166 7,603 Riau 1,247 8,677 12,286 7,518 7,794 37,522 Jambi 2,068 4,183 3,775 2,584 4,072 16,682 South Sumatra 8,234 22,948 24,673 18,162 17,323 91,340 Bengkulu 2,022 2,596 1,679 3,611 2,279 12,187 Lampung 1,813 5,204 0 19,187 16,672 42,876 Subtotal 16,384 49,043 48,520 57,578 55,522 227,047 Kalimantan West Kalimantan 0 1,877 6,420 6,478 366 15,141 Central Kalimantan 250 2,228 4,072 4,780 16,891 28,221 South Kalimantan 315 5,769 4,033 3,676 1,581 15,374 East Kalimantan 0 2,102 2,853 2,554 4,369 11,878 Subtotal 565 11,976 17,378 17,488 23,207 70,614 Sulawesi South Sulawesi 450 650 750 1,220 537 3,607 Central Sulawesi 2,000 2,372 5,582 3,273 2,513 15,740 Southeast Sulawesi 1,010 3,928 5,895 4,589 3,803 19,225 worth Sulawesi 75 1,216 1,667 366 830 4,154 Moluku/others 319 1,497 1,996 2,948 2,214 8,974 Subtotal 3,854 9,663 15,890 12,396 9,879 51,700 Irian Jaya 290 2,521 2688 5,362 5,755 16,616 Total 22,093 77,283 98,405 104,237 63,959 365,977 Unassisted Migrants 1,985 3,359 8,961 32,445 122,747 169,497 Grand Total 24,078 80,642 107,366 136,682 186,706 535,474 Source: Ministry of Transmigration. - 176 - ANNEX 1 Table 2 Repelita IV Settlement by Type of Migrant Family and Province Through June 30, 1987 Partially Province Sponsored assisted Spontaneous Total Sumatra Aceh 5,855 40 6,045 11,940 North Sumatra 3,575 285 7,025 10,885 West Sumatra 2,440 310 5,855 8,605 Riau 24,160 1,990 22,465 48,615 Jambi 19,115 555 18,870 38,540 South Sumatra 20,705 4,040 57,695 82,440 Bengkulu 6,800 150 10,350 17,300 Lampung 16,465 1,255 55,545 73,265 Subtotal 99,115 8,625 183,850 291,590 Kalimantan West Kalimantan 13,100 2,035 16,760 31,895 Central Kalimantan 13,145 2,065 17,955 33,165 South Kalimantan 9,575 1,210 15,590 26,375 East Kalimantan 9,125 495 23,195 32,815 Subtotal 44,945 5,805 73,500 124,250 Sulawesi South Sulawesi 3,275 370 4,280 7,925 Central Sulawesi 8,905 1,065 3,965 13,935 Southeast Sulawesi 6,640 550 4,145 11,335 North Sulawesi 2,515 5 2,190 4,710 Moluku/Others 4,160 460 4,150 8,770 Subtotal 25,495 2,450 18,730 46,675 Irian Jaya 9,400 40 5,125 14,565 Total 178,955 16,920 281,205 477,080 Source: Ministry of Transmigration. -177 - ANNEX 2 Table I Sites Covered in BPS Survev Households Arrival On February Sample Critical Site Province date arrival 1985 size Farm layout site Repelita II Sites Tidal Sei Rateh Riau 07/78-06/80 714 705 39 nucleated/combined No Upang Delta S. Sumatra 09/74-03/75 525 546 40 linear No Tanah Grogot E. Kalimantan 04/73-02/76 151 202 20 linear No Subtotal 1,390 1,453 99 Upland Pematang Panggang S. Sumatra 05/75-12/76 3,633 4,219 201 linear No Babulu Darat F. Kalimantan 01/79-05/79 933 1,173 40 linar Ves Sepaku F. Kalimantan 07/75-07/76 537 362 2n combined No Malonas C. Sulawesi 12/76-02/76 980 926 100 linear/nucleated No Subtotal 6,143 6,680 361 Upland with Trees Baturaja S. Sumatra 10/76-06/7( 1,7n2 2,524 100/a linear No Subtotal 1,702 2,524 100 Repelita II Total 9,235 10,657 560 Repelita III Sites Tidal Air Sugiban S. Sumatra 04/80-08-84 4,486 4,506 221 linear Yes Terusan Tengah C. Kalimantan 03/80-02/82 1,600 1,657 120 nucleated No Pangkoh C. Kalimantan 04/80-01/85 4,906 5,019 140 nucleated No Saka Lagon/ Sunggai Muhuir F. Kalimantan 11/79-10/81 769 780 40 linear/nucleated No Stubtotal 11,761 11,962 521 Upland Belias Riau 01/81-11/84 4,010 3,877 180 nucleated Yes Teluk Kuantan Riau 01/80-12/84 4,176 4,122 60 linear/nucleated No Pematang Panggang S. Sumatra 02/80-06/83 6,039 6,640 281 linear/nucleated No Sunggai Waras S. Sumatra 02/82-06/82 800 818 59 nucleated No Natulicin S. Kalimantan 03/80-08/82 4,239 4,309 178 linear/nucleated No Kumai/Hanjalipan C. Kalimantan 06/80-11/80 547 294 20 combined No Teluk Dalam F. Kalimantan 03/81-04/81 1,925 1,81q 159 nucleated Yes Lahumbuti S.F. Sulawesi - - - 60 linear/nucleated No Subtotal 21,736 21,839 997 Upland with Trees Baturaja S. Sumatra 04/79-06/81 1,731 2,077 80 linear No Tepong Tandung /b Riau 04/82-11/84 500 500 20 linear No Batulicin /h S. Kalimantan 01/82-04/84 1,203 1,054 20 linear No Subtotal 3,434 3,671 120 Repelita III Total 36,031 37,432 1,638 TOTAL 46,166 4P,089 2,198 /a Baturaja 19 housholds classified as estate by BPS, the others as upland. /b PIR Projects. Comparison of Income, Net Value of Production and Expenditure in Transmigrant Communities and in Sending and Receiving Provinces Mean Net value of prod. Net value of Mean monthly _(Rp'000) production (X) Expenditure (X) Exp. as income Sample income Food Tree Other Food Tree Other Non- % of in 1985 size (Rp) crops crops agri. crops crops agri. Food food income prices Sending Areas - 1984 West Java 3,452 67,702 11,776 2,568 1,756 17 4 3 70 30 90 73,660 Central Java 3,527 53,537 15,062 4,495 3,895 28 8 7 66 34 85 58,248 U.K. Yogyakarta 1,275 66,970 14,956 1,892 5,275 22 3 R 62 3R 83 72,863 East Java 4,155 58,764 14,885 2,890 4,637 25 5 8 63 37 84 63,935 Bali 1,461 56,208 20,241 6,389 10,710 26 8 14 66 34 78 83,R27 NTB 1,506 61,738 14,325 1,989 5,686 25 4 10 70 30 83 61,154 Total 17,241 61,738 13,099 2,985 4,060 21 5 7 66 34 85 67,171 Receiving Areas - 1984 P Riau 380 104,951 11,147 48,401 19,776 11 46 19 75 25 84 114,187 co South Sumatra 883 79,305 21,959 14,194 4,699 28 18 6 74 26 95 86,284 1 Central Kalimantan 271 76,894 22,642 15,082 8,344 29 20 11 75 25 91 83,661 South Kalimantan 229 75,091 26,835 1,649 12,605 36 2 17 76 24 95 81,699 East Kalimantan 393 89,865 15,556 4,708 9,681 17 5 11 72 28 86 97,773 Central Sulawesi 195 58,407 23,170 13,522 15,318 40 23 26 82 18 79 63,547 Southeast Sulawesi 480 86,346 13,515 10,905 13,592 16 13 16 70 30 81 93,944 Total 2,831 83,396 18,730 15,935 10,647 22 19 13 74 26 88 sn,735 Transmigrants in Receiving Provinces - 1985 Riau 300 65,513 12,624 436 8,099 19 1 12 75 25 95 71,278 South Sumatra 981 52,888 11,955 2,805 5,527 23 5 10 78 22 87 57,542 Central Kalimantan 320 39,131 8,483 25 7,679 22 0 20 84 16 97 42,575 South Kalimantan 239 55,369 11,905 150 8,240 22 0 15 78 22 9] 6n,241 East Kalimantan 100 107,893 38,606 5,021 18,786 36 5 17 77 23 63 117,388 Central Sulawesi 100 60,231 10,470 817 4,717 17 1 8 74 26 78 65,531 Southeast Sulawesi 160 89,062 19,360 533 15,106 22 1 17 75 25 57 96,899 Total 2,200 58,343 13,218 1,661 7,748 23 3 13 78 22 84 63,477 Sources: Sending and receiving areas - 1984 Susenas; transmigration areas - BPS Transmigration Income Survey, 1985. Source of Tncome hv Farm Model and Location Sfte inc. as Non- Gov't Non- r.ov t a 7 of ave. Annual Site Sample Agric. Wages agric. asat. Transfer Total Agric. Wages agric. asst. Transfer income /a 11S8 Repelita rI Sites Tidal Sel Rateh 34 41,045 7,077 14,391 1,40R (2,8R7) 61,034 67 1? 24 2 (5) 105 666 11pang Delta 40 26,811 13,586 17,966 0 18,529 76,890 35 18 23 0 24 132 R39 Tanah Grogot 20 28,992 37,250 45,5n0 61 n 111,9113 26 33 41 n n 192 1,220 Subtotal 99 32,659 15,803 22,120 567 6,348 77,607 42 20 2R 1 A 134 R4R lJpland Pematang Panggang 201 8,351 14,035 11,961 3,858 3,776 41,581 90 34 26 Ft 72 454 Babujlu Darat/Sepaku 40 42,238 31,413 6,700 6,539 3,365 92,255 46 34 7 9 4 199 1,006 Malonas 100 18,068 14,935 14,0R6 2,579 9,976 59,648 30 25 23 4 17 103 656 Subtotal 341 15,176 16,337 11,731 4,032 5,547 92,623 29 31 22 6 i1 In 57R UTpland w/Trees gaturaja 100 32,761 9,034 16,560 4,415 1,R05 66,995 49 14 2F 7 3 115 726 Subtotal 0 32,761 9,"34 1A,580 4,415 1,605 66,599 49 14 2R 7 3 115 726 Repelita II Total 54n 21,674 14,887 14,904 3,468 5,001 99,934 36 25 25 6 8 103 695 15 Repelita III Sites Tidal Air Sugihan 221 17,097 3,474 19,320 5,155 (461) 44,986 36 6 43 12 (1) 77 486 Terusan Tengab 120 16,832 9,862 9,261 2,212 2,550 36,717 46 16 25 6 7 63 401 Pangkoh 120 16,165 1,667 8,487 7,722 412 34,995 47 5 25 22 1 99 376 Saka Lagon/Sunggai Muhur 40 9,628 16,075 3,751 9,OQO 3,9R7 38,531 25 42 10 13 10 66 420 Subtotal 501 16,214 4,619 13,073 5,06' R24 3c,791 41 12 33 13 2 66 434 IJpland Belias 140 11,049 17,70(9 0,403 ln,6R7 5,967 54,415 20 33 17 20 1n 94 S94 Teluk Kuantan 60 30,447 29,953 16,(08 9,355 720 R6,494 35 15 19 11 1 149 943 Pematang Panggang 261 23,RO8 12,122 14,071 2,603 1,310 53,914 44 92 26 5 2 93 558 Sunggai Waras 59 16,353 30,483 8,415 9,030 5,936 7n,210 23 43 12 13 6 121 766 Central Kalimantan 60 19,644 5,61R 21,343 10,358 791 97,954 34 10 37 16 I i00 632 Ratulicin 176 18,283 9,410 18,603 9,605 3,620 59,792 31 16 31 16 6 103 657 Teluk Dalam 20 50,193 13,700 11,617 144 275 75,869 66 I 15 0 n 131 62R Lahumbuti 179 35,091 15,079 17,419 23,251 1,663 92,524 3R 16 19 25 2 159 1,009 Subtotal Q77 23,294 14,R07 15,051 10,049 7,509 6S,R')1 35 23 23 15 4 113 716 Upland w/Trees Batiraja R0 20,646 11,711 14,652 9,3n0 2,256 58,57n 35 2n 25 16 4 101 639 Tepong Tandung 20 14,979 61,903 11,102 3,333 R,925 99,642 19 62 11 3 9 172 1,nR9 6atulicin 2n 5,999 41,3O0 2,667 0 (551) 49 ,415 12 84 5 0 ( 65 S39 3 Subtotal 120 17,195 25, 28 02,n63 6,75R 2 0900 63,473 27 39 19 11 S 110 697 F_ Repelita ITI Total 1,598 20,616 12,37Q 14,207 6,236 2,065 57,9505 36 22 25 14 4 99 627 Total 2,138 20,693 13,012 14,383 7,n33 2,807 S8,11R 36 22 25 12 5 100 634 Sou rce: RPS Transmigration Income Survev, 1989. - 180 - ANNEX 3 ANNEX 3 - TABLE OF CONTENTS Table 1 Price Structures for Agricultural Inputs and Outputs Table 2 Price Structures for Tree Crops Table 3 Economic and Financial Analysis of Upland Low Input Food Crop Model Table 4 Economic and Financial Analysis of Upland Low Input with Cattle Model Table 5 Economic and Financial Analysis of Upland Diversified Food Crop Model Table 6 Economic and Financial Analysis of Upland Sustained Input Food Crop Model Table 7 Economic and Financial Analysis of Upland High Input Labor Intensive Foodcrop Model Table 8 Economic and Financial Analysis of Upland High Input with Cattle Food Crop Model Table 9 Economic and Financial Analysis of Tidal Low Input with Severe Pest Problems Table 10 Economic and Financial Analysis of Tidal Low Input with Reasonable Pest Control Table 11 Economic and Financial Analysis of Tidal with Improved Food crops Table 12 NES/PIR Rubber Economic Analysis, 2 Ha Rubber, 0.5 Ha Houselot and Garden Area. Table 13 NES/PIR Oil Palm Economic Analysis, 2 Ha Oil Palm, 0.5 Ha Houselot and Garden Area Table 14 NES/PIR Coconuts Economic Analysis, 2 Ha Hybrid Coconuts, 0.5 Ha Houselot and Garden Area Table 15 PMU-Planted Rubber Economic Analysis, 2 Ha Rubber, 0.5 Ha Houselot and Garden Area Table 16 PMU-Planted Rubber as Second Stage Development for Existing Transmigrants, Rubber Planted on 1.5 Ha Table 17 Projected Paddy Yields at Appraisal from World Bank-Financed Trans- migration Projects and Yields in Models Presented in Chapter III. ANNEX 3 - 181 - Table 1 Page 1 of 2 Price Structures for Agricultural Inputs and Outputs /a (1986 constant prices) Operation 1986 1987 1990 1995+ Rice (Prices used in economic analysis are import parity-based through 1984. From 1985 onwards, they are an unweighted average of the import and export parity-based price.) Import Parity Price Export price, Thai 5% brokens, FOB Bangkok (US$/ton) 205 178 362 342 Quality adjustment 90% 184 160 326 308 Freight and insurance + 32 32 32 32 CIF price Palembang = 216 192 360 340 Port handling and storage losses /b + 26 26 26 26 Transfer to wholesalers + 4 4 4 4 Transfer mills to wholesalers - 3 3 3 3 Ex-mill price Palembang = 250 225 391 367 Conversion to paddy 60% 150 135 235 220 Milling and cleaning costs - 7 7 7 7 Paddy transport to Palembang by wholesaler - 24 24 24 24 Economic farm-gate price (US$/ton) = 119 104 204 189 Economic farm-gate price (Rp/kg) = 130 115 224 208 Export Parity Price Export price, FOB Jakarta (US$/ton) 205 178 362 342 Quality adjustment 90% 184 160 326 308 Port handling, storage and losses - 26 26 26 26 Transfer to wholesalers - 4 4 4 4 Transfer mills to wholesalers - 3 3 3 3 Ex-mill price Palembang = 152 127 293 275 Conversion to paddy 60% 91 76 176 165 Milling and cleaning costs - 7 7 7 7 Paddy transport to Palembang by wholesaler - 24 24 24 24 Economic farm-gate price (US$/ton) = 60 45 145 134 Economic farm-gate price (Rp/kg) = 66 50 160 147 Average of Import and Export Parity Price = 98 83 192 178 Financial Farm-gate Price (Rp/kg) (average price received by farmers, not BULOG floor price) = 120 150 200 200 Maize Export price, FOB US Gulf port (US$/ton) 98 79 121 118 Adjusted export price, FOB Palembang /c + 15 15 15 15 Port handling + 10 10 10 10 Transport wharf to wholesaler + 6 6 6 6 Transport farm to wholesaler - 21 21 21 21 Economic farm-gate price (US$/ton) = 108 89 131 128 Economic farm-gate price (Rp/kg) = 118 98 144 141 Financial farm-gate price (Rp/kg) 130 130 130 140 Cassava Economic and financial farm-gate price US$/ton 17 17 17 17 Rp/kg 19 19 19 19 Groundnuts Export price, CIF Rotterdam (US$/ton) /d 345 301 336 324 Freight and insurance + 25 25 25 25 CIF price, Palembang = 370 326 360 349 Port handling, storage, transport to wholesaler + 16 16 16 16 Transport, farm to wholesaler - 24 24 24 24 Economic farm-gate price (US$/ton) = 362 318 352 341 Economic farm-gate price (Rp/kg) = 398 350 387 375 Financial farm-gate price (Rp/kg) = 450 450 450 450 - 182 - ANNEX 3 Table 1 Page 2 of 2 Operation 1986 1987 1990 1995+ Urea World price, FOB Europe (US$/ton) 130 159 206 220 Premium for Asian markets + 16 16 16 16 Economic price, FOB Palembang = 146 175 222 236 Port handling + 10 10 10 10 Ex-factory economic price (US$/ton) 156 185 232 246 Ex-factory economic price (Rp/kg) = 172 203 255 271 Handling, transport and storage - factory to farm + 40 40 40 40 Economic farm-gate price (Rp/kg) = 212 243 295 311 Financial farm-gate price (Rp/kg) = 120 130 150 200 TSP World price, FOB Florida (US$/ton) 140 140 144 158 Freight and insurance to Palembang + 35 35 35 35 Handling, transpor, storage - port to farm + 40 40 40 40 Economic farm-gate price (US$/ton) = 215 215 219 233 Economic farm-gate price (Rp/kg) = 236 236 241 256 Financial farm-gate price (Rp/kg) = 120 130 150 200 Pesticides Economic farm-gate price (Rp/kg) 7,000 7,000 7,000 7,000 Financial farm-gate price (Rp/kg) 1,500 1,500 3,000 5,000 1986 1987+ Labor /e Hired Labor Financial daily wage rate Rp 1,200 1986 wage rate grows in real terms with projected growth in per capita gross domestic income. This growth is assumed to be: 1987 = 1.3% 1988 = 1.1% 1989 = 2.8% 1990 = 2.5% 1991-95 = 3.5% 1996+ = 3.7% Economic daily wage rate 0.65 * financial daily wage rate Family Labor Economic Shadow-priced estimate of income foregone in Java. This is assumed to be Rp 222,000 per year in 1986, growing thereafter in real terms with growth in per capita gross domestic income. /a All economic prices except cassava, pesticides, and unskilled labor based on World Bank price projections as presented in the "Half-Yearly Revision of Commodity Price Forecasts, January 1986. Prices converted at an exchange rate of US$1 = Rp 1,100. Prices are smoothed between the years shown on the Table. /b Derived from BULOG records. /c Includes a US$15 transport premium for Asian markets. /d The Bank no longer publishes prices for groundnuts. These prices have been derived from prices for groundnut oil at assumed extraction rates from kernel of 46%. /e See Chapter 3 for discussion of how labor is handled in the economic and financial analysis of the food crop models. ANNEX 3 - 183 - Table 2 Page 1 of 3 Price Structures for Tree Crops /a Projected Palm Oil, Kernel, and FFB Financial and Economic Prices (Rp/kg in 1986 constant prices) Palm oil Palm kernels Export value Smallholder FFB FOB /b FOB /c of FFB /d Export /e,f 1985 519 303 120 60 1986 266 117 60 30 1987 292 119 65 33 1990 581 178 128 64 1995+ 565 195 126 63 /a Based upon World Bank Commodity Price Forecasts, January 1986. Prices for the years not shown (1988, 1989, 1991-1994) are assumed to move to the next projection linearly. /b IBRD commodity forecasts adjusted for freight and insurance to obtain FOB, or US$50/ton less. /c FOB price adjusted from CIF prices by deducting US$65/ton. /d By virtue of extraction rates of 21% for palm oil and 3.5% for kernels. /e Deductions equivalent to 50% of FFB value. /f Economic price assumed to be 10% higher. ANNEX 3 - 184 - Table 2 Page 2 of 3 Projected Copra Financial and Economic Prices /a (Rp/kg in 1986 constant prices) Copra Ex-factory Farm-gate prices /c Rp/kg price /b Paid by CIF NW Europe (10% moisture) PTPs/NES /d SCDP /e 1985 443 264 218 185 1986 303 124 103 87 1987 274 96 79 67 1988 343 165 136 116 1989 412 234 193 164 1990 481 303 250 212 1991 476 299 246 209 1992 472 295 243 206 1993 468 290 239 203 1994 464 286 236 200 1995 460 282 262 197 /a Based on World Bank Commodity Price Forecasts, January 1986. /b Inter-island freight, port handling, adjustment for losses, processing costs, adjustment for mosture content, US$162/t in 1986. /c Conversion to economic prices is 5% higher. /d Deduction of general charges, selling expenses, and profit margin or 17.5% of ex-factory price. /e Prices paid by middlemen and KUD, etc. to smallholders about 70% of ex- factory prices. - 185 - ANNEX 3 Table 2 Page 3 of 3 Projected Rubber Financial and Economic Prices /a (Rp/kg in 1986 constant prices) CIF New York FOB Sumatra RSS1 RSSI /c Farm-gate prices /b (US/kg) (Rp/kg) NES /d SRDP /e 1985 96 918 617 543 1986 95 908 610 537 1987 93 891 598 527 1988 106 1,033 694 611 1989 119 1,175 790 695 1990 132 1,318 885 779 1991 133 1,328 892 785 1992 134 1,338 899 791 1993 135 1,348 906 797 1994 136 1,358 912 803 1995+ 137 1,368 919 809 /a Based on World Bank Commodity Price Forecasts, January 1986. /b Economic prices are assumed to be 7% higher when withholding and other taxes are added back to FOB values. /c Freight and insurance paid at US cents 12.5/kg. /d Four percent discount from RSS1 to account for SIR 10-20-50 mix as com- pared with RSS1 prices and 30% discount as smallholders assumed to receive 70% of maximum FOB sales value of the product. /e Assumed to receive 12% less than NES farmers to account for production of SIR 20 and 50 grades and profits to middlemen. ECONOMIC AND FINANCIAL ANALYSIS OF UPLAND LOW INPUT FOODCROP MODEL (all prices in 1906 constant Rp) Year 1979 1980 1981 1982 1933 1984 1985 1986 19S7 199E 1989 1990 1991 1992 1993 1994 1995 1996 197 1998+ Commodities Rice Yield (h/ha) 1 0.9 0.8 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 Area (ha) 0.4 0.6 0.8 1 I I I I I I I I I I I I I Economic Price (Rp/kg) 184 176 165 117 98 03 119 15 192 189 186 183 180 178 179 179 178 Financial Price (Rp/kg) 152 149 152 127 125 150 150 150 190 190 190 190 190 180 1t0 180 180 Corn Yield (t/ha) 0.7 0.6 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 Area (ha) 0.1 0.15 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Econosic Price (Rp/kg) 135 167 170 149 110 98 113 128 144 143 143 142 142 141 141 141 141 Financial Price (Rp/kg) 120 140 150 130 130 130 130 130 130 130 130 130 130 140 140 140 140 Cassava Yield (t/ha) 6 6 U 6 6 6 6 6 6 6 6 6 6 6 6 6 6 Area (ha) 0.15 0.25 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 Economic Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Financial Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Peanuts Yield It/ha) 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 Area (ha) 0.05 0.06 0.00 0.08 0.00 0.08 0.08 0.00 0.08 0.08 0.00 0.08 0.08 0.08 0.08 0.08 0.08 Economic Price (Rp/kg) 469 504 501 535 39B 350 362 374 387 385 382 380 378 375 375 375 375 Financial Price (Rp/kg) 400 400 430 450 450 450 450 450 450 450 450 450 450 450 450 450 450 Vegetables and Fruit ('000 Rp) 50 60 75 75 15 90 90 100 100 100 100 100 120 120 120 120 120 Small Livestock ('000 Rp) 12 18 24 30 36 36 36 36 36 36 36 36 36 36 36 36 36 Costs of Production Planting Material ('000 Rp) 9.2 13.3 18.1 20.4 20.4 20.4 20.4 20.4 20.4 20.4 20.4 20.4 20.4 20.4 20.4 20.4 20.4 Urea (t) (first 3 years provided free by 601) 0.1 0.1 0.1 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 Economic Cost-Rp/kg 250 223 269 266 212 243 259 275 295 298 302 305 308 311 311 311 311 Financial Cost-Rp/kg 99 96 104 95 120 130 ISO 150 ISO ISO 150 150 I50 200 200 200 200 TSP (t) (first 3 years provided free by 601) 0.1 0,1 0.1 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 1 Economic Cost-Rp/kg 240 230 237 222 236 236 238 239 241 244 247 250 253 256 256 256 256 00 Financial Cost-Rp/kg 99 96 104 95 120 130 150 150 150 ISO 150 ISO 150 200 200 200 200 Pesticide (kg) 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Economic Cast-('000 Rp/kg) 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 Financial Cost-(OVOO Rp/kg) 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 3 3 3 3 5 5 5 5 Hired Labor (days) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Total Labor Required (days) 200 200 10 I1O 160 160 160 160 160 160 160 160 160 160 160 160 160 Financial Cost of Hired Labor ('000 Rp/day) 1.20 1.20 1.20 1.20 1.20 1.22 1.23 1.26 1.29 1.34 1.39 1.44 1.49 1.54 1.59 1.65 1.72 Other ('000 Rp) 45 45 5 5 5 5 5 5 5 S S S S S S 5 5 Production (t) Rice 0.400 0.540 0.640 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 0.750 Corn 0.070 0.090 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 0.100 Cassava 0.900 1.500 1.8 1.000 1.000 1.000 1. 800 I1.8OO 1.800 1.800 1.800 1.800 1.800 E .800 1.800 1.800 1.800 Peanuts 0.013 0.0(5 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 Economic Value of Field Crops ('000 Rp) 106 147 167 14B 127 113 142 171 200 198 196 193 191 189 189 189 189 Financial Value of Field Crops ('000 Rp) 91 128 155 151 ISO 169 169 169 199 199 199 199 199 192 192 192 192 Dross Fare Income ('000 Rp) Economic 168 225 266 253 230 239 268 307 336 334 332 329 347 345 345 345 345 Financial 153 206 254 256 261 295 295 305 335 335 335 335 355 348 348 348 348 Total Production Costs ('000 Rp) Pt nz 12 Economic, Including Income Foregone, Java 340 341 310 281 280 284 207 294 301 310 319 328 337 347 358 369 380 n Financial, Including only Hired Labor 77 01 47 36 38 39 40 40 40 43 43 43 43 51 51 51 51 AGRICULTURE ALONE Net Economic Farm Incoee ('000 Rp) (excludes economic value of 601 eupenditures) -172 -116 -44 -29 -41 -45 -19 13 35 24 13 1 10 -2 -12 -23 -35 Net Return tn Household ('000 Rp) , 96 144 228 220 223 256 254 264 294 291 291 291 311 297 297 297 297 AGRICULTURE AND OFF FARM WORK Off Fare Income: Financial (Rp '000) 174 190 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 Off Farm Income: Economic (Rp 000) 15B 173 200 200 200 203 205 211 216 223 231 239 248 256 266 276 286 Net Economic Household Incoae IRp '000) -14 57 156 171 159 158 186 223 251 248 244 241 257 254 253 252 251 Net Return to Household (Rp 000) 270 334 448 440 443 479 480 496 532 537 546 555 584 579 589 600 611 OTHER NOT DEVELOPMENT COSTS PER HOUSEHOLD ('000 Rp) Base Case 472 2227 1657 10 10 9 9 More Remote 590 2784 2072 12 12 11 11 Very Remote 708 3340 2486 IS 15 13 13 ECONOMIC NET BENEFIT STREAM, A6RICULTURAL PRODUCTION ('000 Rp) Base Case -472 -2227 -1657 -IBI -125 -53 -38 -41 -45 -19 13 35 24 13 1 10 -2 -12 -23 -35 More Remote -590 -2784 -2072 -184 -128 -56 -40 -41 -45 -19 13 35 24 13 1 10 -2 -12 -23 -35 Very Remote -708 -3340 -2486 -186 -130 -5 -42 -41 -45 -19 13 35 24 13 1 10 -2 -12 -23 -35 IRR AND NPV AT 10X IRR NPV ('000 Rp) Base Case negative -3826 More Remote negative -4711 Very Remote negative -5595 (calculated over a thirty year period) ECONOMIC NET BENEFIT STREAM, AGRICULTURAL PRODUCTION AND OFF FARM WORK (000 Rp) Base Case -472 -2227 -1657 -24 47 147 162 159 15B 186 223 251 248 244 241 257 254 253 252 251 More Remote -590 -2784 -2072 -26 45 144 160 159 158 186 223 251 248 244 241 257 254 253 252 251 Very Remote -708 -3340 -2486 -28 42 142 158 159 158 186 223 251 246 244 241 257 254 253 252 251 1 IRR AND NPV AT 101 IRR NPV ('000 Rp -. Base Case 2X -2311 These IRRs and NPVs assume that the full cost of infrastructure investments is inputted to the project. More Remote 02 -3195 In fact, part of the benefits accrue to the local population and spontaneous migrants (see para 3.05 Very Reeote -1 -4080 in main text). IRRs which take these factors into account are shown in the text (Table 3.4). (calculated over a thirty year period) Notes (values in '000 Rp unless otherwise noted) Total Family Labor Supply (Days): 350 364 379 394 409 426 443 461 479 49B 518 518 518 518 518 518 518 Potential Off Farm Incose (1): Financial 220 220 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 Potential Off Fare Incone (I): Economic 200 200 200 200 200 203 205 211 216 223 231 239 248 256 266 276 286 Family Labor valued at subsistence income in Java 270 270 270 270 270 274 277 284 291 302 312 323 334 346 359 372 386 Subsistence Incone in Java, Economic terms 222 222 222 222 222 225 228 234 240 248 257 266 27t 285 296 307 318 Assumptions about growth in per capita GDY: 0.02 0.02 0.02 0.02 0.02 1.3X 1.12 2.62 2.52 3.52 3.51 3.52 3.52 3.52 3.71 3.72 3.7X Total labor available for off farm Nork after farming activities (days) (I): 150 164 199 214 249 266 283 301 319 338 35B 358 358 358 358 358 358 Off fare income, economiuc 158 173 200 200 200 203 205 211 216 223 231 239 248 256 266 276 286 Off farm income, financial 174 190 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 (1) Maximus amount of off farm work available assumed to be 190 days. inl-. ' ECONOHIC AND FINANCIAL ANALYSIS OF UPLAND LOW INPUT hODEL WITH COW (all prices in 1986 constant Rp) Year 1979 1980 1981 1982 1983 1984 1985 1986 1987 1980 1999 1990 1991 1992 1993 1994 1995 1996 1997 1998+ Commodities Rice Yield (t/ha) I 1 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 Area (.2 ha rice replaced w/ forage) 0.4 0.6 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 Economic Price (Rp/kg) 184 176 165 117 98 83 119 15 192 189 186 183 180 178 178 178 178 Financial Price (Rp/kg) 152 149 152 127 125 150 150 150 190 190 190 190 190 180 190 180 180 Corn Yield (t/ha) 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 Area (ha) 0.1 0.15 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Econoeic Price (Rp/kg) 135 167 170 149 118 98 113 128 144 143 143 142 142 141 141 141 141 Financial Price (Rp/kg) 120 140 150 130 130 130 130 130 130 130 130 130 130 140 140 140 140 Cassava Yield (t/hal 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 Area Ihal 0.15 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 Econoeic Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Financial Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Peanuts Yield (t/ha) 0.25 0.25 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 Area (hal 0.05 0.06 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08 Economic Price (Rp/kg) 469 584 501 535 398 350 362 374 387 3895 382 380 378 375 375 375 375 Financial Price (Rp/kg) 400 400 430 450 450 450 450 450 450 450 450 450 450 450 450 450 450 Vegetables and Fruit ('000 Rp) 50 60 75 75 75 90 90 100 100 (00 100 100 120 120 120 120 120 Small Livestock and Cattle Offspring ('000 Rp) 12 18 24 30 36 60 36 84 350 225 201 221 217 259 257 252 252 Costs of Production Planting Material (incl forage seed) ('000 Rp) 9.2 13.3 23.1 25.4 25.4 25.4 25.4 25.4 25.4 25.4 25.4 25.4 25.4 25.4 25.4 25.4 25.4 Urea (t) (first 3 years provided free by G08) 0.1 0.1 0.1 0.04 0,04 0.04 0.04 0.04 0.04 >04 0.04 0.04 0.04 0.04 0804 0.04 0.04 Econoemic Cost-Rp/kg 250 223 269 266 212 243 259 275 295 298 302 305 308 311 311 311 311 Financial Cost-Rp/kg 99 96 104 95 120 130 190 190 10 150 150 ISO ISO 200 200 200 200 TSP (t) (first 3 years provided free by 601) 0.1 0.1 0.1 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 0.04 1 Econoeic Cost-Rp/kg 240 238 237 222 236 236 238 239 241 244 247 250 253 256 256 256 256 I1 Financial Cost-Hp/kg 99 96 104 95 120 130 150 150 150 150 150 190 190 200 200 200 200 Pesticide (kg) 8 8 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 Economic Cost-('OOO Rp/kg) 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 1 Financial Cost-('000 Rp/kg) 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 3 3 3 3 5 5 5 5 Hired Labor (days) 0 0 0 0 0 0 0 0 0 8 0 0 0 0 0 O 0 Total Labor (days) 200 200 170 170 150 150 150 150 150 150 150 150 150 190 150 150 198 Financial Cost of Hired Labor ('000 Rp/day) 1.20 1.20 1.20 1.20 1.20 1.22 1.23 1.26 1.29 1.34 1.39 1.44 1.49 1.54 1.59 1.65 1.72 Other (including plow and medicine) ('000 Rp) 45 45 30 to 10 10 10 o1 10 10 10 10 10 10 10 10 10 Production (t) Rice 0.400 0.600 0.960 0.960 0.960 0.960 0.960 0.960 0.960 0.960 0.960 0.960 0.960 0.960 0.960 0.960 0.960 Corn 0.070 0.105 0.140 0.140 0.140 0.140 0.140 0.140 0.140 0.140 0.140 0.140 0.140 0.140 0.140 0.140 0.140 Cassava 0.900 1.500 1.500 1.500 1.500 t.500 1.900 1.900 1.900 1.500 1.900 1.900 1.900 1.500 1.900 1.900 1.00 Peanuts 0.013 0.015 0.024 0.024 0.024 0.024 0.024 0.024 0.024 0.024 0.024 0.024 0.024 0.024 0.024 0.024 0.024 Economic Value of Field Crops ('000 Rp) 106 160 223 175 149 130 167 204 242 239 236 233 230 228 228 228 229 Financial Value of Field Crops ('000 Rp) 91 139 206 179 178 202 202 202 240 240 240 240 240 232 232 232 232 Gross Fare Income (1000 Rp) Economic 168 238 322 280 260 290 293 38B 692 564 537 554 567 607 605 600 600 Financial 153 217 305 284 289 352 328 386 690 565 541 561 577 611 609 604 604 O H Total Production Costs ('000 Rp) (D i: Economic, Including Incoee Foregone, Java 382 383 354 305 304 30B 311 318 325 334 343 352 361 371 382 393 404 Financial, including only Hired Labor B6 90 80 49 51 52 53 53 53 59 59 59 59 71 71 71 71 0 AGRICULTURE ALONE Net Economic Farm Incose ('000 Rp) lexcludes econosic value of 681 expenditures) -.214 -144 -32 -26 -44 -2B -to 70 367 231 199 202 206 236 223 207 196 Net Return tD Household ('000 Rp) 87 146 246 235 238 300 274 332 637 506 482 502 518 539 537 532 532 ASRICULTURE AND OFF FARN WORK Off Fare Income: Financial (Rp '000) 174 190 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 Off Fare Income: Economic (Rp 000) 158 173 200 200 200 203 205 211 216 223 231 239 248 256 266 276 286 Net Econoaic Household Incose (Rp 000) -56 28 168 174 156 175 187 281 583 454 426 442 454 492 489 483 482 Net Return to Household (Rp 000) 261 336 466 455 458 523 499 564 874 751 736 765 790 821 830 836 847 OTHER NOT DEVELOPMENT COSTS PER HOUSEHOLD ('000 Rp) Base Case 472 2227 1657 294 10 9 9 More Remote 590 2784 2072 296 12 11 11 Very Resote 708 3340 2486 299 15 13 13 ECONOMIC NET BENEFIT STREAM, AGRICULTURAL PRODUCTION ('000 Rp) Base Case -472 -2227 -1657 -507 -154 -41 -35 -44 -28 -18 70 367 231 195 202 206 236 223 207 196 More Resote -590 -2784 -2072 -510 -157 -44 -37 -44 -28 -18 70 367 231 195 202 206 236 223 207 196 Very Remote -708 -3340 -2486 -512 -159 -46 -39 -44 -28 -18 70 367 231 195 202 206 236 223 207 196 IRR AND NPV AT 10% IRR NPV ('000 Rp) Base Case -I% -3356 More Remote -21 -4240 Very Resote -31 -5124 (calculated over a thirty year period) ECONOMIC NET BENEFIT STREAM, ABRICULTURAL PRODUCTION AND OFF FARM IORK ('000 Rpl Base Case -472 -2227 -1657 -350 18 159 165 156 175 187 281 583 454 426 442 454 492 489 483 482 More Remote -590 -2784 -2072 -352 16 156 163 156 175 187 281 593 454 426 442 454 492 489 483 482 Very Remote -708 -3340 -2486 -354 13 154 161 156 175 187 281 583 454 426 442 454 492 489 483 482 IRR AND NPV AT 10% IRR NPV ('000 Rp) Base Case 5% -1841 These IRRs and NPVs assume that the full cost of infrastructure investments is inputted to the project. Bore Remote 4% -2725 In fact, part of the benefits accrue to the local population and spontaneous migrants (see para 3.05 Very Remote 2% -3609 in main text). IRRs which take these factors into account are shown in the text (Table 3.4). Notes (values in '000 Rp unless otherwise stated) Total Family Labor Supply (Days) 350 364 379 394 409 426 443 461 479 498 518 518 518 518 518 518 518 Potential Off Farm Incoee (1): Financial 220 220 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 Potential Off Farm Income (1): Economic 200 200 200 200 200 203 205 211 216 223 231 239 248 256 266 276 286 Family Labor valued at subsistence incose in Java 270 270 270 270 270 274 277 284 291 302 312 323 334 346 359 372 386 Subsistence Incose in Java, Economic terms 222 222 222 222 222 225 228 234 240 248 257 266 276 285 296 307 318 Assumptions about growth in per capita BDY: 0.01 0.0% 0.0% 0.0% O.O0 1.3% 1.1% 2.8% 2.5% 3.5% 3.5X 3.5X 3.51 3.5% 3.71 3.7I 3.71 Total labor available for off fare work after farming activities (days) (1): 150 164 209 224 259 276 293 311 329 348 368 368 368 368 368 368 368 Off farm income, economic: 158 173 200 200 200 203 205 211 216 223 231 239 248 256 266 276 286 Off farm income, financial: 174 190 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 (1) Maximum asount of off farm Nwrk available assused to be 190 days. U ECONOHIC AWD FINANCIAL ANALYSIS OF UPLAND DIVERSIFIED (MARKET ACCESS REQUIRED) F3DDCRDP MDDEL (all prices in 1986 constant Rpl Year 1979 1980 1981 1982 1983 1984 1985 1986 1907 1908 1999 1990 1991 1992 1993 1994 1995 199h 1997 1998+ Commodities Rice Yield (t/al) I I I I I I I I 1 1 1 I I I I i Area (ha) 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 Econosic Price (Rp/kg) 184 176 165 117 98 83 119 155 192 109 186 183 180 178 178 178 178 Financial Price (Rp/kg) 152 149 152 127 125 150 150 150 190 190 190 190 190 180 180 180 180 Corn Yield It/ha) 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 Area (ha) 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.I 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Economic Price (Rp/kg) 135 107 170 149 118 98 113 128 144 143 143 142 142 141 141 141 141 Financial Price (Rp/kg) 120 140 150 130 130 130 130 130 130 130 130 130 130 140 140 140 140 Cassava Yield It/ha) 6 6 6 6 6 6 6 6 6 6 A 6 6 6 6 6 6 Area (ha) 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Economic Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Financial Price IRp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Peanuts Yield It/ha) 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 Area (ha) 0.05 0.1 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 Economic Price (Rp/kg) 469 584 501 535 398 350 362 374 387 385 382 380 378 375 375 375 375 Financial Price (Rp/kgl 400 400 430 450 450 450 450 45D 450 450 450 450 450 450 450 450 450 Vegetables and Fruit ('000 RpI 60 200 300 400 500 SOO 500 500 500 SOO 500 500 500 500 500 500 500 Small Livestock (1000 Rp) 25 56 113 133 133 133 133 133 133 133 133 133 133 133 133 131 133 Costs of Production Planting Material ('000 Rp) 34 46 21 16 11 11 11 11 11 11 1I It 11 11 Urea (t) (first 3 years provided free by 6D1) 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Economic Cost-Rp/kg 250 223 269 266 212 243 259 275 295 298 302 305 308 311 .311 311 311 Financial Cost-Rp/kg 99 96 104 95 120 130 IS 1SO I50 150 150 150 150 200 200 200 200 TSP (t) (first 3 years provided free by 601) 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Econoeic Cost-Rp/kg 240 238 237 222 236 236 238 239 241 244 247 250 253 256 256 256 256 0 Financial Cost-Rp/kg 99 96 104 95 120 130 150 150 150 150 ISO 150 150 200 200 200 200 kD Pesticide (kgl a 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 Economic Cost-('000 Rp/kg) 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 1 Financial Cost-l(OOD Rp/kg) 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 3 3 3 3 5 5 5 5 Hired Labor (days) 100 76 41 6 0 0 0 0 0 0 0 0 O 0 0 O 0 Total Labor (days) 450 440 420 400 380 400 400 400 400 400 400 400 400 400 400 400 400 Financial Cost of Hired Labor ('000 Rp/day) 1.20 1.20 1.20 1.20 1.20 1.22 1.23 1.26 1.29 1.34 1.39 1.44 1.49 1.54 1.59 1.65 1.72 Other ('000 Rp) 55 60 25 25 25 30 30 30 30 30 30 30 30 30 30 30 30 Production (t) Rice 0.400 0.400 0.400 0.400 0.400 0.400 0.400 0.400 0.400 0.400 0.400 0.400 0.400 0.400 0.400 0.400 0.400 Corn 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 0.070 Cassava 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 Peanuts 0.030 0.060 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 0.090 Economic Value of Field Crops ('DOO Rp) 120 140 146 128 106 94 III 127 145 143 142 140 139 138 138 138 138 Financial Value of Field Crops (1000 RpI 104 116 133 123 122 132 132 132 148 148 148 14B 148 145 145 145 145 Dross Fare Income (1000 Rp) Econoeic 205 390 559 601 739 727 744 760 778 776 775 773 772 771 771 771 771 Financial 189 372 546 656 755 765 765 765 781 781 781 781 781 778 778 778 778 Total Production Costs ('000 RpI n0 | Econosic, Including Income Foregune, lava 543 536 458 422 404 410 424 434 444 454 464 474 485 496 506 517 528 Financial, including only Hired Labor 261 248 149 99 96 105 113 113 113 125 125 125 125 161 161 161 161 o w AGRICULTURE ALONE Net Economic Farm Income ('000 Rp1 (excludes econooic value of 601 expenditures) -339 -140 101 239 335 309 320 326 333 322 311 - 299 287 275 265 254 242 Net Return to Honsehold ('nun Rpl -52 144 417 556 65Y 660 652 652 668 656 656 656 h56 617 617 617 617 A6RICULTURE AND OFF FARM WORK Off Fars Income: Financial (Rp 000) 0 0 0 0 34 30 51 74 99 127 15 164 169 175 182 189 195 Off Fars Income: Economic (Rp '000) 0 0 0 0 31 28 46 67 90 115 144 149 154 159 165 171 178 Net Economic Household Income (Rp '000) -339 -140 101 239 366 337 366 394 423 438 454 448 441 434 430 425 420 Net Return to Household (Rp '000) -32 163 438 558 694 691 703 726 767 783 814 820 826 792 799 806 S13 OTHER NOT DEVELOPMENT COSTS PER HOUSEHOLD IOO Rp) Base Case 472 2227 1657 10 10 9 9 More Remote 590 2784 2072 12 12 11 11 Very Remote 708 3340 24B6 15 15 13 13 ECONOMIC NET BENEFIT STREAM, AhRICULTURAL PRODUCTION ('000 Rp) Base Case -472 -2227 -1657 -348 -150 92 230 335 309 320 326 333 322 311 299 287 275 265 254 242 More Resote -590 -2784 -2072 -351 -152 90 228 335 309 320 326 333 322 311 299 287 275 265 254 242 Very Remote -708 -3340 -2486 -353 -155 87 226 335 309 320 326 333 322 311 299 287 275 265 254 242 IRR AND NPV AT lOX IRR NPV ('000 Rp) Base Case 2% -2335 Nore Remote 1% -3219 Very Remote 0X -4103 (calculated over a thirty year period) ECONOMIC NET BENEFIT STREAM, AGRICULTURAL PRODUCTION AND OFF FARM WORK ('000 Rp) Base Case -472 -2227 -1657 -348 -150 92 230 366 337 366 394 423 438 454 448 441 434 430 425 420 More Remote -590 -2784 -2072 -351 -152 90 228 366 337 366 394 423 43B 454 44B 441 434 430 425 420 Very Reeote -708 -3340 -2486 -353 -155 87 226 366 337 366 394 423 438 454 448 441 434 430 425 420 IRR AND NPV AT 10X IRR NPV ('OO0 Rp) Bas Aase NP -('0020 These IRRs and NPVs assume that the full cost of infrastructure investments is inputted to the project. More Remot 3% -2705In fact, part of the benefits accrue to the local population and spontaneous migrants (see para 3.05 Very Remote 2X -3589 in main text). IRRs which take these factors into account are shown in the text (Table 3.4). (calculated over a thirty year period) Notes (values in '000 Rp unless otherwise noted) Total Family Labor Supply (Days): 350 364 379 394 409 426 443 461 479 498 518 518 518 518 518 518 518 Potential Off Farm Income (11: Financial 220 220 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 Potential Off Farm Income (1): Economic 200 200 200 200 200 203 205 211 216 223 231 239 248 256 266 276 286 Family Labor Valued at subsistence income in Java 270 270 270 270 270 274 277 284 291 302 312 323 334 346 359 372 386 Subsistence Income in Java, Economic terms 222 222 222 222 222 225 228 234 240 248 257 266 276 285 296 307 318 Assumptions about growth in per capita 6DY: 0.0% 0.0% 0.0% 0.0% 0.0X 1.3% 1.1% 2.8% 2.5% 3.5% 3.5% 3.51 3.51 3.5% 3.7% 3.7% 3.7% Total labor available for off fare work after farsing activities (days) (11: 0 0 0 0 29 26 43 61 79 98 118 118 118 118 11 I1 118 Off farm income, economic: 0 0 0 0 31 28 46 67 90 115 144 149 154 159 165 171 178 0ff fare income, financial 0 0 0 0 34 30 51 74 99 127 158 164 169 175 192 188 195 fl) Maxioum amount of off farm work available assumed tD be 190 days. 0 ul a,, a,0 D m4 ° m CCC04 04 nC C -C C °C w O CC° CC -0 z ZC o o4 Fm-n 00 o >Zn wCnO _ R, R 5n R'f 5O R' 0 E -v CO , CC, E, C ° DC a n a r _ n 0 00O 0000 e r 0 CC o L n _O <5 n-en- 00 * -C n °--- 00 C-C - n n - r- r o- a c mDC m a-. mDm -, °0 _ o '°0 u 0O~ C iOs 00'°Sq> gC n aonCDn C. n a 0 CC n .r nSC.CmaOmS C 0' - - - a s a - - - - - cCC CC O l - . - n o CO O ve_-nO ,0- Cm ,04C mCCC 04- C 04r CD 0 04 -.0- = CO S- S400400C00 SCC 0C a. , 4C o a 3n ° a° _C 00-C °C - -n C~00 C - 0 - 0 - o '~~~~~~~~~~~~~~~~ZZ C I CCa n f o 00 C _C C_ C O l e fo l ; a .-a ar 0-. r S~- - - 0ci S0…nol _ moooa o - n ls - = er-Cc S0 ° a a- - _ 4 O OC 0 O W CO OC ClO 0 9a mV Cr lmSO _ 0 o o0OC a cc - a c w _ om0 -e O. - oa _a-- or'--in l rm je --n s1~°~ _-o - 0 W- C ° no n- °SJw ;o C04 co oC r- Cr a l O 0 -oi lmmelco°_t CCC - C SC - 00 C-C - - - -- - - 0-- -- Z- - - - -O C o n O 0 a--- n nn a C rC CC _W o a a -0 m Co O. VC l_soo_ov 0-44 04-4 ~ ~ 44- -44C44C, C!4 …-o 04 0404 CC 4CC 040te Z~-C44 - 44 --Z ! ; 44C 44C 0- 4C… 440 -J0- CC 0-44 --C --a-Ca- - 4440- 4 -4 -AC - -0- -04404'…0-.0--'-- C- 44 0 04 C 44 04 -004 04- C 0--- - - - -44- -4 - 4- - -4-C-C -4- - -4… 04 --0- 0-C 0404 -00 0444 - --0--a--0-- -- -4 -44 -4 C -004404'…0-4'--- CCC CC 0-C 004 0.0040- 44040 4444.- -44-0-'-a - - --4-4-' -C - - 0-44 0- 404 04 CaC-a- C- C - 0 - - --CC--- -- C-- - - -C4Ca CC-44 4 44 a - - - 04- 0 0444 440- C4404 - -44 -4-4a- - 044. -0 0---0-C CCS. 4444 CC 04CC C-'- - - - -4 - - - - - - - - - 0-44-C…C 04. C C 0444 04 44 C C C 0 04Z -4CC4 C4 40 44 000 -4 C -C 044 44 0 - Z.4 4 4 C 044- Ca 0444 44 -C-OC-C - -44 - -4C4 a - 0 -.. -40 C CC-C44 4- - --O-CC- z ~~~~~~~~~~~o -E ~~~~~Z--- ---- 9 auqLaj - 6T- C x_ Nm _ AGRICULTURE AND OFF FARM aORK Off Farm Income: Financial (Rp 000) 220 161 91 tO0 127 148 169 196 224 246 254 263 272 282 292 303 314 Off Farm Income: Economic (Rp 000) 200 146 83 99 115 134 154 178 203 223 231 239 248 256 266 276 286 Net Economic Household Income (Rp 000) 19 21B 183 292 322 323 410 508 604 623 634 637 649 667 670 680 679 Net Return to Household IRp 000) 2B5 407 457 536 672 790 817 862 948 982 1010 1029 1058 1071 1085 1107 1119 DTHER MOT DEVELOPMENT COSTS PER HOUSEHOLD I 000 Rp) Base Case 472 2227 1657 10 10 9 9 More Remote 590 2784 2072 12 12 11 11 Very Remote 708 3340 2486 15 15 13 13 ECONOMIC NET BENEFIT STREAM, AGRICULTURAL PRODUCTION ('000 Rp) Base Case -472 -2227 -1657 -191 62 92 IB4 207 IB9 256 330 400 400 403 398 402 410 404 404 393 More Remote -590 -2784 -2072 -194 60 89 IB2 207 189 256 330 400 400 403 39B 402 410 404 404 393 Very Remote -708 -3340 -2486 -196 57 87 IBO 207 IB9 256 330 400 400 403 39B 402 410 404 404 393 IRR AND NPV AT 102 IRR NPV (I000 Rp) Base Case 40 -1900 More Remote 30 -2784 Very Remote 2X -3669 (calculated over a thirty year period) ECONOMIC NET BENEFIT STREAM, AGRICULTURAL PRODUCTION AND OFF FARM WORK ('000 Rp) Base Case -472 -2227 -1657 9 209 174 283 322 323 410 508 604 623 634 637 649 667 670 680 679 hore Remote -590 -2784 -2072 6 206 172 281 322 323 410 50B 604 623 634 637 649 667 670 680 679 Very Remote -708 -3340 -2486 4 204 170 278 322 323 410 508 604 623 634 637 649 667 670 680 679 IRR ANDN NPV 0T101 188 NV ('000 Rp) These IRRs and NPVs assume that the full cost of infrastructure investments is inputted to the project. Base Case 8l -594 In fact, part of the benefits accrue to the local population and spontaneous migrants (see para 3.05 More Remote 72 -1479 in main text). ITRRs which take these factors into account are shown in the text (Table 3.4). H Very Remote 5X -2363 w (calculated over a thirty year period) Notes (values in '000 Rp unless otherwise noted) Total Family Labor Supply (Days): 350 364 379 394 409 426 443 461 479 498 518 SIB 518 518 518 B518 51 Potential Off Farm Income (1): Financial 220 220 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 Potential Off Fare Income (1): Economic 200 200 200 200 200 203 205 211 216 223 231 239 248 256 2U 276 286 Family Labor valued at subsistence income in Java 270 270 270 270 270 274 277 284 291 302 312 323 334 346 359 372 386 Subsistence Income in Java, Ecnoumyc teres 222 222 222 222 222 225 228 234 240 24B 257 266 276 285 296 307 318 Assumptions about growth in per capita BDY: 0.00 0.00 0.0 0.0 0.0 0.3X 1.'l 2.80 2.51 3.5X 3.52 3.50 3.50 3.51 3.72 3.70 3.72 Total labor available for off farm work after farming activities (1): 200 139 79 94 109 126 143 161 179 198 218 218 218 218 218 218 218 Off farm incose, economic: 200 146 83 99 115 134 154 178 203 223 231 239 248 256 266 276 286 Off fare income, financial: 220 161 91 108 127 14B 169 196 224 246 254 263 272 282 292 303 314 3 (11 Maxieum asount of off farm work available assumed to be 190 days. ECONOMIC AND FINANCIAL ANALYSIS OF UPLAND HIGH INPUT LABIR INTENSIVE FOODCROP MODEL (all prices in 1986 constant Rp) Year 1979 1980 1981 1902 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 19984 Commodities Rice Yield It/hal 1 1.2 1.5 1.5 1.5 1.5 1.5 1.5 1.0 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 Area (ha) 0.4 0.75 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Economic Price IRp/kg) 184 176 165 117 98 83 119 155 192 189 186 183 (80 178 178 17B 178 Financial Price (Rp/kgl 152 149 152 127 125 150 150 150 190 190 190 190 190 180 180 180 180 Corn Yield It/ha) 1 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 Area (hal 0.4 0.75 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I Economic Price (Rp/kg) 135 167 170 149 lID 98 113 128 144 143 143 142 142 141 141 141 141 Financial Price (Rp/kg) 120 140 150 130 130 130 130 130 130 130 130 130 130 140 140 140 140 Cassava Yield (t/hal 8 10 10 t0 10 10 10 10 10 I0 (0 10 IV 10 18 Io 10 Area (tha) 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 EconoDic Price (Rp/kgl 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Financial Price (Rp/kgl 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Peanuts Yield (t/hal 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 Area (hal 0.4 0.75 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Ecanomic Price (Rp/kg) 469 584 001 535 398 350 362 374 387 385 3B2 380 378 375 375 375 375 Financial Price lRp/kg) 400 400 430 450 450 450 450 450 450 450 450 450 450 450 450 450 450 Vegetables and Fruit ('000 Rpl 50 60 75 75 75 90 90 100 100 100 100 100 120 120 120 120 120 Stall Livestock ('000 Rp1 12 20 25 30 40 40 40 40 40 40 40 40 40 40 40 40 40 Costs of Production Planting Material ('000 Rp) 38 71 95 95 95 95 95 95 95 95 95 95 95 95 95 95 95 Urea (t) (first 3 years provided free by 601) 0.108 0.202 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 Economic Cost-Rp/kg 250 223 269 266 212 243 259 275 295 298 302 305 308 311 311 311 311 Financial Cost-Rp/kg 99 96 104 95 120 130 150 150 150 IS0 150 ISO 150 200 200 200 200 1 TSP (t) (first 3 years provided free by 601) 0.108 0.202 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 0.27 , Economic Cost-Rp/kg 240 238 237 222 236 236 238 239 241 244 247 250 253 256 256 256 256 Financial Cost-Rp/kg 99 96 104 95 120 130 15O 15O 150 150 150 150 150 200 200 200 200 Pesticide (kg) 14 26 35 35 35 35 35 35 35 35 35 35 35 35 35 35 35 Econoaic Cost-('OO Rp/kg) 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 Financial Cost-('000 Rp/kg) 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 3 3 3 3 5 5 5 5 Hired Labor (days) 0 86 221 206 191 174 157 139 121 102 B2 82 82 82 82 82 82 Total Labor (days) 300 450 600 600 600 600 600 600 600 600 600 600 600 600 600 600 600 Financial Cost of Hired Labor ('000 Rp/day) 1.20 1.20 1.20 1.20 1.20 1.22 1.23 1.26 1.29 1.34 1.39 1.44 1.49 1.54 1.59 1.65 1.72 Other ('000 Rp) 52 57 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16 Production (t) Rice 0.400 0.900 1.000 1.500 1.000 1.000 1.500 1.500 1.500 1.000 1.000 1.000 1.000 1.500 1.500 1.500 1.500 Corn 0.400 0.900 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 Cassava 1.600 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 Peanuts 0.240 0.450 0.600 0.600 0.600 0.600 0.600 0.600 0.600 0.600 0.600 0.600 0.600 0.600 0.600 0.600 0.600 Economic Value of Field Crops ('000 Rp) 271 610 790 713 565 490 569 649 731 724 718 711 705 699 699 699 699 Financial Value of Field Crops ('000 Rp1 235 478 704 655 652 689 689 689 749 749 749 749 749 746 746 746 746 Dross Farm Income ('000 Rp1 Economic 333 690 890 81B 680 620 699 789 871 864 858 051 865 859 859 059 859 Financial 297 558 804 760 767 819 819 029 809 009 089 089 909 906 906 906 906 Total Production Costs ('000 Rp) 0 01 Econseic, Including Income Foregone, 2aua 463 693 000 071 048 040 844 044 043 040 830 048 862 876 890 904 918 ODv Financial, Including only Hired Labor 132 309 405 462 457 445 438 421 401 433 411 415 419 520 525 529 534 AGRICULTURE ALONE Net Economic Farm Income ('000 Rp1 (excludes economic value of G6I expenditures) -131 -3 2 -53 -160 -228 -144 -95 20 25 23 2 3 -17 -31 -4i -59 Net Return to Household ('000 Rp) (05 268 339 297 310 374 381 40B 400 456 47B 474 490 386 381 377 372 AGRICULTURE AND OFF FARM WORK Off Farm Income: Financial (Rp '000) 58 0 0 0 0 0 0 0 O 0 0 0 0 0 0 0 0 Off Farm Income: Economic (Rp '000) 53 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Net Economic Household Income (Rp '000) -78 -3 2 -53 -168 -228 -144 -55 28 25 23 2 3 -17 -31 -45 -59 Net Return to Household (Rp '000) 244 288 375 297 310 374 381 408 488 456 478 474 490 386 3B1 377 372 OTHER MOT DEVELOPMENT COSTS PER HOUSEHOLD ( 000 Rp) Base Case 472 2227 1657 10 10 9 9 More Remote 590 2784 2072 12 12 11 11 Very Remote 708 3340 2406 15 15 13 13 ECONOMIC NET BENEFIT STREAM, AGRICULTURAL PRODUCTION ('000 Rp) Base Case -472 -2227 -1657 -141 -13 -7 -62 -1JB -228 -144 -55 28 25 23 2 3 -17 -31 -45 -59 More Remote -590 -2784 -2072 -143 -15 -9 -64 -168 -228 -144 -55 28 25 23 2 3 -17 -31 -45 -59 Very Remote -708 -3340 -24B6 -146 -18 -11 -66 -168 -22B -144 -55 28 25 23 2 3 -17 -31 -45 -59 IRR AND NPV AT 10 IRR NPV ('000 Rp) Base Case negative -3960 More Remote negative -4844 Very Remote negative -5729 (Calculated over a thirty year period) ECONOMIC NET BENEFIT STREAM, AGRICULTURAL PRODUCTION AND OFF FARM WORK ('000 Rp) Base Case -472 -2227 -1657 -88 -13 -7 -62 -168 -228 -144 -55 28 25 23 2 3 -17 -31 -45 -59 More Remote -590 -27B4 -2072 -90 -15 -9 -64 -168 -22B -144 -55 28 25 23 2 3 -17 -31 -45 -59 Very Remote -708 -3340 -2486 -93 -IB -11 -66 -168 -228 -144 -55 28 25 23 2 3 -17 -31 -45 -59 IRR AND NPV AT 10 IRR NPV ('000 RD) These IRRs and NPVs assume that the full cost of infrastructure investments is inputted to the project. In fact, Base Case negative -3924 part of the benefits accrue to the local population and spontaneous migrants (see para 3.05 in main text). IRRs More Remote negative -4808 which take these factors into account are shown in the text (Table 3.4). Very Remote negative -5693 (Calculated over a thirty year period) 1- Notes (values in '000 Rp unless otherwise noted) Total Family Labor Supply IDays): 350 364 379 394 409 426 443 461 479 498 518 518 518 518 518 518 518 Potential Off Farm Income (1i: Financial 220 220 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 Potential Off Farm Income (1: Economic 200 200 200 200 200 203 205 211 216 223 231 239 248 256 266 276 286 Family Labor valued at subsistence income in Java 270 270 270 270 270 274 277 284 291 302 312 323 334 346 359 372 386 Subsistence Income in Java, Economic terms 222 222 222 222 222 225 228 234 240 248 257 266 276 285 296 307 318 Assumptions about growth in per capita DY: 0.00 0.00 OOX O .OX 0.0d 1,30 1I1Z 2.80 2.5 3.51 3.50 3.50 3.50 3.5X 3.71 3.7X 3.70 Total labor available for off farm mork after farcing activities (1h 50 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Off farm income, economic: 53 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 U Off farm income, financial: 58 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 (1) Maxieuu amount of off farm work available assumed to be 190 days. o -o 1 ECONOMIC AND FINANCIAL ANALYSIS OF UPLAND H1GH INPUT WITH CATTLE MODEL (all prices in 1906 constant Rpl Year 1979 1980 1901 1982 1983 19B4 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998+ Coamodities Rice Yield (t/ha) 1 1.25 1.5 1.75 2 2.3 2.3 2.3 2.3 2.4 2.4 2.5 2.5 2.5 2.5 2.5 2.5 Area (ha) 0.4 0.8 1 1.2 1.35 1.6 1.6 1.6 1.6 16 1.6 1.6 1.6 1.6 1.6 1.6 1.6 Economic Price (Rp/kg) 104 176 165 117 9e 83 119 155 192 189 186 183 180 17B 178 178 178 Financial Price (Rp/kg) 152 149 152 127 125 IsA 150 150 19A 190 190 190 190 1DA IAO ISO 180 Corn Yield (t/ha) 0.75 1 1.25 1.5 1.75 2 2 2 2 2.1 2.1 2.2 2.2 2.2 2.2 2.2 2.2 Area (ha) 0.4 0.8 I 1.2 1.35 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 Economic Price (Rp/kg) 135 167 170 149 118 90 113 128 144 143 143 142 142 141 141 141 141 Financial Price (Rp/kg) 120 140 150 130 130 130 130 130 130 130 130 130 130 140 140 140 140 Cassava Yield (t/ha) 10 11 12 13 14 15 15 15 15 16 17 18 18 18 1s 18 is Area (ha) 0.1 0.2 0.25 0.3 0.4 0.4 0.4 0.4 0.4 0.4 0.4 A.4 0.4 0.4 0.4 0.4 0.4 Economic Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Financial Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Peanuts Yield (t/ha) 0.6 0.7 o.8 0.9 1 1.1 1.1 1.1 1.1 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2 Area (ha) 0.4 0.8 1 1.2 1.35 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 Economic Price (Rp/kg) 469 584 501 535 398 350 362 374 387 385 382 380 378 375 375 375 375 Financial Price (Rp/kgl 400 400 430 450 450 450 450 450 450 450 450 450 450 450 450 450 450 Vegetables and Fruit ('000 Rp1 (.2 ha) 10 37 39 41 41 50 75 100 125 100 150 175 200 200 200 200 200 Small Livestock and Cattle Dfftake ('000 Rp) 12 20 25 30 40 60 40 84 350 225 201 221 217 259 257 252 252 Costs of Production Planting Material ('000 Rp) 1/ 130 120 145 170 335 200 200 200 200 360 200 200 200 200 200 200 200 Urea (t) (first 3 years provided free by 601) 0.13 0.25 0.31 0.38 0.44 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 I Economic Cost-Rp/kg 250 223 269 266 212 243 259 275 295 29B 302 305 308 311 311 311 311 Financial Cost-Rp/kg 99 96 104 95 120 130 ISO 150 150 150 150 150 150 200 200 200 200 TSP (t) (first 3 years provided free by 6011) 0.1 0.2 0.25 0.3 0.35 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 8.4 0.4 Economic Cost-Rp/kg 240 238 237 222 236 236 238 239 241 244 247 250 253 256 256 256 256 Financial Cost-Rp/kg 99 96 104 95 120 130 150 150 150 150 150 150 150 200 200 208 200 KCI (t) 0.1 0.125 0.15 0.175 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Economic Cost-Rp/kg 176 169 180 184 IBI 18l le1 181 190 190 190 190 190 196 196 196 196 Financial Cost-Rp/kg 99 96 104 95 90 120 150 150 150 150 ISA ISA 150 200 200 200 200 Pesticide (kg) 17.5 35 43.75 52.5 61.25 70 70 70 70 70 70 70 70 70 70 70 70 Econoaic Cost-('OOO Rp/kgl 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 Financial Cost-(lOOO Rp/kg) 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 3 3 3 3 5 5 5 5 Hired Labor (days) 0 0 0 19 72 124 107 89 71 52 32 32 32 32 32 32 32 Total Labor Required (days) 13B 275 344 413 481 550 550 550 550 550 550 550 550 550 550 550 550 Financial Cost of Hired Labor (1000 Rp/day) 1.20 1.20 1.20 1.20 1.20 1.22 1.23 1.26 1.29 1.34 1.39 1.44 1.49 1.54 1.59 1.65 1.72 Other (1000 Rp) 100 100 30 30 30 30 30 30 30 30 30 30 30 30 30 30 38 Production (t) Rice 0.400 1.000 1.500 2.100 2.700 3.680 3.680 3.680 3.680 3.840 3.840 4.000 4.000 4.000 4.000 4.NO0 4.000 Corn 0.300 0.800 1.250 1.800 2.363 3.200 3.200 3.200 3.200 3.360 3.360 3.520 3.520 3.520 3.520 3.520 3.520 Cassava 1.000 2.200 3.000 3.900 5.600 6.000 6.000 6.000 6.000 6.400 6.800 7.200 7.200 7.200 7.200 7.200 7.200 Peanuts 0.240 0.560 o.0oo 1.080 1.350 1.760 1.760 1.760 1.760 1.920 1.920 1.920 1.920 1.920 1.920 1.920 1.920 Estate Crop Economic Value of Field Crops ('000 Rp) 246 678 918 1166 1187 1349 1551 1752 1962 2067 2057 2098 2082 2065 2065 2065 2065 Financial Value of Field Crops (1000 Rp) 212 527 817 1061 1359 1874 1B74 1874 2021 2152 2160 2218 2218 2214 2214 2214 2214 Gross Farm Incoame ('000 Rp) Economic 268 735 982 1237 1268 1459 1666 1936 2437 2392 2408 2494 2499 2524 2522 2517 2517 (D >I Financial 234 584 BBI 1132 1440 1984 1989 2058 2496 2477 2511 2614 2635 2673 2671 2666 2U6 6 6 Total Production Costs ('000 Rp) Economic, Including Income Foregone, Java 649 B12 873 1005 1284 1296 1294 1297 1302 1458 1294 1306 1320 1334 1346 1358 1371 Financial, including only hired labot 289 328 314 383 656 627 632 613 592 834 649 .651 652 R49 05) 053 OSS AGhRICULTURE ALONE Net Economic Farm Income ( 000 Rp) (excludes econoxic value of 601 expenditures) -381 -77 108 232 -1 163 371 639 1136 934 1115 1188 1180 1190 1176 1159 1146 Net Return to Household ( 000 Rp) -32 299 624 749 784 1357 1357 1445 1904 1643 1861 1964 1983 I124 1820 IB13 1ll AGRICULTURE AND OFF FARM WORK Off Fars Income Financial (Rp 000) 220 103 40 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Off Fars Income Economic IRp 000) 200 94 36 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Net Econosic Household Incose (Rp 000) -181 17 145 232 -16 163 371 639 1136 934 1115 1188 1180 1190 1176 1159 1146 Net Return to Household (Rp 000) 188 402 664 749 784 1357 1357 1445 1904 1643 1861 1964 1983 1824 1820 1813 1811 OTHER MOT DEVELOPMENT COSTS PER HOUSEHOLD ('000 Rp) Base Case 472 2227 1657 294 10 9 9 More Remote 590 2784 2072 296 12 11 11 Very Remote 708 3340 2486 299 15 13 13 ECONOMIC NET BENEFIT STREAM, AGRICULTURAL PRODUCTION ('000 Rp) Base Case -472 -2227 -1657 -675 -86 99 223 -16 163 371 639 1136 934 1115 1188 1180 1190 1176 1159 1146 More Remote -590 -2784 -2072 -677 -89 97 221 -16 163 371 639 1136 934 1115 1188 l11O 1190 1176 1159 1146 Very Reeote -708 -3340 -2486 -680 -91 95 218 -16 163 371 639 1136 934 1115 1138 1180 1190 1176 1159 1146 IRR AND NPV AT 10 IRR NPv ('000 Rp) These IRRs and NPVs assume that the full cost of infrastructure investments is inputted to the project. In fact, Base Case 10 -109 part of the benefits accrue to the local population and spontaneous migrants (see para 3.05 in main text). IRRs More Remote 8X -993 which take these factors into account are shown in the text (Table 3.4). Very Reeote 70 -1878 (calculated over a thirty year period) Notes (values in '000 Rp unless otherwise notedi Total Family Labor Supply (Days): 350 364 379 394 409 426 443 46l 479 498 518 518 518 518 518 518 518 Family Labor valued at subsistence incose in Java, Financial teres 270 270 270 270 270 274 277 284 291 302 312 323 334 346 359 372 386 Subsistence Incose in Java, Economic terms 222 222 222 222 222 225 228 234 240 248 257 266 276 285 296 307 318 Assuaptions about growth in per capita 6DY: 0.0 0.0 0.0 000 OO. 1.3X 1.1 2.80 2.5X 3.50 3.50 3.5X 3.5Z 3.5X 3.71 3.71 3.7% 1/ Includes application of lime. Liming: 1982, 2t; 1983 .5t; 1984 .5 t; 1985 .5 t; 1986 4t; 1991 4t. Lime is costed at Rp 40,000 per ton (quarry is assumed to be on site). ID M o0 ECONOMIC AND FINANCIAL ANALYSIS OF TIDAL LOW INPUT FOODCROP MODEL WITH SEVERE PEST PROBLEMS (all prices in 1986 constant ROp Year 1979 1980 1901 1982 1983 1984 19B5 1906 19 Y 190E 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998+ Commodities Rice Yield (tlha) 0.8 0.9 0.75 0.75 0.75 o.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 0.75 Area ihal 0.4 0.6 0.7 0.8 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 Economic Price (Rp/kg) 184 176 165 117 9B 03 119 155 192 109 186 183 IgO 178 17B 178 178 Financial Price IRp/kg) 152 149 152 127 125 150 150 150 190 190 190 190 190 180 18O 180 180 Corn Yield (t/ha) 0.6 0.5 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 Area (ha) 0.1 0.15 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Economic Price (RpJkq) 135 167 170 149 11 98 113 12B 144 143 143 142 142 141 141 141 141 Financial Price (Rp/kg) 120 140 150 130 130 130 130 130 130 130 130 130 130 140 140 140 140 Cassava Yield (t/ha) 4.5 4.5 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 Area (ha) 0.1 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Economic Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Financial Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Peanuts Yield (t/ha) Area (ha) Econosic Price (Rp/kg) 469 584 501 535 39B 350 362 374 387 385 382 380 378 375 375 375 375 Financial Price (Rp/kg) 400 400 430 450 450 450 450 450 450 450 450 450 450 450 450 450 450 Vegetables and Fruit ('000 Rp1 40 50 60 60 75 75 BO 90 90 100 100 100 112 112 112 112 112 Small Livestock ('000 RpJ 12 20 25 30 40 40 40 40 40 40 40 40 40 40 40 40 40 Costs of Production Planting Material 1'000 Rp1 5.2 8.3 10.5 11.7 13.B 13.8 13.0 13.B 13.8 13.0 13.B 13.8 13.8 13.8 13.8 13.8 13.8 Urea (t) 0 0 V 0 0 0 0 0 0 0 V 0 0 0 0 0 0 Econosic Cost-Rp/kg 250 223 269 266 212 243 259 275 295 29B 302 305 30B 311 311 311 311 Financial Cost-Rpikg 99 96 104 95 120 130 150 1i0 150 150 150 150 150 200 200 200 200 TSP (t) 0 0 0 0 0 0 0 0 0 0 V 0 0 0 0 0 0 Economic Cost-Rp/kg 240 230 237 222 236 236 23B 239 241 244 247 250 253 256 256 256 256 Financial Cost-Rp/kg 99 96 104 95 120 130 150 150 150 150 150 150 150 200 200 200 200 Pestiride (kg) 5 5 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 EcDosoic CDst-(lOOO Rpikg) 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 Financial Cost-('000 Rpikg) 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 3 3 3 3 5 5 5 5 Hired Labor (days) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Total Labor (days) 250 200 100 180 180 160 160 160 160 160 160 160 160 160 160 160 160 Financial Cost of Hired Labor ('000 Rp/day) 1.20 1.20 1.20 1.20 1.20 1.22 1.23 1.26 1.29 1.34 1.39 1.44 1.49 1.54 1.59 1.65 1.72 Dther 1'000 RDJ 45 45 5 9 5 5 9 9 5 5 5 5 5 5 5 5 5 Production (t) Rice 0.320 0.540 0.525 0.600 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 Corn 0.060 0.075 0.080 0.080 0.080 0.080 O.OBO 0.080 0.080 0.00 0.080 0.080 0.080 0.080 0.080 0.080 0.00 Cassava 0.450 0.900 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 Peanuts EcDnosic Value of Field Crops ('000 Rp) 76 125 123 105 76 6B B5 103 121 119 1I1 117 115 114 114 114 114 Financial Value of Field Crops ('000 Rp) 64 108 115 109 89 101 101 101 119 119 119 119 119 115 115 115 115 Gross Farm Incoae ('000 Rp) Economic 128 195 208 195 191 183 205 233 251 259 258 257 267 266 266 266 266 Financial 116 178 200 199 204 216 221 231 249 259 259 259 271 267 267 267 267 °9 |3 Total Production Costs ('000 Rp) 114 Economics Including Incose Foregsne, Java 300 311 266 267 269 272 275 281 287 295 304 313 322 332 342 353 365 o ` 0 Financial, Including only Hired Labor 58 61 22 23 25 25 25 25 25 31 31 31 31 39 39 39 39 ASRICULTURE ALONE Net Economic Farm Income ('000 Rp) (excludes ecoonomic value of 601 expenditures) -1I0 -116 -56 -72 -78 -B9 -69 -48 -36 -36 -46 -56 -55 -66 -76 -87 -99 Net Return to Household ('000 Rp) 59 117 178 177 100 191 196 206 224 220 228 228 240 228 228 228 228 A6RICULTURE AND OFF FARM WORK Off Farm Income: Financial (Ro '000) 116 190 220 220 22(0 223 225 232 237 246 254 263 272 282 292 303 314 off Farm Income: Economic (Rp 000) 105 173 200 200 200 203 205 211 216 223 231 239 248 256 266 276 286 Net Economic Household Income (Rp 000) -75 57 142 128 122 113 136 162 180 187 185 183 193 191 190 188 187 Net Return to Household (Rp 000) 174 307 398 397 4001 414 421 438 461 474 482 491 512 510 521 531 543 OTHER MOT DEVELOPMENT COSTS PER HOUSEHOLD ('000 Rp) Base Case 472 1509 2252 1952 10 10 9 9 ECONOMIC NET BENEFIT STREAM, AGRICULTURAL PRODUCTION ( 000 Rp) Base Case -472 -1589 -2252 -2132 -12o -68 -B -B7 -89 -69 -48 -36 -36 -46 -56 -55 -66 -76 -87 -99 IRR AND NPV AT 10% IRR NPV ('000 Rpn Base Case negative -5376 (calculated over a thirty year period) ECONOMIC NET BENEFIT STREAM, A6RICULTURAL PRODUCTION AND OFF FARM MORK ('000 Rp) Base Case -472 -1589 -2252 -2027 47 132 119 113 113 136 162 1B0 187 185 183 193 191 190 188 IB7 IRR AND NPv AT 10 IRR NPv ('000 RP) These IRRs and NPVs assume that the full cost of infrastructure investments is inputted to the project. In fact, Base Case -2% -39?7 part of the benefits accrue to the local population and spontaneous migrants (see para 3.05 in main text). IRRs (calculated over a thirty year period) which take these factors into account are shown in the text (Table 3.4). Notes (values in '000 Rp unless otherwise noted) I Total Family Labor Supply (Days) 350 364 379 394 409 426 443 461 479 498 51B 518 518 518 51B 51B 518 Potential Off Farm IncoDe (1): Financial 220 220 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 Potential Off Farm Income (1): Econosic 200 200 200 200 200 203 205 211 216 223 231 239 248 256 266 276 2B6 Family labor valued at subsistence income in Java 270 270 270 270 270 274 277 284 291 302 312 323 334 346 359 372 38b Subsistence Incoee in lava, Economic terms 222 222 222 222 222 225 228 234 240 248 257 266 276 285 296 307 318 Assumptions about gromth in per capita GDY: 0.0Z 0.0% 0.0% 0.0% 0.0% 1.3% 1.1 2.8% 2.5% 3.5% 3.5% 3.5% 3.5% 3.5X 3.71 3.7% 3.71 Total labor available for off fare work after farming activities (days) (1): 100 164 199 214 229 266 283 301 319 33B 358 35B 358 358 358 358 358 Off fare income, economic: 105 173 200 200 200 203 205 211 216 223 23'1 239 248 256 266 27b 28b Off farm income, financial 116 190 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 (11 Maximum amount of off farm work available assumed to be 190 days. 0 ~h N H ECONOMIC AND FINANCIAL ANALYSIS OF TIDAL LOW INPUT WITH REASONABLE PEST CONTROL (all prices in 19B6 constant Rp) Year 1979 1900 1901 1902 1903 1904 1905 1986 1907 1900 1909 1990 1991 1992 1993 1994 1995 1996 1997 1990+ Cosmodities Rice Yield (t/ha) 0.0 0.9 1 1.2 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 Area (hal 0.4 0.7 1 1 1.3 1.3 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 Econonic Price (Rp/kg) 104 176 165 117 90 03 119 155 192 109 106 103 10O 170 170 170 170 Financial Price iRp/kg) 152 149 152 127 125 150 150 ISO 190 190 190 190 190 100 100 1ao 180 Corn Yield (t/ha) 0.6 0.7 0.0 o.0 0.0 0.0 0.0 o.0 0.0 0.0 0.0 0.0 0.0 0.0 0.6 0.0 0.0 Area (ha) 0.1 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Economic Price IRptkgl 135 167 170 149 le0 90 113 120 144 143 143 142 142 141 141 141 141 Financial Price (Rp/kg) 120 140 150 130 130 130 130 130 130 130 130 130 130 140 140 140 140 Cassava Yield (t/ha) 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 Area (ha) 0.1 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 Econoeic Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Financial Price (RpIkgl 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Peanuts Yield It/ha) Area (ha) Economic Price IRp/kg) 469 584 501 535 390 350 362 374 307 305 302 300 370 375 375 375 375 Financial Price (Rp/kgl 400 400 430 450 450 450 450 450 450 450 450 450 450 450 450 450 450 Vegetabls and Fruit ('000 Rp) 40 50 60 60 75 75 0o 90 90 100 100 000 112 112 112 112 112 Seall Livestock ('000 Rp1 12 20 25 30 40 40 40 40 40 40 40 40 40 40 40 40 40 Costs of Production Planting Material (1000 Rp) 5.2 9.5 12.5 15.5 10.0 10.0 10.0 (0.0 10.0 18.0 1.0 10.0 10.0 10.0 19.0 10.8 18.6 L'realt 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0° Econosic Cost-Rp/kg 250 223 269 266 212 243 259 275 295 290 302 305 300 311 311 311 311 O Financial Cost-Rp/kg 99 96 104 95 120 130 150 150 150 150 150 150 150 200 200 200 200 O TSPIt 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0O Economic Cost-Rp/kg 240 230 237 222 236 236 230 239 241 244 247 250 253 256 256 256 256 Financial Cost-Rp/kg 99 96 104 95 120 130 150 ISO 150 150 150 150 150 200 200 200 200 Pesticide (kg) 4 4 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Economic Cost-(OOO Rp/kg1 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 Financial Cost-('OOO Rp/kg) 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 3 3 3 3 5 5 5 5 Hired Labor (days) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Total Labor (days) 250 250 280 300 300 300 300 300 300 300 300 300 300 300 300 300 300 Financial Cost of Hired Labor ('000 Rp/day) 1.20 1.20 1.20 1.20 1.20 1.22 1.23 1.26 1.29 1.34 1.39 1.44 1.49 1.54 1.59 1.65 1.72 Other ( 000 Rp) 45 45 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 Production (tl Rice 0.320 0.630 1.000 1.200 1.625 1.625 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 2.000 Corn 0.060 0.140 0.160 0.160 0.160 0.160 0.160 0.160 0.160 0.160 0.160 0.160 0.160 0.160 0.160 0.160 0.160 Cassava 0.600 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 1.200 Peanuts Economic Value of Field Crops (1000 Rp) 70 157 215 107 201 173 279 353 430 424 410 412 406 401 401 401 401 Financial Value of Field Crops (1000 Rp) 67 136 199 196 247 207 344 344 424 424 424 424 424 405 405 405 405 gross Farm Income (1000 Rp) Economic 130 227 300 277 316 208 399 483 560 564 s58 552 550 553 S53 553 553 Financial 119 206 284 206 362 402 464 474 554 564 564 564 576 557 557 557 557 Total Production Costs (1000 Rp) Hit Economic, Including Incose Foregone, Java 301 305 259 262 265 260 271 277 203 291 300 309 310 320 330 349 361 o 1 Financial, Including only Hired Labor 56 61 26 29 32 32 32 32 32 35 35 35 35 39 39 39 39 O A6RICULTURE ALONE Net Economic Fare Income (1000 Rp) (lecludes economic value of GOI expendituresl -170 -70 41 15 51 20 126 206 277 272 250 243 239 225 215 204 193 Net Return to Household ('000 Rp) 63 146 250 250 330 371 432 442 522 529 529 529 541 510 51B SlO 510 AGRICULTURE AND OFF FARM WORK Off Fara Income: Financial (Rp '000) 116 132 114 108 127 148 169 196 224 246 254 263 272 282 292 303 314 Off Farm Income: Economic (Rn '000) 105 120 104 99 115 134 154 176 203 223 231 239 248 256 266 276 286 Net Econosic Household Income (Rp '000) -65 42 145 114 166 154 282 384 480 496 4B9 482 4B7 482 481 480 478 Net Return to Household (Rp '000) 179 278 372 366 457 518 601 638 745 775 783 792 813 800 811 822 833 OTHER MOT DEVELOPMENT COSTS PER HOUSEHOLD ('000 Rp) Base Case 472 1589 2252 1952 10 10 9 9 ECONOMIC NET BENEFIT STREAM, AGRICULTURAL PRODUCTION '000 Rp) Base Case -472 -1589 -2252 -2122 -88 31 6 42 20 128 206 277 272 258 243 239 225 215 204 193 IRR AND NPv AT 10% IRR NPV ('000 Rp) base Case -2X -4102 (Calculated over thirty year period) ECONOMIC NET BENEFIT STREAM, ASRICULTURAL PRODUCTIDN AND OFF FARM NORK ('000 Rpl Base Case -472 -1589 -2252 -2017 32 135 105 157 154 282 384 480 496 489 482 487 482 481 480 478 IRR AND NPV AT IOX 1RR NPV ('000 Rp) These IRRs and NPVs assume that the full cost of infrastructure investments is inputted to the Base Case 31 -2865 project. In fact, part of the benefits accrue to the local population and spontaneous migrants (Calculated over thirty year period) (see para 3.05 in main text). IRRs which take these factors into account are shown in the text (Table 3.4). Notes (values in '000 Rp unless otherwise noted) Total Family Labor Supply (Days) 350 364 379 394 409 426 443 461 479 498 518 518 518 518 51B 518 518 Potential Off Farm Income (1): Financial 220 220 220 220 220 223 225 232 237 246 254 263 272 282 292 303 314 CD Potential Off Farm Income.(l(: Econosic 200 200 200 200 200 203 205 211 216 223 231 239 248 256 266 276 286 Family labor valued at subsistence income in Java 270 270 270 270 270 274 277 284 291 302 312 323 334 346 359 372 38b Subsistence Income in Java, Economic terms 222 222 222 222 222 225 228 234 240 248 257 266 276 285 296 307 318 Assumptions about growth iv per capita 8Y9: 0.0% 0.01 0.0% 0.0% OO. 1.3X I.1X 2.8% 2.51 3.5X 3.5X 3.51 3.5% 3.5% 3.71 3.71 3.71 Total labor available for off farm work after farming activities (days) (1): 100 114 99 94 109 126 143 161 179 198 218 21B 218 218 218 218 218 Off farm income, economic: 105 120 104 99 115 134 154 178 203 223 231 239 248 256 266 276 286 Off fara income, financial: 116 132 114 IOB 127 14B 169 096 224 246 254 263 272 282 292 303 314 (1) Maximum amount of off farm work available assumed to be 190 days. (D > H* W o o h0 FAR BUDGET: 9 amps-Improved (Meinr Inpntt) RICE VALUE (BO24): Average of Import and Export Parity Price fran 1985 owards YEAR Cramdities 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Rice Yield (t/ba) 1.5 1.75 2 2.4 2.4 2.4 2.4 2.4 2.4 2.4 2.4 2.4 2.4 2.4 2.4 2.4 2.4 Area (ha) 0.4 0.7 1 1.3 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 Econ Price (Rp/kg) 184 176 165 117 98 83 119 155 192 189 186 183 180 178 178 178 178 Fincl Price (Rp/kg) 152 149 152 127 125 150 150 150 190 190 190 190 190 180 180 180 180 Com Yield (t/ha) 0.8 1 1.2 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 Area (ha) 0.1 0.2 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 Fcam Price (Rp/kg) 135 167 170 149 118 98 113 128 144 143 143 142 142 141 141 141 141 Finl Price (Hp/kg) 120 140 150 130 130 130 130 130 130 130 130 130 130 140 140 140 140 Cassava Yield (t/ba) 8 10 12 15 15 15 15 15 15 15 15 15 15 15 15 15 15 Area (ha) 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Fcn Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Fixl Price (Rp/kg) 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 Peanuts Yield (t/ha) Area (ha) Fcan Price (Rp/kg) 469 584 501 535 398 350 362 374 387 385 382 380 378 375 375 375 375 Firl Price (Rp/kg) 400 400 430 450 450 450 450 450 450 450 450 450 450 450 450 450 450 Estate Yield Ct/ha) Area (ha) Fcon Price (Rp/kg) Fincl Price (Rp/kg) Vegetables/Fruit ('OCO Rp) 40 50 60 60 75 75 80 90 90 100 100 100 112 112 112 112 112 Livestorc ('00D Rp) 12 20 25 30 40 40 40 40 40 40 40 40 40 40 40 40 40 Coats of Production Planting l4aterial ('0CO Rp) 5.2 9.5 12.5 15.5 18.8 18.8 18.8 18.8 18.8 18.8 18.8 18.8 18.8 18.8 18.8 18.8 18.8 0 Urea (t)(first 3 years provided free by GOI) 0.10 0.10 0.12 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 0.16 N Ecrnaairi Cost-Rp/kg 250 223 269 266 212 243 259 275 295 298 302 305 308 311 311 311 311 Financial Cost-Rp/kg 99 96 104 95 120 130 150 150 150 150 150 150 150 200 200 200 200 TIN (t) (first 3 years provided free by GOI) 0.05 0.05 0.06 0.08 0.08 0.08 0.C8 0.18 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.18 0.08 Fconomic Cost-Rp/kg 240 238 237 222 236 236 238 239 241 244 247 250 253 256 256 256 256 Finruial Coat-Hp/kg 99 96 104 95 120 130 150 150 150 150 150 150 150 200 200 200 200 KC13 (t) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 BElmai Cwot-Rp/kg 176 169 180 184 181 181 181 181 190 190 190 190 190 196 196 196 196 Finawcial Cost-Rp/kg 99 96 104 95 90 120 150 150 150 150 150 150 150 200 200 200 200 Pesticide kg) 4 5 6 6 7 7 7 7 7 7 7 7 7 7 7 7 7 Economic Cost-('000 Rp/kg) 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 Fiinaial Cost-('000 Rp/kg) 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 3 3 3 3 5 5 5 5 Kieserite Ocg) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 FBrnasic Cost-('00 Rp/kg) 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 Finarcial Ccst-'OOO p/kg) 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17 Hired Labor (days) 0 0 0 6 41 24 7 0 0 0 0 0 0 0 0 0 0 Total Labor (days) Z50 300 350 400 450 450 450 450 450 450 450 450 450 450 450 450 450 Finarcial Cost-('OOO Rp/day) 1.20 1.20 1.20 1.20 1.20 1.22 1.23 1.26 1.29 1.34 1.39 1.44 1.49 1.54 1.59 1.65 1.72 Other ('OCO Rp) 45 45 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 Producticn (t) Rice 0.600 1.225 2.000 3.120 3.840 3.840 3.840 3.840 3.840 3.840 3.840 3.840 3.840 3.840 3.840 3.840 3.840 Com 0.080 0.200 0.360 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 0.450 Casssava 0.8C0 1.OCO 1.200 1.500 1.500 1.500 1.500 1.500 1.500 1.500 1.500 1.500 1.500 1.500 1.500 1.500 1.500 PNants DH r Estate Crop I-. 0 - Eacnomic Value of Field Crops ('000 Rp) 136 268 414 461 458 391 536 681 831 819 807 795 784 775 775 775 775 Financial Value of Field Cropa ('000 Rp) 116 230 381 483 567 663 663 663 817 817 817 817 817 783 783 783 783 Gross Eamr Incue ('000 Rp) Econcmic 188 338 499 551 573 506 656 811 961 959 947 935 936 927 927 927 927 Financial 168 300 466 573 682 778 783 793 947 957 957 957 969 935 935 935 935 Total Production Cots ('000 Rp) Economic, Including Incure Foregone. Java 338 346 333 355 385 380 372 375 384 193 403 413 423 433 444 455 466 Fincl. Hired Lebor 71 76 50 65 117 100 84 75 75 86 86 86 86 112 112 112 112 4GR:CtL'URUE ALONE Net Economic Fare7 Income i-0100 Ro; lcmludes economic value Ot GOI expendituresi -149 -6 166 :9' 1SB ;2' 284 4b S, b 5S5 544 52'2 513 494 484 473 461 461 Net ;etLrq tc Household ( 000 R8) 1,2 23B 416 508 Sb5 b7O 699 71B 871 B7.1 6! 871 883 823 823 823 823 Off Farm Income: Financial (RS '0021 116 74 03 0 0 ' 0 13 36 62 9' 94 98 1 01 105 109 113 084 Farm Iicome: Econom-: (RDo 000d 105 67 3C 0 C 0 0 13 33 57 87 86 e9 02 95 99 102 ASRICULTURE AND OFF FAPM WORK Net Economic Household Income (RD '00Qi -44 59 19O '95 I1S 126 284 448 609 622 b27 608 602 586 579 572 564 564 Net Return to Household (RD 000: 2238 312 467 508 565 678 699 771 908 937 962 965 980 924 928 932 936 OTHER MOT DEVELOPMENT COSTS PER 06 Base 472 1589 2252 1952 10 10 9 9 ECONOMIC NET BENEFIT STREAM, AGRICULTURAL PRODUCTION Base -472 -1589 -2252 -2101 -18 156 18b 179 126 284 476 576 565 544 522 53 494 484 473 461 461 IRR AND NPV AT 10O IRR NPv Base 3X -2018 ECONOMIC NET BENEFIT STREAM. AhRICUETLTRAL PRODUCTION AND OFF FARM WORh Base -472 -1589 -225' -1996 49 186 186 179 12o 284 448 b09 b22 627 608 60'2 586 579 572 564 564 8RR AND N°v AT 102 IRR NPV These IRRs and NPVs assume that the full cost of infrastructure inrestments is iqldptted to the project. In fact, Dart of Base 4X -2435 the benefits accrue to the loca; population and spontaneous migrants (see Para 3.05 in main tect!. IRRs which take these factors into account are snoown In the text (Table 3.4: Total FamilV Lator SuppIy kDavs) 350 304 579 394 409 426 44- 461 479 49B 518 518 51 518 518 518 51B Off Fare, Income: Fhnan;;al 2.2' 226 ,2b 220 .20J 223 225 232 237 246 254 263 272 282 292 3fi3 314 00f Farm In-ooms: Econom7ic 200 200 200 200 200 2037. 205 311 2,b 223 331 239 248 256 2b6 27b 286 Suosisteoce income in uava, Financial 270 27;: 2 70 '70 2'70 274 277 284 291 302 312 323 334 346 359 372 386 Suosi3tence Income :r Java, EconDmic 222 222 22 222 222 225 228 234 240 248 257 2u6 276 285 296 307 318 Assun ptions a0out oer caoita 8D0: 0.02 0O.'.2 0., 0 0 .0 Il 1 ' i 1. ". 2.83 2.5 3.52. . - 3I. .. 57. 3.52 3.72 3.72 3. 72 18054 i.Ou ,22 !,oi'H 1.00 ' - 1.i0 1.00 IcC 1,'0 122 05 1.08 1.12 1:6 1.20 1.24 1.28 1.33 1.38 1.43 Total IaDor avail fcr off farmi Nork: 100 b4 29 1 ; j ' 11 29 4B 68 68 bE 68 be 68 68 co Days reonired f4r off farmi worn as,ffeed to be 190 X olt tarT income, economIc; 105 67 T0 J C17 63 86 B ' 9 59 2 w off farr, :ncome, ftnancial 1l6 74 23 C i t 3 '6 62 91 94 90 101 105 109 113 i NES SCHEME ECONOMIC ANALYSIS 2 HA RUBBER, .5 HA HOUSELOT ORD GARDEN AREA (CONSTANT 1986 '000 RP) Year 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Production ST 0.9 1.4 1.8 2.15 2.45 2.7 2.05 3 3 3 3 2.9 2.15 2.b 2.4 2.2 1.95 1.7 1.5 1.3 1.1 0.85 0.8 0.7 Selling Price 955 962 969 970 984 984 984 984 984 984 984 984 984 984 984 984 984 984 984 984 984 984 984 984 Value of Sales 860 1347 1744 2090 2411 2657 2804 2952 2952 2952 2952 2854 2706 2558 2362 2165 1919 1673 1476 1279 1082 836 787 689 Operating Costs Inputs 209 209 200 209 209 209 209 209 209 209 209 209 2109 209 209 209 209 209 209 209 209 209 209 209 Household Labor 11) 248 257 266 276 285 296 307 318 318 318 318 318 318 318 318 318 318 318 318 318 31B 218 318 318 Investment Cnsts Hoasehold Labor (1) 222 222 225 228 234 240 Fertilizers/Agroches. 625 113 151 285 303 322 Materials Equipment 404 62 36 27 27 106 Agricultural Mach. 44 44 36 Overheads/Mgmt. Fee 12) 252 76 50 53 52 69 Cash Payment 73 Houasing 930 Roads 251 251 251 Buildings 42 42 42 Others 151 151 151 Total CostS 2064 1890 949 593 616 737 457 466 475 405 494 505 516 527 527 527 527 527 527 527 527 527 527 527 527 527 527 527 527 527 Net Benefit Stream -2064 -1890 -949 -593 -616 -737 402 880 1269 1613 1916 2151 2288 2425 2425 2425 2425 2326 2179 2031 1834 1637 1391 1145 949 752 555 209 260 161 Inner 30 year period) *ET BENEFIT STREAM, RIES POOJECTS FOR TRAOSMIBOANTS (INCLUDES MINISTRY 8F TRANSM16RATION COSTS FOR NIES PROJECTS) Additinoal M0T Costs 1166 B 0 0 0 Net Benefits from .5 ha houselot/garden 13) 25 53 09 97 91 90 104 123 138 136 135 134 143 142 142 142 142 142 142 142 142 142 142 142 142 142 142 Net Benefit Stream -3230 -1898 -931 -548 -535 -640 493 971 1373 1736 2054 2280 2423 2550 2567 2566 2566 2468 2320 2173 1976 1779 1533 1287 1090 893 697 451 401 161 IRR= 131 NPV(l0%)= 2418 (over 30 year period) NET BENEFIT STREAM, NIES PROJECTS FOR TRARSMISRANTS WITH OFF FOARM WOOK STARTING IN 1988 Economic Value to Off Farm Work (4) 205 211 216 164 192 221 248 256 266 276 286 286 286 286 286 286 286 286 286 286 286 270 270 270 270 270 270 Net Benefit Stream -3230 -1898 -931 -343 -324 -424 657 1163 1595 1984 2310 2554 2699 2844 2853 2852 2852 2754 2606 2458 2262 2065 1019 1573 1360 1163 967 721 671 431 1RB= 15% lfP0(l0Bi= 4,086 (uver 30 year period) (1) Shadow priced subsistence income in Java. Land clearing is smoothed uver the first two years. 12) Overheads are 150 of the direct financial costs of field crop establishment, excluding agricultural machinery (financial costs iv Chapter 111, Table 3.81. cn (3) Proxy for this is one half of the net benefits of the low input foodcrop model, after the opportunity cost of labor is excluded. -D > (4) Transaigrant houasholds assumed to work an average of 269 days per year four maintenance and harvesting of rubber once rather starts to produce, and about 00 dayslyear on their garden. The yoUng tTanSmigrant family starts to work off farm in yEar 4 and works 190 days per year in off farm work for 3 yLars until harvesting of rubber begins. Then off farm work declines for three years until the faxily labor supply has grown to the level to supply 190 days ner Year for off farm work, after meeting labor requiresents for rubber and the garden. NES/PIR OIL PALM ECONOMIC ANALYSIS 20HA01IL PALM, .5 HA HOUSELOT AND BARDEN AREA (CONSTANT 1906 '000 RPI Year 1985 I9D6 1987 19BB 19D9 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2002 2004 2005 2006 2007 2000 2009 2010 2011 2012 2013 2014 Production MT FFB 12 20 24 29 30 32 32 32 32 32 32 32 32 30 30 20 26 26 26 24 24 22 22 22 20 20 Selling Price 00 71 71 70 70 09 09 09 69 09 09 09 09 09 09 09 69 09 09 69 69 09 09 09 09 69 Value of Sales 720 1420 1704 1960 2100 2200 2200 2206 2208 2200 2200 2208 2200 2070 2070 1932 1932 1794 1794 1050 1056 1518 1516 1518 1380 1380 Operating Costs Inputs 333 333 333 333 333 333 333 333 333 333 333 333 33 33 33 333 33 33 33 33 33 33 33 33 33 3 Household Labor 11) 234 240 240 257 266 270 285 290 307 318 318 318 318 318 318 318 318 310 318 310 310 318 318 318 318 310 Investment Costs Hoasehald Labor (1) 222 222 225 228 Fertilizers/Agrochem. 794 303 511 416 lach Service/Plant. Hat. 326 97 52 52 Agricultural Mach. 30 30 30 Overheads/ Mgmt. Fee 12) 233 97 101 90 Cash Payment 73 Housing 930 Roads 251 251 251 Buildings 34 34 34 Others 191 151 151 Total Casts 2119 2120 1300 792 507 573 581 590 599 009 610 029 040 051 051 051 651 651 051 051 051 651 051 651 651 651 651 051 051 651 Net Benefit Stream -2119 -2120 -1360 -792 153 847 1123 1370 1501 1599 1590 1579 1568 1557 1557 1557 1557 1419 1419 1281 1281 1143 1143 1005 1005 867 807 867 729 729 18R 141 NP (01 )l0l 2,345 lover 30 pear period)I NET BENEFIT STREAM, NES PROJECTS FOR TRANSMIGRANTS (INCLUDES MINISTRY OF TRANSM16RATION COSTS FOR NES PROJECTS) n.j- Additional NOT Costs 1100 a 0 0 8 Net Benefits from .5 ha houselot/garden 13) 25 53 89 97 91 90 J04 123 138 136 135 134 143 142 1421 142 142 142 142 142 142 142 142 142 142 142 142 142 Net Renefit Stream -3284 -2128 -1343 -747 234 944 1214 1460 1005 1722 1720 1715 1703 1091 1700 1699 1699 1501 1501 1423 1423 1285 1285 1147 1147 1009 1009 1009 871 871 100 130 NPO )1HZl= 2007 loner 30 year period) NET BENEFIT STREAM, NES PROJECTS FOR TRANSMIGRANTS 8170 OFF FARM WORK STARTING In 1909 Economic Value to Off Farm Work 141 144 171 211 231 239 248 250 200 270 280 200 286 280 286 280 286 280 200 206 286 280 286 286 286 286 286 Net Benefit Stream -3284 -2128 -1343 -747 378 1114 1425 1091 1844 1970 1904 1981 1979 1977 1980 1989 1985 1847 1847 1709 1709 1571 1571 1433 1433 1295 1295 1295 1157 1157 IR8 150 NPV )10I)= 3585 loner 30 pear period) (1) Shadow priced subsistence income in Java. Land clearing is smoothed over the first twn years. 12) Overheads are 150 of the direct financial costs of field crop establishment, excluding agricultural machinery Ifinancial costs in Chapter III, Tahle 3.B). 1 13) Proxy for this is one half of the net henef its of the low input foodcrop model, after the opportunity cost of labor is excluded. 14) Transmigranit households assumed to mork an average of 103 days/year for maintenance and harvesting of oil palm after (D trees begin to produce, and about 80 days/year on their garden, leaving time to pursue off faremacrk starting year 5 on site. In year 5, settler works about 130 days off farm, rising to 150 the next year, 180 days the following year, and 190 days per year thereafter (the maximum number of days of murk assumed to be available). NES/PIR SCHEME ECONOMIC ANALYSIS 2 NA HYBRID COCONUTS, .5 HA HOUSELOT AND BARDEN AREA (CONSTANT 1986 '000 HP) Year 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2009 2009 2010 2011 2012 2013 2014 Production MT FF8 1.28 1.60 3.04 5.60 6.60 7.40 7.40 7.40 7.40 7.40 7.40 7.40 7.40 7.40 7.40 7.40 7.40 7.40 7.40 7.40 7.40 7.00 7.00 7.00 7.00 Selling Price 262 259 255 251 247 244 244 244 244 244 244 244 244 244 244 244 244 244 244 244 244 244 244 244 244 Value of Sales 335 414 775 1406 1630 1806 1806 1806 1806 1806 1806 1806 1806 1806 1806 1806 1806 1806 1806 1806 1886 1708 1708 1708 1708 Operating Costs Inputs 353 353 353 353 353 353 353 353 353 353 353 353 353 353 353 353 353 353 353 353 353 353 353 353 Household Labor (1) 248 257 266 276 285 296 307 318 310 318 318 318 318 318 318 318 318 318 318 318 318 318 318 318 Investment Costs Household Labor (II 222 222 225 228 234 240 Fertilizers/Agrochem. 473 133 190 209 227 227 Tools, Mach./Plaot Mat. 545 70 27 36 36 36 Agricultural Mach. 36 36 36 OverheadslMgmt, Fee (2) 270 74 59 59 60 34 Cash Payment 73 Housing 930 Roads 226 226 226 Buildings 42 42 42 Others 178 178 178 Total Costs 2066 1910 982 532 557 537 601 610 619 629 638 649 660 671 671 671 671 671 671 671 671 671 671 670 671 671 671 671 671 671 Met Benefit Stream -2066 -1910 -982 -532 -557 -201 -186 165 787 1001 1168 1157 1146 1135 1135 1135 1135 1135 1135 1135 1135 1135 1135 0135 1135 1135 1037 1037 1037 1037 HRR= 9% NPV (1011= -596 (over 30 year period) --- - - - - - - - - - - - - - - - - - - - - - - - - - -- - - -- --- - - -0-- - - - - - - - - - - - - - -- - - - - - - - - - - - - - - - - - - - - - - - - - - -- - - - - - - - - - - - - - - - - - - - - - - - - - NET BENEFIT STREAM, NES PROJECTS FOR TRANSMI8RANTS (INCLUBES MINISTRY OF TRASHSISRATIDN COSTS FOR %ES PROJECTS) Additional NOT Costs 1166 a 8 8 8 Net Benefits from .5 ha houselot/garden (31 25 53 89 97 91 90 104 123 138 136 135 134 143 142 142 142 142 142 142 142 142 142 142 142 142 142 142 142 Net Benefit Stream -3232 -1918 -965 -487 -476 -lOS -96 256 891 1125 1306 1293 1281 1269 1277 1276 1276 1276 1276 1276 1276 1276 1276 1276 1276 1276 1179 1179 1179 1179 IRR= 92 NPV ll01(= -875 (over 30 year period) NET BENEFIT STREAM, NES PROJECTS FOR TRANSNIORANTS WITH OFF FARM RORK STARTIN6 IN 1988 Economic Value to Off Farm Work (41 205 211 216 223 231 239 240 256 266 276 286 286 286 286 286 286 286 206 286 286 28,6 286 286 286 286 286 286 Net benefit stream -3232 -1918 -965 -282 -265 III 127 487 1130 1373 156,2 1559 1557 1555 1563 1562 1562 1562 1562 1562 1562 1562 1562 1562 1562 1562 1465 1465 1465 1465 IRR= 112 NPV (10l(= 856 (over 30 year period) (11 Shadow priced vubsistence income in Java. Land clearing is smoothed over the first two years. (2) Overheads are 152 of the direct financial costs of field crop establishment, excluding agricultural machinery (financial costs in Chapter 1II, Table 3.9), (3) Prosy for this is one hulf of the net benefits of the low input foodcrop model, after the opportunity cost of labor is excluded. o (4) Transmigrant households are assumed to work an average of 166 days/year on maintenance and harvesting of coconuts, andJ. about 80 days/year on their garden, which permits them tg work the maximum of 190 days a year of off farm employment from year 4 onwards. %M PLANTED RUBBER ECONOMIC AN9ALYSIS. 2 HA ROBBER, .5 HB HOUSELOT AND GARDEN AREA I (CONSTANT 1996 '000 AP) Year 1985 1986 19B7 1999 0989 1990 1990 1992 1993 0994 1995 1996 1997 19498 1999 2000 2001 2000 2003 2004 2000 2006 2007 2009 200R 2000 20001 2012 0013 2014 Production MT 0.95 1.00 1.70 2.00 0.30 2.60 0.80 2.90 2.10 2.60 2.40 2.20 2.00 1.60 0.00 0.30 0.00 0.90 0.70 0.70 0.70 0.70 0.70 0.70 Sellin; Price 933 962 909 976 9B4 R8R 904 RAN 94 98 904 904 RBR 99 4 99 4 9 B4 994 94 994RA RN 904 964 RA RAN R94 904 Value of Sales 012 1251 0647 1952 2263 2009 2300 2755 2906 2558 2411 2065 0969 1701 0000 0279 0033 996 730 699 699 699 6AM 69 Operating Costs Inputs 200 250 200 250 250 200 250 200 200 250 200 250 200 200 200 250 200 250 120 lO5 125 125 125 125 Household Labor (I) 249 257 266 279 290 296 307 318 319 009 319 319 309 319 319 319 310 009 319 309 3109 318 311 399 On'stustnt Costs Household Labor (1) 222 222 220 209 239 240 FerbItileesfAgrochee. 099 001 227 322 340 390 MaterialsEqucipment 931 1A 19 9 27 00 Agricultural Rach, 19 OR 09 OseheadslMget. Fee (2) 203 74 61 03 07 67 Reads 67 67 67 Buildings 07 07 07 Others 67 67 59 Total Costs 1662 639 692 613 659 727 499 507 006 026 535 546 057 599 06B 568 568 060 568 568 068 569 069 568 442 443 443 443 443 442 Net Benefit Strea.. -0662 -634 -6R2 -613 -659 -727 314 743 0130 0426 0729 2012 2090 2197 2139 1990 1943 0597 1400 0203 957 701 460 207 293 246 246 246 246 246 IRRo 090 NPO)1OZt' 2,622 (over 30 year period) NET BENOV09 006000 INCLUDING SETTLEMENT COSTS COOT ALTERNBTOYE ORE: Additional NOT Costs (31 472 1906 0607 00 00 9 9 Net Aweleits from .0 ha hiuselot garden (4) 25 53 99 97 91 90 104 023 129 036 030 024 042 042 042 042 142 042 142 040 042 042 042 042 942 Net Benefit Stream, Alternative Bee -472 -1906 -0607 -0672 -644 -676 -069 -069 -630 404 934 0235 0049 1966 2049 2333 2320 2290 2002 19A4 0739 0541 0340 1009 933 607 459 437 387 W1 049' 90 NP0(10010 -671 (over 30 year period) COST ALTERNA0IVE TWO: Additional EcEtlgeest Costs, ME$ Levels (3) 1032 1231 3099 Ba Net Benefts free .0 ta houselot garden (4) 25 03 99 97 91 90 004 023 (39 026 005 134 043 042 142 042 042 042 042 042 042 042 042 942 142 942 642 042 Net Benefit Strea, Alternatioc T.o -3194 -10995 -976 -567 -577 -630 404 934 1235 0049 0966 2149 2332 2320 0290 2032 0994 0739 0540 0343 1009 953 607 459 407 397 397 397 397 397 099' 110 NP0(0701= 773 (over 30 year. per iod) NOT 09NOV07 STREAM, INCI.UDIN9 OFF FARM NORE ETARTINSI 048099 Economic Value to Off Fare Nork 05) 205 201 206 164 192 200 049 206 266 276 296 296 296 296 266 296 296 2936 286 296 296 270 270 279 279 279 Net BeneRit Streaen Alternative One -472 -1006 -1657 -0672 -644 -676 -069 -364 -419 900 996 0426 0770 2113 2405 2599 2096 0569 2409 2270 2024 0927 0630 1364 0038 892 745 702 657 657 270 279 279 OARR' 010 4PV1001' 419 (noer 30 year period) Met Benef it Stream, Alterntive T.. -3194 -1965 -976 -567 -372 -419 620 998 1429 0771 2003 2400 2599 2596 2066 2409 2270 2024 1927 1630 1384 1138 992 740 722 657 607 657 657 907 08R' 130 4P0(l00(= 0,275 (nuer 30 year period) (1) Shadow priced subsistence, income. in Java. Land ciearing is smoo..thed aver the First two years. (2) Overheads are 05% of the direct financ. .ial casts of field crap establishmet, e-cldong agrirsiceral machinery (fina-cial roars in Chapter III, Table 3.0). (3) Since this model, has oot hoes. used to settle tra-smigraor, and it is so.t clear what institutiono will provide the settlemen. .t infrastruerure what the associated Costs will he, two aiter-tiv-s are presented. The first use. Ministry of Trannruigratiom standard costs far settlement with the land clearing a-d agricoltoral impel package halved, to reflect the tedoced foodorep area (0.5 ha cleared instead of 1 ha cleared for foodcrops). The second alte,rnative uses NOV leve invest.eet for housing, roads, buildings, rash paynent to settler, and other. Settlers provide their ow foodarop agricultural inputs. MHOT proides imvetremts tar ancillary serices se .. ttlement and relocation only (US$1,362 as is NOES Projects). The second alternative is less eaPensive, and show. that the rate of rotors is slightly higher if settle- nest costs Can, be c..ntained to NES level. (4) Pramy far this in soe-half of the met bsfr of the low input foadcrop model, after the opp-rt-ity cat of lahor Is macladed.isB 5) Transmigrant hoon...hoids asso.ne.d to wok as avrage of 269 days per year for macnt-aance and (harvesting of rubber onc robber starts to prodoce, and shout 00 days/yea -o their garden. The ya.mo.g croosnigrat fasily starts to work off farm In year 4 and work. i9o days per year In oft farm work far 3 yearun til harvesting of ruhher begins. Thea off fare work decli. es far three years until the famnily labor sapply h.. grows to the level to sopply 090 days per year for off fats work, after meeting labor reojlrenonts for robber and rho garden. PHU1 PLANTED RUBBER AS SECOND SATGE DEYELOP9ENT FOR EXISTING TRANSTNIORANTS. RUBBER PLANTED ON 1.9 HA (CONSTANT 1986 '000 RP) Year 1985 1986 1987 I9AA 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2095 2096 2007 2008 2009 2010 2011 2012 2013 2014 WNitlDUt Project Scenario Net AqniCUltUre InEOce from I ha foodtrops 194 180 181 209 292 2761 273 270 2118 2985 282 282 282 282 282 282 282 282 282 2B2 283 282 282 283 283 283 283 283 283 283 Off farm income IDA 173 200 200 200 2013 205 211 216 223 231 239 248 250 266 276 286 286 286 286 286 28A 286 286 286 286 286 286 28A 286 Nith Project Scenario, P40 and settlers clear I ha reserve land and replace .5 ha oN foodcrop with rubber Production MT, on 1.5 ha. 0.61 0.93 1.21 1.43 1.A4 1.85 2.00 2.00 1.96 1.85 1.75 1.57 1,43 1.2B lIA0 0.93 0.75 0.64 A.53 0.5A 0.50 A.5A A.50 A.50 Sellmng Price 955 962 969 976 984 984 984 984 984 984 984 984 984 984 984 984 984 984 984 984 984 484 984 984 Value of sales 578 891 1174 1391 1613 1823 1963 1963 192B 1823 1718 1542 1402 1262 1087 911 736 631 526 491 491 491 491 491 Operating Costs Inputs 188 188 188 lAB 188 188 188 188 188 188 188 188 188 188 188 188 188 188 914 94 94 94 94 94 Nousehnold Labor (11 Investment Costs Household Labor (ll Fertilizers/Oqrochem. 440 113 170 242 255 255 Materials Equipment 324 13 13 7 20 60 Agricultural Mach. 19 19 19 OverheadsiNgmt. Pee (2) 190 5b 46 40 92 30 Roads 87 67 67 Buildinqs 17 17 17 Others 67 87 59 Total Costs 1123 352 391 289 318 365 188 188 188 198 188 188 188 188 188 188 188 188 188 188 188 188 188 188 94 94 94 94 94 94 '. C) Net Benefit Stream, Treecrops -1123 -352 -391 -289 -318 -365 391 703 986 1203 1425 1635 1775 1775 1740 1635 1530 1355 1215 1074 899 724 548 443 432 397 397 397 397 397 00 Barden Production (3) 97 91 90 104 123 138 136 135 134 143 142 142 142 142 142 142 142 142 142 142 142 142 142 142 142 142 142 142 142 142 Net Benefit Stream mith tarden Production -1026 -281 -300 -184 -195 -227 527 839 1120 1346 1566 1772 1917 1917 1882 1777 1672 1496 1356 1216 1041 885 690 585 574 539 539 539 539 539 With Pruiect M4inas Nithoot Project Net Benefit Stream -1219 -442 -481 -393 -442 -503 254 568 852 1081 1283 1494 1634 1634 1599 1494 1388 1213 1073 933 757 582 407 302 290 255 255 255 255 255 4RR= 165 NPV(OXlO= 2,214 (over 50 year period) Including Off Parm Nork 205 211 216 164 192 221 248 258 268 276 286 246 286 286 286 286 286 286 286 286 288 270 270 270 278 270 270 With Project Minus Nitbout Projeot Met Benefit Streas, locloding Ofif Fvrs Work -1577 -615 -681 -388 -431 -409 214 550 898 1085 1308 1520 16i62 1663 1619 1504 1388 t213 1023 933 757 582 407 302 275 248 240 240 240 240 I4RR 153 151 NPOIIozlz 1,811 lover 30 year period) (0 Coot of labor is the opportunity cost of labor, or the witlnout project scenario benetits from food crops and oft far. income. (2) Overheads are 05% of the direct financial coors of field crop establishment, excluding agricultural machinery (financial roots in Chapter 111, Table 2.0). (3) Proxy for tlnis Os one-half of the net beniefits of the low input foodcrop moodel, after the opportunity cost of labor in excluded. (4) Transmigrant households assumed to wnork am average of 200 days per year for maintenance and harvesting of 1.5 ha of robber once robber c o starts to produce, and shout 00 days/year on their garden. The young transmigrant family starts to work off farm in year 4 and works 190 ~ days per year in off farm work for 3 years until harvesting of rubber begins. Then off farm work declines for three years until the family labor supply requiremnets for rubber and the garden.M (5) Rate of retuin declines sligintly when off form work is incorporated into the amalyls, because the settlers lone some benefits of off farm o work wlv9 labor is shifted to treecrops. ANNEX 3 - 209 - Table 17 Projected Rice Yields at Appraisal from World Bank-Financed Transmigration Projects and Yields in Models Presented in Chapter III Year of Staff Rice Yield At Project or Model Appraisal Report Full Development (t/ha) Project Upland Transmigration I (Ln. 1318) 1976 2.10 Transmigration II (Ln. 1707/Cr. 919) 1979 1.70 Transmigration III (Ln. 2248) 1983 2.00 Transmigration IV (Ln. 2288) 1983 1.50 Tidal Swamp Reclamation I (Ln. 1958) 1981 2.40 Swamp Reclamation II (Ln. 2431) 1984 2.50 Models in Chapter III Upland low input 0.75 Upland diversified 1.00 Upland low input with cow 1.20 Upland sustained input 1.80 Upland high input with cattle 2.50 Tidal, severe pests 0.75 Tidal low input 1.25 Tidal improved 2.00 -20- ANNEX 4 Table 1 Population and Transmigrants in Receiving Provinces ('000) Provincial population Increase Growth rate Transmigration population Province 1980/a 1985/a 1971-80 1980-85 1980-85 (%) 1971-80/b 1980-85/c 1971-85 Aceh 2,611.3 2,999.9 602.7 388.6 2.81 9.6 61.0 70.6 N. Sumatra 8,360.9 9,452.0 1,739.1 1,091.1 2.48 1.8 37.1 38.9 W. Sumatra 3,406.8 3,779.6 613.6 372.8 2.10 34.8 23.2 58.0 Riau 2,168.5 2,513.4 527.0 344.9 3.00 29.3 177.9 207.2 Jambi 1,446.0 1,755.3 439.9 309.3 3.95 96.0 107.8 203.8 Bengkulu 768.1 946.9 248.8 178.8 4.27 41.7 61.1 102.8 S. Sumatra 4,629.8 5,423.1 1,189.2 793.3 3.21 141.3 379.0 520.3 Lampung 4,624.8 6,089.7 1,847.8 1,464.9 5.66 133.3 188.2 321.5 Sumatra 28,016.2 32,959.9 7,208.1 4,943.7 3.30 487.8 1,035.3 1,523.1 W. Kalimantan 2,486.1 2,771.5 466.2 285.4 2.20 23.7 131.9 155.6 C. Kalimantan 954.4 1,123.9 252.4 169.5 3.32 9.4 109.8 119.2 S. Kalimantan 2,064.6 2,285.5 365.5 220.9 2.05 41.0 91.4 132.4 E. Kalimantan 1,218.0 1,603.0 484.2 385.0 5.65 29.6 55.5 85.1 Kalimantan 6,723.1 7,783.9 1,568.3 1,060.8 2.97 103.7 388.6 492.3 N. Sulawesi 2,115.4 2,359.0 396.9 243.6 2.20 11.2 18.8 30.0 C. Sulawesi 1,289.6 1,549.8 375.9 260.2 3.74 51.5 75.5 127.0 S. Sulawesi 6,062.2 6,547.7 881.6 485.5 1.55 36.5 25.0 61.5 S.E. Sulawesi 942.3 1,091.9 228.2 149.6 2.99 37.9 92.1 130.0 Sulawesi 10,409.5 11,548.4 1,882.6 1,138.9 2.10 137.1 211.4 348.5 Maluku 1,411.0 1,617.4 321.4 206.4 2.77 4.3 35.1 39.4 NTB 2,724.7 3,045.1 521.2 320.4 2.25 NA 6.9 NA E. Timor 555.3 608.4 NA 53.1 1.84 NA 1.0 NA Irian Jaya 1,173.9 1,332.0 250.5 158.1 2.56 10.6 75.6 86.2 Total 51,013.7 58,895.1 11,752.1 7,881.4 2.91 743.5 1,753.9 2,497.4 /a BPS, Statistik Indonesia 1984, Table 3.1.2. /b P. Gardner, Provincial Population Projections (Jakarta: UNCHS/GOI NUDS Project, 1985), Table 4. /c Departemen Transmigrasi, Daftar Proyek Transmigrasi Yang Dibina Tahun 1985/86 (To August 1985). 211- ANNEX 4 Table 2 1970-85 Transmigrants As % 1980 Provincial and Kabupaten Population /a Sponsored Popula- % Trans- Sponsored Popula- % Trans- Province/Kabupaten transmigrants tion 1980 migrants Province/Kabupaten transmigrants tion 1980 migrants Aceh 60,957 2,611,271 2.3 South Kalimantan 91,442 2,064,649 4.4 Aceh Utara 7,533 625,296 1.2 Kota Baru 78,913 193,650 40.8 Aceh Timur 9,271 423,418 2.2 Barito Kuala 11,442 169,952 6.7 Aceh Barat 23,878 288,422 8.3 Tabalong 1,087 124,939 0.9 Aceh Selatan 16,093 275,458 5.8 Aceh Tengah 4,182 163,341 2.6 East Kalimantan 55,513 1,218,016 4.6 Kutai 32,784 368,501 8.9 North Sumatra 37,070 8,360,894 0.4 Samarinda (KM) 2,375 264,718 0.9 Tapanuli Selatan 37,070 757,159 4.9 Balikpapan (KM) 2,586 280,675 0.9 Bulongan 6,069 176,923 3.4 Riau 177,880 2,168,535 8.2 Berau 11,699 45,903 25.5 Indragiri Ilir 10,312 398,276 2.6 Indragiri Ulu 34,246 229,182 14.9 Total Kalimantan 388,644 6,723,086 5.8 Bengkalis 38,368 566,671 6.8 Kampar 91,582 362,867 25.2 North Sulawesi 18,817 2,115,384 0.9 Kepulauan Riau 3,372 425,277 0.8 Gorontolo 12,146 502,695 2.4 Bolaang Mongondow 6,671 299,696 2.2 West Sumatra 23,153 3,406,816 0.7 Sawah Lunto 17,594 224,446 7.8 Central Sulawesi 75,531 1,289,635 5.9 Pasisir Selatan 4,288 315,954 1.4 Banggai 27,396 268,203 10.2 Solok 1,271 355,539 0.4 Donggala 17,818 581,772 3.1 Poso 21,794 266,708 8.2 Jambi 107,819 1,445,994 7.5 Buol Toli-Toli 8,523 172,952 4.9 Bungo Tebo 24,457 302,386 8.1 Sarko (Sarolangun) 48,585 217,653 22.3 South Sulawesi 24,991 6,062,212 0.4 Tanjung Jabung 20,922 216,897 9.6 Luwu 14,424 503,757 2.9 Batang Hari 13,855 237,604 5.8 Mamuju 10,567 99,796 10.6 Bengkulu 61,112 768,064 8.0 Southeast Sulawesi 92,114 942,302 9.8 Bengkulu Utara 49,207 178,250 27.6 Kendari 59,583 306,675 19.4 Bengkulu Selatan 7,465 236,775 3.2 Kolaka 4,566 144,446 3.2 Rejang Lebong 4,440 288,256 1.5 Ruton 21,662 317,124 6.8 Muna 6,303 174,057 3.6 South Sumatra 378,959 4,629,801 8.2 Banyuasin 250,012 591,074 42.3 Total Sulawesi 211,453 10,409,533 2.0 Ogan Komering Ilir 42,595 564,080 7.6 Ogan Komering Ulu 27,100 750,799 3.6 Maluku 35,139 1,411,006 2.5 Lahat 32,204 484,893 6.6 Maluku Tengah 35,139 443,940 7.9 Musi Rawas 19,937 367,037 5.4 Lematang Ilir 7,111 430,834 1.7 West Nusa Tenggara 6,878 2,724,664 0.3 Dompu 6,878 95,827 7.2 Lampung (S) 188,178 4,624,785 4.1 Lampung Utara (SS) 207,053 882,479 23.5 Irian Jaya 75,604 1,173,875 6.4 Lampung Tengah 4,949 1,690,947 0.3 Jayapura 12,071 151,308 8.0 Manokwarl 9,980 84,757 11.8 Total Sumatra 1,035,128 28,016,160 3.7 Sorong 24,881 134,833 18.5 Merauke 22,229 172,662 12.9 West Kalimantan 131,890 2,486,068 5.31 Panisi 6,443 177,619 3.6 Pontianak 10,734 608,893 1.76 Sanggau 14,540 323,499 4.49 East Timor 965 555,350 0.2 Sambas 92,792 603,104 15.39 Bobonaro 965 61,98n 1.6 Ketapang 10,870 253,828 4.28 Kapuas Hulu 2,954 128,647 2.30 All Receiving Areas 1,753,811 51,013,674 3.4 Central Kalimantan 109,799 954,353 11.51 Kapuas 56,201 258,473 21.7 Kotawaringin Timur 23,900 83,166 13.0 Kotawaringin Barat 27,730 94,544 29.3 Barito Selatan 1,940 69,020 2.8 Barito Utara 28 63,601 0.0 /a Repelitas II, III and IV to August 1985. Notes: KM = Kotamadya; S does not include resettlement within province; SS includes resettlement from other kabupaten in Lampung. Source: Department of Transmigration, 1985. - 212 - ANNEX 4 Table 3 Population by Place of Birth, Place of Previous Residence, Place of Residence Five Years Ago and Place of Present Residence (Urban and Rural) Present residence Other Sumatra Java Kalimantan Sulawesi Islands Total Place of Birth Sumatra 24,825,271 718,420 25,474 22,651 19,879 25,611,695 Java 2,906,014 89,999,829 374,097 167,393 137,364 93,584,697 Kalimantan 19,594 121,807 4,165,294 9,758 4,626 6,321,079 Sulawesi 145,417 136,745 123,413 10,l8,689 139,614 1n,653,R78 Other Islands 30,789 114,883 11,282 74,185 10,126,012 10,357,151 Abroad 39,703 59,310 10,220 7,990 7,525 124,7X8 Not stated 29,139 65,976 7,116 9,882 11,112 123,225 Total 27,995,927 91,216,970 6,716,896 10,400,568 10,446,132 146,776,4J3 Place of Previouis Residence Sumatra 24,729,720 767,199 19,676 24,093 14,265 25,556,953 Java 2,849,120 89,621,756 373,046 196,381 154,013 93,194,316 Kalimantan 25,481 156,685 6,142,569 25,364 6,183 6,356,282 Sulawesi 127,617 151,183 117,960 10,008,948 141,535 1n,547,243 Other Islands 38,146 150,707 10,780 88,719 10,083,534 10,371,886 Abroad 35,658 56,06n 9,368 9,673 8,520 119,279 Not stated 190,185 313,380 43,497 47,470 38,082 632,594 Total 27,995,927 91,216,970 6,716,896 in,400,548 10,446,132 146,776,4;3 Place of Residence 5 Years Ago Sumatra 22,544,460 261,137 12,949 12,451 10,670 22,841,667 Java 810,340 78,296,777 131,438 77,794 67,463 79,403,812 Kalimantan 10,432 43,860 5,446,630 7,855 2,246 5,511,023 Sulawesi 21,269 38,891 38,173 8,674,190 44,703 8,817,226 Other Islands 15,116 76,806 4,868 43,006 8,709,324 8,849,120 Abroad 4,154 10,269 1,243 2,611 3,265 21,542 Not stated 35,387 77,630 8,310 11,663 8,127 141,117 Total 23,441,158 78,803,370 5,663,611 8,829,570 8,845,798 125,585,506 Note: Population 5 years of age and over. Classification of Forest Land, May 1984 Total Conversion forest and Park and Limited Definitive definitive Other other utilization Province Protection reserved production production forest (ha) Conversion utiliza- Ha Percentage Province area (ha) area (ha) forest (ha) forest (ha) forest (ha) (3+4+5+6) forest (ha) tion (ha) (8+9) by province (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) Aceh 5,539,000 1,051,400 666,800 1,375,700 188,300 3,282,200 192,700 2,064,100 2,256,800 41 North Sumatra 7,168,068 1,391,100 253,900 1,349,900 531,500 3,526,400 253,700 3,387,968 3,641,668 31 West Sumatra 4,229,730 1,206,600 599,700 539,900 596,800 2,942,800 437,700 849,230 1,286,930 30 Riau 9,456,156 741,800 267,200 2,764,200 2,772,900 6,546,100 1,754,100 1,155,960 2,910,060 31 South Sumatra 10,277,500 774,700 796,500 333,000 2,124,000 4,028,200 1,186,500 5,062,800 6,249,300 61 Jambi 5,100,000 1,147,500 493,000 974,000 - 2,614,500 1,013,200 1,472,300 2,485,500 49 Bengkulu 1,978,870 465,500 249,900 242,000 34,100 991,500 193,600 793,770 987,370 50 Lampung 3,200,000 315,000 356,000 - 573,000 1,244,000 - 1,956,000 1,956,000 61 West Java /b 4,630,000 229,500 196,400 - 547,980 973,800 - 3,656,200 3,656,000 79 Jakarta /b 59,000 - 15 - 1,100 1,115 - 57,885 57,885 98 M Central Java /b 3,421,000 65,500 3,000 - 605,100 673,600 - 2,747,400 2,747,400 80 u Yogyakarta /b 317,000 3,200 200 - 13,200 16,600 - 300,400 300,400 95 East Java /b 4,791,9700 255,800 245,300 - 847,100 1,348,200 - 3,443,770 3,433,770 72 West Kalimantan 14,680,700 2,047,100 1,336,700 2,988,700 1,323,000 7,695,500 1,508,700 5,476,500 6,985,200 48 Central Kalimantan 15,300,000 800,000 729,400 3,400,000 6,068,000 10,997,400 3,000,000 1,302,600 4,302,600 28 South Kalimantan 3,700,000 432,700 66,000 200,600 1,330,400 2,029,700 284,700 1,385,600 1,670,300 45 East Kalimantan 21,144,000 3,643,900 1,968,600 4,826,100 5,513,100 15,951,700 3,500,000 1,692,300 5,192,300 25 North Sulawesi 2,751,501 285,400 326,600 741,200 230,500 1,583,700 699,400 468,401 1,167,600 42 Central Sulawesi 6,803,300 1,156,900 616,700 1,364,100 1,028,000 4,165,700 335,000 2,302,600 2,637,600 39 Southeast Sulawesi 3,814,000 420,800 273,400 827,100 668,900 2,190,200 699,400 924,000 1,623,800 42 South Sulawesi 6,292,650 2,004,100 189,600 993,100 165,000 3,351,800 259,400 2,681,450 2,940,850 47 Bali 563,286 84,100 32,000 5,700 3,900 125,700 - 437,586 437,586 78 West Nusa Tenggara 2,015,315 481,700 134,800 222,800 224,100 1,063,400 195,900 756,015 951,915 47 East Nusa Tenggara 4,738,920 677,600 131,900 399,000 278,100 1,486,600 2,801,600 450,720 3,252,320 69 Maluku 8,572,800 1,550,400 441,000 2,075,600 1,029,900 5,096,900 436,400 3,039,500 3,475,900 41 Irian Jaya 41,066,000 8,648,500 8,311,800 4,732,300 7,123,500 28,816,100 11,775,400 474,500 12,249,900 30 East Timor 1,468,937 435,300 38,800 170,500 45,200 689,800 10,000 761,137 771,137 53 Total 193,871,707 30,316,100 18,725,215 30,525,300 33,866,600 113,433,215 30,537,400 49,101,092 79,638,492 41 Source: FAO/World Bank Cooperative Program. Indonesia Forestry Project Working Paper #1. October 1985. Preliminary figures. Planned and Actual Plantings for Ongoing and New Smallholder Tree Crops Development Programs - Repelita III and IV Repelita IV Total Repelita III 1984/85 1984/85/a % 1985/86 1986/87 1 97/88 1988/R9 (1984/85 actual + Planting programs Planned Actual % Plan Actual Plan Plan Plan Plan plan 1985/86-1988/89) Rubber NES, foreign aided 59,870 54,972 92 18,462 R,886 48 i1,1i0 8,400 6,030 - 33,416 Transmigration III - - - I,0no 1,000 100 1,000 - - - 2,000 PIR Khusus 61,100 17,875 29 30,486 3,680 12 36,400 30,000 9,noo 6,000 85,080 PIR Lokal 7,000 4,725 68 2,275 - 0 4,000 2,0n0 - - 6,000 Subtotal NFS/PIR 127,970 77,579 61 52,223 13,566 26 51,500 40,400 15,030 6,no 126,496 SROP I planting/replanting 28,100 26,400 94 9,280 4,772 51 5,815 16,750 20,500 20,000 67,837 PRPTE planting/replanting 150,9R2 78,257 52 - - - 23,293 /h /b /b 23,293 PFRSB (West Sumatra) 7,600 4,994 66 806 409 51 - - - - 409 Subtotal PMU planting 186,682 109,651 57 10,0R6 518 51 2q,log 16,750 20,son 2),000 91,539 Subtotal rehabilitation (PRPTE) 50,134 893 1.8 _ - /b /b /h b Total Rubber 364,786 186,116 52 62f39 1 30 80,608 57,150 35,530 26,000 218,035 Oil Palm Ongoing Programs NES, foreign aided 14,840 16,099 96 12,482 3,411 29 10,110 10,350 4,298 - 28,569 PIR Khusus 17,000 5,440 32 21,300 7,423 35 8,00o 9,500 6,780 3,000 34,703 PIR Lokal 21,742 13,059 60 9,072 2,000 22 3,873 - - - 5,873 Subtotal 55,582 34,598 62 42,853 3,04 30 22,1R3 19,850 11,07R 3,000 6q,145 New Programs NES - - - - - - - 8,750 6,050 5,600 20,400 PIR - - - - - - 500 8,000 26,5on 35,0no 70,000 PIR Akeelerasi - - - - - - - 14,500 22,000 23,500 60,000 PIR Swasta - - - - - - - 28,500 74,000 108,500 211,000 Subtotal _- - 500 59,750 12R,550 172,60 361,400 Total Oil Palm 5 5,82 34,598 62 42,853 13,034 30 22,683 79,600 139,628 17_,60_ 430,545 Coconuts /a NES, foreign aided 5,370 6,278 117 3,857 1,823 47 3,500 2,980 1,000 3,000 12,303 Transmigration IV - - - 1,000 500 50 3,000 4,000 - - 7,500 SCDP planting - hybrids 33,399 17,693 53 8,10n 4,768 59 11,240 - - - 16,008 PRPTE planting - hybrids 9,656 6,966 72 - - - 5,113 1n0000/b 15,000/b 25,000/b 55,113 PRPTE planting - talls 168,423 125,874 75 - - - 26,164 10,0007w 35 00O7F 40 o0o7i 131,164 PPPTE rehabilitation 262,449 169,218 65 - - - - 10,0007W 15,000/b 20,000/b 45,000 Total Coconuts 479,297 326,029 68 12,957 7,091 55 49,017 56,980 66,000 88,000 267,088 Total All Crops _899665 549,743 61 118,119 38,872 152,308 193,730 239,158 289,6)00 915,668 /a Data provided by Team Khusus and the managers of each planting program, October 1985. The following programs in effect during Renelita III -- SRIP rubber rehabilitation, SCOP coconut rehahilitation, SCIP planting of talls -- are not planned for continuation in Repelita IV. /b Annual targets not finalized; 246,000 ha of smallholder plantings of tall and hybrid coconuit and 107,200 ha of smallholder coconut rehabili- tation are targeted for Repelita TV. These figures were estimated prior to recent budget constraints and may be subject to significant downward revision. -215 - ANNEX 7 Page 1 Description of Transmigrant Surveys Reviewed Large Scale Surveys 1. BPS Income Survey of Transmigrants, 1985. Survey of 2,200 transmigrants in Sumatra, Kalimantan and Sulawesi. Surveys were undertaken in 110 villages in about 22 locations in seven major receiving provinces, and in each village 20 households were randomly selected for interviews. Upland, tidal and upland sites with tree crop development were included in the survey. The survey was based on the National Social-Economic Survey (Susenas) to permit comparisons between transmigrant and nontransmigrant households. 2. Smallholder Cattle Development Project Socioeconomic Impact Survey, 1985. Survey of 1,200 households on upland sites. Of the 1,200 households interviewed, 600 had received cattle and 600 had not. Information was collected on gross cash income, including income from off farm work. 3. MOT Income Survey of Transmigration Areas, 1985. Survey of 1,500 sponsored transmigrants in Sumatra, Kalimantan and Sulawesi on upland and tidal sites. Information was collected on yields and annual on farm and off farm income. Small-Scale Surveys 4. Perry, Douglas et al. "The Economics of the Transmigrant Farmer: A Survey of Sitiung II and Sitiung III, West Sumatra", October 1983, Balai Penelitian Tanaman Pangan. Study of 30 transmigrants in rainfed upland sites in West Sumatra. Information collected on farming and off farm activities, labor inputs into upland rice, yields, and livestock. 5. Levang and Riskan Marten. "Sembamban I, Agro-economic Survey of a Transmigration Center on South Kalimantan", Indonesia-Orstom Transmigration Project, Jakarta 1984. Study of 60 transmigrants in Sembamban I, a site settled between 1979 and 1981. Information includes yields on rice, peanuts, soybeans, maize and cassava, marketed crops, non-agricultural activities and other sources of income. 6. Euroconsult-ID Consultants, Joint Advisory Services, Quarterly Report No. 6, April-June 1983, Transmigration II Project Jambi Province. - 216 - ANNEX 7 Page 2 Report contains results of a survey of 180 farmers in Singkut, a site settled between 1974 and 1979, and 77 farmers in an area of Kubang Ujo settled between 1981 and 1982. Information collected on rice yields, use of fertilizers and insecticides, fruit and tree crop developments, cash income and types of off-farm employment, farm labor, and marketing. 7. Chapman, Barbara. "Diet and Production Survey of Sitiung, West Sumatra, Research Memos 2 and 3." Tropsoils Project, December 1984. Survey was done in April 1984, and 77 households were interviewed about agricultural production, off farm income, and quality of diets. 8. Khan, Marshal. "An Inquiry into the Agricultural Economy of Singkut", Jakarta, December 1980. This survey was conducted in Singkut, a site with settlement that started in 1974. Information was collected from 78 households on the value of crops produced, net farm earnings, and off farm income, including a breakdown of types of off farm work. 9. Stuart, Robert. "Employment in Singkut, 1983". Euroconsult, 1983. Survey of 182 farmers in Singkut. Information includes detailed breakdown on off-farm employment. 10. U.H.K. Golaszinski, "Employment and Income Survey, Middle-Mahakam- Area, 1984". Employment and income survey of 388 households in eight villages in East Kalimantan. The sample included transmigrants, Kutainese and Banjarnese. Information was collected on agricultural production, livestock, off-farm employment, and income levels. 11. Levine, et al. "Transmigrants and Tolakis in Kecamatan Tinanggea Southeast Sulawesi, Socio-economic Survey Results for the Southeast Sulawesi Transmigration and Area Development Project". Resources Management International, Inc. Survey of 45 families in a transmigration village and 60 local families in an upland area of Southeast Sulawesi during December 1982- January 1983. Information collected on land and labor use and on and off-farm income. 12. Davis, Gloria. "Migrant Welfare in Five Transmigrant Communities." Working Paper, World Bank, 1980. Survey of agricultural production of 592 farmers in transmigrant communities dependent on rainfed agriculture (Sitiung, Baturaja, Rimbogujang, and Way Abung) and 219 farmers in transmigrant tidal sites. - 217 - ANNEX 8 Page 1 Methodology for Demographic Analysis Contained in Chapter IV The potential population impact of sponsored transmigration on the inner and outer islands depends upon several factors. These factors include the age and size of sponsored transmigrant households and their fertility and mortality levels and the rate at which sponsored transmigrants or their offspring might return to the inner islands. In addition, the impact depends upon the number of spontaneous transmigrants who are attracted, in turn, by the sponsored movement and the opportunities that the transmigration program creates, and the fertility, mortality and return migration rates of spontaneous transmigrants. A cohort component population projection was done to illustrate the long run population impact of alternative levels of sponsored transmigration along with high, low, and zero levels of associated spontaneous transmigration for the period 1980 to 2020. Age-Sex Structures of Population and of Transmigrants The age-sex structures for the population of the major islands in 1980 provide the base for the population projections, and this is taken from the 1980 census. Transmigrants were assumed to have the same young age-sex structure as a sample of 1120 sponsored migrants arriving in Baturaja in 1983/84 (Table 1). Household size of sponsored transmigrants is assumed to decline over the period of the population projections due to projected declines in fertility. Household size is assumed to average 4.1 persons per household from 1980 to 1984, 4.0 from 1985 to 1989, 3.8 from 1990 to 1994, and 3.6 thereafter. Projected Fertility and Mortality Rates Fertility and mortality rates for 1980 were derived from the 1980 census. To simplify the analysis, which is more concerned about the impact of migration than fertility, fertility and mortality rates were assumed to be the same for all provinces in the inner islands (Java, Bali and Lombok). A separate projection of fertility and mortality was used for the outer islands (the outer islands refer primarily to Sumatra, Kalimantan, Sulawesi, Irian Jaya and Moluku). Fertility rates are significantly lower in the inner islands than in the outer islands of Indonesia, and this is reflected in the analysis. Fertility and mortality rates are projected to decline over the - 218 - ANNEX 8 Page 2 Table 1: AGE-SEX STRUCTURE FOR SPONSORED MIGRANTS UPON ARRIVAL TO BATURAJA,/a 1983/84 (percent) Age Males Females 0-4 9 8 5-9 8 6 10-14 6 4 15-19 4 5 20-24 5 8 25-29 7 6 30-34 6 3 35-39 3 2 40-44 3 2 45-49 1 1 50-54 1 1 55-59 1 0 60+ 0 0 Total 54 46 /a Settlement site financed by Transmigration I. Sample size is 1120 people. period to replacement levell- by the year 2010 for the inner islands and by the year 2020 for the outer islands. Projected Total Fertility Rates (TFR) and life expectancies are presented in Table 2. For the purposes of the analysis, transmigrants are assumed to adopt the fertility and mortality levels of the destination area after their move. 1/ Replacement level fertility and mortality, where the net reproduction rate (NRR) is one, indicates that at these levels of fertility and mortality, childbearing women, on the average, bear only enough daughters to replace themselves in the population. Indonesia has a strong family planning program in place relative to the level of development. Nevertheless, the assumptions about declines in mortality and especially fertility are quite optimistic. - 219 - ANNEX 8 Page 3 Table 2: PROJECTED TOTAL FERTILITY RATES AND LIFE EXPECTANCIES FOR THE INNER AND OUTER ISLANDS, 1980-2020 1980-84 1985-89 1990-94 1995-99 2000-04 2005-09 2010-14 2015-19 INNER ISLANDS Total Fertility Rate/a 4.16 3.70 3.17 2.76 2.48 2.20 2.17 2.15 E(o) males/b 53 55 57 59 62 65 67 68 E(o) females /b 56 58 61 63 66 69 70 72 OUTER ISLANDS Total Fertility Rate /a 4.94 4.33 3.63 3.11 2.88 2.65 2.43 2.2 E(o) males /b 55 58 60 62 63 64 64 65 E(o) females /b 58 61 64 66 67 67 68 69 /a The total fertility rate represents the number of children that would be born per woman, if she were to live to the end of her childbearing years and bear children at each age in accord with prevailing age-specific fertility rates. /b E(o), or life expectancy, is the number of years a newborn male or female would live if subject to the mortality risks prevailing for the cross-section of population at the time of his or her birth. Assumptions about the Levels of Spontaneous Migration Associated with Sponsored Migration The demographic impact of spontaneous migration to rural areas in the outer islands is clearly as important as sponsored movement. This is discussed in Chapter IV, paras. 4.04 to 4.06, with evidence from the 1980 census. Overall the ratio of rural spontaneous transmigrants to sponsored transmigrants in 1980 is 1.2 to 1 (see Table 4.1 of main report). This figure excludes Lampung and North Sumatra because these provinces received many migrants prior to 1940. The ratio is much higher (5.5 to 1) if Lampung and North Sumatra are included. As many micro-level studies of transmigration communities have shown, almost all of these spontaneous settlers have moved as a result of chain migration. With the exception of the government's sponsored program, virtually no one moves to transmigration areas without friends or relatives in the destination area who can provide them with information about what to expect, a place to stay and other help getting started in the new area. The move would otherwise be unacceptably risky for a poor laborer from the inner islands. - 220 - ANNEX 8 Page 4 In order to mathematically model this chain migration in the demographic projections, a variety of functions were tested relating sponsored transmigration to spontaneous transmigration. A simple model was chosen, whereby a transmigrant moves, and then for 10 years this move induces the move of a fixed proportion of transmigrants per year. These migrants, in turn, pull other migrants to the outer islands. For the demographic projections, three levels of spontaneous migration were used. The first assumes that sponsored migration induces no spontaneous migration, a very unrealistic assumption, but included just to show the population impact of purely sponsored movement. The second assumes that a sponsored migrant induces .05 more transmigrants annually for the ten years following the move. The third assumes that a sponsored migrant induces .075 more transmigrants annually for the ten years following the move. The second and third assumptions fall within the movements of spontaneous migration from sponsored migration to rural outer island areas that we can estimate from the 1980 census. Assumptions about the Level of Sponsored Transmigration Demographic projections with several different levels of sponsored transmigration were made in the course of the exercise. Given current information of land availability, implementation capacity, and financial resources, two scenarios are presented in Chapter IV, Table 4.3. One is high, the maximum level of transmigration which could be achieved under the sponsored program if land availability issues are resolved, financial resources are made available, and implementation capacity in the tree crop subsector is significantly improved. The second scenario, considered to be intermediate, appears at present to be more realistic. The high scenario assumes the sponsored movement of 300,000 families in Repelitas IV and V, falling to 240,000 families in Repelita VI. The intermediate scenario assumes sponsored transmigration levels of 200,000 families in Repelitas IV and V, falling to 100,000 families in Repelita VI. Assumptions about Labor Force Participation Rates The labor force participation rate is assumed simply to be 53% of the population age 15 and over. Summary Table 4.3 in the main report presents the summary figures from the demographic projections. The demographic impact of transmigration is compared with population growth assuming zero migration. In other words, in the without migration scenario, population in the year 2020 is simply a function of natural increase alone (fertility and mortality), all forms of migration are assumed to be zero. Sponsored transmigration at levels of high and intermediate levels is projected to have a significant impact on population growth and labor absorption. This demographic impact includes not just the sponsored transmigrants themselves, but their children and any spontaneous migration that is induced by the sponsored program. - 221 - ANNEX 9 Page 1 Summary TORs For Phase II and III Studies Phase II Studies: Reconnaissance and Master Plan A. General This is a rapid reconnaissance of areas identified for transmigra- tion and is often referred to as the screening process. One objective is to assess the suitability of the area for a particular form of development, so that a decision can be made as to whether it is worthwhile undertaking more detailed surveys. A second objective is to carry out socio-anthropological studies of the local people living in and around the settlement areas to be developed and determine their attitudes towards transmigration. A third objective is to prepare a master plan for developing the sites found to be suitable including a preliminary development plan. It is assumed that aerial photographs at scale 1/50,000 and 1/30,000, rectified semi-controled photo mosaics at scale 1/100,000 and topo- graphic maps at scale 1/50,000 with 25 m contour intervals will be supplied to consultant firms by Bakosurtanal through Bina Program prior to the implementa- tion of studies. B. Summary of Activities The Phase II studies for one settlement unit (SKP) of about 1,500-2,000 families (KK) covering a gross area of about 20,000-25,000 ha are summarized below. The type, density and distribution of field observations are given in Table 1. (a) Preliminary interpretation of aerial photographs at scales 1/50,000 and 1/30,000. (b) Investigation of land units and preparation of a land unit map at a scale of 1/50,000, based on air photo interpretation, field observa- tions of slopes and soils along traverses a maximum of 2 km apart over the whole SKP. (c) Investigation of climate and hydrology based on analysis of existing climate and hydrological data supplemented by additional field and map observations of the drainage network and flooding hazard. (d) Preparation of present land use and forest status maps at a scale of 1/50,000 based on air photo interpretation, field investigations and local enquiry to establish land availability. ANNEX 9 - 222 - Page 2 (e) Preparation of land suitability maps at scale 1/50,000. (f) Social-economic studies to establish the activities of the existing populations, their access to outside markets and/or traders and the type and value of goods traded. (g) Socio-anthropological studies based on direct interviews with local people and government officials to assess the number of inhabitants in the area, the tribal and administrative organization, the age structure, the literacy, the health conditions, the land occupied, the desire to participate in the transmigration program, the method of participation, and the willingness to release land. (h) Preparation of agricultural development proposals taking into account local people's attitude and participation. (i) Preparation of an SKP master plan at a scale of 1:50,000 and modification of the 1/100,000 regional structure plan if necessary. (j) Preparation of a preliminary 10 year development plan for settlement units. (k) Reporting. (1) Formal and on-the-job training of the national associate staff in the preparation of Phase II studies. The composition of the study team and man-month requirements to carry out a package of 24 Phase II studies per consultant contracts is given in Table 3. Phase III Studies: Settlement Structure Plan and Economic Feasibility Phase III studies will be undertaken only for those SKP which, following the reconnaissance Phase II studies, are found to be suitable for further development and which have been approved by the Project Working Group and cleared by the Provincial Authorities. The planning and surveys done in Phase III studies will follow the type of development proposed in Phase II studies. In most sites to be developed, standard TOR will apply but in areas with particular problems or when new development models are considered, special TOR will be proposed by the consultant and approved by the Bank and Bina Program. The objectives of Phase III studies are to: (a) Confirm that an area is technically suitable for transmigration. (b) Prepare physical plans and cost estimates for the settlement and in particular define the areas within which land clearing should take place. (c) Ensure that the proposed development and agricultural activities are economically and financially feasible. - 223 - ANNEX 9 - ~~~~~~Page 3 (d) Ensure that the proposed settlement is socially and environmentally viable. (e) Prepare detailed development plans for each settlement unit. It is assumed that Phase II maps and reports have been completed and that recent air photographs of scales 1/50,000 and 1/30,000 and topographic maps at scale 1/50,000 are readily available through Bina Program. The Phase III studies for one SKP covering a gross area of about 10,000-12,000 ha and consisting of 4-5 villages (SP) of 300 to 400 families each are summarized below. The type, density and distribution of field observations are given in Table 2. (a) Preparation of a 1/20,000 scale topographic map with, where necessary, surveyed control based on traverses no more than 3 km apart. Topographic detail is derived from field data anl,aerial photographs or from any suitable available control data.- (b) Reconnaissance slope surveys: overall density of one observation per 5 ha, data obtained by measurement every 50 m along traverses (rentisans) not more than 1 km apart. Data to be presented and slope map prepared at 1/20,000 scale using air photo interpretation. (c) Soil survey at semi-detailed level: overall density of one observa- tion per 25 ha, data obtained from observation every 250 m along rentisan not more than 1 km apart. Soil map prepared at 1/20,000 scale. (d) Present land use survey: overall density of one observation per 5 ha, data obtained from observations every 50 m along rentisan not more than 1 km apart. Present land use map prepared at 1/20,000 scale using air photo interpretation. (e) Forestry investigations to establish volumes of marketable timber with a sampling error of 10% at 95% significance level. Minimum sample of 0.5% by area. Proposals for utilization of timber and forest products in the light of existing laws and regulations. (f) Water resource investigations primarily to establish and prepare designs for potable water supplies for village sites and for drainage and flood protection requirements for the SKP. (g) Preparation of land suitability map at 1/20,000 scale. 1/ This activity will not be required if radar/laser maps at scale 1/10,000 with 5 m contour intervals are available. ANNEX 9 - 224- Page 4 (h) Preparation of agricultural and other development proposals for incoming transmigrants and local people. (i) Preparation of preliminary SP structure plan at 1/20,000 scale showing blocks of land within which more detailed surveys for houselots and arable land should take place and demarking any areas required for the use of local people. (j) Survey of slopes within area of houselots and arable land: one observation per 1 ha; data P tained from observations every 50 m along rentisan 250 m apart.- (k) Access and main village road alignment survey at 1/20,000 scale. (1) Preparation of final SP structure plan at scale 1/20,000 showing the land allocations for incoming transmigrants and local people, the village boundaries subdivided into hamlets of 20-50 families, the blocks for houselots, the first arable land, the second farm land, the reserved lands, and the areas to be cleared. Alignments for access and village roads will also be shown. (m) Cost estimates for settlement development and assessment of the land use rights given up by local people. (n) Study of regional and environmental setting. (o) Prepare an economic vnd financial feasibility study including farm budgets, cost estimates and benefits to settlers. A comparative analysis of costs and benefits of maintaining the area as productive forest under current or proposed management systems. (p) Prepare detailed settlement development plans for the first five years of settlement and indicative development plans for the following five years for each settlement unit. (q) Recommend, in the most satisfactory way possible, the way of parti- cipation of indigenous people in the transmigration development projects and the most equitable compensation arrangements. (r) Reporting. (s) Formal and on-the-job training of the national associate staff in the preparation of Phase III studies. The composition of a study team and the man-months requirements for the implementation of a package of 16 Phase III studies per consultant contract are given in Table 3. 2/ This activity will not be required if radar/laser maps at scale 1/10,000 with 5 meters contour intervals are available. ANNEX 9 - 225 - Table 1 Type, Density and Distribution of Field Observations In Phase II Studies Observations Minimum Type Density Distribution Slope /a - Clinometer 1/10 ha 1 every 50 m along traverse measurements Traverses 2 km apart over whole SKP soils - auger boring 1/200 ha 1 every 1 km along traverse or profile pit Traverses 2 km apart over description whole SKP Present land use - field observation 1/10 ha As for slope measurements - local enquiry variable All village heads. Some farmers Socio- - field observation variable All village and districts heads Anthropology - local enquiry Some villagers. Hydrology - river flow direc- variable All river intersects along tion and flood traverses hazard - potable water variable Existing wells supply - tidal fluctuation variable All major tide-affected rivers /a These observations will not be required if radar/laser ground profiles are available. Note: Traverses may be along cut lines (rentisan) or along existing roads or footpaths where these give representative sections at an appropriate spacing. ANNEX 9 -226- Table 2 Type, Density and Distribution of Field Observations In Phase III Studies Observations Minimum Distribution Type Density Maximum spacing Slope reconnaissance /a - clinometer 1/5 ha 1 every 50 m along traverse measurement Traverses 1 km apart Slope confirmation within land clearing block a - clinometer 1/1.25 ha 1 every 50 m along traverse measurement Traverses 250 m apart Soils - auger borings or 1/25 ha 1 every 250 m along traverse pit descriptions Traverses 1 km apart Present land use - field observation 1/5 ha 1 every 50 m along traverse. Traverses 1 km apart - local enquiry variable All village heads. Minimum of 50 farmers Forest inventory 0.5% by area Within each major forest type Water resources - river flow variable All river intersects along directions and traverses flood hazard - tidal fluctuations variable All tide-affected rivers - potable water supply variable Existing wells Additional bore holes Surface water storage /a These activities will not be required if radar/laser maps at scale 1/10,000 with 5 m contour intervals are available. - 227 - ANNEX 9 Table 3 Settlement Planning Man-Month Requirements per Average Phase II and III Contract (Assuming 24 Phase II studies and 16 Phase III studies carried out over a total contract period of 30 months) Phase II studies Phase III studies Total Specialists Foreign National Foreign National Foreign National Project manager 12 12 18 18 30 30 Team leader 18 18 24 24 42 42 Topographers 18 20 36 42 54 62 Soils specialists 18 20 36 42 54 62 Planners 18 18 12 12 30 30 Sociologists 12 12 12 12 24 24 Agronomists 12 12 18 18 30 30 Tree crops specialists - - 12 12 12 12 Foresters 12 12 12 12 24 24 Agroeconomists 12 12 18 18 30 30 Aerial photographic int. 12 12 12 12 24 24 Cartographers 12 12 12 12 24 24 Civil engineers 12 14 18 24 30 38 Water engineers 12 12 12 12 24 24 Training specialists 12 12 18 18 30 30 Miscellaneous /a 8 16 30 33 38 49 Total 200 214 300 321 500 535 Average per study 8.3 8.9 18.8 20.0 - - Total for 14 packages 2,800 3,000 4,200 4,500 7,000 7,500 /a Includes short-term assignments of the following specialists, among others: mar- keting, storage and processing, rural institutions (cooperatives), agricultural credit, regional planning, hydrogeology, health, nutrition, education and fish- eries. DISTRIBUTORS OF WORLD BANK PUBLICATrIONS ARGENTMNA FRANCE KENYA SOLUTH AFRICA Caros Hirsch, SRL World Bank Publications Africa Book Service (E. A.) Ltd. For single rities: Galeria Guemes 66, avenue d'lena PO. Box 45245 Oxfond University Press Southerm Africa Floinda 165. 4th Floor-Ofc. 4531465 75116 Parts Nairobi P.O. Box 1141 1333 Buenos Aires GERMANY, FEDERAL REPUBLIC OF KOREA, REPUBLIC OF Cape Town 8000 AUSTRALIA. PAPlUA NEW GUINEA, FUI, UNO-Verlag Pan Korea Book Corporation For subscriprion orders: SOLOMON ISLANDS, VANUATU, AND Poppelsdorfer Alle 55 PO. Box 01. Kwangwhamun International Subscription Service WESTERN SAMOA D-5300 Bonn I Seoul PO. Box 41095 Iufo-Line Craighall Info Lix ~~~~GREECE KUWAIT Joaihannsbri 22 Overseas Document Delivery KEME MEMRB Johanxsburg 2024 Boxv 506.2GPO 24, Ippodamou Street P.O. Box 5465 SPAIN Sydney. NSW 2001 Athens-11635 Mundi-Prensa Libts, S.A. AUSTRIA MbALAYSIA Castello 37 Gerold and Co. GUATEMALA University of Malaya Cooperative Bookshop 28001 Madrid A-ll IWen Librerias Piedra Santa Limited A-GlOlmtblen 31Cento Cultural Piedra Santa PO. Box 11 27, Jalan Pantai Baru SRI LANKA AND THE MALDIVES Graben 3t It calle 6-50 zona I Kuala Lumpur Lke House Bookshop BAHRAl!i' Guatemala City OP.O. Box 244 PMRN5Bo 2750 HONG KONG, MfACAU hfEXIC 100, Sir Chittampalam A. Gardiner Masiatha MananP a Town 317 Asia 2000 Ltd. Apartado Postal 22-860 Colom 2 6 Fl., 146 Prince Edwand Road, W, Col. PE/A Pobne SWEDEN BANGLADESH Kowloon 14060 Tlalpan, Mexico D. E For single titles: Micm Industies Development Assistance Society Hong Kong AHCE Frines Kungl. Hovbokhanded MicrDAS lrdttisMOROCCORgensau12Bo 66 (MIDAS) __tHUNGARY Societe d'Etudes Marketing Marocaine Regeringsgatanl2,Boxl6356 G.PO. Box 80Kultura 2 Rue Moliere, ld. d'Anfa S 0 7So" Dhaka PO. Box 139 Casablanca For subsciption orders: BELGiltlh 1389 Budapest 62 THE NETHERLANDS Wennergren-Williams AB Publicatons des Nations Unies PDIA InOr Publikaties Box 30004 Av. du Rol 202 Allied Publishers Private Ltd. Noorderwal 38 S-104 25 Stockholm 1060tBrussels 751 Mount Road 7241 BL Lochem SWITZERLAND BRAZIL Madras-600002 NEWZEALAND LihrtePayot Publicacoes Tecaicas Interacionais Ltda. 15 J.N. Heredia MarB Hills Library and Information Service 6 Rue Gnnus Rns Peuxoto Gomude, 209 Ballard Estate Private Bag Case postal 381 01405 Sao Paulo, SP Bombay-400 038 Now Market CH 1211 Geneva It CANADA 13/14 Asaf Ali Road Auckland TANZANIA Le Diffuseur New Delhi-I 10002 NIGERIA Oxford University Press C.P. 85. 1501 Ampere Street Botucherville Quebec 17 Chittaranjan Avenue University Press Limited PDO. Box 5299 JIB 5E6 Calcutoa-700 072 Three Crowns Building Jericho Dar es Salaam Private Mail Bag 5095 THAILAND COLOSBIA Jayadeva Hostel Building 'badan Central Department Store Eauace Ltda 5th Main Road Gandhinagar TnORWa Carm6No. 51-21 Bangalore-560 009 NOWY306 Silom Road Caerta 6-E Taurro-KarI Johan, A.S. Bangkok Bsogota D.E. 3-5-1129 Kachiguda Cross Road PO. Box 1177 Senuum TRI11DAD & TOBAGO, ANTIGUA, Apartado Aemo 4430 Hyderasbad-SOC027 Oslo I BARBUDA, BARBADOS, DOMINICA Cali Valle Pranthana Flats, 2nd Floor OMAN GRENADA, GUYANA. JAMAICA, COSTA RICA Near Thakore Baag. Navrangpura MEMRB Information Services MONTSERRAT, ST. KITTS AND NEVIS, Librena Tejos Ahmedabad-380 009 PO. Box 1613, Seeb Airport LUCIA, ST. VINCENT & GRENADENES Calle 11-13 Patiala House Muscat Systematics Studies Unit Av. Fernandez Ourll ~~~~~~~~~~~~~~~~~~~~~~~~~~~55 Eastern Main Road Av. Fer-adez Guell 16-A Ashok Marg PAKISTAN Curepe Satn Jose Lucknow-226 001 Mirma Book Agency Trinidad. West Indies COTE D'IVOIRE INDONESIA 65, Shahrah-e-Quaid-e-Azam Enut d'Edition ex de Diffusion Afficainms Pt. Indira Limited P.O. Box No. 729 TURKEY iCEDA) Jl. Sam Ratutlangi 37 Lahore 3 H Hst Kitapevi A.S. 469, Istiklal Caddesi 04 B. P 541 Jakarta Pusat PERU Beyoglu-Istarbul Abidjan 01 Plateau P.O. Box 11 Editorial Desarrollo SA CYPRUS IRELAND Apartado 3824 UGANDA MEMRB Information Services TDC Publishers Lirrta Uganda Bookshop PO. Box 2098 12 North Frederick Street THE PHILIPPINES Bampala Nicosia Dublin I Natonal Book Stoap DENMARK ISRAEL 701 Royal Avenue UNITED ARAB EMIRATES SamfundsLineratur The Jerusalem Post Metro Manila MEMBR Gulf Co. P.O. Box 6097 Rosenoerns Alle 11 The Jerusalem Post Building POLAND ShaBah DK-1970 Fredeiksberg C. PO. Box 81 ORPAN DOMINICAN REPUBLIC Romema Jerusalem 91000 Palac Kultury i Nauki UNITED KINGDONf Editora Taller, C. por A. ITALY 00-901 WARSZAWA Microinfo Ltd. Restaurarion ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~P.O. Box 3 Resciuricio. Licosa Commissionaria Sansoni SPA PORTUGAL Alton. Hampshire GU 34 2PG Apdo. posta 2190 Via Lama,mora 45 Liveria Portugal England Santo Domingo Casella Postale 552 Rua Do Calmo 70-74 EGIPT, ARAB REPUBLIC OF 50121 Flornce 1200 Lisbon VENEZUELA Libreria del Fste Al Abram JAPAN SAUDI ARABIA, QATAR Aptdo. 60.337 Al Gasaa Street Eastern Book Service Janr Book Stom Caracas 1060A Cairo 37-3, Hongo 3-Chome. Bunkyo-ku 13 PO. Box 3196 MsLAND Tokyo Riyadh 11471 YUGOSLAVIA Jugoslovenska Kujiga Akateeminen Kirjakauppa JORDAN SINGAPORE, TAIWAN, BURMA, BRUNEI YU-I 1000 Belgrade Tig Republike PS.-Box128 Jordan Center for Marketing Reasech Information Publications Helsinki 10 PO. Box 3143 Private. Ltd. ZIMEtABWE Jabal 02-06 Ist Fl., Pei-Fu Indstrial Textbook Sales Pvt. Ltd. Amman Bldg., 24 New Industrial Road Box 3799 Singapore Haram IBRD 20722 0~~~~~~~~~~~ IBRD 20724 1BRD020725PIA 40 40~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- - IBI San~~~~~~~~~q A~~~~~~~~~.hB n k ~M ALAY.SEIATN.*ambn If A PRE 197H - ROAS ~Mednd t;000 j\ 0000NATIONAL CAPITALS, PROVINL-CE, CAPITAL IV0;0 00 0 f -. - PROVINCEw A LBUDRE AMPUR YAWA RIVERS ~ ~ ibIg --00000 t ;i t - ITERNAT BOUNDARIES . .. - 4 >GSNA RE i i St 0 - t;0 -15j tS0 3 ; 00 :1 980 . 1020 R0M I A8 U8 it;0 ; ; 500; 0;0 00 00 t;t0 ti:TRNsG4ON7SCHEES 00 00 \0000 0 y 00 0t 0 0000t 00 0000Renqot A009 t ;0000000 000000 t0000000 00 00X;;00i * 00tO0T009,9090000000000000000WSuttg0040000000000000000000000000000000000000Xjdng0000 0 0000 00 l000 000 $0 X 00 - tiOA SX 02g00000000000000;$00 ;000t00000000000000 ;00;0)0' 0000 ;0 t0 $ l00000f000 00 0 i40 ;000 - 0 j:i-XU 0 0tf0000 t; 4 PRo INCElCAPIAtSt)02t 0000000 0 E o iCtt7X;(X00 00000000 ,`B . n kW ;g tt ) 00:a: 00 : 000 00 0 00 i ;;- *- R I;C O D I 0;)0(090000l30aggf0000; ;0 a 00(g Cl ;0000 lf;;g i0ag0000;) ggf^ jf000 0;t 000 ;;y 02 A R;000 30 0 0 0000000t0 000 : li00--[NERNtiNALBON:ARI5 000000 0000000000 00 00000 t000000p j$50 t T Palembang 0 ;;0 ; t 0000 \ 0 00000300000 0 :Xt000Xt00:0000000 t0;ItOE JR bu0000000X02lingg0004;f f? 00 0000 u00 -8 J RAN00000 i0t0;0000 i ; i;0 ii;; 0iittig;; 000ti;;0i0; i0 gte$ |0t0:|0;00000t0i00|;00000000000000gg $00 ;0 :S00 ::f: $S: 0 At: $;| ::: tEf : ( iD 0 a$ 0 :; :: :: 05 $;S :(: SS :E tS 0) ::0 :: :tl 0;; t;( S;X dA) it ); (;: :ff :::2: :V0:u: 70 3 0 ::: ifif :;;: fiC ::; ;:tS Sit ;;S} (C00 fll t;f: i;E )W: :ff ;XlY :;V44 1) A; t0fC:;ft;0:ff :aiS00i: d-di S;020CtE:00t; tS:;4 ;fadlf;:0SSiV.t0MiLE:g ;f;40Ct:tSgbEVSS;Et}k000;fEi;C;0SSi?tSr02 o A . t;0; fida0 005000000 :00 V }:0000;f 00ff0S;;000ff0;2XCVd00 ;00000X;01ii; 00X000ff$t0000000000j00 dE0;g000 d0k';SCAS'S00,$0000000 ASSdgN -00gW0a S;f ;d S : wtA; SSX ;;: udi: f(00fiS Sd' titugSS 0\Sl:;St tSS;S00 AE ;'Si;: S:S;0000yS:SSaiA;'A);i;;Sl;tS SSS:00; iSS ;;0l ;0SS ;0S N003etruh Cu;040 ;00f SEetS iSi'g8 ttStf lE $:00DS(i:CC g0t0 f:(ft 05S:H4:;0:'V;S gS :iEtt;ft; 4580t S gE SSSSTRA 02 SE0000 SDSS A~~ ~AC M1988 IBRD 20723 tpH.lPNES IBRO 20724 IBRO 20725 0: 0 ;SERI BEGAWAN 0~ ~ ~ ~ ~ ~ ~~SIwo IflCBRflI72 . : r | ; 000000000 ; 0l$ 000000 ; 00 j ';0 - 40X0. S00 :: 'AUSTRALIA 100A R!S00z 00 4 ~~Poniaa . aa; .tii, 00t | 1 0 2 0 9 Y * * A % L i M A,NT AN .0 aO t0S;00 - - --0 KALIMAN TAN X KLI>tSiETANt0000000 t000AL dung ~ ~ ~ ~ ~ ~~a Plnkrya adu a B A R A'T J A W. $:~* KAJ NTN ELTA ;S : : 0z:0 0 0 0; 00 000 2 :;40 a : : 7 ----- -- ~~~~~~~~~~~~~~~~~~Sur olya ot YOGYAKA-RT' Bnyoni BA NSATN AABAT TRANSIMIATNSETRRVW TRANSMIGRATIN SCHEMES~ - ROD015 300 450 PRE - 1979 o s AINLCPTL RVNECPTL POST - 1979 RIVERS0 530 --PROVINCE BOUNDARIESMIE --INTERNATIONAL BOUNDARIES 1080 112 LLL ~~~~~~~~~~~~~~~~~~~~~~MARCH 1988 IBRD 20724 1200 ~~~~~~~~~~~~~~~~~~~~26- INDONESIA -0\TRANSMI;G0RATION SECTOR REtVIEW SULAWESI AND MALUKU ISLANDS TRASMIRATONSCHEMES& ROADS/ PRE -< 19790500 z00 0 00000 0$00 o PROVINCE CAPITALS, OTHER TOWNSN POST 1979 ~~~~~~~~RIVERS - '-ROVNCEBOUNDARIES - INTERNAT IONAL BOUNDARIES < 00 000 :;<, A S 0 X .. < 00 0 0 j : 4 ; h * !~~A A U 1K U;: -0~~~~~~~~~~~~~~~0 a