UNN-48 ETURN TO REPORTS DESK WITHIN ONE WEEK AGRICULTURAL SECTOR SURVÉY INDONESIA (in four volumes) VOLUME. IV ANNEXES 11 THROUGH 15 April 10, 1974 This report ias prepared at the request of the Government of Indonesia. The viéws and recommendations contained in the report are those of the authors rather"than those of the World Bank Group. ,The Government of Indonesia and the authors ae respoisible for the releåae and distribution of the report. CURRNCY EQUIVALENTS ,. Us$ 1.00 SRp 425 1 Rupiah =$0.002 1,1000 Rupåah $2 .110 .. . (9 CURNC QIRET S$1.0St. 2 9 NOTE The Agricultural Sector Survey report is baoed on data collected by a Bank mission to Indonesia in March 1972. Some of its major findings and conclusions have been overtaken by events. Therefore, this report is not intended to represent the Bank's current assess-ment of Indonesial's agricultural sector. Nevertheless, much of the technical data presented in the Annexes remains valid and it should be useful to the' Government of Indonesia, IGGI members and to other countries, agencies and institutions interested ini the development of Indonesia. April 10, 1974 ANNEX 11 Page 1 INDONESIA FARM STRUCTURE AND ORGANIZATIONS A. Introduction 1. Farm orgaiizations in Indonesia are shaped by the special conditions which prevail there. These include a few large government or private estates; millions of smallholders whose farms, though small, are not of uniform size; an exceptionally high man/land ratio on Java, and therefore extremely small average farm size, with a'relatively low ratio in the Outer Islands; a closely knit village structure within which most farms operate; an absence of institutions providing distribution and marketing services to small farmers; and a variety of prohibitions against private individuals providing these services. Within this context, the definition of a "farm" or "farmer" is not precise, and therefore the definition of "farm organizations" is vague. 2. Indonesian agricultural policy in the area of farm organizations reflects this diversity and includes: (a) The BIMAS programs, which at times have been organized as mass action efforts giving little or no consideration to the needs or wishes of individual farmers, but now are shifting to programs designed to deal more directly with those individual farmers they can, or are designed, to reach. Until recently, BIMAS programs were concerned with rice farmers in good response areas, but are now branching out to other commodities, mainly maize and other dry season crops. (b) A number of private cooperative ventures such as the German organized and managed Friedrich Ebert-Stiftung venture in North Sumatra, and six Japanese corn cooperative ventures in East Java. (c) A general government effort since 1967 to rejuvenate the cooperative movement. The most promising new programs being the BUUD (Badan Usaha Unit Desa) village unit cooperative experiments, developed as pilot projects near Jogjakarta and now being established on a national basis. (d) A variety of private ventures, such as the Mitsugoro operation in Lampung, designed to stimulate production of crops, especially corn, for export; or the Japanese model farm in South Sulawesi near Pare-Pare experimenting with a variety of possible export crops. ANNEX 11 Page 2 (e) Estate rehabilitation programs designed to reestablish the productivity of these large-scale enterprises through capital investment, modern management and modern production methods. (f) The transmigration program which until now has been essentially concerned with reestablishing in the Outer Islands the system of agriculture prevailing on Java--providing settlers' families with about 2 ha of land, one of which is, or is supposed to be, irrigated. (g) The granting of leases to private individuals or firms for a variety of ventures such as cattle ranches on the Outer Islands. Similar ventures are being established under various aid programs. (h) The special arrangements in the sugar industry, in which the mills are PNP's (state enterprises), but have special lease arrangements with villages so that blocks of land belonging to individuals are set aside for fixed periods to grow sugar. (i) Traditional use rights of indigenous groups particularly in the Outer Islands. 3. Each of these approaches reflects the quite different coniitions and factor proportions which prevail in Indonesia, as well as the per- sistence of production methods developed decades ago; but they also reflect different philosophies about how agricultural production can or should be stimulated in Indonesia. Conditions on the Outer Islands, where much of the land is still essentially unsettled may provide the possibility for a variety of choices. In the settled areas, however, especially Java and Bali, and the densely populated areas of the Outer Islands, there are few choices. In these areas, farm size, land ownership and village structure will be essential factors shaping the nature of farm organizations and their effectiveness. B. Land Tenure and Farm Size 4. The 1963 Census of Agriculture indicated that 70% of all farms in Indonesia had less than 1 ha of land, while 71% of the land belonged to the 30% of the farms with 1 ha or more. Farms of less than 0.1 ha were not counted. Only 2.5% of the farms (those from 5 to 25 ha) had 25% of the land. 5. The average size of farm varies considerably among provinces and islands. The smallest average size farms are found on Java--0.6 to 0.8 ha-- especially in the Jogjakarta and Djakarta areas. On Kalimantan, depending on province, average farm size ranged from 3.1 to 5.3 ha, and in many parts of Sumatra and Sulawesi it averaged 3 ha. ANNEX 11 Page 3 6. There is considerable fragmentation and, although it is difficult to quantify, share cropping and other rental agreements exist. 1/ It is obvious that landholdings are unequally distributed and that a great proportion of the farms are very small. 7. Sizeable differences in land ownership, or control over land are noticeable in most villages. Random observation by mission members indicated a generally similar pattern throughout the country--a few large landholders (5 to 10 ha), a number of medium size holdings (1 to 2 ha), many very smallholdings, and a number of villagers with no direct claims to land. In one village in North Sumatra, the range of landholdings was said by the Lurah (the village headman) to range from 0.2 ha to 3.0 ha. The total land of the village was 150 ha, there were 170 families, 90% of whom owned some land, and 900 people lived in the village. In another village in East Java, the Lurah said that of the 1,200 families in the village, 900 were farmers and 700 owned some land. Land ownership ranged from 10 ha to 0.25 ha. Two hundred farmers were said to own 1 ha or more, while the total wet season area was said to be only 300 ha, and the dry season area 400 ha. 8. Much the same picture, but on a generally larger overall scale was evident in South ySulawesi, and on a considerably smaller scale in Central Java. In the heavily overpopulated Jogjakarta Special District, one hamlet (Miri) has 164 families with 691 people. Only 29.52 ha are available to them. Of the 164 families, 60 own no land, anJ 77 own 020 ha or less, for a total of 9.7 ha, or 33% of the land. The remaining 67% of the land belongs to 27 families, 3 of whom own 27% of the land. But the largest single farm in Miri is less than 3 ha. In such an area, unequal farm size is not as important as the absolute shortage of land. 9. Villagers have a number ot sources of income in addition to land. Some villagers with small landholdings possess draft animals and plow for others, some keep poultry or ducks, some are tradesmen, craftsmen, or work entirely outside the village. Some farmers with small plots rent their land to others while they themselves rent larger plots from still larger landowners. A variety of customs and practices regulate by whom and how harvesting and other activities are done, which may alter the relationship of landholding and income, making it more equitable then it appears from land ownership data. 2/ 1/ The Agro-Economic Survey of 11,000 farms indicated that 33% of the "land was leased or shared in other ways." These were relatively large farms by Indonesian standards, over 1 ha. The same survey also noted that the average number of land parcels for these farms was 2.3, and that "the smaller size of the farmer's sawah, the greater the fragmentation." (Preliminary Analysis of Rice Farmers in 37 Villages, Collier and Sajogyo, 1970). 2/ Customarily harvesting is done by women or poorer villagers who receive a share of the crop. This share diminishes as the relative surplus of labor increases, ranging from one-half to one-ninth of the crop. ANNEX 3 Page 4 C. Present Organization of the Village Economy 10. Nevertheless, relative economic status is closely linked to land ownership or control over land, and a major link in this chain is the village leadership. "... Most of Central Java's village chiefs are large landholders." 1/ Village leaders tend to be owners of relatively large farms themselves or have other sources of income, and village leadership confers access to still more land. A 1969 survey of two Kabupatens in Central Java (Table 1) indicates that 75% of the village headmen owned wet rice land (sawah) and 50% other land. 11. Obviously not all village headmen are large landowners, but the majority become so when they have achieved that position because of: (a) the Bengkok (land-use rights) land which they have the right to work and "to collect the product of that land as long as they retain their position," 2/ and (b) the Tanah Lungguh (village treasury) land. Table 2 indicates that 34% of the income of village headmen and other village officials comes from the land provided by the village for their use. Table 3 indicates that the average size of the land allotments associated with the village headman position was 6 ha in one Kabupaten and 17 in another. Some of this land might not be sawah, the most valuable wet rice land, but even so these are "large" holdings for Central Java. Table 4 shows the size distribution of Bangkok and Tanah Lungguh lands in the villages surveyed. Table 1: VILLAGE HEADMEN SAWAR AND OTHER LAND OWNERSHIP Amount of Hectares Sawah Other Number Percent Number Percent None 94 24.3 196 50.6 0.1-0.5 108 27.9 64 16.5 0.6-1.0 66 17.0 52 13.4 1.1-1.5 24 6.3 8 2.6 1.6-2.0 43 11.7 27 7.0 2.1-3.0 15 4.1 9 2.3 3.1-4.0 12 3.2 5 1.3 4.1-5.0 6 1.6 11 2.8 5.1-10.0 12 3.2 6 1.6 More than 10.0 5 1.5 7 1.9 No response 2 2 1/ Smith, Theodore M., "Village headmen as Agents of Change," The Indonesian BureaucrAcy: Stabilityi Change add Productiviy, unpublished Ph. D. Dissertation, University of California, 1971, p. 59 2/ Although the position of Lurah is elective, most hold the position for life or for more than a decade, and those elected are from a selective list. ANNEX 11 Page 5 Table 2: SOURCES OF INCOME OF THE LURAH AND OTHER VILLAGE OFFICIALS Source of Income Number of Responses Harvest assessment 2 Taxes and administrative fees 7 Bengkok (land-use rights) 34 Pologoro (administrative fees) 17 Land tax fees 5 Salary 3 Subsidy or bonus 12 Bondo desa 2 Source: Based on Survey Struktur Pemerintahan Desa, Semarang, Goenor's Office, 1969, pp. 91-96, (mimeogratphed). The respondents constituted 34 out of Central Java's 35 regents (bupatis and walikotas) 12. This explains why considerable efforts and funds are spent by those attempting to become village headmen. Smith's survey indicates that before becoming headmen about 30% were village officials, 16% farmers, 10% government officials, and 9% each were traders and military men. This nay understate the role of men with military backgrounds especially on the Outer Islands. A report of the Central Java legislature states that: "In practice the candidates selected are usually not people who are competent but people who are rich. This is because the practice is for the villager to select not a candidate based on his ability, but based on the amount of gifts (he gives them). This is even more true in the prosperous villages such that the selection of the Lurah becomes a conflict arena of great activity (jang ramai sekali). The main reason for such conflict stems from tanah bengkok and tenah lungguh that falls to the successful candidate." 1/ It also noted that "the plum of all this land encourages many candidates to go deeply into debt in providing gifts of all kinds to prospective supporters." 2/ 13. Pride and sense of larger village purpose undoubtedly play a role in the desire to become village headmen. Traditions, adat law, and a whole netowrk of factors--especially political-military--influence the actions of village leaders, and stability requires a variety of practices designed to preserve a degree of equity. 3/ Nevertheless, inequalities are 1/ Smith, p. 49. 2/ Smith, p. 29. 3/ There is undoubtedly a close working relationship between Lurahs and the military-governmental leadership at the province level. ANNEX 11 Page 6 evident in-the relationship of smaller to larger landholders--the ownership of small village enterprises such as rice hullers, the marketing and credit system at the village level, and the prices different groups receive for their produce. Table 3: BENGKOK PRIVILEGES IN TWO KABUPATENS OF CENTRAL JAVA Kabupaten Karanganjar Kabupaten Demak Size of Size of Village Bengkok Village Bengkok Official in Hectares Official in Hectares Lurah 6 Lurah 17 Kamituwa 3 Kamituwa 4 Tjarik (Clerk) 3 Tjarik 12 Kebajan (Messenger) 2 Kebajan 3' Djogotirto 1 U1u2 1.5 Pamong Tani 1 --- --- Modin (Summoner to prayer) 1 Modin 3 Dj ogoboj o 1 Kepentengan 3 --- Bekel 4 Source: Based on Survey Struktur Pemerintahan Desa, Semarang, Governor's Office, 1969, pp. 91-96, (mimeographed). The respondents constituted 34 out of Central Java's 35 regents (bupatis and walikotas). Table 4: LURAH BENGKOK AND VILLAGE TANAR LUNGGUR HOLDINGS IN CENTRAL JAVA (in hectares) Amount in Hectares Lurah Bengkok Rights Village Tanah Lungguh No. % No. % None 9 (6.1) 2 (1.4) 0.1 to 0.5 1 (0.7) 5 (3.4) 0.6 to 1.0 0 (0,0) 6 (4.1) 1.1 to 1.5 2 (1.4) 6 (4.1) 1.6 to 2.0 6 (4.1) 8 (5.5) 2.1 to 3.0 9 (6.1) 12 (8.3) 3.1 to 4.0 35 (23.7) 3 (2.1) 4.1 to 5.0 23 (15.8) 6 (4.1) 5.1 to 10.0 50 (34.4) 25 (17.0) 10.1 to 20.0 12 (8.3) 38 (26.6) 20.1 to 30.0 -- -- 13 (9.1) 30.1 to 40.0 -- -- 11 (7.7) 40.1 to 50.0 -- -- 7 (4.9) More than 50.0 -- -- 6 (4.2) Source: Primary data from field st;rvey work, 1970. ANNEX 11 Page 7 D. The Influence of Farm Size and Ownership on Production, Marketing and Distribution 14. Production credit is a critical determinant of the adoption of new production methods, even if the new method is simply a better variety of seed and fertilizer. The adoption of new high-yielding varieties, fertilizer and plant protection agents require an increased cash outlay, but the landless or small landholding farmers already must often borrow in advance of the crop. In one village visited by the mission in East Java, the price received by farmers for a kg of corn varied by as much as 100% (from less than 10 to 20 Rp). The lower price was received by those farmers who borrowed ahead of the crop at high interest rates, and sold their crop in advance of the harvest at a low price. Farmers who sold at harvest time received 14 to 15 Rp/kg, and those who could hold their crop for later sale or sold directly in the retail market received 20 Rp. How this affects the prices received and the cost of credit is dealt with in detail in a survey of corn marketing in East Java, where the price range under the idjon system is given as 5 to 7 Rp/kg. 1/ 15. The gap between prices received and prices paid varies among the different crops. Theoretically the government supported floor price for rice provides minimum protection to all rice farmers where the program is actually operating, and the government subsidies on credit and fertilizer are supposed to be equally available to large and small farmers alike in program areas. For livestock products such as eggs and poultry, and other crops which individuals are able to market directly at retail, the prices received are presumably the same for all sellers. These are of necessity, however, rather small quantities. This gap seems largest in the production and sale of crops which require credit to produce and depend on markets outside the village itself. 16. Larger farmers are not as likely to resort to borrowing, and are themselves sources of credit to and buyers of the produce of smaller farmers. They are also better able to withhold their crops from market until prices are better. This double edged advantage--on the cost side and on the price side--reinforces the advantage of larger farmers in adopting new technology. Also a new practice on one of the parcels of land belonging to a larger farmer does not compromise him if it proves to be a failure, whereas such a failure could seriously affect a very small farmer. 17. Whether it is theoretically possible for small farms to adopt modern production techniques, there is considerable doubt that farms of much less than 1 ha can in practice be an efficient economic unit, without adequate credit input distribution and marketing institutions to assist them. What is more to the point, can such farms adopt new technology fast enough to compete with larger farms or other agricultural enterprises? Japan and Taiwan developed with small scale agriculture, but both had a much longer history of infrastructure development and a far more advanced system of governmentally 1/ Corn Marketing in Banjuwangi Regency, Malang, 1971, p. 17. ANNEX 11 Page 8 supported institutions--for credit, extension and marketing--designed for small farms. These do not exist in Indonesia at present. Furthermore, both Japan and Taiwan experienced land reforms. There is also considerable doubt that the very small farms in Japan are economically viable today, even with the extremely high Japanese subsidy for rice. 18. Although small farm size may not present a significant production problem in rice, the traditional crop most of which is not marketed, the technical problems of growing high-yielding cotton, corn, soybeans and a variety of other crops on very small farms, when the object is for sale outside the village, may be different. The real cost of developing a marketing and distribution system capable of serving these farms would be substantial. When these are added to the credit and price barriers facing smaller farmers, there is at least a strong doubt that they could compete, either against larger farms or with the outside world. 19. Technological and organizational developments in such agricultural industries as broiler and livestock production have developed to such an extent that small scale village production does not compete in feed con- version, quality of product, continuity of supply, or price. Developments in the Philippines in the broiler industry, and in egg production around Indonesian cities (see Livestock Section), provide an example of how rapidly such enterprises can develop. Often it is not large farmers who enter these activities, but those with substantial control over capital and contacts with the enterprises which possess the technology and the supply and marketing connections. These activities appear to have been beyond the capabilities of even sizeable Indonesian farms with current leveln of agricultural services. E. Existing Farm Organizations 20. Average farm size in Indonesia dictates that many elements of what might be considered "on-farm" activities--storage, drying, initial processing, etc.--be presently carried on at one step removed from most farms. Larger farms may possess all or some of those facilities and perform the services for smaller farmers, but many of them are performed by merchants and traders. Many of the credit and marketing services are also performed by these same individuals. 1/ Although a great many people, both farmers and manufacturers, are involved in the marketing and distribution system 2/, the relatively slow progress in agriculture, distrust of "middle men", a desire to improve the efficiency of production, marketing, and distribution, and a desire to capture more of the benefits of these activities for farmers as such, has resulted in a number of different experiments in farm organizations. 1/ Credit is usually provided by larger farmers, those renting land. or merchants. Much of the initial marketing is performed by small peasants or those with no land at all, while later stages of processing and marketing are done by larger farmers and merchants. 2/ Geertz, Clifford, Peasant Markets in Java. ANNEX 11 Page 9 Cooperatives 20. The cooperative movement in Indonesia has had a checkered history. "Mismanagement and politicalization" were especially heavy burdens in the late fifties and early sixties. Inflation and then the monetary reform in the mid-sixties added to the burden. With "The Law on the Basic Regulations for Cooperatives, No. 12, 1967," the government commenced to weed out unsatisfactory cooperatives and to establish new, stronger ones. Between the end of 1967 and the end of 1968, the number of cooperatives dropped. from 64,096 to 14,745. They increased to 17,225 by 1971. In essence, the effort has been to combine some, start new ones, and eliminate others. Between 1967 and 1970, savings in co-ops rose from 207,231,090 to 1,753,043,665 savings shares. Their turnover reached Rp 39 billion. 1/ 22. Since 1970, the co-ops have been utilized in the BIMAS program. The "ultimate goal (is) to enable the cooperatives to take over the basic functions of the unit Desa (village unit), particularly marketing (including processing) and the supply of farmer's inputs." The Jogjakarta Special District has been chosen as a special pilot project area for the development of the BUUD (Badan Usaha Unit Desa) or Village Unit Program. 23. There is some recognition of the need for capital and good manage- ment in published statements of the government and the Directorate of Cooperatives. Much of the government's effort, however, remains in the area of education and propaganda. The total budget for cooperatices in 1971/72 was Rp 416.5 million, about US$1 million, and 72% of this was for "educational" purposes. 2/ 24. Two important elements in the government's future plans for cooperatives are relevant to subsequent discussions: a. Within the framework of participation in the developmeat plan, cooperatives are expected to contribute not only to increasing the national product and consequently the national income, but also simultaneously to make the distribution of income more equitable. b. There are three objectives which are considered most strategic for the cooperative role at the present stage of development, 1. to increase the members' income; 2. to participate in creating employment opportunities and, 3. to support the supply of foreign currency. 1/ Cooperatives in Indonesia, unpublished document of the Directorate of Cooperatives, Djakarta, 1972, p. 4. 2/ Ibid, p. 8 ANNEX 11 Page 10 2.4. In viewing the obstacles to effective cooperative development, the Directorate of Cooperative demonstrates some perceptive insights. While recognizing the weakness in management, financing, and members' parti- pation, it notes that "when we examine further, these problems originate from the weakness of the social economic structure of our community itself." On the economic side, it is recognized that to be successful there must be economic gains to each member, and that "the big problem confronting us in the development of cooperatives among the poor people, 'is their low econo- mic standard'." 1/ BUUD and Other Cooperative Type Experience 25. Management. In the BUUD pilot projects good managers have been hard to get--"the good men go out of the village." The government has established the BUUD's and regulates the method of payment. The president of one cooperative visited receives Rp 1,500 per month, and a small commis- sion on the co-op's turnover. But "the volume of business is so small that the present salaries are much below what they should be." An effort is now being made to group cooperatives from several villages into more economically viable units. 26. For these reasons, and because of existing village economic and political realities, government-sponsored co-op management leans heavily on the existing village leadership, and upon those in the village who can afford to devote time to the co-op, or have the ability. A number of the co-op leaders visited in Central and East Java were Lurah's or other sub- stantial farmers. They were also, at times, essentially village merchants, or persons in the village who were already in the distributing and marketing business. 27. The cooperative pilot project in North Sumatra run by the Friedrich Ebert-Stiftung cooperative organization, indicated that the "lack of management skills at the village level is such a handicap that only one primary co-op per Kabupaten was really workable," whereas the present law requires one per village. In attempting to obtain co-op employees or leaders from the Lurah, they have found that those suggested were often close associates of his. They in turn have submitted names of others in the village, but have faced difficulties in "breaking through the village formal and informal relationships". 28. An analysis of corn cooperatives in Banjuwangi Regency in East Java concluded that important causes of co-op failures were: a. the lack of skill of managers and members; b. the double function of the manager; and c. the lack of integrity on the part of the managers. 2/ 1/ Ibid., 2/ Corn Marketing in Banjuwangi, Regency, Malang, 1971, p. 29. ANNEX 11 Page 11 29. The "double functions" of managers in this case is the common problem in Indonesia of people having multiple jobs. This, and the low income of managers to start with, raises the possibility of the managers augmenting their income fcom co-op resources. 1/ 30. Participation and Size of Holdings. Although data from which valid generalizations can be drawn is lacking, there seems to be a strong tendency for co-op membership to fall into two categories. Either relatively large farms predominate, and there is some measure of progress, or very small farms predominate and the co-op fails or does not produce significant results without continued outside assistance. In North Sumatra the Ebert- Stiftung co-ops had members whose farms average 0.5 to 1.0 ha which is relatively small for that area. "The larger farmers are reluctant to join the co-ops." In the BUUD co-ops visited in Jogjakarta, where farm size is extremely small, the prevailing average farm size of co-op members appeared to be relatively large. In Kabupaten Bontul for example, most of the members of the co-op had 0.5 ha the Lurah, also a member, had 1.5 ha, and the largest holding was 2.0 ha. These are "large" farms for this area. In one BUUD co-op, 25% of the farmers were in the co-op, and in another 20%. 31. In general, there appears to be a feeling that the very small farmers are so poor that there is little prospect of their making a sizeable contribution to production or deriving significant benefits from the co-op. When larger farms dominate, the smaller farmers seem to be fearful that their interests will not really be served. When smaller farms dominate, the larger farmers see few advantages to co-op membership. 32. Services Provided. At present, the main thrust of agricultural cooperatives in Indonesia, on the economic side, is the cooperative provision of some production inputs--seeds, fertilizer, credit and extension services, and in some cases, marketing. The BUUD operations involve all of these, but the cooperative organization as such provides only a limited amount of storage and initial processing, such as rice hullers. Credit is provided by Bank Rakjat Indonesia (BRI), extension by the Ministry of Agriculture, and fertilizer by P. N. Pertani, Pusri, or local retailers. In BUUD co-ops, the fertilizer and credit are provided at the government subsidized fixed rate. The current effort is to develop a BUUD with ancillary services for every 4 - 6 villages. 2/ 33. The East Java Corn Project (Japanese) and the Ebert-Stiftung pro- ject in North Sumalra (German) both involve foreign consultants, the 1/ In the Makmur co-op in East Java the management group used the proceeds of a co-op loan for their own purposes. Since the loan could not then be repaid, the co-op was unable to obtain a second needed loan. 2/ Subsidized fertilizer is also provided by other co-op ventures, including the German and Japanese ones mentioned below. ANNEX 11 Page 12 provision of tractors, trucks and some basic processing machinery--corn dryers, peelers, rice hullers, etc. The Ebert-Stiftung co-op in North Sumatra has stopped providing credit "because the 'adat' relationships in the village do not support the co-op and repayments is only by force. " It does supply fertilizer, chemicals, extension and tractor services. According to the co-op manager, the operation is making a profit. It plans to under- take marketing activities in corn and vegetables, but at present "the farmers offer the co-op the worst quality of their output." The "farmers have (credit and marketing) relations with the merchants which can not be cut at once." 1/ 34. For the most part, therefore, the existing cooperatives are essentially production co-ops with some limited credit and marketing features. They do not alter significantly the internal relationships in the village, and in most cases they are only marginally competitive with the private sector. Given their limited numbers, and the difficulties outlined above, the cooperatives do not at present represent a significant alternative farm organization to the existing small unit farms in their present village structure. 35. There sp-ems little likelihood that much success can be achieved in the stated government objective of raising the level of income of the poorest peasant unless the participation of all or most of the farmers in existing villages is obtained. The resources of the poorer farmers are too limited, they are customarily obligated to larger farmers or merchants, and they are obviously dependent upon the village leadership. Unless the co-ops are able to provide a dependable, clearly superior range of economic services to all members propaganda, education and admonitions to work cooperatively are not likely to produce noticeable results. Some method must be found to either integrate the existing village leadership into cooperatives in such a way that their contribution is seen to be clearly economically beneficial by all members, or co-op leadership will have to be developed through a combination of village and outside staff which is able to compete economically with the existing village leadership and merchants. Other Farm Organizations 36. In addition to smallholders, a wide variety of other types of farms and farm organizations operate in Indonesia--private and government (PNP) estates, and other private farms. Customarily these are viewed as quite separate production units with separate distribution and marketing linkages. These large-scale production units contrast sharply with the prevailing very small-scale farm units, are often removed from direct connection with the village structure, and yet there are relevant ltnkages. 1/ The Japanese co-ops and the corn BIMAS projects in East Java provides "package" of seed and fertilizer which is repaid in kind with 450 to 500 kg of corn per hectare at the end of the crop. ANNEX 11 Page 13 37. Long-term leases are granted to a variety of enterprises to undertake specialized agricultural production in Indonesia. Many of these, like most of the estates, are selfcontained activities with little impact on smallholders, but some act as catalysts to surrounding smallholders. An example is the Mitsugoro (Japanese) project in Lampung Province in South Sumatra. The object of the project is to increase corn production primarily for export through its own mechanized farm operation and purchasing the output of surrounding farmers. As presently conceived the project consists of four blocks of land in different locations in the province, ranging in size from 100 ha to 3,000 ha. The strong export focus of the enterprise has resulted not only in a concentrated effort to introduce higher yielding corn varieties and more productive methods of land preparation and cultivation, but to develop sufficient infrastructure to move the crops from the farm to the export market--including highway improvement, storage, drying and handling equipment, improving the port of Pandjang, and bulking cargoes for lower cost shipment. The project has the added advantage of providing the necessary capital and management ability, as well as knowledge of the major export market, all of which are scarce in Indonesia. 38. The Mitsugoro farm is not designed to be the major source of corn production or of corn exports in the Lampung region. By 1975 projected corn production in the province is 200,000 tons of which the Mitsugoro farms are supposed to supply 50,000 tons for export, while 150,000 tons for export is expected to come from surrounding farms. 1/ 39. Observations by the mission indicate that such activities have a significant impact on the surrounding farmers. Production practices used on the "pilot" farms--clearing of alang-alang grass, fertilizer, etc.,-- are adopted at least in part by the surrounding smallholders. Aside from the substantial employment provided by the farm itself (which provides funds for cash inputs by farmers on their own plots), infrastructure and storage and handling improvements benefit smallholders through improving the market for their corn production, and generally improving both the distribution and marketing system. Because the object is to produce an exportable commodity, not only on the farm itself but in the surrounding area, the management of the enterprise has included in its program facilities for handling both kinds of output. A similar impact is noted in the "cooperative" pilot projects in East Java, which have involved some similar elements, especially the strong export focus. 40. On the surface this particular approach appears to be in conflict with the prevailing factor proportions and farm size in Indonesia--they involve large blocks of land and production and handling methods are essentially capital intensive. Whether there is ultimately a conflict, however, depends on whether the project proves to be an effective catalyst 1/ Maize Handling Project in Pandjang Lampung Province, Indonesia March 1971. ANNEX 11 Page 14 in stimulating increased corn production and marketing among the smallholders, or whether similar farms end up replacing smallholder production. The recent interest of the Government in attracting foreign enterprises to develop rice estates is another example of this type of approach. 41. Another "catalyst" private farm operation is illustrated by the Westenberg farm at Kebun Djeruk near Tebing Tinggi in North Sumatra. This private farm of about 50 ha has for the past 20 years operated on a profit from the sale of its products, (planting material, seeds, rubber, etc.). It combines many attributes of a private farm, and research station and extension center in one unit. Much of its activities are directed at developing crops such as corn, sorghum, rice and dwarf coconuts, and other farm activities, such as fish ponds and intercropping for livestock grazing, which are suitable to smallholder operations. Observations in surrounding villages indicated that villagers were aware of activities at the farm and often obtained new seeds and planting materials from the farm as well as production ideas and advice. 42. The Westenberg farm has recently been transformed into a foundation with headquarters at Kebun Djeruk and hopes to develop similar farms in other locations on the same basis of self-sustaining enterprises designed to develop and disseminate more productive smallholder profuction practices. The existing management of the farm considers the availability of a block of land and assets sufficient to maintain demonstration plots and "adequate for the production, processing and storage of improved planting material, seed, fish, etc.," a necessity if similar farms are to be self-sufficient and perform a similar func- tion to that of Kebun Djeruk. 43. Although estates are separate farm organizations with direct linkages to the markets for their output and separate relationships to the system of dis- tributing inputs, it is noteworthy they they too can have a beneficial impact on surrounding smallholders, "In certain parts of North Sumatra and West Java farmers have been influenced by the proximity of large rubber plantations, but in Atjeh, Riau, Djambi, West Sumatra, South Sumatra, and in most of Lampung there are very few plantations and therefore no influence on the farmers. Use of improved planting materials has occurred in areas near the plantations while in the other areas almost none of the better clones and planting materials exist." 1/ 44. The operation of the sugar industry in Indonesia represents still another type of farm organization, but with direct and explicit links to the village unit structure. "Land is owned by the villages in small units. The size of these varies, but almost all are between 1/3 and 1-1/2 hectares. The plots are leased to sugar factories in such a way that several contiguous leased units are farmed in sugar cane as one integrated field, 1/ Collier & Werdja, Improving Smallholder Rubber Djakarta, May 1972, pp. 3-6. ANNEX 11 Page 15 which varies in size from about 4 to 100 hectares." 1/ Although many abuses have been attributed to the sugar lease arrangements--from the point of view of both the mills and the farmers, they indicate a long tradition in Indonesia of combining small blocks of land into larger ones, without altering the ownership arrangements, if production methods call for it. This tradition is strongest in some of the most densely settled, traditional village areas. F. A Strategy for Developing Farm Organizations 45. Because most agricultural production comes from small farms in Indonesia, and in settled areas land and capital are scarce while labor is abundant, there are strong economic, social and political grounds for developing better farm organizations to stimulate production and income increases on these farms. Scale Advantages 46. The existence and perpetuation of estates, the granting of leases for specialized export production, and attempts to develop cooperatives all suggest that there is a presumption in Indonesia that economic advantages accrue to larger farm units than to smaller ones, or that larger units possess an organizational advantage over combinations of smallholders. 47. Whether these advantages exist in fact, has not been clearly demonstrated. It is argued that a half hectare of smaller farm can be as efficient as a larger one. Yields on small units are often higher than on larger farms. This might be true for rice production and a limited number of other traditional crops, when the purpose of production is essentially to provide for the needs of a family but there is room to doubt it if the object is to greatly accelerate the production of crops of a more specialized sort essentially for sale outside the village. Even if there are no apparent advantages at the production level as such--if yield levels per unit of land are about the same--the above suggests that there may still be sizeable organizational and marketing and distribution advantages particularly if rapid changes in technology are to be introduced over a significant area. This is particularly true in terms of obtaining efficient management and access to technology and funding. 48. For Indonesia, it is not simply a question of farm size, but of the organizational structure in which small farms operate. Clearly many operations which would be considered "on-farm" in other countries are not appropriate to many of Indonesia's small farms. With limited credit, extension capability, and infrastructure development, practical considerations suggest that the smaller and more remote farmers will be less likely to gain from these limited services than others. 1/ Indonesia Sugar Study, Vol. II, p. 8. ANNEX 11 Page 16 49. In addition, the vast majority of smallholders are concerned primarily with production for their own consumption. On their own farms, and as a group, they lack the economic leverage of a larger single unit enterprise. Present cooperative organizations are able to make only a marginal contribution to providing this organizational structure. 50. There is no doubt that emphasis should be placed on extending as rapidly as possibly the distribution and marketing services needed to the existing farm units. Since "on-farm" services such as initial processing, grading, packaging and storage are needed, but are not likely to be appropriate to most farms, such services can continue to be provided by the private sector or by cooperatives whichever is the most efficient. Extension services can probably be best provided by the government, although at present they are quite inadequate and.are likely to remain so for a long time to come. A system of rural banks, along the lines of BRI or private banks should be extended as rapidly as possible. Input supplies--seeds, fertilizer, etc.,-- should also be largely handled by the private sector or cooperatives. 51. However, in view of the weaknesses of government services for extension and credit and the time it would take to build up effective institutions, such efforts should be augmented by programs to bring in resources - both financial, managerial and technical to accelerate the move from traditional subsistence oriented, smallholder production to a modern market oriented production. 52. Such a strategy could employ government, private or cooperative methods, as are now being experimented with, and it could involve elements of improved equity in the distribution of the benefits of production and marketing gains or it need not. Examples of alternative approaches are the following: 53. In the proposed development of cotton production in East Java and Lombok, contracts between villages and ginneries could be drawn up on lines similar to those in the sugar areas. The villages could provide the land and labor, and the ginnery could provide the necessary technical knowledge, insure that seeds and inputs are provided, and serve as a marketing and distribution link for the crop. If some machinery were required this could be provided by the ginnery, the village as a co-op, or by private inAividuals on a custom basis. Provisions could be made for gradual acquisition of common facilities by the producers through some form of profit sharing. What may be true for cotton may very well be true for a variety of other cash crops as well. 54. A certain amount of machinery or equipment--some tractors, some drying, hulling or shelling equipment, and some storage facilities is used in most new co-op or private ventures undertaking the development of expanded output of cash crops. These seem to be needed in both existing farm areas as well as new areas. Also there seems to be a fairly close connection between some organizational unit with a market focus and the general success of new ventures. It is this "missing link" that co-ops are trying to provide and that the Mitsugoro project provides. ANNEX 11 Page 17 55. If projects in new areas are found to function considerably better with such services in the hands of a single organization, then newly developed projects in unsettled areas could combine both smallholder and one or a few larger units in the same development. A newly developed area could be esta- blished around a central operation which combined an appropriate size pro- duction unit of land, sufficient production and marketing facilities, equip- ment, and management ability to survive, if it operated in close conjunction with surrounding smallholders. Smallholder settlements could be established around such a unit availing themselves of the production and marketing services, and providing needed labor and additional production. In this way a newly settled area would be more than a place for people to live and hopefully feed themselves, but an economically productive unit with links to the market which would give it leverage that an unorganized group of small individual farms would not have. The central core unit (nucleus estate) could, again, be private, PNP or cooperative but with provisions for profit sharing and owhership participation by the smallholders or laborers. 56. Since the vast majority of the output of most crops comes from smallholders, the PNP's should not be viewed as separate units to be developed on their own, or at the expense of smallholders. The existing PNP's in some areas, and those which may be newly established should operate--or experiments should be carried out to see if they can operate-- as the focal point of smallholder activities in the same crop in their area. They could operate in the same fashion as that outlined above providing centralized marketing and production services as well as training to smallholders. 57. The largest potential demand for eggs, poultry and meat is on Java. At present production of these commodities is carried out in Javanese villages along traditional lines. Much of the egg market in larger cities is, or is in the process of being, captured by large-scale operations. Broiler production in large units is not yet a significant factor, as it is in the Philippines, but the factors which have given rise to its development there and elsewhere can be expected to emerge in Indonesia with rapid increases in urbanization and income growth. Whether to be efficient such operations must be large, modern and capital intensive, or whether smaller scale poultry and other livestock activities can compete should be tested. In either case, efforts should be made to develop organizational structures in the village, or a combination of villages, to assist in developing such production methods, and in linking domestic food supplies to those production units. 1/ 58. The object of the above strategy is to provide a method of allocating the scarce resources of infrastructure and management in such a way that smallholders benefit as well as individuals or organizations with larger resources. It also provides a method of combining land, labor and 1/ In the Philippines the broiler industry has developed largely on the basis of imported feeds despite the demonstrated potential for greater production of corn and sorghum on small farms. ANNEX 11 Page 18 capital in ways which are not possible now. Presently the choices appear to be between large operations or very small ones, with the latter concentrated especially heavily on Java, Bali and a few heavily populated areas. The Village Structure and Equity 59. Both the efficiency and equity of village or slightly larger units, centers on the ability and degree of equity which exists in the village units, and the amount of support the government gives to its policy of developing greater equity in the distribution of the benefits of agricultural production. Since the village headmen and Kabupaten leaders are economically and politically powerful, and since the Lurah in most cases has a strong personal stake in the success of agricultural production, it is logical that his cooperation in any village program is needed. Whether these officials take advantage of new technology and services for their own gain, or for the gains they can produce for the entire village depends not only on them, but on the strength of other institutions centered at the village or slightly larger level--banking, extensions services, input suppliers, and those providing storage and marketing services. Indonesian villages are said to combine many elements of equity, and much stress is placed on the cooperative labor effort of "Gotong Rojong." It should be possible to capitalize on these elements in the village if they in fact exist, but these cooperative efforts are not likely to be significant unless the benefits associated with them are significant. 60. There is inequality of land ownership in Indonesia and this does produce inequality in the distribution of income in agriculture. Land reform has not been given serious consideration in Indonesia in the first plan. The second plan with its emphasis on better income distribution should give serious consideration to the following issues: (a) Unless a great many, probably millions, of farmers were moved out of Java, the average size of farm would be very small, probably considerably less than 1/2 hectare, because of the absolute shortage of land on Java leaving most farmers dependent on supplemental off farm employment. (b) The infrastructure and institutions to provide adequately for the distribution and marketing needs of millions of uniformly small farms does not now exist, and is not likely to be developed rapidly. (c) Since conditions prevailing in the Outer Islands are quite different from Java. developments should aim at larger farm units and farming systems suitable for the ecological conditions. 61. The farm organization strategy indicated above could involve elements of greater equity in income distribution, and it would have the advantage of generating greater income from which this distribution could take place. Elements of cooperative ownership, either participating or non-participating, could be associated with any core unit of production - ANNEX II Page 19 PNP, Pilot farm, or cooperative. If separate blocks of land were to be set aside for specialized production of cotton, corn, soybeans or other crops, as is now the case for sugar, the distribution of the income derived from that element of production in the village could be shared, partly on the basis of land contribution and partly on the basis of village membership. The land contribution of villagers could be progressive, with much of the land presently falling to the Lurah and village leaders going into the specialized operation. The income presently received by village leaders from the land, could be replaced by income from the village production operation. Tenure reform could be incorporated in such a program, and a variety of social services could be keyed to village or larger units. 0. Recommendations 62. The Government believes that the present cooperative organizations, including the BUUD's should have a much stronger economic focus. To be effective they must provide better services to their members. Government has initiated efforts aimed at improving the storage and initial processing services they provide.to enable them to buy, store and market the output of all members. 1/ At present improving the credit, input, and extension services should be done through the organizations which already provide them, BRI, private firms, and the Department of Agriculture. Whether general purpose co-ops should be developed, should be determined after these improvements. Better management is especially crucial. Each cooperative should have at least one full time manager whose sole or major source of income comes from the cooperative. Greater government and foreign assistance should be provided to cooperative experiments, including salary payments if necessary, for at least a trial period. 63. Some well funded and well organized village unit production and marketing experiments should be carried out to determine: (a) Are there combinations of land use arrangements in village or larger size units which provide clear production and marketing advantages? (b) What methods can be developed within different villages to bring about the formation of these optimal size units, and what is the impact on employment and income distribution of such arrangements? (c) Which organizational form is the most effective in producing these results--private or government firms, cooperatives, or government agencies? 1/ The present BUUD co-ops are not able to handle more than a fraction of the output of the village, and they sell to BULOG, whereas farmers can often get better prices from merchants. ANNEX 11 Page 20 (d) Can organizational forms be developed which can combine small and larger size farms in the same organizations, and what are the production and income effects? (e) Which organizational structure or combination of structures produces the best incentive to develop unsettled areas? 64. Specifically, at least one BUUD co-op, one private venture such as the Mitsugoro project, one of the cooperative projects in East Java, and at least one new transmigration project should be given additional support and its progress carefully monitored and studied. The costs of providing adequate management and services to each operation, and the benefits from production and marketing should be closely studied. At least one new cash crop experimental production unit in cotton, corn, or another cash crop should be established under each type of nanagement--co-op, private and PNP employing the system of setting aside special blocks of land such as now exist in the sugar industry. The production, marketing and income gains or losses of these ventures should be studied in relationship to progress in villages where these systems are not tried. 65. A number of these experiments should be carried out with the same amount of planning, funding and foreign economic and technical assistance as is given to other economic activities. This assistance and funding should be provided for a long enough period to demonstrate clearly the results--5 to 10 years. The question of the economic viability of the millions of small Indonesian farms in their existing village complex is not academic. How these villages and farms can be most productively developed, and what devices can maximize the development of the Outer Islands of Indonesia, are among the most important immediate and long-range problems facing Indonesian agriculture. Intelligent, factual alternatives, and the costs and consequences of each, should be known. 66. While these experiments are underway, every possible effort should be exerted to permit the private sector to extend its services in the area of credit, distribution, storage, handling and marketing, since these services are needed and are not appropriate at the level of many small farms. In addition support should be provided for at least minimal smallholder programs which provide better planting materials, extension services, and initial processing at locations as close to the smallholder producing areas as possible. Serious consideration should be given to supporting the type of self-sustaining demonstration farm activity already proven successful at Kebun Djeruk in North Sumatra. 67. The results of these experiments should provide a reasoned basis for developing a long-range farm organization strategy for Indonesia. Those experiments which proved unsuccessful, or produced unacceptable social or economic consequences, even when well funded and managed, could be abandoned, and future government policy could direct its attention to those which proved successful. ANNEX 12 Page 1 INDONESIA AGRICULTURAL SECTOR SURVEY SUPPLY AND DEMAND FRAMEWORK A. Introduction 1. Many elements necessary for the development of a satisfactory supply and demand framework for Indonesia's agriculture remain obscure despite much recent study. Such critical variables as the current level and growth rate of population and the actual level of rice yields, and pro- duction, are in doubt. Many other serious statistical inadequacies make a reliable analysis of past developments for 'major crops difficult. Pro- jections, therefore involve qualitative judgments based on many assumptions. 2. During 1971-72, major studies of Indonesia's Rice, Fertilizer, and Sugar industries were completed; and a major evaluation of the various Rice Production Intensification programs was carried out. The first and last of these studies analyze the current and past situation, and project the future, using relatively low current population and relatively slow rate of population growth. The other two studies, as well as most other project- ions studies dealing with Indonesia, assume a larger 1970 population which is projected to grow much more rapidly. This alone accounts for the quite different levels of current and future consumption,.production and trade of major crops found in different studies. 3. In 1970, a sizable upward adjustment was made in the rice pro- duction figure--probably the most critical statistic in agriculture-- without adjusting the series backward 1/. Preliminary results of the 1971 population Census, on the other hand, points to a 1971 total population figure of some 119 million instead of the 124 million generally assumed. This roughly 5 percent increase in rice production, coupled with the approxi- mately 5 million person reduction in population estimates, produced a sizable "statistical improvement" in the presumed level of rice production and an even more significant "improvement" in per capita rice consumption. 4, Coming at the mid-point of the current Five-Year Plan, with its strong emphasis on increased rice production and consumption aimed at domestic self-sufficiency, these alterations would appear to reduce substan- tially the apparent need for additional rice production, and raise the level of success achieved in recent rice production programs. This may account 1/ Government is currently undertaking a new revision of the series. ANNEX 12 Page 2 in part for announcement in March 1972 that the government was reducing the rice production target. If the rice yield and production series were adjusted backward, the yield and production pattern for earlier years would change significantly, and imply a substantially lower rate of growth in recent rice output than generally has been assumed. Subsequent events during August-November 1972 appear to support the view. Domestic rice prices rose sharply during this period in line with the general sharp increase in world rice prices as a result of generally short grain crops throughout Asia and the Soviet Union. The domestic Indonesian picture was also affected by the relatively low level of government-held stocks, a prolonged drought which affected dry season grain production and a growing belief that the 1971/72 rice crop may have been overestimated. Government has managed to accelerate rice imports and government releases of rice into the major urban markets in an effort to dampen price increases. However, rice prices for the 1972/73 crop reflected the higher prices of cereals world wide. 5. In addition to studies narrowly focused on one or a few crops in Indonesia, Indonesia's agriculture has recently received substantial study by organizations concerned with world production, consumption and trade in agricultural commodities. FAO recently published its detailed projections of production, consumption and trade of crops and livestock products for Indonesia for the decade 1970-1980, and the University of Indonesia is carrying out a study of the long-term supply, demand and trade prospects for selected agricultural commodities in Indonesia to 1975, 1980 and 1985. Although much the same ground is covered in each, there are sizable differences in currently assumed and projected results. 6. These studies lend to the study of Indonesian agriculture and its future prospects a degree of "numerical specificity" which can easily obscure the shaky statistical foundation upon which they all rest. More- over, they undertake to project the future--usually the decade 1970-80-- on the basis of a period of profound economic adjustment in Indonesia (1967-71). The full effect of that adjustment is not clear, and the earlier period (before 1967) was one of little significant agricultural growth. There remains considerable disagreement at this point over how much weight should be given to trends evident since 1967, and how much should be given to trends derived from a longer period stretching into the past. 7. It should also be noted that none of the recent studies found it possible to probe deeply into, or develop new analyses about, many important variables affecting demand and supply response. Their conflicting results are, therefore, often due to different assumptions based on, or conclusions drawn from essentially the same material, inadequate as that might be. 8. This Annex does not seek to add still another set of conflicting numbers. The object is rather to bring together the important assumptions or conclusions of the existing studies, compare them, and to use them as appropriate in the evaluation of the agricultural sector mission. ANNEX 12 Page 3 9. Section B of this Annex presents a rough quantitative assess- ment of recent levels of output and projections for 1975 and 1980. Sections C and D discuss briefly the issues of population and income growth. Section E presents a brief discussion of developments by crop, which can be con- sidered in conjunction with Section B. Greater detail by major groups of commodities -- rice, field crops, tree crops, livestock and fisheries and forestry -- is given in Annexes 4, 5, 6, 13, 14 and 15. ANNE 12 Page, Table I B. Agricultural Sector Mission Expectations 1980/1971 1970 1971 1975 1980 Annual % of eiportedT Increase Population: Total (millions) 119.2 130.0 145.0 Annual % increase 2.08 2.4 GDP - Annual 1% increase ---------- 6% --------- Range ------- 5 to7% ------- GDP per capita (% increase), ---------- 39 --------- Range 2.9 to 4.9 ------ Rice Production: High (million tons milled) 12.0 12.7 15.5 18.1 14.0 Low (million tons milled) 12.0 12.7 14.5 16.6 3.0i Rice Area: High (million hectares) 8.2 8.5 9.7 Low (million hectares) 8.2 8.5 9.3 -"9 Rice Yield: High (quintals/hectare) 14.7 15.0 18.16 2.4 Low (quintals/hectare) 1.7 15.0 1.9 Per Capita Rice Production: (Net: minus 6;9 for seed and losses) High (kgs. per capita) 102.1 102.1 110.1 120.0 1.8 Low (kgs. per capita) . . .102.1 102.1 10.7 110.1 . 0.8 Rice Imports: High (thousand tons) 950 600 300-400 0 . - Low (thousand tons) 950 600 300-400 300-100 - Corn. and Sorah Area (1,000 ha) 3,018 2,687 3,500 4,00 4.5 Production (1,000 tons) 2,888 2,138 4,000 5,000;. 10.6 Yield (quintals/ha) 9.6 7.9 11.5 12.5 5.2 Exports (1,000 tons) 250 500 1,-000- Cassava: Area (1,000 ha) 1,431 1,51.8 1,500 1,00 . .1 Production (1iOO0 tons) 10, 451 9,839 11,000 11,500 * Yield (quintals/ha) 73.0 64.8 73.0 75.0 1.7 Exports (1,000 tons) 2001/ 1001/ Sweet potatoes: Area (1,000 ha) 355.8 366.7 370.0 380.0 0.4 Prcyduction (1,000 ton,') 3,030 2.579 3,1)5 3,500 3.4 Yield (quintals/ha) 85.0 70.3 85.0 91.0 2.9 Peanuts: Area (1,000 ha) ,02 13 450 500 21 Production (1,000 tons) 293 299 360 500 5.9 Yield (quintals/ha) 7.3 7.2 8.0 10.0 3.7 S Exports (1,000 tons) 501/ 751/ Soybeans: - Area (1,000 ha) 684 630 700 -1,000 5.2 Production (1,000 tons) h88 h22 560 1,000 10.6 Yield (quintals/ha) 7.1 6.7 8.0 , 10.0; 4.6 Exports (1,000 tons) 501/ 200-001/ - 1/ Product equivalent ANNEX 12 Page 5 1980/1971 1970 1971 1975 1980 Annual % of .(ReportedT Increase Rubber: Area (1,000 ha) 1,969 1,972 2,000 2,200 1.2 Production (1,000 tons) 778 810 850 1,000 2.4 Yield (quintals/ha) 4.0 4.1 4.3 .5 1.0 Exports (1,000 tons) 800 900 - Sugar: Area (1,000 ha) 135 135 150 200 4.5 Production (1,000 tons) 715 833 1,100 .,500 6.7 Yield (quintals/ha) 53.0 62.0 73.0 75.0 2.1 Exports (1,000 tons) 0. Imports (1,000 tons) 118 150 50 Coffee: Area (1,000 ha) 354 - 368 375 3.6 Production (1,000 tons) 182 192 232 272 Yield (quintals/ha) 5.1 6,3 7.2 Exports (1,000 tons) 97 (100) 120 150-160 'loves: Area (1,000 ha) 83 - 100 128 5.2 Production (1,000 tons) 12.5 - 20.0 30.0 7.9 Yield (quintals/ha) 1.5 - 2.0 2.3 Exports (1,000 tons) 0 - 0 0-5 - Imports (1,000 tons) 8.8 - 3-5 Tobacco: Area (1,000 ha) 171 - 175 190 . 1.1 Production (1,000 tons) 71 77 100 120. 4.5 Yield (q,-intals/ha) 2 - 5.7 6.3. 4.1 Exports (1,000 tons) 14.2 - 20 30 - Imports (1,000 tons) 1.9 - 3.0 5.0 - Pepoer: . Area (1,000 ha) 46.4 (1957)- 50.0 5.0 1.8 Production, (1,000 tons) 15.7 )7. hO.0 30-60 - Yield (quintls/ha) 3. 8.0 5.5-10.9 Exports (,,,000 tons) 2.3 37.2 25-30 25-50 - Tea: Area (1,0co ha) (120) - ,10 100 - 1.8 Produftior (1,000 tons) ( 85) - 90) 95 1.1 Yield (quin al-/1,) 7.1 - 8.0 9.5 3.0 Exports (1,CCO tons) 35 - 35 - Coconlits: Aroa (1,000 hA) 168 3 750 2,000 1.7 Production (1,000 t-nM J,2)0 1,300 1, 0c3) 1,000 3.3 Yield (qitntal./ha) (7. 8.6 9.0 - Exports (1,0,10 tr) 168.6.1 200-2501 300.-3U0 - Palm . n K "1ls: Area (i,000 -n 123 126 150 200 5.3 Production (1,000 tons) - oil 207 210 400 600 11.0 - kernels b3.3 5.0 0.0 1 - 9.8 Yield (quintals/h1a) - oil (16.8) 16.7 27.0 30.0 6.8 - kernels (3.5) Exports (1,000 tons) - oil (1,n) 200 300 - - kernels (35) 45 55 - Cotton: Area (1,000 ha) 2 2 10 - 30 l00-1 0 - Production (1,000 tons lint) - - 3 -,_9 40-60 - Yield -(quintalz/ha) - - - 3.0 4.0 - Imports (1,000 . nt) 40 60 91-97 90-110 - 1/ Product nauivniqnL. ) stimatod. '/ 1970 crop was abnormally low. ANNEX 12 Page 6 C. Population 10. A wide discrepancy exists among the existing studies over the level and future development of population in Indonesia. The results of the 1971 census announced in June 1972 indicated that the 1971 population was 119 million rather than the 124 million assumed in most studies, and that the rate of population growth between 1961 and 1971 was only 2.1% rather than the 2.5% assumed in the Five-Year Plan and in most studies. In contrast the population growth rates for the 1970's in recent studies have ranged from just above 2 to 2.8% annually, and the assumed 1970 population figure has varied between 115 and 120 million. 11. Because there is still some doubt about.the results of the 1971 census, and because so many major studies have just been completed, con- siderable care should be used in interpreting results particularly with reference to the major differences between them with-respect to population and income assumptions. The Rice Marketing study (RMS) and the Rice Production Intensification Program Evaluation study (RPIPE) assume a 1970 population of around 115 million people for Indonesia, and a growth rate to 1980 of about 2.1 percent per year. These two studies relied upon the 1970 "house count", a limited sample survey taken for election purposes. This "house count" has now been confirmed by the census results. The Sugar, National Fertilizer study (NFS), FAO and Indonesian Supply and Demand (IS&D) (162) studies, all assume a 1970 population of close to 120 million, and a growth rate to 1975, 1980 and 1985 of between 2.5 and 2.8 percent per year. This results in discrepancies between these two groups of studies of about 8 million peo)le by 1975 and of between 12 and 13 million people by 1980. 12. For rice the new lower population estimates coming out of the 1971 census mean that 1 million tons less milled rice will be needed by 1975 to achieve the government target of 120 kgs. per capita and 1.6 million tons less would be needed to reach this per capita level by 1980. 13. Recent studies also differ about the pattern of population growth. The RPIPE and RMS studies assume the rate of increase in population will rise from 2.07 in 1970 to 2.43 in 1980. The NFS study assumes that the population growth rate will rise from 2.63 during 1970-74 to 2.8 from 1975 and 1980. The Sugar study assumes the annual population growth rate will rise from 2.4 percent in the 1960's to 2.7 percent during the period 1970-75, but will fall thereafter to 2.5 percent by 1981, and to 2.4 percent by 1986. The most recent IBRD "low" population growth rate, which is used in all ANNEX 12 Page 7 three of its projections, is still different--being 2.7 percent to 1973 and then dropping precipitously thereafter to 2 percent by 1980. 14. In contrast with the RMS and RPIPE studies, therefore, the 1971 census results make the population figures of the FAO, Sugar, NFS, IS&D and IBRD studies too high. The Sugar study, appears to assume the largest absolute level of population now and in the future (Table 1). D. Income 15. Estimates of Gross National Product (GNP), Gross Domestic Product (GDP) or National Income (NI) in Indonesia are presently based on very in- adequate statistics. According to the Indonesian Central Bureau of Statistics (CBS), annual rates of growth of Gross National Expenditures between 1960 and 1967 reached a high of 5 percent in 1960 and a low of -3 prcent in 1963, varying between 1 and 3 percent for the other years. Since population was growing at around 2 percent, per capita income growth was slight or negative. Prom 1965 to 1969 per capita income is reported to have increased at an average annual rate of 1.8 percent, due largely to accelerated growth in 1968 and 1969. 16. FAO calculations of Gross Domestic Product (GDP) show annual rates of growth of 2.1 percent from 1955 to 1960, 2.0 percent from 1960 to 1965, and 3.4 percent from 1965 to 1970. The FAO per capita rates of growth of GDP for the same periods are -0.1, -0.3 and 0.9. 17. The first Five Year Plan (1969/70-1973/74) did not stipulate a rate of economic growth, but "estimated that the targets in the various sectors will yield a 5 percent rate of economic growth." With population growth at 2.5 percent, it was assumed that "this would produce a 2.5 percent rate of per capita economic growth." 18. Rates of economic growth druing 1968-71 have been estimated at between 5 and 8 percent depending on the source. The Central Bureau of Statistics' figures for National Expenditures in constant prices indicate an annual growth rate of 7 percent in 1968 and 6 percent in 1969. The IBRD Economic Report of March 1971 indicated a 5 percent GDP growth rate in 1969/70 and "approximately 8 percent" in 1970/71. The IBRD report of November 1971 indicates a rate of growth of GDP of "around 6 to 7 percent" during 1971, but subsequently, in the same document, concludes hesitantly that "if a guess were to be hazarded it might be that Gross Domestic Product had grown by not less than 6 to 7 percent, and perhaps by more." The latest IBRD economic report estimates growth of GDP between 1968-72 at 6.8% per year in real terms. 19. The range of projected rates of economic growth during the 1970's vary from the modest trend extrapolations of FAO's "low" projections (3.0 percent per year for 1970-75 and 3.5 percent for 1975-80) to the very high IBRD growth rates of 7.2 and 8.1 percent for annual increases in total GDP (Table 2). In per capita terms, the difference in projected growth rates is ANNEX 12 Page 8 Table 2: INDONESIA - Population Projections in Major Studies Year BPS FAO Sugar RPIlE NFS RMS IS&D ----------------- 7- --rUillions ------------------------ 1961 97.1 96.4 1965 107.0 105.7 105.7 1966 109.6 108.3 1969 118.1 118.1 116.5 1970 119.8 115.3 115.3 119.5 1971 127.7 117.7 1972 127.7 120.1 1973 122.7 1974 125.4 1975 136.8 128.1 138.4 1976 130.9 138.4 130.9 (141.8) 1977 145.6 134.0 1978 137.0 1979 140.1 1980 157.1 143.4 143.3 17.2. 1981 158.7 1982 165.2 1986 1987 187.0 1991 1992 210.1 1/ The 1971 Census results would imply a population growth rate only rate only slightly higher than that assume4 by the RPIPE--1971 population of 119.2 compared with 117.7. ANNEX 12 Page 9 from FAO's low trend projections of 0.3 and 0.7 percent for the first and second halves of the decade, to the IBRD rates of 4.5 and 5.8 percent for the same periods. Table 3: INDONESIA: PROJECTED RATES OF GROWTH OF MAJOR STUDIES 1970-75 1975-80 1970-75 1975-80 Total Per Capita IBRD (37) Alternative I GDP (factor cost) 7.2 8.1 4.5 5.8 GDP (factor cost) 7.0 8.0 4.3 5.7 GNP (market prices) 7.5 8.0 4.6 5.7 Alternative II GDP (factor cost) 7.2 8.3 4.5 6.0 GNP (factor cost) 7.0 8.3 4.3 6.0 GNP (market prices) 7.5 8.3 4.6 6.0 Alternative III GDP (factor cost) 6.9 7.9 4.2 5.6 GNP (factor cost) 6.7 8.0 4.0 5.7 GNP (market prices) 71 8.0 4.4 5.7 First Five Year Plan 5.0 2.5 FAO Commodity Projections 1970-1980 Trend 3.0 3.5 0.3 0.7 High 4.5 5.5 1.8 2.7 Indonesia S & D 6.0 6.0 3.3 3.3 Rice Marketing 4.6 4.6 2.5 2.5 National Fertilizer 5.2 6.3 2.5 3.5 Rice Intens. Eval. 5.0 2.92 3.5 (1971-77) (1977-92) Sugar Study 7.0 6.5 4.3 3.9 20. Since these projections were made additional evidence has become available indicating considerable growth prospects, at least over the next several years in the mining construction, industries and forestry sectors. Moreover, agricultural investments in rehabilitation of estates and irrigation facilities can be expected to mature during the second plan period and strengthen yield increasing trends already evident from increased use of high yielding varieties, fertilizer and other inputs. The general view now held at working levels in government for growth prospects during the second plan period 1974-78 appears to be for a growth rate of about 7% per year overall implying a rate of about 4.5% per year for agriculture and about 9.5% per year for all the other sectors. ANNEX 12 Page 10 21. A large part of the recent growth in agriculture has been due to increased rice production. With the statistical adjustments in rice output upward and population downward and a number of other factors pointing toward slower growth in rice production in the future than that reported during 1968-71, the contribution of rice production to increases in total agricultural production can be expected to fall off during the 1970's. To sustain the rates of growth of 4.5% per year for agricultural GDP sizable increases in the output of other crops must be achieved. E. Commodity Expectations in Brief 22. Rice accounts for roughly 50% of the value of agricultural produc- tion excluding livestock, fisheries and forestry. Its value in any agricul- tural production index exceeds all other crops by a wide margin. In 1970 the Indonesian government revised its production series upward by more than one- half million tons of milled rice. This adjustment was based on new crop cutting samples of the Central Bureau of Statistics 1/. No attempt has been made to revise earlier figures. This adjustment lent a sharp upward bias to reported rice production in 1970 and 1971 compared with production reported for earlier years. The full impact of this upward revision is on the rice yields reported for Java. The Indonesian Department of Agriculture reports area and yield for rice for all islands, and has recorded yield figures for Java considerably higher than the CBS for more than a decade. The upward revision by the government in 1970 is about half the average difference between these two series during the 1960's. 23. As a result, although rice production has indeed increased rapidly since 1968, the rate of increase has not been as rapid as reported. Part of this increase involved picking up slack in the rice economy, part was a real increase in rice production and part was a statistical adjustment. As noted earlier, this statistical adjustment, combined with the new lower population estimate makes the availability of rice in Indonesian prior to 1968 much larger than had been assumed. Rice consumption per capita is now some 5 to 10 kgs above recorded historical consumption levels. 24. Projections of the demand for rice in Indonesia for 1975 range between 13.5 and 15.2 million tons, and for 1980 between 15.7 and 19.3 million tons. These can be compared with the reported 1971 production of 12.7 million tons. Such large discrepancies exist among the major studies on the locational and area-yield combinations appropriate to achieve a given level of output, that it would be unwise to base a large-scale transport, processing and marketing program on any one of them. Most important, they all rest on the assumption that the rice area will expand significantly on Java. Since expansion on Java has ceased since 1969, the short-run possibi- lities for increasing rice acreage on Java should be carefully reconsidered. 1/ The series are being further revised in connection with the preparation of the second plan. ANNEX 12 Page 11 25. The possibilities for increased rice production in Indonesia during the 1970's are more likely to involve a rate of growth of output of 3 to 4% than the average 8% increase reported from 1967 to 1971. This would result from a 1 to 1.5% area and 2 to 2.5% yield increase. It would produce 14.3 to 15 million tons of rice by 1975, and 16.6 to 18.1 million tons by 1980. A 4% rate of growth of rice production is within reach for Indonesia if steps are taken to improve the movement of rice from surplus (outer islands) to deficit (Java) areas, and to generate greater demand among poorer consumers. These rates of growth of rice production would provide from 110 to 120 kgs. per capita by 1980 with no imports. 26. Maize is the second largest user of land in Indonesia, but its value in total production ranks below rubber and cassava. Maize area and production have declined in the face of intensive programs for rice. Production is concentrated on Java, and yields are extremely low. It is an important factor in the diet in areas where it is grown, but is apparently being replaced by more plentiful rice supplies. Very little is reported fed to livestock, and small amounts are exported. Various pilot projects in East Java and South Sumatra, and experience in North Sumatra, have demonstrated that with better seeds, fertilizer and improved credit and marketing opportunities, farmers can raise production substantially. 27. Japan and smaller markets in Asia have shown an interest in importing maize from Indonesia, but internal market costs (real and disguised) and port and ocean freight charges make Indonesia's maize more expensive than that of other suppliers. The opportunities for corn as an export area are considered good. Exports of 500,000 tons in 1975 and 1 to 2 million tons are possible by 1980 with intensive corn programs. More emphasis will have to be placed on yield improvements on Java, and yield and area expansion outside Java, than is presently in prospect however and marketing and transport costs would have to be reduced. 28. The area and production of both peanuts and soybeans have moved upward fairly steadily in Indonesia. Some upward t:rend is noticeable for peanuts, but yields generally have not improved significantly for either crop. Price increases in rural markets in excess of 50% since 1968 indicate a strong domestic demand for these crops, despite the modest increase in output. Export prospects for both are good. A sizable yield increase in both crops is possible. Indonesia could expand its exports of both crops significantly to take advantage of the large world demand. 29. Indonesia's rubber production has grown relatively slowly in the past decade and half, with an appreciable increase in output reported in the past 3 years. Output increases during the 1960's came primarily in the svmallholder sector while output on estates stagnated or declined. Rubber yields in the 1960's were considerably lower than in the 1950's, with most of the increase in output resulting from expansion of area under small holder trees. Projections of Indonesian rubber production and exports by FAO and the International Rubber Study Group indicate modest increases in both, with Indonesia projected to fall progressively behind Malaysia and other countries as a supplier of rubber to the world. Despite competition in the elastomer market, a sizable growth in the future demand for natural ANNEX 12 Page 12 rubber is projected. The mission feels that with Indonesia's favorable growing conditions it should be able to compete effectively in the rubber market. to this end, greater support should be given to improving the efficiency of existing smallholder rubber, especially through programs to provide distribution of better planting material, better extension services, and at least some minimum credit and marketing assistance and nucleus estate programs to increase new plantings of rubber. The recent sharp increases in rubber prices and improved long run prospects for natural rubber should provide a strong impetus to such programs. 30. The recently completed Indonesian Sugar Study has been well done and the general conclusions and approach to increasing Sugar pro- duction appear realistic. Effective implementation of the rehabilitation program in Sugar should be carried out but more careful consideration should be given to the questions of the future level of demand and the best locations for production. The projected increase in domestic demand in the Sugar study is high compared with other projections and especially so now that population growth is believed to be lower. The Sugar and National Fertilizer studies suggest different areas for expansion of the sugar area. Obviously, before large area expansion efforts are undertaken, further surveys and trials should be carried out, not only to determine the best agronomic locations for expansion, but also the lowest cost economic regions. The possibilities should then be carefully compared with costs and benefits of further rehabilitation and expansion possibilities in Java. 31. Estate coffee production (mostly arabica) has stagnated in the past decade, while smallholder coffee (mostly robusta) has expanded rapidly. Domestic coffee consumption has apparently increased substantial- ly. Sizable stock accumulations are estimated for most years of the 1960's. Export since 1965 have averaged about 100,000 tons or more, well above the level of the first half of the decade. Substantial domestic demand increases will take place and the opportunities for exports may be greater than those projected by FAO in view of recent declines in Brazilian output. Expansion in the smallholder area may slow down, but is not likely to be halted. There is an urgent need for upgrading the efficiency and quality of smallholder coffee. 32. Tobacco production in 1971 is reported to be not significantly different from 1960. Estates presently account for close to 10,000 tons, and the remainder is made up of smallholders (about 55,000 tons) and virginia (about 10,000 tons). Smallholder tobacco yields are reported to be about one-fourth the level of 1953-57 (110 kgs. in 1963-67 compared with 400 kgs. in 1953-57). Rapid increases in Indonesian consumption of tobacco are projected in all studies. Most projections of Indonesian exports of tobacco foresee little increase over recent levels, not because world demand is limited but because other suppliers, mostly developing countries, are ex- pected to do better in supplying the world demand than in Indonesia. Wita expanding domestic demand, and opportunities for exports, production possi- bilities are good if the problems of heavy and illegal taxes, insufficient processing capacity and inadequate extension are corrected. A lower but ANNEX 12 Page 13 more stictly enforced, duty on imports of tobacco products, might well provide higher government revenues and better protection against imports. High domestic taxes on tobacco products, might well lead to a,considerable amount of undeclared transactions in tobacco and its products in a country such as Indonesia where tobacco is grown by many smallholders. Serious consideration should be given to how effective the present tax is and its impact on produc- tion, or declared production. 33. The area and production of cloves have doubled since 1960. Indonesia is the world's largest producer of cloves, but also the largest importer. Price increases in 1953/54 and especially recently have sparked large area increases. As a result, a large number of trees in the 16 to 20 year age brackets are now coming into full production, and the recent heavy planting will add large additional supplies in the future. The Clove Production Processing and Marketing Study in 1971 concluded that by 1977 or before Indonesia will be self-sufficient in cloves. A fluctuating production cycle is expected to produce intermittent export surpluses, and prices are expected to drop sharply. 34. Tea area and production differ substantially depending on how different Indonesian reporting organizations estimate smallholder area and production. Tea production and consumption probably have increased due to expanded smallholder increases. FAO and IBRD project very modest increases in tea exports by 1980. With slow growth in world demand, prospects for Indonesian tea do iot look favorable but sizable increases in exports domestic demand can be expected. 35. The fact that prices received in the Netherlands and at the London auctions (for Indonesian tea) are lower than for any other tea-producing country attests to the poor quality of Indonesian tea. This type of comment, made by importers of other Indonesian agricultural commodities, indicates the extent to which inadequate quality control, grading and initial proces- sing--rather than supply or demand constraints--handicap Indonesia in export markets. Serious attention should be given to these problems for a number of Indonesian export commodities as well, since Indonesia's share of future export markets hinges on improvements in these areas and on low cost, competitive trading. 36. Sharp declines in Indonesia's pepper production and exports have been reported for 1970 and 1971, blamed largely on Phythophthora foot rot in Lampung, where the largest concentration of black pepper is located. Apparently more is involved in the decline in pepper production and exports than disease problems, and a study of existing processing and marketing problems for pepper is needed. The loss of nearly $10 million in foreign exchange from reduced pepper exports from 1968 to 1970 underscores the need for such a study. 37. The coconut industry has stagnated for the past decade and a half with the area at between 1.5 and 1.6 million hectares and production between 1.1 and 1.3 million tons. All but about 17,000 hectares belong to small holders. Some recent progress has apparently been made in rejuvinating ANNEX 12 Page 14 small holder area, but estates have been reluctant to participate. In addition to producing some $30 to $35 million in export earnings from copra, coconut products are an important source of income to small farmers because of the ftuit and because of the wide use of coconut materials in rural con- struction and household products. The sharp rise in fresh coconut (71 percent) and coconut oil (27 percent) prices in rural markets since 1968 indicates the strong demand for these products of wide domestic use, and this demand is likely to continue into the future. The National Fertlizer Study concluded that "analysis for a wide range of coconut prices, showed that fertilizer use is highly profitable even at low prices for coconuts." Further expansion of the present, limited small holder rejuvination project is highly import- ant. Broad distribution of new planting material should be considered, since 60 percent of the present area is reported to have trees 40 years or older. Special attention should be given to the dwarf coconuts presently being adopted in North Sumatra. 38. Expansion of palm oil production is taking place. Under the estate rehabilitation projects substantial new plantings and replanting of average trees is being undertaken and new plantings of smallholder oil plantings w111 be financed under the Smallholder Development Project. Indonesian's favorable natural conditions, coupled with an intensive production and marketing effort should make it possible for Indonesia to increase palm oil production exports beyond the level projected in most studies. Projects for establishing new necleus estates and smallholder production are now under active consideration. 39. Cotton experimental projects in East Java and Lombok indicate substantial scope for development of cotton production in these areas. Plans are now underway to establish the feasibility of commercial scale develop- ment in these areas and to test out possibilities in other areas. If these developments are favorable an area from 50-100 thousand hectares could be developed by 1980 which would provide a significant portion of Indonesia's domestic requirements. Important production and organizational problems must be overcome however. ANNEX 13 Page 1 INDONESIA AGRICULTURAL SECTOR SURVEY THE RICE ECONOMY Introduction 1. Rice is the major component of the diet, the major user of cultivated land, and the major factor in farm income and wage costs. Regard- less of whether self-sufficiency is attained or not, Indonesia will have to increase rice production substantially in future years. Failure to do this would mean slow economic growth, higher rice import needs and most importantly a perpetuation of low incomes for a major segment of the Indonesian population. 2. Despite the great importance of rice in Indonesia, statistics on area, yields and production are far from satisfactory. The data which are available have been cast into especially serious doubt by the abrupt and in- complete revision of production figures in 1970, and by differences between central government and provincial data. Although inadequate statistics are a common problem in developing countries, the anomalies in recent Indonesian rice statistics are serious enough to affect significantly any evaluation of recent progress and projections for the future. 3. Very rapid increases in rice production have been reported in Indonesia since 1967 (Table 1). The output of most other field crops has not increased significantly. The major factor said to explain these in- creases in rice output has been an abrupt increase in yields, mainly on Java, brought about by rice production intensification programs. These programs have included larger supplies of fertilizer, high-yielding rice varieties, subsidized credit, and farmgate support prices. 4. However, a review of the statistics provided by the Indonesian government, and the adjustments that have been made in them by the govern- ment, suggest that: (a) output growth has been rapid, but not as rapid as reported; (b) more rice was produced and consumed prior to 1967 than reported; (c) a large part of the increase can be attributed to a substantial increase in area outside of Java; and (d) the statiqtical basis for evaluating produc- tion performance and prospects in specific areas is considerably weaker than has been generally acknowledged in a number of major studies (Annex 12). These points are critical for evaluating the specific conclusions and assumptions of the three recent studies devoted to rice production possibil- ities in Indonesia. Rice Production Statistics 5. For many years rice yields and production figures have been gathered for Java and the Outer Islands by the Indonesian Ministry of Agri- culture (DIPERTA), and for Java and Madura only by the Central Bureau of ANNEX 13 Page 2 Statistics (CBS). The latter were based on yield reports for purposes of land tax (IPEDA) calculations, and generally were considered understated. The DIPERTA statistics for Java and Madura indicated higher yields, and thus higher production, 1/ but were not used in official statistical publi- cations, and are not used now. In 1970, because of a new crop cutting system, a shift was made by the CBS to a higher yield and production series. The old CBS series would have recorded total milled rice production in 1970 of 11.4 million tons. The figure for 1971 is on the basis of this new series also, but figures prior to 1970 are still used to calculate yield and production levels (Table 1). 6. Although this adjustment is generally known, little attention has been given to its implications. It involves a major statistical upward adjustment in yields on Java--where most of the thrust of recent rice programs has been concentrated--because Java is the only area that the CBS reports on. Table 1: RICE PRODUCTION 1961-71 (CURRENT REPORTS) ('000 m tons) Paddy Milled (dry stalk) (52% basis) 1961 15,900 8,268 1962 17,111 8,898 1963 15,276 7,943 1964 16,192 8,420 1965 17,072 8,877 1966 17,960 9,339 1967 17,398 9,547 1968 19,550 10,166 1969 20,465 10,642 1970 /1 23,065 11,994 11,417 /2 1971 Ti 24,454 12,716 -- /1 New, upward revised series. /2 Old series. Source: CBS. There is little disagreement about the area figures collected by CBS or DIPERTA. Work sheets from DIPERTA indicate the extent of differences for Java between its series and that of CBS from 1956 to 1969 (Table 2). With 1/ The two sources provide similar estimates of cropped area. ANNEX 13 Page 3 the exception of 1963, the DIPERTA series for Java aad Madura exceeds the CBS series by 1 million or more tons of dry stalk paddy from 1960 to 1969. 1/ For most years, it is about 1.7 million tons higher, and in two years it was more than 2 million tons higher. In 1968, the DIPERTA figure was over 3 million tons higher than the CBS series. 7. There is no basis for assuming that the high DIPERTA figures are correct, with respect to magnitude, but the adjustment by CBS in 1970 of one-half million tons of milled rice (1 million tons dry stalk paddy) is more than 50% of the average difference between the CBS and DIPERTA figures for earlier years. This indicates a strong movement in the direction of the DIPERTA series. 8. If the DIPERTA series is used to analyze past trends, it causes the increase in yields reported for Indonesia from 1960 to 1971 to drop from 25% to 13%, and the yield increase reported for Java to drop from 44% to 22% (Table 3). A major drop was reported in the rice area between 1960 and 1963, and all of the drop was reported on Java, 700,000 ha. Between 1963 and 1971 rice production increased 60% and 63.5% in Java according to the existing series, while the DIPERTA figures indicate a 53% and 51% increase. Half of the production increase for Indonesia is explained by increased area, and over 37% of the increase on Java is explained by the same factor. With the DIPERTA series area changes account for more of the production increase. Table 2: RICE PRODUCTION DIFFERENCE BETWEEN CBS AND MINISTRY OF AGRICULTURE SERIES FOR JAVA 1956-70 ('000 tons dry stalk paddy) Ministry of Year Agriculture CBS Difference 1956 9,638 9,146 492 1957 9,964 9,116 848 1958' 10,110 9,544 566 1959 10,326 9,719 607 1960 11,434 9,723 1,711 1961 10,386 9,236 1,150 1962 11,510 9,895 1,625 1963 9,287 8,545 742 1964 10,457 8,413 2,044 1965 11,860 9,422 2,438 1966 11,741 10,006 1,735 1967 11,306 9,569 1,737 1968 13,852 10,659 3,193 1969 13,011 11,295 1,716 1970 14,043 1/ Fifty-two percent dry stalk paddy would be milled rice, so the difference in milled rice average around 800-900,000 tons. ANNEX 13 Page 4 Two Views of Rice Production Growth in Recent Years 10. Although the DIPERTA figures may be inflated and CBS figures understated, the changes in production they indicate provide a somewhat more plausible view of the past then the existing series when combined with fertilizer use statistics, and other relevant material. 11. The National Fertilizer Study (NFS) 1/ noted the anomaly of expanded fertilizer use in the early part of the sixties with no apparent corresponding increase in yields. It suggested that "recent programs wiay have encouraged over-reporting." With the DIPERTA series, another explanation presents itself. There are two periods of sizable yield increase: 1962 followed by stabiliza- tion (except for the 1963 shortfall), and 1968, followed again by relative yield stability. The 1963 figure undoubtedly has a large weather explanation. This was also the period of conflict with Malaysia and an agricultural census year. The existing series shows relatively little significant increase until 1968, when yields move up drastically and more or less continously. Although the "fit" is not terribly good, the upward movements in yields on Java reported by DIPERTA--most of the fertilizer was used on Java in the earlier years as well as later--corresponds better with the fertilizer use figures than does the old series (Table 4). 12. The Rice Intensification Study (RPIPE) concludes that, for most parts of Indonesia, 1968 was an exceptionally good year. 2/ for West Java it was "the outstanding rainfall year." For Central Java it was "likewise an excellent rainfall year." For East Java, "both yields and rainfall were best in 1968." For North Sumatra the "best year on the record is 1968, followed by 1969." For South Kalimantan "the best year in terms of yield and rainfall was 1968, and in South Sulawesi "the correlation between lowland yield and wet season rainfall is very high for 1966-69, the best year being 1968, followed by 1969." 13. The 1958-63 period was the period of the Padi Centra program, which is generally agreed to have been reasonably successful, although it experienced numerous difficulties. 3/ Better extension services and larger supplies of fertilizer were associated with this period. The period 1964-67 is associated with the government's pushing of the Bogor students' DEMAS program, a modest but successful intensification program, into a large-scale BIMAS program in 1965-66. 4/ The difficulties encountered in this program were compounded by reduced fertilizer supplies, an unfavorable fertilizer/rice price ratio and rapid inflation. 1/ Final Report (Draft) Annex II, p. 42. 2/ Rice Production Intensification Programs Evaluation (Draft) 1971, (RPIPE). Chapter III. 3/ RPIPE, Chapter II, pp. 3-5. 4/ RPIPE, Chapter II, p. 6. ANNEX 13 Page 5 Tabole 3. RICE AREA, YIELD AND PRODUCTION FOR INDONESIA AND FOR JAVA, 1969-71 Area YinIdP •: CiTicial ofic_. Year :out- : series ___TA DIPLRTA. Ind:Java side - nesia: Java :indo~: J a7a - d ~: JaFa:.ap-: Java: In ~•A nesia. ne-,ia. .nesia. .nes ia. Million ha--- uiteas F.ectares •ilion Tor:s 1960 : 7.3 4.3 3.0 23.1 22.5 25,5 26.5 16.9 9.7 18.6 11.4 1961 : 6.9 4.0 2.9 23.2 23.1 24.8 26.3 15.9 9.3 17.1 10.5 1962 : 7.3 4.1 3.2 23.5 24.2 25.6 28.0 17.1 9.9 18.7 11.5 1963 : 6.7 3.6 3.1 22.7 23.4 23.9 25.6 15.3 8.5 16.o 9.2 1964 : 7.0 3.7 3.3 23.2 23.0 26.0 28.1 16.2 8.4 18.2 10.4 1965 : 7.3 4.0 3.3 23.3 23.7 26.7 30.0 17.1 9,6 19.5 12.0 1966 : 7.7 4.1 3.6 23.3 24.6 25.6 28,5 18.o 10.0 19.7 11.7 1967 : 7.5 4.0 3.5 23.1 23.6 25.5 28.3 17. 9.6 19.1 11.3 1968 : 8,0 4.3 3.7 24.4 25.9 28.5 32.3 19.6 10.7 22.8 13.9 1969 8.0 4.3 3.7 25.5 28,0 27.8 32.6 20,5 11.3 22,2 1t.0 1970 : 8.2 4.3 3.9 28.2 30.0 28.2 30,9 23.1 13.3 23.1 13.3 1971 : 8.5 4.3 4.2 28.8 32.1 28.8 32.3 24.5 13.9 24.5 13.9 1971/ 1960 :16.4 0. 25 44 13 22 45 43 32 22 1971/ 1963cj .26.7 19.4 26.8 38.0 20.5 26.2 6o,0 63.5 53.1 51.1 a/ Dry stalk paddy. ANNEX 13 Page 6 14. In 1967-68--when the second large area increase on Java developed, when the weather was best, and when fertilizer supplies increased sharply- the government also fashioned a much more reasonable BIMAS program, put into effect the Rumus Tani which aimed at maintaining the fertilizer/rice price ratio at 1 to 1 by subsidizing fertilizer, expanded credit and continuing to support rice ,prices. All of these factors lend support to a large increase in production in 1968. 15. By 1968/69 the BIMIAS Goton. Rojong program was instituted, the weather was no better, and in many cases not as good as in 1968, and the area expansion on Java was halted. All these things lend plausibility to a stabilization or decline in yields in 1969 as indicated in the DIPERTA series, by contrast with the increases reported by the CBS series. Changes in Ric6 Area 16. Increases in the rice harvested area have been an important factor in rice production increases. They have been especially important in the production increases since 1967, although this fact is not commonly noted. 17. The rice area remained fairly stable at 6.5 - 7 million ha between 1953 and 1963. It was above 7 million ha only in 1959, 1960, and 1962. Between 1963 and 1971, however, it is reported to have increased 1.8 million ha, and 1 million of these came between 1967 and 1971. By 1965 the rice area exceeded the previous peak of 1960, and since that time has risen an additonal 1.2 million ha (Table 4). A large part of this, presumably, came about without benefit of new or rehabilitated irrigation programs but rather through,spontaneous development of wet rice fields and terraces on newly opened land in the Outer Islands. 18. No competitive crop, such as sugar, nonirrigated rice or maize shows any significant reduction in area. 1/ The area of Ladang (unirrigated) rice has remained fairly close to 1.5 million ha since 1962. The maize area, although it fluctuates greatly greatly from year to year, was low (2.6 million ha) in 1963, when the rice area was lowest, and has fluctuated between 2.5 and 3.5 million ha in most subsequent years. The area planted to sugarcane increased during the same period that the rice area was growing (Appendix table A-1). 1/ There has been a "decline" in the maize area, but large annual fluctua- tions make it hard to identify how much, if any, of this may have gone to rice. See Annex 7. ANNEX 13 Page 7 Table 4: FERTILIZER USE AND RICE YIELDS, 1960-71 Use on :FoodCros Yields : : Indonesia : Java NPK : Year CBS Diperta CBS Diperta '000 tons Nutrients 1960 3.0 23.1 25.5 22.5 26.5 1961 23.4 23.2 24.8 23.1 26.3 1962 73.0 23.5 25.6 24.2 28.0 1963 70.7 22.7 23.9 23.4 25.6 1964 105.4 23.2 26.0 23.0 28.1 1965 90.3 23.3 26.7 23.7 30.0 1966 80.6 23.3 25.6 24.6 28.5 1967 48.9 23.1 25.5 23.6 28.3 1968 119.9 24.4 28.5 25.9 32.3 1969 192.5 25.5 27.8 28.0 32.6 1970 197.3 28.2 28.2 30.9 30.9 1971 234.4 28.8 28.8 30.9 32.3 1972 288.3 Table 5: CHANGES IN PADI SAWAH AREA, 1960-71 ('000 ha) :: : Bali Kali- and Year :Indonesia: Java Sumatra : mantan :Sulawesi:Nusa Tengarra 1960 5,975 4,006 873 363 415 317 1961 5,584 3,668 856 351 433 277 1962 5,836 3,703 934 402 500 296 1963 5,329 3,301 935 346 445 301 1964 5,484 3,295 1,001 414 483 289 1965 5,875 3,650 1,039 399 457 322 1966 6,011 3,682 1,126 455 482 264 1967 5,994 3,636 1,005 449 495 330 1968 6,364 3,857 1,152 458 561 334 1969 6,544 3,947 1,280 421 569 325 1970 6,743 3,946 1,295 450 712 327 1971 6,955 3,913 1,517 478 676 370 ANNEX 13 Page 8 19. The reported decreases and subsequent increases in rice area took place in padi sawah (wet rice) land (Table 5). The padi sawah area in Indonesia was almost identical in 1967 to that of 1960--5,975,000 ha in 1960 and 5,994,000 ha in 1967. During that period it increased steadily in Sumatta, almost 200,000 ha. With fluctuations, it increased also in Kalimantan and Sulawesi. The declines in rice area from 1960 through 1963 appear to be almost exclusively in the padi sawah area of Java and Madura, and Bali and Nusa Tenggara. Between 1960 and 1964 the padi sawah area on Java fell 711,000 ha. By 1969, 652,000 ha of this area had been recovered, and there has been a slight decline since. 20. Between 1967 and 1968, the padi sawah area increased abruptly by 370,000 ha, of which 220,000 took place on Java, 75,000 on Sumatra, and about 66,000 on Sulawesi. In 1969 the padi sawah area moved up again by another 180,000 ha, of which 90,000 was accounted for by Java and 130,000 by Sumatra. Since 1969 the padi sawah area has increased another 400,000 ha, with all of the increase coming outside Java, mostly on Sumatra and Sulawesi. Future Area Expansion Possibilities 21. All of the recently completed rice production studies anticipate sizable future area increases on Java. That the area has leveled off since 1968 makes this a questionable expectation, since the major impact of rehabilitation projects will be to increase yields rather than expanding double cropping. The 500,000 ha increase in wet rice area outside of Java, since 1968, reinforces the conclusion that the Outer Islands will be the major source of continued significance area expansion. 22. Between 1967 and 1971, when the padi sawah area is reported to have increased 1 million ha, two provinces where especially rapid area expansion took place stand out--North Sumatra, where it increased over 200,000 ha, and South Sulawesi where it increased 154,000 ha. Almost half the padi sawah area expansion is reported to have taken place in Sumatra, one-fourth in'Java, and one-fourth in South Sulawesi (Table 5). 23. But there is something wrong with those statistics too. Provincial official statistics for South Sulawesi indicate a padi sawah area in 1970 of 480,129 ha, compared with a central government official figure of 599,349 ha. There seems to be no way of reconciling the 120,000 ha discrepancy. This discrepancy is especially crucial since South Sulawesi is identified as a major area of future rice production growth, and is now a surplus area. Whereas the central government figures imply no essential yield increase in Sulawesi, a sizable yield increase is implied in the provincial figures (Table 6). This type of discrepancy contributes to the impression created by the existing statistics that Java was the only areas where major yield increases took place between 1967 and 1971 (Table 7). A10EX 13 Page9 Table 6. RIGE AREA, PRODUCTION AND YIELD, SOUTH SULAWESIa/ S Pai Rndenan. ; ~Payen Gadu Padi Ludan- Produc : Produc :Produc: re . .ield Area .. Yield Area . Yl tion : :on: tion : 196Q1 :327 756 23.1 18,8 109 22.3 33.8 41.5 12.3 1962 : 380 756 25.2 50.0 115 23.0 47.9 ·68.1 153y : 330 720 21.8 38.8 104 26.8 43.2 60.7 378 966 25.6 28.0 71 25.4 54.1 78.4 1965 31, 758 21.2 71.9 182 25.3 46.9 59.8 1966 : 355 851 24.0 38.1 99 26.0 58.31 81.5 1967 362 947 26.2 37.1 88. 23.7 50.4 -68.3 196 : 428 1,179 27.5 31.11 100 31.8 50.4 77.7 1969 : 411 1,1403 31.1 54.5 158 29.0 50.6 68.5 13.5 1970 : 23 1,144 34.1 56.8 181 31.9 49.8 65.8 13.2 Ce-tral Gover-nt Gtatistics : Provincial Governumen- on yet rice area, yield and Statistics of v,et rie- prodtion in othSulawici : area, yield and productifon Are a Produ,ction . Yicid Area Production: Yeld 1 : 404 975 24.1 393 950 24.2 1967 : 00 1,035 25.9 399 1,035 25;9 968 : 459 1,279 27.8 459 1,279 27.8 196 : 46r7 1,201 25.7 467 1,561 33.4 1970 599 1,539 25.7 48c 1,625 33.8 1971 554 1,632 .29.5 Table 7. AREA, YIELD AND PRODUCTION OF PADI SAWAH (Central Government Statistics) a/ 1967 1971 or ~Pro po Region Area : Ye1. d Pro Area Yield. 29. . d on 1 ,085 29.1 3,158 1,5]6 31.4 4,837 Javra : 3,636 24.7 8,993 3,91- 34.1 13,3h8 Krli)antan: 449 ~19.2 863 478 19.3 974 Sul;esi : 1195 25.6 1,264 676 28.9 1,955 S Baui. & : T gara : 330 31.0 1,o:. 370 32.4 193 a/ Area in 1,000 ha; production in 1,000 t; yield in m quintal/ha ANNEX13 Page 10 Rice Consumption 24. The pattern of rice consumption in Indonesia is affected by the divergent views on both population and production. A larger production or lower population raises the level of calculated per capita consumption. Both of these adjustments have now been made by the Government. Table 8 compares the existing and DIPERTA production series, deducts from each 6% for losses, adds net imports, and then divides both series by the midyear population assumed until the 1971 census results were made available and by the midyear population implied by the 1971 census. 25. Changes in rice stocks did not significantly alter rice available for consumption through 1963. In 1964 almost a half million tons of rice was added to stocks, and almost the same amount was released from stocks in 1965. 1/ This would have had the effect of reducing consumption in 1964 and raising it in 1965, so that consumption in those two years was probably around 93 kg per capita with the old population series and 94 kg with the 1971 census series. Stock changes in subsequent years amounted to around 100,000 to 150,000 tons, which had little effect on annual consumption levels, except in 1968 when stocks of 350,000 additional tons were accumulated. 2/ 26. The DIPERTA series elevates the level of apparent consumption in Indonesia considerably, and the new population series raises it even more. In only two years does consumption fall below kg per capita--1963 and 1967. The former, as noted above, was the year when the rice area reached its lowest ebb adn yields were low. The latter was also a year of area reductions, about 450,000 ha, and yields on Sumatra were lower than at any time during the decade. For the years after 1964, the difference is from 7 to 10 kg per capita per year higher, a very significant difference for a rice-eating, poor country. 27. This adjustment takes on considerable significance when it is remembered that rice "consumption per capita was about 90 kg during eight of the prewar years from 1921 to 1941 and reached 95 kg in peak years..." 3/ By either the CBS or Department of Agriculture production series, and using the "high" population level, the level of rice consumption in Indonesia during 1970-71 appears to have been higher than at any point in this century, and 1/ Long term Projections of Indonesia's Import Demand and Export Potential for Selected Agricultural Commodities, University of Indonesia (3rd Progress Report) Draft, p. 38A. 2/ These two periods of large stock build-up add additional support to the Department of Agriculture's production series which shows especially large production increases in the two years of large stock build-up. 3/ Rice Marketing in the Republic of. Indonesia, Leon A. Mears, University of Indonesia, 1971, p. 46. Table 8. RIGE AVAIIABILITY, POPULTION AND APPARENT PER CAPITA RICE C0SUMPTION, 1961-71 Net pr cetion Net Per Capita suzppl Consumption Sttstical Net : : Mid year serics. Imports Official D population Official alida DIPETAERTAe DIPERTA Official 'DIPERTA Series •Serips Series Series - i:illion tons - - - - 1,000 tons - - - - Nillions - - - Kilograms - - b/ c/ d/ e/ d/ e/ 1961 7.8 8.4 1,064 8.9 9.5 96.4 96.4 92.3 98.5 952 8.4 9.1 1,096 9.5 10.2 98.5 98.5 96.4 103.5 963 7.4 7.8 1,076 8.5 8.9 100.8 100.5 84.3 84.6 88.3 88.5 -4 : 7.9 8.9 1,025 8.9 9.9 103.2 102.6 86.2 86.7 95.9 96.4 : 8.4 9.5 193 8.6 9.7 105.7 104.8 81.4 82.0 91.8 92.6 1966 8.7 9.6 306 9.0 9.9 108.3 107.0 83.1 84.1 86.6 92.5 967 8.5 9.3 346 8.8 9.6 110.9 109.2 79.3 80.5 86.6 87.9 9.5 11.2 625 10.1 11.8 113.7 111.4 88.8 90.0 103.8 105.9 1 9 9.9 10.8 604 10.5 11.4 116.6 113.6 90.0 92.4 97.8 100.3 70 11.3 11.3 950 12.2 12.2 119.6 115.8 102.0 105.3 102.0 105.3 1971 : ~11.9- 11.9 (600) 12.5 12.5 122.6 118.0 102.0 105.9 102.0 105.9 a/ 9 4% of reported milled rice production. b/ 'r.3ler serics. c/ 1971 ecnsus end of year population of 119.2 million. d sed on earlier population. e/ Based on 1971 census. aO ANNEX 13 Page 12 the Department of Agriculture series indicates that it was higher during 1968-71. The new population series raises consumption to even higher levels. Reevaluation of Recent Rice Production Progress 28. The major recent studies dealing with rice were too far along to be changed before the adjustment in the 1970 figure, and only one, the RPIPE study, considered the difference between the CBS and DIPERTA series. It assumed that the DIPERTA series was inflated, but did so prior to the revision in the 1970 rice production figure. All of them, therefore, analyze the past and project the future on the basis of the existing data. None of them appears to have seriously tackled the question of i.nconsistent or contradictory provincial-central government data, and only the RMS and the RPIPE studies use figures close to the new lower population figures. 29. The upward revision of yield and production figures in 1970 clearly calls for a corresponding revision in the data for earlier years. The magnitudes of such a revision may not be as a great as suggested by the DIPERTA figures, but given the size of the 1970 upward adjustment, over one-half mil- lion tons of milled rice, the inference is that an appropriate revision would be at least half the difference between the DIPERTA figures and the present CBS figures through 1969. This would produce a considerably more sober view of recent progress than that implied in the existing figures. 30. Three important conclusions follow from the above: (a) a very large part of the increased rice production reported since 1967 has come from a major increase in the wet rice area, 1 million ha; (b) the levels of rice yield, production and consumption on Java were higher during the "troubled years" 1964-66 than has generally been assumed; and (c) a large part of the present increased supply of rice, which has permitted-the government to main- tain rice price stability (with sizable imports), came in 1968, and was in large part due to factors which stimulated an increase in both yields and area. Existing Projections 31. Past and future demand, supply, and trade in rice have been treated in great detail in the major studies just completed. A summary and comparison of the major studies has also been completed by Cummings. I/ Actual develop- ments in the future will undoubtedly fall somewhere between the projections of these recent studies. However, they do not all arrive at the same con- clusions, and none of them seriously considered the issues raised above. There- fore, it would appear that they all assumed less rice available and lower rice yields than actually existed. Also, the changes in area since 1970, the cut- off year for most of them, have deviated significantly from what they anticipated would happen. I/ Contribution of Three Recent Studies to Improving the Information Base on Which to Plan the Rice Sector in Repelita II, Ralph Cummings, Djakarta, April 3, 1972. ANNEX 13 Page 13 Demand 32. Although the recent studies use alternative projection techniques, and assume different population and income growth rates, their proj(ctions of the future aggregate demand for rice in Indonesia fall within a narrow range. With the exception of the FAQ projections, which use generally much lower rates of income growth, all estimate demand within a few hundred thousand tons of 15 million tons by 1975 (Table 9). 33. The Rice Marketing Study projected demand for each of the 26 provinces for each year from 1970 through 1980, based on estimated consumption levels in the base year. It projected demand for a low, medium and high rate of income growth, of which only the "medium" growth is noted in Table 8. The RPIPE divided the population and economy into a rice (rural) and nonrice (urban) sector, with separate population, income and consumption assumptions, The IS&D analyzed rural and urban consumption surveys and consumption patterns of Java and the Outer Islands, and concluded that none of these methods pro- duced significantly different demand results. 34. Because the RMS and RPIPE studies used lower and slower population growth assumptions than did the others, their estimated present and future levels of per capita consumption are higher (113-116 kg in 1975) than in the other studies (108-110 kg in 1975). None of the studies accepted the nutritional goal of 120 kg per capita, embodied in the Five-Year Plan as a realistic goal for 1973. But all three major rice studies project a level of self-sufficiency by 1975, with the RMS and the RPIPE projecting surpluses in that year, and the NFS projecting a slight surplus by 1980. By 1980 all the studies, except the FAO projections, assume a level of per capita consumption within 5 kg of the first Five-Year Plan target kg per capita. 35. The FAO projections assume a much lower level of consumption (98- 100 kg in 1975 and 100-105 kg in 1980) and rice imports of 885,000 tons in 1980. These are augmented by 1.6 million tons of other cereals, mostly wheat. The supply section of the IS&D study had not yet been completed at the writing of this .report (July 1972). 36. There was essential agreement among all studies that the income elasticity of demand is around 0.64 and a price elasticity of demand of -0.5. This does not appear to change the aggregates no matter how demand is projected, although poorer and rural people have a higher income elasticity than those with higher incomes. Subsequent work in the Ministry of Agricul- ture and BAPPENAS which takes account of the distribution of increased .n- comes concludes that the income elasticity of demand may be considerably lower. 37. The three major studies all couched their results in terms of levels of production which would provide sufficient rice to satisfy the projected domestic demand by 1975. That is, the supply projections are not ANNEX 13 Page 14 independently arrived at. The FAO projected demand and supply separately, and its slower growth of production resulted in continuing imports. 38. General agreement on the quantity of rice demanded, and a sufficient supply to provide it, conceals a remarkable difference in the projected rice production pattern in Indonesia over the coming decade (Table 10). At the aggregate level, the RMS projects a rice area of 9.0 ha by 1975 and 9.8 million by 1980, and the NFS study projects an area of 8.4 and 9.3 million ha for 1974 and 1980, respectively. FAQ projects the 1980 rice-area of 9 million ha, which is the extrapolated 1955-70 trend. 1/ These differences in projected area are reflected in sizable differences in projected yield. The 1974 aggregate yield of the NFS (1.74 tons of milled rice per ha) is similar to the 1980 yield projected by the RS (1.78). Table 9: PROJECTED DEMAND FOR RICE IN INDONESIA (million m tons) Study Year NFS RPIPE RMS IS&D FAO 1969 11.74 1970 12.12 12.05 12.0 11.7 1974 14.60 14.32 14.01 1975 14.92 14.51 15.12 13.5-13.7 1980 19.3 17.4 18.40 15.7-16.4 1/ Agricultural Commodity Projections 1970-1980, FA0, Rome, 1971, Vol. 1, p. 108. ANNEX 13 Page 15 Table 10: COMPARISON OF RICE PROJECTIONS, 1975 and 1980 RMS RPIPE NFS 1975 1980 1975 1974 1980 Production (million tons) Java 7.84 9.20 - 8.20 10.40 Other Islands 6.87 <3.29 - 6.31 8.87 Indonesia 14.71 17.49 15.41 14.60 19.35 Rice Hectarages ('000 ha) Java 4,575 4,833 4,525 4,402 4,672 Other Islands 4,436 41987 4,680 4,004 4,630 Indonesia 9,011 9,820 9,255 8,406 9,300 HYV Hectarages ('000 ha) Java 1,600 2,800 1,367 1,461 2,411 Other Islands 12265 2 110 416 415 1 366 Indonesia 2,865 4,910 1,783 1,876 3,777 Average Yielde (tons/ha) Java 1.71 1.90 - 1.88 2.24 Other Islands 1.55 1.66 - 1.58 1.91 Indonesia 1.63 1.78 1.66 1.74 2.08 39. All three major rice studies project sizeable increases in the rice area on Java as well as the Outer Islands. However, none project the rapid increases envisaged in the First Five-Year Plan, which called for a total increase from 7.6 to 9.3 million ha between 1969 and 1973, with 1 million ha of this coming on Java (from 4.3 to 5.2 million ha). 40. It is difficult to account for the differences in projected area, yield and production patterns on a province-by-province basis. The RPIPE study projects a pi:oduction increase on Java from 1970 to 1975 of 7.3% per year, while the NFS projects an 8.2% increase for 1970-74, and the RMS projects a 5.3% increase. The NFS projects a uniformly more rapid yield increase, 6.3%, compared with 4.1% in the RPIPE and 3.7% in the RNS. On the other hand, the NFS calls for a very small area increase, 0.5% per year for Indonesia, whereas the RMS projects a 1.8% annual area growth rate, and the RPIPE projects a 2.4% rate of increase. ANNEX 13 Page 16 41. Other anomalies for critical growth areas emerge when the projected results for South Sulawesi, North Sumatra and South Kalimantan are compared. The NFS study projected an output increase of 12% annually for North Sumatra, during 1970-74, with a 3.6% area increase and an 8.1% yield increase. The RPIPE has an almost identical area increase but a miniscule yield increase of about 1.6%, while the RMS has almost the reverse growth pattern for area and yield. Future Supply and Demand Constraints 42. Two issues are raised by these conflicting results which deserve special notice. First, all of the studies project sizeable yield increases on Java between 1970 and 1975, ranging from a lot of 3.8% per year in the RMS to 7.3% annually in the NFS. The upward revision in the yield levels on Java in 1970 should presumably reduce the projected yield increases for Java in these studies. 1/ Second, when a decline since 1969 in Java's rice area is combined with this adjustment, it would be logical to assume that the growth potential of Java from 1972 to 1975 would be much reduced. 43. Of far more importance are the interprovincial and interisland implications of these different projections. Observations in Indonesia in 1972 indicate that in certain key areas such as North Sumatra, South Sulawesi, and South Kalimantan, the incentive to continue to expand rice production as rapidly as in the past few years is declining. In South Sulawesi, where production expanded rapidly since 1969, serious difficulties have been encountered in distributing these added quantities of rice to other areas of Indonesia. As a result, it has been difficult to maintain the floor price of rice, and officials indicate that prices in rural areas have dropped well below the floor level at peak harvest time and during periods when BULOG is not buying rice. It is hard to project continued rapid growth there if this situation continues. In North Sumatra, some officials argued that the rapid expansion of production in recent years was due to a sizeable price incentive. This incentive is said to be rapidly disappearing because of the declining real price of rice, and maintenance of a ceiling price which is too low given the higher production costs of the region. 44. The ability of the economy to absorb continued rapid increases in rice production is certainly more limited than has been assumed. Some of the reasons for this have been cited--undercounting production in earlier years, a higher level of consumption at present than was previously assumed, and a need to reappraise how much rice is actually demanded by Indonesian consumers. The overestimation of how much rice will actually be demanded in Indonesia may very well have come from the attachment to a 0.65 income 1/ With the exception of the NFS, which chose as base the average 1965-69 area, yield and production figures, considering these averages more meaningful than single 1970 figures. ANNEX 13 Page 17 elasticity. Although the income elasticity of poorer segments of the popula- tion may be even higher, there is a growing realization that not all Indonesians have shared in the income growth of the past few years. Also, the upward revision in rice production figures in 1970 distorts the growth rate over previous years, implying that agricultural production and income have not increased as rapidly as assumed. Thus, total income growth would have been less rapid, and so the income effect on consumption would have been less. Finally, to the extent that national income growth has been the result of growth in nonfarm income--petroleum, industry and other sectors-- little of this is likely to have spread out to the vast majority of low income rural consumers and been translated into increased demand for rice. Mission Projections 45. If the rice production figure for 1971 of 12.7 million tons (milled) is correct, projecting it into the future at 3%, 4% and 5% per year produces the results set forth in Table 1. Table 11: PROJECTED TOTAL AND PER CAPITA PRODUCTION OF RICE 1971-80 (milled basis) 3% Growth 4% Growth 5% Growth /2 Per Capita /1 /1 /1 Pop.- Production (Net) Gross Net- Gross Net-- Gross Net-- (Million) 3% 4% 5% (kg) 1971 12.7 11.9 12.7 11.9 12.7 11.9 116.5 102 102 102 /3 /3 1972 13.1 (107)- (107)- (107) 1973 13.5 1974 13.9 1975 14.3 13.4 14.9 14.0 15.4 14.5 125.8 106 110 114 1976 14.7 1977 15.2 1978 15.6 1979 16.1 1980 16.6 15.6 18.1 17.0 19.6 18.4 141.7 110 120 130 /1 Minus 6% for seed and losses. /2 Mid-year estimates calculated from the series in RMS and RPIPE. /3- Consumption per capita including imports. All demand and supply projections in the available studies are contained within these three rates of growth. A 3% annual increase in output, although it seems modest by comparison with most projections except FAO's, and by comparison with recent claimed rice production growth would, if modest imports are continued, raise consumption above the presently high level, ANNEX 13 Page 18 which were already presenting some demand ind marketing problems in late 1971 and early 1972. Output growth of 4% a year would produce 120 kg per capita by 1980 with no imports. A 5% growth rate would produce a consumption level of 130 kg per capita, which would greatly exceed the first plan target of 120 kg, and may be well beyond the level of effective demand. 46. Depending on the policies adopted by the government (see Annex on Marketing and Prices), a 4% rate of growth for rice production is possible. This would be about equal to the rice output &rowth rate during 1964-66 and 1969-71 if the rice production figures for the years before 1969 are raised half the difference between the CBS and DIPERTA series. It would be only half the output growth of 8% on the average for 1968-71 implied by the existing CBS figures. 47. But, as noted earlier, past rice production growth was the result of both area and yield increase. 1/ Since the rice area on Java has not increased since 1968, area expansion may not be as easy as assumed in existing studies, unless government takes effective steps to facilitate interisland trade and make rice growing outside Java more economically attractive. 48. A 2% area increase would exceed that found in any of the studies. An annual increase in the rice area of 1% would imply a 1980 rice harvested area of 9,255,000 ha, which is equal to the area projected in the RPIPE study for 1975, slightly less than the 9.3 million ha projected for 1980 in the NFS, and 600,000 ha less than the projected 1980 area in the RMS. It should be possible to sustain an area increase of 1-1.5% annually, but less of this may come on Java than is projected in major studies or was anticipated in the Five-Year Plan. With an area increase of around 1.5% per year, yield increases of 1.5%, 2.5% and 3.5% per year would be required to raise output by 3%, 4% or 5% per year. A sustained yield increase above 3% would be a remarkable achievement and should not be counted on. 49. The magnitude of output increases, the realism of consumption levels of 120 kg and the attainability and advantage of self-sufficiency in rice, depend far more on how the government deals with already evident problems of interisland and interprovincial movements and with income distribution problems, than it does on aggregate supply and demand considera- tions. In view of these difficulties and of the levelling off of the rice area on Java, the prospects for achieving a sustained rate of rice output growth of more than 4% do not seem good. On the other hand, should the government take steDs to foster more rap%d output growth on the Outer Islands, and move the rice to Java and other deficit areas promptly and efficiently, the prospects for a rate of growth in output of 4% or more seem considerably better, and imports could be reduced to zero except in years with particularly unfavorable weather. 1/ If the DIPERTA series is used, area expansion accounts for more of the growth from 1966 to 1971 than does yield. If half the difference is used, the contribution of area and yield are about equal. ANNEX 14 Page 1 INDONESIA AGRICULTURAL SECTOR SURVEY FIELD CROPS Introduction 1. The m o'r field crops occupy roughly 6.5 million ha, compared with the 8.5 million ha devoted to rice, and the roughly 19 million ha devoted to all crops. Individually, field crops account for at most 10-11% of the value of crop production, compared with more than half for rice. The generic Indonesian word for field crops other than sugarcane is polowidjo (second crop). 2. The principal field crops are maize and cassava, which are the main staples after rice, and sugar, once a major export and now an imported commodity. Although these crops are at times grown exclusively or separately from rice, they are in most cases literally second crops grown after the rice crop is harvested. The case of sugar is more complicated and involves an elaborate rotation system with rice: blocks of village land are formed in rotation for the exclusive production of sugar, then revert back to small- plot rice. Cassava and maize are often grown as insurance crops against the failure of rice, but they are also the "poor man's" crop, grown in those areas where rice does not grow well, or where--in the case of cassava--not enough rice can be grown on the land available. 3. As a consequence, there is is a close relationshl,p between rice and polowidjo crops. As rice production increased rapidly after 1967 the area and production of maize and cassava fell. Consumers apparently preferred rice, at prices that were falling relative to the general price level. The rice price did not fall relative to maize and cassava prices, however. The strong inducements--support programs--to grow rice since 1968 undoubtedly caused some shift in land use to rice from polowidjo crops. Relatively more favorable returns from rice than from sugar also explain part of the decline or stagnation in sugar production in recent years. 4. To a far greater extent than rice, polowidjo crops are concentrated on Java. Only about half the rice area is located on Java, but all of the sugar, 75% of the maize, 80% of the cassava, 60% of the sweet potatoes, 50% of the groundnuts, and about 90% of the soybeans can be found in Java. Furthermore, the overwhelming proportion of these crops are grown in the two Javanese provinces of East and Central Java, mostly the latter. 5. There is, therefore, a critical question of competition for land on Java among these crops. Although extremely dense population preasure on Java is widely recognized, little attention has been directed to the question ANNEX 14 Page 2 of how expansion of one crop affects the production of others. The obvious desire to expand food production on Java resulted in large projected area increases for all of these crops, as well as rice, in Java during the First Five-Year Plan. These area increases did not materialize. In fact, the area in rice stopped growing after 1968 and the area in maize has declined. For most other polowidjo crops it has not grown. Despite this, most of the recent projections studies have projected major increases in the area of rice, maize and sugar on Java during this decade. Recent experience suggests that land expansion possibilities on Java are more difficult than generally recognized, and that far greater attention should be focused on yield increases on Java, and area and yield increases outside Java. 6. Since 1968, the real price of rice has fallen, that of maize and cassava risen about 15%, and for other polowidjo crops increased 50% or more. The failure of output of polowidjo crops to respond to these price signals should be a matter of considerable concern. Since growth of the rice economy is likely to slow (Annex 13), continued agricultural growth will depend partly on growth in other food sectors where little growth has taken place recently. In summary, it appears that the production and demand potential exist in the polowidjo sector, but the problems of achieving growth of output are likely to be much more complex than was the case for rice. A higher priority needs to be put on this important smallholder crop sector. Maize and Sorghum 7. Whereas rice accounts for roughly 50% of the value of agricultural production (excluding livestock, fisheries, and forestry), maize ranks fourth (6.5%), after rubber (10%) and cassava (8%). It is the second largest user of land, however, and an important food crop in some areas, especially East and Central Java. Some maize is presently exported, and considerably larger quantities can be exported in the future. 8. The maize area fluctuates from year to year with a consistent pattern of low area in odd years and higher area in even years. Over the past two decades, the area has grown with a trend of 2.5%, but during the mid-60's the upward movement stopped. 9. It is argued that maize is an inferior food to rice and, since .it is not used as a livestock feed in Indonesia, increased supplies of rice since 1968 have reduced the demand for maize and thus the area devoted to it. On the other hand, the strong bias in favor of rice in government programs since 1967 has militated against maize production--it would have had to surmount a high credit and fertilizer cost barrier to compete with rice. The relatively slow growth in maize prices since 1968 lends support to the view that demand pressures for it are not strong when plentiful supplies of rice are available. 10. Maize is grown in almost all provinces. The largest areas are in Central and East Java and South Sulawesi, with lesser areas in West Java ANNEX 14 Page 3 and East Nusa Tenggara, Small area declines since 1966 (the peak year of the decade) have taken place in Bali and Nusa Tenggara, Sulawesi, and Sumatra, but major declines took place in the major growing regions of Central and East Java. Table 1: MAJOR MAIZE GROWING AREAS, 1966-71 ('000 ha) 1966 1971 Central Java 944 486 West Java 207 118 East Java 1,544 1,077 South Sulawesi 476 410 East Nusa Tenggara 172 172 Total 3,778 2,687 11. The area decline from 1966 to 1971 is exaggerated, because 1966 was a peak "even" year and 1971 was a low "odd" year. The average maize area during the five years 1966-70, roughly 3 million ha, was above the 1961-65 average, 2.87 million ha. The average for the three-year period to 1971 was only 2.7 million ha, however, and these lower volumes are significant because some of the increase in rice area in this period may have come from this source. However, it is premature to say that a new trend has been developing. 12. The highest yields are obtained in North Sumatra, where relatively little maize is grown. In 1971 the yield there on 19,000 ha was 2.2 tons per hectare, three times the average yield for Java and Indonesia. This yield growth developed since the mid-sixties. Prior to 1965, North Sumatra yields were not significantly above those of Java, or Indonesia as.a whole. Yields in West and Central Java have been in the range of 1-1.2 tons/ha during the 1960's, with Central Java showing occasionally sharp seasonal shortfalls. East Java, with the largest area (usually more than 1 million ha), has yields below 1 ton/ha and often as low as 0.6-0.7 ton. Existing Projections 13. Outside of the Five-Year Plan projection of area and production, the only other projections for maize are those of FAO 1/ and the National Fertilizer Study (NFS). 2/ The NFS accepted the Five-Year Plan targets for area and production, and essentially extrapolated them to 1980. It derived the demand for maize by extrapolating 1952 per capita provincial consumption estimates to 1969, and then projecting them to 1974 and 1980, assuming that no essential change in per capita consumption would take place. 1/ Agricultural Commodity Projections 1970-1980, FAO, Rome, 1971. 2/ National Fertilizer Study, Indonesia, Final Report (DRAFT), September 1971. ANNEX 14 Page 4 14. To this domestic food demand, the NFS added export target levels for for maize for 1974 and 1980, and an assumed domestic mixed feed demand for the same years, to derive total demand (Table 2). Table 2: NFS PROJECTIONS OF MAIZE DEMAND, 1974 and 1980 ('000 tons) 1974 1980 Human Consumption 3,560 4,350 Exports 700 1,600 Feed Supplements 65 150 Total Demand 4,325 6,100 It then calculated the fertilizer needed to produce the implied yield increase, and assumed that most of the fertilizer would be used in designated maize export areas. Table 3: NFS ASSUMED MAIZE EXPORTS BY PRODUCING REGIONS ('000 tons) 1974 1980 East Java 500 1,000 Central Java 100 250 South Sulawesi 50 175 Lampung 50 175 Total 700 1,600 16. The FAO projections for 1970-80 produce quite different results. Total domestic demand for maize is projected to grow from 2.8 million tons in 1970 to about 3.2 million tons in 1975, and 3.8 million tons in 1980. Per capita maize consumption is projected to recover slightly from the drop between 1965 and 1970, and to be around 24 kg in 1975 and 1980. Area and yield assumptions are not available for the FAO projections, nor are specific trade projections. However, FAO projects a net import of 323,000 tons of "coarse grains", which suggests that there will be either net maize imports or no significant increase in exports from the present level of 100,000- 200,000 tons. 17. The Indonesian Supply and Demand Study (IS&D) 1/, which has not as yet produced its supply conclusions, projects a strong movement out of maize to rice in the coming decade. It concludes from an analysis of the Second National Socio-Economic Survey, 1964, that maize is an inferior food, the 1/ Long-term Projections of Indonesia's Import Demand and Export Potential for Selected Agricultural Commodities. University of Indonesia, Djakarta, in process. ANNEX 14 Page 5 per capita consumption of which decreases as expenditure rises. For Java and Madura it calculates an expenditure elasticity of -0.48, and for the Outer Islands -0.79. It projects per capita consumption of maize to fall from 24.9 in 1975 to 22.8 in 1980, and 20.3 kg in 1985. On the Outer Islands it projects per capita consumption to decline from 9.1 to 7.8 and 6.2 kg for the same years. The IS&D projection of total food consumption of maize are set forth in Table 4. Table 4: IS&D PROJECTED DEMAND FOR MAIZE, 1975-85 (million tons) Year Java-Madura Outside Java Indonesia 1969 2.0 0.4 2.5 1975 2.2 0.5 2.6 1980 2.3 0.5 2.7 1985 2.3 0.3 2.7 Evaluation of Projections 18. The projected 1980 demand for maize for human consumption ranges from a low of 2.7 million tons in the IS&D, to a high of 4.3 million tons in the NFS. Projections of future maize trade range from small implied imports (FAO) to exports of 1.6 million tons (NFS). In no study does maize figure significantly as a domestic livestock feed. 19. The area projections of the Five-Year Plan and the NFS require a 4.3% annual increase in the maize area from 1965-69 to 1973/74, and a 6.1% production increase. The implied yield would be 1.1 tons/ha in 1973/74 compared with a 1971 yield of .8 tons, and yields for most of the sixties of around .95 tons. The NFS projects the maize area to expand by 4% per year to 1980 and concludes that at constant yields this would produce 5 million tons of maize. It applies to the projected 4.86 million ha in 1980 sufficient fertilizer to raise the total production of 6.1 million tons, implying a yield of 1.25 tons/ha. These projected trends contrast sharply with the essential stagnation of the past decade, and especially movements in area since 1967. 20. In 1966 the maize area was almost 3.8 million ha, which is close to the NFS and Five-Year Plan 1973/74 projected level, but as noted before, the maize area swings sharply from year to year, whi'h must reflect weather and cropping patterns. Such a well entrenched pattern is not likely to change quickly, and it is reasonable to expect large annual fluctuation in area, production and trade in the future 21. The magnitude of the annual fluctuations in the maize area has varied between 600,000 and 1 million ha, and neither the Five-Year Plan nor the NFS takes this fluctuation into account. If the NFS projected area (4.19 million ha) is to be taken as an average area for 1980, which it must, then the implied area fluctuations would be 4.2-5.5 million ha compared with fluctuations in recent years of 2.5-3.5 million ha. ANNEX 14 Page 6 22. In looking to the future, a significant fact is that despite a rise in the maize price relative to the price of rice between 1968 and 1971 of about 15%, maize production has declined or remained relatively constant. The subsidies for rice have undoubtedly contributed to the stagnation in maize production, but the aggregate demand does not appear as strong as for some other field crops. 23. On the other hand, despite essentially no change in area, yield and production during the sixties, the longer run trend (1951-1971), if extrapolated to the future, would produce an area of about 4 million ha by 1981-85 and a production of about 4 million tons. An aggressive policy toward maize production--at the least elimination of the discrimination vis-a-vis rice--could raise production of maize well above the trend line and produce 5 or 6 million tons by 1980. 24. The question of area is critical, however. The major projected increases in the maize area in the Five-Year Plan, and in the NFS are on Java. From 1965/69 to 1973/74, the total projected area increase is 1 mil- lion/ha, of which 600,000 ha are to come in Java. Thus, for maize and rice alone, the NFS projects area increases on Java of 800,000-plus ha in the first half and another 800,000-plus ha in the second half of the seventies, but the area in both these crops has declined on Java since 1969. Since the RMS, the RPIPE and the Five-Year ?lan all project even larger increases in the rice area on Java than does the NFS, it is reasonable to assume that had they dealt with maize and other crops the area problem would have been even more acute. This is not to say that area increases cannot take place on Java through rehabilitation and improvement of irrigation systems to provide more water to secondary crops, but far too much of a burden has been placed on Java and too little scope has been given to expanded area outside Java, and too much stress has been placed on area increases on Java and not enough on yield increases. The gap in the projections of maize consumed as food between the FAO and NFS studies on the one hand, and the IS&D study on the other, is more than 1 million tons by 1980. It would appear from other studies and from observations that the IS&D projected consumption is too low, but there does not appear to be much evidence that per capita consumption will rise. To the extent that the IS&D study proves to be correct--and this depends heavily upon the availability and price of rice-- reduced consumption of maize as food could increase the supply for export or feed use from zero to 1 million tons. 25. The prospects for increasing maize output, if they are tied to area increases on Java, and to the present modest programs are not particularly promising. Exports between 500,000 and 1 million tons by 1975 and 1980 might be achieved: (a) if the present pilot projects in Lampung, East Java and elsewhere are continued; (b) if modest improvements in the marketing system are made; (c) if smallholders are allowed access to credit and ferti- lizer and better seed. 26. Far more dynamic possibilities exist for maize, and possibly sorghum, if a major maize production expansion program were to be adopted. It would have to involve: (a) major stress on producing yield increases ANNEX 14 Page 7 on Java; (b) major stress in the Outer Islands on area and yield expansion; (c) a concerted effort to improve marketing and processing of maize at key export points; and (d) rapid expansion of core maize production and marketing centers along the private and cooperative lines presently being tried in Lampung and East Java (see Farm Structure and Organization). 27. An aggressive maize program is justified by the large and growing demand for feed grains, especially in Japan, the major market. It is also justified by expanded poultry and hog production which will demand consider- ably more grain feed than is presently projected for this use. (Indonesia should not repeat the Philippine experience of expanding egg and broiler production largely on imported feeds.) Sorghum possibilities also exist in Indonesia, and experiments with this crop should be pushed since preliminary results of field tests indicate considerable scope for ratooning and good possibilities for inter-cropping in rubber and oil palm planting schemes. However, with roughly 3 million ha already in maize, and most of it on Java, a major effort on maize is called for. 28. Large, consistent volumes of maize for export will not be. forthcoming for some time. Thus first reliance should be placed on improving the efficiency of the present marketing system, selling surpluses on foreign markets which can use smaller, uncertain supplies at competitive prices. Cassava and Sweet Potatoes 29. Land use, yields, and production of cassava and sweet potatoes depend to a considerable extent on the supply of rice, since there is broad general agreement that they, especially cassava, are inferior substitutes for rice. Despite this, and despite an extremely rapid projected increase in rice and maize production in the Five-Year Plan, a significant expansion in production of both crops was also planned (Table 5). 30. There is little relationship between the reported production figures and the Five-Year Plan (Table 6). 1970 and 1971 were evidently poor growing years, because reported yields are especially low, 64.78 quintals/ha (6.5 tons) for cassava and 70.34 quintals for sweet potatoes. Yields for cassava usually average around 75 quintals. For sweet potatoes, yields were around 65 quintals from 1960 to 1965, about 50-60 quintals in 1966-68, and were reported to be 82 and 83 quintals/ha in 1969 and 1970. 31. The drop in reported production in both crops in 1970 and 1971, and the absence of any significant increase in cassava prices, seem to correspond with the sharp rise in rice supplies, but the area figures and sharp yield drop would indicate that the level of production, at least for 1971, was below expectations. Existing Projections 32. The FAO and NFS project sharply different demand patterns for these two crops. FAO assumes an income elasticity of .20 while the NFS ANNEX 14 Page 8 assumes eleasticities of -.1 for 1974 and -.2 for 1980. Because of this, and because the NFS assumed more rapid growth of income and rice production, it projects 1974 demand for cassava at 9.9 and 1980 demand at 11.2 million tons (about the level of total production in 1968-69). The demand for sweet potatoes is projected to grow at the same rate as cassava, 2.5% per year, which would amount to about 3.75 million tons by 1980. This total demand for cassava and sweet potatoes of about 15 million tons compares with FAO's demand for starchy roots of 16.3-18.8 million tons in 1980. 33. The NFS study assumes that the projected increase in output by 1980 can be obtained exclusively through yield increases, so the projected 1980 area is identical to the reported 1970 area, 1.4 million ha for cassava, and 350,000 ha for sweet potatoes. 34. There is good reason to believe that cassava will remain in part a "poor man's crop". The extent to which poor farmers, presently dependent on cassava, will be able to consume additional rice, even if it is available at prevailing or lower prices, is questionable. The 1970 area for both these crops was exceptionally low, and the area under cassava increased in 1971 despite larger rice output. It does not seem warranted, therefore, to assume that the area in cassava is likely to decline in the near future. Most of the cassava--1.2 million of 1.5 million ha--is grown on Java, and much of it on less desirable land. Thus, not much cassava land is likely to be available for expansion of other crops. Table 5: PROJECTED AREA AND PRODUCTION OF CASSAVA AND SWEET POTATOES (THE FIRST FIVE-YEAR PLAN) Cassava : Sweet potatoes Year : Area : Production Area : Production ('000 ha) ('000 tons) ('000 ha) ('000 tons) 1969 1,482 12,287 478 3,363 1970 1,511 12,655 489 3,325 1971 1,532 12,963 500 3,393 1972 1,533 13,271 507 3,431 1973 1,568 14,283 510 3,807 Source: (163). However, there is reason to oelieve that cassava production on newly developed land particularly outside Java, may be more significant than indicated in the statistics. Cassava is a good crop on newly cleared land since it tends to keep out "alang alang" (imperata cylindrica) which is a major weed pest. Moreover, in reclaiming land which has been taken over by "alang alang" cassava can play a major role in preventing reinfestation after the initial clearing of the land. The recent emergence of a number of pelletizing plants in Lampung to process cassava for export support this view. ANNEX 14 Page 9 Table 6: REPORTED AREA AND PRODUCTION OF CASSAVA AND SWEET POTATOES, I1963-71 Cassava Sweet potatoes Year : Area : Production: Area Production ('000 ha) ('000 tos)(00h ( 0tons) 1963 1,598 11,678 484 3,069 1964 1,578 12,262 620 3,958 1965 1,754 12,643 416 2,651 1966 1,513 11,232 401 2,475 1967 1,524 10,747 360 2,143 1968 1,503 11,355 403 2,364 1969 1,467 11,034 369 3,621 1970 1,390 8,955 350 2,947 1971 1,518 9,839 367 2,579 Table 7: EXPORTS OF CASSAVA PRODUCTS, 1965-69 Cassava : Tapioca Year : Chips : Flour : meal --------------Tons---------------- 1965 84,210 74,704 54,043 1966 144,894 37,372 32,520 1967 110,597 30,312 7,327 1968 152,582 7,796 857 1969 99,684 5,934 2,225 35. Although data for 1970 and 1971 are not available, Table 7 indicates that exports of cassava products, although small, were significant. 1/ The major markets were the European Common Market countries of the Netherlands, Germany, Belgium and Luxembourg. There is every indication that this demand remains strong because of the policies of the common market toward feed grains, which make cassava chips a competitive feed component. 1/ The export value of these products reported by the Indonesian govern- ment is grossly underestimated. Even at $30 a ton, cassava chips in 1968 should have earned about $4,555 million. Combined with flour and tapioca meal, the export earnings from cassava products in 1966 and 1967 should have amounted to $10 million or more. ANEX 14 Page 10 36. The sharp decline in cassave production in 1970 and 1971 (Table 6) probably reduced supplies for export in these years. This, like the drop in maize production, must be attributed in part to the rice programs, and the resultant losses in foreign exchange earnirnge should appropriately be charged against the savings attributed to reduced rice imports (which were not, as a matter of fact, reduced significantly during these two years). 37. Since cassava and sweet potatoes are both responsive to fertilizery thought should be given to moderately increasing production of sweet potatoes primarily because of their food value, and cassava primaily as an export crop. This would benefit farmers who cannot benefit from present rice programs. Indonesian cassava is free of mosaic and appears to be much better than that found in African countries. There are varieties which produce yields two to three times those presently obtained, but to get such yields would require these varieties, much more fertilizer, and possibly better soils than those presently used. A careful study of the export market for cassava products should be undertaken, not only with a view toward Europe, but also with a view toward whether cassava products may become a more important feed source for the Japanese market in the future. 1/ If the market proves to be a good potential one, then serious attention should be given to developing centers for processing and channelling inputs and production advice in major producing areas, and to developing better marketing and export links. 38. Because of the large area devoted to cassava on Java, there is a strong temptation to consider raising yields for the purpose of conserving land, or for allowing land to revert to forest or other covers to facilitate water conservation. Both are desirable objectives. However, such programs should be considered in light of the alternative crops that can be grown on present land devoted to cassava, or the costs of moving farmers off this land. If good alternative uses exist for the land presently devoted to cassava, and this can be demonstrated through practical production trials, then a major effort should be launched to demonstrate how, and to assist smallholders, to raise and market those crops. The weakness of the existing governmental structure to carry out such a program, however, makes this a low priority issue unless good alternative crops exist. Groundnuts and Soybeans 39. The Five-Year Plan proposal for increased production of groundnuts and soybeans reported developments in these two crops, and the area location of both crops are given in Tables 8-10. 1/ The Japanese Government until recently discouraged cassava imports to protect Japanese sweet potato farmers. The ban has now been lifted. ANNEX 14 Page 11 Table 8: FIVE-YEAR PLAN FOR GROUNDNUTS AND SOYBEAN arvested Area Production Percent Increase (million ha) (million tons) 1969/70 1.27 0.95 2.6 1970/71 1.35 0.99 4.2 1971/72 1.49 1.08 10.9 1972/73 1.70 1.21 12.0 1973/74 2.02 1.40 15.7 Table 9 : AREA, YIELD AND PRODUCTION OF GROUNDNUTS AND SOYBEANS, 1963-71 Groundnuts Soybeans Area Yield Production Area Yield Production (000 ha)(t/ha) (000t) (000 ha) (t/ha) (000t) 1963 352 .67 235 539 .65 350 1964 372 .70 261 571 .69 392 1965 350 .67 243 583 .70 410 1966 388 .68 263 605 .69 416 1967 351 .69 241 589 .71 416 1968 394 .73 287 677 .62 420 1969 372 .72 276 553 .70 389 1970 400 .75 301 686 .57 391 1971 413 .72 299 630 .67 422 Table 10; AREA OF GROUNDNUTS AND SOYBEANS, 1966-71 ('000 ha) Groundnut Area Soybeans Area Other Other Java Islands Indonesia Java Islands Indonesia 1966 320 68 388 512 93 605 1967 287 64 351 497 92 598 1968 324 70 394 543 134 677 1969 290 82 372 476 78 554 1970 298 162 400 598 89 687 1971 317 96 413 529 101 630 The trend in both area and yield of groundnuts has been upward for the past two decades, around 1.7 percent for area and 1 percent for yields, whereas for soybeans the trend in area has been 211 percent upward with negligible change in yields. The groundnut area on Java fluctuates from ANNEX 14 Page 12 year to year but has remained fairly constant at from 290,000 - 320,000 ha in recent years. Increases have taken place in the Outer Islands. 40. The NFS concludes that a sizable yield increase in both crops is possible from new varieties, improvements in presently poor cultivation practices and some fertilizer. This is a plausible conclusion. The NFS projects that the groundnut area will expand to 500,000 ha by 1980, most of this coming in the outer islands, so that the area and production pattern in 1980 would be as follows: Java Outer Islands Indonesia Area Prod Area Prod Area Prod ('000 ha) ('000 ha) ('000 ha) ('000 ha) ('000 ha) ('000 ha) 1974 300 219 145 106 445 325 1980 337 246 163 119 500 365 For soybeans the NFS expects "a reduction in area but a more intensive cultivation." 41. The NFS projects domestic demand for groundnuts to grow from about 240,000 tons in 1965-69 to 320,000 tons in 1974, and 400,000 tons in 1980, and it projects a 5 percent annual growth rate in exports of groundnuts and groundnut oil. For soybeans, it projects a modest increase in total demand in 1980 that "will be about 100,000 - 150,000 tons above present consumption." 1/ Thus, it raises the 1970 level of production (and assumed consumption) of 390,000 tons to 423,000 tons in 1974 500,000 tons in 1980. The FAQ projected demand for pulses, nuts and seeds, which cannot be disaggregated, is equally pessimistic, from 1.9 million tons in 1970 to 2.3 million tons in 1975, and 2.6 million tons in 1980. Since the income elasticity of both is low (0.3-0.4), and both these projections assumed a larger population than the 1971 census implies, these studies would presumably adjust demand downward. 42. With continued neglect, and dependence on the domestic market, this might well be the pattern of future production and trade. Indonesia's share in the world market for the two crops is so small, that exports could be increased if production costs were competitive and if reasonable quality control were provided. The major world supplier of groundnut products is West Africa. Reduced supplies from this area in the late 1960's produced high prices and induced suppliers of other oils to invade the market, mostly in Western Europe. There is no doubt that reasonable quality nuts and oil could have found ready buyers there during this period. Once established in that market, a continued competitive position could have been retained. 1/ NFS, p. 198. ANNEX 14 Page 13 43. The demand for soybeans and meal in the past half dozen years is well known. There is no doubt that the United States will continue to supply much of the world market for these commodities, but the strength of the demand is such that Indonesia could have exported sizable quantities had they been available at competitive prices. 44. The strong rise in the domestic price of both these crops since 1968 (roughly 50%) suggests that the domestic demand is also strong. Despite the strong foreign and domestic demand, however, output has grown slowly. his has been partly a function of the concentration of resources on rice, but to a greater extent it has been due to a general neglect of these smallholder crops. 45. A number of pilot projects--cooperative or private--should be started in the present growing areas, and in areas which show potential for future growth. These should combine at least sufficient land to conduct worthwhile practical experiments with new varieties, rotation trials and better cultural practices, but at the same time provide sufficient production to market profitably. These projects should also serve as a nucleus of basic seed, fertilizer, and extension services to farmers in the surrounding area, and, as a focal point for marketing the produce of smallholders, (See Annex 11, Farm Structure and Organization). Sugar 46. Of the three recent studies dealing with sugar in Indonesia, two--the NFS and Sugar 1/--tied their production increases to demand pro- jections. FAO projections of the late sixties (Commodities Projections for 1975-1980) have been increased in its projections for the Second Development Decade, due essentially to higher income growth assumptions, and a revised level of per capita consumption. Since the population levels assumed in the three studies are not significantly different, it is not clear why their levels of assumed consumption in 1970 differ so much. Now that the 1971 census has produced much lower population levels for the decade 1961-71, and a lower growth rate than is assumed in these studies, each will have to be modified. The population component of demand should logically be reduced significantly, and the per capita consumption level in the base year raised. 1/ Indonesian Sugar Study, Tate and Lyle Technical Services Ltd. and Bookers Agricultural and Technical Services Ltd., March 1972. ANNEX 14 Page 14 Table 11: PROJECTIONS OF SUGAR DEMAND FAO NFS Sugar Study Commodities DD II Projections Commodities Projections 1970-80 1970-80 Per Total Per Total Per Total Per Total capita ('000 capita ('000 capita ('000 capita ('000 (kg) tons) (kg) tons) (kg) tons) (kg) tons) 1970 7.4 676 7.7 708 7.0 824 7.92 1,012 1972 1975 L 7.6 790 8.1 869 8.5 1,135 1975 H 7.8 813 8.6 921 9.75 1,420 1977 1980 L 8.1 983 8.8 1,104 10.0 1,475 1980 H 8.6 1,055 9.9 1,257 11.50 1,901 1982 13.67 2,560 1987 15.71 3,302 1992 47. The FAO projections assume an income elasticity of demand for sugar of 1.35, the Sugar study assumes 1.0 for 1970-75; 0.9 for 1967-80; and 0.8 afterward. The NFS study apparently adopted an elasticity of demand with respect to income of 1.5. Table 12: WHITE SUGAR PRODUCTION, AND DISPOSITION, 1960-71 Changes in Apparent Annual stocks on Consumption Percent Year Production Eports Imports estates (1-2+3+/-4) Charge -------------------------(tons)----------------------------- (%) 1960 667,000 34,800 - -75,590 707,790 -3.0 1961 639,000 142,713 - +34,733 461,572 -34.8 1962 583,000 33,279 1,756 -39,865 591,445 +28.1 1963 650,284 105,049 70 +47,747 497,558 -15.9 1964 647,000 104,627 113 -84,758 626,018 +25.8 1965 770,900 75,400 340 +32,747 672,093 +7.4 1966 605,000 27,262 23,504 +15,179 586,063 -12.8 1967 659,000 - 37,524 -41,596 739,020 +26.1 1968 600,595 - 107,757 +31,422 676,930 -8.4 1969 732,288 - 77,780 - 2,330 812,298 +20.0 1970 715,047 - 118,164 -50,634 883,845 +78.8 1971 833,000 - 150,000 n.a. n.a. n.a. ANNEX 14 Page 15 48. To become self-sufficient in sugar, the NFS study projects that Indonesia would have to raise output by 761,000 tons between 1970 and 1980, while the Sugar study assumes that output would have to increase by 1.1 million tons from 1971 to 1982. The Ministry of Agriculture has announced that sugar output would be increased by 1.2 million tons over the next 10 years, apparently based on the recommendations of the Sugar study. 1/ 49. The NFS assumes a yield of 10 tons per ha by 1980, which would require an additional 66,000 ha, to be achieved as f6llows: 1970-75 1975-80 Estates 20,300 40,000 Smallholder 2,860 2,860 23,160 42,860 Total 66,020 The Sugar Study concludes that despite sizable yield improvements some 50,000 - 60,000 ha of new land will be required on Java, and an additional 40,000 ha on the outer islands to reach its production level by 1982. These would add from 90,000 - 100,000 ha to a 1970 sugarcane area the Sugar study calculated at 135,000 ha, including 80,000 ha of factory and 50,000 ha of smallholder cane. 50. The projected increased area on the outer islands is the same in both studies--40,000 ha--and the assumed 1970 level of smallholder cane area is almost identical (about 55,000 ha). But whereas the Sugar study records an estate sugar area of 80,000 ha for 1970, th. NFS records only 66,700 ha. The NFS further states that the sugar area on estates "was 87,000 hectares more in 1965 than in 1969." 2/ The NFS study assumes that 20,000 ha of this lost 87,000 ha can be recovered by 1975, which is essentially all the additional area it anticipates bringing into production on Java. Thus, the NFS anticipates having only 87,000 ha of estate cane on Java by 1975, and maintaining that level through 1980, whereas the Sugar study expects the eane area on Java to rise from 80,j00-ha into 1970 to 130,000 - 140,000 ha by 1980. 51. A further source of difference is that the NFS and the Sugar study have different conclusions about where in the outer islands the 40,000 ha should be developed. The NFS projects an increase of 21,000 1/ Indonesia News and Views, March/April 1972. 2/ (NFS, p. 218). The Central Bureau of Statistics data does not indicate estate sugar area changes of this magnitude--it recorded 70,200 ha in ........... ANNEX 14 Page 16 ha of estate cane on Java from 1970 to 1974, and a 3,000 ha increase in smallholder area on Java. On the outer islands it projects a decrease of 3,000 ha of smallholder cane, and no estate cane from 1970 to 1975. But for 1975 to 1980 it projects that the cane area will increase by 40,000 ha, all of it on the outer islands, all of it estate cane and all of it in Atje (Northern Sumatra) and in South Kalimantan. 52. The Sugar study, on the other hand, specifically rules out Kalimantan and "most of Central and Northern Sumatra" as being within the equatorial region of evenly spread rainfall with no significant dry season. Thus, the Sugar study excludes any significant area expansion in the two areas where the NFS concentrates all of its outer island estate development. 1/ The Sugar study anticipates major increases in the cane area in South Sulawesi, Timur and Java, areas which for the most part are ruled out by the NFS. 53. It is noOt obvious that the Sugar study is entirely correct. It may be correct on agronomic grounds, but the opportunity costs of the land areas involved may alter the picture considerably. The areas indicated for expansion in the NFS have considerably fewer demands on them than South Sulawesi and especially Java. A quick decision on this matter is not called for, given this large discrepancy. 54. The Sugar Study suggests a series of experimental plots in various areas, and these are being undertaken with assistance from British aid to determine both the technical and economic potential of different regions. But given the strong difference of agreement, both Kalimantan and North Sumatra should be included. 55. Future Demand, Supply and Trade. The Sugar study may be a bit high in its projected domestic demand, so that a level of 1.3 - 1.5 million tons of sugar would be more reasonable for 1980. This would be between the recent "high" projections of FAQ and the NFS projection, but considerably below the level projected by the Sugar study. 56. There seems little doubt that Indonesia can and should expand sugar production, and that seth production could be achieved efficiently. The critical questions are how and where. Movement to ratooning and better incentive rental arrangements would do much to encourage higher yields. There seems little ,doubt that the incentives to grow rice in the recent past have been detrimental to sugar production. 57. However, as the sugar study discusses, there is a conflict between aggressive rice and sugar self-sufficiency programs, especially when they both concentrate on area increases on Java. The conflict is even stronger when so much of the planned increases L other crops call for sizeable area increases on Java. It is in this light that far more attention should be given to the possibilities for increasing sugar production outside Java, and 1/ (Sugar, Vol. I, pp. 12-13). ANNEX 14 Page 17 not necessarily on an estate basis. Many of these areas may not be as technically suitable as Java for sugar, but the opportunity costs of the land should be considerably lower. Investment Program 58. The Sugar study proposes a ten year program for the rehabilitation and expansion of the industry. IDA and the ADB have recently made loans financing the first stage of this program. Investmeats of the order of $140 million are planned for the major rehabilitation and expansion of a substantial number of existing factories. Cotton 59. Efforts to grow cotton in Indonesia have thus far met with limited success. The cotton area reached a peak in 1943 of 33,300 ha under Japanese encouragement and a lack of imports. They reached another lower peak in 1960 of 26,000 ha, and fluctuated between 10,000 - 20,000 ha during the first half of the 1960's. At present the area of about 2,000 ha in East Java and Lombok has been intensively developed with Dutch and French tech- nical assistance demonstrating the feasibility of growing cotton. 60. Cotton imports by Indonesia ranged from 20,000 to 65,000 bales (500 lb) for most of the sixties. In 1967, however, they jumped to 158,400 bales, and in 1970 reached 240,000 bales, most of which was provided through PL 480 from the United States. This increase in imports reflects higher income, concessional supplies, and rapid advancements in the textile industry, most of which began to have their full impact by 1968. 61. Estimates of Indonesian import demand Lor cotton was estimated at 410,000 bales (100,000 tons) by 1975 in an October 1971 study by the Unido Textile Industry 1/. Since then substantial investments in new spinning capacity make it likely that demand by 1980 could be of the order of 150,000 tons. 62. No study has proposed that Indonesia will be able to supply all of its needs in cotton. The results of recent experiments indicate that Indonesia can produce medium staple lengths efficiently. These make up a little less than half the expected demand for different grades of cotton. 63. A more critical factor is under what organizational form will cotton be grown. The present proposals, discussed in Bogor in February 1972, call for the development of 100,000 ha of cotton over a 10-year period, of which 60,000 ha would be in East Java, 7,000 ha in Lombok, 5,000 ha in Bali, and 28,000 ha in other parts of the country: Sulawesi, Sumbawa, Sumba, Flores, Timor, Lampung, and near Lake Toba in North Sumatra. 1/ Page 3. ANNEX 14 Page 13 64. Because cotton is a labor intensive crop, demanding 300 to 400 working days per hectare, and many especially labor intensive practices such as multiple pickings and careful weeding would have advantages in Indonesian conditions, present proposals involve an elaborate system of production on areas averaging 1/4 ha or less. Roughly 24,000 farmers with 6,000 ha would thus be involved in providing sufficient cotton for each production area surrounding one gin. To produce the hoped for eventual production of about 174,000 bales of medium staple cotton would require about 130,000 ha and half a million farmers. 65. To coordinate the activities at the farm to ginnery level, it has been proposed that cooperatives be established in the production areas to provide inputs to farmers, organize the transport of cotton to the gins, and provide the necessary production credit. These cooperatives are in turn to be coordinated by a National Federation of Cotton Growers Cooperative Societies. 66. It has also been proposed that the Cotton Board, established in 1970 by a Presidential decree, be entrusted with the responsibility for organization and implementation of the cotton program. As a National Cotton Board, it would be responsible for selling planting seed, buying seed cotton, and taking care of all gins, bags, seed dusters and seed crushing units. It would require considerable capital investment funds, and it is proposed that the National Cotton Board include representatives from the Ministries of Agri- culture, Finance, Industry, the National Bank of Indonesia, the Directorate General of Transmigration Cooperatives, BAPPENAS (the national planning organization), Provincial and Regional Government agencies, the BOGOR Research Institute, the textile manufacturers association as well as from farmers' organizations and farmers' cooperatives. 67. The need for smallholder projects is obvious, especially in many of the locations proposed in the project. However, it is questionable whether a project as complex as the one proposed will have the kind of success anticipated. Since cotton can be grown in rotation with rice and other crops, and the agroeconomic survey has demonstrated that it would compete if farm-gate prices are allowed to reflect world prices, it should be, wherever possible, a smallholder crop. 68. Since both techniques of production and the organizational struc- ture would be new, however, both should be approached experimentally. The gins are an obvious focal point for production advice, credit, and input supplies as well as processing, distribution and marketing. Experiments in different production areas should be undertaken with one gin owned and operated by cooperatives, another by a PNP similar to that in the sugar industry, and a third by private enterprise to determine which method provides the best organizational structure. 69. The technical problems of producing and marketing cotton may require some modification of smallholder units to be efficient, as seems to be the case in sugar. Experiments with combining blocks of smallholder land ANNEX 14 Page 19 to minimize the costs of extension, insect control and movement of inputs and outputs should be included in the above proposals. The inter island shipping bottlenecks must be given serious thought in a project with far- flung island implications such as that being proposed, these bottlenecks are already a serious impediment to goods movement. 70. The problems involved in a high level, national coordinating organization, juch as that proposed, will place a heavy burden on Indonesia's already overburdened and limited managerial capabilities. The possibilities of contracting the management and organization of the operation to a private firm, or independent operational freedom to each gin-production unit, should be carefully explored. Tobacco 71. Tobacco area and production statistics are especially poor and do not permit a meaningful analysis of area and yield developments. The production series in Table 13 was developed by series complied by Zainul Jazni for the years 1960-68 with figures provided by the M1inistry of Agri- culture in early 1972. The NFS records a 1970 tobacco area of 171,850 ha for 1970 which, if it follows the pattern of other NFS estimates, was the average area for 1965/69. Its projection of area is given in (Table 14). Table 13: TOBACCO PRODUCTION, 1960-71 (tons) Year Smallholders Virginia Estate Total 1960 45,022 23,431 7,277 75,680 1961 54,611 21,433 5,794 81,838 1962 52,615 19,703 6,517 78,835 1963 60,492 21,808 6,286 88,586 1964 43,066 9,035 8,043 60,144 1965 62,321 17,829 6,425 86,575 1966 68,700 20,200 11,999 99,899 1967 47,750 10,762 9,461 67,963 1968 37,575 5,668 9,880 53,123 1969 55,900 17,000 (9,700) /1 (72,900) 1970 37,000 14,000 (9,820) 71 (70,820) 1971 54,000 11,000 (9,940) 71 (76,940) /1 Source: 1960-68. Zainul Jazni; "Tobacco in the Indonesian Economic," Agro Ekonomika, Vol. 1, No. 2, July 1970; 1969-71 smallholders and Virginia (2); 1969-71 Estates- Five-Year Plan. ANNEX 14 Page 20 Table 14: PLANTED AREAS OF TOBACCO (ha) 1970 1974 1980 Estates 15,650 15,650 17,000 Virginia 36,700 36,750 41,000 Smallholder 119,600 119,600 137,000 Total 171,850 172,000 185,000 Source: NFS, Appendix Table 11-38. 72. Jazni indicates that the tobacco yield has declined from 400 kg in 1953-57 to 380 kg in 1958-62 and to 110 kg in 1963-67. This he attributes to "inflation, heavy taxes, illegal taxes, insufficient processing, and in- adequate extension." Most of the estate tobacco area is in North Sumatra (4,250 ha), Central Java (6,550 ha) and East Java (4,859 ha). Smallholder production is spread out over many of the provinces of Sumatra, South Sulawesi, Bali and Nusa Tenggara, with the largest concentration in Central and East Java (90,000 - 100,000 ha). 73. Domestic Consumption. Considering the inadequate production data, consumption estimates are crude. With a detailed consideration of the various uses of tobacco in Indonesia, the IS&D study concluded that the elasticity of demand for tobacco is about 1.0 and it projects domestic consumption to grow sharply from 48,800 tons in 1969 to 67,000 tons in 1975, 87,000 tons in 1980, and to about 1.14,000 tons in 1985. The FA0 projects Indonesian production, consumption and trade to grow as follows: 1964-66 1970 1980 ('000 tons) Production 72 70 100 Exports 16 4 10 Consumption 56 66 90 Total Per capita 0.53 0.55 0.57 74. Although FAO and the IS&D studies agree fairly closely about tobacco consumption in total and per capita in the future, the IS&D study assumes considerably lower present consumption, and thus more rapid consump- tion growth during the decade. This is partly because the IS&D study assumes larger exports than are reported. ANNEX 14 Page 21 75. Trade. Data on tobacco exports and imports differ considerably depending on source (Table 15). The discrepancy between the export figures of Jazni and CBS is accounted for presumably by smuggling. It is generally agreed that a sizeable amount of cigarette imports (90 - 95%) according to some estimates is also smuggled into Indonesia. This conclusion seems justified considering the plentiful supply of foreign cigarettes in towns throughout Indonesia at from US$.25 - .35 per pack, despite an import duty of roughly 140% (70% plus a surtax of 100% of the duty) and no registered importers of cigarettes. Table 15: INDONESIAN TOBACCO TRADE Year Exports Imports Jazni7-T CBS/2 CBS ------------1,000 tns- 1960 14,812 3,400 600 1961 16,192 3,400 1,500 1962 19,221 11,000 1,400 1963 18,157 11,900 900 1964 21,263 18,700 - 1965 19,841 12,400 - 1966 22,036 12,400 - 1967 28,352 9,000 - 1968 20,322 8,100 21,300 1969 n.a. 16,400 20,735 1970 n.a. 14,200 1,873 /1 Jazni, Tobacco in the Indonesian Economy. /2 Central Bureau of Statistics. 76. Reported import costs of tobacco and manufactured tobacco products were $7.4 million in 1969 and $1.5 million in 1970, compared with reported export earnings of $23.8 and $21.0 million for the same years. These export earnings were significantly above the 1967 and 1968 levels of $14.7 and $13.7 million, respectively, but not above the level of the mid-1960's, and were below the level of the late 1950's. 77. Sizable stock accumulations of tobacco were reported in 1970 and 1971 (70,000 100-kg, bales in East Java were reported as the ending 1970 stocks, and this had been reduced to 50,000 bales by March 1971) which in part accounts for the sharp drop in reported imports in 1970. 78. Future Growth in Demand and Supply. There is apparently little if any attention given to the question of the health hazard of smoking in Indonesia, and the FAO and IS&D studies are both probably correct that there will be a sizable increase in domestic demand with continued strong income ANNEX 14 Page 22 growth. The projections would indicate something close to a doubling of domestic demand in the next 10 to 15 years. Relatively favorable prospects for tobacco trade are projected by FAO for the 1970's;-although not as favorable as had been thought in earlier projections. 1/ 79. The U.S. share in world tobacco leaf exports fell from 40% in 1950 to close to 20% in 1965, due to growing competition from lower cost suppliers many of them in Asia. This situation is likely to continue with relatively high support prices in the U.S. Also favoring developing country suppliers is the continued trend toward a lower proportion of high quality tobacco for blending in filter tipped cigarettes. 80. FAO projects a more rapid increase in tobacco exports for Korea and Thailand than for Indonesia and the Philippines. The Philippines is expected to recover its 1964-66 export position by 1980 while Indonesia is expected to have considerably lower 1980 exports (10,000 tons) compared with the FAO assumed 1964-66 exports of 10,000 tons. This conclusion appears to follow from the realization that domestic Indonesian demand will be growing rapidly, and the assumption that there will be supply and marketing constraints. 81. The NFS projects an area expansion in the 1970's of around 2% per year, mostly on the outer islands, and correctly notes that there are sizeable possibilities for yield increases. 2/ The limitations on supply, however, appear to be due more to marketing and quality control problems than to area constraints. As with a number of other crops, the demand (both domestic and foreign), appears strong enough to permit sizeable output increases if Indonesia can effectively market its tobacco at competitive prices. At present not only can PL 480 tobacco be obtained, but cheaper tobacco is apparently also imported from the Philippines and India. 3/ 82. Despite the fact that "improved processing facilities" and "better extension" would undoubtedly improve the competitive position of Indonesia, there appear to be a variety of taxes, disguised charges, and other controls which play a significant role in retarding Indonesia's competitive position. 83. If protection for domestic production is desired it would be far better to have lower import duties which are enforced, than the presently high ones which apparently are not. By the same token, central government revenues from tobacco could probably be improved by deriving these from expanded production and exports through a lower tax strictly enforced and the elimination of other "charges" on tobacco which impede both production and exports, and probably do not benefit the government directly. 1/ Commodity Projections for 1970-1980, Vol. 1, p. 213, FAO, Rome 1971. 2/ NFS, pp. 235-238. 3/ Report of the U. S. Agricultural Attache, May 31, 1971. ANNEX 14 Page 23 84. Improvements in Production Practices. Wrapper Deli is produced on PNP IX, in North Sumatra, following methods established 100 years ago and which no longer seem economically justified. Less than 5,000 ha producing around 2,000 tons of tobacco leaves for cigars are grown every year, but to avoid soil infestation by nematode and bacterial diseases the land is put under fallow without any crop during tL following five years. Thus, over the six-year rotation, the effective yield on 30,000 ha is only about 67 kg per ha per year. Uncultivated fallow is considered a must to produce a very high quality. It is questionable that such practices are necessary and, if they are, it represents a misuse of land in a country where the domographic pressure is important. These 30,000 ha surrounding the city of Medan are in an excellent location to produce other crops. The soils of PNP IX are dacitic soils, among the best in Indonesia, and it could be possible to find a much better use. 85. A tobacco area of at least 185,000 ha (the NFS projection) seems likely by 1980. A serious effort to raise smallholder yields should produce a total production level of 50 - 75,000 tons above the current 70,000 - 75,000 ton level. Domestic demand will absorb about 90,000 tons leaving export supplies of 45,000 to 60,000 tons above those projected by FAQ. ANNEX 15 Page 1 INDONESIA AGRICULTURAL SECTOR SURVEY A. Tree Crops Introduction 1. The tree crops in this annex constitute the major export crops of Indonesia. Prior to World War II this sector was dominated by estate production, most of it owned and operated by non-Indonesians. Sugar, and to a lesser extent tobacco, discussed in the field crops annex, were also estate crops. 2. Prior to World War II, the estate organizations in Indonesia -- especially rubber and sugar, but also tea, pepper and oil palm were among the most advanced in the world. Indonesia was a major center of tropical research and a major source of planting material and production methods for the rest of the tropical world. The combination of the depressed 1930's, World War II and Japanese occupation, and the revolutionary post war years, culminating in the expulsion of foreign estate owners, managers and technical personnel, have virtually crippled the Estate sectpr. Area, and especially yields, have dropped sharply, so that now production is less than half what it once was, and yields often less than half. 3. In many areas, North Sumatra for an example, the period between 1930 and the present has witnessed the continuous encroachment of small- holders onto land once held by estates, which partly explains the rapid rise of smallholder production in many of the traditionally estate crops, as well as the rapid rise in the smallholder area in some food crops. 4. Compared with conditions four decades ago, the estates in 1967 had lost: a) their managerial and technical personnel; b) their highly developed, on-going research and practical experimentation in new planting methods and cultivation practices; c) their contacts with, and connecting links to, the market for their products and others engaged in similar production around the world; and d) their land. 5. The estates represented a tremendous sunk cost, but a greater portion of this once substantial asset had also been lost. In many respects the lost land was far less important than the lost talent and research in planting materials and production methods. New managerial and technical personnel can be obtained from abroad at a cost, but the on-going research on new planting materials and production methods will be much harder and more time-consuming to restore. ANNEX 15 Page 2 6. The effects of the past four decades and the fact that much of the estate sector was owned by the government or tightly controlled by it, made it important to develop a strategy for the future of the PNP's based on existing conditions. The strategy currently being evolved involves questions as to: a) whether to restore the estates to their former position; b) whether to abandon those portions of the estates which have been for all practical purposes lost to smallholders, or reverted to jungle, and restore what remains of the estates; or c) to develop the estates as integral parts of the Indonesian economy relying on them as a nucleus around which to develop and integrate the smallholder sector. 7. On practical grounds the first alternative seems to be ruled out. The second makes sense, if the objective is to restore a highly efficient export-oriented segment of the agricultural sector. This may be necessary in a number of cases because of the limited possibilities for smallholders in surrounding areas. 8. The last alternative has considerable merit. The large estates were at least in part a historical accident, but there is a strong presump- tion that they have production and marketing advantages normally not avail- able or appropriate to the smallholder sector. Whether this is due to their size and their infrastructure, or whether the needed services simply do not exist for the smallholders is of great importance. 9. There are some clear advantages to the estate form of production -- as opposed to smallholder production - because of economies of scale. How- ever, many of the advantages of the estates result from organizational, research and production and marketing factors, which could be shared with the smallholder sector. Thus, there is a strong case for developing them as focal points to expand and improve smallholder production and marketing. The estates could become the marketing and distribution pipeline for small- holders, as well as providing research and extension services and on-the-job training to smallholders. B. Rubber Background 10. Rubber comprises more than 10% of the total cropped area of Indonesia, and is by far the most important tree crop in terms of area planted and value, with 778,000 metric tons being produced from an estimated 1,969,000 hectares in tapping in 1970. 742,800 tons valued at US$213.4 million was exported in the same year. These exports constituted 55.8% of the total value of agricultural exports in that year, or 21.1% of the value of all exported produce. 11. Two-thirds of the rubber output is in the hands of smallholders, who are estimated by the Directorate General of Estates to have had 1,813,200 hectares under rubber in 1970, although much of this was not in tapping for ANNEX 15 Page 3 a variety of reasons. The remaining third comes from estates under both public (PNF's) and private, domestic and foreign ownership. All three groups of estates are in production to the extent of about two-thirds of their planted area, and are giving average yields ranging from a high of .75 ton/ha. of PNP's to a low of .49 on estates under private national ownership. The smallholders, in comparison, are obtaining only .32 ton/ha., calculated on the basis of total area planted to rubber, but this would be increased to a figure comparable to that of the private estates under the ownership of Indonesian nationals if the same proportionate area in production is assumed for smallholders as for estates. 12. Apart from indicating the general order of magnitude, very little reliance can be placed on these yield figures for the smallholders and pri- vate national estates, because of the difficulty of determining a precise figure for the area of smallholder rubber, both total and in production. There is also a probability of serious under-reporting in the production figures from the estates under national ownership. This could account for the much poorer yields calculated for these areas as compared with public and foreign-owned estates. The figures are, however, sufficient to illustrate the extremely low yields--less than half those of the PNP's--presently being obtained by rubber smallholders, who constitute the most neglected sector of the industry. This neglect is particularly apparent from the very limited extent to which the smallholders have been able to undertake replantings with improved planting materials, a development which can open the way to other improvements which collectively lead to better yields and higher net incomes for the cultivators. 13. The economic disruption stemming from political disturbance and inflation during the past two decades led to a considerable decline in over- all production, with a low of 582,000 tons in 1963. Since then, production has improved until by 1971 it had risen to 810,000 tons--almost back to the 1951 figure of 828,000 tons. To the extent that yield and area statictics can be relied on, Indonesia has roughly maintained the level of production through area increases with an annual linear trend of over 5% from 1951 to 1971 compensating for a 2.8% annual trend decline in yields. A major part of the increase has been from the smallholder sector. The Demand for Rubber 14. Total elastomer usage increased about 73 percent between 1961-63 and 1970, but the use of natural rubber increased only 32 percent, dropping its share of the market from 45 percent to around one-third. 1/ However, all rubber supplies have been sold, including releases from stockpile by the U.S. of from 1 to 8 percent of annual production. 15. FAO projects the demand for elastomers to grow at from 6.2 to 6.6 percent per year in the seventies, rising from consumption in 1964-66 of 6.2, 1/ Agricultural Commodity Projections, 1970-1980, FAO, Rome. ANNEX 15 Page 4 and in 1970 of 8.6 million tons to 15.8 to 16.4 million tons by 1980. The' demand for natural rubber in this expanded demand is projected by FAQ to rise from 3 to between 3.8 and 4.9 million tons by 1980, depending on whether world incomes rise at the trend level or faster, and depending on the rela- tive success of natural rubber in competition with synthetic rubbers. 16. Because a rubber tree takes 5 to 7 years from planting to produc- tion and synthetic rubber plant can be put on stream 3 to 4 years after the investment decision, the natural rubber industry has a very strong influence on the rate of new investment in the synthetic industry. In the past, decisions to build new synthetic plants have generally occurred during periods when new rubber plantings and replantings have lagged or in periods when natural rubber prices have been high. An aggressive policy on the part of Malaysia and Indonesia to expand rubber output to maintain its share of the market would tend to discourage excessive new investments in synthetic production capacity since between them Malaysia and Indonesia produce more than 70% of the supply of natural rubber. The Supply of Rubber 17. By comparison with most other rubber producers, Indonesia's pro- duction increase as projected by FAO, is very modest (Table 1). The FA0 projections can be compared with those of the International Rubber Study Group for 1975 which also project very slow growth for Indonesia compared with Malaysia, Thailand, and for that matter, all other parts of the world (TabLe 2). Table 1: NATURAL RUBBER PRODUCTION AND PROJECTIONS OF FAO FOR SELECTED COUNTRIES 1962-63 1964-66 1970 1980 Average Average (Estimated) Use of "ethrel" excluded included ---------------------000 tons----------------------- World 2,128 2,359 2,950 4,407 4,800 Far East 1,948 2,161 2,704 4,032 4,400 Malaysia 833 951 1,300 2,130 Indonesia 649 694 800 900 Thailand .190 215 285 485 Ceylon 161 120 162 250 Africa 152 163 216 321 340 Latin America 29 34 30 54 60 18. The general pessimism in these studies with respect to Indonesia comes from the very aggressive estate and smallholder rubber production programs of other suppliers, and the slow progress in replanting and other unsatisfactory cultural practices of Indonesia. As a result, Indonesia is projected to fall progressively further and further behind, dropping from 30 percent of the world's production in 1964-66 to 20 percent by 1980. 4 Table 2 Natural FIbber Production Projections, 1968-1975 (1000 long tons) Rate of P R 0 J E C T E D Growth 1968 1969 1970 1971 1972 1973 1974 1975 195557-77 Malysia 1p100 1,218 1,292 1,362 1,435 1,484 1,536 1,580 4.7 Indonesia 738 743 748 768 787 807 826 836 0.9 Thailand 221 220 225 235 245 255 265 275 4.1 Sri Lanka 143 150 157 163 169 175 180 185 3.6 Other Southeast Asia 184 214 229 243 257 270 286 297 4.0 Africa 180 21h 232 245 262 274 286 295 5.7 Latin America 34 36 38 hO h3 __h 45 _5 2.2 Total 2600 21795, 2,921 3,056 398 3,309 3.h24 3_13 3.5 Source: IRG, Loncton. (Internationel ibber Study Group) ANNEX 15 Page 6 Table 3 Production of Estate Rubber in Indonesia, 1959-1969 (000 tons) Department Central Bureau Central Bureau of of o. 2/ 4/ Financi Year Agriculture Statistics Statistics I.R.S.G. - Note 1959 223.3 238.3 224.0 224.0 1960 219.4 219.4 215.6 215.6 1961 229.3 231.4 223.2 223.2 1962 217.3 218.4 210.2 209.3 1963 216.4 217.4 208.9 208.3 1964 231.7 233.7 224.2 223.2 1965 227.8 228.1 219.7 219.7 1966 223.4 216.7 208.4 208.4 1967 100.6 200.6 200.6 194.4 201.0 1968 205.9 305.9 206.2 205.9 206.0 6/ 1969 230.O240.06/ 226.6 228.7 210.1 227.0 1970 238.0 6/ Source: 1/ Department of Agriulture, Facts and Figures 1969, hal, 40. 2/ Central Bureau of Statisticec C.B.S. Statistical Pocketbook of Indonesiaq 1963 dan 1964-1967 dan Indikator Ekonomi (jAn 1 70 u/t Dec 1970). 4/ I.R.S.G. Rubber Statistical Bulletin, 24, 12, September 1970. Department of Agriculture and Nota Keuangan dan Rantjangan Anggaran Pendapatan dan Belandja Negara Tahiti 1971/1972, p. 204. Ministry of Agriculture, Jakarta, February 1972. ANNEX 15 Page 7 19. There is relatively little disagreement among the five different organizaLions reporting estate rubber production in Indonesia (Table 3), but the figures reported for smallholder production diverge significantly (Table 4). The production figures in Appendix Table A-1 are about at the mid-point of this range. There is also some considerable disagreement about the level of exports depending upon reporting organization (Table 5). 20. These differences in reported production and trade make it dif- ficult to be precise about the level of domestic rubber consumption in Indonesia. Crude estimates suggest that it ranged around 25,000 tons during the sixties. Tire and tube production, a major user )f rubber, which declined in the mid-sixties, has increased from around 220,000 tires and 155,000 tubes in 1966, to around 320 thousand tires and 237 thousand tubes in 1970. These are, however, only relatively modest improvements over the 1964 level of 307,150 tires and 223,397 tubes. As domestic tire and tube production increased in the late 1960's, imports of tires and tubes also increased. Continued substantial economic growth can be expected to add significantly to this demand, as transport vehicles are a pressing need, and other domestic uses can also be expected to grow. But the significant factor in demand will remain the export market. Table 4: ESTIMATE OF SMALLHOLDER RUBBER PRODUCTION IN INDONESIA 1965-1969 ('000 tons) Year Dept. of Agriculturei/1 C.B.S./2 I.R.S.G./3 Financial NoteA 1959 478.9 478.9 480.6 1960 403.5 403.5 404.6 1961 470.0 470.0 459.0 1962 480.0 480.0 472,0 1963 490.0 490.0 373.6 1964 500.0 480.6 425.2 1965 510.0 503.5 496.9 1966 536.0 527.9 507.3 1967 530.0 500.3 567.6 538.0* 1968 500.0 531.2 546.1 529.0* 1969 510.0/558 . 544.7 528.8 545.0*/558.2** 1970 571.- 1971 599.0- Source: /1 Department of Agriculture Facts and Figures 1969, p. 40. /2 Central Bureau of Statistics. /3 I.R.S.G. Rubber Statistical Bulletin, 24, 12, September 1970. k Department of Agriculture & Nota Kauangan dan Rantjangan Angaran Pendapatan dan Belandja Negara Tahun 1971/1972. pp. 20* and 210**. /5 Ministry of Agriculture, Jakarta, February 1972. Table 5 Export of Smaliholder and Estate Fubber Indonesia, 1959-1969 (1000 tons) Rubber Export Statistics Central Bureau of Statistics2/ B. T.E. S. Estate Smallholder ToW Estate Smallholder Total I.R.S.G.- Estate Smallholder Totali- 215.4 479.8 692.2 232.7 459.5 692.2 692.2 178.0 378.0 656.0 189.9 366.1 556.0 586.5 223.3 429.7 653.0 236.9 419.8 656.7 677.2 237.0 440.1 677.1 194.0 466.2 660.2 209.7 567.5 777.2 660.2 209.7 482.3 692.0 186.5 359.3 545.8 203.3 379.2 582.5 561.0 194.3 378.3 572.6 224.8 409.8 634.6 240.7 418.3 659.0 627.4 232.9 418.1 651.0 204.3 483.8 688.1 220.5 487.3 707.8 694.6 224.8 429.5 654.3 238.0 441.8 679.8 690.0 246.8 465.3 712.1 194.8 437.7 632.5 211.3 440.2 651.5 738.0 209.7 431.7 641.4 196.8 522.9 719.7 230.1 539.8 769.9 729.0 216.9 511.1 723.0 179.1 576.0 755.1 715.9 1/ Facts and Figues 1969, hal. 40. Bogor & B.P.S. Statistik Ekspor Karet Indonesia 1968. 2/ £ Statistics. 3 Statistics, Economiic 1ýdicators (Jarnary 1970-December 1970) ET/ 1 Bulletin, 24, 12y September 1970. Nomic Studies, VI, 2, (July 1970), pp. 120-135. ANNEX 15 Page 9 21. The world demand for rubber is substantial and the critical factor determining Indonesia's share in it will be her ability to compete as a low cost supplier, not only with Malaysia, but also with synthetic rubber pro- ducers. Natural rubber producers have not kept pace with sharply rising demand for rubber and have gradually lost ground to synthetic rubber. Post- war increases in synthetic production have been triggered largely by the failure of the natural rubber industry to expand. As already indicated, an aggressive expansion policy on the part of Indonesia and Malaysia could do much to discourage new investments in synthetic capacity. The technology is available to produce natural rubber competitively. 22. To the extent that yield and area statistics can be relied on-- since smallholder production dominates, but is very inadequately recorded-- Indonesia has roughly maintained the level of production through area in- creases with an annual linear trend of over 5 percent from 1951 to 1971 compensating for a 2.8 percent annual trend decline in yields (Appendix Table A-1). 23. There is general agreement that Indonesia has a natural environment which should give it a comparative advantage in rubber production. The factors which could sharply boost yields are relatively well known--modern planting materials, fertilizers, herbicides, stimulants, tapping techniques and leguminous ground covers. The NFS concludes that it is in this area and not in extension of the planted area that significant production increases should come. 1/ It thus projects the estate and smallholder rubber area to grow from 1970 to 1980 as follows: Table 6: NFS PROJECTED RUBBER PLANTED AREA ('000 HECTARES) Estates Smallholders Young Mature 1969/70 144 317 1,700 1974 93.5 370.5 1,900 1980 93.5 370.5 2,000 24. With such an area planted to rubber (2,370 thousand hectares), and there seem to be few constraints on area expansion since this would be outside Java, even a maintenance of the 1966-70 yield level (390 kilograms or 860 lb.) should produce a level of output which would exceed that projected by FAO for 1980. 1/ NFS, pp. 208-210. ANNEX 15 Page 10 Table 7: AVERAGE RUBBER YIELDS OF ESTATES AND SMALLHOLDERS Estates Smallholders -----------kg------------- 1953-57 645 378 1958-62 548 315 1963-67 581 338 Source: Five-Year Plan, Vol. 2-A. 25. Improvement of estate rubbar yields can be anticipated,over this decade, but their effect on boosting total output depends on what happens to the yield of smallholder rubber. A yield increase on the estates of 100 kg/ha would add 37,000 tons of additional rubber, but a yield increase of only 25 kg on the smallholder area would add 50,000 tons of production. Conversely any drop in smallholder yields would offset yield increases on estate lands. 26. An aggressive rubber program should provide considerably more rubber in Indonesia.than present studies project. At the very least, better planting materials, some extension, and some improvement in marketing and processing facilities for smallholders should produce significant improve- ments. Because of the low opportunity cost of smallholder labor, improve- ments in the collection, proceEiing and marketing of smallholder rubber and a reduction of the official and unofficial taxes levied on it, should make it a low cost product in the world market. 27. Because of the active policies of Malaysia and other countries, the future level of Indonesian exports will hinge to a very great extent on how competitive its rubber is. The mission considers a larger increase in exports than that projected by FAO and the IRSG quite possible with an active smallholder program. Production, Processing and Marketing Improvements 28. An overhaul and expansion of the smallholders extension service is obviously an urgent necessity. Improved clonal material, much of it originating from within the Indonesian estate sector, is in most cases unavailable ta the smallholders, who have been replanting with unselected seed from their own or neighbors plots. The incidence of overdense stands, profuse weed growth, poor tapping standards and primitive processing methods, could be reduced if extension advice of good quality, together with minimal credit for inputs and replanting are more readily available. However, the problems of moving from seedling rubber to improved clones on a meaningful scale cannot await the establishment of such an improved and expanded extension service. The cheapest way of fffecting the change would be to provide budwood from tested clones to bud-graft seedling rubber. In the ANNEX 15 Page 11 absence of reliable bud-grafting services, distribution of budded stumps could be initiated in areas close to estate nurseries. In more remote areas, provision of improved planting material could probably be facilitated by the use of green stump buddings-a recent technological innovation which has been successfully applied in other countries. 29. There is a marked lack of information concerning the condition of smallholder rubber, but sufficient is known to allow the planning of an appropriate development strategy. A large proportion of the smallholder rubber, particularly in the more accessible areas in Northern Sumatra, is exhausted or nearly exhausted, in that the remaining tapping panels have very thin, low-yielding bark. However, there are significant areas of sleeping rubber which has not been opened for tapping and of rubber with some 4-5 years of tapping life on virgin high panels. Although these latter categories of rubber have been maintained badly by any standard, it is very likely that economic responses to ethrel stimulant could be obtained. Strategy should be to exploit such rubber to exhaustion, over some 4-5 years, while replanting adjacent already exhausted areas. Over this period of time, application of fertilizer would generally be unnecessary. In Malaysia and Sri Lanka, use of etirel is considered economic where an additional annual yield of 100-150 kg/ha is obtained. In the categories of rubber mentioned above, yield increases of 200-400 kg/ha should be attained. 30. In the major rubber smallholder areas the next steps should be: (a) Survey and identification of majjr areas where ethrel could be applied for the next 4-5 years. (b) Investigations into concentration of ethrel, place of application on the tree,' frequency of application, and timing of application, all in relation to yield responses with smallholder rubber. (c) Provision of finance and advice for ethrel application and for improved cultural practices. (d) Organization of a replanting scheme in adjacent areas. (e) In addition to ethrel application, immediate and economic yield improvements could be obtained through control of the inter-line undergrowth and the use of correct tapping techniques. 31. Over-dependence of the smallholder on one crop, particularly where this is an export crop subject to considerable price fluctuations, is to be avoided, and the mission therefore wishes to stress the importance of increasing the area and diversifying the range of subsidiary crops on all rubber smallholdings, at the same time that improvements are being affected in the condition of the main crop. The research and extension services ANNEX 15 Page 12 should demonstrate the production of these subsidiary crops to the small- holders with emphasis on crops which can be inter-cropped with rubber during the first 2-3 years of immaturity. 32. In remote areas where marketing conditions are especially difficult (the average smallholder is estimated to receive less than 30% of the f.o.b. value of his crop) the encouragement of collective processing and marketing of ribbed smoked sheet could be especially effective in increasing the net return to the smallholder. Low cost group processing centers should be established in suitable locations. Following the example of private opera- tors in Jambi, with attention to cleanliness, smallholder rubber can readily be processed into RSS 2 and 3 grades. 33. In the more accessible areas, and particularly in the neighborhood of recently constructed block rubber factories designed to process small- holders' crops, the assistance should lean more toward providing a price incentive to encourage the growers to improve the quality of slab or of air dried sheet, delivered to the factories or towards developing an effi- cient latex purchasing and delivery system. The present system whereby block rubber processors invest heavily in machinery for cleaning unneces- sarily dirty and adulterated rubber is extremely wasteful. The processors should be urged to provide extension services from their factories to the growers aimed at improving raw material quality. Both growers and processors would benefit from such action. 34. In addition, substantial areas, perhaps of the order of 250,000 ha over the next five years, should be newly planted to rubber. Part of this expansion would permit an increase in the size of individual small- holdings in existing rubber growing regions to a more economic level and part would be brought about by opening up forested land in other regions, particularly in the main regions of transmigration potential in Sumatra and Kalimantan, where rubber is one of the most suitable crops in respect to climatic and soil conditions. The development of nucleus estates, owned by foreign and local planting enterprises, which provide a market for the outgrowers rubber and at the same time serve an important demon- stration and extension role, should be seriously cousidered in these areas. The use of private sector expertise under long-term management contracts on public sector nucleus estates could prove to be a highly practical arrangement. C. Coconuts Background 35. Although there are various statistical series on coconuts (Copra) production, they all point to an essentially stagnating industry (Tables 8 and 9). The coconut area has been reported at between 1.5 and 1.6 million hectares since the late 1950's, and production between 1.0 and 1.4 million ANNEX 1 Page 13 Table 8 Area, Production and Average Yield of Coconuts (Kelana): Indonesia (1955-1971) (Copra Equivalent) Smallholders Estates Total Average Average Average Year Area Prod. Yield Area Prod. Yield Area Prod. Yield ('000 ha) ('000 mt) (Quha) ('000 ha) ('000 mt) uha ('000 ha) ('000 mt) (Qu/ha) Average 1951-54 1,264.7 998.0 7,89 25.3 17.34 6.98 1,290.0 1,015.3 7.87 1955 1,514.1 1,039.1 6.86 23.2 16.2 6.98 1,537.3 1,055.3 6.86 1956 1,545.1 997.7 6.46 24.7 20.1 8.14 1,569.8 1,017.8 6.48 1937 1,597.0 1,092.7 6.84 25.0 19.8 7.92 1,622.0 1,112.5 6.86 1958 1,651.2 1,777.7 6.53 15.9 7.2 4.53 1,667.1 1,084.9 6.51 1959 1,561.6 1,267.8 6.12 16.2 7.0 4.32 1,577.8 1,274.8 8.08 1960 1,609.8 1,239.0 7.70 15.8 8.8 5.57 1,628.6 1,247.9 7.66 1961 1,593.5 1,360.3 8.54 19.8 8.8 4.44 1,613.3 1,369.3 8.49 1962 1,583.8 1,386.8 8.76 20.8 66.6 3.17 1,604.6 1,393.4 8.68 1963 1,533.2 1,379.0 8.85 20.7 77.7 3.72 1,573.9 1,386.7 8.81 1964 1,558.1 1,182.4 7.59 1965 1,539.9 1,235.2 8.02 1966 1,479.3 1,127.8 7.64 1967 1,471.6 1,094.2 7.44 1968 1,590.4 1,131.2 7.11 N/A 1969 1,625.1 1,211.5 7.89 1970- 1,684.0 1,280.0 7.60 1/ 1971- 1,700.0 1,300.0 7.65 Source: Central Bureau of Statistics and Directorate of Estates. 1/ Estimated. Table 9 Estim.te of Productior of Copra Equivalent, Indonesia (1959-1969) ('000 tons) _r Department ofAr Central Bureau of Statistics BapDenas Financial Nota 1959 1,267.8 1,274.8 1,267.8 1960 1,239.1 1,2h8.0 1,239.1 1961 1,360.6 1,369.3* 1,360.5- 1962 1,386.9 1,393.h 1,386.8 1963 1,379.0 1,379.0* 1964 1,193 - 1,182.4 1965 219 - 1, 235.2* 1966 350 3 1,12948** 1967 1, 248.4 1,09h.2* 1,09h.1 1,248.0 1968 1,275.0 1,1312* 1,131.2 1,275.0 1969 1,300.0 1,281.5* 1,281.5 1,220.0 Source: 1/ Department of Agriculture, Facts and Figures 1969, pp. 49.1-50.1. 2/ Central Bureau of Statistics: * Production oi' Copra equivalent, state and smallholder. * Production of åopra equivalent, smallholder. 3/ Business News, XV, 2046 (18 January 1971). Dejartment of Agriculture and Nota Keuangan dan Rantjangan Aiggaran Pendapatan dan Belandja Negara, 1971/72, p. 204. ANNEX 15 Page 15 tons. Virtually all the coconut area is in the hands of smallholders. The government has apparently undertaken a program to rejuvenate the coconut industry. Efforts to encourage government estates (which have only about 17,800 trees) to replant apparently resulted in little success, but 43,800 hectares of smallholder area was reported replanted in 1970/71, and plans call for replanting 848,000 hectares. Although yields apparently rose from the mid-fifties to the early sixties, they appear to be now at about the level of the early 1950's. 36. The sizable increase in copra exports reported for 1968-70 (Table 10) may simply be the result of generally stabilized economic and social cr.litions since the mid-sixties, but it may also be evidence that the upward trend in the reported coconut area and production since 1967 is real. 37. The importance of the coconut industry, no less than the rice industry, lies in the fact that it is a large industry almost exclusively in the hands of smallholders. Coconut trees are in all provinces with large concentrations in Java, Riau, South Sumatra, and North Sulawesi. In addition to the roughly $30 to $35 million in foreign exchange from copra, coconut products are the main source of income to a large number of farmers, especially poorer ones, not only from fruit products, but also the construction products it provides--fibers, platiug, weaving and roofing material, and as a source of sugar. To the extent that it is a stagnating industry, an important segment of farm income is also stagnating. Table 10: EXPORT OF COPRA FROM INDONESIA 1959-1969 ('000 tons) /1 /2 Year Dept. of Agriculture- Central Bureau of Statistics-- 1959 133.3 139.4 1960 170.4 167.8 1961 238.5 247.4 1962 109.8 109.8 1963 108.5 108.4 1964 175.5 149.5 1965 123.8 123.8 1966 119.4 119.4 1967 112.2 113.9 1968 155.4 217.0 1969 200.0 157.0 1970 - 168.6 Source: /1 Department of Agriculture, Facts and Figures, 1969, pp. 49.1-50.1. /2 B.P.S. 1959 - 1964 Statistical Pocketbook of Indonesia, 1963 & 1964 - 1967. 1965 - 1969 Business News, XV, 2046, (18 January 1971). ANNEX 15 Page 16 Government Programs 38. Numerous FAO and ECAFE study groups have proposed improvements in the coconut industry, but little has materialized directly from these efforts. Bilateral agreements between Indonesia and the Philippines--the major producer to whom Indonesia has lost ground since World War II--were signed in 1963 and again in July 1971. The former agreement was not successful 1/, and the latter, which permits the processing of Indonesian copra in the Philippines, in part only legitimatizes the 50-60,000 tons of Indonesian copra the Asian Coconut Community estimates is regularly smuggled into the Philippines from Sangir and Talaud islands, and from Sabu. 39. Repeated general admonitions to improve extension, credit, in- frastructure, processing and marketing facilities were followed by a more precise "Suggested Policy for Improvement of Copra Trade" by the Ministry of Trade on August 31, 1971. The short-run proposals include: 1. Reduction of the Copra Rehabilitation tax from Rp. 4 to Rp. 1 per kg. 2. Reduction of the cess from Rp. 5 to Rp. 2.51 per Kg. 3. Elimination of the Surveyor's report (Rp. 150/ton). 4. Transportation subsidies and construction of Copra ware- houses by the government. 5. Farmer incentives in the form of hoes, knives, etc. 6. Proposed revision of the "London Contract" and "Uniliver terms" to improve copra trade. In the longer run it is proposed that Indonesia 1. Should close out copra exports, permitting oil exports only. 2. Reduce domestic consumption of coconut oil by substituting tallow and other vegetable oils. 3. Encourage investment in facilities for production of coconut oil and desiccated coconut, and the processing of shells and oil. 1/ Study on Regiotal Plan Harmonization and Economic Co-operetion in Coconut and Coconut Products, Bangkok, October 1968, p. 147. ANNEX 15 Page 17 40. Of these, the first three proposals seem especially warranted. Improvement of transport and credit seem more important than subsidies, unless these subsidies are clearly related to planned transport improvements. Some phasing out of copra by exporting oil is already taking place with a new coconut oil plant in North Sulawesi and this should continue, but because it is economically warranted and not for its own sake. 41. Phasing out domestic consumption of coconut oil by substituting animal fats and other oils seems a long way off. The sharp rise in the prices of coconuts in rural markets since 1968 (71 percent) and in coconut oil (27 percent) indicates a strong domestic demand for these products which is likely to continue well into the future. Domestic sources of alternatives are limited, and other uses and exports are likely to compete with domestic demand. Palm oil for instance, the logical alternative, will be in strong demand as an export earner. The rise in the real wholesale price of copra since 1967 reinforces this indication of a significant domestic demand (Table 11). 42. The NFS projected no increase in the coconut area, but recommended a number of cultural practices which have improved yields significantly in other countries. It concluded that "analysis for a wide range of coconut prices, showed that fertilizer use is highly profitable even at low prices for coconuts". 43. The government is already undertaking to eradicate Sexara tree- hoppers, in the Sangri and Talaud islands which is delaying replanting programs. Further efforts in the direction of improving disease control and other cultural practices for coconuts seem needed. Attention should be given to the dwarf coconut which is reported by Westenberg in North Sumatra to have good production characteristics. Recommendations to Improve Production 44. It would seem, therefore that sound economic as well as larger income distribution and employment of objectives would justify more emphasis on this large smallholder crop than it has received. This should take the form first of new planting materials, since 60 percent of the area is reported to have trees over 40 years old, and permitting better access to credit and fertilizer presently restricted to rice. 45. The average yield per hectare of 700 kg of copra, which represents only 25 to 30 nuts per tree per year is very low. Nevertheless, possibilities exist to improve coconut cultivation, In the short-term, it is necessary to improve the quality of copra which could increase the income of farmers by 40% without disturbing market prices for coconut oil; this includes improve- ments in marketing methods to ensure that the farmer receives the benefit of the improvements he makes. In addition, improvement of inter-cropping with soil tilling, weed and pest control and use of fertilizers would benefit coconuts and intercropped products. A 50% increase of coconut production has Table 11 Esttta, Consurption of Gopra 24uivalent per Gapita, Consumption Expenditure per Capita, Real Ubolesale Price of Copra in Indonea, 1_960-1969 2/ Estimated Consumption of ro ra Real Wholesale Price of Copra - Equivalent per CapitaConsumption National Deflator Cost Department of Financial Erpenditure", Income Deflator Price of Year Agriculture 0.B.S. ota per Caita per apiua - Index of Food Living Index -------------- kgT -------------Rp) ----------- (Rp/kg) --------- 1960 11.343 11.370 3,306 3,611 6.55 6.55 1961 11.643 11.550 3,h83 3,730 4.66 5.35 1962 124963 12.963 3i643 3,716 3.98 4.75 1963 12-.605 12.606 3,h23 3,55 5.83 6.78 1964 9.861 10,006 3,369 3,953 4.01 4.95 1965 10.646 10.514 3,368 3,5h7 3.S6 h.h0 1966 11,360 9.331 3,239 3,556 4659 5.26 1967 10.240 6.835 10.222 3,442 3525 6.90 8.24 1968 9.845 8.039 9.304 3,485 3,671 6.86 8.69 1969 9.135 9.645 9.117 3,611 3,804 8.72 9.6 Source: 1/ Tables 9, 10 and population figures from Central Bureau of Statis«tics. 2/ Central Bureau of Statistics. ANNEX 15 .½ Page 19 been reached in practical experience abroad. (Long-term trials at Central Research Station at Kasaragod, India). Since coconuts are grown on a wide range of soils with different fertility, any fertilizer program should concentrate in the first instance on poorer soils. There are considerable areas of rich alluvial clays where fertilizer would be uneconomic but where improved drainage may be required. 46. In the long run, replanting/new planting with high-yielding coco- nuts can easily treble the present average yield. Until hybrid material is available, planting material should be produced following mass selection among tall palms around each village. Selection should be done on the basis of both yield per palm and weight of copra per nut. This latter character has high heritability and concent:ation on this character should lead to selection progress. Nursery selection on the basis of vigor should also be carried out, as this technique has been shown in W. Africa to reduce the non-bearing period from 8 to 6 years and to increase yield by about 25%. A further reduction in the non-productive period could be obtained through the use of the polybag nursery technique. 47. The eventual aim should be to produce tall 1warf hybrid material from seed gardens planted with selected tall and dwarf progenies. Material produced in this way could yield 3 tons of copra per ha. At the same time as laying down seed gardens with F1 material, trials of the same F1 material should be set up for comparison of progeny performance. It will then be possible to eliminate the unsatisfactory progenies from the seed gardens. Dwr-f material exists on Nias Island and could also be obtained from abroad. Suitable tall material also exists in Indonesia. Early action to set up these seed gardens in several locations is necessary. Consideration should also be given to production of commercial hybrid material using pollen collected from selected dwarf palms on female inflorescences of selected tall parent palms., Coconut breeding has been neglected and a country-wide program should start as soon as possible and should be integrated with the UNDP support for rehabilitation of the coconut industry. 48. In addition to improved husbandry techniques and selection of high grade planting material much more attention should be given to inter- cropping possibilities during the first 3-4 years of planting and after 25 years when ground shade lessens. Cocoa is a promising inter-crop at this latter stage. Livestock pilot projects should also be implemented making full use of the information available from Sri Lanka. 49. Improvement of coconut cultivation is not difficult but requires funds, trained personnel and continuity. Since coconuts were never a sig- nificant estate crop, there is no history in Indonesia of research and extension, and farmers were not able to borrow estate technology. As a consequence, a special effort is called for in this area. An important area for initial rehabilitation and replanting efforts is North Sulawesi where there are 190,000 ha of coconuts with an average density as low as 80 palms/ha and average palm age is 50 years. ANNEX 15 Page 20 Demand Conditions 50. Export earnings from coconut products are significant, and although competitive, the world demand for copra and coconut oil is moderately good. A serious export oriented effort by Indonesia could substantially improve export earnings from these products. The rapid increase in domestic prices for coconuts and oil since 1968 indicates a strong and growing demand which is not being satisfied by domestic production. With the improved production practices mentioned above, smallholder incomes could be improved significantly even if larger production resulted in lower prices. Because of the signifi- cance of coconuts to the mass of small, and often very poor, farmers, much stronger efforts are nieded to improve smallholder production and marketing. D. Palm Oil and Kernels Background 51. By contrast with most crops, palm oil and kernel production is largely an estate crop. Of the roughly 121,100 hectares of oil palm reported in Indonesia in 1969, 118,164 were reported to be in PNP's and private estates. 1/ 1/ National Fertilizer Study, Appendix II, Tables 11-31. ANNEX 15 Page 21 Table 12: INDONESIA: AREA, YIELD AND PRODUCTION OF PALM OIL AND KERNELS Average Yield Area Production ('000 tons) QQu/ha Year 1,000 ha Oil Kernel Oil Kernel Average 1950-54 92.7 144.6 37.0 15.60 3.99 1955 101.4 165.8 41.9 16.35 4.13 1956 103.0 165.9 41.4 16.01 4.02 1957 103.0 160.2 39.5 15.55 3.83 1958 104.5 147.7 35.3 14.13 3.38 1959 102.9 137.5 33.2 13.36 3.23 1960 104.3 141.2 33.1 13.54 3.17 1961 104.9 145.7 34.3 13.89 3.27 1962 105.0 141.5 32.8 13.48 3.12 1963 105.9 148.3 32.6 14.00 3.08 1964+ 107.7 160.9 34.2 14.94 3.18 1965+ 107.8 156.7 32.5 14.54 3.01 1966+ 110.8 174.4 35.0 15.74 3.16 1967+ 109.7 174.0 34.7 15.86 316 1968+ 113.6 188.2 40.3 16.57 3.55 1969+ /1 121.1 200.3 40.5 15.59 3.44 1970 /T 123.0 131.5 /2 207.0 216.8 /2 43.3 48.7 /2 1971 Ti 126.0 210.0 45,0 /1 Department of Agri,:ulture, January 1972. /2 Director General of Estates, 1972: + = preliminary. Source: Central Bureau of Statistics and Directorate of Estates. 52. As with so many of the statistical series encountered in earlier sections, there are inexplicable differences between statistical series on oil palm production and trade (Tables 12, 13 and 14). The statistics pro- vided the mission by the Indonesian Department of Agriculture in February 1972, which overlapiwith the other series of 1969, do not, obviously, fit any of the series for oil palm, and are especially out of line with the Department of Agriculture's own series. 53. If one assumes that the figures provided by the Department of Agriculture are revisions not of its own series but that of the Central Bureau of Statistics (CBS), then significant progress in area and production has taken place since 1965 - output growing from 157,000 to 210,000 tons. On Government-owned estates (PNP's), production was reported to have increased from 125,582 tons in 1969 to 170,304 tons in 1971, and is projected to grow to 278,000 tons by 1975. This increase in 2roduction has resulted from relatively high prices and rehabilitation programs. 二 Table 14 Pro etion rt Stock ~nd Estimate 45 T- 0 Oil in Indonesia (1959-1969) k IUM tons) Change Production- Year Production in Estimated ä&ort pqm rt Stock Sto ck, Gonsumption 1959 137.5 103.0 34.5 12.6 37.8 19 5, 0 141.2 106.5 34.7 9.5 -3-1 29.9 1961 145.7 117.3 28.4 7.9 -1,5 33.8 1962 111.5 100.1 41.4 15.5 7.6 41.2 1963 148.3 109.7 38.6 12.8 -2.6 14 6 1964 160,9 133-3 27.6 25.9 13,0 29.8 1965 156.7 125.9 30,8 26.8 l.o 6.4 1966 171.4 177.1 -2.7 17.8 -9 4 28.2 1967 174.o 133-3 4o3 30.2 1295 33-3 1968 18m 152.4 27.6 24.5 -5.7 53.0 1969 188.8/ 119.6 69.2 9-9 -14 -6 200 1/ 1970 207 1971 21.0 Source: Caloulated frorn data or ceritral 3,reau of Statistics. 1/ Ministry of Agriculture., JarluarY,1972. ANNEX 15 Page 24 54. Despite the differences in reported trade figures, it is evident that most of the palm oil and kernels produced flow out as exports. Implied domestic consumption has been around 30 to 40,000 tons for the past decade (Table 15). Future Demand, Supply and Trade 55. FAQ projects substantial increases in "fats and oils" and oilcakes and meals exports for Indonesia in the 1970's (Table 15). Table 15: FAQ PROJECTION OF FATS AND OILS AND OILCAKES AND MEALS FOR INDONESIA 1964-66 1970 1980 ------(000 tons)---------- Fats & Oils Production 618 740 994 Consumption 351 437 583 Trade 267 303 411 Oilcakes and Meals Production 266 309 356 Consumption 50 36 66 Trade 216 273 290 However, FAO also projects rapid increase in domestic consumption of fats and oils, an assumption supported by most other projections. FAO assumes an in- come elasticity of demand for vegetable oils of 1.0, and the Indonesian Supply and Demand 1/ (IS&D) study estimates even higher--1.33. Present levels of per capita consumption are estimated by FAO at around 3 kgs. and projected to reach 3.4 kgs. by 1980. 56. The IS&D study, for which neither the demand nor supply sections for fats and oils are complete as yet, has produced tentative conclusions which indicate a considerably stronger domestic demand than that projected by FAO. All major studies dealing with the world demand for fats and oils have concluded that a strong demand for fats and oils exists in developing countries, constrained only by low income levels. 57. If income grows as fast as currently projected by the mission, then it is reasonable to assume that the domestic demand projections of VAO are much too low. Unless vegetable oil production--not only palm oil, but coconut oil and other oils as well--is expanded rapidly in Indonesia, the projected exports of vegetable oil may not materialize. 58. The NFS concluded that "growing conditions in Indonesia, in particular North Sumatra, are among the best in the world. The mission 1/ University of Indonesia, Djakarta, (Indonesia). ANNEX 15 Page 25 generally agrees with this appraisal. The expansion plans of the PNP's call for a 50% increase in area during the Five-Year Plan. The NFS, be- cause of the general world surpluses of vegetable oil projected by 1980, especially palm oil, assumed that the palm oil area would not increase beyond the expansion called for in the Five-Year Plan. 59. The mission, however, feels that there has been a general tendency to underestimate domestic demand. This is reflected in domestic price movements in rural markets for coconuts and coconut oil. The assumption of a more rapid rate of income growth than that assumed by FAO also supports a stronger domestic demand. 60. Furthermore, part of the basis for the assumed modest expansion of Indonesian vegetable oil exports of FAO and the NFS, is the assumption of spectacular export increases by Malaysia. FAQ projects Malaysia's ex- ports of fats and oils to grow from 267,000 tons in 1964-66 to 414,000 tons in 1970 and to 1.8 million tons in 1980. 61. The international market for fats and oils changes sharply from year to year. Prices were high in the early sixties, and fell sharply when Russian sunflower oil exports expanded in 1964-66. Prices then rose sharply in the late sixties when Russian and West African vegetable oil exports declined. This condition is likely to prevail in the future. 62. The natural conditions prevailing in Indonesia, the strong domestic demand, and the possibilities for a fluctuating but still good world demand, all favor strong emphasis on increasing Indonesian palm and coconut oil production. A stronger movement into smallholder production is warranted. As with many of Indonesia's potential exports of agricultural commodities, the future world market will be competitive and prices may be lower than now prevail. Thus competitive prices and the quality of the marketed product will be the major determinants of what share of these markets falls to Indonesia. Production Recommendations 63. The fact that the palm oil industry is completely under the con- trol of large scale producers, which include the most efficient estate enterprises in the country, accounts for the relatively high average yields by Indonesian standards. These enterprises have been able to respond rapidly to the improved market prospects for palm oil, as compared with rubber, that have prevailed during the recent period of economic recovery, and this explains the marked increase in planted area which has taken place during the last three years, largely on old rubber land. 64. The foreign-owned private estates, many of which are closely associated. with similar enterprises in other countries where considerable technological progress has been made in breeding, fertilization, pest and disease control, general field management, harvesting and processing of the ANNEX 15 Page 26 crop, are able to tap this expertise and, should they wish, to transform their old properties into model estates which could be the equal of those anywhere in the world. 65. Similarly, the Government-owned oil palm estates (PNP's) are undergoing a substantial program of rehabilitatiin and new planting which is already having a beneficial effect on yield and output. The practice of engaging consultants to advise on matters of agricultural and financial policy, and to prepare plans and monitor progress, should be a most valuable aid to strengthening the management of these properties. The quality of this management has improved considerably since 1968 when the PNP's were reorganized along regional lines. The permanent adoption of a planting advisory and inspection system to cover the needs of the PNP's, similar to the system generally used in foreign-owned estates, has much to commend it. The temporary use of resident consultants, who should as much as possible assume an executive or technical assistance role, should be allowed to con- tinue for several more years at least. 66. Experience in other countries has shown that oil palm is a crop which can be grown profitably by smallholders provided they are able to dispose of their fruits to a nearby processing factory. The possibility exists in the neighborhood, of many of the existing and proposed public and private sector oil palm estates for encouraging the cultivation of the crop among surrounding smallholders. Existing prices should be a sufficient inducement to encourage some of the more progressive farmers to diversify into this crop, while minimal extension advice and a supply of reputable seedlings would be additional inducements. 67. Oil palm is a crop particularly well suited to the even-rainfall areas of Sumatra, where the bulk of the crop is presently grown, and of Kalimantan. Large areas of land suitable for further planting exist in both these regions and, as in the case of rubber, there are good prospects for developing nucleus oil palm estates on vacant lands. Foreign investors are known to be interested in such possibilities, and it is up to the Govern- ment to establish the conditions under which such private sector groups should be permitted to operate. 68. The Smallholder Development Project for North Sumatra, recently financed by IDA, is a first step toward improving smallholder rubber production and also provides for smallholder oil palm production in an area where existing PNP processing facilities can be utilized. The IDA credit also finances technical assistance for a management unit within the Ministry of Agriculture which,would carry out the project. It provides financing and technical assistance for planting, replanting and primary processing facilities, as well as making additional land available to smallholders and improving communications and marketing. ANNEX 15 Page 27 E. Coffee Background 69. There are major unexplained differences in both estate and smallholder production depending on the reporting organizations (Tables 16 and 17). Data provided the Mission by the Department of Agriculture in Jakarta in March 1972 for 1969, 1970 and 1971 do not correspond with any of the published series (Table 18). They show a sizeable increase in produc- tion for each of the three years, but are said to refer only to smallholders. Nevertheless they exceed the combined production estimates of smallholders and estates in any other series. 70. Despite these discrepancies, there is clear agreement that estate coffee has stagnated in the past decade while that of smallholders has expanded rapidly. Estate coffee production was recorded in all series at about 19,000 tons in 1959. Only the Department of Agriculture series records it as above that level in 1969, and the other series put it at about 14,000 tons for that year (Table 16). It was reported to have increased 12% to 15,884 tons in 1970. By contrast, smallholder coffee production was generally agreed to be about 65 thousand tons in 1959, and between 143 and 162 thousand tons in 1969 (Table 17). 71. Since domestic coffee consumption is customarily arrived at by deducting exports and stock changes from production, little can be said for the consumption data as an indicator of demand, since neither the production nor export data is consistent over time. ANNEX 15 Page 28 Coffee Table 16: PRODUCTION,1 OF ESTATE COFFEE IN INDONESIA, 1959-1969 ('OO tons) Dept. of Central Bureau Central Bureau /1 /2 /3 /4 Year Agriculture- of Statistics-- of Statistics-- Financial Nota- 1959 19.0 19.0 19.0 1960 13.3 18.3 18.3 1961 13.6 19.1 19.1 1962. 12.3 12.6 12.6 1963 18.4 19.3 19.4 1964 7.4 7.6 7.5 1965 14.1 19.8 19.8 1966 19.5 12.3 12.4 1967 13.8 18.8 19.0 19.0 1968 13.4 13.4 13.6 13.0 1969 20.9 14.0 14.1 14.0 Source: /1 Department of Agriculture, Facts and Figures 1969, hal. 48a. /2 C. B. S. /3 C. B. S. 1959 - 1962 Statistical Pocketbook of Indonesia, 1963, 1964 - 1947. 1963 - 1969 Economic Indicators (Jan. 1970 - Dec. 1970). /4 Dept. of Agriculture and Nota: Keuangan dan Rantjangan Anggaran Pendapatan dan Beland;a Negara Tahun 1971/1972. p. 204. ANNEX 15 Page 29 Table 17: ESTIMATE OF SUA.LLHOLDER COFFEE PRODUCTION IN INDONESIA 1959 - 1969 ('000 tons) /1u e C /3 Year Dept. of Agriculture C.B.S.- Financial Nota- 1959 65.3 65.2 - 1960 77.9 77.9 - 1961 78.3 88.4 - 1962 99.1 99.1 - 1963 126.6 126.5 1964 80.2 111.5 1965 91.5 120.7 1966 116.9 129.2 1967 117.0 140.3 143.0* 1968 90.0 143.7 143.0* 1969 146.9/161.5 /4 143.3 143.0*/161.5** 1970 169.0 /4 1971 178.0 /4 Source: /1 Department of Agriculture, Facts and Figures 1969, p. 48a. /2 Central Bureau of Statistics. /3 Dept. of Agriculture and Nota Kauangan dan Rantjangan Anggaran Pendapatan dan Belandja Negara Tahun 1971/1972 p. 204* p. 210**. /4 Source (2) Department of Agriculture (March 1972). Table 18 Prodction, Export, EstimRted Stocks and Estimated Consumption of Coffee in Indonesia (1959-1969) (000 -tons) Changes Produ ction- Estimated in Estimated Year Production Exports ExporAs Stocks Stocks Consumtion 1)59 84.2 38.4 45.8 29.3 1960 96.2 41.4 54.8 53.2 23.9 30.9 1961 107.5 65.5 42.0 56.8 3.6 38.3 1962 111.6 57.2 54.5 43.3 -13.6 68.0 1963 145.9 80.9 65.0 62.1 18.9 46.2 1964 119.0 62.4 56.6 56.8 -5.3 62.0 1965 140.5 108.3 32.2 62.3 5.5 26.7 1966 1,1.6 97.5 44.1 72.8 10.5 33.6 1967 159.3 133.1 26.2 68.0 -4.8 30.9 1968 157.3 84.7 72.6 64.5 -3.5 76.1 1969 157.4/162.0 104.3 53.1 50.4 -14.0 67.2 1970 157.4/169.0 97.4 2/ 1971 157.4/178.0 Source: Derived from Gentral Bureau of Statistics data. 1/ Estimated stocks = stocks of estate coffee + estimated stocks of smallholder coffee. 2 Source (2) Department of Agriculture, Jakarta, January, 1972 (smallholders only). ANNEX 15 Page 31 72. Data on area, yield and production of coffee of the Directorate General of Estates indicates a strong upward movement in the smallholder area from 132,500 ha in 1954 to 315,000 ha in 1969/70. Smallholder yields are estimated to have risen from around 400 kg/ha in 1958 to almost 600 kg/ha in 1969/70. 1/ By contrast, the estate coffee area has fallen from 49,400 ha in 1954 to 38,796 in 1969. Yields on estates rose from around 300 kg/ha in 1961 to around 548 in 1967, but were lower in 1968 and 1969 (400 and 428 kg/ha, respectively). 2/ 73. The higher yields of smallholder trees are explained by a larger proportion of Robusta coffee and younger trees than on estates. 3/ Also the estimated number of trees on sTvallholder areas is larger than on estates. Although tree density on smallholder plots is considered to be greater than optimal (between 1,600 and 2,000 as opposed to 1,100), apparently the loss in yield is less than proportional to the increase in total tree density. Future Supply, Demand and Trade 74. FAO, in its most recent projects for the Second Development Decade, projects the world demand for coffee to increase at an annual rate of 2.5% during the 1970's, the same rate projected by the International Coffee Council. 4/ Demand in the planned economies is projected to grow fastest (7%), with the developing countries second (3.4%), and the developed countries slowest (1.7%). For Indonesia it projects domestic demand to grow from 60,000 tons in 1970 to about 80,000 tons in 1980. Indonesia's exports are projected to grow from 81,000 tons in 1970 to 108 000 tons in 1980, with Indonesia dropping from 3% of world trade in 1964-66 to 2.7% in 1980. 5/ 75. FAO concludes that, as in the past, the world coffee market will be one in which surpluses exist and the International Coffee Agreement will strongly .affect trade, despite the difficulties experienced by Brazil in the past several years. 6/ 1/ Long-Term Projections of Indonesia's Import Demand and Export Potential for Selected Agricultural Commodities, (IS&D) University of Indonesia, Djakarta, (in process). 2/ Ibid, p. 154. 3/ Ibid, p. 145. 4/ Agricultural Commodity Projections 1970-1980, FAO, Rome, 1971 Vol. 1, p. 220. 5/ Ibid, p. 222. 6/ Ibid, p. 218. A14NEX 15 Page 32 76. The NFS adopts a strongly pessimistic view about the medium- and long-term demand for coffee which it sees as "one of potential over-supply, at lower prices than now prevail." It concludes that, "any further expansion of Indonesia's coffee area should therefore be discouraged", with the slow growing demand being satisfied by productivity improvements in the estate and smallholder sectors. It does not foresee rapid use of fertilizer on the smallholder sector "because of insufficient organizational and credit facilities." 1/ It, therefore, projects no essential change in the area of estates or smallholders between 1970 and 1980--39,000 ha for estates and 315,000 ha for smallholders. Almost all of the estate coffee area is in East and Central Java (34,408 and 3,624 ha respectively out of a 1970 total of 38,796). 77. The Indonesian Supply and Demand Study (IS&D), not yet in final form, has produced two sets of domestic demand projections and a number of alternative supply projections. It projects demand with an income elasticity of 0.2 (the r_-O assumption for Indonesia) and of 0.5 (the FAO assumption for "Asia"). Table 19: IS&D PROJECTED GROWTH OF COFFEE DEMAND AND PRODUCTION Year Demand Production 0.2 0.5 High Medium Low --------------Thousand Tons---------------- 1969 48.3 58.8 1975 71.4 75.4 231.5 231.5 231.5 1980 83.7 92.1 290.7 271.8 253.9 1935 99.0 113.8 356.0 314.7 275.9 The differences in the supply projections are due to differences in the assumed rate of increase in the smallholders coffee area. 78. The estate coffee area is assumed to remain constant at 33,000 ha, with yields rising from 502 kg/ha in 1975 to 582 kg/ha in 1985, which would produce estate coffee production of 16.6, 17.9 and 19.2 thousand tons in 1975, 1980 and 1985. For smallholders, however, three rates of growth for area are assumed with the "high" rate being an extrapolation of the past trend, the medium rate being half the past trend, and the "low" rate holding the area constant at 328,700 ha. Yields were assumed the same in each case (Table 20). 1/ National Fertilizer Study Indonesia, NFS, Annex II, p. 222. ANNEX 15 Page 33 Table 20: IS&D ASSUMED GROWTH OF SMALLHOLDER COFFEE AREA AND YIELD Year Area Assumptions Yield Assumptions High Medium Low 1,000 Hectares Kg./Ha. 1975 328.7 328.7 328.7 654 1980 379.9 335.7 328.7 718 1985 431.2 378.7 328.7 781 Source: IS&D, pp. 147-149. 79. It seems highly unlikely that the strong upward movement in small- holder coffee plantings will be halted even if that were to be government policy, and as is argued in the NFS. The rate of increase might very well be cut to half what it has been (the medium projections of the IS&D), however, if demand and marketing conditions in Indonesia do not improve. 80. The implications are that regardless of domestic and world developments, Indonesia will have supplies of coffee for export considerably above the 108,000 tons projected by FAO--between 160 to 244,000 tons. Exports reached or exceeded the FAG projected level in 1965 and 1967, and were close to it in a number of recent years, not counting smuggling. 81. The strong upward movement of the area and yield of smallholder coffee despite little, if any, assistance, and a reportedly low price to producers compared with the f.o.b. price, suggests that the incentive to produce is strong despite quota restrictions and past difficulties in nonquota marketing abroad (which are said to account for a large stock accumulation). 1/ 82. Given the problems of the ICA and the potentials for exploiting spot markets in the world, the strongly pessimistic view of export prospects for coffee in existing studies does not seem warranted. The IS&D study's domestic demand projections also support the conclusion that demand prospects for coffee are better than have been assumed. Therefore an area expansion by smallholders to about 335,000 hectares by 1980 (the medium IS&D projection) is expected. Total production by 1980 will probably be between the IS&D studies "low" to "medium" levels of 254 to 272 thousand tons. Domestic consumption can be expected to be between 84 and 92 thousand tons. This would provide exportable supplies of between 160 and 180 thousand tons. Whether these quantities will be exported or not depends upon improvements in the quality of coffee exported and on improvements and reduced costs in the marketing of smallholder coffee. 1/ The farm gate price of coffee is reported to be about Rp 125,000 per ton compared with a price to the export of Rp 303,000. Of the Rp 178,000 difference, Rp 61,400 is made up of eight kinds of levies and charges by the government. ANNEX 15 Page 34 Production Improvement Recommendations 83. Estate plantations are old, some of them being planted between 1910 and 1920 and even before 1900. If reasonably well managed, they can always produce a few hundred kgs of coffee per hectare, but if they were replanted with new high-yielding varieties, they would produce 3 to 4 times more. Already some PNP estates have replanted part of the very old coffee plantations with cocoa. 84. Most of the smallholder coffee plantations are young, but due to lack of upkeep require renovation. Robusta type coffee has a natural tendency to have many stems, the best way to improve yields is to adopt quinquennial multiple stem pruning with a sap-drawer, to obtain trees whose stems form a funnel-shaped goblet producing abundant fruit on the primary branches. Advantage should be taken of this cutting-back to renovate the coffee trees. The vacancies should be filled in to regulate the shade density which must be reduced or suppressed, and weeding and manuring should be carried out. 85. If the demand for coffee proves to be insufficient to support these improvements on the existing area, some of the coffee land can be diversified to other crops. Clove is one of the crops which can replace coffee below 300 m, and coconut at sea level, in addition to cocoa on the best latosols. Arabica plantations can be maintained above 1,000 to 1,200 m if Indian selections resistant to hemileia are grown. These varieties have already been introduced in some collection plots in Indonesia. F. Cloves Background 86. The area and production of cloves have expanded sharply over the past two decades (Table 21). Table 21: INDONESIA: CLOVE AREA AND PRODUCTION Year Area Production Exports 1,000 Ha 1,000 Tons 1,000 Tons 1952 10.9 3.0 2.7 1956 21.1 4.0 12.7 1960 36.6 6.0 6.8 1961 46.8 7.1 9.0 1962 48.6 6.7 5.2 1963 50.7 7.9 5.4 1964 61.5 12.6 8.9 1965 67.1 13.7 4.6 1966 69.2 17.4 6.9 1967 72.3 11.3 13.8 1968 74.0 13.1 10.2 1969 76.2 12.0 7.4 1970 82.9 12.5 8.8 ANNEX 15 Page 35 Clove production, originally limited to the Moluccos, is now widely diffused in Indonesia with large areas in Sumatra especially North Sumatra (14,000 ha), Java (17,600 ha in Central and West Java) and North Sulawesi (11,058 ha). 87. The expansion in area and decline in production since 1967 is partly explained by the especially rapid increase in new seedling planting which has not yet begun to produce. It may also be due,in part to a four-year production cycle involving one heavy crop, two medium and one poor crop. 1/ 88. Indonesia is the world's largest importer of cloves, with most of imported and domestic clove production going into the Kretec cigarette industry. The rapid expansion of demand for Kretec cigarettes, and the imposition in 1968 by the Government of Zanzibar of a L 55 per 100 lb export tax on cloves, caused the cost of imported cloves, and thus domestic cloves, to shoot up dramatically. Table 22: CLOVE PRICES TO KRETEC MANUFACTURERS Year Local Imported -----------Rp/kg---------- 1966 60-65 65 1967 120-125 120-125 1968 225-900 300-900 1969 1,400-2,100 1,650-2,700 1970 1,600-1,900 2,100-2,400 1971 1,600-1,900 1,900-2,400 89. In response to this price increase, and an earlier increase in 1953/54, a rapid increase in clove seedling planting took place. As a consequence there are a large number of trees in the 16 to 20 year age category of heavy and growing full production, and a large number of new trees which will commence first production. An additional 20,000 ha were to be added in 1971 to the 83,000 ha existing in 1970. Despite disease problems and poor cultivation practices, the 1971 study of clove Production, Processing and Marketing in Indonesia concludes that, even with a major allowance for possible overestimation and production problems, Indonesia will be self-sufficient in clove production by 1977, or before, producing about 24,500 tons. Future Demand and Supply 90. The clove study correctly notes the major implications following from these developments. Unstable production from year to year, a charac- teristic of the crop, will produce possible shortages and surpluses for 1/ Report on Clove Production, Marketing and Processing in Indonesia, (W. B. Williams) December 1971, paragraph 5. ANNEX 15 Page 36 export in the future. To effectively exploit these surpluses, better storage and grading facilities will be required to balance supplies from year to year, and to exploit opportunities in the export market. The Malagasy Republic and Zanzibar can both be expected to respond strongly to competition from Indonesia. 91. High prices and other factors have resulted in too little concern for quality in Indonesia, and considerable adulteration of cloves. Indonesian cloves are usually discounted in foreign markets. Too early picking, too high an oil content, and insufficient drying are the major problems. 92. A rapid increase in Indonesia production can be expected to produce a sharp drop in domestic and world prices. Thus, improvements in processing and marketing may be harder to achieve in the face of falling prices, than were improvements in production in the face of sharply rising prices. 93. It is questionable, however, whether the government should under- take the supply management and storage construction program ("two 5,000 ton godowns as close to Semarang and Surabaja as possible") suggested in the clove study. 1/ If, as the clove study indicates, one cigarette factory "has recently erected buildings covering 6-1/2 hectares as part of a policy of expansion", it should be possible for the private trade to develop and manage the needed storage and processing. If proper drying and grading are as profitable as they would seem, the government should be able to influence the development of such facilities through tax credits or adjustments'. 94. A major factor undoubtedly influencing both too early harvesting and insufficient drying is the current high rate of interest coupled with high current clove prices. The need for cash by smallholders, and the absence of facilities which would provide them with a higher portion of the price of good quality cloves undoubtedly acts as a disincentive to farmers to provide better quality cloves. 95. The NFS study indicates that significant yield improvements can be expected from fertilizer applications, and it and the clove study suggest many cultural improvements which can raise yields. The NFS study is not as optimistic as the clove study that the increase in planted area will produce self-sufficiency, and indicates that yield improvements are necessary. 96. The NFS projects the clove area to rise from 77,100 ha in 1969 to 94,000 ha in 1974 and to 128,000 ha in 1980. Areas of major expansion are North and West Sumatra, Central Java, North Sulawesi and West Irian. If this is indeed the path of expansion, it would seem to call for a more 1/ Paragraph 28. ANNEX 15 Page 37 diffused supply of storage and processing facilities than those indicated in the clove study. Government assistance, if it is needed, should take the form of facilitating the development of storage facilities close to smallholder producing areas. 97. Indonesia should be able to produce 30,000 tons of cloves by 1980. Prices can be expected to be much below current levels, and fluctuating production will undoubtedly produce fluctuating export surpluses. G. Pepper Background 98. At the beginning of World War II Indonesia was the number one exporter of pepper in the world, with 60,000 tons of pepper per year and 80% of the world's pepper supply. Most of the plantations were destroyed during the Japanese occupation and the production decreased drastically, mainly in Lampung where 97.5% of the vines were destroyed. After the war, pepper growing developed again and exports of 42,000 tons were attained in 1959. Production rose to almost 50,000 tons in 1967 and 1968, but dropped sharply to 10 to 15 thousand tons in 1969, 1970 and 1971. 99. The main growing areas are Lampung and Bangka, a small amount being grown in West Kalimantan near the Sarawak border. In Lampung, the most important growing area, 30,000 ha are under cultivation, with 20,000 ha in North Lampung with 3,000 mm of well-distributed rainfall per year, 5,000 ha in Central Lampung with less rainfall and a pronounced dry season, and 5,000 ha in the mountainous South Lampung. In Bangka 5,000 ha are grown under a yearly rainfall of up to 3,000 mm. 100. Today Indonesia is the third largest exporter after India and Sarawak (Malaysia). Brazil started growing pepper after World War II and its production increased rapidly, but presently it is having trouble with a virus disease. World consumption estimates of 140,000 tons in 1980 compared with present world production of 95,000 tons give scope for production and trade increases in coming years. Recent Production and Trade Experience 101. There are two conflicting production series for recent pepper production, one by the Central Bureau of Statistics and Ministry of Agriculture, the other by the Pepper Marketing Board (Table 23). The forecast of the Directorate of Plantations indicated still different production figures for 1969-71 as follows: 1969 - 5,408 tons 1970 - 10,000 tons 1971 - 11,000 tons ANNEX 15 Page 38 Table 23: PEPPER PRODUCTION AND EXPORTS Year CBS - Ministry of Agriculture Pepper Marketing Board Production Exports Production Exports (originAl pepper) (crude pepper) -('000 tons)---------------------------- 1960 12.2 12.7 - - 1961 13.7 19.1 - - 1962 55.4 11.0 - - 1963 57.9 28.0 - - 1964 46.4 23.2 30.8 23.2 1965 46.5 (11.0)- 12.3 17.4 12.3 1966 36.3 20.7 28.0 20.6 1967 16.5 37.4 49.0 37.2 1968 46.7 24.5 46.7 24.4 1969 50.4 15.3 11.1 14.9 1970 n.a. n.a. 15.7 2.3 1971 n.a. n.a. 15.7 2.3 /1 Reported by U.S. Agricultural Attache as also a CBS figure for this year. 102. Of the two major series, the IS&D study decided that the Pepper .Marketing Board figures more closely matched actual developments in Indonesian pepper production, because they matched the export data better, and because they fit the four-year production cycle of small, medium, peak and small harvests. There is general agreement about the export figures, and that production has been severely damaged by Phythoptera Palmivara Var Piperis disease in the Lampung area. 103. On the basis of the Pepper Marketing Board production figures, and the cost of Living Survey in Jakarta, the IS&D study developed the domestic consumption estimates and projections for p'Apper in Table 24. The figure for 1969 is considerably below the 8.1 thousand tons of consumption one would obtain from a simple moving average subtracting exports from reported production. Table 24: INDON4ESIA'S ESTIMATED PEPPER CONSUMPTION Total Consumption Per Capita Consumption (tons) (kgs.) 1969 5,188 0.0445 Projected: 1975 6,558 0.0474 1980 7,906 0.0503 1985 9,625 0.0536 ANNEX 15 Page 39 104. The NFS indicated a 1970 planted area on 46,408 ha allocated as follows: South Sumatra 7,038 East Java 774 Lampung 33,800 East Kalimantan 800 Bengkulu 1,471 West Kalimantan 136 West Sumatra 746 Sulawesi 320 West Java 1,000 Bali 136 Central Java 67 Nusa Tenggara 19 Various other estimates exist placing the area at about 40,000 ha, almost all of it in Lampung (black pepper) and in Bengkulu (Banka) (white pepper), but they presumably excluded the smaller areas. 105. Conflicting reports exist about the factors causing the sharp decline in recent production. A Workshop on Export Development in Jakarta in 1970 blamed it on disease in Lampung, lack of fertilizer and poor maintenance. The NFS noted that the workshop also cited, "the unsatisfactory financing arrangements of pepper growers are largely responsible for the poor Lampung pepper crop..." and recommended to change Indonesia's export policy in the sense that pepper sales to Singapore be allowed in order to ensure the best prices for the growers." 1/ 106. There is apparently more involved in the declines in reported pepper production than simply disease problems. Wholesale and retail pepper prices in Jakarta were considerably lower in the first six months of 1971 than they were in 1970, which would indicate that the large reported shortfalls in production and exports have not had a significant effect on domestic prices. 107. The NFS cites numerous examples of striking improvements in yields produced through better cultivation practices and use of fertilizer in countries whose production conditions seem no better than those of Indonesia. It would appear also that the current disease problems in Lampung can be corrected. For these reasons, the mission has concluded that a significant recovery in pepper production in Indonesia is possible. The loss in export earnings of about $10 million a year since 1969 warrants special attention being given to both production and marketing of pepper. The world demand for pepper appears to be such that it could absorb Indonesian exports of from 25 to 50 thousand tons, and Indonesian export earnings could be improved if greater care were exercised in quality control. 1/ (NFS, p. 239.) AliiiEX 15 Page 40 Production Recommendations 108. In Lampung which produces black pepper on live poles the ecological conditions are excellent for pepper growing, provided the poor podsolic soils are carefully avoided. But it is grown extensively with cuttings of poor quality, incorrect growing methods, little cultivation, and almost no fertilization. This results in plants which have low resistance to the very contagious Phytophtora fungus infection which has developed during the last 10 years. 109. In Bangka, due to better growing methods and application of some fertilizers, it is possible to produce on poor podsolized soils white pepper, but the nutritional imbalance is reflected by the so-called "yellow disease", which denotes physiological exhaustion of the plants; nematode infestion sometimes considered asithe primary cause, is more a consequence of an imbalanceci nutritional regine. 110. Pepper must be grown with very intensive fertilization and excellent growing methods. These are now incorrectly applied or not used at all because they are considered too expensive. This is mainly due to the fact that the farmer is largely dependent on the "Idjon system" (see ;,arketing and Prices) where cash is advanced against a part of the unharvested crop at extremely unfavorable terms of repayment in kind. The effective interest rate may be as ,high as 500% per annum. The farmer receives only 50 to 60 11p per kg of black pepper, compared with a guaranteed price of Rp 200 fixed by the Pepper Marketing Board. H. Tea Background 111. Before World War II Indonesia was the third largest supplier of tea in the world with an export of 72,000 tons in 1938; during the war half of the plantations were destroyed and replaced by food crops. Since then new plantations have been developed, and in 1971 65,000 ha of the Tea estates and 33,000 ha belonging to smallholders produced respectively 46,000 tons of black tea and 24,000 tons of green tea. Black tea is for export, through a joint marketing office whose director is appointed by the Minister of Agriculture,' green tea is for local consumption. 112. An International Tea Agreement set up in 1970 and extended to 1971 allocated a quota of 35,000 tons per year to Indonesia, but it has not been enforced. Estate production is already above the proposed quota. Since it is not clear what the future of the Tea agreement will be, a rehabilitation program financed by the Bank to improve quality by better picking, yields by fertilization, and renovate the factories is called for. Fertilization improvement requires foliar analysis which is new for tea, but an immediate response to fertilizer applications has largely increased the yields. Since tea requires acid soils, ammonium sulfate is applied ANNEX 15 Page 41 instead of urea used for most of the crops in Indonesia. With a present production cost of 150 Rp/kg of black tea and a sales price of 273 Rp/kg for 1972, very substantial benefits have been obtained by estates. Tea production is relatively a simple crop presenting fewer problems than other crops, and the future of tea estates seems good. 113. Smallholder tea plantations have more problems. Except for some 7,500 ha which are reasonably good, 25,000 ha suffer from lack of upkeep, very poor pruning, food crops grown with tea plants and too intense shade. For these 25,000 ha the average yields of green tea are low despite thtl poor quality produced. During recent years the acreage of smallholder,. tea plantations is said to have decreased. Recent Performance 114. Two alternative views of the development of tea production, consumption and exports are set forth in Tables 25 and 26. The first table results from use of data provided by the Director General of Estates and the Department of Trade, whereas the latter is the result of usin CBS data. The major difference is in smallholder production and domestic tea consumption. Area and yield data are far from satisfactory, Table 27 is the series considered by FAO to be "the most reliable", with preliminary figures provided the mission by the Ministry of Agriculture in Djakarta in February 1972. 115. lable 27 indicates a continuous drop in the tea area of both smallholders and estates from 1955 to 1968, with some increase in 1969. Yields would appear to have increased more rapidly on smallholder areas than on estates. (Smallholders produce green and black tea, whereas the estates produce black tea.) The higher yield of tea on the smallholders area is strongiv disputed by the JFS (p. 227) and by the U.S. Agricultural Attache. Estimates of the total production area of tea range from the 95,000 ha by the U.S. Agricultural Attache, to an average area of 127,900 ha for 1965-67 in the IBRD Review of the World Tea Economy (p. 61). The NFS study assumed a 1970 area of 123,000 ha, whereas the CBS figures used by FAO indicate 110,000 ha in 1968. Domestic Consumption 116. Domestic consumption derived by subtracting exports from produc- tion and accounting for changes in stocks would suggest that there was a general decline in tile 1960's until a sharp upward movement in 1969 and 1970 using the DG of Estates data (Table 25), and a significant increase in consumption using the CBS data (Table 26). Given the uncertainty about smallholder production which provides the major portion of the local supply, tnese estimates should not be taken too seriously. Future Developments 117. After attempting to construct numerous consumption estimates, the IS&D study concluded that the FAO income elasticity of demand of 0.6 and price elasticity of -0.4 Zor tea were probably correct, and projected domestic demand as follows: ANNEX 15 Page 42 Table 25 Production &.Per Capita Consumption for Tea (1960-1970) Arnnu al Production Closing Consumption Year Non-Estate Estate Tota Exports Stock Total Per Capita .. Tt on ----------.---------.-----. -- kg) --- 1960 37,300 46,053 83,958 39,800 2,800 41,358 0,328 1961 36,600 421999 78,599 36,200 2,600 39,799 0,413 1962 37,300 49,527 86,827 32,700 3,300 50,827 0,9516 1963 38,600 39,471 78,071 31,600 4,100 42,371 0,423 1964 23,340 51,118 74,458 36,000 2,700 35,750 o,340 195 34,990 44,816 79,806 36,500 4,000 39,306 0,371 1966 34,550 42,445 76,995 37,400 3,300 36,295 0,335 1967 38.230 37,791 76,021 29,600 2,500 44,921 o4O4 1968 33,120 42,956 76,040 37,100 2,400 36,540 0,321 1969 40.260 41,920 82,180 32,900 2,500 46,780 0,h.01 1970 43,641 42,354 85,995 35,000 2,500 48,475 0,4o Notes: A. Sources: 1/ Directorat General of Estate Agriculture 2/ Directorat General of Estate Agriculture 3/ Department of Trade / Department of Trade B. Export: 1960-1970: No quota. As from 1970: Quota system (35,000 tons) C. Total consumption total production - (exports + annual closing stock). Table 26 --Poction. Export and Consumption of Tea 1960 - 1970 .state. Pro- Sallholders Total Pro- Export Annual Closing Total Con- Per Capita Year duction production duction Stock sumption Consumption (tor) (ton) (ton) (ton) (ton) . (ton) (Kg) 19cl 46,061 37,390 83,451 39,800 2,194 41,457 0.440 1961 43 z19 37,700 80,839 36,200 2,580 42,059 0.436 1962 47,221 37,334 84,555 32,700 3,287 48,568 0.493 19b3 L,634 38,562 77,226 31,600 8,177 37,449 0.372 1964 4.,234 41,150 81,384 36,000 2,733 43,561 0.471 1965 46,952 42,304 89,266 37,000 10,518 41,748 0.395 1995 4L,953 45,000 91,953 37,400 3,427 51,126 0.472 1967 33,304 47,000 80,304 29,600 2,494 43,210 0.435 1968 41,o86 61,000 102,386 37,100 1,405 63,881 0.562 1969 O9,300 62,000 101,500 32,900 3,450 65,150 0.559 1970 40,000 65,000 105,000 35,000 4,000 65,000 0.553 Source: 1900 - 1967 from Central Bureau of Statistics, Djakarta 1988 - 1969 from Government Financial Report 1971. 1970 - preliminary figures. ANNEX 15 Page 44 Table 27--Area, Production and 'Average Yeld of Tea (teh): Indånesia 1955-1969 Small Holdings f Estate 1 Total yec,ar ; -- t Area Prod. Averg- e Are a I Prod. . Aver. t Area I Prod. t Aver. IjOha 100mt S yield 1000ha% 100:t Ield 1000ha C10M0-t 9 yleld 61.7 l 26.7 1 4.13 79.8 8 33-7 4.22 * 144.5 60.4 4.218 1955 64.6 21.9 3.39 '0.3 39,3 4.89 144.9 61.2 4.22 9566 21.9 3.67 76.9 40.5 5.27 136.5 62.4 4.57 196o. 1957 60.1 22.2 3.69' 75.1 44.7 5.95 135.2 6b.9 4.95 1958 60.1 23.6 3.93 74.3 45.7 6.15 134.4 69.3 5.16 1959~ 62.9 32.1 5.10 72.9 - 40.6 5.57 135.4 72.7 5.35 1960 64.4 37.3 5.79, 72.9 i.9 5.75 137.3 79.2 5.77 1961 64.5 36.6 5.67 72.4 40.5 5.59 136.9 77.1 5.63 1962 6 37.3 5.77 72.4 43.1 5.95 137.0 80.4 5.87 1963 60.9 38.6 6.34 73.9 37.0. 5.01 134.8 75.6 5c61 1964 57.1 23.3 4.08 69.4+ 43.9 6.33 126.5 67.2 5.31 1965 55.1 35.0 6.35 67.5+ 41.5+ 6.15 122.6 76.5 6.24 1966 55.1 34.6 6.28 65.2+ ·37.6t 5.77 120.3 72.2 6.00 1967 53.9 38.2 7.09 60.9+ 32.9 5.40 114.8 71.1 6.19 1968 50.1 33. 6.61 60.5+ 39.7+ 6.56 110.6 72.8 6<58 1969 57.1 40. I 7.06 69.4 N/A 10 11, /A '. Area refers to planted area. Production refers to dry green tea. + 64 - 69: Estate data are preliminary. VA : Not available Source: Central Bureau of Statiotics and Directorate of Entate. 1/ Ministry of Agriculture, green tea only (2). 2/ Ministry of Agriculture (2). AtNEX 15 Page 45 Table 28: IS&D PROJECTED TEA DEMAND Year Per canita tea consumption Total consumption (Kg.) (Tons) 1970 0.55 66,123 1975 0.61 84,813 1980 0.66 104,998 1985 0.73 131,261 118. FAO's projections of production, consumption and trade differ considerably, however, because the IS&D study concluded that the CBS data were a more accurate reflection of the tea situation in Indonesia than the DG of Estates' data. As a result, whereas the IS&D study assumes a per capita consumption level of 0.55 in 1970 rising to 0.66 in 1980, FAO assumes a per capita consumption level of 0.4 throughout (Table 29). FAO also assumes a slower income growth rate. They both assume a large and more rapidly growing population than now seems warranted by the 1971 census. Table 29: FAO PROJECTED TEA PRODUCTIO:-, CONSUAPTION AND TRADE 1964-66 1970 1980 Production (1,000 tons) 81 84 103 Consumption Total (1,000 tons) 45 51 64 Per/cap Kgs 0.4 0.4 0.4 Exports (1,000 tons) 32 33 39 119. The IBRD Review of the World Tea Economy projects a pattern of future production, consumption, and exports similar to that of FAO. Table 30: IBRD PROJECTIONS OF INDONESIAN TEA PRODUCTION, CONSU6PTION AND TRADE 1955-57 1965-67 1975 1980 I II I II ----------------------1,000 tons---------------------- Production 60.0 82.4 92.0 92.0 100 100 Consumption 24.3 45.8 62.8 56.8 71.2 65.0 Exports 35.7 30.7 29.2 35.2 28.8 35.0 Source: Review of the World Tea Economy, p. 64. Assumption I assumes that per capita consumption will rise from 0.43 to 0.47 kgs between 1965-67 and 1975 and 1980. ANNEX 15 Page 46 120. The NFS also assumes a constant per capita level of consumption, and "a yearly total production increase of 3 percent will be absorbed on the world and the home market together, without great influence on prices." 1/ 121. The gross discrepancy in production, and thus consumption, figures does not permit any easy solution to the problem of large discre- pancies in magnitude. The weight of acceptance seems to fall on the side of the lower DG of Estates figures, but only the IS&D study has apparently attempted a serious demand analysis, and it accepted the higher CBS figures after considering both. The sharp reported rise in domestic consumption in both series, corresponds with the first significant increase in Indonesian incomes in a decade. Whether this increase in incomes elicited increased smallholder production as the CBS figures suggest, or whether smallholder production has simply been under-reported in the past should not be too difficult to determine. However, it is critical to estimating future Indonesian domestic and foreign tea trade. Future Demand and Supply 122. With a 28 percent increase in world demand projected by FAO for the 1970's, and at least the possibility of a more significant demand increase in Indonesia, the prospects for tea production do not look as limited as some of the above studies indicate. As the IBRD Review of the World Tea Economy notes "prices received in the Netherlands and at the London auctions are lower than for any other tea-producing country, a fact which attests to the poor quality of Indonesian tea." 2/ 123. The NFS concludes that the tea "supply could easily be increased", but expects only the planned area increase of 4,000 hectares on North Sumatra and West Java estates. Although this may be true, the lack of output growth on estates in recent years, and the fact that most of the domestic demand is supplied by smallholders, where little has as yet been done to im- prove production, suggest that Indonesia may face greater difficulties than anticipated in satisfying a growing domestic demand and taking advantage of export possibilities, unless more is done to improve smallholder production. 124. Government, with IDA financing has recently initiated a project for rehabilitating smallholder and small private tea estate production with a view toward ensuring sufficient production to satisfy the growing domestic demand and maintain its present level of exports. 1/ (NFS, p. 225). 2/ Review of the World Tea Economy, IBRD, p. 63.