Report No. 9352-TA Tanzania Economic Report Towards Sustainable Development in the 1 990s (In Two Volumes) Volume Il: Background Papers June 11, 1991 Country Operations Division Southern Africa Department FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Currency Unit = Shilling Average Fiscal Year Exchange Rates (July/June) (Tsh/US$) 1965/66 1966/67 1967/68 1968/69 1969/70 1970/71 1971/72 7.143 7.143 7.143 7.143 7.143 7.143 7.143 1972/73 1973174 1974/75 1975/76 1976/77 1977/78 19734/79 7.143 7.014 7.143 7.976 8.368 8.059 7.873 1979/80 1980/81 1981/82 1982/83 1983/84 1984185 1985/86 8.221 8.207 8.647 9.762 12.645 17.875 18.646 1986/87 1987/88 1988/89 1989/90 1990/91 51.163 82.286 119.424 173.505 201.801 Average Calendar Year Exchange Rates (Tsh/US$) 1966 1967 1968 1969 1970 1971 1972 7.143 7.143 7.143 7.143 7.143 7.143 7.143 1973 1974 1975 1976 1977 1978 1979 7.021 7.135 7.367 8.377 8.289 7.712 8.217 1980 1981 1982 1983 1984 1985 1986 8.197 8.284 9.203 11.143 15.292 17.472 32.698 1987 1988 1989 1990 Jtlie 1991 64.260 99.292 143.377 195.056 230.0 FOR OFMICLAL USE ONLY COUNTRY: Tanzania TrILE: Tanzania Economic Report - Towrds Sustainable Development in the 1990s Volume I: Main Report Volume II: Background Papers REGION: Africa SECTOR: Country Economic REPIORT TYPE ŁASS5TIObN LANGUGE 9352-TA ERA Official Use English DATE: June 11, 1991 ABSTRACT: This report analyzes the achievements of Tanzania's economic reform program since the mid-1980s and identifies constraints to continued economic recovery. Building on the directions set forth in the Government of Tanzania's Economic Recovery Program and Economic and Social Action Program, the report considers the key elements of a strategy for attaining sustainable economic growth and renewing social progress - reducing poverty - within Tanzania. Chapter 1 assesses the performance of the economy over three distinct policy regimes during the past two to three decades, thus providing a framework for evaluating, particularly, the results of change and growth over the past five years. Chapter 2 analyzes the period of reform from the perspective of the impact on and welfare gained by a rage of economic agents. Chapter 3 focuses on the impact of public sector performance on macroeconomic stability and discusses the financial and public sector reforms required to achieve greater macro stability. Chapters 4 and 5 analyze the impact of recent adjustment efforts and the agenda for further reform of the agriculture and industrial sectors, respectively. The requirements of parastatal reform are discussed, as are important infrastructure / real sector linkages. Chapter 6 assesses the stalled social progress of the past decade and suggests a strategy for restoring momentum toward social goals. Chapter 7 consolidates the report's analysis and recommendations, provides a framework for projections and argues that a deepening and a hastening of the reform process is required if Tanzania is to achieve its longer term objectives for the economy. This document has a restricted distribution and may be used by recipients only in the perforrnance of their official duties. Its contents may not otherwise be disclosed withown World Bank authorization. Towards Sustainable eMent Ir the 12ft Volume I: Main Report Chapters Acknowledgement Executive Summary . . . . . . . . . . . . . . . . . . i 1. Performance of the Economy . . . . . . . . . . . . . . . 1 2. Impact of Reform on Key Economic Agents . . . . . . . . . . 17 3. Public Sector Performance and Macroeconomic Stability . . . . . . 31 4. The Agricultural Sector: Performance and Linkages . . . . . . . 59 5. Industry and Small-Scale Enterprise Development . . . . . . . . 89 6. Accelerating Social Progress . . . . . . . . . . . . . . . 111 7. Sustainable Growth in the 1990s and Beyond: Scenarios and Conclusions . . . . . . . . . . . . . 131 Volume nI: Background Papers Papers 1. The Tanzanian Recovery, 1983-89 . . . . . . . . . . . . . . . 1 2. National Income Estimates in Tanzania: Methodological and Empirical Issues . . . . . . . . . . . . . 37 3. Private Sector Development in Tanzania: Obstacles and Opportunities . . . . . . . . . . . . . . . . . . 99 4. Enhancing Aid Effectiveness in Tanzania . . . . . . . . . . . . 119 Statistical Appendix . . . . . . . . . . . . . . . . . . 133 Bibliography .................. .. 143 ABBREVIATIONS AIDS Acquired Immune Deficiency Syndrome BIS Basic Industrial Strategy BOP balance of payments BOS Bureau of Statistics BOT Bank of Tanzania CCM Chama Cha Mapinduzi CDC Commonwealth Development Corporation CMS Central Medical Stores CPI consumer price index CRDB Cooperative and Rural Development Bank DRC domestic resource cost ERP Economic Recovery Program ERR economic rate of return ESAP Economic and Social Action Program FIS formal sector ind'vitAal survey GDP gross domestic product GDY gross domestic income ICD Institute of Curriculum Development IFC International Finance Corporation IFS International Financial Statistics IMF International Monetary Fund KILIMO Ministry of Agriculture MDB Marketing De9lopment Bureau MOP Ministry of Finance NBC National Bank of Commerce NDC National Development Corporation NGO non-governmental organization NMC National Milling Corporation ODA official development assistance OGL Open General License PEM protein energy malnutrition PER Public Expenditure Review PEs public enterprises PSAP Priority Social Action Program PSM public sector management QRs quantitative restrictions SAR Staff Appraisal Report SDR Special Drawing Right SGR strategic grain reserves SMEs small scale enterprise TAC Tanzania Audit Corporation TANU Tanzania African National Union TCMB Tanzania Coffee, Cotton or Cashew Marketing Board TDFL Tanzania Development Finance Co., Ltd. TIB Tanzania Investment Bank TOT terms of trade VAT value added tax VCRs value-cost ratios THE TANZANIAN RECOVERY, 1983-1989 Tanzania Economic Report Background Paper #1 THE TANZANIAN RECOVERY, 1983-1989 CONTENTS 1. Money Creation and the Banking System: 1983-88 ........................................ 3 2. Explaining the Perforvance of the Real Economy since 1983 ........................... 8 3. Peasant Supply Response, 1977-88 ......................................... 13 4. Trends in Living St?ndards, 1983-89 ......................................... 20 S. Conclusion ......................................... 34 AZA-MAN KECOVER 1. Money Creation and :he Banking Systan: 1983- In centraly controlled economnes, although financial institutions usually have the nomenclature of those found in market economies their functions can be radically different. Because of these differences, both the process of money creation and the consequences of financial liberalization are quite unlike those of a market economy. In Tanzania during the period 1983-88 there was a major reform program which ostensibly included fiscal retrenchment and financial liberalization. We argue that neither of these can be well understood without taking into account the characteristics of the centrally controlled banking system. As a result of fiscal retrenchment the cumulative domestic financing requirement over the five years was only Tsh 14.8 billion, equal to 6.8% of government expenditure. Despite this, the money supply increased by over Tsh 60 billion in the same period. This appears to imply that either 'outside' money is highly geared, each shilling increasing the money supply by a factor of four, or that the money supply is scarcely related to the budget. We will argue that neither of these inferences would be correct; once allowance is made for the nature of the Tanzanian financial system there is oneto- one correspondence between the money supply and the budget. The Tanzania financial liberalization took the form of increases in interest rates on bank deposits and loans. Such an interest rate increase is commonly seen as the essence of financial liberalization. Indeed, financial repression is commonly defined in terms of negative real interest rates which 'reduce the attractiveness of holding claims on the domestic banking system" (McKinnon (1989) p.137). We will argue that given the Tanzanian financial system, raising nominal interest rates does not achieve any of the gains associated with liberalization and is indeed counterproductive. These gains can be realized only if financial liberalization is coordinated with fiscal reform. 1.1 Money Creation For our purposes the most important aspect of a centrally controlled banking system is that lending is determmed by government priorities rather than market criteria. Within such a system it is natural for the government to rank its own expenditures ahead of those of private agents. As a result, in Tanzania the commercial banks (which are publicly owned) have lent predominantly to the public sector 1' Within the public sector lending has often been determined by need rather than by viability. That is, loans are made to cover the operating deficits of public agencies rather than being earmarked to asset formation. For example, around two-thirds of all bank lending outstanding at the end of 1987 was to cover the operating deficits of the crop marketing parastatals. During 1988 the government officially took over 40% of these liabilities, yet by the end of the year the indebtedness of the crop marketing parastatals to the banking system was higher than at the end of 1987 2'. As of the end of 1988 only 13% of the ast of tho commercial banking sytmm were chims on tho 'private sector", (Banl of Tanania (1988), Table 12). But dho private sector in this definition included the cooperative societies which are better regard as public entitis. Dank of Tamania (1988), Table 14, and unpublisd infonmation frxm the Dank of Tanzania. 4 Thus, although the commercial banks take in deposits from private agents, they do not use them to finance lending to private agents. In the past, therefore, the banking system has not performed any intra-private financial intermediation function. There is no pyramid of credit and hence no banking multiplier. Because virtually all bank deposits have as their counterpart borrowing by public agencies, "outside money" has a radically different definition from that in a market economy. Instead of being only currency outside the Central Bank, outside money includes almost all bank deposits. Conversely, currency held by the commercial banks, which vould normally be a liability of the government, is not part of outside money since in Tanzania the balance sheets of the commercial banks should be consolidated with that of the government. In short, there is no inside money, and outside money is constituted by private claims on the commercial banking system plus private holdigs of currency. Further, a function of the Bank of Tanzania is to lend to the cormmercial banking system to finance that part of its expenditure function which is not covered by Its debt sales function. Between the end of 1983 and the end of 1988 net lending by the Central Bank to the commercial banks was Tsh 34.5 biillion ' whereas 'ending by the commercial banks to the government and parastatals was Tsh 50.9 biil.on v. Hence, most of the subsidies channelled through commnercial bank "lending" appear 'o have been financed by the Central Bank. Thus, although superficially Tanzania appears to have a conventional financial system: commercial banks regulated by a Central Bank, the system in the past often was a device by which the public secor financed expenditures by debt sales and printing money. We now show that once the financial system is corsolidated into the government accounts, there is a close correspondence between the public sector deficit and the increase in the money supply. In turn, this increase appears to be closely related to the observed rate of inflation. Bank of Tanzania (1988), Tables 11 and 12. Bank of Taania (1988), Tablo 12. 5 Table 1: The Budget, the Money Supply and Infladon 1983 or 1984 or 1985 or 1986 o 1987 or CuauLtivo 1983/4 1984/5 198516 1986/7 1987/8 Total Totd Government (1) Expenditure (FY) 23.9 26.7 33.2 55.5 76.9 216.2 Commeritl bank lending: to publo admnstaion (CY) (2) 0 0 0.1 0 0 0.1 to aricultund mukedng 0.8 0.9 2.2 4.7 30.1 29.3 (CY)(2) Tota (FY) 0.8 1.6 - 1.2 12.7 23.9 37.8 Comnmemial bank inst 0.7 0.7 1.4 1.5 4.2 8.5 payments (CY),3) Conmunial bank inteest 0.7 1.0 1.4 2.8 4.7 10.6 payments (O)M_ _ __ _ _ _ _ _ __ _ | .djused governuent 25.4 29.3 33.4 71.0 105.5 264.6 (apenditure (FYXI) ,l _ . . . Revene: Total Revenue |FY)(1) 14.2 18.0 20.8 31.4 46.4 130.8 Otherltems(FY)(1) 1.9 1.7 3.2 11.3 9.4 27.5 Interest on Bank lending to public adminisation and agricultural marketing (FY)(5) 0.1 0.3 0.6 4.2 2.5 3.3 Financing: Foreign fiuancing (FY)(1) 2.5 2.7 2.9 8.6 17.7 34.4 Non-bank borrowing (PY)(I) 0.9 1.1 1.3 2.9 2.5 8.7 Totnl Residual Monetary fulancing (FY)(6) 5.8 5.5 4.6 17.0 27.0 59.9 Actual inc. in money supply (CY)(7) 1.1 8.8 11.4 16.1 23.4 60.8 pY = Piscal Year, CY = Calendar Year (1) Tanzian Economic Trends, Vol 1(4) Tablo 7. (2) Bank of Tanzania (1988), Table 14. (3) Bank of Tanzania (198b), Table 15, for time and saving deposits; iterest rates on ssvings deposits from Table 9. (4) Adjusted govenmet exponditum equals total governmetd expenditu plus conmercial bank lending to public administration and agricultuml nmaketing, plus commercial bank intered payments on tim and savings deposits. Adjustments to a fiscal year bis are simple avegos of calendar yea. (5) Iraat paynts arm calculated s the cumulative commeial bank landing to public adminisation and the agnculural maketing boards imes the upper range s"iot term lending rate from Table 9 of tho Bank of Tanzania (1988). 6 (6) R"sidua mmzly rm.is is aduMed expditur minms thi thre rwe ims at two rmXing IXms. ( From BDnk of Tamnla (1988), Table 5, the componn being curreny in olmuuaon outside th bank, demand deposit, time dqei d svins depoi. The adjustment of the government accounts through their consolidation with those of the banking system makes an enormous difference to the residual monetary financing item. Table 1 attempts this consolidation by adjusting the government accoun.es in three ways. First, as we have argued, commercial bank 'lending' to the governmnert and other public entities to finance recurrent expenditure should be treated as part of expenditure. Of this lending the main components can be readily identified, namely, lending to the government itself and to the agricultural marketing boards. These loans, which account for 68% of all lending to the public sector during the period, clearly financed recurrent expenditure. However, some of the remaining loans to the public sector are for asset acquisition and therefore should not be treated as a disguised item of exrpndlture. Since we do not have information on how much of the lending to industrial parastata., falls into this category we treat it as though it were all genuine lending. Second, commercial bank payments of interest on time and savings deposits should be regarded as domestic debt service payments and so z.- uld also be treated as part of expenditure. Tlird, these two adjustments double count one component of expenditure, namely, bank lending to public agencies to meet their debt servicing, for this is merely an internal bookkeeping transaction within the public sector. We have therefore added back to govermnent revenue the debt service payments of public agencies. The details of these adjustments are set out in the notes to Table 1. Residual monetary financing is then the difference between adjusted expenditure and revenue less other means of financing (almost entirely foreign). These three adjustments substantially change the estimate of residual monetary financing: summing over the fiscal years 1983/84 to 1987/88, the residual is increased from Tsh 14.8 billion to Tsh 59.9 billion. Monetary financing of a budget deficit leads either to a depletion of the foreign exchange reserves or to an increase in the money supply. The Government of Tanzania has been in no position to buy back its currency with foreign exchange reserves since by 1983 its reserves were largely depleted. Hence, we should expect a fairly close correspondence between monetary financing and the increase in outside money. As shown in Table I this is indeed the case. Over the five years the increase in the money supply is Tsh 60.8 billion compared with the Tsh 59.9 billion residual monetary financing item. The year to year correspondence is less close, presumably reflecting errors introduced by the conversion from Walendar to fiscal year for some items, minor fluctuations in reserves and the incompleteness of our consolidation of the banking system into the government accounts. The increase in the money supply, which superficially appears to be so much larger than the fiscal deficit, is therefore in fact well explained by it. Finally, we demonstrate that there is a reasonable correspondence between the money supply (defined as above) and the price level. In Table 2 we construct a measure of the transactions demand for money. If the velocity of circulation is constant then the growth rate of transactions demand follows directly from the inflation rate (i) and the growth rate of real income (g), being (1 + i)(1 + g) - 1. This assumption of a constant velocity appears justified. Two 7 factors on which the velocity of circulation normally depends are the rate of inflation and per capita real inWome. During the period both of these were fairly constant N. The exception to the constancy of velocity is 1984. However, dudng 1984 the relaxation of price controls substantially removed what had been acute shortages of consumer goods. As we have previously shown (Bevan, et al. (1987, 1987a, 1989 and 1990)) such shortages induce an increase in cash balances. Consequently, the return to market clearing enables agents to reduce cash balances to normal levels and during this process of adjustment the velocity of circulation rises. Table 2 shows that during 1984 there was indeed a marked increase in the velocity of circulation: the money supply grew by far less than the constant velocity prediction of transactions demand. Thereafter, there is a close correspondence between the grow*th of the money supply and our constant-velocity prediction of the growth in money demand f. Hence, if we regard real growth in the economy as independent of monetary policy, changes in the money supply in excess of this growth are associated with equal changes In the price level. This begs ue question as to the direction of causality. Potentially, either inflation or the budget deficit could be exogenous. Inflationary expectations could drive up the price level; this in turn might increase the budget deficit (if revenues are less fully indexed than expenditures). This increase in the deficit might sufficiendy increase the money supply to accommodate the increase in money demand. Alternatively, the budget deficit might be considered as exogenous, the resulting changes in the money supply causing changes in the price level. However, in Tanzania despite quite rapid inflation there is no indexation in the labor market, wage revisions being infrequent. Hence, goverment expenditure is not heavily influenced by inertial inflation. It therefore seems more piausible to regard the budget deficit as determining inflation rather than the other way around. Sinc ml CDP growth wu broadly in n wih population growth, per capita ralnm was almot constant. Haen, even if th income elasticity of the demand for qme is substantially differat fom unity, the real GDP gtowth during ths period should be expected to mime the demand for money propotiontely. It should be noted that no p-cise analysis betwen the money supply and prices can be donr in Tanzania because of the absenco of a sie for GDP at maket prices, the only price seies being for usban conmer which is dominated by food. 8 Table 2: Predicted and Actual Growth of Money Demand (Increase during year %) 1984 1985 1986 1987 1988 Money Supply (1) 3.7 29.0 29.2 32.0 36.7 Predicted Money Demand (2) 40.8 36.8 36.5 35.0 33.5 Demand Deposits -18.6 24.7 39.4 29.0 49.2 Time and savings Deposits 13.1 41.5 6.1 32.8 26.4 Currency Plus Demand Deposits -0.1 23.0 41.7 31.6 38.9 Source: Bank of Tanzania (1988), Tables 5 and 25. (1) Currency outside the banks plus bank deposits (2) See text: constant velocity of circulation assumed. Inflation rates are the national consumer price index, real GDP is from unpublished Bureau of Statistics data. 2. Explaining the Performance of the Real Economy since 1983 Table 3A summarizes the performance of the real economy since 1983. It therefore covers the period of important policy changes and these changes therefore dictate the disaggregation which is appropriate. First, because there was a series of large devaluations we need to distinguish those parts of the economy which stood to benefit, namely, the sectors producing internationally tradable goods: manufacturing and agricultural export crops. Second, because there was a large trade liberalization, we need to distinguish between those tradable activities which had benefitted from protection and those which had not. The former group is broadly captured by the manufacturing sector. However, there are substantial variations between manufacturing firms which cannot be incorporated into our macroeconomic analysis. The latter group is primarily that part of agriculture which produces readily exportable products. Third, there were substantial changes in government recurrent expenditure. The sector which most closely approximates to this expenditure is public administration. Fourth, there were large changes in both public and private expenditures on investment. This we identify through purchases of capital goods, distinguishing between those which are tradable and those which are non-tradable. The latter is to a large extent captured by construction which is our proxy. The above categories leave a residual, namely, those activities which produce output (other than capital goods) which is not readily tradable internationally. Within this group we distinguish between agriculture (food) and other activities (such as transport). Food is included in this category because very high transportation costs make it difficult for Tanzania to export it. In the absence of protection or subsidies on food, Tanzania would probably be self-sufficient. In fact, at the start of our period Dar es Salaam was largely fed by food imports because domestic food prices were kept below world levels. By the end of our period Tanzania had restored self- sufficiency (indeed there ware some subsidized exports). Hence, to the extent that the increase in food production displaced imports, it should be regarded as a tradable activity. However, as we show in Section 4, only a small fraction of extra food production is accounted for by this 9 import-displacement so it Is a better approximation to regard food production as a non-tradable activity. Table 3A: The Performance of the Economy: Output 1983-88 l____________________ 1976 1983 1984 1985 1986 1987 1988 GDP (1) 108.3 100 103.4 106.1 109.3 113.5 118.2 GDP per capita 111.0 100 100.6 100.4 100.6 101.6 103.0 Exports (2) 180.3 100 104.6 106.3 101.3 105.6 104.3 Manufacturing value 133.7 100 102.7 98.7 94.7 98.7 104.0 added (1) _ Public administration 66.1 100 100.2 102.1 91.0 91.5 94.4 etc. (1) Capital goods: 103.6 100 164.3 221.0 200.0 192.9 168.9 Traded (3) Non-traded (4) 161.0 100 120.2 109.5 128.4 131.3 137.0 Non-tradables: agriculture (all) (1) 91.2 100 104.0 110.3 116.6 121.7 127.2 Service etc. (5) 95.0 100 103.0 104.6 112.7 116.7 121.5 GDP per capita excluding public administration (1) 119.4 100 101.2 101.3 104.0 105.3 106.9 (1) Bureau of Statistics, unpublished data. (2) From Foreign Trade Statistics, 1987, Tables 28 and 29, for 1983-86. 1987 and 1988 from Unit Value Indices of Table 1 and dollar value of agricultural exports and imports (CET Vol 1 (4) Table 6). (3) Source National Accounts of Tanzania, 1976-87, Tables 13 and 15, "equipment". (4) Non-tradable capital goods is construction value added. Source as in note (1). (5) Source as note (1); sectors are wholesale and retail trade, restaurants and hotels; transport and communication; finance, insurance, real estate and business services. All are value added. In addition to the above policy changes the performance of the economy was affected by four favorable external shocks, two of which were temporary and two of which are probably permanent. First, there was a temporary improvement in the terms of trade due to the 10 coffee boom of 1986. Second, there was a probably permanent fail in oil prices. Third, because the policy refoms restored donor confidence there was a resumption in aid flows. Finally, in response to the own-funded imports scheme there was a large repatriation of previous capital outflows. Between them, the above policy changes and external shocks must largely account for changes in economic performance. As shown by Table 3A, since 1983 there has been a modest increase In per capita GDP partially reversing the rapid decline from 1976 to 1983. There are good reasons for believing that in the absence of policy changes this decline would have continued. As argued by Bevan at al. (1987), rural shortages had led to a process of cumulative contraction: farmers reacted to shortages by reducing the production of export crops and this in turn, by reducing the amount of foreign exchange available for imports, further reduced availability of consumer goods in rural areas. Hence the counterfcual cannot be a stagnant level of output: without the policy reform the rapid decline of output would have continued. Hence, the improvement is more impressive than the aggregate GDP growth indicates. Taking the sectors in order, the most remarkable feature about the export sector is that there appears to have been only a minor recovery in output despite the devaluations. On a per capita basis, the volume of exports has actually fallen below its 1983 level. Output in the protected tradable sector (manufactring) declines until 1986 (as a result of the negative effect and those liberalizations as the output of protected investments) and then recovers rapidly (at 4.8% per annum). This is a major reversal of trend: in the previous decade manufacturing value-added had declined by 3.4% per annum. The public administration sector is virtually the mirror image of manufacturing: rapid decline from 1985 reversing a decade of rapid growth. The Table shows a massive increase in investment since 1983. The two components, imported and non-traded capital goods, behave very differently. There is an early surge in imports of capital goods which gradually tapers off, being replaced by growth in the construction sector. This is indeed the classic feature of an investment boom: it is much easier to increase capital goods imports quicldy than it is to expand construction activities. Construction (in the Table listed as the non-traded capital goods) is the most rapidly growing sector of the economy. Finally, output in both of the non-tradable sectors (foed and services) grows rapidly, indeed more rapidly than all other sectors of the economy except for construction. Having described the changes in output the task is now to explain them. T'here are three possible reasons for changes in output: changes in weather conditions (about which we will have little to say), the return to market-relaxing conditions (which would affect output even if prices did not change) and relative price changes which reflect policy changes. In Table 3B we show relative price series for the various sectors, the numeraire being the shilling cost of $1 of imports. This is simply the shilling cost of imports in a year divided by its dollar cost. The relative price of exports fob is therefore an export unit value index. Tbis peaks in 1986 due to the coffee boom (an effect which incidentally is not identified in the published export unit value series). After that the index declines to a level slightly below that in 1983. However, this is not the price index appropriate for producers. The producer price index shows a substantial fall (40%) during the period. Thus, much of the incentives for export production provided by the devaluations were not transmitted to producers. Since in the short term there is litte scope for transferring resources from the urban economy into export agriculture, the chief potential for expansion of agricultural exports is resource transfers from the rest of the agricultural sector. However, our price series suggest that by 1988 the price of 11 exports had actually fallen relative to agricultural prices generally. Hence, although the fob price of exports rose by almost 50% relative to agricultural output in general, producers had an incentive to switch out of export crops into other agricultural activities. Thus the gains from devaluation have accrued not to peasants but to the marketing agencies that have benefitted from higher fob prices without having to pay higher prices themselves. Although the devaluation has not benefitted export producers, the trade liberalization has massively eroded the profitability of the protected tradable sector. The relative price of manufactures has fallen to one-third of its 1983 level and this presumably accounts for the contraction in output between 1983 and 1986. However, the trade liberalization has facilitated the expansion of those parts of the manufacturing sector which were not net beneficiaries of the system of protection. 12 Table 3B: Prices Relative to the Domestie Currency Cost of $1 of Imports 1983 1984 1985 1986 1987 1988 Exports: FOB Price (1) 100 89.8 105.2 115.1 79.8 93.0 Producer Price (2) 100 119.4 130.4 85.7 64.3 60.6 Manufacturing value-added 100 112.6 115.5 66.1 39.6 33.4 (3 ) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Capital Goods: Non-traded 100 104.8 125.1 69.7 42.9 40.5 (3) Non-tradables: 100 115.2 141.2 78.9 61.7 65.9 agricultu.- (all) (3) _ _ Services etc.(3) 100 116.0 131.2 66.5 50.6 44.8 Memorandum Item: shilling 100 105.3 120.1 279.3 491.9 650.1 cost of $1 (index) (4) _I__ (1) From TET Vol. 1(4) Table 6(a) a price index for the main six agricultural exports can be derived in dollars. This has been converted into shillings at the mean export exchange rate. (2) From Table X in Section 3.2 (3) Bureau of Statisdcs, implicit GDP deflators derived from unpublished National Accounts data in current and in 1976 prices. Construction is used for non-traded capital goods. (4) The shilling cost of $1 of imports is derived from the Economic Survey 1988, Tables 10 and 12. Indeed, our output series already shows rapid growth of the manufacturing sector post-1986. Even this must under-state the true recovery of manufacturing because the series misses the effects on output of the entry of new firms. Hence, the positive effects of trade liberalization on manufacturing are already outweighing its negative effects. The relative price of construction changes in a way similar to that for other non-agricultural non-tradables. For these two groups prices rise relative to manufacturing but fall relative to agriculture. To summarize the relative price changes, the price of agricuiture (both food and export crops) rises relative both to manufacturing and to non-food non-tradable. This is fully consistent with the changes in output described in Table 3A: agriculture expanded all during the period. However, our price series suggest that by 1988, the price of exports had actually fallen relative to non-tradable agriculture (food). The favorable shocks of foreign aid and own-funded imports increased income in terms of tradables and some of this was spent on non-tradables, tending to increase their relative price. 13 To conclude, first, the performance of the economy is explicable in terms of relative price changes induced by the policy changes made during the reform period. It is encouraging that the reforms have halted the decline, that the non-tradable sectors have expanded considerably, and that manufacturing is now growing rapidly. However, the export potential of the economy remains to be realized. We will show in Section 3 that this is substantial. 3. Peasant Supply Response, 1977-88 Since the 1984 reforms, there appears to have been a substantial increase in the volume of peasant sales of crops. This is more remarkable because it reverses the trend of the pre-reform period. Between 1977 and 1985, the volume of official sales of export crops per capita of the rural population declined at around 5% per annum whereas between 1985 and 1988 it expanded by 3% per annum. The recent growth in the volume of food sales has been far more dramatic, particularly through non-official channels: combining official and non-official channels, food sales per capita of the rural population have increased by 79% between 1983/84 and 1987/88. Overall the volume of crop sales per capita of the rural population has expanded at an annual rate of 10% during this four- year period. This expansion is not explained by relative prices. Since the reforms, the real producer price of crops has on average declined by 20%. To understand the post-reform recovery it is necessary to understand the pre-reform decline, which, we will argue, was caused by a deterioration in the availability of consumer goods in rural areas. Because parts of the analysis are quite intricate, we begin with an overview. ReW Prduw QpiBf 81 l s /~~a Volume of CmpSats Flgure 1: The Path of Crop Supply 1 977-B8 Our argument is sununarized in Figure I which shows thie supply curve of marketed crops. Comparing 1977 and 1988 thie crop price is now lower in real terms and crop supply per 14 capita of the rural population is lower. Hence, in the first instance, we can think of the economy as having moved down its supply curve. There are two provisos to this. It is not safe to conclude that the economy can move back up this supply curve simply by restoring prices. During the Intervening years agricultural infrastructure has deteriorated so that it would probably be unable to handle sales at their 1977 level. This has, in ei,Vct, steepened the supply curve. Further, the supply changes during the period 1977-89 have not been movements down the supply curve but rather movements off it and then back onto it as depicted by the dotted line. There was a movement off the supply curve during the period 1979-84 because of shortages in rural markets for consumer goods. These shortages came about for three reasons. First, the national supply of consumer goods declined: the allocation of foreign exchange for consumer imports and the domestic production of manufactured consumer goods both fell severely. Second, although this decline in supply would normally have caused a large increase in the relative price of consumer goods, in Tanzania there was a cost-plus price control system already in place. Because the system was cost-plus, it did not permit prices to increase in response to shortages. The effect is depicted in Figure 2: instead of the price rising to the new market clearing level p*, it remained at the controlled level p and shortages emerged. Thirdly, these shortages were particularly acute in rural areas. Official allocations of consumer goods came increasingly to be skewed towards the cities. For example, whereas in 1977 Dar es Salaam had been allocated only 22% of Matsushita radios by the Board of Internal Trade, in 1982, with total supply more than halved, its allocation was raised to around 40% (see Bevan et al (1989), Table 3.6). Of the goods which left Dar es Salaam, a substantial proportion leaked onto the parallel market, however, there is evidence that these supplies were also skewed towards urban centers. For example, of the supply of soap to towns other than Dar es Salaan around 40% was probably via the parallel market, compared with only around 30% in villages 1'. Such a bias in the parallel market is consistent with evidence that it was easier to operate as a black marketeer in urban than in rural areas (see Bevan et al. (1989a)). To an increasing extent then peasant farmers faced shortages in the market for consumer goods. Prior to shortages peasant sales of crops were based upon the expectation that the resulting income could all be spent on desired purchases. However, if a household is repeatedly unable to make its desired purchases due to shortages, it will revise downwards its planned cash income by choosing to sell fewer crops. Hence, in times of shortage, crop sales become determined by the availability of consumer goods instead of by the normal relationship to prices. Further, once there are shortages, the relationship between prices and supply is altered. Whereas under normal conditions of market-clearing raising crop prices (relative to goods prices) will raise supply, in conditions of shortage the opposite will happen. Faced with 2/ Bvan ct al. (1989) Table 10.7 gives a breakdown of the lekp of mp suppliea fiom the factory gate to the village hop. Of the soap sent to regional headquarters 36% lealkd away betwn the three recowded dispatch points (factory, regional headquates and ditict headquarters) and their official r.axt detaton (rgonal headquter, distrct headquarters and village shop, respectively). Presmably, thee leakages to a lar extent oonsiuted the supply to tht part of the paallel maet outside Dar es Salaam. Of thae leakages, nealy half occured between the factory d the regional headquartera and a fiuther quarter betwen there and tho dinsrt headquarte. It seems likely that, given the inadequacies of the transport sysan, black maet supplies were sold to consumn whore the leak took placeo rathr than being tansported long distances. If this is o, then the above leaages probably supplied towns. At most, those lakages between the dispatch from district headquartes to village shops may have constituted the unofficial supply reaching the villages. This is the assumption used to convat the dat of Table 10.7 to the figur cited in the text. IS constrained desired expenditures and hence a limit to the amount of cash income which is likely to be usefil, higher crop prices will be offset by lower sales in order to maintain income constant. Ibis reversal of supply response to price is shown as relationship (3). This can be related to our original story as depicted in Figure 1. As shortages of consumer goods intensify the supply function gradually shifts to the left (SI, S2, S3) and becomes negatively sloped. pvo= of Conmer 2o Ooo~~~~~~~o wth Price Controls crees Shortage B3etweeni 1977 and 1984 the supply of export crops per capita fell by 30% and the real producer price of export crops fell by 26%. On our account, under these special cJircumstances without te fall in crop prices the decline in crop sales would have been even larger. Econometric tests of supply response in 17 regions during the period 1978-84 have found the volume of crop sales to be positively and significanly related to the supply of consumer goods to rral areas and negatively (though i osisnffcanf y) related to rer crop prices, supporting the above account (Bevan, et al. (1989) Chapter 10). The above influences upon crop supply operate through the peasant's need to finance expenditure. An additional motive for generating income is to add to cash balances, these being both the medium of exchange and the principal financial asset available to peasants during the period. Conversely, should desired cash balances fall, the peasant can reduce income (i.e., crop sales) and maintain expenditure while money balances are being run down. This is shown as relationship (4) in Figure 3. We distinguish between two types of change in money balances, real and nominal. - 16 - Iflation (6) j Demand for nflation I -) I Money Balances | (5) / (4) Availability of l (2) Volume of Consumer Goods I > Crop Sales (3)/ /(1) I I Real Producer l Prices Figure 3: Influences on the Volume of Crop Sales The demand for real money balances is normally related to real expenditure and hence to real income. Between 1977 and 1983 peasant eal cash income per capita fell substantially. Using a comparison of budget surveys, Bevan et al. (1989) find a fall of 38.7%g'. The expectation is therefore that real cash balances per capita would also have fallen. This would have enabled peasants temporarily to reduce their cash income by more than the fall in their expenditure. The magnitude of such an effect depends upon the size of peasant cash holdings relative to income. Peasant cash holding was around 75% of cash incomeW. Hence, had peasants reduced their money holdings so as to restore this ratio of money to income, the 38.75 fall in income would have permitted the equivalent of a one-year reduction in cash income of 29% (38.7% x 0.75). Had this been spread over six years, cash income would have been reduced by 5% throughout. Hence, crop sales, as an important part of cash income, would also probably have been reduced by around this amount. WJ Drived from Bevan at al. (1989) Table 3:14. 221 Mean cash income in 1980 per peasant household is known frm a large sample ruml wuvey (Collier, Radwan and Wangwe(1986)). Mean cash holdings from Ceatral Bank and Census data. 17 and the number of households (from the Census). Since peasants make up 80% of the population, it is likely that their holdings are closely related to the Tanzanian average. Although peasants have lower cash incomes than the average household (and so would tend to hold less money on this count) they receive income much less frequently and less p-edictably than wage earners, the other major group in the economy. The latter effect implies that per shilling of income they would tend to hold considerably more cash than wage earners. These two differences from the average Tanzanian household are therefore qualitatively offsetting. The resulting series on real cash balances per capita Is shown in Table 4. Between 1977 and 1981 there was a 21.2% increase in mean real cash balances. If cash income did indeed decline by around 39% during this period then the velocity of circulation halved in the peasant economy. (1.212/0.613 = 0 6). Bevan, et al. (1989) provide an explanation for such a decline in velocity as a by-product of the shortage of consumer goods. They argue that, faced with random shortages of goods, peasants would need larger cash holdings per shilling of intended expenditure in order to avail themselves of those occasions when availability was atypically good. This relationship between expenditure and cash balances is shown in Figure 4. With the onset on shortages they show that it is quite possible for money balances to rise even though expenditure falls. However, if expenditure continues to fall, money holdings must eventually decline: in the limit, if there are no goods so that expected expenditure is zero, there is no need to hold money. Starting from market-clearing in 1977, real balances indeed rose as expenditure declined until 1981. Rel RetNsp R hla p wlih Moy Shoftges MaSt-Cng Planned E*ednure Flgure 4: The Demand for Real Money Balances: 1977.89 18 ble A4: Changes In Rel Cash Balances and the Inflatlon Tax Real Cash Cash Balances Change Inflation Total Savings as Nominal Real Per Cap in RCBpc Tax (1977 RCB = 100) of Cash Income 1977 100 100 11.8 5% 1978 119.0 108.6 8.6 6.2 14.8 8% 1979 156.6 123.2 14.6 12.4 27.0 16% 1980 209.0 122.8 -0.4 28.6 28.2 17% 1981 266.5 121.2 -1.6 25.1 23.5 14% 1982 328.2 112.6 -8.6 27.1 18.6 16% 1983 363.6 95.5 -17.1 24.0 6.9 6% 1984 419.5 82.0 -13.5 25.3 11.8 9% 1985 521.1 68.6 -13.4 20.5 7.1 5% 1986 697.3 65.1 -3.5 16.8 13.3 10% 1987 963.1 66.7 1.6 15.0 16.6 14% 1988 1266.0 67.1 0.4 15.9 16.3 12% 1989 67.2 Notes to Table 4 Nominal Casb Balances from Bank of Tanzania Economic and Quartely Bulletin December 1988 Table S Column 6. This gives end of year figures. Our series is for the annual average. Hence, for example, we aveae he end-1976 and end-1977 figures to get the averg money holding in 1977. Mhe conversion to a per capita basis uses the inter-censal population grwth rate of 2.8% p.a. The conversion to constant 1977 prices uses for the period 1977-83 the National CPI, Bank of Tanzania (ibid) Table 25 (a) (i). The inflation tax is shown in constat 1977 prices. For example, the price level in 1981 was 25.7% above that in 1980. Real balances in 1980 aveged 122.8 (indexed on 1977). Inflation eroded these balnces to 97.7 duiAg 1931 (122.8/1.257). Hence, the inflation tax was 25.1. For the pediod 1983- 89 we us the rual cost of living index constucted and described in Section 4.1. The conversion to the equivalent rate of taxation on cash income in the final column is cieved in two stages. First, cash balances are estimated to be 75% of cash income in 1980 as discussed in the text. Since in 1980 28.2 is added to balances of 122.8 (i.e., 22.9%) the equivalent tax rate is 17% (22.9 x .75). Cash incomes for the period 1977-83 relative to that in 1980 are based on per capita sales of six peasant export crops. Cash incomes for the period 1983-88 are from Table 9. Ibereafter, there was a substantial decline. By 1983 they were slightly below their 1977 level. Bevan et a]. describe such a pattern of money holdings as the 'honeymoon effect' of shortages: the onset of shortages is accompanied by an initial phase during which peasants are willing to sell crops to the government in order to build up money holdings. However, this phase does not last; once peasants choose to reduce their holdings the effect on crop supply is reversed. Hence, this analysis implies that whereas until 1981 crop supply was boosted by the change in real money balances, post- 1981 they were reduced. This helps to 19 account for the severe decline in crop sales during the early 1980s. The analysis implies that this effect would have continued as goods availability deteriorated. The effect of shortages on real money demand (shown as relationship (5) in Figure 3) poses a serious problem for the transition back to market-clearing. By 1986 goods were much more available in rural areas so that peasants could reduce money holdings per shilling of expenditure. As Table 4 shows, this they indeed did: real balances fell by around a third between 1983 and 1986 and have since been stable. The third column of Table 4 in each year between 1982 and 1985 there as a large reduction in real balances. In terms of Figure 4, the adjustment back onto the 1977 ratio of money to expenditure has been achieved partly by increased expenditure but mainly by reducing real money holdings. This rapid reduction in real money holdings implied that ash income, and hence crop supplies, would temporarily be reduced. This was, however, avoided by offsetting changes in the price level to which we now turn. The demand for money can rise for two reasons: either the desired holdings of real balances can increase (as it did 1977-81) or, the price level can rise, requiring households to add to their money balances in order to keep them constant in real terms, an effect described as an inflation tax. This is shown as relationship (6) in Figure 3. In Tanzania the inflation tax was initially modest (see column 4 of Table 4). However, during 1980-82 it became very powerful for two reasons. The rate of inflation increased, and this interacted with the "honeymoon effect": since peasants had increased their real money balances a given rate of inflation was eroding a larger stock of money. Thereafter, as peasants started to reduce reai balances, the take from the inflation tax fell despite a generally rising inflation rate. The net effect of the inflation tax and the change in the demand for real balances is shown in the last two columns of Table 4. The first of these is simply the addition of the two effects. The last column expresses this as a percentage of cash income in each year. This is of interest since the income peasants need to devote to adding to money balances instead of to expenditure is analogous to a tax on cash income and so the column can be thought of as the implicit income tax which the government was levying on peasants, achieved by the changes in money holdings. To arrive at this percentage we need an annual series for peasant cash income 1987-88. As a proxy we use an index of crop sales per capita of the rural population. This column reveals substantial swings in the rate of implicit taxation. At the start of the period the rate was modest: 5% in 1977 and 8% in 1978. It then rose sharply, averaging 16% in the years 1979-82. Even during this phase there was a change in the composition of the tax from the "honeymoon effect" to inflation. In 1983 and 1984 there was a large fall to an average of 7%. Finally, during 1986-88 the rate rose again to an average of 12%. That is, despite the low level of real balances during this period the inflation rate was sufficiently high to restore the implicit tax rate to close to its previous peak. These large swings must have contributed to the changes in crop supply. In particular, the sharp fall in the tax rate during 1983-84 helps to explain the severe economic collapse of those years while the restoration of tax rates by means of inflation explains why the economy was able to overcome the otherwise major transition problem posed by the run down in real balances. To summarize, there are six relationships influencing peasant crop sales. These are in two groups, a monetary effect and a 'real' effect (availability and relative prices). During the 20 phase of decline in crop sales, 1977-84, the volume of crop sales per capita fell by around 30% 'W. The simple interpretation of this decline, namely that it was caused by the fall in the real producer price of crops, we have argued is incorrect for the special circumstances of shortages of consumer goods which prevailed during this period. Statistically, a better explanation is provided by the contraction in goods supplies. However, the value of the more complex approach to supply response really comes into its own when we turn to the recovery phase 198318S87/88. During this four-year period there was a spectacular recovery in the volume of crop sales per capita of the rural population of 47%. Yet the average crop price fell in real terms by 20% during the same period. (Details of both the quantity and price series are given in Section 4). Clearly, whatever, explains the remarkable growth in crop sales which reversed a long period of fairly continuous decline, it was not real producer prices. Yet returning to Figure 1, it is what would be predicted by the more complex theory. Between 1983/84 and 1987/88 the rural economy moved back to market-clearing conditions and hence reverted to something approaching the original supply curve (S-S). Tbis was achieved pardy by larger supplies of consumer goods, finantced by own-funded imports and by aid, and pardy by the abandonment of price controls which raised the price level. The rise in the price level eroded real money balances, as discussed above, and lowered real crop prices (despite being partly offset by nominal increases in producer prices). Note that the increase in the volume of crop sales has primarily been due to the annual crops and particularly to grains. That peasants chose this particular composition for their Increased sales may reflect temporary climatic improvements which favored these crops, and/or the deterioration in marketing infrastructure for the export crops relative to that for the grains (where competition has been permitted among trading agencies), and/or the deterioration in the tree stocks of coffee, cashew and tea). Again, the relative switch towards the annual crops is not a result of relative price changes since grain prices have fallen relative to the export crops. This is not, of course, to deny that price effects are important, but rather to indicate that to date they have been swamped by other effects. 4. Trends in Peasant Living Standards, 198349 We now present four approaches to the measurement of trends in peasant living standards taking as our base year 1983, the first three concerning trends in real cash expenditure on urban-supplied goods. The fwrst approach is to take cash balances as a proxy for peasant expenditure. The second is to measure peasant incomes from some large components. The third is to measure the national supply to consumers of major goods. The fourth approach focuses upon peasant consumption of food. However, we start with a discussion of the measurement of the peasant costs of living. 4.1 The Peasant Cost of UAving There is no cost of living index in Tanzania specifically for peasants, there being only the National Consumer Price Index (CPI). However, for some purposes this index is quite inappropriate for two reasons. First, although the typical peasant household has a consumption pattern similar to the national average, It has a radically different expenditure pattern because of subsistence consumption. Further, the peasant economy in aggregate has a yet more different .jQI The six major oxport emps only. 21 pattern of net purchases because most of the food which the typical peasant household buys is purchased directly or indirectly from other peasant households. Since food has a weight of 64% in the (CPI) these differences are important. The (CPI) is dominated by changes in food prices yet the peasant economy in aggregate virtually does not purchase food. This is particularly important when measuring the peasant terms of trade with the rest of the economy: the price of the goods which peasants sell to the rest of the economy should clearly be deflated by the price of the goods which they buy from it rather than by the . We have therefore constructed a peasant-speciflt, (CPI) based on the prices of those goods which peasants purchase from the rest of the economy. From the 1977 Household Budget Survey the Bureau of Statistics provided unpublished data on the consumption pattern of rural households. From this we identified goods supplied from outside the peasant economy (such as soap and bicycles). Virtually all of these goods were included in the price gathering process for the construction of the (CPI). Hence, by going back to the underlying price data It was possible to construct a price index for rural purcbases from the rest of the economy on 1977 weights. This was undertaken for the period 1983-88. The resulting index is compared with the (CPI) in Table S. From the onset of the reforms in 1984 to 1988 prices of the goods which peasants purchase from the rest of the economy have risen by 9.8% relative to the (CPI). Hence, estimates of real producer prices which are generally based upon the (CPI) substantially overestimate prices by the end of the period. The second reason why the (CPI) is likely to mis-state the true change in the cost of living is that it does not take into account changes in the range of choice available for the typical differentiated product. For example, even soap, a relatively standardized product, is differentiated by scent, by quality and by function (aundry or toilet). Between 1978 and 1984 the range of choice narrowed as part of the general contraction in supplies of consumer goods, whereas since 1984 it has widened again. The range of choice affects the cost of living in two ways. First, since consumers have different requirements, the wider the range of choice the closer are the types of goods available to what different people ideally want to buy. If people who want laundry soap have to make do with toilet soap the true cost of washing clothes has been increased. Second, for many differentiated products, the typical consumer wishes to buy more than one variety. People will wish to buy both laundry soap and toilet soap. 22 Table S: Consumer Prices for the Peasant Economy and the National Consumer Price Index Compared National Price of Rural CPI Purchases from Urban Areas 1983 100 100 1984 136.1 130.7 1985 181.5 188.9 1986 240.3 258.9 1987 312.3 339.7 1988 409.7 431.8 Source: National CPI as Table 4, Price of Rural Purchases derived from unpublished Bureau of Statistics data. We now attempt to quantify the effects on the cost of living of the post-1984 widening of product choice. We use the theoretical framework developed by Dixit and Stiglitz (1977). They .;how how for a given total expenditure a consumer will get more welfare as variety is increased. An alternative way of expressing this is to say that in order to achieve a given level of wtiare a consumer will need to spend less as variety is increased: widening choice lowers sla.ost of living. The magnitude of this effect depends upon how closely one variety of a pmr4 2uit can substitute for another: for example, how willing the consumer is to use laundry soap istuid of toilet soap. This is measured by the 'elasticity of substitution' which can range between zoo and infinity. When the consumer is completely indifferent between varieties the elasticity of substitution is infinite. Clearly, in such a case greater variety would not lower the cost of living 7n Table 6 we show the reduction in the cost of living implied by various increases in the nu '-ooer of varieties for various values of the elasticity of substitution 11. jjl In their approch differert varieties aro modelled as CES (constant elasticity of substitution) substitutes in danad. The mumber of potenial varieties nay be vew large but not all are available. If differt vauicties ane treate symmetrically, t consmer will choose to buy the ame quantity of all available varieties. 23 Iable 6: The Reducion In the Cost of Uving Resulting from Greater Varlety (%) Elasticity of Substitution 1 2.5 5 10 Increase in number of Varieties Double 50 24 13 7 fourfold 75 42 24 13 six fold 83 51 30 17 For example, if the number of varieties doubles and the elasticity of substitution is 2.5 (the sort of magiltude that might be expected) then the implied reduction in the cost of living is 24%. Such a reduction in the cost of living would in tum imply that for the same expenditure the consumer would be 32% better off (1/1(1-0.24) = 1.32). To apply this theory to Tanzania we need data on the range of product choice available to the typical consumer. For this we have adapted the underlying data gathered by the Bureau of Statistics for the Consumer Price Index. Our analysis is confined to the range of choice among different varieties of soap but this is the single most important differentiated product in the expenditure pattern of peasants according to the 1977 Household Budget Survey. The Bureau samples of 16 varieties of soap in twenty urban centers each quarter. Hence, each year there are 80 shopping visits. When a particular variety is not available this is recorded by the statistical officer. Using this information, we have calcuated for 1983, 1986 and 1988 the average number of varieties of soap found during these 80 visits. The results are shown in the first column of TaLs 7. 24 Table ?: hIprovements In Consumer Cholce, 19888 Availability of Varieties (1) Availability of Goods (2) 1983 0.563 4.8 1984 4.7 1985 - 6.7 1986 2.544 6.1 1987 - 7.3 1988 3.825 7.4 Notes and Sources: 1. Refers to how many varieties of soap were available on the typical shopping trip. Data for soap varieties is available only for 1983, 1986 and 1988. See note 2 for definition of shopping trip. 2. Refers to eight goods: cooking oil, margarine, toilet soap, laundry soap, matches, khanga, kerosene and cigarettes. Figure refers to how many of these eight goods were found to be available on the average "shopping trip' by an enumerator of the Bv-reau of Statistics. The number is the unweighted average of eighty shopping trips per year; one per quarter for each of twenty regions. Source: Derived from unpublished data gathered by the Bureau of Statistics for the construction of CPI. This example suggests that wider choice has substantially lowered the cost of living and hence raised living standards. If the soap data are representative (and while for other goods the increase in the number of varieties may not have been quite as dramatic, for many goods other than soap one would expect the elasticity of substitution between varieties to be lower) the change between 1983 and 1988 would have raised welfare by more than 20% even if one adopted as high a value for the substitution elasticity as 10. If the substitution elasticity has a value of 5 (still a rather high value) the improvement would be 47%. This would imply a 32% fall in the cost of living. So far we have only considered the widening of choice through an increase in the number of varieties of a good. However, choice can be increased in two ways: not only can the number of varieties increase but so can the number of goods. In our previous example, soap was treated as a single good and we zonsidered changes in the number of its varieties. At a higher level of aggregation we can similarly consider not how many varieties are available but how many different goods are available. In many ways this is more important because the scope for substitution say between soap and matches is much lower than that between different varieties of soap and so non-availability has more deleterious effects upon living standards. Figure 5 Quarterly Data on the Availability of Eight Goods on the Typical Shopping Trip, 1983-1988 8 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1983 1984 1985 1986 198 198 26 We attempt to quantify the widening in the choice of goods using the same Bureau of Statistics data on shopping visits. We consider eight Important consumer goods (each usually represented by several varieties). A good is available if at least one of its varieties is available. The results in Table 7 reveal a marked improvement in the availability of the range of goods. For example, between 1983 and 1988 the number of different goods available increased by 54%. The quarterly data (shown in the graph) reveal an even more dramatic turn-around. Availability deteriorated sharply between the first and third quarters of 1983 (from 6.1 to 3.8) and continued at this low level until the second quarter of 1984. From then on there is a rapid and fairly continuous improvement. In the case of the increase in the number of varieties we were able to indicate the order of magnitude change in living standards and the cost of living which it implied. To do the same thing for goods rather than varieties is more problematic since there are more likely to be large differences between goods in their substitutability. Ideally, we would need to estimate a 'demand system'. Since this is not feasible on the data we can only follow the same procedure as that used for varieties. Again the magnitude of the effect is dependent upon the substitution elasticity chosen. Here we choose a value of 2.5 which is very high for this level of aggregation and so likely to understate the fall in the cost of living (welfare improvement). The improvement in availability between 1983 and 1988 (54%) would then imply a welfare improvement for given expenditure of 19%, or equivalently stated, a 16% reduction in the cost of living. The effects which we have discussed above are cumulative. The wider choice of varieties may (using an elasticity of 5) have lowered the cost of living by 32%. The wider choice of goods may have further lowered the cost of living by 16%. Hence, together, they imply a fall in the cost of living of 43% (I - (1-.32)(1-.16)) for a given level of cash expenditure this implies an increase in the welfare derived from it of 75%. This is in addition to the more conventional changes in living standards which we will be analyzing below, namely increases in cash income and increases in subsistence consumption. Clearly, the figure of 75% is little more than a numerical illustration. However, it does suggest that there have been substantial gains from wider choice, gains which are Ignored in the usual statistical standard of living analysis. 4.2 Money Balanes and Rural Expenditure If the velocity of circulation is approximately constant then money balances are an indicator of expenditure. As discussed in Section 3, mean cash balances per capita of the national population constitute a reasonable proxy for mean holdings per capita of the rural population. The assumption that the velocity of circulation is constant would not, we have argued, be valid during the period of shortages (1979-84). During shortages the demanL for money rises relative to expenditure. We have already noted such a rise, followed by a sharp fall following the restoration of market-clearing. Hence, the years which are candidates for the procedure are 1977-78 and 1986-89 (although 1985 was probably a year a market-clearing, it is likely still to have been part of the adjustment process to lower real balances). On the basis of real balances per capita for these years two points are worthy of note. First, we can compare real balances in 1986 with those for the period 1977-78. This provides some information upon the change in real cash income between the onset and the end of the period of shortages. This comparison implies a decline in real cash expenditure of 37.6% (1-[65.1/104.3D). Note that this is very similar to the estimate based upon a comparison of budget surveys discussed in Section 3 (38.7%). Second, during 1986-89 there has been no significant increase in real cash balances 27 per capita. This in turn would imply that real cash incomes in the peasant economy have not recovered substantially to date. 43 Income from Sales of Crops In this section we attempt to estimate peasants' real income from crop sales. We distinguish three categories: sales of export crops, sales of food crops through official channels, and sales of food crops through unofficial channels. The results are set out in Tables 8 and 9. Sales through official channels are unproblematic. However, food sales through unofficial channels are estimated indirectly. Our procedure is to estimate urban food consumption using the 1977 Household Budget Survey data of urban grain consumption per capita and the inter- censal growth rate for the urban population. That is, we assume constant urban per capita grain consumption during the period. Table 8: An Estimate of Open Market Grain Sales: 1982/3 - 1987/8 1982/3 1983/4 1984/5 1985/6 1986/7 1987/8 Urban Pbpulation 2904250 2969874 3038545 3109799 3184381 3262240 Urban Food Consumption: Maim '000 tonnes 487 498 509 521 533 546 Rice o00 tonnes 99 102 104 106 109 112 NMC Sales: Maize '000 tonnes 209 254 220 157 76 120 Rice '000 tonnes 72 79 58 28 63 62 Open Market Sales: Maize '000 tonnes 278 244 289 354 457 426 Rice a00 tonnes 27 23 46 78 46 S0 Assumptions: 1. Urban population is estimated from the National 1988 Census figure, and respective regional gowth rates applied to each urban center. 2. Assume a per capita grain consumption of 2000 cal/day. Maize volumes obtained by assuming an 89% share of grain consumption coming from maize and a maize calorie content of 3530 per kg. Rice Volumes obtained by assuming an 18% share of grain consumption coming from rice and a rice calorie content of 3450 per kg. The maize and rice share are deducted from the 1976n7 HBS which indicated 100 kg/year/pc of maize and 20 kg/year/pc of rice consumption in DSM. 3. NMC sales source: HDB 1988. 4. Open market supplies: residual of urban consumption - NMC official sales. 28 Tabl I: Crop Sales 198213-1987/88 1982/83 1983/4 1984/5 1985/6 1986/7 1987/88 Official Grain Sales: Maize '000 tonnes 85 71 85 178 178 229 Rice '000 tonnes 21 22 12 16 11 43 Official Price Maize 1750 2200 4000 5250 6300 8200 Official Maize Price Index 100 126 229 300 350 469 Official Price Paddy 9000 4000 5000 8000 9000 14400 Official Paddy Price Index 100 133 200 267 300 480 Open Market Grain Sales: Maize '000 tonnes 278 244 289 364 457 426 Rice '000 tonnes 27 23 46 78 46 S0 Open Market Price Maiza 4740 9530 10a10 9710 10590 14960 O.M. Maize Price Index 100 201 222 205 223 316 Open Market Price Rice 13500 29400 65300 38S00 37100 46900 O.M. Rice Price Index 100 218 484 265 276 347 All Grain Sales: Total Maize '000 tonnes 354 915 374 542 630 655 TotalRice '000tonna 48 45 S8 94 S7 93 Maize Quantity Index 100 87 109 149 179 180 Rice Quantity Index 100 92 120 195 118 191 Average Maize Price 4033 7876 9030 8245 9413 12698 Average Maize Price Index 100 195 224 204 233 312 Average Rice Price 9646 17950 53701 34063 32613 35418 Average Rice Prie Index 100 186 657 353 338 367 Grain Sales Quantity Index 100 88 107 160 160 183 Grain Sales Value: Maize m. Tah 1458 2477 3977 4468 5934 8258 Rice m Tsh 467 802 3115 3217 1859 3282 Total m Tsh 1933 3279 6492 7686 7793 11540 Maize Index 100 169 230 305 405 663 Rice Index 100 172 667 689 398 703 Total Index 100 170 336 398 403 597 Export Crops Sales 1/: Quatity Index 100.0 100.9 102.3 89.9 119.4 120.8 Price Index 100.0 131.6 174.3 254.2 325.2 415.8 Value m Tsh 1986.2 2651.8 3550.7 4580.7 7794.5 9886.3 Value Index 100 134 179 231 392 498 All Crops: Sales Value Index 100 152 257 314 398 547 Quantity Index 100 94 105 125 139 151 Price Index 2/: 100 162.6 207.8 237.6 283.4 376.1 /1: All taditional exports (coffee, cotton, cashew, tea, tobacco, pyrmthr, cardamon), excluding sisal. /2: Lspeyres price index, base 1982/82-100. Table Cont.l 29 Table 9, Cont. 1982/83 1983/84 1984/85 1985/86 1986187 1987/88 Non Food CPI 100 130 183 239 306 373 Real Official Maize Price 1750 1690 2186 2199 2062 2197 Real Off'l MaiZ Price Index 100 97 125 126 118 125 Real Off'l Rice Price 3000 3072 3279 3351 2946 3858 Real Ofrl Rice Price Index 100 102 109 112 98 129 Real O.M. Maize Price 4740 7819 5744 4067 9466 4008 Real O.M. Maize Price Index 100 154 121 86 73 8S Red O.M. Rice Price 13500 22578 95685 16127 12143 12567 Ral O.M. Rice Price Index 100 167 264 119 90 93 Real Average Maize Price 4093 5048 4935 3454 3081 3376 Real Ave. Maize Price Index 100 150 122 86 76 84 Real Ave. Rice Price 9646 13785 29347 14268 1067S 9490 Real Ave. Rice Price Index 100 143 304 148 111 98 Real Gain Sdes Value: Maize m Tah 1466 1902 1846 1872 1942 2213 Rice m Tsh 467 616 1702 1348 609 879 Totd m Tsh 1933 2518 3548 3219 2551 3092 Maize Index 100 130 126 128 132 151 Rice Index 100 132 364 289 130 188 Totd Index 100 130 184 167 132 160 Export Crops Sales: Real Price Index 100 101.1 9S.3 105.S 106.4 111.4 Real Vdue Index 100 102.5 97.7 96.6 128.4 133.4 All Crops: Sales Value Real Term Index 100 116 141 132 130 147 Real Price Index: 100 12S 114 100 93 101 Official sales are then deducted, leaving open market purchases as a residual. The details of the calculations are set out in Table S. If we regard the income from crop sales during the crop year 1982/83 as being spent during 1983, we can deflate the resulting sales value index ('all crops") by the peasant cost of urban purchases (Table 10). 30 Table: rPeasn Real Crop Incom Nominal Crop Rura Purchases from Real ncoke frnOM Sale (total) Urban Arms Prce Index Crop Sales p.c. J' 1983 100 100 100 1984 152 130.7 113.8 1985 257 188.9 130.3 1986 314 258.9 113.6 1987 398 339.7 107.4 1988 547 431.8 113.6 This income series again suggests that there has been little growth in real income since 1983. On a per capita basis between 1983 and 1988 real income rose at an annual rate of 2.6%. 4A Supplies of Consumer Goods Our third approach to trends in living standards is to estimate changes in the supply of consumer goods. Whereas our two previous approaches applied only to rural living standards, this approach applies to the national level: we are unable to distinguish between rural and urban households. The supply of non-food consumer goods comes from two sources, domestic production and imports, and we consider these in turn. The series on domestic production measures the volume of consumer goods output and is shown in Table 11. Note that despite the overall revival in manufacturing production post- 1985 the output of the consumer industries continues to decline throughout the period. This may be illusory, because the series is confined to large firms operating at the start of the period. Many of these firms have been adversely affected by trade liberalization whereas the policy changes have assisted new and small firms which are not part of the index. Data on the recorded volume of consumer imports are published but are subject to biases from over- and under-invoicing. We atteipt to correct for this by measuring the changing incentives to mis-invoice. The details of our calculation being set out in the footnotes to Table 12. The resulting series is shown in Table 11. We construct a weighted average of these two series as an indicator of the total availability of non-food consumer goods to the economy. In per capita terms by 1988 consumption is around 21% higher than in 1983. W Ung tho natl population g8wth rate of 2.2% p.a. 31 Table 11: Suppfli of Non-Food Consmer Good 1982/3 1983/4 1984/5 1985/6 198617 1987/8 Conmsun3 Manufacues J 100 99.0 104.8 109.0 87.9 (81.0) Adj. Cons. Imp. J# 100 169.4 145.8 138.1 133.0 184.6 Cons. Availabilityz' 100 138.4 127.8 117.0 113.2 139.0 Per Capita consumpt. 2' 100 134.6 120.2 107.7 101.4 121.1 4.5 Peasant Food and Consumption In Table 13 we compare our estimate of food sales to urban areas with MDB estimates of grain production maize plus rice, thereby deriving rural consumption residually. The result implies that rural food consumption increased by half between 1983 and 1988. 13V 1985 peducton value-weightd index of aU conwmer caegories in th Industrial C;omoditieB Qusey Rbport for 198547. For 198345 mcm is as Table 3A above. Tec ISIC caegoris inluded are 'food, bevenrgs and tobacco'; textile nd lather, paper and pawr product. For 1987-8 th sa is confined to beer, cigarets and taxtes, a in the Quantity table. M/ PFm Table 12. 2V Weiglhtd by value. mpors valued at Tuh 188 per $ (blaok make rate in 1988) domestc producdon valued at 3.255 ibm its 18 ui value (the inra in conumr prices for domestic manufact betwen 1985 and 1989(1)). z1 lbT int a popuation gmwt rawt is 2.8% p.a. 32 Iable 2: Consumer bnports Adjusted for Over and Under-Invoicing 1982/3 1983/4 1984/S 1985/6 1986/7 1987/8 Recorded Consumer Imp. (10) 100 160.5 132.4 124.7 110.7 154.5 % own-funded 0 39.6 64.2 60.8 63.3 SS.6 Net Returns per $1 if: over-invoice off.imp (1) sh. 25.2 50.1 87.3 102.5 81.4 SS.9 Index in real purchasing power 100 146 191 169 103 54 Under-inv. own-funded (2) sh. 3.3 4.6 5.3 9.8 19.3 29.8 Index in roal p.p. 100 101 87 122 185 217 Over-invoice rate % (3) 20 29.2 38.2 33.8 20.6 10.8 Under-invoicerato % (4) 3 3 2 3 5 6 Off. funded cons. imp. (5) 100 9.9 47.4 48.9 28.S 68.6 Adjusted Off. -funded imp (6) 83.0 75.0 34.3 36.S 23.6 61.9 Own-funded imp (7) 0 63.6 85.0 7S.8 82.2 85.9 A4justed own-funded imp (8) 0 65.6 86.7 78.1 86.8 91.3 Adjusted total imports (9) 83.0 140.6 121.0 114.6 110.4 153.2 Index 100 169.4 145.8 138.1 133.0 184.6 (1) The net incentive to over-invoice imports which are financed by official purchased foreign exchange is determined by two factors. The difference between the official and the black market exchange rate provides an incentive to over-invoice, but this is offset by having to pay a larger import duty. By over-invoicing an import costing $100 by $1 the net profit is given by (1 +t)e-b, where t is the tariff rat, e the official exchange at, and b the black market rate. The tariff rte used in the calculation is 30%. T black mta;t series used is Tsh 39.6 per $ in 1983, then 70, 110, 145, 165 and 18S. In the next row this is deflated by the national CPI. (2) In the cam of own-funded imports there is no incentive to over-invoice bcause the importer is using his own foreign exchange. However, the incentive to avoid tariffs by under-invoicing remains. Therefore the incentive to under-state the dollar cost of imports of $100 by $1 equals tLe where the symbols are as defined in note (1). This gain is again deflated in the next row by the CPI. (3) We assume that the over-invoicing rate is proportional to the incentive to over-invoice. We further assume that in 1983 the over-invoicing rate on consumer imports was 20%. (4) We assume that the rate of under-invoicing would be the same as the rate of over-invoicing for a common value of the incentive to mis-invoice. (S) Total consumer imports in 1983 are set to 100. This row shows that portion which is officially funded. The proportion of consumer imports which are own-funded is derived from unpublished Bank of Tanzania data on import licenses. We assume a common utilization rate of these import licenses. (6) The data are adjusted by the over-invoicing rate. (7) As note (5). (8) The data are adjusted by the under-invoicing rate. (9) Sum of adjusted series. (10) For 1983-86 the volume series is from Foreign Trade Stadstics, 1987, Table 27. This is extended to 1988 assuming a 3% p.a. increase in the dollar unit price. 33 Table 13: F-'d Production, Sales and Subsistence Consunption i983 1984 1985 1986 1987 1988 Grain production (000t) (1) 2001 2295 2520 2758 3003 3415 Total sales to urban (OOOt) (2) 412 360 432 636 687 748 Rual consumption (OOOt) (3) 1589 1935 2088 2122 2316 2667 Index per capita 100 119.2 125.8 125.1 133.6 150.4 (1) From Marketing Development Bureau. (2) From Table 9. (3) Residual of production and urban sales. 4.6 The Approaches to Trends in Peasant Expenditure Compared We bring together the three approaches in Table 14. Table I4: A Comparison of Real Income Changes 19U3-88 Real Money National Non-Food Real Income from Balances pc Consumption pc Crop Sales p.c. (1) Gross Net of Tax 1983 - 100 100 100 1984 - 134.6 113.8 110.2 1985 - 120.2 130.3 131.7 1986 100 107.7 113.6 108.8 1987 102.5 101.4 107.4 98.3 1988 103.1 121.1 113.6 106.3 1989(1) 103.2 The series on real balances and on real income from crop sales both refer only to peasant real cash income whereas that on non-food consumption refers to the whole population. The real crop income series should overstate the growth in peasant expenditure since from Table 4 we know that the incidence of the quasi-income tax levied through changes in the demand for money tended to rise over the period. Real income net of this tax Is shown in the last 34 column. For example, in 1983 the tax incidence was 6% so real post tax Income was .94 x 100, and in 1988 the tax was 12% so real Income was .86 x 113.6, which indexed onto 1983 is 106.3 Our analysis has suggested that despite a large increase in peasant crop sales, peasant real expenditures on urban gorIs have only Increased slightly. By contrast, their consumption of food has increased substantially. Tlis implies that the pattern of peasant consumption has switched in favor of food. By contrast, our assumption in Table 7 of constant per capita urban food consumption implies that the urban consumption pattern has switched away from food. These switches are consistent with movements in relative prices. Table 15 summarizes the changes for representative peasant household. In rural areas maize became cheaper relative to urban-supplied goods because of the fall in the real producer price of maize (the relevant price for peasants). Hence we would expect substitution into food and away from supplied goods. Table I1: Food and Urban Goods Consumed by Peasants (per capita, 1983=100) 1983 1988 Food (Quantity) 100 150 Urban Goods (Quantity) M' 100 108 Food/Urban Goods 100 139 Price of Food/Urban Goods a 1.0 0.72 The implied price elasticity of demand for the two groups of goods is around unity (which is credible). In urban areas, by contrast, the relative price of food has almost certainly risen due to the removal of the subsidy on sembe, which benefitted almost exclusively urban consumers. Hence, in urban areas we would expect substituton out of food into other goods. Note, however, that this does not imply an absolute fall in urban food consumption because overall expenditure has risen. Thus, given what we know about changes in relative prices, it must be the case that the constimption of food has increased relative to other goods in mral areas and decreased in urban areas. This provides at least qualitative support for our estimates. S. Condusion Our central findings are as follows. First, the rapid and continuous decline in living standards which took place between 1978 and 1983 has been arrested. Nationally, living standards have risen significantly in terms of both food and non-food consumption. Secondly, 21/ Fomm peasa pot-tax crop income in Table 14. 2s/ Price of urban goods from Table 5, producer price of muze fmm Table 9. 35 in rwail areas, while food consumption has increased substantially, non-food consumption has been broadly constant per capita. These findings must be qualified, in particular because there is no direct survey evidence post-1983. Some of the data are of uncertain quality and in some cases we have had to make assumptions. However, by using three distinct approaches, income, expenditure and output, we have generated a number of opportunities for our interpretation of the evidence to conflict. lhe results of these various approaches appear to be consistent. 36 References Bank of Tanzania (1988), Economic Bulletin, Vol.XVIII no.3 Bevan D.L. Bigsten, A., Collier, P. and Gunning, J.W., (1987) 'Peasant Supply Response in Rationed Economies', World Development, No.4, April, pp.431439. - (1987a), East African Lessons on Economic Liberalization, Thames Essay, No.48, Aldershot: Gower, for the Trade Policy Research Center. Bevan, D.L., Collier, P. and Gunning, J.W. (1989), Peasants and Governments: an Economic Analysis, Oxford: Oxford University Press. (1989a), 'Black Markets: Information, Illegality and Rents', World Development, Vol. 17(12), December. (1990), Controlled Open Econonies: a Neoclassical Approach to Structuralism, Oxford: Oxford University Press. Collier, P., Radwan, S. and Wangwe, S. (1986), Labour and Poverty in Rural Tanzania, Oxford: Oxford University Press. Dixit, A. and Stiglitz, J.E., (1977) 'Monopolistic Competition and Optimum Product Diversity', American Economic Review, 67, pp. 297-308. McKinnon, R.I. (1989), 'Financial Liberalization and Economic Development', Greek Economic Review, Volume 11, no. 1, pp.137-176, reprinted in Oxford Review of Economic Policy, Volume 5, no.4 NATIONAL INCOME ESrIMATES IN TANZANIA Methodological and Empirical Issues Tanzania Economic Report Background Paper #2 National Income Estimates In Tanzania Tablg of Contents Executive Summary . . . . . . . . . . . . . . . . . . . . . 39 I. Introduction .42 H. Review of National Accounts Methodology . . . . . . . . . . . 44 III. Methodological Adjustments to the National Accounts . . . . . . . 46 IV. The Parallel Economy in Tanzania . . . . . . . . . . . . . . 47 V. Institutional Action Program for Statistics . . . . . . . . . . . 53 Sub-Annex 1. Review of National Accounts Methodology . . . . . . . . 54 Sub-Annex 2. Methodological Adjustments to the National Accounts . . . . 58 Sub-Annex 3. 'Hidden Economy' Adjustments for the National Accounts . . . 65 Sub-Annex 4. Components of EUROSTAT's National Accounts Adjustments . . 67 Sub-Annex 5. Future Work Program . . . . . . . . . . . . . . . 69 Statistical Appendix Table 1: GDP by Industrial Origin: Mission Adjustments . . . . . 71 Table 2: GT)P by Industrial Origin: Official and Mission Adjustments ................ . 72 Table 3: GDP by Expenditure: Official and Mission Adjustments ............ . . .. 77 Table 4: Agriculture . . . . . . . . . . . . . . . . . 79 Table 5: Informal Food Crops . . . . . . . . . . . . . . 83 Table 6: Cardamon . . . . . . . . . . . . . . . . . . 84 Table 7: Oilseeds . . . . . . . . . . . . . . . . . . 85 Table 8: Construction . . . . . . . . . . . . . . . . . 86 Table 9: Hunting . . . . . . . . . . . . . . . . . . 89 Table 10: Mining . . . . . . . . . . . . . . . . . . . 90 Table 11: Handicrafts . . . . . . . . . . . . . . . . . 92 Table 12: Housing . . . . . . . . . . . . . . . . . . 93 Table 13: Tourism . . . . . . . . . . . . . . . . . . 94 Table 14: Gross Domestic Investment . . . . . . . . . . . . 95 EXECUTV SUMMARY 1. The official national accounts of Tanzania reasonably reflect the economic activities that are covered by conventional means given the resource constraints and basic data problems. The main challenge of the national accounts section of the Bureau of Statisdcs Office (BOS) is to extend the conventional framework to include parallel market activities, while at the same time to incorporate recent work on data and methods from the ongoing work with the technical assistance of EUROSTAT. 2. In the 1970s the parallel market probably accounted for 10-15 percent of GDP. The official figures represented a fair estimate of economic activity, especially since the input/output table and the household income and expenditure survey were done in the mid-1970s. From 1980 onwards, evidence points to fairly rapid growth in the parallel economy, and even through the period of liberalization starting in 1984, the sector continued to expand and remained significant. Very little of this rapid structural change, however, has been captured in the national accounts. 3. According to the recent publication, The Second Economy in Tanzania, by Bagachwa and Maliyamkono, the official national accounts are underestimated by about 30 percent for 1986. The accuracy of these estimates cannot be confirmed by the mission, but given indirect measures such as: (i) the apparent major discrepancies at the household level between reported incomes and expenditures required for basic sustenance; (ii) partial information from surveys of the informal sector; (iii) the level of own account imports (which have a counterpart in unrecorded exports of local production); and (iv) anecdotal information on informal employment in virtually all economic sectors, this seems to be reasonable. Only since the mid-1980s, as some of the restrictions have been removed and an extended range of consumer goods has entered more freely into the domestic market, have the severe shortages common in the past disappeared, and the size of the parallel economy has remained more stable. 4. Participation in the parallel economy in Tanzania is widespread because state intervention has not allowed markets to function properly. Fixed price policies, across-the-board subsidies, and incomes policies that began in the 1970s resulted in widespread supply shortfalls and the emergence of economic agents who brought producers and consumers together informally and for the most part "illegally". Shortages resulted in rapid price increases, with inflation averaging about 30 percent per year since the early 1980s. 5. The parallel economy comprises all operations and transactions that do not come under either direct or indirect state observance or control. The informal sector forms part of this economy, often supplying many of the goods for the parallel market. Informal sector activities tend to go mostiy unrecorded because they are small scale, and difficult to identify. Most activities in the parallel economy are small scale private agricultural and manufacturing activities that are genuinely productive, and trading activities that earn economic rent due to inadequacies and inefficiencies in the system form another main parallel market activity. 6. Apart from the main urban areas of the country, cross border trade is relatively common where access roads and tracks are better maintained than the internal road and communications network. There is an added incentive for cross border trade when a premium can be gained from the foreign exchange differential by selling directly or through barter trade. ITe customs departnent assumes that illegal cross border trade is substantial. - 40- 7. Reliable data on the parallel economy are difficult to collect, so the tendency has been to produce national accounts estimate based on existing data sources and established estimation procedures. Bank mission estimates attempt to go beyond the official estimates by extending some work done by BOS in EUROSTAT projects, and by making estimates for 'hidden economy" activities that are believed to be the most egregiously under-reported. Mission estimates were based on partial data in an attempt to fill in known gaps and reduce the existing bias. It is hoped that the comments and mission estimates will be a point of departure for generating better estimates through improved basic data collection and methodological refinements. 8. The Bank's main objective in terms of statistics development in Tanzania is to ensure that macroeconomic measures are robust enough for meaningful analysis and policy formulation. The official national accounts of Tanzania are currently only marginally sufficient for macro analysis. Some of the key components of GDP are subject to wide margins of error. For example, agriculture covers about 65 percent of GDP, but since the emphasis is on measuring export crops and a few main staples, about 40 percent of agriculture production is measured by proxy indicators such as population growth. It can be easily demonstrated that methodological choices in the agriculture sector have the effect of reducing agriculture value added levels and altering growth rates considerably. Another high priority area for improving estimation procedure is the level and change in investment. Due to the existing paucity of data on construction activity and lack of details of machinery and equipment imports, many of these numbers reflect little more than artfil guesswork, given also problems with measuring appropriate deflators, official estimates of real investment should be used with caution as well. 9. With the participation of EUROSTAT technical assistance experts and enhancement of computer operations as well as dissemination capabilities by Statistics Sweden, considerable efforts have been made by the national accounts office to improve the present situation. However, to achieve greater progress in national accounts coverage and enhance their reliability, the Tanzanian authorities will have to: 3 devote more resources to statistics; - begin more relevant and focussed in-house training with local experts and other outside resources; * devise a work program that includes surveys designed to encompass paraliel economic activities; * complete and incorporate all EUROSTAT projects into the national accounts; * increase the range of crops covered in agriculture surveys; and improve agriculture estimates by measuring area planted and harvested (either directly or using farmer surveys); * collect data on formal and informal employment, recognizing these activities may not be mutually exclusive; * revise the national accounts base year to a recent year in which conditions were more normal and prices were relatively stable; * incorporate estimates of Zanzibar in GDP estimates of Mainland Tanzania. 10. Finally, to promote the activities noted above and to determine further priorities for the national accounts work program, it has been suggested and we would endorse the idea of the -41- Planning Commission and Bureau of Statistics organizing a workshop of producers and users of national accounts information. In the first instance, such a workshop should involve government officials, the academic community and other Interested parties. NATIONAL INCOME ESIMATES IN TANZANIA I. Inroductlon 1. The official national accounts of Tanzania are compiled following broadly the conventional guidelines defined according to the UN System of National Accounts. The aggregates and their component breakdowns comprise categories that are internationally recognized, although the bases of the estimates themselves may differ from those in other countries because of the different assumptions and imputations that have been made, particularly because of the predominantly rural nature of the economy. A commodity flow basis of estimation (rather than enterprise survey returns or tax records) is used as the primary method for determining gross output and value added for some sectors such as construction. 2. Compared with many African countries the official national accounts tables of Tanzania are compiled with reasonable care and attention to detail, and senior national accounts staff are competent and dedicated. In conventional terms, the coverage of reported transactions is fairly comprehensive. With closer scrutiny, of course, it is possible; to incorporate quite a number of specific improvements to the estimates and to refine the existang aggregates to account for extra information about such components as recorded trade flows, intermediate consumption, etc. not previously available to the Bureau of Statistics (BOS). BOS staff have been refining the data and methods for nadonal accounts in conjunction with EUROSTAT sponsored projects. Progress has been made in identifying areas of weakness, but there is room for improvement. 3. Although the official national accounts of Tanzania are prepared in a traditional way following accepted guidelines as a comprehensive set of economic accounts describing the size and structure of the Tanzanian economy, the existing series are limited because they do not capture all transactions. In addition, they do not value correctly some of the transactions that are recorded and included in the national accounts. This is because many of the activities which are missing are 'hidden' from official enquiry, in most cases deliberately, while many others are misreported or under-recorded. The activides and associated transactions concerned belong to the "secondary economy" and take place in the parallel market. This necessarily restricts the usefilness of the present official series for development planning, policy analysis and program formulation, despite the fairly accurate reflection the national accounts provide of the official, formal economy. However, since the 'official' economy remains the dominant sector and continues to represent the principal stage on which the more important, large scale transactions (that are clearly more difficult to hide) take place, this review looks first at how well the government's data describe this part of the economy. 4. By their very nature, however, the official national income figures of Tanzania do not adequately take account of those economic activities that arise, informally, when there is, for example, significant disruptions in public services or breakdowns in the organized system of distribution, including transportation. Without a living wage from employment, people have to find alternative means of economic survival, selling the goods they produce and acquiring the food and other essential commodities they require for daily survival. They do this through the parallel market and, sometimes, illegal (in the sense of contravening state regulated activities and controlled prices) economy. -43 - 5. Agriculture is the principal sector, and while total outputs - or rather production sales - are reasonably well covered for major food crops, annual measures of intermediate consumption are not readily available and have to be assumed or approximated. Hence the gross value added figures are also affected, although less so to the extent that inputs form a relatively small percentage share of total output. 6. The same remarks can be made about the mining and manufacturing sectors. However, to the extent that recorded or reported, output sales do not reflect actual total sales (because, for example, the former are figures of sales reported by some monopsonistic national marketing board or controlling co-operative) the sector outputs will be undervalued. Such 'outputs', furthermore, may not accurately reflect the true physical production at the farm gate, mine, or factory level. 7. There are thus essentially two separate issues:- (a) To what extent should the existing official numbers be amended and modified to take account of new data obtained by the mission and the somewhat different approaches, for example, in assessing intermediate consumption that can be used to determine value added (and which also take account of previously unused information)? (b) How much do the official figures understate true sales of goods and services, and how different are these from the real current output measures? 8. Both issues draw attention to potential discrepancies in the official figures in the physical quantity estimates (and hence in the constant price output series) and to the underlying valuation (price) basis of the respective aggregates. 9. As in most developing countries, Tanzania shares in the lack of resources, wide geographic diversity and inaccessibility, the small scale of some economic activities, etc., and difficulties arise in collecting, compiling, and analyzing the national accounts. The issues associated with compiling national accounts are usually known best by the national accountants in the country, and Bank missions rely on their experience to formulate suggestions and adjustments that enhances the accuracy or overall coverage of the accounts. The mission review and subsequent adjustments to the national accounts were done with the hope that an exchange of information and ideas would benefit both the Bank and the Tanzanian authorities. 10. Part II of this chapter briefly reviews the current practices in compiling the national accounts of Tanzania, and discusses strengths and weaknesses of the official estimates. Details of output and expenditure accounts are in Sub-Annex 1. Part III summarizes how the official numbers can be improved using a more refined estimation procedure, re-doing the calculations at the more detailed commodity level in many instances, and applying new knowledge that has become available, e.g., as a result of the recent Eurostat consultancy arrangements on the national accounts, since the last official estimates were drawn up. This represents a disaggregated, sector focussed review of the data which emphasizes the significance of the micro data sets. Details of methodological adjustments are in Sub-Annex 2; "hidden economy" adjustments are noted in Sub- Annex 3. -44 - 11. Part IV takes a broader perspective by including estimates of the 'hidden economy". The aggregates are refined and corrected as far as the information allows in order to properly reflect the reality of economic activity and the scope of transactions regularly occurring. 12. In both instances, care must be taken not to incorporate unwittingly the assumptions that have been adopted in deriving new estimates, especially where they rest on only the shakiest quantitative foundation and by their nature rely little on supporting hard empirical evidence. Tbis applies not only in trying to determine more appropriate benchmark figures, but also in attempting to monitor more realistically the changes occurring from year to year since in the past fixed parameters have been applied to derive estimates, and such measures are clearly inappropriate in a country like Tanzania with its path of development over the past decado and a half. 13. Part V gives an overview of the institutional requirements and changes needed to improve the coverage and quality of national accounts in Tanzania. 14. The report also discusses In greater depth the methodological and data adjustments to the otfficial national accounts, and details of a future work program in the annexes. Also attached are statistical tables that show all mission estimates. II. Review of National Accounts Methodology 15. The national accounts office in the Bureau of Statistics (BOS) has recently written informative notes on the sources and methods of the latest GDP series. Ihis note and the Eurostat consultant's reports have provided a point of departure for the mission's review and adjustments of the national accounts. Eurostat's national accounts project (begun in 1986) has been based on a program of improving the naonal accounts by using existing basic data. A brief description and status of these projects is in Sub-Annex 4. 16. Only selected partial information from these projects have been incorporated into the official national accounts series. It is important to note that these revised national accounts will only marginally improve the reliability of the national accounts since the basic data upon which they are estimated are weak in some areas. The BOS note on sources and methods acknowledges this shortcoming, and the mission agrees that it is important to develop improved basic data from surveys, especially in sectors in which the informal economy is large. 17. A review of some of Tanzania's key strength and weakness in national accounts data and methodology in the existing national accounts are given below. 18. The key weaknesses of data inputs into the national accounts include: * uncertain estimates of agriculture output for many crops; * under-recording of informal economic activities; * lack of actual construction data; * reliance on official prices rather than transaction or market prices; e trade data -4S * lack of recent HBS data. 19. Key data strengths of the national accounts include: * availability of data from public sectors such as government services, utilities and public transportation and communication; * availability of data from large scale manufacturing; * new agriculture survey data available. 20. Key methodological weaknesses include: * outdated base year (1976); * outdated input/output relationships (1976) * weak estimation procedures for private and rural construction; * inconsistencies between current and constant price estimation methods. 21. Key methodological strength include: - UN SNA concepts and definitions followed reasonably well; 3 Some improvements implemented recently. 22. Price deflato are impliit deflators for most sectors since constant price estimates are calculated on the basis of physical indicators. This results in some deflators that change dramatically, and are inconsistent with other price changes in the economy. For example, the implicit utilities deflator increased by .8% in 1983, and by 85% in 1984 (overall inflation was about 20-30 for those two years). Since the estimation of current and constant prices used different methods (current price estimated by income method; constant price estimated by physical indicators) that may not have been 'nternally consistent, the deflators are odd looking. In this case, it might have been best to ase the physical indicator such as kilowatt hours and gallons of water produced for the constant price series, and then inflate by the average cost per unit to obtain the current price estimates. The deflator would be reasonable, and the nominal series would be consistent with production estimates. 23. In developing countries, price collection is weak, so using physical indicators to estimate constant price series is a common practice. The choice is between using physical indicators of the output series or the input series. Where the ratio of output/input is high, as in agriculture and manufacturing, it is best to use output as the physical indicator. If output is homogeneous, such as in the electricity sector, the output indicator is also good to use. Where output is difficult to identify, as in construction, transport, and services, inputs serve as the best physical indicator for the constant price series. However, when inconsistencies arise between current and constant price series, it may be necessary to review the data in both series to determine which gives the best estimates. Ibis is an area that deserves special attention in the Tanzanian accounts. 24. Valuation of national accounts should be at prices at which economic transactions take place, but in Tanzania, price controls at one time were widespread and transactions or "market" prices often differed from official prices. The national accounts of Tanzania have been based -46 - largely on the official prices because other prevailing prices were often unknown, or illegal and not officially recognized. 25. The Harrison Report (June 1988) looked into this problem for some sectors, and in agriculture, calculated that the official price may be reasonable to use for the bulk of the main cash crops. In some cases the market price is below the official price because official purchasing agencies could not make purchases for financial or logistics reasons, and farmers were forced to sell at a lower price. However, since smuggling occurs in many regions, the prices must have been higher than those offered by the official buyer. Moreover, the existence of a parallel market implies that goods are selling at higher prices. Much of the "rents" associated with the parallel market, however, could be attributed to the commerce sector, leaving the productive sectors reasonably well estimated in terms of prices. III. Methodological Adjustments to the National Accounts 26. The estimates provided by the Bank mission, especially the agriculture value added series, gives an important illustration of how the newly developed series on gross output can be used and integrated into the national accounts. Crop-by-crop intermediate consumption estimates were applied to the gross output series, and the results show a somewhat different picture of the level and growth of the sector in the national accounts. It is important to produce these data and analyze the results to find out if they make economic sense. Various data and methodological issues may only arise once the figures are compiled, thus speeding the process of revision and final integration into the main national accounts. 27. Other Bank estimates of the "hidden economy" based on fragmentary data (such as the trophy hunting or unrecorded rental income based on the Bagachwa study) can be used as a point of departure for refinement by researching alternative/improved data sources and better methodologies. It may be possible to establish or to improve communications with other government agencies in order to develop information for where it is known to be weak. It is especially worthwhile to check into alternative sources and methods in areas for which published official data are not available, such as special University studies or surveys. 28. Methodological adjustments in the industrial origin accounts have been made to agriculture, construction, and other services (subsector of public administration and services). 29. In current prices value added, the adjustments give a slightly lower GDP at factor cost than the official figures. This is mainly due to agriculture value added estimates that are about 5-10 percent lower in the adjusted figures. 30. Mission estimates of construction value added is about 10-20 percent higher for most years than the official estimates. Public administration and services value added have been increased by about 5 percent in the adjusted series. 31. In constant prices value added, the adjusted series overall growth in GDP at factor cost is slightly lower than the official accounts; this is driven by the adjusted agriculture series which shows substantially lower growth rates than the official series during 1986-88. Earlier years show similar trends in the official and the adjusted series (official GDP (at factor cost) growth -47 - from 1980-88 is 1.8%, MA series is 1.4%). Sector by sector methodological adjustments by the mission are discussed in Sub-Annex 2. The statistical appendix includes both official and miss!bn estimates of the GDP by industrial origin and by expenditure. IV. The Paraliel Economy in Tanzania 32. Chaeristics. The parallel economy in Tanzania is a part of the official economy and inter-related with it. Like the official economy it is also integrated into the international economy, perhaps more realistically and more successfully, since this component of the parallel economy has been growing whereas the official economy declined for a significant period. The parallel economy Is now estimated to have reached 30 percent of GDP (Bagachwa). These activities form no part of the presently calculated GDP, even though most parallel market transactions are for cash. Thus, the official national accounts data do not describe the underlying economic reality of the country. The official figures do not capture all transactions and they do not value certain transaction which are included in GDP appropriately. Many of the activities of small producers, especially those involving cash food crops remain "hidden". Furthermore, the official data on prices for most consumer goods explain less than half of the actual prices paid (Bagachwa). Policy analysis based on such data must necessarily be incomplete and unsatisfactory. 33. Origins. The government's earlier response to external economic problems which it could not resolve was to increase the range of protective domestic economic policies and to impose more internal controls. Given already low levels of subsistence, most people were forced into the second economy simply to make ends meet and to preserve their families from the risks of further declining living standards. Transactions on the parallel market, though mostly small scale and in commonly required goods, became pervasive. The effective controls and regulations of the market, however, took several explicit and implicit forms: (i) Official Controls and Interventions; These were introduced to support newly established institutions to regulate production, and introduce a more equitable distribution of output, through an extensive system of licensing, and in an attempt to control economic "sabotage" by statute. (ii) Official marketing Arrangements. The establishment of artificial markets for a number of crops, the payment of low official producer prices and the setting of a wide range of administered prices (for goods which rapidly became unobtainable through official channels) provided a strong incentive for producers to seek alternative and more profitable outlets for their (mainly agricultural) output. (iii) Deterioration of Economic and Social Infratructure. Tbe failure of the government to provide satisfactory marketing arrangements combined with its inability, through lack of resources, to maintain the nation's infrastructure and transportation networks imposed distributional constraints on producers that encouraged them to resort to their own devices in marketing output, especially where such items were perishable or deteriorated rapidly with time and storage. (iv) Shorgg of basic gQgds and services. Basic necessities are required on a regular basis by households to fulfill basic human needs; when unavailable through the usual formal channels, -48 - people acquire them by various means 'under the counter". The government's response to the growing recognition of its inability to effectively manage the economy was to extend the number of controls, Increase the degree of prohibition and the severity of punishment for offenders, and raise the level of taxes. At the same time, inflation remained high. In effect, this only provided a further incentive to the growth of the parallel economy. But these actions have also had the additional undesirable effect of placing more (unofficial) power in the hands of those controlling and policing the system who stood to gain greater rents from exploiting their position through basically unproductive activities. 34. In Tanzania, the formal official economy has not completely satisfied the needs of the population, and many of the participants in the parallel economy bypass expensive, unnecessarily complex and time-consuming licensing, purchasing and production regulations simply to survive. 35. It was essentially the failure of the public sector in general to provide the normally expected range of public services and communal goods, and to adopt a supportive rather than regulatory role (including nationalization) with respect to the private sector that led to the establishment and entrenchment of the parallel economy in Tanzania. Over time, much of what has gone on in the parallel market has been institutionalized. 36. Participation in the parallel economy is widespread and pervasive over a wide range of activities, although, in general, according to well informed observers, the transactions are for the most part small scale and involve independent operators. 37. Parallel markets, including the so called "black market", occur when significant parts of the economy are controlled goods and services can no longer be freely produced and sold, or offered for sale (and thus purchased), within the existing national institutional economic framework. Such a situation arises in part because economies are not sufficiently efficient nor developed enough to permit certain types of production and trade to take place in open markets in the formal, monetized sector. These economies usually devote a significant share of output to non-monetary rural household production for own consumption, and to small scale informal production and service operations. Parallel markets also occur when, through the impact of state intervention markets are not allowed to function properly, if at all. Supply shortfalls appear and economic agents emerge to bring producers and consumers together at a different equilibrium level within a differentiated but not completely separated market. 38. Parallel markets and associated unreported informal activities therefore develop when a gap between effective demand and potential supply in the controlled formal economy is artificially introduced by policy. Whilst unintentional, such a situation was created in the 1970's by the Tanzania government, paradoxically, by deliberate official action to redress what it saw were growing inequalities in society. These were reflected in the increasing differential between incomes observed in the urban wage sector and in the village economy and this policymakers believed would undermine the main principles of African Socialism. The unexpected, and therefore unplanned for, repercussions of fixed price policies, across the board subsidies, and incomes policies led to widespread shortages, and an associated rapid increase in prices in the uncontrolled market, including the prices of those goods which were officially fixed. These economic difficulties were compounded by serious cash flow problems and the lack of adequate "cash cover" in parastatal organizations and the government. In practice, this meant that -49 - suppliers and employees could not be paid, nor - in the case of the marketing boards - goods acquired by parastatal organizations. The problems were aggravated by distributional bottlenecks and domestic transportation difficulties. 39. The distinction between the official economy and the paral;el economy are reflected in differences in the types of labor used, the growth of self employment and differences between capital technology and its utilization. This also leads to differences in the relative factor shares of capital and labor. These differences are manifested in Tanzania's case in the contrast between, on the one hand, a small number of fairly large state controlled production operations and, on the other, a large number of small, privately operated establishments. Each type of activity, for the market it attempts to serve, operates below optimum capacity and is usually established on a scale that can be either above or below what is technically optimum. There is a notable lack of intermediate medium scale operations to bridge the traverse between the two groups. 40. A variety of factors have led to the development of the parallel economy and the informal sector in Tanzania. The parallel economy as it has emerged in Tanzania can be defined as comprising all operations and transactions that do not come under either direct or indirect state observance and control. The informal sector necessarily forms part of this economy, often supplying many of the goods for the parallel market, though it Is not synonymous with it. Informal sector activities go mostly unrecorded, not necessarily because they are illegal or contravene some local statute or by-law, but because they are usually small scale, difficult to identify and often family Qcinship related) businesses. Other operations may be more truly "parallel", having been set up to by-pass cumbersome and frequently counter-productive registration and licensing regulations. 41. There are probably several categories of activity belonging in the parallel economy: (a) Those like small scale private agricultural and manufacturing activities that are genuinely productive, and contribute to an increase in the total output of goods and services. (b) Quite a few productive activities that may not get recorded (because they go through unofficial channels) which are relatively large. (c) Activities that merely facilitate or transfer output, but which earn an economic rent on account of the inadequacies and inefficiencies of the system. These activities, which are in some sense "productive trading" exploit difficulties in the distribution process by shifting goods from one market to another to bring buyers and sellers together at a price above that fixed by administrative flat. (d) Those that are illegal in that they by-pass trading, finance and other exchange restrictions. Some may involve both production and sales and include transactions that deal in contraband or prohibited items. 42. Informal sector operations in Tanzania fall mainly in the first category. Parallel market transactions cut across all levels of enterprise activity, freeing up goods and resources bound in by artificial controls. -so - 43. In the past, in many African economies, these parallel markets were regarded initially as representing a minor, temporary pathological and essentially parasitical growth on the formal monetary economy. But increasingly, as the state lost control of the economy, these markets developed in response to people's basic needs and the failure of the formal official system to deliver the goods and services required. They grew to become areas where important, permanent (and in some countries, dominant) submodes of production take place and, for the majority of households, the parallel market is now where most real economic activity occurs. 44. Before liberalization in Tanzania, an important share of the country's GDP and a large part of household income was generated in the parallel market. This sector achieved an effective dominance over many activities, particularly in agriculture and small scale mining but also including domestic transport and distribution activities. It hence exercised considerable control over both internal output and external trade flows. Measures were taken at that time with the objective to restrict these parallel activities - rather than to free up operations in the state controlled economy. These exacerbated the problem and further undermined the recognized 'legal' economy. Paradoxically, they only helped to strengthen the parallel market. Estimated Magnitudeu f Tanzania's Hidden Economy 45. The characteristics of the parallel economy make it difficult to acquire reliable evidence concerning its scale of operations and to accumulate hard data on the various transactions involved. Information on this sector is thus fragmentary and incomplete. It is rarely garnered from official sources. 46. The thesis put forward here is that in the mid and late 1970s the parallel market was not very large and that it probably accounted for no more than 10 percent of GDP (Green, 1981). It was concentrated mainly in the area of food production and distribution. At that time, the official national accounts probably represented fairly faithfully the magnitude of economic activity in the country. From 1980 onwards, however, there is much clearer evidence that the size of the parallel market grew rapidly and that, even through the period of liberalization, this sector has remained significant and still continues to expand. Only in mid to late 1980s, as an extended range of consumer goods has entered more freely into the domestic market and the severe shortages common in the past have disappeared, has the size of the parallel economy remained more stable. It continues to represent, nevertheless, an important part of the domestic economy, not least because the new institutions and channels of commerce developed during this period are proving more difficult to dismantle. 47. Table A below gives some estimates in US dollars of the probably current magnitude of parallel market operations in goods, excluding commerce (distribution) and related transport and storage activities in 1988. 48. Assuming a conservative 30% distribution and transport margin, the total value of unreported trade transactions could be have been high as around $380 million or 37,500 million Tanzanian shillings at average 1988 exchange rates. The intermediate consumption elemenc of this gross output is unlikely to be very significant for the commodities in question. Given that most of the expatriate component is more in the nature of a transfer which therefore contributes mainly to disposable income rather than to GNP, the contribution of the parallel economy in the -51 - early 1980s to the overall real economy on the most conservative estimate is in the region of $325 million or about 32,000 million Tanzanian shillings, i.e. around only 11 percent of an enlarged real GDP, which represents respectively 12% and 13% of the official GDP. Over time the share has probably fluctuated but only slightly, around a steadily rising trend until it was believed to have been around 30% of GDP at factor cost in 1988. -52 - ble A: TANZANIA ADDMTIONAL COMPONENTS VALUE ADDED 1985-8 1.0 Goods Aicultur 1.1 Cudamon 90% outside 'official' market. (10% - 760 tons) ($4 million '85; $2 million '86; $7 million '87) 1.2 Non-traditional crops, mostly food 1987 $10 milion 2.0 Hutr Trophes (ivory/ides) Ivory $135 million - 1988 Rhino $ 13 million- 1988 (Official sales $3 million) 3.0 Minitn Gold $14 million; other minenals - $10 million 4.0 Ma- ufaiuxn HandictaPws Wood. shell, painting, oraments (sold through Kenya) 1987 - $9 million (for 5 million peak) - say $15 million. 5.0 Sicxces 5.1 Housing Rent $800 - 2,000 p.m. per housing unit. $35 million 1988 Total 'Black Matket" Rent $32 miUlion 1987 $26 million 1986 5.2 Toudan Paralel market from tourism = $70 million 1988 Av. daily exp. = $82-85 (50% = of which lodging) Fo,^otnote: 1) Net Under invoicing imports 1985 Under $177 million 1986 Over = $65 million (This seems odd unless it was to facilitate capital flight). 2) Net Under invoicing of exports 1986 = $49 million 3) Own funded imports 1984 = $252 million 1988 = $638 million Sour: Bagachwa. -53 - 49. These calculations, however, do not take into account unreported transactions in the services sector (apart from the transport and distribution component mentioned above). In addition, therefore, it would be necessary to include the transportation of people (in dala dala and other private trucks and buses), informal banking and credit services, including those serving the tourist industry. In terms of value added these possibly account for further $50 million plus, giving a 4 to 5% addition to the total parallel market share of GDP. V. Institutional Action Program for Statistics 50. Considerable efforts have been expended by the European Communities (EC) and Statistics Sweden to strengthen the capacity of the Tanzanian Statistics office. The former has focussed primarily on the improvement of the national accounts, while the latter has paid more attention to the enhancement of computer operations and dissemination capabilities. 51. The EC, through a series of more than a half a dozen successive ccnsultancies, has laid the foundation for reconciling the production and expenditure estimates in the official set of national accounts. These relate to the formal, official economy. Apart from some still outstanding problems concerned with balancing the public sector accounts, this work is more or less complete. It provided the main foundation on which the Bank recently made its own adjustments to the official series. The statistics office in Dar-es-Sa!aam, with whom the EC consultants worked closely, has no dispute about these numbers, but little has been done with in the adjustments by the planners and policy makers in Tanzania. Furthermore, data improvements and enhanced statistical capability need to be institudonalized and broaden within the BOS, with emphasis given to devoting more resources to improving statistics and the national accounts in particular. The problem is that there is much economic activity going on which is not captured by the official statistics and, apart from the need to broaden the comprehension of national accounts statistics - the foundation of which is now reasonably well developed - there should be some consideration of how the conventional framework ought to be extended to incorporate unreported parallel market activities and under-recorded informal sector transactions. 52. Finally, to promote the activities noted above and to determine furither priorities for the national accounts work program, it has been suggested and we would endorse the idea of the Planning Commission and Bureau of Statistics organizing a workshop of producers and users of national accounts information. In the first instance, such a workshop should involve government officials, the academic community and other interested parties. -54 - Sub-Annex 1 Present Practice of Compilingi National Accounts in Tanzania 1. Aggicult is the main sector in Tanzania (about 60 percent of GDP), and production data is estimated for six major food crops (maize, paddy, wheat, sorghum and millet, cassava, and beans), which are obtained from the Marketing Development Bureau (MDB). Data on official sales of export crops are obtained from the Annual Economic Surveys (AES), and producer prices are from MDB reports. Data for about 25 other crops (accounting for about 40 percent of agriculture crop output) are extrapolated by rural population growth from the 1976 Household Budget Survey (HBS). The sector also includes estimates of livestock, hunting, fishing, and forestry. Constant prices are estimated by valuing quantities of production at 1976 producer prices. 2. The major shortcomings of measurement in agriculture value added are: (a) MDB crop production estimates are not based on actual measurements, (b) export crops sales only include official sales, not production (unofficial sales, own use, etc. are not included), (c) prices for most crops are official prices, (d) no reliable production data are available on the minor crops which comprise about 40 percent of crop output, and (e) intermediate consumption (seeds, fertilizers, fuel, etc.) of 10 percent is assumed to be the same for all crops instead of using crop by crop ratios. 3. Mining includes stone quarrying, salt mining, and diamond and other (gold, gemstones, tin, mica, etc). Constant prices are estimated by applying 1976 sale prices to quantities of mineral output. The Eurostat project has identified some problems with the estimates of stone quarrying (input/output data for 1976 differs substantially from the national accounts estimates); revisions in the national accounts are planned. Estimates of other minerals are from reported sales. For some minerals, such as gold, the official price is far below the international price and partial evidence suggests that much production is sold through unofficial channels, and is therefore not captured in the national accounts. Constant prices are estimated by valuing quantities produced at 1976 prices. 4. Manufacring data are derived from the annual survey of industrial production of manufacturing establishments employing ten or more persons. Estimates of gross output for establishments employing five to nine persons are based on a 1978 survey, and are unadjusted for subsequent years. Partial evidence suggests that between 1978-1986, small-scale manufacturing (1-9 employees) declined significantly from 1986-1989, small-scale production increased dramatically. Estimates of output for firms employing less than five employees are based on the 1976 input/output studies, and are assumed to be one third of the total value added of all manufacturing establishments. Constant prices are extrapolated from the base year via physical indicators of the volume of production. National accounts estimates of this sector are weak since the informal economy is thought to be quite large in small scale manufacturing. 5. Electricity and water data are collected by the Tanzania Electric Supply Corporation (TANESCO) and the National Water Authority. Constant prices are derived by -55 - applying physical measures of energy and water production to the base year estimate. The data used in the national accounts are derived from the accounts of these entities. 6. Constructin estimates include private and public construction projects. Data on construction for central and local governments are derived from their accounts, and for parastatals information is obtained from questionnaires. Information on private construction is derived from extrapolating the 1976 benchmark figure by the increase in cement consumption. Although using other proxy indicators would be desirable, it is difficult to obtain consistent series on specific material inputs into construction. Concrete products account for about 25 percent of material inputs. Other inputs include wood (13 percent), paint (12 percent), and hardware and windows (8 percent). If domestic production and imports can be established for the 1976-88 period for these inputs, this method of estimation can be refined. In the case of rural own account construction, rural population growth is used to extrapolate the benchmark year estimate. Since an important direct source of information for private construction has not been available since 1972, estimates are weak, and are probably understated. 7. Commerce covers wholesale and retail trade, and restaurants and hotels. Estimates of gross output and intermediate consumption are available from the 1976 input/output survey, and subsequent years have been extrapolated on the basis of current price gross output in the goods producing sectors of the economy (agriculture, mining, and manufacturing). Constant price estimates use the 1976 price series of gross output of the producing sectors to extrapolate subsequent years. Data on hotels and restaurants are derived from accounts of the Tanzania Tourist Corporation aTC). Estimating the gross output of the commerce sector is difficult given the absence of systematic data sources, especially since informal activities such as street vending and small scale food caterers are included. The Eurostat report makes some specific recommendations on possible methodological changes (for example, trade margins on imported goods can be separately calculated and added to the domestic goods margins), but the results are similar to current methodology. This sector is underestimated to the extent that the productive sectors and imports are underestimated, and to the extent that other small scale informal catering activities exist. 8. Transport. storage. and communications data for state owned enterprises such as Air Tanzania (ATC), Tanzania Railway Corporation (TRC), Tanzania Harbors Authority (THA), Tanzania Zambia Railway (TAZARA) and Tanzania Posts and Telecommunications (MTC) have been compiled from their annual accounts. The constant price series is extrapolated from the base year by the number of passengers and freight transport, and with a combined index of letters carried, telephone lines in use, and telegrams and telexes. For other parastatal and private enterprises, the benchmark 1976 input/output estimates have been extrapolated on the basis of diesel consumption (constant price estimates) and inflated by the cost of living index for transport for the current prices. Use of a fixed input/output coefficient in the transportation sector would produce reasonable value added estimates during the period of stable oil prices (1976-79) but would have over-estimated value added when oil prices increased during 1980- 1986. 9. Finance, insurance real estate. and business services includes financial institutions, insurance carriers, real estate and owner-occupied dwellings, legal services, accounting, and data processing. Current price estimates of public finance institutions are -56 - calculated by the income approach; constant prices are extrapolated by the number of persons employed each year. The 1976 benchmark of imputed rental value of owner-occupied dwellings has been estimated from the 1976 HBS based on per capita rent and population estimates for Tanzania. The constant price series uses population growth to extrapolate from the base year estimates. Current price estimates inflate the constant price series by an index of rent quotations in urban areas. Eurostat has recommended that this sub-sector be revised to take the 1988 census into account, and to substitute the urban price index for one that is more appropriate for rural properties. Cross-checking the data with rural own account construction is also recommended. Business services are probably underestimated since the employment surveys do not include self- employed persons, and many small new businesses are probably missed. Employment data are only available through 1985. 10. Public administration and other services includes the various organizations engaged in the administration of central, regional, and local government activities. Data are obtained from budget documents and wage bills from the agencies. The 1976 price series are based on quantitative indicators such as number of students enrolled, number of patients treated in hospitals, and the number employed in various agencies. 11. GDP by expenditure accounts are compiled only in current prices. GDP estimates from the industrial origin accounts are used in the expenditure accounts; this is the usual practice in most developing countries. Private consumption is calculated as a residual. This is also quite common, but since the GDP by origin in recent years may be underestimated by as much as 30 percent (see Maliyamkono and Bagachwa, 1989) private consumption as a residual is also underestimated. 12. General government consumDtion data are from the budget, and export and imports of goods and non-factor services are from the Balance of Payments compiled by the Bank of Tanzania. Exports and imports are converted from US$ by the official exchange rate. There is a substantial amount of unrecorded external trade in Tanzania, thus distorting the share of trade in GDP. To the extent that both exports and imports are unrecorded or under-reported by the same order of magnitude, the net effect may not alter the overall resource balance, but the shares of trade in GDP and the relative openness of the economy would be underestimated. 13. Gross domestic investment includes construction and machinery and equipment which are broken down into private (residual) and public investment. Data on public investment in machinery and equipment are from accounts of the central government, parastatal enterprises, and non-profit making institutions. Private investment is a residual between the total and public investment. Official data for total investment in machinery and equipment is estimated by adjusting imports of machinery and equipment by trade and transport margins (25 percent). The official estimates do not include domestically produced machinery and equipment, and it is likely that under-reporting of imported capital goods occurs, affecting estimates of gross domestic investment. Estimates of capital formation in construction are noted in the previous section. 14. Exports and imports of goods and non-factor services are derived from the Bank of Tanzania's balance of payments data. Data are converted from US$ to shillings with the official annual average exchange rate. -57 - 15. Constant price estimates of GDP by expenditure are not calculated by the national accounts office in Tanzania. -58 - Sub-Annex 2 Methodological Adjustments to the National Accounts 1. Because of the many economic changes occurring in Tanzania, the mission placed a great deal of emphasis on the data side. It is always important to know how the data are collected and compiled, to know the limitations for analysis and to help explain possible anomalies in economic measures. New economic models developed by the Bank require both greater detail and a broader range of measures than before, and understanding relationships between socio-economic sectors and agents requires accurate data. 2. Before mission estimates were made, official source data and notes were reviewed to understand what the accounts represent in terms of coverage and methodology. Adjustments are based on information collected during the mission and are undertaken to take into other account available official information not yet incorporated into the national accounts, and some estimates based on a review of other studies related to the national accounts. 3. Since the mission estimates are based on information gathered during a relatively short mission, they are based on fragmentary data, and should be reviewed by national authorities with their special insights. Mission estimates may illuminate some areas that had not been previously covered due to lack of official statistics, and serve to stimulate thinking about how important unrecorded or under-recorded sectors can be better covered. ALricultur 4. Gross output estimates from the Bureau of Statistics' EUROSTAT project provided the basic data for the adjusted series from 1976-85. Data from Bureau of Statistics worksheets were used for the 1986-88 series for all food crops except for 1986-88 for some Household Budget Survey (HBS) crops. For these years, quantities have been adjusted by using population growth, as was the case for the pre-1985 series. S. The methodological adjustments to the official series were: (i) value added/gross output series based on MDB estimates were used to go from gross output to value added instead of using an overall ratio of .90 as in the official series; (ii) the extra 10 percent adjustment in the official series for undercounting has been eliminated, as per the Harrison 1988 suggestion; (iii) cassava production is based on new MDB estimates; and (iv) as noted above, some HBS crops quantity measures have been adjusted. 6. The major weakness of the agriculture estimates (both official and the adjusted series) is that 40 percent or more of production (barley and other cereals, sweet potatoes, yam and cocoyam, potatoes, cooking bananas, other starches, dry peas, lentils and other pulses, shelled groundnuts, vegetables and fruits) is not measured by MDB, and is estimated by the Bureau of Statistics from benchmark figures extrapolated by population growth. Also, oilseeds estimates, like export crop figures, are based on official sales, not on production. These crops may be underestimated by a factor of four or more. There are some new estimates for some of these -59- crops done by the agriculture statistics section, but they do not cover all the crops. Moreover, the BOS survey does not cover cash crops, and although the sample is representative of agricultural zones in the country, it does not represent the farming population. Agriculture surveys to improve these estimates should be done along with the other crops monitored by MDB. Even the 'best" MDB agriculture estimates used in the national accounts are not based on actual measurements of area planted and harvested, and export crops estimates are based on sales to official agencies, rather than on production. MDB estimates are from the Early Warning unit in Tanzania. The data are derived from a model that takes into account estimates of production by extension workers, the weather, and fertilizer availability. Crops covered include: maize, millet, paddy, beans, wheat, bananas, cassava, and sweet potatoes. 7. A comparison of the official and adjusted series for agriculture value added shows tiat, for most years in current prices, the official series is about 5-10 percent higher than the adjusted series. The real growth of agriculture value added in the adjusted series is lower for 1986, 1987, and 1988 (2.3 percent, 3.8 percent and .9 percent) than the official series (5.7 percent, 4.4 percent and 4.5 percent). The differences between the official accounts and the adjusted series for agriculture value added can be explained by reviewing the performance of crop by crop growth rates. As the table below shows, in 1986, many crops such as seed cotton, sugar cane, cashewnuts, wheat and legumes performed poorly. In 1988, declines in coffee, tobacco, cashewnuts, other cereals, and legumes resulted in the low overall growth. -60 - A!riculture Value Added In 1976 Pr1.es - (Percentage Change) Mission Adustgnla 1986 1988 Seed cotton -34.3 20.2 Sisal fibre -6.5 0.3 Tea leaf -7.5 -1.6 Coffee 12.8 -14.8 Sugar cane -11.6' 14.7 Tobacco 4.5 -239 Cashewnuts -32.4 -40.3 Pyrethrum -12.1 14.7 Maize 2.9 6.3 Wheat -14.5 -6.9 Paddy 16.7 9.8 Other cereals 7.3 -21.7 Legumes -16.6 -4.5 Oil seeds -6.2 3.9 Plantains 2.8 2.8 Other hor. crops 3.1 2.8 Coconuts -9.9 2.8 Other crops 4.6 2.3 of which cassava 6.7 1.6 Crops total 0.6 -0.2 Livestock 2.8 5.5 Forestry 9.4 2.0 Fishing 16.7 1.6 Hunting 52.0 35.7 Total Adjusted Agriculture 2.3 0.9 Total Official Agriculture 5.7 4.5 Cantructim 8. Construction value added in the official accounts is estimated from buildings (residential, rural/own account, and non-residential) and other works Oand and roads, water and other). Adjustments to the accounts were made first on the constant price series, and then inflated with either the official deflator if it was appropriate, or by a newly developed deflator as noted below. - 61 . 9. Official estimates of residential construction are based on the 1976 figure and extrapolated by the increase in cement consumption. The real change, however, does not correspond with the figures on cement production; sometimes they move in opposite direction. This subsector was re-estimated using cement production figures (TET No. 2, p. 42). 10. Official ural own account construction in the official account has been estimated by applying the rural population growth to the 1976 output figure. The special notes in TET No. 2 advise that this method may not adequately measure rural construction because it does not take into account the improvement in quality of rural housing. Rural housing is improving (more metal roofs, bumt bricks and cement rather than traditional methods and materials are being used). Rural housing has been re-estimated by (1) using new census rural population growth (about 2.4 percent instead of 2.8 percent), and (2) building in a "quality improvement" factor by adding .16% per year to the population growth. 11. Ng_idntial construction is estimated from the accounts of the central government and parastatal enterprises, so the official accounts are well covered. 12. Much of the investment activity of the peasant farmer's labor to improve future productivity through land improvement, storage construction, and development of tree crops is not included in land improvement in the official accounts. Therefore the land component has been re-estimated by taking 15 percent of the gross output of tea, coffee, and cashew nuts and adding this to the official estimates of capital formation of the subsector land. 13. Other canital formation was adjusted to take into account community based construction of schools and community centers that are not included elsewhere. Ten percent was added to the roads, water, and other subsectors. 14. The current price adjusted series for construction was estimated by developing a construction deflator for residential and nonresidential construction, and using the official deflators for "other works". The building deflators are from the Building Research Unit estimates of building costs indices. These deflators were similar to the official series for most years. 15. Value added for construction in both the official and adjusted accounts is estimated by applying the ratios of value added to gross output developed by the 1976 I/O. Rural own account construction value added/gross output ratio is .43; all other construction is .33. 16. A comparison between the official value added in construction with the adjusted value added shows that the adjusted series is about 10-30% higher for most years. The major increases are in 1986-89; this is consistent with the belief that the construction sector has been underestimated, especially in the past few years. The adjusted constant price series generally moves in the same direction as the official series; however, the 1988-89 adjusted series resulted in higher growth than presented in the official accounts. -62 - Constrion Value Addgd (7 prices Annual Growt 1286 1987 1988 Adjusted Accounts Construction 17.3 2.3 5.0 Residential -42.0 22.4 118.3 Rural own account 2.6 2.5 2.5 Other 27.6 -21.1 59.9 Official Accounts Construction 21.9 5.0 12.2 Residential 15.7 14.6 18.6 Rural own account 3.7 3.9 4.1 Other 25.9 -18.5 50.1 17. The main difference is in private residential construction. Mission estimates of residential construction are based on the TET Vol. 1, No. 2, July 1988, which also states that the considerable expansion and high levels of cement supply suggest that both the growth and level of construction are underestimated. GDP by Expenditure Accounts 18. The GDP by expenditure accounts have been adjusted only for gross domestic investment. Although some adjustments may be necessary in the external accounts (to account for own account import funds) these adjustments have not been made in the accounts because the data are not available before 1984 and a substantial break in comparability with previous years would affect other %;omponents such as private consumption. Further work needs to be done to develop a consistent series of the external transactions. 19. C,ital formation. or gross domestic investment in the expenditure accounts includes buildings, other works, and machinery and equipment. Adjustments to the buildings and other works are described above in the construction section. The adjustments noted here pertain to the machinery and equipment category. 20. The adjusted series of machinery and equipment also uses imports of machinery and equipment (marked up 25% for trade and transport margins), and adds to this gross output of domestically produced machinery and equipment. These data for 1976-85 are from the exercise of gross output estimates (part of the EUROSTAT program); 1986-88 are estimated to increase at the same rate as manufacturing value added. Refinement of the 1986-88 estimates could be done by substituting machinery and equipment production for overall manufacturing if the data were to become available. Constant prices of imports are derived by deflating the current price US$ series by the MUV index (adjusted for the exchange rate). -63 - 21. A comparison of the official and adjusted series shows that the levels of machinery and equipment are substantially different; the official series are generally lower than the adjusted series by 15-30 percent for most years except for 1988 in which the official series is higher by about 30 percent. The adjusted series is higher because of the inclusion of domestically produced machinery and equipment. Deflators are similar from 1976-43, but the adjusted series (based on the MUV index) is higher after 1984. By 1988, the official series is at 1,107, while the adjusted series is 1,494 (1976=100). 22. As is the case in any country with high inflation, there is difficulty in deflating the current price series of investment goods. For machinery and equipment, the official accounts were deflated by the exchange rate index for 1986-88 since specific deflators were not available. This is the major contributor of price changes in shillings (depreciation of the shilling was about 90 percent for 1986, 100 percent for 1987 and 50 percent for 1988). However, the MUV index in USS increased by 18 percent in 1986, and about 10 percent in 1987 and another 8 percent in 1988. Since this wasn't included in the official deflation method, the official series deflator for 1988 is about 25 percent lower than mission estimates. .3. The current price series is similar for these years, so differences in deflators result in differences in the constant price series for machinery and transport equipment. Capital Formation- Selected Components (1976 prices) Annual Growth 1984 1985 Adjusted Series Mach. & equipment 1.7 17.6 Transport equip. 8.2 3.9 GDI 3.7 6.4 Official Series Mach. & equipment 64.3 34.5 Transport equip. 3861.2 34.5 GDI 36.7 20.7 24. For GDI overall, the official and adjusted series levels are quite similar; the official series is about 10% higher for most years, with only 1988 being about 10% higher in the official series. The two series shows very different growth rates for some years. 25. The adjusted series deflators for imports (based on the MUV, adjusted for the exchange rate) and for domestic series (from the BOS worksheet on gross output) increase at a higher rate than the official series for these two years, making the real increases smaller. The overall official equipment deflator increase is about 12% for 1985, whereas the adjusted series is about 15% for the imported component, and 21% for domestic machinery and equipment and -64 - an additional 100% for domestic transport equipment. The domestic component's increase is what pushes up the overall deflator in the adjusted series. 26. The constant price expenditure accounts have been estimated by using the GDP from the adjusted industrial origin series (adjusted to include net indirect taxes) and deflating the general government consumption, exports of goods and nonfactor services, imports of goods and nonfactor services, and gross domestic investment with private consumption as a residual. The deflator for general government consumption should be used with caution since it has been derived by using the implicit deflator for public administration. The physical indicator used to estimate the constant price series of public administration value added is number employed. With problems of absenteeism and curtailed working hours, real labor input is likely to be much less than the present physical indicator suggests. The external trade deflator is based on the Bank's trade deflators unit values for 1976-88. The gross domestic investment deflator has been developed from mission estimates of domestic deflators for capital formation in buildings and domestic machinery and equipment and imported machinery and equipment. - 65 - Sub-Annex 3 "Hidden Economy" Adjustments tQ the National Accounts 1. "Hidden Economy" (HE) adjustments were made to the following sectors: (a) agriculture, (b) mining, (c) manufacturing, (d) trade, hotels and restaurants, and (e) finance, real estate, and owner occupied dwellings. Table 2 shows estimates of "HE" adjustments. See below for brief descriptions of the adjustments. Agriculture 2. Adjustments were made for oil seeds, cardamon, other food crops, and trophy hunting based on the Bagachwa study. (See Tables 5, 6, and 7). Mining 3. Gold is one important component that is believed to be underestimated. Customs realizes that there is much gold smuggling out of Tanzania since the official pAice of gold is one fourth to one half the international price. Newspaper articles quote officials of Dar Tardine Tanzania (DTM) that they will be purchasing 50 to 100 kilograms of gold per month, compared to the official purchase of 1.8 kilograms during 1988. Using estimates from the Bagachwa study, value added for gold and diamonds and gemstones have been adjusted (HE) for 1976-88. Gross output is estimated by assuming the price is about 70 percent of the international price, and value added/gross output is .95 (from BOS) estimates. (See Table 10). Manufacturin 4. Estimates for handicrafts items have been incorporated into the HE adjusted series. These estimates are based on the Bagachwa study. Other adjustments to include unrecorded small scale metal working, etc. enterprises should be made if appropriate data become available. (See Table 11). Trade. Hotels and Restaurants 5. Estimates for tourism and distribution have been added to the HE adjusted series. The Bagachwa study shows that only about 25% of tourist expenditure is officially recorded. Based on this assumption, and applying the "hidden' expenditures to the number of visitors and length of stay, a revised series was estimated. (See Table 13). Finance, real estate. owner occupied dwelling 6. Estimates for unrecorded rental housing (mostly to expatriates) have been made in the HE adjusted series. The Bagachwa study estimates that unrecorded rents were about $25 million in 1988, $22 million in 1987, and $18 million in 1986. Assuming that previous - 66 - years were about $2 million less each year down to a minimum of $6 million per year, estimates of the series from 1976-88 were made. (See Table 12). -67- Sub-Annex 4 Comnonents of Eurostat's National Accounts Proect 1. Since 1986, the national accounts office has benefitted from support from Eurostat to improve the national accounts. Given the limited resources in the national accounts office, the project has focussed on using existing dsta sources and improvements in the methodology. The sub-projects include (1) compilation of output and intermediate consumption of all industries based on the 1976 110, (2) calculation of value added at market prices, (3) implementation of double deflation method, (4) improved deflation method for activities of general government, (5) new method of compiling consumption of fixed capital at replacement costs, (6) evaluate compilation methods for insurance, financial institutions, etc., (7) direct computation of private final consumption based on commodity flow method, (8) compilation of fixed capital formation by the commodity flow method, (9) computation of private final consumption at constant prices, (10) compilation of gross output by sector in current and constant prices, and (11) computation of the increase in stocks at annual average value and at constant prices. 2. The reports from the Eurostat missions have been thorough in providing information and adjustments to the consistency of the 1976 I/O and the national accounts, as well as documenting methodological and data inconsistencies in the national accounts. The national accounts office should strive to complete as many of the projects as soon as possible and incorporate the results into the national accounts. The mission recognizes that the projects listed above are quite ambitious given the available resources. In order to complete the projects within a reasonable time, it might be best to decide which projects will deliver the highest benefits that can be accomplished within the given time frame. These could then be integrated into the national accounts rather than awaiting the results of all sub-projects. Other sub-projects could be incorporated at a later stage without posing serious problems of revision. The information in the various sub-projects is interesting and useful only insofar as they represent the most meaningful picture of recent economic performance that can be analyzed and digested by policymakers. 3. The specific EUROSTAT projects include the following: Koszerek oject: Derivation of price indices for exports and imports. Scheduled to have been completed by April 1989 but not yet submitted to SOEC. Revised Munnsad Project: Measurement of output at constant prices. Subrnitted in time series to SOEC; updated version to have been submitted in August, 1989, with versions in gross output data in current prices. Harrison Proiect (1988): Collection of gross output at current prices in time series format. Time series revision carried out during Mr. Staglin's visit. Harrison Project (1989/fl: Calculation of sales by industry and domestic supply at current prices in time series format. Completed in August, 1989. Harrison Proiect (1982/ID: Derivation of estimates of household consumption in time series format. Scheduled for completion in September, 1989. -68 - StagUn fict: Estimation of fixed capital formation by the commodity flow- method for 1976 and 1985; completed in September, 1989. -69 - Sub--AnnMI Futum Work ProFmm 1. The special funding for this national accounts project with Eurostat ends shortly, so it is Incumbent upon the Tanzanian authorities in the Bureau of Statistics' national accounts office to plan for the next phase of improvements in the national accounts. 2. As mentioned above, the current project to improve national accounts is generally limited to using existing data sources and improvements in methodology. There are, however, gaps in the existing basic data sources, so it will be necessary to develop new sources of basic data. This was well understood and articulated by BOS officials, and is mentioned in the note on sources and methods on the national accounts. New sources of basic data can be developed through gteater use of administrative data (BOS may offer technical assistance to other ministries in the collection of these data), censuses and periodic surveys. There are already some surveys in the pipel!3e such as the industrial census and labor force survey which are presently ongoing and the Household Budget Survey which is tentatively scheduled for 1991/92, so other projects to improve the basic data will have to be coordinated with planned projects. 3. Improvements in basic data collection, compilation, and analysis are needed in several areas. Priority areas can be identified both by data producers and users to ensure that social and economic policies are based on timely and relevant data. This is especially the case during transitions and adjustment, and is made more difficult since the link between social and economic data becomes more crucial. 4. Some impootant areas for future work include: dricultur-(a) improved crop coverage and data based on actual area planted and harvested, (b) estimates of marketed and non-marketed production, (c) estimates of other agriculture income, (d) breakdown by sex of ownership, hours worked, etc. for women in development focus. The new agriculture surveys being undertaken by the Bureau of Statistics agriculture section could be coordinated with survey work of other agriculture agencies to agree methods of collection and analysis and to avoid duplication and the proliferation of agriculture output estimates. Given agriculture's importance in the Tanzanian economy, and the difficulty in collecting accurate data, this sector deserves special attention. In this context we understand the BOS will carry out an agricultural census after the completion of the household budget survey. This will up date the previous 1971/72 census. informal/hidden/unrecorded/parallel economy-difficult to define and to measure, but important to quantify to fully understand the economy and its changing structure. Various studies have described the multi-faceted informal economy, but due to the lack of good benchmark data and frequent surveys, little has been done to quantify it. Specially designed surveys are needed to begin including estimates in the national accounts. Coverage of commerce and other services sectors will be improved as a result of work in this area. construction and capital formation-estimates are important since they play a key role in economic growth. As administrative information on private construction becomes more -70- readily available, collection and compilation systems need to be developed to process infonnation. Data on other agriculture construction and improvements may be picked up from agriculture surveys. employment data-covers only the "formal" sector, or only about 6% of the working age population (15-64 years), and the data are available after a lag of about 4-5 years. Since the informal economy is quite diverse, it is important to collect aad analyze data on all persons in the labor force in a timely manner for improved social and economic policy formulation. GDP by expenditure in constant prices-are not officially calculated by the national accounts office. Mission estimates are made since analysis is done on real variables. Private consumption is still a residual, but the national accounts office is working towards developing an independent estimate of private consumption in current prices, and deflator for this component may be developed for the constant price series. Base year revisio-the current base year of 1976 is out of date by more than 10 years. If the weights are fixed from a distant past as the economy evolves, real changes in the economy are not well represented. 5. Finally, it will be important to gain as wide a perspective on the future work program as possible. In this context we would endorse the idea to hold a workshop at which both producers and users of Tanzanian National accounts information could discuss priority areas for data collection and dissemination. Certainly it would be important that the primary participants be staff of BOS, the Planning Commission and other government ministries and bodies (e.g. MDB) as well as other users of national accounts data such as professors and others at the university, the local consultant community, etc. At that workshop it could be discussed whether to plan further such seminars, perhaps with the participation of representatives of organizations and universities outside of Tanzania. Such a meeting should be chaired by the Planning Commission. - 71 - Tbbl 1 CDP ff7 DOBWAL GUM9 11 NJ333N3 1J WD 1W 198 IIK 100 234 :19 1S9 I8 193 UP kbV darial O*rign (11 Ion mu Tdh) caatim, ex. 9,149 20,625 32,72 14,M7 17 3 31,6119 3,8 U,6= 4,6 00,28 0,495 214,31 12,795 lin'lt 24 gn 96 21 832 41 884 34 37 41 US 687 1,8 1,9 lhuirw1,o 2,71 8,474 4,8 4.5 4,881 4,7M 4,829 6,27 6,M9 7,A 9,8 9,968 21, Ete & r ZS 2 4 31 275 423 423 4Z 14 R1 1,42 1483 2,Q9 4Ae2 CaGnrolon 38 1,231 1,32 1,83 1,4 1.778 2,119 1,60 2,A 2,11 8,301 4,856 0,43 Tr4aheosL 8,= 8,817 4,43 4,W99 5,477 6S28 7,.16 9,24 21,669 25.829 2,69 23,965 47.21 T-uoc. 1,65 1,798 1,9017 2,218 8,829 81,3 S8;86 8, 8 4,789 7,3 7,797 14,1S 28,2W Fninm, 4c. 2511 2,484 2,72 SAM8 8,788 4,W7 4,916 5,872 6,.31 6,.U 6,68 12,388 16,171 PAb adn,*c 2,480 2,781 8a, 8,84 4,2 4,90 5,74D 7,7# 9.37 11,12Y 12,94 1833,4 28,86 of hudc: d;h W. 881 878 444 8 68 78 1s 962 1,21 1,=2 1,1n 1,369 1,494 - b d(420e4d) c(568) (d8o) 064) (18) (t,2w) (18 (1,6U (1,G1 (1.8) C,5) (8,444) (C22. [P fec fkField s)/A 22,0 3,229 2,16 833,8089, 45,871 64,29 563J 3,177 29,27 141,70 194,6A 2,436 Oficial MP fc 21,A6 23,68 23,2 S2,8W7 87,454 481,5 62,95 602,O 76,148 2188 10,788 IW,939 2n,718 Send mc 644 Q*12 74 0 6,7 8 9,262 12,8 25,484 28,642 12,9 31.,212 41,17 la IN up ft (dt.U a B)p 22,5 23.68 S1;S52as,3 4S,716 3,24 a8,9 76,9 97.192 285,256 264938 ,13583 17 CP fc O)qP fbPa/4 (M (164) 1,79 5.391 7.8 29,S64 26 0229 1 ,034 18,1 3. 452 40, 8 88 O,OM *4aparn of CP ft lo1le_ M & B ilbbmiy clal 3J 23O 23.7 22.2 224.7 2.0 23.5 1n.8 22.8 13.1 129.8 2S.9 2S5 M a B u.*k* Field b. 98.7 97.9 23.1 118.4 17.9 223.1 12.7 215.7 12.1 323. 23.4 2.2 289.6 UEP BY tstrlal Origin (l II lanmof 19M Tah)2a/ AiowIlbws oc. 9,149 Q.1 9,2J1 9,43 9,73 9,918 9,3 0,938 9,28 90,0 11U9 11,S 126 Mlinir2 2 2 2 3S Z 2 229 28 28 294 28 28 172 26 2s Wbrawfa*ard 2,971 2,732 2,88 2,92. 2,835 2,618 2,456 2,22a 2,262 2,183 2,24 2,193 2,812 Ehc. A winr 219 244 28 818 40 417 43 428 439 481 4G2 82 637 Cinicuon N8 1,P2 894 1,51 1.4 1.914 1,964 6s 79 729 87 1.38 TndmhU./lrst 2,968 2,83 2,86 2,960 2,94 2,8U 2,7B1 2,65 2,69 2,711 8.39 8,23 8,438 T.uwt/Camu. 1,ff 1,3 1,669 1,684 1,828 1,06 1,694 1,478 1,48 1,19 1,104 1,tS 1,69 Fl, etc. 2,e2 2,121 2,94 2,7 2,614 2,54 2,7W 2,8 8,3 8,232 8,617 8,38 4,Z5 P=n,dca 2,485 2,32 8,12 8,889 8,83 8,71W 8, 3786 8,712 0,8, 8,418 8,487 8,68 do ddh,: dth r. 381 N87 429 40 479 412 896 43 467 so 541 s6 648 -p*W h.& dg. C (40* (485) (OR) 1) 8) PM 7M Q5 PM () ( (0 P te Q* Field ea)A 98 22,838 92,118 23,38 24,781 24.22 24,174 23,467 24,13 24,08 3,88 3,GM 3.,U8 Official UP ?a 21,62 21,739 22, 22,869 23,419 23,81 3,439 22,8i2 28,= 24,278 25,9 2S3,2 13,69 Sou:d m= /I aun j 64 0 2,12 4,87 5,7t 6,73 6,37 4,881 5,748 6,887 7,816 N % UP ft (f a/8 22,20 21,789 24,851 1,3 29,210 29,3 8,868 V7,763 23,44 8,6156 82,4 85,6 82,1#l GOP fe 0f-CP fcO/4 -27.8 4b.6 1498.67 81.18 443.27 X.32 A 72 4U.67 84 87 .28 798.6998 L.872 8689.8 38W4.5 rbmi on d DP ft ginth i1_- Official P ftc 0.4 2.1 2.9 2.6 -0.6 0.6 4.4 8.4 2.6 83.0 8.9 4.1 UP ft (u F-ied a) 0.1 2.4 4.1 2.1 40.2 0.4 4.0 2.9 2.9 2.1 6.0 6.7 GP fC (IFeOM A) -2.6 12.1 28.6 6.6 2.1 1.7 -4.5 5.9 4.1 5.0 8.6 4.1 Srm nd rdtm: A1 Brk minion e1I by rlnludlr uc ologmaIl d end anmr 4i- t. Bek fitld tl. irc lnucd ulr 6bdglcal mnd demy q lila ftr: agrlcultu, mlnirg, huudicr , uabuftion, ad tourlum ssbme bamdd on " 9b*n no tla*1wqw by US.0bDn. .. 1uu, nd 0.0. 143, ~br, 199 (dut) /1* bnicn mii. 1i by mca lncludirB mcU,dol. mu zmmd .um antam. 82 'Thl Ssad conue 7n Tuma,llimknen@ A Bspdu1, 2899, Tdul 2.8 1imtm for 291068; 288748 ammod to hmw inarmad to 185 o? official t7in /a Ofcial UP f.c. plu IbI id /4 k=w not mm for b in 1F asbimats. lhu diffwmu In B1* field etinta md tIm lV l l1ko mud B*dmelabsim dhould be IncluXd In smob nob etmatd by bi fld dlsw imu dml. uVb, aP on, mudu. iSbn ^arutpi Plrl by P t.. il* o r - 72 - TaI- 2a CP BY W DOM1tX RIs l4a L X1 11O114 AJBN (A I I| ta ofcwn T P4p 1 of 5 Official Accou*a Akrlculbo, ote. 9, 11,11 2,13244,7W 38,5 p,N :5,446 32,787 41,25 61,281 94,W132,9411 8,25 Minirg 214 94 8 29 W9 29 U 20 87 2S1 474 645 7M r1moacrbi. 2,61U 8,7 8,93 8,00 4, 4,11 4,851 4,98 5,=61 6,665 8,1i 9,497 11,93 Elec. &A -r 2s 54 5 2V 424 42 42 514 6 1,i71 1,498 2,9Q 4,U2 Contruntoln ON 1,3 1,661 1,22 1,438 1,2 ,s8= 1,212 1,68 2,0t 8,181 8,45 4,61 Tm l / 2,89 0,41W 3,63 4,844 4,718 5,479 6,614 0,146 15,447 14,12 19,478 23,#11 88,724 Truiqurt/cat. 1,M85 1,795 1,917 2,13 8,829 8,18 8, 83,W 4,70 7,?11 7,77 14,124 18,7 Finmm, dc. 2,84 2,419 2,66 2,978 8,744 4,7 4,851 5,0M S,A 0sU 8,1 21,51 12,5 PIh an,*c 2,8 2,8 2,98 2,67 8,842 8,Q3 4,72 5,445 7,872 6,614 19,736 10,145 12,9W 18,M of ichtd thr 218 24 8 3 41 457 581 65 7 M 817 T8 9S4 7IA.d brA dg. (424) (GM (do 0) (6) (1,15 (1,$) (L,2O (1,51) (1,95 (2,544 8,444)22,9M UP * f.c. 21, ,6 25, 5 20 32,K7 17,454 41,961,548 61,608 78,148 19,83 140,M 12,919 271,718 Minion Eutb heiuinm n -i 11,Mhadlcwu* f 4/)V Agricalture, de. 8,546 9,09811,42 38,51 15, 19,141 24,8i 81,468 18,9 67,4M 78,929138,54I,143 Uinlra 214 48 22 294 82 23 US 249 837 2S 474 e6 7m Mdacbiln 2,611 87 8,s8= S,= 4,7 4,S1 4,11 4,090 5,9il 6,05 8,51 9,457 11,Us Elec. A mabar S9 24 51 276 424 423 4E 614 a1 1,i,1 1,468 2,93 4,0 Camtrucbl0n 97.8155.56 131.611.7 18.7 3775.5 2128.6 319.7 M.2 291 893.8 466.2 6486.8 Tmd1ie6IN f/w* 2,03 81 3 8,30 4,84 4,718 5,476 6,914 8,14 31,447 14,196 9,47 23,931 88,724 Tewiqurt/cai. 1,05 1, 1,97 2,128 18AO9 $,1=8 8,89 8,37 4,78 ,51 7,77 14,124 18,27 Finin etc. 2,84 2,419 2,63 2,98 8,744 4,987 4,89 6,252 6,9 6,63 8,67 , 12,987 Peba On.c 2,408 2,781 8,1 8,614 4,1 4,95 5,74J 7,718 9,7 11,167 18,94 18,194 28,618 of viii: hma 8o 1 M 88 444 6. 781 825 91 1,11 1,1 1,671 1,86 1,484 * W dg. (424) O) (1 0 4) () (1,35) (1,83) (1,) (1,511) (1,931 (2,544) 0,444)02,994 W L *c. 21,18 24,7 6,7 881,574 V7,3 48,116 61,4* 91,;1 75,78 39,2W 134,6 1B182,6 356,M Mimic. bbimb (Jsudlrn HiOn Eianv gQ)8/ Aorioultum, dc. 9,149 1,675 12,872 14,78 17,Vs8 1,5393,87 8,662 40,68 6,258 80,496 14,e1 172,79 fllnlra 94 2671 S 861 418 854 867 481 888 687 1,Sg 1,389 lRIacbrw4e 2,97 8,474 4,079 4,8M 4,881 4,788 4,603 5,147 6,2015 7,6 9,=8 9,948 1,682 Else. WANr 2U1 5 28 2VS 424 423 421 614 S1 1,671 1,498 2,699 4,82 Canlrubcban 978 121i 1,1 1,811 1,654 1,7I 2,U9 1,60 2,S8 2,901 8,63 4,M6 6,480 Trho8sleA/n, 8,38 8,s 4,425 4,979 6,477 6,28 7,86 9,184 11,639 15,619 21,68 23,4 47,911 Teient/M. 1,6 1,798 1,97 2,1138 8,89 3,188 8,89 8,317 4,769 7,I1 7,7 14,32 18,7 FI.inm, ic. 2,51 2,454 2,723 S,A 8,78 4,5W 4,97 5,872 6,51 6,911 8,e67 12,W3 15,17 Pib aun,nc 2,46 2,781 8,1 8,14 4,167 4,5 6,745 7,78 9,W 11,157 10,4 13,894 16,63 Of dhl1h: oth Sm W 878 444 On1 68 781 ON 942 1,1 1,11 1,271 1,862 1,484 kpAhd brk dg. (424) 1 ( 0 4)O (16) (1,15 (1,81) (1,20 (1,511) (1,98) (2,6U) (6,444)(12,8" P at f.c. 22, 32,,23 2,3,1815 83,1 S9,CZ1 45,1 54,197 65,09 80,177 119,287 141,79824,0W 280,485 Official: Fe*=u* $bm AOriault, 45c. 41.6 48.8 4.8 46.6 44.4 48.8 60.8 62.8 E2.8 68.7 E9.8 61.7 06.6 Minir 1.0 0.9 0.6 0.9 0.9 0.7 6.5 0.4 0.4 0.2 0.8 9.8 6.8 hhwufactwi, 18.0 12.8 18.5 22.0 10.9 10.8 8.8 7.8 7.6 6.2 6.1 4.9 4.1 Elb. I auto 1.O 1.0 0.9 9.9 1.1 1.J 0.8 9.8 0.7 1.0 1.1 1.6 1.6 Cenatiu:bIo 4.1 4.8 8.7 8.8 4.0 8.7 8.6 2.9 2.1 1.9 2.2 1.8 1.7 TrOA1itJI/t01 38.1 1.8 318.6 18.4 12.6 12.6 18.0 18.0 18.4 13.1 18.8 12.4 14.8 Trimmo /m. 7.8 7.0 6.7 6.5 8.1 7.1 6.6 6.6 6.1 6.5 6.6 7.8 8.7 Flrwico *a. 9.4 9.4 9.4 9.2 39.0 10.8 9.8 8.4 7.7 6.2 6.8 6.7 4.8 PFb a&tn,gc 10.8 10.1 10.1 10.8 10.6 10.8 19.4 11.6 11.6 9.9 7.2 6.7 6.9 Of dilih: oth_ M 1.0 e.9 1.0 1.2 1.1 1.1 1.0 1.0 0.9 0.7 0.8 0.6 0.4 ImWd brk di. 4.O -2.1 4e.4 -4.6 42.6 42.6 4Q.6 4.1 -1.9 -1.7 -1.8 4.8 -4.7 (M f.c. 1O 1O 13 13 1 10 1 13 13 23 13 23 1W.0 P_wtp Sum re# len 6stim SE) lgaltb, *r. 49.0 45.0 41.1 48.1 42.8 44.4 48.8 98.8 98.1 54.6 57.1 69.6 62.4 MIn It 1.9 1. 0.8 0.9 0.9 0.7 0.5 0.4 0.4 0.2 0.4 0.4 0.8 hkredarIu.2 18.2 18.8 18.9 12.8 11.1 10.4 8.5 7.9 7. 86.8 6.8 5.2 4.8 Elcc. Aubar 1.0 1.0 0.9 0.9 1.1 1.0 0.8 0.8 0.7 1.0 1.1 1.6 1.6 letnaoIln 4.6 6.1 4.8 4.4 4.6 4.1 4.1 2.5 2.7 2.6 2.9 2.6 2.6 T1*sIu/iL. 13S.8 18.8 14.0 18.8 12.7 12.7 18.2 18.2 18.8 18.5 14.4 18.1 15.2 - 73 - POP 2 oe s 16 101t 78 29 1339u 198Xt61 urn im u r un u Tport/eanu. 7.9 7.2 0.9 6.7 6.1 7.8 6.0 5.7 6.8 6.7 5.8 7.7 7.2 Ft,n etc. 9.5 9.8 9.7 9.4 10.1 10.4 9.6 8.6 8.0 6.8 6OA 0.1 6.1 Pib Jin,tc 11.6 11. 10.9 U.S 2.8 11..6 1.1 12.4 21.9 10.6 7.9 7.8 6.5 of lchdi: oth tr 1.6 1.6 1.6 2.0 1.7 1.7 1.6 1.6 1.5 1.1 0.9 0.7 0.6 Ib i bW* dg. -2.0 -2.2 -2.6 2.7 4Q.6 -2.6 4.6 -2.1 2.0 -1.7 -1.9 -8.6 =6.1 CaP at f c. IN lo 2 lo is 1 16 us - 2m 16 2EJ P'rwitgm Shmr QIalcn Eutiat. w ) AgrisultAw, utc. 49.5 4J.7 41.6 48.6 48.6 45J 48. 61.3 0.7 54.7 6.8 68.6 61.6 Mining 1.1 1.0 0.9 1.0 1 0.8 0.6 0.5 0.5 0.8 0.6 0.0 0.6 IiWfacbrlrg 18.1 18.2 18.6 22.1 10.9 31.4 8.5 7.8 7.8 6.4 0.4 6.1 4.2 Ebc. *rati 1.0 1.0 0.9 0.8 1.1 0.9 0.8 0.8 0.7 1J 1J 1.5 1.4 Contrnaan 4.8 4. 4.0 4.1 4.2 8.9 8.9 2.8 2.6 2.4 2.8 2.4 2.8 Trm*j%ftlvsIs 14.8 14.7 16.0 14.8 18.8 8.7 14. 14A 14.6 14.2 15.8 14.5 16.8 Trupsommn. 7.5 6.8 6.6 6.8 7.0 6.9 6.8 5.8 6.0 6.4 5.5 7.8 6.6 Finno, eto. 9.2 9.4 9.2 9.0 9.6 10.0 9.2 8.2 7.8 6.8 6.1 6.4 6.4 P&b adun,es 10.9 10.4 10.8 20.6 10.0 10.9 10.6 1.7 U1.2 10.1 7.5 0.9 6.9 of idich oth tw 1.5 1.4 1.5 1.8 1.0 1.6 1.6 1.4 1.4 1.1 0.9 0.7 0.6 31ut b* dig. -1.9 -4.1 42.8 4Q.5 42.4 -2.5 4Q.5 -24 -1.9 -1.6 -1.8 4.8 -4.6 CDP at f.o. INA 1. 1. .0 A iS 10 1 0.0 16. 1A 2INA aP by ad;trihl Origl Cal I hI Io 1 N Th Arlculturs, dA. 9, 9 , 9,23 S.63 g.. g.1 gui 9,614 10,82 10,11 U,57 126 12,M Miniln 214 m1 uM 12 9 13 191 74 1 174 154 14919 emuaoturlro 2,611 2,641 2,71 2,11 2,6118 2,164 2,16 2,23 2,015 1,611 2,0t5 2,107 Else. LAtwr 219 244 818 46 427 40 418 48 41 41 2 W &-unu- o4 01 73 a 9 US 63 am 610 6 1 71 m 75w Tem*4WmIu/rsL 2,80 2,71 2,m 2,03 t,o= 2,75 2,6 2,1' 2,66 2,63 2,63 8,212 8,31 Twt/omnam. 1,6156 1,03 1,63 1,68 1,62 1.6 1,668 1,478 1,42 1,1 1,0 1,61 1,652 ˘inmntdo~a. 2,e 2,191 2,63 2,83 2,468 3,63 2,1 2,7 2,684 8,01 8,82 8,8 8,10 l " an,uA 2,842 2,41W 2, 8,216 8, 83s 8,63 8,658 8,654 8,61 8,226 8,248 8,848 dohl;: :oth r 22 248 210 9 2113 254 5 234 87 88 849 219 3pkW hk dt. (G (GO (4) 0 (14 ) (CM 1O 0M (05) ( MP P.o. 21,322,75 ,22,930849,419 3,8061 2t,41 92,0112 2 4,1,28 3.623, 3S,9W 7,69 Mission Eint=Am (Inluding Mmft&IoW AMJab Dq)t/ bolzultwo, te. 8,646 5,66 8,46 0,760 8,684 9,218 9,2 9,83 9,60 10,14 10,451 10,03 10,M88 Minir 9214 31 29 209 16 196 12 174 13 174 154 149 28 I_rf.ctulng 2,811 2,01 2,718 2,821 2,888 2,82 2,*4 2,15 2,23 2,015 1,91 2,0 2,10 Fic. A brw 219 244 96 US 46 417 425 41 4180 481 452 82 W8 Corutmon 97 1,02 609 1, t 1,4 1,114 1,044 63 7 719 8Y O7 1,038 TrdihotnhJ/rm. 2,68 2,7S2 2,797 2,83 2,3$ 2,75 2,66 2,612 2,645 2,03 2,63 8,112 8,238 Tiwpt/cnm. 1,65 1,02 1,09 1,684 1,828 1,3 1,694 1,478 1,432 1,1 1,4 1,iS 1,0s3 Firm ctc. 2,e8 2,03 2,3 2,888 2,483 2,820 2,73 2,617 2,94 8,04 8,813 8,895 8,6m Pwb Mn,tc 2,415 2,612 8,23 8,83 8,169 8,73 8,6 8,66 8,712 8,06 8,418 8,487 8,0 ofihld:othmr 881 S 87 426 4 479 412 86 46 457 63 641 648 548 Iq#d b.* dig. (4G (4G (450 (Wi) (55 0W) 14 OM CM CM (OM (GM GP tf.c. 21,=321,W8 21,21522,M 33,2a9, ,233, 22,6 23,178 29,89 24,27 35,182 3,828 UMilon Eml (Incluing Hi1dn k4ud A4iubmt IQ)8/ Abrculb", etc. 9,149 9,178 9,21 9,4S90,73 9,188 9,83 9,068 20,6S 10,845 11,3 2139 12,0849 MlMni 26 ES 249 62 218 213 194 26 19 172 167 US ikuiPcturlri 2,971 2,72 2,60 2,9il2 2,13 2,618 2,485 2,21 2,252 2,23 2,10 2,12 2,812 Else. LAtur 219 244 26 US 46 417 42 418 48S 48 432 522 S3 GlnstruUon 978 1,03 6N 1,' 1,Ot4 1,614 1,044 am 7W7 719 8W7 1,03t TM*hotslu/rsto 2,90 2,3 2,89 2,940 2,944 2,811 2,781 2,65 2,098 2,711 8,83 8,20 8,489 T /cs. 1,01 1,862 1,63 1,0t4 1,813 1,3 1,694 1,478 1,482 1,611 1,814 1,6C1 l,F6 Finu~s, ctc. 2,01 2,23 2,244 2,877 2,614 2,6C4 2,71 2,988 8,03 8,12 8,517 8,83 4,22; P1b shin,dc 2,490 2,612 8,2 B,309 8,8Y 8,7 8,097 8,6 8,712 8,8i1 8,416 8,407 8,637 of tdi: othr 881 876 429 4B 479 412 8U5 42 467 6l 641 648 543 kutAd brm dg. (4G (4G (495) (1) 51) 0 (W) MM CM 0 0) "0 (am (GM^ P^ f.c. 22,8 22,81 22,3,8 23,610 24,818 24,272 4,174 2,447 24,23 24,81 23,807 2,62 23,118 - 74 - w By MM. amm am m Offclet Aaesfsa ArIoultum, da. 1.1 -1.7 0.8 8.9 10 1.8 2.9 4.0 .0 5.7 4.4 4.5 Mitnfti 7.9 -U2.2 5.8 -6.6 2.1 6.J -0.9 00 4.5 -2.5 -4.2 -7.4 lMuIa*lws -4A 8.4 8.8 -4.9 -21.2 -4.8 -4.7 2.7 4.9 -4.0 4.2 6.4 Elc. &A a 21.4 2V.2 u.2 C5.6 4.8 0.7 -1.7 6.8 56 4.0 0.8 2.9 conb"lin 8.5 -14.4 12.8 O6 -4.6 4.5 41* 20.2 4.9 17.8 2.8 5J Tid4helui_ -z4 *.5 1.6 O.J -1.0 -L1 4.1 1.1 0.6 U1.1 6.2 4J T-A. 4Q 2. 4.8 0U.S 49.1 2.6 -18* O.6 1.8 4.8 8.1 0.6 Finl kc. 2.0 6.7 6.9 5.z 1.9 6.6 4.8 6.9 2.1 0.9 2.8 8.1 I'lb aWn,d 6A 21* O. -2.1 11.4 0.1 -6.4 6.2 1.9 -20.3 6A L.1 of diidg : oth wr. 14.1 28. 4.2 12.9 -14A. 4.4 1.2 14.4 232 -7.7 0.8 -3.7 rqx* be* dV. CDP at .c. .4 2. 2 2.6 .5 6.0 4.4 84 2. 8. 8. 4.1 MIsion E*lms (ua c uI A1 4u WJ)V jlbmw I,. 4.2 4.5 8.2 2.8 l.1 0.1 1. 2.5 .2 2.6 87 0.8 Inl trg 7.9 -28. 6.0 -4.5 2.1 .5 40.6 6.9 4.6 -21.5 4.2 -7.4 obw 4.6 8.4 8.8 -4.9 -11.2 4.8 40.7 2.7 -8.9 -4. 4.2 6.4 Elw. & mft 1.4 2V.2 21.2 25.6 4.8 0.7 -1.7 6.8 56 4. 8.8 2.9 CnictJon 5.5 -28.4 21.0 6. -4.7 ao 41.9 14.9 4*A 21.9 4.9 2L. TebA"OIMl* 4A6 0.6 1.6 6J -4* 4.1 4.1 1.1 0.6 21.1 5.2 4. Tripovcem. 4.6 2.6 4.8 21.8 -.1 2.6 -28* 60 1.8 4.8 8.l 0. Fltevem, ac. 2.0 L. 5.9 6.2 1.9 0.8 4.8 6.9 2.1 8.9 2.8 L8 X 8n SW,c 7* 19.7 7.9 -1.2 23.1 0.1 o4. 6.7 S. -9 0.65 5.8 at hi;g: oth ur. '4.2 28.4 4.0 18.9 -14.1 -4.2 1. 14.4 232 -7.7 0.2 6. -po bm& dg. MP IPA 0. LO2.7 2.02 6.1 OA . .2 2.7 2.9 1.6 8. 8.1 ilsioen Eilm 7mlsdiu HNIn 6Sn. AbO4u-11 S)E A-WlaUl, dc. 6.8 0.6 8.7 2.6 2.8 40 1.1 2.0 0A 2.4 4.6 4.4 U1inlr 4. -2L8S 5.8 4.7 2.1 0.5 4.1 6.2 4.6 -0.7 -24. -7.8 b,srtwur4 46X 8.4 8.8 -4.9 -21.2 4.8 40.7 2.7 -4.9 -4.1 4.2 6.4 Elsa. A mr lA.4 27.2 1.2 26.6 4.8 0.7 -1.7 0.8 GA 4. 8.8 2.9 Ca-ucnlen 5.5 -18.4 U1.9 0.8 -4.7 ,0 45.9 14.9 -6. 21.9 4.9 22. T:i1b:/=m 4S. 0.8 1.6 0.1 -4.5 -4. 4.4 1.2 60 22.6 8.2 4.2 T 4J 2.6 4.8 21.8 4.1 2.5 -28* 0.0 1.6 4.8 8.l 0.5 Flaw. st 2.8 5.4 5.0 6.6 2* 7.$ 4.8 0.9 2.0 1.2 8.8 20.4 Pub Mn,dc 7J 29.7 7.9 -1.2 20.1 4.1 48 6.7 8.0 -16. 0.6 5.8 01 shldh: th _r. 14.2 18.4 -5.9 18.9 -14.1 -4.2 1.8 14.4 3.2 -7.7 0.2 0.0 JqA *n&dr. 9A 5*.0 8.8 80. 8.4 21.5 7.8 5.4 5.0 21.2 2.7 6.7 CDP at 6. 0.1 2.4 4.1 2.1 4.2 4.4 -4A 2.9 2. 2.1 6.0 5.7 Official Aawaft 2d 3 2 2 122 17 214 9T4 S10 490 68 7M IM 1414 Snlb . 230 2 321 14W2 V4 2 U28 148 218 144 25 428 618 W wlr 3 9124 141 27 23t 2llD in9 232Z2 215 6W4 Ela.aV 23 236 91 65 2 la - 2s4 22 m W SW 71 ca la n 12 212 2 W4 145 21 22 1 220 2i2 848 444 481 ma TuweIwbsli#Az_. 23 M2 189 W1 2 23 25 812 US US W68 775 21f TruE_'tSom. 23 W3 2. 1 28 190 W 2 2 8 25 4S w18 m1 23 Fl do0 W lie WZ W7 la Vs8 in 2a 219 26 m 87 F=n,ftc 2 3 9a M 16 32 2 12 3 M 24 2 814 SSB 481 at sild: olh 3 23 2 34 254 128 2J8 B 235 241 M 25 251 818 bpfl W d. 2 1 128 24 288 22 25 25 I3 2B W 27 748 i C 23at2f.28 13 141 2t0 28 221 274 85 445 18 74 INS Wlonon E _m (nahdlitr Mw logc J4Aubt ;DA)/ Z*A91 g2 210 25 25 17? 237 2U0 8 Ul8 BES 7W7 21S 14i8 tI4l,% 3l5 IN = 142 14 25 28 1J 4 2I1 144 2 438 62 _uIa-vlv 23 226 11 287 0 2I 29 189 S 275 - 429 49l 64 Elsa.&Ss er 2n 4 I1t s8 23 201 2I 22O 21 242 we as 75 Ca,alon 13221 322 289 25 175 2B 232 m1 86 446 W ON Tib lmtsIi/fa~. 23 32 UP l 23 23 - 812 US 6U8 6 73 U2? - 75 - Pqp 4 of6 1in iw ins in nlow Im W n 1111 984 un ur lw urn Traupwtanm. 18 0 16 218 32 2 12 2S 2S3 3 4z 6MO M1 15 Flnmm ,tc. 28 23 222 2 261 18 181 12 28 219 246 8 370 Pub lmin,*W 288 106 9 15 15i 2U5 15 23 268 22 8MO 86 4 of wich:dotdh_ 2 Is 2s 2l4 a 2 U18 S 20 8 26U 2S 256 2EI 28 . bwk dg. 18 UB 14 1 182 24 36 2 US U2 748 14 MP at f.c. 21 11 2 IN 2 25 21 275 8 441 8 7? 88 MIsic.n EatI (inlcki,r Hlclin r 4uut DIJ)A / rlaultu, et. 13 11 28 25 178 gS 37 8 SO 6 725 M 148 tllnrgl1 20 22 8 10 IN 11 2 23 274 US 698 I3 lacbrirg 20 22 141 JR 25 2 1 8 2 25 In 408 45 34 Elsc. A r 28 30 . S 108s 281 30 22I 23 232 US 78 Cajmcbin 28 121 23 2I9S 2S 2N 2S 22 271 10 448 U eU Tr hoslu/ emt.. lo 28 25 2O 28 z n 844 48 5 721 9 I Tmumtc i. 8 l3 28 S 123 288 2 3B S S 486 US IIU UN88 Flnw., dc. 18 116 U22 122 2n1 2n 181 2 2U 9 26 28 a F whinatis 2 206 98 186 25 IN 12 2 26 U0 MO M 467 flvihi: othmv 18 2 206 114 13 1278 U 228 2S U 296 21 2 2-rpit dbisg. 0 2U8 142 1M8 32 204 3 2 0 S 2 U 748 14 CP at fc. 2M 238 132 141 16 18 22 210 M 444 1 783 9W7 A4cou t !ucldirg M11_m PNMlei E;mbl (Wil l Ian of iAht Tab (r a4suet) Dg lturm, otc. 6.6 782.4 94.2 21.6143.1 .8 14.2 2L38 218.8 2981.2 6A A.8 2M6 oi _ b 40 614 8 86 2015 28 2781 18 288 3 88l SW ow a'p (wd 44.4 61.7 75 175.5 246.9 28.1 25.6 23.6 .4 Z 8.9 178.2 89.5 474.6 ratna. (fcadc 27 88 44 66 2 is 99 its 2in 188 24 4 G7m tftW hzl*M 64.4 68.7 5.2 68.1 48.2 4.7 66.7 78.5 218.9 116.8 2USA.4 8.8 887.1 Milnir 32.4 27.9 98.1 48.8 88.6 64.5 Ł8.4 77.6 96 84.2 198.1 422.8 617 lpid a our 32.4 27.9 80.1 48.8 88.6 64.5 68.4 77.6 9# 84.2 198.1 422.8 61 kMvAt8*w4rg lo U gi 23 23 64 2S 209 2V8 8M U 45? IN 9 hladoat 2 V 88 18 U 224 2US 248 278 M 83 487 59 86 Else. a wŁ4 Catuctviton Trad4intal,/rL 89.1 43AI 8 3WX.2 .79 7A.4 757.27 R8. If 4 1834.22. 206 .72 43* s.81 taurim 10 1 2 256 24 2J2 188 185 5 38 eV 1441 2421 di1tr1b. mg. 216.1 842.8840.37479.79 60.4 6MM 6 S 3.8 2122.8 U61 I5.7232.67 8.81 FInmi, ctc. 46.8 44.8 41.6 44.4 44.8 9.6 83.6 12.8 3.7 251.8 9 .7 227 2234.1 hasmi, 46.8 44.8 41.6 44.4 44.8 6.6 83.6 123. 13.7 51. 6D.7 2.8 2.284.1 I'lb aIniun,tc of whid,s: 6h ar. 3pvA bw* du. MP at f.c. 2296 1486.7 IJ7X MA .4 238.1V M.21887.94 448.8 Ml 019.22 11481.6 518.5 (aply ^dlstr) M.62 12.1 1.9 11.8 188 11.9 .1 28.8 84 8WM A2.4 O.9 1 37 Mlllow of 17 Tab t#4tA4dak) Ar1lctw1, &a. .6 664 2 68.6 741.8 696.7 6.8 69.8 .6 at.9 6A.6 X4.6 2181.1 oll Iu 43 m 118 6 8 1l 567 69 - 63 1 1 6 odw cup (card 44.4 41.8 88 48.8 65.6 47.5 38.6 82.9 48.5 1S 21.6 8J.6 80.8 wnktd. (fooadc : 27 S 88 88 a as as as 48 sO 41 s9 tFht htlrg O4.4 68.7 87 48.2 48.7 46.2 81.8 22.9 8.1 83.9 87 42.1 415.8 M1nlrh 32.4 25.8 2.8 21.6 19.9 23.2 30.8 19.0 19.9 28.5 17.9 18.4 17.1 pid A other 3.4 5.8 3.8 21.6 19.9 23.2 23.8 1'.9 19.9 18.5 17.9 28.4 17.1 lb AlIF IN 2Q1 25 2 138 2IN 21 120 1 li3 u38 us8 2S ludifa 2821M 2 52S1 26 288 18 1 23 32 238 238 238 12S El.c. A Aar T bm*,lq/iat 20O S 32 1M1 25S 80 6e S 68 49 32 W S tarwim 2s so 2 l115 5 6s8 68 a 49 32 1 2S distribuin 256.2 419.6 9.8 718.8 978.4 12M.6 511.2 049.6 4S.6 6215.6 8.9 24.1 A.6 Frin c. 45.8 89.8 86.8 8.6 5 .7 84.6 48.8 88 .6 118.1 199.2 492.5 725 hlark 45.8 89.8 a5.8 88.5 9.7 86.6 48.8 97.6 318.1 29.2 492.6 725 Ptb anun,dc of ahldi: ot r. -pA . bv CP . f.c. - 76 - /1 Flg,. for IS ad M am, sAJ$ toe sIsn. = 1 l 1l mtuJs forswhgisltuo, ambrulcn, otiu mvlm So SO=t1 A w-I la 1d1n Omi 4u*a for: ol m a, n, ar tod crp, trq*p hlul, VW A adores hsla,wfte, tourIm% dist1.bt1An, ad rntal hosalr. Dos not Inlaudo uN Wm tImotbas In tv_potstlan of pwbpl In *Alu wd ethr prlva trade md busm, infesl bmdcr1 1vlm. A4ud bmd an "A Sb* re nonanoditlaml out by M.S.D. 1chiS. N.E. Luur, wed 0.D. lQm, Spr,M So amt1Uml a dad - 77 - Table 8: CDP B1 0OrE OCUL M MISMIJ JZAIrS Pop 1 of 2 193 1977 178 19 19 18 2 1 988 14 165 1M 27 I O~~XaM. .S . B 0' By EG' DMM Q8±DB O Gar M.) Privat aAptlon 1377 17979 23388 S47 88 83 46 113 6M8 M1 2 1BM 21# Ounl gv'lt. acin. 89 4M Q S 494 6M G4 9448 18844 IBM8 2981 2 87887 EWbt gwrif 8W SW 4698 5181 1548 94 45481261 em 74;S 14Si 298 4WD Xport gfs 6841 O6M OM5 MM15 11? 0168 8898 8781 2548 174# NM 71M 21911 G d. Im. 7W 7104 54 94 6 2"3 26 M 188 1 81a 417 6am Graw fthud lm. 51S3 6 783W 2 83 8832 1M ; 778 27 237 459 OM brss In stod 438 881 764 88 I 1498 1410 185 14 2 2487 85 41 CDP a op. 24419 29 3 189 8 4228 49 GM 7M 89M 11 24 ZB 511478 li factor Inim -28 -8 -4 -71 -132 -I3 -2 -21 -178 -7 .-M -1M -15 OP at op 24819 279 M14 8922 429 4895875 GM m 8 729 197 23 g7 m 729 29 WY (aF -18.0 -172.0 -44.0 -71.6 -22.0 -178. -1. -184. -499.0 -D11.O -. -25.9 -2.0 Ow Mf"( ESI) 248 29M9 35 8i212 4M 4M 579M 7 83 228920 M "8 2MM741 2- 9S OF CPaIRaAL Private apissblon A8.0 62.8 72.6 79.8 77.1 75.4 72.6 7d.2 77.2 77.2 79.7 88.7 3.6 Gn'll ov't. =m. 28.8 14.9 17.4 16.4 18.0 12.4 18.8 28.4 16.6 2S.4 14.8 18.4 12.2 Eorts Wife 21.7 19.6 14.6 14.1 13.2 12.2 7.8 7.2 7.1 6.2 9.1 18.6 15.8 npsrte Vf 23.9 22.8 29.7 3.9 3.8 20.7 15.2 12.4 15.2 24.6 23.2 88.8 41.7 aross tmu Inv. 22.9 3.1 2S.2 3.1 2.0 29.6 21.0 18.6 15.8 16.7 19.6 22.8 21.2 Gras tsd Inv. 22.1 2.1 22.8 23.7 0.5 17.8 18.6 21.0 18.6 14.0 18. 21.1 19.8 nem In std 1.8 8.0 2.4 2.4 2.6 8.1 2.4 2.6 1.9 1.7 1.8 1.7 1.8 CUPat op. 1.0 W 2.0 2W8.0 1.6 1. 188.0 180 28.0 188.0WX 20 280 .0 Nd; tactor In=c -0.4 -4.8 -0.1 4.2 -0.8 -0.4 4.4 4.8 0.2 0.6 -2.1 -6.0 4.a OW at up 99.6 99.7 99.9 99.8 99.7 99.6 99.6 99.7 99.8 99.4 97.9 #i.0 94.0 MMN EUM115 rJV P Privxte macnon 1a811.0 18.0 27.A G a in8.0 .892.0 5.0 8a 999.01 272.0 288.0 2Me dn'l wvet. =m. 898. 4.0 SM.0 580 5440 6.0 0M8.0 9448.02 . 1 .0 2J)81.03J 2J3 8787.0 Eapr01.0nts 47. WU.0 480A 181A15M.094. 456.0 51J1. 62.0 748.0 1A.0 2.0 4.0 Iruoeb Wife FB41.0 6.6 J5.0 9.0127.021 0 8982.0 817.08.0 17489.0 835.0 7959.02911.A Ors dm. lm. l.O0 42.0 7m84.0 118.0 11671.0288.02 .0 28.0 141%5.0 1993.089 A.0 .a=0 65. Cras tiud Inv. M2.0 7X.0 89.02 5.0236.02 1.022.9 94.0 2.0 17278A. 8104.05928.J1898A.0 Ifame In stods 4388 81 34.0 8.01 8. 1498.0 1410.0 .0 264. A287.0 88. 432.0 CDP st op. 2.0 Z791.06 8W4.0 96 A 4138.0 48416.0 67.0 a3 88870 11781.02t8. .0 2.0 84.0 Ni factorw I-n -238.0 -172.0 -44.0 -nX -1122 -176. -21 -18.0 -4899 -1721. -MA.-283.0 -in.. aP at up 239.9 27748.0A81J0.0 11.0 41724.04 AMA.0 6 .0 8A 89 3 MW.0 263010 8.0 I4.0 274842.0 9WRE OF CDP OMaR 1FC-! MIDlEZ Privae qmaplan 8a.7 Ea.8 65.7 68.4 72.2 68.9 69.7 75.6 7.0 76.8 77.1 W9.9 95.7 mn'lI 2v't cns. 1.8. 15.4 17.8 16.7 18.1 12.6 14.1 18.6 26.0 15.7 25.8 14.0 12.8 ot ts gtfe 28.0 3.2 14.9 14.4 18.2 12.4 7.9 7.8 7.8 6.8 9.6 14.1 16.6 Zqmt gts 24.8 28.6 . V7.4 3.6 21.0 15.6 12.6 25.6 14.6 24.0 88.0 48.9 Grme dmL In. 24.9 29.2 81.0 32.9 27.9 2V.1 28.6 18.1 16.4 18.4 22.1 29.0 19.9 Gra f1hd Inw. 28.1 3.1 28.6 30.4 A.4 24.0 21.8 18.6 14.6 14.7 2.6 27.2 17.6 1cams In stucl 1.8 8.1 2.4 2.4 2.6 8.1 2.6 2.6 1.9 1.8 1.6 1.8 1.4 DP at op. 28.0 18.0 288.0 2.0 2O8 8 08.0 188.0 18.0 108.0 8.0 208.0 8.0 188.0 Nb factor Inamrn -0.6 -.6 0.1 0.2 0.8 -0.4 -0.4 40.3 4.6 -1.6 -2.1 48.4 -7.0 OP at lp A.4 99.4 9.9 99.8 99.7 99.6 99.6 99.7 99.4 98.6 97.9 93.6 98.0 CDP Of PDOrM Q 4 OF Oa * 2I97 IS" PrIvtn aramptlcn 1277.02 .7 1725.0 22.4 18.7 2 .1 18.6 1792.6 14.2 17991.8 1772.2 194.2 192.7 Gm'l gw't. ama. 8.0 44.5 518.0 6S77.8 414.6 483.1 6477.2 425.2 188.1 97.8 797.8 71.6 78.9 SiquortuMf W7.0 46.0 44M.6 87.5 498 497.8 4A.1 8245.5 2817.2 229.2 23.6 8S1.2 89.8 orts Ffs 6B1. 48.1 8179.9 69.9 68.5 68.1 4M.5 4610.8 38.0 GM0.5 591.7 M9.9 MA2.0 Gros daIn. . 59. 54.5 6414.8 68.8 62 890 1.2 649.4 4441.7 513. 887. 6748.2 58.2 81.9 Gras flaw 1w. M530 6.8 M 84.9 6221.5 87.2 6517.2 1542.8 85.4 4740.1 338.2 535.0 56W9.2 6324.1 a. In stud 488. 735.8 38.9 481.8 61.8 644.0 7.6 63.4 49.6 418.8 8.2 421.0 407.9 CDP a op. 24419.0 244.6 2498.8 2 .1 933.8 2M.5 2S72.7 257.7 5920.0 274.8 233W.1 2948.4 96.6 Nfator fa In -I1.9 -74.4 -85.0 -8.2 -70.0 -48.4 -168.0 -77.1 -A.4 -168.1 -82.8 -1485.6 -18.2 O at p 2482. 2686.2 2418.4 21Z2.9 255.2 2M9 .1 218.6 2l526 57.6 299.2 27774.8 2799.9 2923.6 - 78 - Pop 2 of 2 I915 in INS I19 190u 1= I im 1m ISiS ISIS w im W Y -15 -145.3 -84.178 -.193 -21.6 W8.4AS -15.4 -.248 -14658 -6.831 4M2.6K -1789.89 -372. Of PN( E) 2428 24215.1 24956.1 2R.8 32.2 205.1 23M 25M6 .4 12.9 2M71a 2M4.48 2781.8 292.75 Private amplaen 14811.01440.815 6.01458.2 176.5 14M.8 14488.4 I.8A 1.7 1747.6 1M.8 17 1 .2 1973.1 unli g,Vt,. ae. 8.0 4.646 5188.0 6 77.8 41.6 43.1 6W7.2 410.2 513.1 6W.8 77.8 7811.6 763.9 Eiporta bfg 6291.0 431.6 445.5 8Y78.6 458.4 4970.8 843.1 84.6 2317.2 S29.2 208.5 3S5.2 819.8 pot MgSUR1.0 M48.1 8179.9 NS.9 8481.6 38.1 4711.6 46101. 418_ 6 .6 9.7 19.9 OM2.6 Grow dn. lmw*6. 3.0 1.0 m74.5 84.6 7867.9 7714.8 720.7 829.1 610.6 560.2 a0.7 6779.4 651.2 rss fiud mw. 2.0 6847.0 72211.6 m9.6 n9.9 716.8 85380.7 459.1 4008.6 642.2 8.7 63A.4 59.2 Jnrm In stods 438.0 734.0 18.A 426. 438.0 62.0 511.0 61.0 442J. 414.6 80.0 423.0 410.0 CP at Bp. 2433.624W7.2 2 .05295.2 2M8.9127V.85928.6545.8332.6 899.8 27W7.7 23.5 2983.5 Ni factor Imo -118.0 -148.7 44.8 -1.8 -70.2 -65.0 -101.5 47.0 -149.6 -. 49.6 -121.2 -21210.2 OP at up 2S.0 29 f8.5 2648.8 248.9 2X3148.7 2812 2 .56 2.99 2.1) 3N2.2 2138.0 2S3.8 27188.3 lll OF MP BY EOrUE *SD N 04GRaM.. GO ar Privas omsmpbicn -0.9 12.9 -6.8 13.9 -12.6 -6.6 17.8 2.4 -1.9 -1.6 9.7 1.8 h'1 Pvw't. am. 0.4 28.3 10.6 -.4 12.5 18.1 -2.7 23.8 8.6 42.0 -8.0 4.4 Eqot gM% -13.6 5 8.8 -12.1 6.6 21.6 -80.4 -6.1 -18.2 11.1 -0.7 11.6 0.6 D*Dot g's 1.8 87.6 -18.1 -8.2 -12.6 -17.6 -14.9 14.7 14.4 12.0 2.2 8.2 Gross ds. Im. 18.9 -2.0 2.8 6.7 -L9 8.7 -81.6 18.9 8.6 1.9 -2.7 20.4 Gros fiwAd lm. 12.6 1.8 4.0 -8.2 -6.8 11.8 -98.0 24.6 10.1 2.6 4S.4 22.8 Ina in aod. 67.9 -27.9 -12.9 18.0 23.4 -18.6 18.4 -29.2 -7.0 -5.4 8.8 -8.1 MP at op. 6.6 2.8 2.7 2.7 -1.1 4.8 40.8 4.4 0.7 4.7 4.0 5.1 Nafactor Inam Nt ia M Wi W Ni Ni Wi M W OP at up 0.1 2.6 2.6 2.6 -1.1 40.4 4.7 4.6 0.8 8.1 0.8 4.1 1111%01 113 Prlvat anaiption -1.4 9.1 46.6 18.8 -16.8 0.1 18.8 6.2 -1.2 -6.6 6.9 12.6 a'l 1cv't. am. 0.4 28.8 10.6 -2.4 12.5 18.1 -3.7 23.8 9.6 42.0 -8.0 4.4 E~ipoe giFs -13.6 -4.8 -12.1 6.6 21.6 -9.4 -8.1 -18.2 11.1 -4.7 11.8 0.6 rport gs 1.9 87.6 -18.1 -8.2 -12.5 -17.0 -14.9 14.7 14.4 12.0 2.2 8.2 res d.L im. 18.0 8.9 6.8 9.2 4.8 4.6 -.6 8.7 6.4 8.9 8.1 -15.9 Om. f1md Inv. 14.1 18.6 7.2 -10.5 8.6 4.8 -89.9 7.8 7.6 10.1 8.1 -18.8 Incun.. in stade 67.8 -8.1 -17.2 14.8 15.2 -0.1 15.6 -5.1 4.8 -6.8 8.6 -3.1 DP at op. 0.8 2.8 8.6 2.0 40.8 -0.9 -1.7 4.0 0.8 8.7 8.8 4.8 Natfactor ;nww K% M i H% a Na a Ma 1a Ma Wi a OW at up 0.8 8.8 8.6 1.9 0.4 0.9 -1.6 8.7 4.1 2.9 0.7 8.6 1 IBlDf lS{l Private conmarba 2.0 118.9 188.8 158.6 178.9 236.6 276. 814.4 871.6 514.8 718.8 957.9 1456.4 Ow1'l gov't. acn. 2.9 10.6 108.7 18.9 13X.2 2A8.6 148.9 25.9 58.6 881.6 57.2 41.1 498.8 Ezorb #gl 1.0 122.8 15.6 122.6 15.6 1320.6 181.6 157.6 226.4 238.2 518.8 9.08 12.7 b7o.t off 18.0 110.6 118.9 146.7 171.1 171.1 188 218.6 236.6 832.8 63B.4 1.2 51.8 Grow de.L li. 19.6 115.0 13.2 148.7 157.8 1W9.9 198.8 216.9 .4 835.4 542.4 868.9 979.8 Cres tiud inv. 1.6 114.7 126.6 140.4 15.6 132.8 182.2 203.7 212.8 28.8 563.1 896.4 917.2 iens.. in sto MP d; op. 9.0 112.7 13.6 138.6 13.6 196.8 21.1 274.8 .9 441.4 60.8 728.4 98.7 Ni; factor nin 189.0 126.7 3.6 188.6 13.6 16.8 21.1 274.8 13.9 441.4 S8.8 738.4 998.7 oP atop Official cbflator DP at f.c. 189.5 118.21 128.74 141.44 13.98 15.48 23.18 278.61 83.83 445.19 W2.9D 742.99 29.8 Cwwb piese MP b emdb. sta the officlal wud s6W riles. JuFAdc be for:sga d9tm c I IInvsbn &d privats nmuptlen (rmids . Nb tfr mn coe atbd to be canisbu* wiUth 9W ( ,JWIIP 1976. 8 7 I Caatut prlie smim dsflat by: gtirsi goAmut (pblie ainistaion wots ad datlor), aite OP (M WI 19646), larpwb Off (M= IPI for 97688, Wrld TWal 1909O), (M mid I an tloet frem misslon eti ats awk to. OP ad WY ftin OP f.c. dSlator fraU aduetndntrial origin aftitss. Se thsr table for 1lI. - 79 - Table 4: AGIUM P#p 1 of 4 1918 197 1I7 197 low8 196 I" I=8 198 135 13 137 118 io of GMC WM aP pM uSod Cotn 289 872 7 8 MO 54 468 S4 634 19 18o OM 68 .1 Simi fibr 25 278 MS 2 2 27 7 23 23 26 22 45 048 784 M.IA Tft In?d9 68 13 15 13 118 U 12 287 234 815 8 4e m rW2n GOOD 713 048 7 8 448 711 624 747 139 17 26S 21 t 25 .6 wr Ceu U 121 13 181 141 111 UK 235 3 45 8 45 781 1.6/A Tobo 84 10 127 12 141 16 195 219 uS me 480 7l 7 m O.8 Cinwb 96 21 84 84 1141941 245 US 2m 24 D 0 8.9M PyrJInMA 29 2 1 10 9 18 1 14 18 27 20 as a 0.9i Maim U1 181 14 142 17M 1M9 AM s 4s MO 1117 1M 1910 .7 wub ag 77 69 8 117 149 2 145 2 87 45 as 38 1.8A Flo*8 814 484 814 437 88 788 26WM 1424 O8 GM OMW 10.69 OIM cIrle - a6 38 19= 18 Oa 12m 187 23 am15 4884 540 s21 . /I Laos= 6 M 8 12 18M3 1888 1 871 2817 445 04.815 74U1.90 26.24 8.M 0il SON& / 65 781 9 I 177 1 8 9 12 8s Om 4w 0448 .3A PI*at . 468 699 78 98 112 127 19N 2L3 214 92 SIR O8 MM / Othw hor. ap 740 04 1'# M14 M17 24 89 451 4i4 544 1 1479 A87 A.M0 A Slkm 124 23 23 1m 9 211 a81 547 61 70 790 182 W .70 OUIme we 15 W12 -M 2 2748 87M 494 69 781 M17 146 999 8 9.9 t2 of : 478 674 7 3 7 756 9 1 10 145 17 O7 8 0 6 782 8. /2 WI&L aS 7545 6m94 i 111 law49 gm gm gm A487 4618 G4 I 194 aS W[IFUf CP) own OF 1UM Sa m atton 8 4 4 8 4 8 2 2 8 8 2 4 4 Simi fibr 4 8 2 2 2 2 1 1 1 1 1 1 1 Tem id 1 1 1 1 1 1 1 1 1 1 1 1 a Cdffe 10 7 6 a 8 4 8 8 a 8 4 8 2 S5r =& 1 1 1 1 1 1 1 1 1 1 1 1 1 Tcbom 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 0 PyriOm 0 0 0 0 0 0 a 8 8 I I 8 I I ibim 15 15 12 12 12 U 12 U 14 18 19 is 16 u_t 1 1 1 1 1 1 1 1 1 1 1 1 0 P **b 6 4 4 8 8 2 4 4 6 6 7 6 7 OUu.mr nla 7 7 9 11 7 * 6 0 7 7 7 a 4 Laau 9 6 U1 10 12 10 9 7 7 is 9 8 8 OlI Ssb 8 9 9 9 S0 9 7 9 9 7 6 6 6 Planutiu 8 7 8 8 8 8 9 8 7 6 6 09 1 Ottw hwr.capo 10 10 10 12 12 I6 17 lB 14 12 10 1i I6 C4com* 1 1 1 1 1 1 2 2 2 2 1 1 1 Othw crops 17 19 19 21 23 24 24 26 24 23 21 22 23 cf v:mu.s 0 8 7 8 0 a 7 7 7 8 10 7 6 THUTWS 123 18 2323 232 23 lo 23 23 23 am rA W ) SJ a 298 872 O U 88 8 41 2W 213 2717 E 23 42 511 SIi Tfbh 2u6 28 238 204 1 8 I4 1S2 U O 81 7t 8t Oa TM f 9w 68 n 67 67 68 ˘0 07 69 e2 67 a8 51 G0ffwM M 76671486O e eM 72 678 684 749 7n 698 &Wco2 92 82 09 101 78 W 67 11 01 s8 77 88 Tcbsm 84 107 I 10 97 65 94 74 64 77 s 1 78 Caslows 96 06 98 51 51 8 40 42 42 27 18 81 18 Pyuotlvm 28 30 17 20 1 18 1 1e 9 10 9 8 9 Lill 113 18 1172 17 2 1471 l n 1 m 1m 173 179 1an Mt1 SO64 6; 70 87 9 S6 618 74 e8 n 72 67 FW* 841 814 w8 2m2 2' SUB OM 425 49B 6MO M6 O _ rmla 571 e sn 740 85 - o 62 748 30 884 on 893 9 I_ da6 634 62 6a e m ane 6on 615 138 410. 4 M2.0M 994.S6 OTt Src is 6 8 eM 73 73738045 7065 79 704 73 768 PITanle 483 48 510 mo ss s5 571 889 64 787 727 747 7m8 Othehw. alps 740 7M 78 769 797 812 eV87 MS1 7 08 1 U1 CoaM St01 61 e6 82 68 C6 67 68 2 78 as 84 as Other aN 15n 188 19 143 18 91 119 1W 172 I 191 8I74 1917 of :- 478 69 I 54 48 62 eM 7W7 718 769 on 684 64 1W ON S p.O.) 4 74 8 38 694 687 88 O44 91 g172 1319 2310 - 80 - Pip 2 of 4 19 27 298 299 2188 91 21 18 14 1 298 1 16 rio8 m of GF01 RA1B8 SJd anl IA 6Ei.4 -18.6 -1.1 7.7 40.6 -2.8 -2.8 9.7 21. -.34.8 219.1 30.2 Siml fibr -7.9 -0.6 -14.8 6.7 -14.2 -17.8 -2.8 -17.2 -16.7 4.6 20.0 0.8 Tom let 28.6 21.2 -4.8 -1.8 -6.7 -4.9 18.1 -13.4 6.8 -7.6 -0.6 -1.6 Coffee -28.8 8.9 -8.6 -4.4 47.6 -5.0 8.0 -7,8 -0.9 12.8 2.8 -14.8 Supr wo 1.7 -12.7 21.6 1.8 -4.7 27.8 -10.1 27.4 -18.8 -21.6 -4.0 14.7 Tdo 2B.7 -0.9 -4.1 -4.6 -1.8 -1.4 -0.0 -28.1 19.1 4.6 27.2 -0.9 CmIuant -9.7 -24.65 -.1 -0.1 8.0 23.0 6.8 -0.4 -M.6 -2.4 88.1 -4.8 ly,dnm -21.1 -17.8 -37.8 1.9 22.6 -6.4 -25.4 -9.6 6.4 -12.1 9.0 14.7 Wibi 14.8 -12.0 17.4 0.8 6.6 -20.1 40.2 17.4 6.6 2.9 3.4 6.8 Wmt -7.2 -14.1 27.8 24.8 8.4 6.6 -8.9 27.6 12.2 -14.6 1.4 4.9 Po* -9.2 28.2 -92.8 11.1 -8.8 08.0 9.4 1.7 19.4 18.7 12.9 9.8 Othr _ouml 12.4 29.7 84.0 -Q2.9 -28.7 80.7 -28.8 35.9 -0.4 7.8 8.4 -21.7 1_2AWW -14.1 4.7 6.9 24.1 4.5 6.6 1.2 -2.0 82.6 -2B.6 25.6 -4.6 all Snc5 6.9 8.1 8.4 7.7 -4.6 -12.0 8.7 -4.9 8.0 48.2 7.6 8.9 Plorbtlro 2.8 2.8 2.8 2.9 2.8 2.8 10.8 -4.0 17.0 2.8 2.8 2.8 Os hNW. a 1.9 1.9 2.8 0.9 1.9 1.9 7.2 -8.0 28.2 8.1 2.8 2.8 COMiMA 2.7 0.2 0.8 2.8 2.8 2.8 1.0 -7.7 21.7 4.9 6.0 2.8 Othr crap 7.2 -0.9 6.4 -4.2 9.2 7.8 14.0 -4.9 9.9 4.6 -6.9 2.8 of udi:v. 14.0 -7.1 12.7 -14.6 23.6 28.9 19.6 -3.7 1.6 6.7 -16.7 1.6 lTrAL aI S (l.O.) 8.8 1.2 5.8 0.1 2.4 40.8 8.0 8.0 7.8 0.2 6.6 0.1 DEiA1US (GM anU Sad cAkan 21 82.0 6.0 2.8 148.9 219.8 184.7 236.2 801.2 43.4 A. 845.0 972.1 Siml fibr 21 26.8 S8.9 21.7 25.8 149.9 251.8 291.9 2".6 257.8 648.9 778.8 942.1 TM let 2d 112.6 17.8 287.6 217.4 287.6 28.6 851.6 849.7 611.8 62.6 949.9 27.4 Coffee 21 8.2 n.1 63.5 71.8 77.1 91.6 2.8 2.6 2.1 20.8 81A.6 42.2 Qa me218 120.8 3S2.6 22.7 29.6 2.6 1.4 224.5 .2 448.0 46.9 63.6 7.6 Tdmo 2 1I8.0 20.8 218.7 145.6 27.6 21.8 25.7 7.4 43.2 85.6 78.7 94.0 Cadinub. 18 26.4 18.9 284.6 23.6 871.1 489.4 GW5.1 794.8 2010.1 237.6 2.0 8219.8 py1tDu 21 820.0 26.0 21.0 84.0 O.0 238.2 238.5 2.6 2VO.0 824.6 418.9 644.4 Mis 218 11D.021 1.2 20.8 25.0 125.0 1.6 228.7 275.0 GWA.0 A.8 787.6 12.0 Wt 1120.0 25.1 25.0 134.9 026.0 23 .02093.6 449.4 3.0 720.0 90.0 PF 12 810.0 20.0 20.0 10.0 176.0 230.0 80.0 4M.0 N.0 93.0 93.0 144.0 OUm.. aus 21 21.6 280.8 183.1 .84.6 142.6 149.6 224.8 23.6 895.4 E.2 8M.6 7B3.6 LMgAm 26 218.8 175.7 12.1 197.8 260.1 24.0 224.9 284.1 426.8 189.4 711.8 2I9.8 Oil Smda 123.4 149.8 21.1 179.2 I6.2 232.8 842.0 80.6 425.6 MS6 647.9 818.1 Pltulrn 18 J23.8 254.6 11.8 203.0 228.0 347.1 847.6 8A1.8 411.9 4S8.1 22.56 237.8 Othr hor. crps 1123.0 287.6 277.4 21.6 A.6 422.6 478.1 6.6 78.0 04.4 1497.8 17M1.8 COuMA 218 2a3.3 178.9 243.9 826.2 82.2 56A.4 810.7 632.8 903.1 17.6 280.6 244.7 0th cr.np I 328.7 149.8 177.1 23.8 254.7 821.1 818.? 443.4 S2.1 798.8 20.9 1488.8 of "wo:mma 26 125.0 20.0 02.6 M2.5 176.0 26.0 221.0 O3.0 G3;0 730.0 9Z.0 112.0 VEaE AMED GP Sd acton 125 194 198 23 2S8 24 23810 45 6769 6n 2875 Simi fibrm 24 232 23 178 223 236 15 188 23 297 418 54 688 Tu Ius? 44 67 18 llO 10 12 973 27 178 278 am 42 549 Coffee 664 MS 43 20 SW MO9 9G U8 840 2016 1740 342 28 Sipr am 78 9S2 122 323 95 23 174 810 844 882 428 622 Tobado 26 82 38 88 42 48 65 68 M# 93 141 22 2M1 Cuu 88 66 79 782 621 17624818 2M 8 23m 472 M99 FP uMan 26 19 la 20 9 12 28 14 17 5 27 as 48 maIm 922 I068 980 2251 2368 1447 118 2374 8357 ?7 8788 U= 1i6 "nest M 65 8 74 I6 3 178 22828 817 8n 441 528 P** 81 279 418 230 M 812 1M237 230 8832 4488 7897 Otbr wds 432 15 M 1122 Or 799 12 143 156 84 4519 40 LAr a1 9on 8 14 37 M2 284 29 OM 485 OM 9104 Oil S /j 649 1T 884 977 229 137 284 213 2441 2378 8W 4418 5799 Plubtul 425 675 710 851 I= 1228 1782 198 9 n 2822 8i 7lii 14482 Othr hor.. eq 8 81s 22 I0 254 2211 8145 879 4M 4W7 S 2E0 18 CowrbA 42 87 772 16 141 149 2l89 410 515 MD 2 118 Oth croqs 2.218 118 28S M 39 4726 147B477 12 29 1 272 of wd:ev 449 60 4 871 745 188 1418 2897 2289 8854 5848 1G4 74M 1UT.AL OS 8 71B OM I 1261 182 I2 2388 279 3 2 77?7 2 t \ § 8 N O N _ @ $ b * 0 n _ | R l 0 @i ˘ l 9 a > | S $ | ˘lsl l B II .4l$ "^^^Se° e^@ -° "889l llgt l | lg -^"°°>-oo°sRna-o N88a>|8p^fiS@R.0Qs | 8 § s b NO - Z - O - @ 9 ^ @ ^ s ^ ai" 9 9 " g 8 3 St§ " >|˘ 0ffi|S 0 1 l g l .~~~~~~~~~~~~~~~: ir lal | "^"00-hbkRa@ 8aao8llp|E!s * a iZo ontrr" r c4sU4O.4e.4Qeggp9t%f% Pae _I | ^Z00s*^B;X0 "° sg5a3Esgx p|p;8 !| g ls H eS R @ -r "0 S suw!SSls|!p Ill "000-rn na Sl9>glltl l | Zs s N"^Z - ~~~~~~~~~~~~~~~0--"o-l 8i8||2@ p | g488 ffi .4P0V4m.4:2.4ter..m3w4Xe s ga8st@'@ X l I III'"_w0o-r oSilR@ gfta$ W bSfiglR | Rkka| *-o@^a-u><,..Dss3,abgn*@ Xffi^augsgBWgy!QgRk>"0 N | fi..*i$ !0lggWX S i l - 82 - Pap 4 of 4 19 17 1918 19 M 9O881 12 18 1M84 819 E 1# 1 7 1898 o of WIE AM O) Sad e*bon 65.4 -13.6 -1.1 7.7 40.6 -3.8 -2.8 9.7 1.7 44.8 13.1 0.2 Smti fibs -7.9 -4.6 -14.8 5.7 -14.2 -17.8 -88.8 -17.2 -15.7 6.6 10.0 0.8 T_ltt 18.6 21.2 -4.8 -1.8 -6.7 -4.9 18.1 -18.4 6.8 -7.6 -9.6 -1. Coffee -18.8 8.9 4.6 -4.4 47.3 -3.0 6.0 -7.8 -0.9 12.8 2.8 -14.8 wr eve 1.7 -12.7 21.6 1.6 -24.7 27.8 -10.1 27.4 -18.8 -11.6 -4.0 14.7 Tabs= 23.7 4.9 -4.1 -4.6 -1.8 -1.4 -1.6 -18.1 19.1 4.6 27.2 -3.9 cahinu*e 4.7 -24.6 -21.1 40.1 8.0 -8.6 6.8 4.4 48.6 -2.4 68.1 -6.3 FPyUim -21.1 -17.8 -7.8 1.9 288 -6.4 -16.4 4.6 6.4 -12.1 -90 14.7 1ths 14.8 -12.0 17.4 0.8 6.6 -10.1 4.2 17.4 6.6 2.9 8.4 6.8 Yah -7.2 -14.1 27.8 24.8 8.4 6.6 -8.9 27.6 12.2 -14.6 1.4 6.9 Paft -9.2 23.2 -12.8 .1 -M.8 68. 9.4 1.7 19.4 18.7 12.9 9.8 Cli- *WWI- 12.4 29.7 84.0 -8.9 -13.7 W.7 -18.8 56.9 -8.4 7.8 8.4 -2.7 Lag..ss -14.1 4.7 6.9 24.1 4.6 6.6 1.2 -2.0 82.6 -18.6 16.6 -4.6 Oil SasI 6.9 8.1 8.4 7.7 -4.6 -12.0 8.7 -0.9 8.0 4.2 7.6 8.9 Pli*Irm 2.8 2.8 2.8 2.9 2.8 2.8 10.8 -4.0 17.6 2.8 2.8 2.8 OUihi hor. crtn 1.9 1.9 2.8 0.9 1.9 1.9 7.2 -4.0 18.2 8.1 2.8 2.8 6o=o" 2.7 0.2 0.8 2.8 2.8 2.8 1.0 -7.7 21.7 -4.9 -.O 2.8 Othe crap 7.2 -0.9 6.4 -4.2 9.2 7.8 14.0 -4.9 9.9 4.6 -6.9 2.8 of Wc :iwm 14.0 -7.1 12.7 -14.6 20. 13.9 18.6 -8.7 1.6 6.7 -16.7 1.6 *s 1W Q Q' 2.9 1.6 6.8 0.1 2.8 0.1 8.6 2.6 8.0 0.6 4.6 -0.2 MIUC -10.7 -1.8 2.7 7.9 6.2 -0.6 -0.1 0.9 -6.6 2.8 8.1 6.6 FRO f -12.9 -0.8 *12.1 18.9 4e.7 -4.0 67.8 0.0 7.6 9.4 -A.4 2.0 FRIB 6.6 -12.8 -11.1 14.8 8.2 12.1 -3.6 6.4 18.1 18.7 8.9 1.6 HNaRD 0.0 -.41 38.0 23.1 2S.0 -3.0 -17.9 0.0 8.7 62.0 10.6 86.7 1WAL 3COJIU1 ,1 0.8 4.6 8.5 2.7 2.6 -0.6 1.2 2.4 6.8 2.8 8.8 0.9 1DPAMN OF a NI. OAN IXDB AgIQLttJ VAllE AMEO ICVIL 9,045 11U,81 12,1 14,72 18,63 283a, 28,449 82,737 41,295 61,l31 84,13 2,941 178,288 A5D01 9,01310671 2,39 414,441 8,109 2D 42 3,58 8,31 40, 69 ,37 7, 872 12,75 M12,89 fiaA4L./AsUE 6 268 1US 182 98 168 102 S8 12 13 1us 17 IN OFICML 9,0t 9,16 89,98 95 9,41B 9,51 9,8 9,91410,812 10,DM 1,67 12,5 2,95 ADU61E) 98 M 9, i 9S, 89,1 9,183 9,820 9,757 9,877 0,118 10,764 U1 . 11,423 U,I aIH RA1A3 UFfICML 1.1 -1.7 0.8 8.9 1.0 1.8 2.9 4.0 6.0 6.7 4.4 4.6 AX0151 0.8 -0.6 8.6 2.7 2.6 40.6 1.2 2.4 6.8 2.8 8.8 0.9 OERARRS FICUL 168 122 139 112 177 214 74 83D 4 723 18 1414 ADA1 26 126 188 1S 177 2 1 838 897 5W 7 73 98 1414 A TDl /2 MWVA VAA.O. RAUD. /8 FDI UO S M$ FAN O OM WM ; UT U0(, l1Rf, HIJ A F1I V.A. ROI OFZCAL A.M1 /4 V.A. P OUiA1B 9r0.o. AC6A1 M S!TADI 2N W SP1 /5 SE C:\TZA\O(lrA0 18 FM IR.51wS M OM DATA fIR LMM&, 0IH.HRr GKIPS OM CIALS, RU OM MS. /S SW C:\7U\O)45 F OR AD1J51 3M O1M S SIRE: BIJ OF SIATJSZ W\IIIS; HN 19 98 I; IFFICUL T1. - 83 - Table 6: DFIUAL FOD COPS 1976 1977 1978 1979 1980 191l 19e2 1983 1984 198 196B 1987 1988 VALUE ADDED I-LIOdS aR8NT T9H. Ma ON OFFICIL C RO ) Other care s 432 565 828 1122 876 799 196 1420 2 26536 884 4699 4630 Legumes 69 621 982 IM88 1437 1852 1821 1646 2039 8958 4520 6523 9104 Plantains 485 676 710 651 I 1118 1782 1989 1921 2521 8082 7555 14482 Other hor. crop 686 81e 960 1260 2545 2211 8146 8799 4099 4927 5809 18250 16626 Coconuts 42 a7 77 208 141 149 26 886 412 515 629 650 11s TOTAL 2182 2662 s822 4406 499B 56e2 7899 9220 26474 14651 17164 82568 45877 VALUE ADDED MIULTONS 1976 TH. MM N G CFFICAL Wa08S WOUPW) Other eeals 482 486 629 848 650 581 788 682 859 64 784 758 593 Legumes 69 62 648 565 726 679 724 782 718 962 794 917 978 Plantains 436 447 459 472 486 499 514 s6e 544 e68 654 672 691 Other hor. crops 666 678 691 710 717 781 744 798 734 852 878 9a8 928 coconuts 42 48 48 48 46 46 47 48 44 64 48 45 47 TOTAL 2182 2176 2870 2654 2628 2516 2761 2776 28 8178 else 8298 s813 VAI ADDED (ADJ1D) CRENT 2210 2969 8s56 4461 50 5s69 7998 9836 19865 14834 17818 82912 46450 1976 PRICES 2210 2208 23 268 236 2547 2796 2811 2934 8217 8147 a888 8174 DIFFERENCE CURRENT 27 a2 44 55 82 70 99 115 181 la 214 408 578 1976PRICES 27 27 8 a8 38 81 85 86 8S 40 89 41 89 NOTES: /1 NON TRADmONAL FOOD oRoPS ADJUSTHENT OF 810 MILLION IN 1987; MIS PSRCENTAGE (A9CUT 1.25K) ADDED TO OFFICIAL ESTIUATES OF CTH CaALS, LEOAES, PLAINTADNS, OiER HORT. aROPS, COCONJTS. OIL SEEDS ADJUSTED 85PARATELY; OTER CROPS ALREADY INCLWE A US ADomoN FOR UNCERCDUNTING. FROM BAOAOHWA ESTIMATES - 84 - Tabe 6: CAR4 - Of 1918 1979 A23 1968 19B 1984 19 98 88 qmnm Qir FFaCL 1A 488 Z SU 451 9 48 33.8 O .1 424.1 17.1 21 208.9 3W BI( 5T39.E /2 433 41M 0M 4519 OM6 4835 33 SM1 4241 12I 21.15 2ND OM FRICE (1MM1) I ua M1 44 44#10 eOa -MW GM I lM WFCIL 6.2 6.9 8.8 3.8 .6 20.8 14.7 1.4 3.7 19.2 2.8 45.f 5.5 WC 831)E 62J. eo.5 87.8 21.0 237.6 28.4 4.8 I4.6 237.2 18.7 3.0 45.2 55.9 aGs 0J1fr w fIFOiL 6.2 4.8 4.2 6.4 7.7 6.6 8.9 8.9 6.1 1.5 2.6 8.6 8.6 B*K EIWE 62.0 48.4 42.1 54.1 76.7 56.6 3.2 88.5 5.0 16.8 5.8 86.6 85.9 *:X MEOD (P / XCFICL4 4.9 6.7 8.8 19.8 27.8 19.8 18.9 18.7 5.4 9.7 19.2 48.7 62.7 VW4C OWN1E 49.4 67.4 83.4 12.8 278.2 197.9 18.6 187.8 263.8 83.6 12.4 4a7.2 68.8 ':lEAlOE IP WFICML 4.9 4.6 4.9 5.1 7.8 6.8 8.7 8.7 4.8 1.4 2.4 8.4 8.4 IMW ESIMMT 49.4 46.9 40.0 51.4 72.8 68.8 7.2 35.6 48.8 14.6 24.1 18.8 84.2 VALjE MM DEW a, 44.4 5.7 75.9 17.5 264.9 178.1 15.6 3. 229.4 85.9 178.2 80B.5 474.6 9 44.4 41.8 88.9 48.8 65.6 47.5 8.6 82.9 43.6 18 n.6 35.5 85.8 /1 a WsrnB ,J 5I1E 1HT N OF : R JS OJI IF CIaL Wa1E WVLE M O/ M WO1FUr MM 10 GE.96 - 85 - Table 7: 0RSES OM E2t "r Al4ur In QatlttW d rum CApb for NbJmtl Aoa.u bvlsia 193 1577 = I= lm n 2I323 19 108 2no 1I8=493926 113o OFictel EB E*M. Q(d) 90tlm 9M OSf 713 23 2191 19 94396K 4649 n7W rn 2r 2=. Son 1 78 674 482 19B 42S 42r3 OM823 49 an GrVwdwh Ml 422 1 39 GM 172 254 2M 96 57S 75 40 65 Rol led 0rauut 9S24 MG 26749 O7 2 29 M5 OM 7 1MM 0 WS 41 8OM SM* SU ON8 60 3 69 IM 198 438 1 58 6012 13 NO Cotn Sod 21 IW a# S2 27 IM X6P2 W ON M M W f 6 6.2 -s s etm OMOar I8 2j 41 6S SmI U7 63M2 638 GMS 87426 M2 Wa MO11GM 27 247 96995 4379 4573 Orcwd,Ai 415 183621751289 478 2867 2914 OM 4568 WS lO 4UW Sit led G eFmtu 1SM6 42 5 314614 Z666-2 11 s 94463 2U9 S4-- 1784 4B12 6 UM N 46 OM 3 t34 U6482 6 81Ma 4; am Uw CGo dSnued I61431= 44 VW3 48Ł 45 4WS 36455471381 O 49_0 Prim (tAM1) SIm2 4344334 4 4 573 3141 VW 2736 2uadiAs 3 4110 In4W 434UW 4GM I432623 U 9hl led rWdtl 39 27M 6 OM 4111 4I M 73 U US N 2M 24 Soban 21#0 22a 99 22x1 _= 22 a S OM 144 M M 0 M Cdm' Soed 73 73 263 M6 I63 2I 10 23 43 43 4 U 743 &om pt (ml Iam Td) Sflmr 88.9 81&.6 67.2 98.4 1.2 01.9 3.2 21.8 62.6 25.6 18. 28. 193.4 S_ 23.1 283.8 271.6 2.8 21.6 2.09 8.2 39.7 R77.8 7.1 5.6 75.6 2811.7 Gra _ds" 8.8 6.8 46.9 64.7 289.6 6.7 11.4 21.2 63.1 G.5 23.9 760 W27.2 Sol led 0rsMf 487.V n 6i.. M6.4 749.0 03.2 976.7 M.A .6 23i.224.221.092U8.8 W.8 SAM 8.6 20.6 7.6 288 2.6 .28.6 8.8 7.8 7.6 21.2 44.1 78.7 224.1 Co Sgn 21.4 9.9 16.4 1.4 8.7 6.2 7.7 8.2 8.8 28.8 24A 23A 1J.4 Cip 7.8 0.6 7.8 5.1 2.4 7.1 9.6 6.5 6.6 6.6 10.6 1.0 6.8 T-el 01 l0b ((P) i.2 7M.2 A.2 921.62 4.O 2B.0 23a7.6 2712.0 .1 35.2 4W.1 6es9 CuIp*. ImI llam Tab I=W) -al1w 88.0 38.7 84.8 57.8 92. 61.7 46.1 45.4 21.8 87.0 W.8 256.6 28.4 S8. 2.1 15.6 U4.6 138.9 74.8 2.5 72.S 73.2 2.6 54.7 49.6 B.O 91. 3raw_iA 9.8 6.7 28.4 42.8 96.8 27.9 4.6 8.6 18.7 9.8 U.9 6.5 8.7 Sol led Oiu,t 487.0 449.4 492.0 475.1 4S8.5 6W.8681.5 59.9 56.9 .9 57.4 642.7 W7.9 S-9bum 8.6 9.6 6.7 U.8 9.4 22.0 2.2 4.9 8.8 6.8 9.4 18.9 2.8 CoUm Sid U1.4 9.8 U1.6 8.6 8.6 8.2 8.4 8.6 1.2 8.6 8.7 8.0 8.1 cap,'. 7.8 8.7 6.4 8.8 1.8 4.9 0.8 0.1 0.2 0.1 1.4 0.2 0.6 Total Oll_dm QVf) aM.6 .4 65.0 718.4 768.8 732.6 644.9 38.1 9. 73.2 1.0 677.? 797.6 MM* 01 beds btm,m - Bmid an Kl I Ion Prodmotlan Fltw A GI itow mti1Au 199/84 -293/8 QJI)) Sm Sam 0rowdu*a 13603 CMzan Sed /2 OM OM k he SEblwt.,983 /84 - 25/88 (Ul Slwr 758 s& 4"3 G Sd /2 579 R"tlo Kt I ln)^; S9o aot 6.8 Se. 8.7 ˘rowd%ite 8.1 Sybemi 5.6 cw Sd /8 28.1 Notsi /1 ES emb eod an officlal pna bas; KLDQ esUtet are prodiction flwm ar teors re pu memuIitic to far tor Nttioal Akmuit. tb a4umbei for cap. /j2 2 7 1tein. w1i owl aI 9ow 99w1 Wr art owI oWI zwI wi wi *r9 Him *w 'wS US grow wrtr wm I,m s orn m If=* ar )m z s no mani j,ep elmrow *** *uSIT" I'm Wm e'es1 ron uon rim SOIt lato Wm (in'o AVMirc U8603 slaw VI= el= SW elm 9'*9IWm 'lat S*IM SIAM rell SUE wzuin3m am w NIEIM *dONSMU H1 SkIMY 8G9E9 VIV O*98 VIM 5 SIT 9win B'-S ge1n O911 r 6it1 e'lm SUE AEIEU rusm W1V Wmn elmrnteaz sla rUst Vr6VT eSB ri8 rrn gen,r WWI ue jm mm le*9 elm9 6FL eSlS i'9 0-81 8-11 Ble 9t*i rLST 1 glen Sgln SE 3U4 so a am ast 01 U 9 mu 90s1 aK UK U em am ut 9 mela OL JsTW alSUL W69M FU FBffl 91*9msle OW " S g WWS I aSK 3IIT atls AMfLSUT Give woLS sm6 e'si *'u wsi slow raic riii rmu m rui;' eto ismn3 J aDrt (LO Ng) AWl AWl 3 31r Ti 3Wr 3Wr lWr 3W 3W 3Wr LI' A *Jm S'I 61*6 619 SW 181 0 0S S'L*: 991Z 8031 9 l 911 z-m 5 194 .s *LSL m r9 a8m r9se 8 vm F- 1*6 wep r 31 S r'm aSEl a9tls AfL W AWl AWl 1U1V UM l W 1 XITNr 3 3Nr 3W ANrr Ai-r ass 9951 95 mu am mu gm am sr am am Am Uan 86 (iDt *SR *au ssiNmv aL JOUNIE Wi 591a GUIS else ala- Ilse 511- rot 663P- SIM 9- LITT I's *'SI Ye- Mm `m* 1GWm rmi FEB- 9'S 55a* S'a6* L19 6'E5 gS'S *31- 1'9* 5'9- Fl-, Ow S'4 119- SO 91- 111g 5'Si- SIM 919- 9we fWu ~S`- SI- ~MN mi *81- 8159 8 als 8 1- sla S9U- 911- 111 we 015 '-' win zw-m Ss-O S*Z S'S B'S Fl S'S s'S S9- B'S 9- ZLS Ss S's 9 91t 9W1EU glen *ra wS- SS rot e8'- 9eg- 818 FtlB- 9'L IlS0'8'- 'NW 1 8 'eS- 618 I'M I'S- nI S'S 'ln- sla all GIV 9111- tIE SEGUED WOO S0-8? 11Z I'UB 615 6a Z '-3l- SUE G Vat 6151 11 t WM S 0 'MLA "SOM LWn 6'a9 6'w 9-91 raic rUBz r' won SW r'i vat r'u fUE O ra' rsin Slow3 r13 win rUB ST 13 g SUE Gist SUE w13 wUE SUE am 6r69 aso3 r'u* o'm wis waLs r ris elm r6i1 ron SUE SUE 'WiIN3sOi0 8low s's*sS 8-8B m iW- 8 81Z r'm S'19 Wm S9-- I'm 0an WU ZWZ NM IN" elm39 s 8193 w slaw I'mS 11 m FIR9 SIM riot S''1 0won I,SR WmE 'WIJ18f1 a-Z 19- 63Z 91 Z * L-Z 9Z 631 91 9Z L 13 L-Z E+ gI Z iV NISUD19L NID W Si- rQSV ro r9a E u8 9VW IN emE 6113t 1-81 WSo Soot ast iii s-8- 8nu 9t anX 8- m am a a sn-i lt9n 1*11 Su Mt OUL VW6 apt 3g1 WE amG UK1 6am gmI SUn 9s go no Tag m a9* to rn St mL m m m so-UUm 9913 598 US 6- * tZ6-0 US -93 918 LU S '5 * SJI~16 PI -39 SE 6 9163 U8 31 95 1* *6 6 915 W-80 Zi -tm9 9 am em 1t g's me m w w8am am em 5* am u- -m Bos 95 UK gm St Gm a Mt UK U 6 am Łan am1D $miw gm -?tg ftU9 638 UK 39E951t U91 e365 5f11 K *91t 939 UtI LiVE 90410 *19 '* go n an 8§ m tllll ame ue t an n St sZ8 an 0ttItL `o -Z S ML an ea ist O" 9 S t mg m LTapt 5N *9 I8 63 IB8 09 61 U 98E9 USL 1 91911 091 8i 1 O 38911* US8 UK8 SUE 0889s SLUT OWl eez 9*6 63 936 993 'W IDS 61 gU gm 11 aS * e 18 1 31Z NM* TM Us an 91Z J II l axe1 SER U aS US 98 UK8 on Gm me Ut sE 91 99 t m Em am ae3T a6s3 8 5*1 to M w 381 g1a 951Bt US iw stetiow an so VW as2 Sl S 9It mtl alpB m11 9=c go n QEU gm Y ' S0NM SO OE9 am ong am 8 80Z 8i8 gm gm gm Mt *N sm - 98 9 - 87 Pop2 a 8 153P w1 wu8 im@ ur r r 194285 r w WM wIRl(Jr FaM 1ALEW a E8 n a a a a a a a S Sl Cta 7 m dEm16 7140 6 78 2 m 5 4 7497M CA%J1S a65 6a a 1 a 1 a 4 4 42 27 2 a 2 23 FM DO.ABW U IAO L 0RDO, GO OEII, EIC. 1E A.EfF 7.6 8.8 10.6 29.1 20* 9.4 8.9 29.1 6. 9.8 8* 7.6 7A 228.8 S8.8 W.1 98.0 18.7 23.8 R.8 16.4 2A .5 998 2.4 2.5 98.4 CASWiJ1 14.2 22.8 9.7 7. 7 *7 7.9 6 6.4 6. 4 2.7 8.2 1. l61lL 28S542 9.7 .W82S25.75 U. 27.542 7.27 24.89 2252 =.I22 M.S53 .48 WM AUStS CSIUEI1D4I amo RmI Q8II. 1 BmDTtIS 142 .I94.7 27.9 11.9 248 53.9 m.9 82A. 899.2 41.68881.6 L51.7 1 f 2A m 7.31.2 6.6m m8a. 4.4o .A69.7 64.22 39.2 I 2 .6 5 A.4 I1L CMN AW /2 277 J2L.E4 8446 420.88 432.4 W6A6 .216 321 19Q4.6 5W .2 .2:54.9 298.2 NN4E4TXAL /8 458 9 749 948 23 1445 2 e8 3219 2UN 2 4225 O am It 16 235.53 8.2 2MA. 26.1 98.4 84.5 215.889.8 8579.6 3L6 W9.4 M.9 LUN /4 215*4 8M 6.6. 1 A9.44 8.S 69.24 8S. 448.99 578.184.28 14A2.8 1498.9 ROM, V1R, OHM0 24.711=.6 52iO.8 17.4 M6.9 2UP.2 23.2 2I8.9 212.8 2448 43s.2 86A 769 lINX 081RCX14 )2ffiG.7 8MLA8448 4.7 483.1 524. 615.4 454.857.1 787.6221. 1WO. 133. gClbW I8 ..S SJBDDN 95 2215.4 28. 24.1 2278.6 148.1 4I 6 948.ff 74.9 29.63 .1 28N. 1498.9 DBM11L iS 2 26*2 28*.7 5A 271.1 84.7 25.7 28.9 3X.6 88.8 8M6* 441.7 W4.0 RJE L M. AM / 277WA 23.9 K1.8 29.2 39.5 826.6 8* 888.1 847.2 M.8 872.6 387.1 463 NMEENM& 46 O48 63 690 478 71 93 83 457 8751 6E1 53 OM 015316 2899.7 23.6 557.9 277.0 1.8 2499.9 M8.1 981.99 28.7 95A.42 2215.5 9S1.2 1487.9 14 /7 27C6 23.9 2.4 284. 224.82 .6 3.8 2.9 8.6 8.9 21.7 22.4 256.7 M6, WEU, 01116 / 14.7 21 23 1448 m8 228 1478 738 8MS e6 4 N 7W 13 Mr. M0R (I918 MM 3.73 1 .6 3 WA.1 86.6 82.1 34 .8 62 .9 274.7 :ES. 24.7 2R* 2W84.9 XUStS ECAWS Da 20 1158.85 27.442 I.671 1LA V6.14 4.72 2a4.B7 277.*8 5.67 4M.95 5.6862A FESMSOM ). U.22 117.60 184.54 7.782 I4.57 24.16 33.0.8 12.7 8=.69 868.42 489.82 37.26 tIML .IN A MW 25 1228 U28.2 18.17 14JOA 3.76 212.97 244. 256.28 8M. 42.85690.82 8. NMEhDNM 1f0 lM8 2J18 .2A289.495.85 22.673 1 7.91 21R 849A2 42.79 495.28 O. o003 16 228A 52.28289.6 .82 1M.18 30.9219.10 231.28 8WAS 4.8112.486 G5.GL 0 s2aM.682 l.68 148.83 80.1962.00 0.88 22.918 5. W 488.32 .6 54.71 AmS, WMI, am 30 123.3288 .22 29.78 62.91 21F 214 W21.2 20.8 883 48.8 474.31649.3 lOMM. GtE DIOP 25 221.453 U.9 13.1 23.32 275.6 2.3 SL.M 39.79 83.U 448.63 63.68 M.81 ;IB SOF aNSl8RCr VAMlE MM / EIRXI1 VA1E AlM LL I" S,mD2l 846.2 =A.4 484.6 8.2 f6.8 87.9 63.6 3A.8 254.821485.6 * .42 7 28.7 8.4 1ESKW4T 71.9 3Z.4 6* 228.7 224.4 178.0 6*1* 170.1 8b7.6 4R.7 498.7 719.82164.6 tM.L C1 A.MI 29.1 28.8 148.2 176.5 286.9 217.4 298.6 8S.7 440.6 .7 655.6 18AU8. N3l*fiSDSNTL 254.2 231.7 249.4 814.0 875. 481.5 457.2 29.0 45.9 462.2 849.228.8 219.6 0MI16 6A2* 748.1 7.1 7.6 978.4 .6 82.6 7W7.9 8O8.9 A.4 5A .428.4 89.9 LNG 91.6 2S.8 96.2 1.7 U4.8 212.9 175.9 247.9 U10.8 821.1 436.1 43.5 499.1 MM, WM?, OHM 6 5. 1642.9 M3.8 671.8 654.1 N.7 96.1 W.1 MA.1 64.8 1440.8 2221*269.7 1W W6JCI 0L 191677.8 12E0.5 321.6 U 3.7 28.7 27J5.5 2U.8 2.7 3.2 236863.8 4N6.2 64.8 m tll VALIE aXL W 1Sf) GMD= 845..2 4.6 8.2 442* 456.6 17.8 488.1 88.6 48.6 W.7 472.1 63.8 37.6 FELEBJ& 71.9 72.8 78.6 18.2 90.8 S2.1 18.6 72.9 .6 2X 12.4 147.1 274.6 tIM. L OMU4A 129.1 322.1 226.8 23.6 82.2 28.1 145.2 24.5 29.8 24.6 25.2 288.6 178.8 N ?lESEM.. 254.2 229.1 288.2 6.1 281.1 28.1 2.4 U21.2 25.5 U2.2 183.5 2.7J 23.8 amI= .0 W8.9 621.6 6.8 .8 4.6 58.1 89.4 844.2 82.6 404.8 885 496.5 LI QL* 95.2 06.7 78.2 71.4 117.7 .7 64.6 72.6 80.9 18.8 74.7 S8.1 MOS, A7R, OHM 541*62.7 442.1 4908 7* 878.8 490.5 246.9 21.4 22.6 81. 25.8 6.4 1W. CW6R. MILL I5 16M7* W 1.9 64.1 3.804*29 2.85 4.2 699.9 77.6 729.2 8.9 .8 I.1 - 88 - Plp a of a 183 19w9 us iwv 191# rn 1084 iou# u9rS 108 io am1N 22.8 -0.8 2.6 8.1 18.6 4.9 -5.2 18.6 -8.9 18.7 26.0 4.9 111R. 1.2 1.4 19.6 2.8 27.5 -14.6 -=i 49.8 2.0 15.7 14.6 1.8. RIM Oli AMNr 2.6 2.6 2.7 2.8 2.9 83. 8.1 88 8.6 8.7 8.9 4.1 NW43MOM& 48.6 -3.1 22.9 8.6 4.1 -14.5 -42.8 13.6 -156 28.8 5.8 -.4 Olil Wm6 -8.9 -1.8 9.2 6.9 -18.4 16.4 -48.8 10.9 4.6 5.9 -18.6 5.1 W* -7.A -1.2 12.1 -6.7 4Si. -3.6 -08.4 12.8 22.1 5.6 -.8 27.8 FO0, VIlut, O1 -4.4 -15.4 8.7 U.7 -0.6 23.6 -4.9 .0.4 -14.8 88.7 -17.9 U8.8 l3lL a6rEN V.A. O9M8 6.6 -18.4 1.9 6.8 -4.7 8DA -.9 1.0 -6.1 21.9 6.0 12.2 CUPAWN OF aM.C R)ICUL (XIEiRXflf VAMJ AOMS f. a R D 15 A 88 A6 =A In 39. A489614.0 Il. A l6A1 0 A8 W.0 V mi 581.8487.64A IDJ. 0611LT QMLL T19 97.8 12.6 121.6 18.71 5.7 75.5 22.8e 9.7 in.2 2.0 =.8 4W6.2 84W.8 O1FADJ.51 8J 80 89 U8 U U 888 82 81114 W 74 71 ff. aN Q 176 h 884A 91Q6S7A 8. MA .0 932.60.2 3M.A 6459A a6A a.0 75.0 A.0 757.0 WdJ. OM nmL 197 I" 7.8 I.9 824.1 163.6 304i J18.8 1A04.2 A. 757.8 719.2 878.9 63.81 i.1 OICL I& H RM E 8.6 -14.4 MA -4.8 -4.6 4.6 41.0 S3.2 -0.9 17.8 2.8 6.0 ADASME 011141 WE 6.6 -18.4 Y1.9 6.8 -4.7 SA -8.9 16.0 -6.1 21.9 6A 12.2 A XnIt frin15 prlo. with auep resdetial t&Iator /2 Tnted frn 18 prim with lo cut tamirg laito /38 TIEa fra 1M9 prim wlth ult.I4ort disTar /4 Official Slatar uud to ife1e97 prlm /f Cuiu prdictln uad to aftsolato1916 to us (Cu.t A bled. aaatt for about 2b of mutriole /8 Fra 1978-, nual population goth (a 2.4%) plus .I for ltprved qal it hollri; fram 190-8 Inaums m population grmth kV . for lqwoWd qalltu (m lET 2, p. M) /7 Sinn wil I holsr fuam Imwstuwit ua lawd lpomm , cautrclUon of *t;Op, dIalopwait of taes crps Is not Inrlud, I3 of oamstm prlc grow o_ut of to cife wid FIm fute Is ad. /8 Officil mante do rnt Irnclul d;intte for cmitoW abara;tlon for adhobs, comainty piiIs, ; S pwr yr ad. /8 Ral am cuit autrucn ratlo of alue ackd to Mr=e apA ts 49; for el 1ow It Ies 88.81 A of - 89 - Table 9: tHUNTIN 1978 1977 1978 1979 1980 1981 1982 1988 1984 1986 19o 1987 1986 OFFaAL hhTIN V.A. (MILLIONS TS CURRENT PRICES 29 29 82 a8 88 44 49 74 77 es 74 so 8 1976 PRICES 29 29 20 26 82 40 28 28 28 26 88 42 67 DEFLATOR /1 1.86 1.00 1.60 1.48 1.18 1.10 1.76 8.22 8.86 8.40 8.88 12.51 19.82 SAAO4WA ESTIMATES WROPHY HLTIN (SL MILLION) /2 IVORY 5.8 6.8 6.8 6.8 8 8 8 4 4 4 6 6.66 185 RHINO 7 7 7 7 7 7 7 7 7 7 7 7 18 TOTAL 12.8 12.8 12.8 12.8 10 is 10 11 11 11 12 18.65 148 EXCHANGE RATE 8.88 8.29 7.71 8.22 8.20 8.20 9.28 11.14 16.29 17.47 82.70 64.28 99.29 RErAIL VALLE OF ILEGAL TROPHY HiNTIN TROPHIES TSH CUR. 107.8 10B.1 98.7 106.2 82.0 82.8 02.0 122.5 168.2 192.2 892.4 877.1 14696.2 TROPHIES 1978 TSH. 107.8 106.1 81.7 72.0 72.9 76.8 68.0 88.1 60.2 68.5 01.7 70.1 760.5 VALU ADDED /8 TROPHIES TSH CURR. 64.4 68.7 69.2 68.1 49.2 49.7 56.7 78.6 160.9 116.8 285.4 56.8 8817.1 TROPHIES 1978 TSH. 84.4 68.7 87.0 48.2 48.7 46.2 81.8 22.9 86.1 88.9 87.0 42.1 468.8 NDTES: /1 1986-89 DEFLATOR ESTIMATED BY RATE OF DEVLATION. NATINAL ACCOLM DEFLATOR FOR HU4TDNG IS DECREASING IN TSH TERI) /2 ESTIMATES OF IVORY SEIZWRES GIVEN FOR VARIOUS YEARS; ESTIMATES OF OTIW YEARS USE SI!ILAR FIGURES. RHINO HORNS ESTIUMTED BY USDNG iAGACHWA ESTIMATE OF DECLIE OF RHINO PPUATION BETlIW 1906-08 OF a,0m, OR 876/YEAR. VAU AT 1986 PRICE OF 818,768 EACH. /8 ESTIMATED TO BE 6OF RETAIL PRICE - 90 - Tabl lI WNO Fip 1.of2 1gM 297 198 109 208 191 209 18 S25 238E 105 19fi 0.O-WSWIN 3MWlS ( FIWf=m A1 GU. (1nf C.) A/ am1am mm am am am ami am am 3ms .S am MD WME M.D WS OaLmI) 4. 4.78 6.19 9.2 19.47 14.72 22.60 28.6 21.64 10.1 21.78 14.30 14.3 M.D (5 WAML 8.56 23.6 23.6 48.* 9;.7 78.2 07 W.S 23.9 23.6 21.0 961.4 6.1 MM WM /2 MD (131 WILM) 28.4 12.5 .0.0 83.9 67.0 M1.2 48.9 98.8 74.1 74.6 81J.7 8JIBJ 3.8 VAIJE~ AM/ 9JB4 Rta 22.8 2. 29.9 82.2 98.7 48.7 44.6 9.2 70.4 70.6 2.6 18.7 IiE.7 m 228. 18.4 28.4 18.4 28.4 28.4 18.4 28.4 18.4 28.4 28.4 18.4 18.4 WDX;lL ST'W5 ms aURrw o&S I" DbDO /A 25.7 88.6 12038 2456 S54 238.9 214.4 108. 8S.6 36.8 83.8 6#7.6 W 3ES is 0.8 098 09.6 2.4 1.8 6.2 0. 456. 4.8 2.429 4.83 3.33 1.28 a flRW KL m ! 1 0 DI29 /A 25.7 26.6 207.2 12.2 98.4 23.8 207.9 77.9 1t.4 78.8 s9 65 87 31lwaES *S 0.8 0.23 9.34 1.1 1.182 1. 9. OM UAU 9.43 9.23 9.8SB 10.45 9.84 UMMSDN ED4I /4 0211 3.6 23.8 5.6 88.8 45.8 86.1 82.2 27.1 68.8 83.8 98.8 1.9 2.1 W DNES 9.0 O.9 0.1 0.4 0.2 0.8 9.J 6.1 0.6 0.4 0.6 4.4 6.4 mO M QW llowl ISJIB T0 DID 28.5 27.8 28.1 2.8 24.8 5. 26.2 1.7 25.1 2.8 19.4 8.3 6.8 WElDS 0. 0 92 0.2 9.2 9.2 9. 1.7 0.1 9.O 9.0 1.6 1. VALE AM QmimL 130 AS D14A0 20.1 22.2 21. 25.8 19.7 15.1 18.8 lA.6 22.9 28. 3.8 3.9 57.8 WSIIN 0.0 0.9 0.1 0.8 0.2 0.7 .9 5.7 0.8 0.98 9. 4.1 6.1 WUE AD Qm±m 10781I" mOm 10.1 12.9 6.9 8.0 8.8 6.7 7.0 5.0 6.6 5.1 4.6 8.6 2.4 WEIWD 0.0 O.9 0.9 0.2 0.2 0.2 0.0 1.6 0.1 O.9 0.0 1.6 1.4 Mm M.L B 1MN 20.1 12.2 U.1 28.1 29.9 25.8 28.8 17.4 28.6 28.4 89.5 5.1 d2.4 1lAL MWL 12916 10.1 12.2 6.9 8.2 6.5 6.8 7. 6.6 6.5 5.1 4.6 5.0 8.8 mm= MA1 S ME) 1DITO NO IDLI 154 B 2.4 27.9 3.1 48.8 68.6 64.6 68.4 77.6 96.0 84.2 0.1 422.8 617. MLLD 296M 32.4 26.8 28.8 2L.5 29.9 3.2 98.8 29.9 29.9 8.6 27.9 18.4 2.1 - 91 - Pap 2 of 2 I1 17 in we l u i Is 19 u M =4 I i I m I I EM MOEME CrOI) 0.e? 9.= 7.722 S.217 6.4 *.24 9.2m 1.48 A. 17.472 82.M 64.25 U.2 lON(CN DfflWNUf 31.02 14.72 8.24 =.U 07.87 4w.75 S65. 4247 8.85 W7.18 A7. 446.12 471.5 MD FR1E ( 2I5.01 124.46 241.2 209.0 492.7n 8M.W 8.i 4W7.52 6u e * .78 n2.4 2M.8 4M4.60 alG FRo. 1.00 1.17 1.48 2.41 4.77 8.64 8.84 4. 5.27 5. 2.B 27.44 41.A VME A = wr MMRWM FM OM DVIW 0.48 SLT 0.10 C281U0 0.94 O 0.95 OM O.S6 NIB: A1 sm BMUE OF 314 ml=IN 1I 21J. (D IWD=IO D!V= Bt me lNDlN E IMLS MU WV 18 fR I9m. fRImS WARS ASI M 1B1E SE FWMUCIN. /2 OM IJM AT IC RM1 D M BE 7 OF MIIA E IA VAE ta VAIJ. MM/M WIRff 5TM13 1O1e .15 /4 BW 5 BM1E OF 15 OF DX SMM4 FROM U1114 DWO MM E80 WR; IMS RlE WLI 10 VD7MO NIO WM /, VM.LE MjBS WIFUF ROMi FOR DWNM-.48; OBS1U5-.94 /A 10S65 FRO OM GO OJaFs ll.E4 1984 EWMS= FROM IM M SR", T. 47-4 ,8 O OM W TMAW BY AXMYD n=Me tWE 10= 16m a R˘ql g ~~~~~. . *111g La.?FF Ii!i SE. Ng s WE§! s 5! i |!S; W SE i . ie~5 9" S f "^ w~~~~~. iX WX @I -E -X X F.Sg E I - 93- Table 12: HOUSING 1976 1977 197 1979 106 1961 106 1968 1084 1065 19066 1907 11016 GuOTPUtf USS OdILLION /1 EX-PAT OJSIN 6 6 6 6 6 6 lo 12 14 16 is 22 go GOSM WM TSH (MILLIONS) U0.26 49.74 46.25 49.82 49.2 68.24 02.6 1813.63 214.06 270.52t 6m.8 1418.72 42M.8 VALE ADDED TSN (MILLOB /2 CJ.EN PRICES 45.8 44.6 41.6 44.4 44.8 69.6 68.5 120.8 192.7 251.6 529.7 127.8 2284.1 1976 PRICES /8 45.8 80.8 65.61 3.5 86.7 84.5 48.8 66.6 07.6 116.1 130.2 482.5 7M.5 EMMANE RATE 01.80 6.29 7.71 6.2 96.20 6.26 0.26 11.14 15.29 17.47 82.70 64.26 39.29 REAL EST. DEFt. 1.66 1.34 1.12 1.15 1.4 1.78 1.78 1.82 1.97 2.17 2.66 2.94 8.U OUAT. ACCT.) NOE8a /1 GAAOIA ESTIMATEA THAT IKMIOCZAL PAYkDf W O RR4 FOR 106-66 ISt 26.0, 82.0, AND 85.9 MILLION UNS. ASSUMIN THAT rll NATIONAL ACCUJIT CAPLUS 8iS IN VAUN WF R4AL ISASIN U. ARE LEFT *MWITh 16-416 MILUIO, 1907-622 MIU.I, 1M06-25 MILLION. LISUI A=OU 62 MILON LESS EACH VW EORWO IM TO A MINDAN OW 8 MILLION. /2VAUZ ADDED TO M86 OUITPUW RATIO AWAi TO BE .0 CL OWN PRICES DEFLTED BY REAL hSTAt3/OM OCOPIU DULLING FROM NATMONAL ACCOUWES Table 18: TOURISM 1976 1977 1978 1979 1980 1981 1982 1983 1984 1986 1986 1987 1988 No Visitors 71210 72060 72000 720 78100 880 7000 5200 64000 68100 87872 94000 96900 Length of Stay 7.6 7.5 7.5 7.6 7.6 6.6 6.6 7.7 6.2 6.6 6.3 6 6 Visitor Days 584076 54000 54000 64106 t 0 54800 53950 45000 40400 896800 409509 427593. 470000 480000 Avg. Expenditure 39 48 50 56 62 62 61 60 69 69 70 77 82 Per Day /1 Total Tourist 21.0 28.4 26.9 30.6 84.8 33.6 27.9 23.9 28.8 24.3 30.0 36.2 39.4 Expenditure (USSm) Oth Bugwacha Est 2 37 41 47 58 6o 69 49 42 41 43 63 63 69 (U) Mission Est (US$m) 28.9 a2.2 87.0 42.0 47.8 46.2 88.4 82.9 82.0 83.4 41.3 49.8 C4.2 (Avg. of Both Est) Adj. for NA Est. /8 21.7 24.1 27.7 31.5 3E.4 34.6 28.8 24.7 24.0 26.1 31.0 37.4 40.6 Gross Output Mission Et. (Tsh) 182 200 214 269 291 287 267 275 367 488 1012 2402 4036 V.A. (Mil. Tab) /4 1M9 120 128 156 174 172 160 166 220 263 607 1441 2421 V.A. (iI. 1978 Tsh 109 98 92 101 106 86 68 5a 56 49 92 187 202 Memo: Trade/hotels/rest. 100 122 139 158 166 201 255 812 896 633 58 770 1197 WV USS (1980=100) 68.6 69.9 80.4 91.2 100.0 100.5 99.1 98.5 94.8 98.8 118.3 124.5 132.5 MUV US$ (1976=100) 1.00 1.10 1.26 1.48 1.67 1.68 1.56 1.62 1.49 1.61 1.78 1.98 2.08 Exchange Rate 8.38 8.29 7.71 8.22 8.20 8.28 9.28 11.14 16.29 17.47 32.70 64.26 99.29 Notes: /1 Bagachw estisates of 182 in 1988. Previous years adjusted by change in WV index (1976 a 100). /2 Estimat of 869 millon in 1988, or 1.76 times the estimat derived from the figures of tourists, no. of days, et /3 National accounts for about 20 percent; 76 percent loft unrecorded. /4 Value added/gross output, estimated to be .6. /6 Current price defiated by wholesle, retail trade, hotels and reaurants deflator. - 95 - Table 14: GM PM TIC D1530AES 0ML *MmLI 5mDm 9 21 26 35 1835S 2447 26 24 2M9 16 4I3 79 718 lEWIf. 21 23 288 213 427U 1513 U I 43 86 I6 A .L C4 AJMW 277 8U5 88 45 498 574 aW 760 12 1387 13I 23 NME4M.. 458 6 740 98 U25 1448 1 MO M 113 E 425 OnnER6 2617 12DM 175 20 4 83 26 SW 28 10 2 48 Jll 4 WO 120 88 212 IV 285 J 32 13 2 6 71 S 63 984 M6, VA1, OlHM 1477 17S5 28 284 2S 2872 S Il12I 22 S W S 723 gJv,eIT 36! 8419 4230 US 4M SW G424 4n8 7221914 1Mi 8 4854 1RIM 723 S 1 8 114S U 21 2 12 82I WS 498 7623M 2M997 82DM Oam M 241 MM 2435 822 24 4146 S4 274 84 64 2218 2846 CDlI MIi OMs 73OM OM OM IM M I988M87229 40M OM 04E IN S1M06 438 On 784 88 Y 19 1429 2I 124 2US 2tl 4O22 Ml TOTAL SW17104 O 93 95G MM 2 86 8258 2 382M8408 OM 2D 27 1D 29988 29811883183 1884 185888 1 BMDa 9 22 9E5 U 19 212 22315946 935 1; 28 22 1131I17L 26 218 284 144 212 238 13 61 218 289 98 3 25 ARML OiN AMJ4T 2VT 24El 2S 978 25s a2 84n 849 s 1o 857 87 NMEMRMAL 4fl483 63 3 79 70 9638 447 8975 O8 5lR one 356 267 1 2 1481 3 22 15 743 4 7991N 796 221 LN) 18 18 2 I8 23 23 148 89 1 423 98 D8 D6, 1IR, 0M 1477 1427 17 22 14 24 I8 an1 741 6an 66 m8 etm4T 2 M E 8 23 725SW 265 US 56749aS 4814 4Z4S 1R1wo 72 574 918 2S99 78 M95 79 677 227 89 834 82 3M 2S amE R 2U1 2S 1816 214 232 2523913 WM 2 16285 1 I an MOUS M7O 1 6OM25 SW M 42 6 7511 36W O9M m Iz Srooe 43t 784 Mg 426 4 12 611 6i 442 414 898 42 420 OD lOiN SW 644 8 OM 62 6M GM M8 4082 76M 5 7W 71 72 DCT lU5 UNS 372878I no 1 88 299618 I= iS IS I95 SW Rm nD1T 20A.9 9.9 2.2 19.9 21.1 2 81.83 .7 .2 36.6 88.4 4.9 478.4 W.2 1DBITMAL 203. 923.1 MG .248.8281.4 23.9 Ei 28.42 2.1 W.8 48.2 496.8 83.9 5LW.L OM AMD 20.9120.9 14.7 14.56.6 219.2 23.4 23.8 274.1 89.8 87.6 4.6 498.8 lN04lDE1 4 3.9 12.0 218.2 109.6 2.9 281.0 2El0 37.9 3.0 49.6 48L 496.2 88.9 one 356 2389208.2380 .2 239.8 23.9 IS 2i 22B.0 91.9 83.1 48S.8 495.1 il2.1 W0I 20 123.6 238 140.8 Il.2 23.8 2m.7 22.0 2.1 8.A 489.9 642.9 64.7 NMD, Ml, OHRS 21.0228.0 2.1 9.9 3.9261. Eli 0 2.9 0.0 9.1 48.1 474.8 649.6 E4T 23 .0 1.9 124.2 23.6 24.9 2.7 2.0 IN9. 177J 189.6 S4.1 749.8 2W7.8 1MGWT 8.6 21.912.2 I3.52 4.9 2.891.90 9.9 '77A 91.7 M.$ 42.4 29.1 0W 23.0 921.9 12.2 23.6 14.9 .92.919.9 UFA 189.6 496.6 17.61U47.4 OfI 23 ll22.5 323.4 234.8 13.7 148.7 178.9i1m.8e2.2 283.7 40.8 t8.2 M.2 aw D IN$06 2il38.0 117.8148.9 2.6 26.2 231.6 275.9 811.2 872.2 16.167.7 8n.2 IM.4 (M MTNL 2.9 12.6 21.1 218.6 23.7 23.1 '18.4 2.925A. 248.4 43.8 679.9 25.6 M1Ias 26.8 -21.6 14.8 1. MS.6 -1.8 -8.7 17.2 48.2 20.1 30.9 -6.8 hIETTN. 49.8 8.1 7.6 -6.8 76.8-18.6-7.0 76.4 22.6-2.0 22.4 28.8 lt.L SWN A.4 N 2.6 2.5 2.7 2.7 2.6 2.6 2.6 2.7 2.6 2.6 2.6 2.6 N3k84E11L 48.6 -.1 22.9 8. 14.1-14.6-4.8 128.6-2. 38.8 3.8 -8.4 m IE -4.8 -27.6 21.0 9.6 -M.6 29.6 -8.2 14.1 4.6 27.6 -2.1 59.9 LAO -2.9 -8.7 46.1 -18.4 02.2 -6.8-34.7 46.8 49.6 2.6 -8. 1.6 AOMS, Vl, 0HM 48.4 -25.4 8.7 1.7? 9.6 29.6 -4.9 30.4-14.8 88.7 -47.9 63.8 E4m 128.1 12.8 1.7 -4.6 29.9 4.6 45.9 64.8 84.6 4.5 -4.6 -12.4 I GMU14.7 .91128.2-9.6 26.9-9.6- -7.8 M.2 84.6 4.5 4.6 -12.4 am1 19.6 25.9 -.2 28.9 6.4 2.1 45.5 -43.0 S4.6 -.6 8.6 -22.6 an 10.9 4.2 12.1 -22.8 8.4 4.2 -1.2 46.7 22.6 4.6 -1.1 8.4 ONCE IN SIO6 67.6- 1.1 -17.2 14.8 16.2 -4.1 25.6 -.1 6.8 6.6 8.6 4.1 X WN. 25.8 4.6 9.7 -.6 4.8 8.1 -2.4 80.7 19.6 4.1 4.6 -4.8 r0ims Fgt tSi tW ox UN Di114 (18 38 85.0 70.7 75.6 91.4 I3.N 291.5 132.2 21.1 289.8 24.1 451.8 975.7 15.4 IAN DIN a #9. 23.923.? 128.8 14 .7 158.9 11B.1 272. 2.7 272.0 814.0 S6.1 25.12488.6 - 96 - Pap 2 of 8 9197711318919 19828819B9 191828226 288 SS7 am@ ODM W A & O (iO64m1).0 W6 9.56.9 98.4 01.1 82.6 10.918.2118.729S.012.8 211.6 25.6 UN RIM ENOW9I 13.0 2W7.7 147.6 19#.626.6 198.7 219.7 4.38 .7 478.0 M.6 3.1 787.9 .w DwT 116 a W4. _ GL amm lRAWOW 49.0 79.6203.0 24.981.65 11.9 92.6 66.8 67.7 3.4 223.7 26.8 81.8 o 3142.6 13.7 m.5 2i2.. 2.2 1.9 06.9 I. 1 173.6 28.7 .8 204.2 238.9 + VmD we 191.61 8.7 498.5 8W.742.7 47.6M.726.4 29.5 389.1 W.8 40.0 22.2 ll4ITW M 61.8 8.6 257.5 2J.1 384.4 1n.1 13.5 W.$ 23.7 215.5 14.6 2. 2.6 .6 WAM ) 178 37.1 416. S.6 1.4.75 .4 2A7. 2.8 3.45 7.A 17.6 178.6 SC.WE MTlK 3 .3 6 .n 7.s71 o.95 9.n111.1425.317.472 .7 64.3 98.23 Omm . ---N- MM - aM M M M= nm i?1 lsOlII 193__9o 920216 iW"A MC mjai. ~Mm M M Ms M aM aM aM Ms Mm MM onwmum mmmn.isra *W a^ 2"s M Jwe" " t 2 WM A 6 1M.MIW US 84 47 W 478 412 U 7 123 121S 142 W en. m wm Mw M M M M M eM M muam NW am45 : W-C R0 WWI CW?L OM MLL 030 197AL 1MW lI1LM 10 281 4 3 193 6M1 MS 4 14M 19AL W Q.Ll139 2JtJ UIS J I34UU 1985 t12X1176896 526115264 1989 OSIXC 116 WRISI Ca123 0S .L 236 130. 1fa L W 143M 9 141O 143 9W 9 a7 W M 1M 4 Wirr, W6 M M M7 W M MM74467754m 9M 6 4M O>u w 49j 130 am ma 61r4 pw- 72 6ii mu st 6tl a 6 AFOW L T9 SWt J1 M 2M 274 _ Oa M 2>t UNM M W rnc s 83 77 M W 4M 7 6M 3 M s M M noM M7 W6 0MW ~ ~ ~ ~ ~ I IL .119 am Rm # 9 61mm 6 i729 mm 243 29J11mm u 42 UU MS 1TM 225 2463 193 674 584 96N m 66 2 149 282 183 PfL" 0 N 677322 M46 M482 M 2 4 6 13M 7 26N1 U14 7 8 Mm,-,L m8 1 74M 068 2W 3 244 278 13 21M M M 425 JM BMW 1 27 UPM 3M7 3 2J 33 M3M429 8213 91 MU LI2D JSw" | 35138834589S4214 ""X 481 am, '11w, 01 im231 137 377 2E U 2382 a 1 2e 44 M 857 76mE an M6 8819 M S O 6770 69642 98184 8_ 4OM 8M 1RA aM we - - - - - - - 11795 J2484 28m 982 U!DM M UN M n1oM51617 111K 8UN 3 _u M1716 9no 8M 8 em 6"2 M$ | M M11 1 Mt VIW M M W M M44 Mw 60 R M sioo W M Mu M M M4 M aM W64 MP 248 Ws 04SŁ VSIDM˘ 4 14 li4 MZ1 180 1S1 WJy MOI M M",1AML 0 N us142 78u 2 M12 2-2 22 2 M141 1911 8M 1 W 19B fMMM PVlRi M 17 US O U9 USR 9t M1 UP 1 O 1m 21 Mm M M19 1 148 M 94 41 4 M M 147 LIA4L AWM 277 M 7 2M 13 $9 J 8QO 1 87 M7 M M3 M M NhlANIX 46 m 6119 i. is m9 8 J 47 18? 681 891 510 UNM ~~215; 3 25 35F 214 156 3 s 19 7 in 34 38 RKDS 1B? aiim aml a1i7a 1448 am := i.s 7s 816 u 9 w3 7am1l Bl~~~~7 8311 41 41112 34#RJR 24m am 6467 96137 106 281 lRi~~~17 @14 2116 143212U8 13401W 198823 16 125 ABS Osuma ma01 an0 9588 t4 29 6744 2S9177 1788 W2 6 817 1m U! US 88673377513B 71S tE 98tO98462402966 4989 ;268983 S OIWE DS h;IIX 488 784 ust 421 488 61261160 442 414 89 43 416 0!lON. I U 771 16 789 771 721 39 4157821677 i -97- A D ~~~~1S0 1U9.8 W t.4 1J.8 161J INA1 2.7 234.9 2"77 M.6 40@.9 1R1.98 MOA FllkM 10 IU.2 U17.8 14.6 U7.7 WS. WA MA8 29.7 "2.8 M.4 40.82 W.16 RMAL IIIN AUlNW M 118.2 U18.2 18.1 M4A l i.7 202.9 ql44J 296.1 M.8 4U.8 Q10.82 Mt.62 Nufimm 11S1 2.9 10.1 W.4 30.8 1.9 :IJ 25t.9 21NI "30 4O7 411.18 ZIJ1.2 all 6 10 0A U l.1 WA108.8 . 1SD.9 W.9 =.I 2U.1 3SJ 488. 6.48 S.4 1Z 8040. .6108W1 . . 1899. .9 "2.066M.71 Ra^06 1ER, mmm 20 M2.9 U8.1 lt9.7 100.0 lKAl 2MA .0 27.90.8 J 488.1 474.81 ff9.6B E/Wu MO 3A.6 30. M.1 IRA MA 1W.1 19.6 2W.1 SK1 6.4 t74.1 14A rj#ww W UO11192.4 XE2.1 1J. 17 19.5 218.1 266A SOL .7 1.8 28.1 am1 S MOS. WA 7.4 112.8 140.2 142.9 12.2 V4.5 2i.2 27.1 GH.9 11A MA81 m R0 U114.7 124.6 140.3 USA0 W2.J 1112.1 51|.7 2M2.6 D.8 WA 8i6.44 9t.23 aw IN S1006 I N WA Z8.9 9Z.7 21e.1 MAl 276.9 811.1 W2.1 6|si .8 81.12 im.8 aDI IW7 10 314.9 12.1 USA. 1W.7 W0.9 1Ł8.8 216.8 Z2.S 8JA 642.8 8i8.9 D.251 MO BW-EWW M1# Me 4M MB 4M Ml M 42 MM= OMu17 744 r*w VFK Mw M7 U140 278 = OM 97S 4C 7 18 SUl O amH FAQ 10 2441 MO 40 51Z 8021 MA9 OM W4 SM O4 2W 6 WOM*DeWIIW O 62M 0R9 GM4 WM 9"48 SW O MM~4W= rNtewot ) 9 W 12= IM 84 M 9 =111M 90 190 114J 86 aml 2Ml)S 2U0 OM11 US Ml SW =2 M1i MW _MUM MO3t I85 CR:INL/u161D ~ f 9 " O 7S 74 74 42 Of 08 309 Ul W 08 1 zpFFKx61M9 OM 7MD OM2 8111 OMS 6M ?M| 8M l2 2W9 4019 OM oNaE]NS1106 )41 OM 784 83 US MZ 140 UO I=3 2J OMfi 4W2 < 1~~SW 4 7W4 M$B NO8 M "2 =81= 4=W O alR OM)6|1 729t O WE XIEW 1= MEi OMI W4 MM 6t0 MM1| SaE D SR00 Cle 489 OMa 76 ON 106 1Z 40 1 183 IM # IM 2@ SW OM 432 CDI OM OM OMD WK IU2 1L 72 IM I= U= m low 19sm am" mm 36 L/AD16W41)t~ 0 sa 83 M 83 77 soa so se 98 92 2 X18 aNFArI lDIFO BM 21iB OM1 $M O 2007 O 92X 42g SW6 4=Q 4K4 4ZSS IFWGW PF=AL) M 92 74 ON 8 7da ON 70 6t 2?M MM 285 amR MF>4 21M 2M8 I6 224 2 29@ 1%9 I= VS6 1a 4 1BI# 3M MkD13wWM 2795 S31 4912 4WE OM6 24= W 2704 SW M 236 T*r WN5W lJJ) M X 148 1X 14S2UBS2 10 1M 98 S67 MI 2liE 1NO -3l{ 1DM) IM O ZUS 2818 2W1 2744 2NOV M LB2 2L7 M an Waa GMS GM 670 am1 Gu GM an 4012 GM1 7221 1w7 o am OEM INO MM M" 48 734 Mg 42 4e W2 Ml M 442 44 W 423 410 *=I 1srEL pRaF 6E SW F4 OM10B OM OMf OM8 4W2 OMa 768 ?W9 78 7106 m Q&tsm D SW SW# 722 70MO 71W2 OM 4W 4" GM4 OM M Mg owm NS1 mmp 49 784 618 42 438 Mg Ml1 60 442 414 M 429 420 0DI 7grAL QOD161O M1 MS6 A 7M 7)14 MO1 1I21 GM GM OM MM M (Di pFFIaCL amm HME 16. 8 4. 9.7 -10J6 4.8 8.16. -N 3.7 5fi.6 -8.1 -0.A -8.8 G)l QO)BMl m l AE 10.0 9J 6.0 -9.2 4.8 -6.6 -7.6 8.7 8.4 8.9 8.1 -16.9 OER AMll wo m" D w 13 0 112 124 Ul 247 IS I&N 109 177 20 M@ 749 127 lRi :ICUI) 10 12 IN 4 W| 147 I4 101 M" 17 M"8 426 2 OM a P1A 13 2 Mg 4 IN I 47 W l l30 2 t9 m8 1 40 1471 1147 MtN FD E4JMNr leg S 0 141 161 1E8 1XB 15 2Za OM 00 llS9 1484 IPR8W ILED) m 122 is6 31 108 M7 IN M1 Ws M" M9 MO M Or1 (1DJ ) 10 107 117 US8 M4 144 1* VIt 282 274 678 1=S 149 Onaa AL IC Q 116 W 2M IB4 M5 19 UR9 12 284 40 a8 92 0*E IN SMO aFF U 37 119 4 SO6 M W Ml 872 626 M9 8n HE cot L mm p aL la t UT no 0 no 9 in in l 2a7 Z16 24 4ZI Ml MQ mIF QUM1) 10 14 IN 140 WA4 W0 1112 24 M1i M2 68 M2 tl4 011 NE MM SW) 1l+ lU 7 149 1a4n 2U 2 M SU1 872 M26 On M (Di WrAL QA S6 U 144 12 1" 9 216 W Z 886 649 M 1X8 NM: NMM B=WMO R= FIM 4 14 SEE C:\TIA\WCNS RR aF= *D NJ=W FR GL MAS PRIVATE SECTOR DIVELOPMET 1N TANZANIA: OBS-TACLES AND Tanzania Economic Report Background Paper #3 Private Sector Deveopment In Tanzania; Obstacles aid QQgorunitiefs 1. Introduton. This Annex evaluates empirically the potential of Tanzania's private sector and the obstacles it confronts. In particular, it assesses rigorously the view that the supply of indigenous entrepreneurs is severely limited. The focus on this section is on the small scale private sector, comprising ma uri as well as construction and agro-processing activities. 2. Bacfl nd. In spite of the often hostile or ambiguous attitude of the government towards the private sector through much of the 1970s and early 1980s, Tanzania today has a cadre of experienced entrepreneurs of Indian origin. Additionally, in the wake of economic crisis of the early 1980s, and the subsequent liberalization, myriads of micro-entrepreneurial activities have mushroomed along almost every street in Dar Es Salaam, and throughout Tanzania. In Dar es Salaam alone, the number of micro enterprises are estimated to be at least three times the level of the mid-1980s (Chapter 2). 3. Prima facie, the pervasiveness of micro-entrepreneurial activity suggests that there is no shortfall in Tanzania of individuals with the initiative and discipline to engage in business activities. However, many micro-enterprises were born of the need to survive in a harsh economic climate, and many micro-entrepreneurs may have no purpose other than to endure from one day to another. Under such circumstances, the potential for indigenous-driven private sector development would be more limited. Evaluation of the quality of entrepreneurship and the associated potential for sustained indigenous private sector development thus requires moving beyond aggregate estimates to a firm-level analysis of the various strategies of expansion and survival adopted by entrepreneurs, as a way of leaning what proportion of these entrepreneurs may be committed to progressively expanding their business ventures, and what are the obstacles they confront. Such detailed analysis was conducted during this report's preparation in 1989 through an entrepreneurship survey. Three subsectors of the Tanzanian economy were covered: i) furniture; ii) construction; and iii) horticulture, where more limited information was gathered. In these subsectors, entry was possible at the smallest scale, and where enterprises had the potential to expand incrementally to quite large size. 4. A Framework for Analysis. Rather than viewing entrepreneurship as an attribute of personality that is either present or absent in a single individual, we take the bundle of entrepreneurial functions that must be performed in a successful enterprise as its standing point. 1' This bundle of functions includes: (a) the initiative to identify and respond to opportunities for profit; (b) the discipline and skill to manage an enterprise so as to maximize revenues and minimize costs; I/ This functional appmach has been broadly accepted in reseamh on entrepreneurship in developing cowntries; sf, iuj- example, Leibenstein (1963, 1968), Kilby (1971), and Jones and Sakong (1980). 101 (c) the ability to gain access to finance; (d) the ability to gain access to other inputs - labor, raw materials, equipment - required for production; (e) the ability to market the goods and services produced by the enterprise; (f) the ability progressively to improve the technical capabilities of the enterprise; and (g) the ability to manage relations with the government bureaucracy. Viewed from this broader perspective, there exist three potential sets of obstacles to private sector development in Tanzania: 1. shortfalls in the supply of individuals with the attributes of entrepreneurs [(a) and (b)]; 2. weaknesses in the external economic environment within which enterprises operate and associated high costs of gaining access to resources and markets [(b) - (f)]; and 3. a regulatory environment which makes enormously burdensome relations with the government bureaucracy [(g)t. What type of interventions and policies will be required to support private sector development in Tanzania depends crucially on the relative importance of these three sets of obstacles. 5. The entrepreneurship survey was carried out in three potentially efficient subsectors of the Tanzanian economy in order to evaluate these obstacles. As presented below, the results indicate that there is not a serious shortfall in the supply of indigenous entrepreneurs. Then, we summarize the findings of the survey concerning the magnitude of obstacles confronting enterprises in the regulatory and external economic environment. It is suggested that, although the regulatory environment is exceedingly restrictive and forces entrepreneurs to devote a significant fraction of their time to managing their relations with public officials, regulatory obstacles do not now appear to be the dominant binding constra;t on indigenous businesses. Rather, weaknesses in the external economic environment - especially in the financial sector with associated difficulties for enterprises in gaining access to finance - appear to represent the single greatest obstacle to dynamism on the part of indigenous small and medium enterprises. Entrepgreneurshil In the Furniture Industrv 6. Interviews were conducted with (a quasi-random sample of) 24 furniture firms based in Dar Es Salaam, of which three were medium-to-large sized enterprises whose owners were of Indian origin and the remainder were indigenously-owned small and medium enterprises. Grouping these enterprises according to their year of entry, number of employees, and the year in which their number of employees reached a peak, five distinct segments are evident: (a) Three medium-to-large enterprise, all of which were established more than fifteen years ago, employing an average of 150 people each; 102 (b) Ihree stagnant medium-sized enterprises employing an average of 35 people each, were established prior to 1979 and have failed to grow over the past two (or more) years. The objective of two of the firms was simply to survive at more or less the present level of operations. (c) Four exnanding medium-iIzed enterprises employing an average of about 50 each and which have continued to expand into 1988/89; all four enterprises were committed to continued expansion. The oldest enterprise in this group was established in 1979, and the others in the 1980s. Ail but one employed fewer than five people at entry, and subsequently expanded incrementally to reach their present sizes. The exception was established in 1987 as a co-operative of 25 carpenters. (d) Nine small enterprises, four of which entered the furniture industry in 1988/89, averaging 6 employees with the remaining five entering earlier and expanding in 1988/89. These enterprises, recent entry the industry appears to have been facilitated by the renewed availability of woodworking tools in the Tanzanian marketplare subsequent to liberalization. All nine enterprises scught to expand further: six reported in interviews an ambition to more than double the size of their enterprise in the future, while the remaining three ainied for a 50 percent expansion. (e) Five stagnant small enterprises with an average of 5 employees each. One enterprise has been declining since 1967, three started small and appear satisfied to remain so, and one has had more ambitious, yet unrealized goals. Systematic patterns are evident among the five segments according to: i) the markets they serve; ii) their processes of production; iii) their sources of finance, and, iv) the levels of education of the entrepreneurs. 7. Markets. The submarkets served by Tanzania's furniture enterprises range from the completion of large-volume orders for goverment agencies and the export of processed wood products at the top end of the market, to roadside sales of household furniture to consumers at the bottom-end. Between these extremes firms produced furniture for sale in retail showrooms, supplied high-quality custom designed furniture for higher-income consumers, and produced household furniture on tender to government agencies including parastatals. While the majority of medium and larger enterprises make the largest proportion of their sales to government agencies, only a minority of small enterprises have won orders through government tenders (see Table 1 below). Overall, the small enterprises made the majority of their sales directly from their workshops, although they are not excluded from other niches of the furniture market including sales to government. 8. Paralleling the movement of smaller enterprises into the mid-range of the market, the three larger enterprises in the sample are in the process of moving from the mid-range to the top end of the market. One enterprise has recently abandoned furniture manufacture to devote itself to the production for export of parquet flooring and other components of wood; a second enterprise has reduced the share of wood furniture in total sales from 50 percent to 10 percent, shifting instead into the production of metal office furniture for the domestic market and the export of parquet flooring and other wood components; the third enterprise, although adjusting 103 more slowly, also has made tentative moves to explore export markets. An important implication of these moves up-market by larger enterprises is that, rather than being "crowded out", new market opportunities are likely continually to emerge for small enterprises seeking to expand. 9. Pduction pMC . Table I highlights a significant differences in the production processes of small firms and their larger counterparts: while all of the larger enterprises operate with electrical equipment, only four of the small ventures. Furniture firms with only hand tools are not necessarily inefficient, however, equipment sharing arrangements were not uncommon, with enterprises that owned electrical woodwotking equipment providing access to that equipment for a fee to enterprises that were not so endowed. The combination of hand-tool production, access for a fee to electrical equipment, and the prevalence of payment by piece work (one which reduces substantially the task of monitoring worker effort) appears to be a highly efficient way of orgniizing production to meet the demands of a roadside consumer market for household furniture. Table : Characteritcs ot Five Segments of Tanani's Funiture Indust (1989) Number of Entbrprim with Tenders from Tenders 505 or Etlectuica Loans from Average Parastatas more of Sales Equipment Formal Number of Fmancial Years of Institution' Education of Entrepreneur Med/Large 2V3 2/3 3/3 3/3 Enterprise Stagnant 2/3 2/3 3/3 3/3 10.5 Medium Expanding 314 314 4/4 2/4 8.7 Medium Small with 4.'9 1/9 2/9 3/9 9.4 Potential Stagnant 2/5 15 2/5 1/5 9.0 small _ _ _ _ _ _ _ Total 13/245 9/24 14/24 1224 9.4 * includes financial suppoat from the Small Industries Development Organizion (SIDO). 10. What the above description implies, however, is that significant new technical and managerial demands are associated with the move from a small workshop (with craftsmen producing using hand tools for payment by the piece), to a medium enterprise (which produces on a relatively large volume with electrical equipment). Indeed, as will be explored further below, this transition appears to have been difficult for a number of medium enterprises in the sample. 104 11. Einan . A third systematic difference among the enterprise segments is in their sources of finance. As Table 1 reveals, all the firms in the medium-to-large and stagnant-medium segments received loans from formal financial markets; further, all three firms in the latter group were afforded access to finance relatively early in the life of their enterprise. By contrast, formal sector loans have been forthcoming for only one-third of firms in the remaining tliree segments; in recent years, even enterprises with strong track records of expansion have encountered difficulty in borrowing from the commercial banking sector. As will be elaborated further, difficulties in gaining access to finance emerged in the course of interviews as the greatest obstacle confronting Tanzanian furniture and other enterprises that seek to expand. 12. Entrepreneurs' educatlon. The most striking feature on the educational background of entrepreneurs are the similarities (Table 1). Only in four indigenous enterprises - distributed across the enterprise segments - did entrepreneurs have more than a high school education, although many had secondary education - also distributed across segments. Further, about 60 percent of the indigenous entrepreneurs sampled have been engaged their entire working lives in private sector activities. Entregreneurshin in the Tanzanian Coostruction lndustry 13. Tanzania's National Board of Architects classifies building contractors by class on thl3 basis of equipment owned and the number and skills of each contractor's permanent employees; the best endowed contractors are classified as Class I and the least endowed as Class VII 2'. In practice, all the Class I contractors are international construction and engineering enterprises, most of the Class I! and III contractors enterprises with proprietors of Indian origin, Class IV a mix of Indian and indigenous contractors, and Classes V-VII entirely indigenous operations (Table 2). i4. As indicated by the data, with 1,076 registered contracting companies as of the end of 1989 (and 862 of them in Class V or below) there is no shortfall in indigenous entrepreveurs active in Tanzania's construction industry. Second, there has been a steady flow of new entranits into the industry - somewhat over 100 each year - since the late 1970s. Third, there has been some shift in the composition of new entrants, with a progressive decline in Class V and subsequently Class VI participants, and a corresponding sustained increase in the number of Class VII entrants. Likely reasons for this change include a shift in demand towards household construction and an associated increase in the supply of entrepreneurs willing to start-up activities with low levels of initial invcsment. Fourth, not all enterprises active in the construction industry are included in the table: along with registered operations, there exist also 'hundreds' of unregistered enterprises ' that operate as subcontractors to their registered counterparts. 2 Tanzania's National Boud of Architoect, Quantity Surveyors and Building Contmetors is responsible for the lasafAtion of individual oeterpris. Tho ousftions have a regulatory function: thefe is a eoiling on tho value of a construetion contract for which ach ola is pormnitted to bid. Th lestimats of two interviewees closely involved in the contruetion industry. Constmuetion operators are able to evade Tanzania's cthrwise comprehensive regulatory not by virtue of their ability to operato without any fixed business promimes. 105 TAjgo 2: REGISTRATION OF NEW BUILDING CONTRACTORS 1977-89 NEW REGISTIM I 1977.80 1981-83 1984-86 1987-89 TOWl ____989 IV Clus I 21 a 6 8 41 ClIAU a 8 4 11 1 20 CIanl II 24 22 15 3 59 CIA" IV 44 29 44 9 94 CIaaaV 100 48 4 22 137 CIan VI 72 72 70 25 152 ClsA VIII 87 157 214 239 573 Totd 356 L 327 40 307 1,076 Source: NatlonaW 13rd of Archltecta, Quanfty Survors and BuHdihg Contracoa. Notes: ' Note that thee totals do not ojual the sum of now regatana, 1977-1989; they Include contractors ta rostered prior to 1977, nd exclude contacm that, although enteing between 1977-1989, have subaequendy wlthdrawn from tho contractIng Industy. 15. To complement the aggregate data, interviews were conducted with 14 construction contractors, as a way of leaning about their strategies of survival and expansion, and about the obstacles they confront. Y 16. As with interviewees from the frniture industry, there was a striking similarity in the educational backgrounds of the construction contractors, although here the: all entrepreneurs had completed secondary education; and most of then had a university qualification, mostly in engineering. Again unlike furniture, most of the entrepreneurs had worked for some years in the government or parastatal sector prior to starting up their own operation. 17. Turning to strategies, as building contractors do the world over, they all organized themselves to respond flexibly to a wide diversity of demands in a vey cyclical activity: they maintain a relatively small number of permanent employees, drawing instead a short-term labor as needed; they also keep in-house only a fraction of the equipment they used, renting (in what apparently is a quite well-developed leasing market) the remainder as needed. Further most of 9/ Six of the founten interviewees were Cla 1-rV contrctor, three cla V-VI, and fivo Clas VII or unregsterd. AUl 6 of the Clms II-IV contmeton - along with 2 otn in the sample -- began their opeations prior to 1980, 5 began between 1980 and 1985, ond only I submquently. There contmetors worn idntified non- randonmy: while an effort was made to cover a brad rno of clam, to nsure that infomation was provided feoly nmt interviewees were contated via itoductons from acquairntes of the interviewer. 106 them used subcontractors (which accounted for an est!mated 30 percent of labor costs of an average job) so complete specialized tasks which the primary contractor was not organized to undertake directly. The prevalence of short-term labor, of equipment leasing, and of subcontracting signal that building contractors are embedded in a quite well-developed network cf market relations. 18. Given the centrality of flexibility and market contracting, it is not as straightforward for construction as for furniture to distinguish among enterprises according to their commitments to, and strategies for, expansion. Even so, three distinct strategic orientations were evident in the sample. At one extreme were four highly conservative enterprises: they contracted only for private clients, had not increased their number of permanent employees or their class of registration since entry, and expressed no interest in expanding the scale of their operations; there were no common patterns among these four enterprises in their class of registration. At the other extreme were three highly aggressive enterprises, all of which contracted extensively with government, and all of which were committed to expansion and were investing in skills in-house to bid for road construction and maintenance contracts; two of the three had upgradeu their class of registration, while the third had been Class m since entry; all three - and only these three of the entire sample - had borrowed from commercial banks. While none of the three enterprises in the third group, which had grown incrementally over the years, explicidy had growth as a long-term objective, they had progressively increased their number of permanent employees. Further, they maintained a relatively equal balance of public and private clients. 19. Overall, a majority of the contractors sampled appeared to have significantly enhanced their construction capabilities through participation in the industry, and to be on a trajectory of ongoing expansion. However, as will be examined below, they faced s'abstantial constraints to ongoing expansion, with the character of the constraints varying with a contractor's strategic orientation. Entrepreneurship In the Export of Horticultural Products 20. With the depreciation in Tanzania's real exchange rate, the liberalization of restrictions on import and the liberalization of agricultural marketing for non-traditional agricultural exports, new opportunities for export of nontraditional agricultural products have begun to emerge. Horticultural crops which have export potential include pineapples, chilies, okra beans and other tropical fruits. Exports of these crops from Tanzania are in their infancy: but they have grown rapidly from a small initial base since the ERP was adopted. Interest in horticultural export prospects appears to be widespread - 68 entrepreneurs responded to an effort to recruit members to help start up a Tanzanian Horticultural Producers and Exporters Association. 21. On the basis of surveying a dozen horticultural entrepreneurs (of which all but one were indigenous) ambitious export plans were identified: by 1992, ten of twelve interviewees planned for exports to exceed local sales, and six interviewees had the goal of earning more than $500,000 in export revenue, with three of these targeting revenues in excess of $1 million. Most of them had already made trial shipments to export markets, one had already made a substantial investment in a large-scale pineapple export facility, while 5 others planned to invest in irrigation facilities on their land in the relatively near future or were in various stages of preppring feasibility studies and apply-mg for financing from banking institutions. 107 22. Regarding the strategies and prospects for success of these infant horticultural ventures, two patterns emerge relatively clearly. First, as with construction, the horticultural entrepreneurs surveyed were a well-educated, experienced group: all but one of the respondents had finished high school, while six had completed also a university degree. Of the eleven who provided information, four had been employed in the public sector prior to embarking on their horticultural venture, and seven in the private sector; interestingly, five of the latter group continued to work on other private ventures alongside their horticultural efforts. 23. The second pattern concerns the souring of horticultural products. In principle, horticultural exporters can either sell produce from their own farms or they can operate as intermediaries and thereby provide a vehicle for small farmers to take advantage of export opportunities. In practice, the survey respondents give disproportionate emphasis to own production an apparent response to the need of conforming to requisite size and quality standards for exports, and the weakness of appropriate pub;ic extension support for small farmers. F The consequent likely exclusion of smaliholders from participation in horticultural exports represents a foregone opportunity for moving along a relatively egalitarian path of private sector development. Entrepreneurial Supply: Some Implications 24. The available aggregate data and the results of entrepreneurial surveys in the furniture, construction and horticulture sectors all point towards a similar conclusion, namely that there exists in Tanzania an abundant supply of indigenous entrepreneurs with the initiative and commitment to sustain a development strategy which gives a central role to efficient private enterprises. These indigenous entrepreneurs appear to be a highly diverse group when viewed from the perspective of education and prior employment experience. Further, a substantial fraction of the entrepreneurs of all backgrounds appear committed to increasing substantially the size of their enterprises, and - most clearly in the furniture and horticulture sectors - the opportunities for growth are abundant. Whether this potential can be realized depends importantly on the character of the environment for entrepreneurship in Tanzania. Obsticles to Entrepreneurship in Tanzania 25. Over the past twenty or so years the environment for private enterprise has been less-than- hospitable in Tanzania, although the situation has improved since 1986. If the supply potential of the private sector is to be tapped effectively, there is thus an urgent need to identify and ease those particular obstacles to entry and expansion that most constrain private firms at present. 26. The entrepreneurship surveys for the furniture and construction industries explored the relative magnitude of obstacles faced at the time of entry and inhibiting their efforts to expand. An initial set of interviews yielded a list of twelve potential obstacles to entry and expansion for furniture firms, and a related list of ten obstacles for construction enterprises. The obstacles included difficulties associated with government regulatory and licensing policies (including difficulties faced by furniture firms in gaining access to land zoned as an industrial site, and A/ Insofar as they cannot control to whom small private producers will sell their output, private tmders have Littl incentive to invest themselves in the nequisite support evices. 108 difficulties for construction firms associated with the public tendering process), difficulties in identifying target markets, difficulties in gaining access to requisite inputs (finance, competent workers, equipment, intermediate inputs and technological know-how), and difficulties associated with the high cost of inputs. 27. For each sector, entrepreneurs were presented with the relevant list and asked to rank each obstacle on a scale of 1 to 5 according to its degree of severity. Table 3 summarizes the results obtained after normalizing the results along a scale of zero (east severe) to one (most severe). A As the table reveals, for fms In both sectors the most severe obstacle was a lack of access to flnan , while the least severe obstacles were shortfalls in demand, a scarcity of competent workers, and a lack of technical skills. In between were difficulties asociatd with government regulatory. tendering ad licensing policies, difficuties associated with the high gost of Innuts, and, for construction contractors In particular, difficulties associated with access to equipment and materials. The reasons for these patterns will now be explored in some detail. f/ Normizatin is necesa minc ther is no bads for comparing the aououm of entroepnurs as to the abso magnitudes of veority of the vaious obsales. For example, one entrepronur might choow to rank th entire aet of obsale within the I to 3 ag, whil another - with identical poeeptions of the relative sverity of the obtls - might scale his rnddng in the 3 to 5 rage. Wbile the smilarities between the two entrpnu in tho rlative ank of obstacle. reprseos impoAnt information for policy, nothing of goneral usefulneu for policy can bo leaned from evidonce of diffrnc among enteprenurs in their subjective paroeptn of tho absolut magniude of challenge posed by the obstacle.. Calibrtion of cch frim's asusewnt of tho severity of tho obales on a clo of 0 to I yields a common cardinal scale of the rolative svority of various obstcles which contolb for differences among teprenoun in their subjective perceptions of the absolute lovel of diffioulty posed by the obstacol they confront. 109 IabJl3 ASSESSMENTS BY A SAMPLE OF FURNITRE AND CONSUCTION ENTERIRISES OF TIME RELATIVE MAGNITUDE OF 14 OBSTACLES TO ENTRY AND EXPANSION l NORMALIZED SCORES l____________ ____ FURNITURE CONSTRUCTION OBSTACLES Obstacles to Obstacles to Entry Obstacles to Obstacles to Entry Expansion, 1989 Expansion, 1989 Lack of acce to 0.97 (19) 0.90(14) 0.80(14) 0.81 (13) Finance Cost of Finance 0.72 (10) 0.54 (9) 0.75 (4) Lack of Acces to - 0.70(13) 0.42(11) Tenders Government 0.67 (17) 0.48 (14) 0.63 (14) 0.44 (13) Bureaucracy (Excluding Tender Procedures and Industrial Sites) Lack of Acces to 0.64 (12) 0.84 (11) Industrial Sites Price of Wood 0.63 (14) 0.39 (9) - Cost of Equipment 0.58 (13) 0.58 (9) 0.70 (5) - Lack of Access to - 0.51 (13) 0.86 (10) Building Materials Lack of Access to 0.38 (11) 0.49 (11) - Spare Parts Lack of 0.28 (19) 0.24 (14) 0.31 (13) 0.29 (13) Clients/Limited Market Size Lackt of Access to 0.26 (14) 0.45 (11) 0.71 14) 0.94 (12) Equipment Lack of Access to 0.16 (19) 0.16 (14) - Wood Scarcity of 0.14 (19) 0.07 (13) 0.18 (14) 0.04 (13) Competent Workers Lack of Technical 0.11 (19) 0.07 (14) 0.08 (14) 0.19 (13) S r __ __R.__d _ _ _b___ _s i. _ ._ _ Number of Respond.LIU in brackets. 110 Access to Finance by Private Enterprises In Tanzania 28. A lack of access to finance was identified as the most severe constraint on expansion by eighteen of nineteen respondent furniture firms and eleven of fourteen construction contractors, accounting for the scores of 0.97 and 0.80 accorded the constraint in Table 3. This striking result signals the presence of some critical bottlenecks in Tanzania's financial system. 29. It is helpful to prefice analysis of these bottlenecks by summarizing briefly what is known internationally as to the financing of small and medium enterprises. The evidence is persuasive 2' that the incentives are weak for formal financial institutions to target funding towards SMEs: loan sizes are small, so the costs of processing tend to be high relative to loan amounts; the track record and reputation of a SME borrower is likely to be limited (as is the system of financial accounting), adding further to the costs of loan processing; the probability of failure is high even of well-conceived new ventures. Unsurprisingly in the light of these incentive patterns, SMEs worldwide tend to finance the start-up of their operations with savings from the enterprise proprietor, support from family members, short-term credit provided by suppliers, advances offered by buyers, or (somewhat less frequently) loans from informal, community-based financial intermediaries who are able to draw on locai knowledge as to the reliability of a putative borrower, and on the commnunity sanction of reputai.sn to ensure loan repayment. In general, only once an SME has demonstrated some record of success is it able to tap into the resources of the formal financial system. 30. The results of the survey indicate that the mechanisms of financing of Tanzanian SMEs exhibit some similarities to - but some important differences from - the stylizeu general pattern. The strongest similarity is in the importance of own savings and of related sources: all 34 indigenous SMEs in furniture and construction for which data were available drew on the savings of the proprietor, on retained earnings, on support from family members or from partners to finance their ventures. The most striking differences are in the role of informal financial markets and of trade credits. Only one of the 34 indigenous firms surveyed raised any finance at all from informal market sources; by contrast, non-indigenous entrepreneurs active in Tanzania apparently have access to well-developed informal mechanisms of financial intermediation. Further, not one of the surveyed furniture enterprises have access to suppliers credits or buyers advances. Nor can any contractors procure construction materials on credit from suppliers. 31. The virtual breakdown of the checking system in Tanzania in the face of both pervasive banking inefficiencies (on which more below) and a legal system which affords only protracted and uncertain recourse accounts in part for the absence of trade credits; legally-binding post-dated checks are commonly used as credit and payment mechanisms in countries where firms make extensive use of trade credit. As for the absence of other informal lending mechanisms for indigenous enterprises, here the Tanzanian pattern appears illustrative of a lacuna that is evident throughout East Africa. a/ I.M.D Little, Dipak Mazumdar and John M. Pagp Jr., Small Manufacturinn Entemrises (Oxford University Press for the World Bank, 1987) pp. 279-287 provide a useful smmary of available evidence on SME financing. III 32. In the absence of an informal financial market, the role of the formal financial system is crucial in channelling finance to SMEs. In Tanzania, a number of financial institutions (all state- owned, as are all financial institutions in the country) have the responsibility in principle for providing financial support to SMEs: a small enterprise lending unit has been operating in the National Bank of Commerce since 1982; since 1974 the Small Industries Development Organization (SIDO) has provided financing for the purchase of new equipment; in principle the Co-operative and Rural Development Bank also lends to SMEs. In practice, however, these institutions have not lent resources to a large number of smaller enterprises. Y In fact, only about one third of the indigenous firms interviewed obtained finance from formal financial institutions, and they are largely of medium size. Less than 10 percent of the sample had ever received support from SIDO (and when received, it was prior to 1982). 33. The evidence suggests the difficult relationship between private indigenous firms and formal financial institutions: only one of the 33 firms in furniture and construction had an ongoing loan which was characterized by a "satisfactory' relationship, while 10 had a "difficult" relationship. The other 22 had no loans. One problem lies in the exceedingly burdensome conditions that SMEs must meet if they are to qualify for loans. Loan applicants are required to provide fixed property as collateral; since the banks refuse to recognize leasehold land title (the only form of land title permitted in Tanzania) as collateral, only individuals with the resources to erect physical structures on their land can borrow, and only to the value of these physical structures. Further, along with collateral, applicants are required to provide detailed feasibility studies of their projected investments (even though commercial bank loans take the form of overdraft facilities, typically used for working capital), as well as detailed financial statements. Even with collateral, feasibility studies, and financial statements in hand, firms reported that the processing of loan applications takes an average of six months, and may still require side payments as a precondition for processing. 34. Given the hurdles associated with applying for loans it is hardly surprising that seven firms in the sample were discouraged even from embarking on the loan application procedure, while six other applicants either were denied loans or find themselves still awaiting the results of their applications after an inordinately long period. It is telling that five of the six discouraged furniture firms already had some track record of success and exhibited significant potential for growth - precisely the classes of enterprises that should be able straightforwardly to graduate to formal institutions in a well-functioning financial system. 35. Nor do firms dealing with formal financial institutions confront obstacles only during the process of applying for loans. As state above, virtually all enterprises with bank loans reported that their relationship with their banker continued to be difficult. For three of the furniture firms (all in the medium-stagnant category) these difficulties appear to be related to the precarious financial condition of the enterprises themselves. But of the remaining seven (healthy) firms, two reported that the bank imposed its own arbitrary loan ceiling thereby disrupting the enterprise's business plan even though all prerequisites for a larger loan had been fulfilled, and two others reported that their banker temporarily shut down their overdraft facilities over disputes that were peripheral to the lending relationship itself. I/ For details on the relevant financial sector aspect of the SME development, see Chapter 3, Volume 1. 112 36. In all, the conclusion that emerges from the interviews is that Tanzania's banks do not view their lending to indigenous SMEs in developmental terms, with a goal of identifying and supporting enterprises that have real potential to succeed commercially, and a mutual interest on the part of both the bank and the enterprise in a venture's success. Rather, as indigenous enterprises see it, the banks perceive them to be unwelcome supplicants, to be dealt with only so long as stringent conditions are met and, even then, only so long as the enterprise respects without question the omniscient authority of the lending agency. Regulatory Obstacles to Private Enterprises In Tanzania 37. As the scores in Table 3 revealed, regulatory difficulties, difficulties in gaining access to an industrial site, and difficulties winning tenders ranked high among the obstacles to entry and expansion identified by sample firms. This subsection details the character and economic consequences of these regulatory obstacles. Rather than analyzing comprehensively the full spectrum of laws and regulations affecting private enterprises in Tanzania, the focus here is on the impact of the regulatory regime as it is experienced by interviewees. 38. Four groups of difficulties associated with regulation were identified by firms. Two of these - problems associated with the licensing of enterprises to operate, and problems associated with the assessment and collection of taxes - were shared in common by both furniture and construction enterprises. An additional group of problems for furniture enterprises involved gaining access to industrial sites, while an additional group of problems for construction contractors involved regulatory hurdles specific to the construction industry. 2' 39. Enterprise licensing. The system of commercial licensing in Tanzania is exceedingly comprehensive, with local authorities empowered to require licenses from, and issue licenses to, all commercial ventures irrespective of their size. Enterprises with fewer than ten employees are required to obtain a trading license; enterprises with more than ten employees and enterprises that sell to government agencies are required to obtain a business license. Licenses are renewable annually, for a modest fee. 40. Unlike other countries, where small enterprises often fall outside the regulatory net (and hence sometimes are described as 'informal'), as of 1989 enforcement of Tanzania's licensing requirements appeared quite thorough: all but two J' of the 38 interviewees had been licensed to operate by the Dar es Salaam City Council. There was no evidence that license requirements were used in any sustained way to inhibit entry by new firms. Even so, interviewees reported that - along with the fee itself - compliance with the requirements entailed significant costs. 41. These costs can be traced to the requirement that licenses be renewed annually. This renewal is not automatic, but depends on prior clearance from various tax authorities. In the best of circumstances, obtaining these clearances and thereafter the license renewal imposes a 2/ A fifth group of problems - those associated with the provision of telephone services, electricity and water to enterpriscs -- also emerged repeatedly in interviews. While many of these problems appear to be at least as much administrative as they are technical, in-depth analysis was beyond the scope of the present effort. JQI Both of these were unlicensed construction subcontractors, able to evade the regulatory net by virtue of their ability to operate without any ftxed business premises. 113 significant burden on the scarce time of the enterprise proprietor. In circumstances where not all the necessary clearances have been obtained, the proprietor is at the mercy of the licensing official. In practice, incomplete compliance does not appear to have led to the shutting down of enterprises, but instead functions as a vehicle for officials to extract side payments. Hence the report of about half the interviewees that 'lubrication' is an integral part of the completion each year of license formalities. 42. Tax collection and assessment. Tanzania's construction contractors are liable for payments of income taxes, its furniture enterprises for payments of both sales and income taxes. Sales tax for sales of furniture, payable by manufacturers (and again by retailers if they operate independently of the producers), is formally set at a rate equal to 25 percent of sales revenue; income taxes are set at 50 percent of enterprise profits. 43. The enterprise survey uncovered three noteworthy features of the way in which tax laws are implemented in practice. First, corresponding to the evidence on licensing requirements, even the smallest furniture enterprises in the sample confirmed that they were regularly held liable for both sales and income tax payment; there appears to be no lower bound on the size of firms that fall within the tax net. 44. Second, only one-fifth of the interviewees were willing to characterize the system of tax collection as 'legitimate'; the remaining firms confirmed instead that 'lubrication' was pervasive in their dealings with the tax authorities. Underlying these data there is the broad perception that tax requirements impose an unmanageable burden on enterprises. This perception (together with the low salaries paid to government employees) legitimizes for both owners and tax collectors a process whereby the tax obligations on enterprises are renegotiated (via downward estimates of sales revenue and profits), with associated side payments to the tax assessors themselves. Sometimes the process of renegotiation can be quite antagonistic: one four-person furniture enterprise described how an initial reluctance to 'lubricate' resulted in a sales tax assessment so high that it stood to bankrupt the venture; at that point, the owner had no alternative but to plead with the assessor to accept lubrication in return for scaling down the tax bill. More generally, the combination of unrealistic obligations and the expectation of renegotiation implies that even entrepreneurs who might otherwise prefer to adhere to what is legally required have no alternative but to 'lubricate' the system if they are to survive in a highly competitive environment. 45. The third feature of the implementation of tax laws was the frequency with which even the smallest firms are harassed by tax officials. With sales tax legally payable on a monthly basis, tax officials have the opportunity to visit individual enterprises and 'negotiate' their tax burden twelve times annually. In practice, the number of negotiations is somewhat fewer - averaging 5 per annum - but the burden on enterprises is hardly trivial: all the furniture firms that provided a ranking highlighted the burden of dealing with sales tax officials as the most severe of the regulatory obstacles which they confronted. Income tax is payable annually, so in principle the number of legally mandated visits is fewer. However, individual tax assessors have the right to re-examine the accounts of any enterprise at any time on the grounds that inaccurate information has been filed; thus, over one-half of the enterprises reported receiving more than one visit per annum from income tax officials. 46. Access to industrial sites. Only 5 of 15 firms who provided relevant information in the enterprise survey had their premises on land allocated for industrial purposes. This scarcity of 114 industrial sites in Dar Es Salaam appears to inhibit both the entry of new enterprises and the expansion of existing fims. 47. The high score of 0.84 in Table 3 accorded to lack of access to an industrial site as an obstacle to entry for furniture firms (almost as high as the score for the leading obstacle to entry - - lack of access to finance) signals how severe is the difficulty of identifying a piece of land that can serve as premises for a new enterprise. As for a Mady-established firms on land that is not zoned for industrial use, one potential problem - albek. one which was not supported by survey evidence - is the vulnerability to harassment created by their extra-legal status. A second problem is the disincentive to expansion that follows from an ambiguous land tenure status. With their rights uncertain (and perhaps their existing site inadequate) firms tend to be reluctant to invest in a permanent structure or to acquire equipment that requires the shelter and eciities of a permanent structure. The survey evidence indicated the importance of this problem: four of the ten enterprises without sites volunteered directly that they would expand if they had access to such a site. I! 48. Construction reszulatory and tendering procedures. The combination of: i) safety issues; ii) the inevitable unforeseeable problems inherent in any construction project; and, iii) complications associated with the sequencing of construction work and payment together imply that the construction industry the world over is beset by complex regulatory requirements, often enforced in quite dubious ways. In any country, the best to be hoped for is that the vast majority of construction contracts and associated regulatory enforcemert proceed honestly, with instances of corrupt practices the exception. As the interviews with construction contractors reveal, Tanzania falls well short of this desired state of affairs. 49. Two different sets of regulatory problems specific to the construction industry emerged in the course of interviews. The first set, relevant equally to contractors who serve private and public clients, concerns the bureaucratic steps associated with a construction operation: contractors need to certify the drawings on the basis of which they plan to proceed, to obtain a building permit, to obtain certification of work completed prior to payment at each phase of a contract and, upon completion, to obtain a certification of readiness for occupation. At each step, the individuals responsible for providing the requisite certification are in a position to extract side payments from contractors: all 11 contractors willing to provide information reported the need for lubrication to secure certifications of work completed, 10 reported that lubrication was required to secure issuance of a building permit, 9 reported a need for lubrication to secure certification of occupancy, and S a need for lubrication to secure certification of drawings. 50. The second set of regulatory problems are relevant to construction contracts with government agencies, specifically government tendering, contract renegotiation, and payment practices. To begin with tendering practices, the assessment of the extent to which tendering procedures represent an obstacle to expansion was particularly severe for the firms who currently ./ In fact, a recalculation of the score in Table 3 to include onry those enterprises for whom la^k of access to an industrial site could plausibly serve as an obstacle to expansion suggest that for them the scarcity of poperly allocated land (for industrial use) represents almost as substantial an obstacle to expansion as does lack of acces to fmoe. 115 contract exclusively with private clients. As was explicitly suggested by some firms, this result could reflect lesser likelihood of access to nepotistic contacts among these firms. w 51. While the average delay reported by interviewees between completion of work and certification of due payment is almost two months for government contracts versus two and three weeks for contracts with private clients, and the delay in payment subsequent to certification averages 5 months (versus 7 days), none of the enterprises expressed any desire to reduce their involvement in government contracts, suggesting that renegotiation rarely leaves firms disadvantaged. In fact, numerous industry observers indicated that delay and renegotiation - while leaving unfinished many public construction projects - often can serve well the pecuniary interests of both contractors and their client counterparts. 52. Overall Impact of rqgulatorv obstades. Three major conclusions emerge from this review of the impact on enterprises of Tanzania's regulatory regime. 53. First, although many of the regulatory requirements for enterprises are exceedingly burdensome, in practice many firms are able to mitigate these costs. Other than the problems associated with the lack of access to industrial sites (for the furniture industry) and the potentially distorting influence of corrupt tendering and renegotiation practices on the behavior of enterprises (for the construction industry), their negative impact is less than might appear from the formal requirements as they appear on paper: pervasive lubrication reduces the impact to firms of egregiously unrealistic and arbitrary policies. Obviously, unrealistic policies offset by pervasive lubrication is hardly desirable: corruption is wrong and its existence gives an incentive to engage in unproductive rent-seeking activity; it gives arbitrary advantages to firms that are more successful in cultivating co-operative relations with public officials; and it undermines the capability of the public administration to perform those functions where their contribution is necessary. 54. Second, notwithstanding the above, the overall costs to enterprises of the prevailing regulatory environment are substantial. The official payments, side payments, and time spent navigating the bureaucratic morass can be interpreted as entry-deterring fixed costs, and expansion-deterring variable costs on enterprises. Nor should these costs be viewed only in financial terms: their most important component is the opportunity cost associated with the diversion of scarce entrepreneurial time and effort away from socially efficient wealth-generating activities. 55. Third, the relative burden of the prevailing regulatory environment is significantly greater for the smaller, indigenous enterprises than it is for large, often non-indigenous, ventures. When firms are large enough they can 'grow out' of the costs imposed by an obstructionist bureaucracy: senior executives can delegate the management of relations with the bureaucracy to employees; they can use the influence afforded by their economic power to shield themselves against egregious examples of harassment; and, where necessary, they can cover whatever side payments are required without reducing substantially the working capital available for their business enterprises. In sum, Tanzania's regulatory environment has the paradoxical (and presumably 2IN None of the conractors wth government clients identified government tendering procedures as their greuest obsio to expansion. 116 unintended) consequence of strengthening the competitive position of large enterprises in relation to indigenous SMEs. Other Obstacles to Private Enterprises In Tanzania 56. The remaining obstacles identified in Table 3 can be grouped into three categories: obstacles associated with access to inputs other than finance; obstacles associated with the high price of inputs; and obstacles associated with a scarcity of technical and marketing skills. 57. Access to nonfinanidal inputs. The obstacles to expansion of furniture enterprises associated with lack of access to equipment, spare parts and wood are all ranked low, signifying the existence of well-functioning input markets in these areas ?' oresent, which appear to be associated with the recent internal and external trade liberalization. In the market for wood, large numbers of indigenous entrepreneurs have specialized at a range of levels, with some procuring wood from local forests via their own agents and then retailing in small lots to urban users, and others coming direct from the countryside and wholesaling lots directly to larger urban users. As for equipment and spare parts, these are purchased from traders who in turn import their wares for abroad. 58. Lack of access to equipment and materials were ranked notably higher as an obstacle by construction than by furniture firms. But what is noteworthy for both furniture and construction is the decline between the time of entry and 1989 in the scores for obstacles associated with access to equipment, materials and spare parts, suggesting the importance of the liberalization program (particularly own-funded imports, OGL and deconfinemert) in increasing the availability of inputs for SMEs. 59. The cost of inMuts. In terms of perceived "obstacles" firms ranked relatively highly the price of inputs, in contrast to the lower ranking of their access to inputs. Both findings, are, in fact, consistent with the improved functioning of the market mechanism. 60. While the presence of obstacles involving limitations in access by enterprises to requisite inputs would have signaled a failure in the functioning of the market mechanism, no such implication follows from the presence of obstacles associated with high input prices. On the contrary, reports by enterprises of high input prices can be interpreted as a signal of significant competitive pressu-e and an associated squeeze on profits. Thus the high scores for obstacles associated with the cost of inputs - coupled with evidence that, with the exception of finance, markets for furniture inputs function relatively efficiently - can be viewed as evidence that competition in the unregulated Tanzanian wood and furniture markets sets a ceiling on profits which serves well the interests of consumers. I 61. Scarce technical and marketing skills. The low ranking for obstacles associated with a lack of technical skills, a scarcity of competent workers, and the limited size of the market are LU/ Note alsO the possibility that obstacles associated with high input costs might signal leads and lags in the transmission of macroeconomic policy changes to firsu, with the price of some inputs rising rapidly in response to changes in macro variables (the nominal exchange rate, or the rate of inflation to cite two examples) while output prices adjust more slowly. 117 consistent with the earlier proposition that barriers to entry into both construction and furniture manufacture are low. It is important, however, not to misinterpret these results: most of the firms in the smple evaluated the obstacles in relation to the clientele they were serving at the time, and a production quality that sufficed to meet the needs of domestic consumers. Independent evidence for the furniture industry in particular suggests, however, that the magnitude of the obstacles are very different for enterprises with more ambitious goals. 62. On the technical front, only one of the indigenous enterprises interviewed had organized its workshop on a quasi-assembly line basis consistent with the fulfillment of orders for high volume, and even this enterprise ran Into major technical difficulties as it attempted to upgrade the sophistication of its equipment. On the marketing front, only two of the indigenous enterprises expressed any Interest In producing furniture for export; only one of these was aware that sophisticated wood-curing techiques (not now used In Tanzania) were a prerequisite for the penetration of export markets; and neither enterprise had any clear sense of how they might proceed to learn in more detail about export possibilities. 63. As increasing numbers of progressive indigenous fiuniture enterprises follow the lead of large firms and endeavor to move into new, more demanding markets, they are likely to learn that technical and marketing obstacles are substantially more formidable than they now perceive them to be. Whether they will be able to overcome these obstacles will depend crucially on the quality of the (private and public) technical and marketing support services to which they might then have access. ENHANCIXG AIP EEPEQlYMS IN TANZANIA Tanzania Economi,. Report Background Paper #4 A. THE CHANGING POUCY AGENDA Development assistance in support of ERP has concentrated on balance of payments support to enhance the country's Import capacity and on the process of policy reform. This contrasts with earlier concentration on project assistance. As a result, the focus of discussions of aid policies has shifted from issues of project design and implementation to the macroeconomic policy environment and policy conditionality. But as economic activity has been revived in the productive sectors in response to the more favorable policy environment and increased import capacity, weaknesses in the public provision of infrastructure increasingly constrain economic performance. The government and the donor community also share a concern that during the years of economic decline and during the initial years of the ERP there has been an erosion in the delivery capacity of social services which now needs to be redressed under the ESAP. As a result, the balance of attention has begun to shift back partially to earlier aid concerns; issues related to public sector infrastructure and social service projects must be addressed and public sector institutional constraints resolved. Lessons can be learnt from the earlier experience by donors and Tanzania on project aid. This paper therefore addresses both macro aspects of the aid relationship and project level issues. The historical record of aid performance that provides the context for this discussion is far from positive. B. THE LOW ?RODUCTIVITY OF PAST ASSISTANCE The period 1973-81 saw a remarkable build-up of external assistance to Tanzania. The aid effort was unsuccessful, however, as at the end of the period the economy entered a period of economic decline, in which not only per capita output declined but earlier achievements in the provision of social services were eroded. How far can the absence of positive macroeconomic effects be taken as evidence of inherent flaws in the aid program of the 1970s? In part, failure resulted from exogenous shocks and from domestic policy weaknesses, both of which can be readily documented. However, although some exogenous factors were inherently unpredictable, one lesson of the 1970s is that insufficient attention was given to the impact of public policies in aid design and appraisal. The policy regime cannot be treated as a factor entirely exogenous to the aid program. In practice, policies were sustained by continuing external support and might have been reformed earlier if that support had not been forthcoming. Moreover, such a large part of the government's development program was externally financed that important elements of the institutional and policy structure were directy influenced by aid. While the course of the economic crisis of 1979-84 can be charted in terms of severe macroeconomic disequilibrium, accentuated by an inadequate macroeconomic policy response, very low returns to investmnents made in the previous decade contributed to the underlying weaknesses in real economy. In addition to exogenous shocks and macroeconomic policy weakness, failures at the project level in relation to identification, design and implementation, and in the balance of public investment program contributed to the negative outcome. Aid projects - 121 - tied up scarce local resources, including administrative capacity. The net impact of the many projects which failed was negative. The sobering reflections on the results achieved by the large inflow of aid in the 1970s is that Tanzania's growth would probably have been higher with a much more modest program, if that program had been designed with greater care and implemented with more caution. Ultimately, however, responsibility for the effectiveness of an aid program lies with the recipient. The decision whether or not to accept aid rests with the government, and consequences of success and failure are felt more acutely in the recipient country. In the case of loans, moreover, it is the recipient country which ultimately pays. Even with grants there are always local costs, either explicit pecuniary costs or implicit opportunity costs. The system of aid coordination, appraisal and control has never been sufficiently effective on the part of the Tanzanian govermment. During the past decade a number of initiatives have been launched to improve aid coordination, but these have been insufficient to establish quality control. In fact, the existing system cannot be expected to prevent the government from making the mistakes which resulted in such low returns to aided projects in the past. Consequently, strengthening the system of aid coordination and control should be a government priority. C. CURRENT MACROECONOMIC ISSUES 1. Aid and the Balance of Payments Since 1982 total aid flows net of technical cooperation grants have exceeded recorded export earnings. With the shift in the composition of aid towards import support, begun in 1982/83 and accelerated under the ERP, disbursements have become important not only for investment funding but also for the finance of intermediate and consumer goods ifnports. Initially, import support was begun by a number of bilateral donors as a crisis response to the constraints facing all sectors as a result of the severe decline in import capacity. With the implementation of the ERP, import support has become a systematic feature of the aid program, running currently at around US$ 380 million per annum. It accounts for about one-third of official estimates of total imports (which equalled USS 1,150 million in 1987 and US$ 1,185 million in 1988). The growth of such a large balance of payments support program poses two sets of related questions concerning: (i) allocative efficiency, and (b) longer term sustainability. (a) Allocative Effrciency The growing role of import support and the importance of "own account" (or "own funds") imports are features of a complex foreign exchaDge market which has emcrged, involving at least six different types of foreign exchange transactions: (i) allocation by the Bank of Tanzania using resources generated by trade and service transactions flowing through official channels; (ii) transactions using foreign exchange retained by importers under export promotion retention schemes; (iii) "own account" transactions funded through the parallel market; (iv) goods imported as a component of aided projects; (v) transactions funded under import support schemes, allocated between firms and/or commodities under agreement between the donor in question and - 122- government; and (vi) transactions funded from the OCL scheme, providing for automatic allocation of foreign exchange for imports included in the OGL scheme. The uses of imports supplied under aid are to varying degrees insulated from market pressures reflecting real foreign exchange costs. Project aid only reflects real foreign exchange costs in so far as they are effectively accounted for in appraisal. The implicit exchange rate in the different segments of the market varies not only because the actual rate of exchange varies (notably between transactions at the official and the parallel rates), but also because financing facilities vary. For example the OGL system has required shilling cover at the time of issue of a letter of credit whereas bilateral import support in practice has sometimes involved generous interest free credit, payment being made after delivery of the commodity. Given the segmented structure of the foreign exchange market in effect involving multiple exchange rates, transactions at the official exchange rate and under the liberal financing conditions applicable in practice to some import support programs generate rents. However, these rents are now much less than under the scarcity conditions of the early 1980s, when there was a much greater gap between the official and parallel exchange rates, and firms favored by access to import support were at an enormous advantage. As exchange rate adjustment reduces the gap between official and parallel exchange rates, and a shift to the OGL system increases the availability of fbreign exchange for trade transactions at the official rate, the system is moving towards an increasing degree of equilibrium, at least in relation to the trade account. However, both on the part of some donors and the government there has been some resistance to extending the liberalization of the foreign exchange allocation system. On both sides it apparently reg%t; a measure of skepticism regarding the virtues of market allocation as compared to burea:ratic control. On the donor side ftis in part reflects the need to demonstrate that resources hWve been used for purposes which are seen as desirable in terms of the donor's aid philosophy. The genera o:N of rents by aid programs carries with it serious dangers. At a time when the government has )-.istered great concern about the spread of corruption in the public service, it should be recognia that the existence of opporunities to make gains from rents through aid allocation and procurtc.ent could become a potent source of malpractice. Systems of aid allocation which minit ik the possibility of windfall gain by particular individuals or firms would also minimize the inceitive for corrupt practice. (b) Long Term Suslainability The degree of dependence on external assistance to sustain the inflow of imports poses thorny questions regarding the sustainability of recovery. How long will donors be willing to maintain high levels of import support? Recovery should be expected to be a transitional, finite process. At the same time, there is a need to restore public investment, as infrastructural weaknesses constrain performance. Will sufficient foreign exchange be generated both to fund necessary recurrent imports and expand public investment? How much foreign exchange is required to sustain economic recovery is difficult to judge. In the early 1980s the economy was suffering from a virtual goods famine. There were - 123 - minimal stocks both of consumer goods and inputs required at all stages of the production process. Part of the import demand of the ERP period has been to restore levels of stocks and catch up with rep' icement (e.g. of the vehicle fleet). With goods more readily available and the exchange rate no longer as overvalued as in the early 1980s, the demand for foreign -xchange seems to have stabilized. Traders now claim to be constrained by the availability of shillings and demand for bilateral import support funds and for 'own account' import licenses are both steady. If there is less pent up demand for imports, and recurrent import requirements can be expected to grow in line with growth in domestic income and product - and imports are efficiently allocated - the task of increasing the proportion of recurrent import requirement funded by Taniania's own foreign exchange earnings appears more manageable. 2. Fiscal Dependence In the past, aid was mainly directed to the development budget, recurrent spending being funded from local resources. This was never an entirely satisfactory distinction. Most development spending is only productive if recurrent resources are available to utilize the resulting additional capacity. Compartmentalizing development spending, and neglecting its recurrent implications, can result in the creation of capacity which was unsustainable, as was the case in Tanzania by 1980. Moreover the distinction between investment and recurrent spending is often arbitrary. Maintenance activities and investment are direct substitutes, as demonstrated by the need for rehabilitation after many years of inadequate maintenance. Some recurrent expenditures contribute more directly to expansion in productive capacity than the related 'investmentes (e.g. in education). There is therefore a good case for modifying earlier doctrines which concentrated development assistance on development budget activities, to extend the scope of developmeut assistance to activities which are close substitutes for investment or are developmental in their impact even if they are categorized as "recurrent" in conventional public accounting terms. However, with the growth of import support funding, Official Development Assistance (ODA) has taken on v much broader function, supporting the recurrent budget in a general way. With the expansion of import funding, domestic financial adjustment was necessary to accommodate the counterpart funds (unless the funded imports were simply transferred as a gift to the users). Import support could be accommodated either by: (i) a transfer of the counterpart funds to the government budget; or (ii) the sterilization of the counterpart funds by the central bank. The mechanism adopted has been the first. The consequences of transferring counterpart funds to the government's revenue account depends on the effect on government behavior. If goverment spending and revenue collection are assumed to be unaffected by access to counterpart funds, their utilization as recurrent revenue would keep the national debt below what it would otherwise have been, providing a less inflationary method of funding the deficit that alternatives. In practice, however, budgetary use of counterpart funds is likely to incre2se the real level of government recurrent spending. Given the problems faced by the government because of the constraints on the recurrent budget, the cushioning of downward pressure can be seen as desirable in the short term. However, it is doubtful whether import support counterpart funding at a level more than one third of the recurrent budget will be sustainable in the longer term. There is a danger that ready access to counterpart funds could postpone necessary fiscal reform. - 124 - 3. Development Assistance and Domestic Savings Aid should supplement domestic savings, not substitute for it. There are circumstances in which it is appropriate to support consumption activities as was the case in the early 1980s when the extreme shortage of consumption goods, and resulting repression of consumption levels had a depressive effect on production. However, over the long-term aid should increase the growth in capacity to produce through increased levels and efficiency of savings and investments. Taking the long view, and even abstracting from the special circumstances of the crisis of the early 1980s, the past evidence in Tanzania is consistent with the view of those critics of aid who have argued that aid tends to substitute for, rather than add to the domestic effort. In the 1960s capital formation was built up largely on the basis of a domestic savings effort. By 1970/71 savings had risen to 24% of G.D.P. Subsequently there was a sharp decline in domestic savings except for the brief period of export (tea and coffee) boom, 1976-78. The period of aid build-up (1973-81) was also a period of declining savings. This does not necessarily imply, however, that aid directly caused the poor savings performance. The explanation of the poor savings performance lies in the poor performance of real foreign exchange earnings in the 1970s. It is only possible, in macroeconomic terms, to realize planned domestic savings if there is foreign exchange available to cover the import content of investment activities. To establish a causal macroeconomic link between high aid and low savings it would be necessary to demonstrate that high levels of aid in some way undermined foreign exchange earnings (e.g., if aid flows sustained the exchange rate at higher levels than would otherwise have been the case), or promoted consumption of imports. One reason why the increase in aid was not fully reflected in an increase in capital formation was the low level of private capital formation and private capital flight in the 1970s. Again, there is no simple causal relationship - the decline in private investment activity reflected the impact of a number of government policies rather than "crowding out" by aid finances public investments. There is currendy a revival in private investment underway which is indirectly supported by current aid practices. The provision of aid in the form of balance of payment support rather than funding for the public investment program more readily accommodates the import requirements of private investment activity. D. CONDMONALITY AND POLICY DIALOGUE A new feature of this relationship between Tanzania and the donor community during the 1980s has been the importance of more general policy conditionality, as reflected by the interrelationship between the IMF, the donor community and Tanzania. In the 1960s the IMF played a low key role, and in providing expanded facilities in response to the first oil shock, there was only modest conditionality. The Fund's role has been more active during the present period of reform. Macroeconomic conditionality has focused attention on the management of a range of macroeconomic policy instruments, notably the exchange rate, the budget, and monetary - 125 - instrum^ents such as growth in the money supply and interest rates. These are important policy instruments, particularly for short and medium term macroeconomic management. Extreme disequilibrium in foreign exchange markets and high rates of inflation undermine development activities across the board. Thus firm action in relation to these policy instruments was a necessary condition for recovery, and at the beginning of the reform period focusing policy conditionality primarily on these areas was appropriate. However, as the process of reform deepens, other policy issues can and have taken on greater significance, even for macroeconomic management. Currently, institutional questions relating to agricultural marketing, the financial sector and the management of the public sector (including the civil service) have become more of the primary constraints to extension of the recovery process. As concentration each year on the IMF negotiations focuses government officials' energies and attention on the policy agenda, it will be important for the IMF to continue to ensure that the agenda it helps to define includes these and other emerging important issues. There has been an increasing tendency to develop work involving teams drawn from the donor side, the Tanzanian government and non-government Tanzanian experts. This approach has major benefits. It recognizes the increasing body of Tanzanian expertise, the mobilization of which at an early stage of policy analysis is likely to be productive in improving the empirical basis and the local relevance of policy studies. Joint efforts have a greater chance of internalizing the reasoning and conclusions of policy analysis within the Tanzanian system. This is particularly important where improved policy requires ongoing adjustments in approach rather than a once and for all adjustment in a given policy variable. While the process of policy dialogue has been highly productive in recent years, there are aspects which require continuing attention. The Bank report, Morld Bank/Tanzania Relations. 1961-1987, includes a systematic history of the complicated, and not always happy, policy dialogue between the World Bank and the government of Tanzania over the past three decades. "Policy dialogue" ranges from provision of technical assistance, to advice and persuasion in donor/government interchanges, to negotiated policy conditionality in relation to specific aid agreements. The Bank report offers a number of important insights drawn from study of the historical record. These include: (i) the need to understand the political and social environment and policy objectives on the recipient side, and the need to have a realistic view of the range of policy options practically available; (ii) the advantage of an advisory rather than "mandatory" role; (iii) where inadequate policies make lending unadvisable in the Bank's view, the reasons should be set out cogently and explicitly; (iv) there is a need for improved coordination within the Bank to ensure consistency in advice and conditions; and (v) there needs to be effective coordination between donors involved in policy dialogue. The need for coordination is self-evident. From the point of view of the Tanzanian officials, the sheer volume of work involved in a continuous flow of missions and subsequent flow of reports, in their varying stages of preparation, becomes unmanageable. Existing arrangements for sectoral policy work and dialogue could be strengthened. There is no overall planning of work - initiatives reflect the requirements of the various agencies more than a thought out and agreed position between donors and the Tanzanian government about the priority and proper sequence in which policy issues should be tackled. - 126 - An improvement to existing arrangements would be to adopt a practice as follows: (a) using consultative group meetings not only to discuss the policy framework paper and external financing requirements for the coming year, but also to discuss government proposals on priority issues for further review - and the results of those reviews; (b) the government could then agree with the donors on collaborative work programs involving inputs from the government, multilateral and bilateral donors and other local and foreign experts; on the donor side a 'lead agency' might be proposed by the government and be appointed to be responsible for coordinating donor work on a particular issue, in consultation and collaboration with a "lead ministry" in Tanzania which the government would also identify; and (c) in allocating tasks there would be a need to rationalize and concentrate donor assistance; those donors requested and agreeing to take a "lead role" would need to be very active in assisting the government, including strengthening local capacities, while others would have to consider relying more on the government and the lead donor agency or agencies to guide decisions on involvement and financing; they would also need to exercise restraint in fielding missions and taking up policy issues not on the agenda. The Bank report on World BankITanzania Relations sees the Bank as playing a lead role, sometimes a dominant role, in aid coordination. By and large the donor community accepts the Bank's leading role in relation to macroeconomic work. However, as the Bank report recognizes, in practice the Bank's staff capacity is predominantly located at headquarters. In relation to sectoral work, where local knowledge is particularly important, some bilateral donors have stronger representation in Dar es Salaam. Moreover, in some areas bilateral donors have built up a more substantial and continuous experience of field operations in project implementation than the Bank. It is therefore imperative for donors to take account of comparative advantage within the aid community in allocating tasks. E. IMPROVING PROJECT EFFECTIVENESS The depressing record of past project performance is summarized for the World Bank's portfolio in World BanklTanzania Relatons. Of projects which have been subject to operational evaluation the following conclusions resulted: (a) In the rural sector 12 out of 17 projects had a aegative re-estimated Economic Rate of Return (ERR); and 4 were substantially lower than at appraisal; (b) the two industrial projects evaluated had a negative ERR, while two other industrial projects had capacity utilization of 4% (Morogoro Shoe Factory) and 49% (Southern Paper Mills); (c) five out of 6 transport projects had ERR below estimates at appraisal. However, operations in Telecommunications, Energy and Power and Water Supply and Sanitation were generally successful although typically delayed in implementation; - 127 - (d) of 6 education projects reviewed only two were successfully completed. Others were subject to excessive delays in physical construction and were hampered by staff shortages. Evaluation by other donor agencies suggest that the poor performance of Bank projects were typical of aid performance in general. One problem which emerges from the record is that of donor comparative advantage and delivery capacity. Large-scale, capital intensive, engineering based infrastructural projects, the main focus of ODA in its early days, were appropriate vehicles for aid in that they require foreign expertise and participation in implementation, whatever the source of funding. Successful development in agriculture and the social services depends more on local small-scale activities, is best implemented with technologies at the labor intensive end of the range of possibilities and typically depends on human inputs as much as on physical hardware. High technology inputs may be required in the small scale sector. For example, sustained external support for research is necessary for a breakthrough in agriculture. However, the impact of research depends on adoption at the farm level, requiring extension into the small farm systems through a carefully designed process of local institutional development. Donors should not ignore agriculture and the social sectors, but donor initiatives in these sectors should be more cognizant of the real limitations of their own delivery capacities in the design of programs. In practical terms this means addressing a number of specific issues: (i) the manner and content of initial projects design; (ii) flexibility in implementation; (iii) institutional sustainability; and (iv) financial sustainability. (a) Project Design The design and appraisal of projects has been very much donor driven. Broad priorities and targets (e.g. rural water and primary education targets) have been set by the Tanzanian government but the translation of ambitions into concrete projects has been heavily dependent on donor initiative and technical inputs. As a result of this, the project does not necessarily respond to the priorities of the eventual clients, nor incorporate a realistic view of local capabilities. The govermment planning structure does not effectively articulate local requirements nor realistically assess local capabilities. The aid coordinating agencies of government (int Finance and Planning) have capacity located in ministerial offices in Dar es Salaam but lack the manpower and mandate to play an active role in identifying projects and probing the institutional arrangements for their implementation. Implementing agencies (e.g. line ministries) may have more practical knowledge but are also likely to have their own institutional interests at stake - their role is more likely to be that of advocate than appraiser. It is easy for the project design work on the donor side to concentrate on the "hardware" of the project neglecting local institutional capacities. Institutional "software" tends to be treated formalistically, if at all. Organizational structures are specified within which key actors are assumed to perform in a well-behaved fashion, consistent with project objectives. Donors have implemented massive programs to deliver services to rural Tanzania, which could only be sustained by local communities and users and the local technical cadres in the field. This was most obviously the case in relation to rural water supply, where donors accepted - 128 - ambitious targets to extend the system without taking account of the complex requirements for subsequent operation and maintenance. A program of borehole drilling can be mounted using external consultants and construction firms. Subsequent operation and maintenance depends on the local capacity. The development of that capacity was the crucial constraint which should have influenced the design and pace of implementation of the program, but which was in practice neglected. Bottlenecks have also emerged in the capacity of central institutions to handle the volume of resources supplied under aid, and to transmit them to local operational use. The difficulties of organizing the supply of school text books, involving a central government agency, private contractors and eventual supply to the districts, have not yet been resolved. In some areas, the dependence on external know-how is virtually inevitable. Where donor leadership of project design is the necessary result of the lack of Tanzanian experience and capacity, this should have direct implications for design. In such cases the development of local technical and managerial competence should be a key objective of project design - if not the most important element. Yet in many instances the issue has been neglected. Where there is no realistic scheme to meet the technical and managerial needs of a project, that in itself should be grounds for rejection. If finance is provided for an elaborate project (e.g., Morogoro Shoes) without provision of an adequate back-up program for subsequent management and operation, the donor agency is essentially in the position of funding the delivery of a turn key project. By contrast, relative success in projects involving sophisticated technology has been achieved where there is an existing organizational capacity which is capable of assimilating the project. In areas such as electric supply (TANESCO) and telecommunications (MPTC), long standing public utility corporations with an existing technical capacity provide an appropriate context for high technology investnent. Even where a Tanzanian institutional capacity exists, much of the detail of project preparation and appraisal will be in the hands of expatriate experts, applying donor appraisal criteria. This, in itself, would be no bad thing if the donor exercise were paralleled by systematic appraisal on the Tanzania side. This does not typically happen. While there is frequent debate over project design, reflecting Tanzanian policy thinking and institutional interests, there seem to have been few instances where the Tanzanian side has questioned the economic viability of a project which has met donor criteria. Given the fact that many projects which have passed the test of donor appraisal have nevertheless failed disastrously, and in some cases (when loan funded) the cost is eventually borne by Tanzania, passivity in this regard has been costly. Points which have arisen in enough cases to be treated as general issues include: (a) When ODA is expanding there is a tendency for the project preparation process to be driven by the need to achieve spending targets. Within donor agencies, departments are likely to ensure their command over agency funds if project commitments exhaust available resources, and career success is likely to depend on effectiveness in preparation, packing and negotiation of projectF. There is no mechanism in most agencies by which careers are affected by the eventual operational success or failure of a project (by which time the staff in question are likely to be handling different portfolios possibly at a higher rank). - 129 - (b) Detailed preparation and appraisal is often in the hands of consultants. While they will, no doubt, be concerned for their professional reputations, nevertheless then they typically have an Involvement in the project of strictly limited duration, their immnediate success or failure being judged on the basis of the document which emerges from the appraisal process. (c) Preparation is likely to be stronger on the design of the physical components of the project and its capital investment phase, and less strong on the human institutional side and the operational phase. (d) Although many consultants will be familiar with local conditions, there is still a tendency to design to standards and practices which conform to international norms and neglect issues related to the adaptability of the technology to local conditions and to the problems of subsequent maintenance. Many lessons have been learnt and incorporated in programs and project proposals in such areas of rural water supplies, textbook procurement, etc. In future project design, consideration of these institutional issues should be as explicit as considerations of the 'harder' engineering aspects and should incorporate a realistic assessment of the constraints which impede institutional performance. This may suggest the need for specific institutional changes, or for redesign to increase the prospects of viability even in the light of institutional weakness. (b) Flexibility In Implementation Another weak aspect of project design has been inflexibility resulting from the typical project design/appraisal/implementation cycle. Weaknesses can be described in terms of a Dremature specificity in project design and inflexible commitment in implementation. Many development activities are inherently risky, both because of the need to venture into areas where the required background information is scanty and because the economic environment in which projects are implemented is highly changeable. The appropriate response should be experimental design, developing and modifying the activity in the light of experience and knowledge gathered in the early stage of implementation and adjusting to new circumstances as they emerge. The logic of the public project process tends to demand the specification of a project in great detail in advance of its funding and, once agreed, tends to persist too long with implementation of the agreed plan irrespective of changing circumstances. In agricultural projects, for example, a commitment to work in an area or on development of a particular commodity may be reasonably made on a basis of a broad assessment of market prospects and agroi,omic potential. However, detailed assessment of appropriate inputs, farm level constraints and farmers' response to possible interventions is almost never available on a sure enough basis to plan a detailed multi-year program. Part of the answer is, no doubt, a general effort to improve the knowledge base. However, action can rarely be based on full information and scientific analysis. A practical approach is one of trial and error, in which the design of an activity develops on the basis of field experience. - 130 - Looking back at some of the aid disasters the depressing aspect is not that the initial commitment was made (although the size of many commitments must appear foolhardy in light of inherent uncertainties); indeed development requires risk taking and therefore accepting the likelihood of some failures. What is depressing is the degree to which some activities have been persisted with even in face of the clear warnings of failure. As the funding of projects involves agreements between the Tanzania government and foreign governments or multilateral funding agencies, each with their own procedures and constraints, and as implementation often involves contractual relationships with commercial firms, there may be limits to the flexibility which can be achieved. This suggests that there are areas into which donors might best not venture, leaving them to NGOs or to commercial enterprises. NGOs are better able to operate effectively in relation to small scale, community based projects. Commercial enterprises are more likely to be able to handle the risks of production for competitive external markets. Flexibility could be increased by leaving project personnel or clients room to modify and adjust in those areas where uncertainties are greatest. For example, crop focussed projects tend to be viewed as failures if farmers decide to grow a quite different crop, although their decision may be an astute response to evolving economic conditions. Projects should adapt to the decisions of actors in the economy, rather than expect a disciplined response to a preconceived plan, itself necessarily based on incomplete economic data. (c) Institutional Sustainability The record is replete with instances of projects which have not delivered the desired results because, in the operational phase, the capacity created has not been utilized effectively and physical facilities have not been maintained. At the beginning of the 1970s, Tanzania's organizational capacity was limited by the scarcity of trained manpower. Nowadays this is much less obviously a general constraint, although there is still a scarcity in some specialties. There is even some evidence of graduate oversupply, with some staff in the public sector being overqualified for the tasks they are called on to perform. The contemporary problem is the poor utilization of avaDlable Tanzanian manpower. This is because of poor incentives and motivation at the level of the individual and weak administrative structures. The tendency has been to employ a growing public service at lower and lower levels of real wages supported bv insufficient, and declining real recurrent inputs. Yet technical assistance is not a productive solution to weaknesses resulting from totally inadequate incentives. Inserting high paid technical assistance staff into a work situation alongside qualified but grossly underpaid local staff is a prescription for frustration on both sides. An alternative donor response is to provide piecemeal and indirect incentives to staff involved in "their" projects, through consulting, seminars, overseas trips, etc. Such piecemeal responses distort the incentive structure further. Efforts to maintain performance by special incentives to project staff can have hidden negative effects in shifting attention from non-project activities in an arbitrary fashion and creating additional anomalies in an already confused incentive structure. Efforts by donors to work in the context of a weak institutional structure by either by passing institutions or creating parallel structures has the risk of demoralizing further already weak state institutions. - 131 - A bold solution would be a comprehensive initiati we for donors to underwrite salaries of key cadres at a reasonable subsistence for the appropriate technical and professional levels. This would mean that certain selected categories of staff, agreed between the government and donors, would receive pay sufficient for them to be able to perform their duties. This would not be a distortion precisely because the tasks performed would be identified as essential, necessitating remuneration at levels sufficient to ensure their proper performance. If, at the same time, the government decides to maintain in employment large numbers of non-essential staff at salary levels determined by the limited availability of revenues, that is no reason to pay needed staff inadequately. In relation to weaknesses in administrative structures, Tanzania experienced in the 1970s what could be described as a series of endogenous shocks. Government decentralization, the abolition of local government institutions and of cooperatives imposed a high pace of institutional change on an already fragile system. The temporary disappearance of local government and the impact of decentralization left gaps in the political/administration structure during a phase in which a great deal of emphasis was placed on the public delivery of rural social services. In the process of reorganization, particular capacities of the government virtually disappeared, such as road maintenance - which had in the past been undertaken using labor intensive methods under the control of district engineers. At the same time, despite the formal commitment to decentralization, a number of activities of government became increasingly centralized, both in the sense that an increasing role was played by central institutions within government and in the sense that activities were increasingly confined to public institutions. E-amples include medical stores and procurement of textbooks. The centralization process was itself fuelled by design of projects which at first sight could much more readily be managed by one centralized government institution. The result has been the development of a system with formally powerfut but practically ineffective central institutions and weak or non-existent local institutions needed for the management and maintenance of field facilities. Institutional sustainability is primarily and ultimately dependent upon the Tanzanian side taking action to improve the operation of its own administrative system. Certain main needs are evident. These include: (i) defining the tasks of government at its various levels narrowly enough to lie within likely performance capacity; (ii) a smaller but better paid public service; (iii) a clear definition of responsibility and accountability related to identifiable performance criteria; and (iv) a more coherent and transparent incentive structure. (d) Finandal Sustainability Institutional and financial sustainability are, of course, interrelated. One root cause of institutional weaknesses in recent years has been the recurrent financial constraint facing most public institutions. Donors and the government failed to take adequate account of the recurrent cost implications of development spending. In the future, this issue should be explicitly and seriously considered in all project design and appraisal. Instead, the ad hoc response to the deep fiscal crisis has been a massive donor support to the recurrent budget by the indirect means of counterpart funds discussed above. - 132 - One way in which funds may be assured is through the generation of funds within the program/project. While not all worthwhile projects can generate revenues, and in some cases attempts to generate revenues may run counter to the objectives of the project, the range of possible user contributions to projects is wide. Even where subsidy is seen as desirable, a service provided against a user charge may be better than no service at all. A strategy to improve sustainability should therefore incorporate a systematic exploration of the potential to generate beneficiary contributions to the recurrent requirements of projects. NraNsz ivauTEs StatisticaL Apperiix Table I TANZANIA--MACROECONONIC BALANCES, 1983-1990 percentage of GDP ...................................... ..................................................................................., ........... 1983 1984 1985 1986 1987 1988 1989 1990 .. ._._..... _.............. .............................. ......................................... 1. Foreign Savings Balance on Current Account excluding Net Official Transfers -4.45 -5.38 -5.44 -8.30 -23.19 -25.43 -29.77 -34.20 2. Private Sector 2.1 Gross Domestic Investment 8.29 10.15 10.09 12.39 16.80 13.65 14.55 15.23 2.1.1 Fixed Investment 5.69 8.30 8.36 10.83 15.12 12.51 13.34 13.96 2.1.2 Change in Stocks 2.60 1.85 1.73 1.56 1.68 1.14 1.21 1.27 2.2 National Savings 9.91 8.81 9.90 11.02 -2.29 -4.00 -0.60 -0.60 2.3 Investment minus Savings -1.61 1.34 0.20 1.37 19.09 17.65 15.15 15.29 3. Public Sector 3.1 Gross Domestic Investment 5.31 5.17 5.63 7.13 6.01 4.55 4.85 5.08 3.1.1 Fixed Investment 5.31 5.17 5.63 7.13 6.01 4.55 4.85 5.08 3.1.2 Change in Stocks 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.Q0 0 3.2 National Savings -1.45 -1.10 -0.52 -2.49 2.92 -1.90 -2.35 -2.30 3.2.1 Current Revenues 19.02 17.62 16.52 14.36 19.42 18.70 20.15 20.55 3.2.2 Current Expenditures 20.47 18.71 17.04 16.86 16.49 20.60 22.50 22.85 3.3 Investment minus Savings 6.76 6.27 6.15 9.62 3.09 6.45 7.20 7.38 4. Public and Private Investment minus Savings 5.14 7.61 6.35 11.00 22.18 24.10 22.35 22.67 Memorandum Item: Share of Gross Domestic Investment financed by Foreign Savings (X) 32.61 34.92 34.13 41.62 95.46 71.57 154.25 168.25 SOURCE: INF, 8oT, BDoS, and Mission Estimates. Statistical Appmndix Table 2 TANZANIA--KEY ECOWOCIC VARIABLES,1 193-1990 ........_:....... . ......... ....... .................................... 1983 1984 1985 1986 1987 1988 1989 1990 Incentive Indicators 1. Real Effective Exchane Rate 1.1 Index (1980-100) 172.00 176.00 205.00 142.00 70.00 55.00 1.2 AnmaL Chuue (2) 11.69 2.33 16.48 -30.73 -50.70 -21.43 2. Real Interest Rates 2.1 Short-Teom Deposit Rate (72.16) (75.83) (80.83) (83.63) (86.27) (89.37) (91.50) 2.2 Short-Tem Lending Rate (69.75) (73.74) (79.40) (82.13) (84.88) (88.27) (90.40) 3. Index of Real vbge 3.1 Nfni.in be (1960=100) 63.47 73.06 49.27 47.69 .. 3.2 Average Wage (1980=100) 60.03 47.66 36.95 29.26 .. 3.3 Nfdle Salary(1980=100) 45.92 43.26 .. .. .. 4.3 Top Salary (1980=100) 57.23 52.36 42.92 28.62 .. .. .. 4. Ratios of Domestic AgriculturaL Prices to International Prices 4.1 Coffee 88.41 87.89 63.86 85.45 102.75 82.28 85.10 80.40 4.2 Cotton 84.00 95.59 101.44 90.47 59.45 103.93 105.10 103.90 4.3 Tea 56.35 61.10 73.69 73.94 86.78 80.06 75.40 84.00 4.4 Tobacco 96.53 97.79 90.62 92.95 78.43 75.72 85.20 94.00 External Trade Indicators 5. Volume Index of Major Exports 5.1 Coffee 116.00 126.00 101.00 116.00 97.00 89.00 128.07 138.70 5.2 Cotton 120.00 92.00 70.00 99.00 142.00 164.00 180.56 180.93 5.3 Tea 125.00 84.00 88.00 72.00 90.00 84.00 66.70 74.17 5.4 Tob aco 150.00 132.00 219.00 205.00 234.00 279.00 317.50 374.02 6. Export Shares In Worid Trade 0.02 0.02 0.01 0.02 0.01 0.01 0.01 T. Mamfectured Exports 7.1 ReaL Growth Rate (2 p.a.) (22.85) 18.50 (7.85) 11.15 (1.43) 8.45 -12.00 4.00 7.2 VaLue as Share of Total Exports ( 11.61 8.51 10.09 10.66 17.85 19.50 16.95 15.45 8. Camodity Tem of Trade 8.1 Index (1980 u 100) 90.96 96.21 90.46 103.69 89.35 94.26 95.20 71.82 8.2 Amual change (2) 3.16 5.77 (5.97) 14.63 (13.83) 5.49 1.00 -24.56 ........................................................................................................... 50UR: IF$, SOT, wd NHssfon Estf_tes. - 136 - Statisticat Appendix Table 3 Table : Balance of Payments, 1984-1990 (in mitlions of US$) _______._._ ._.__........... . ...... ......... ..... ...... ...... ........... ... ___................... ............. 1984 1985 1986 1987 1988 1989 1990 _. .... ...... ... .... ... ....__. .. __.................. ..... ........ .. ............................ ......... Exports of Goods & NFS 478.67 430.29 445.91 441.59 485.50 518.00 600.90 loports of Goods & NFS 819.83 936.20 992.20 1,207.09 1,255.10 1,371.00 1550.00 (of which Petroleum) 211.00 224.00 145.00 170.00 293.60 305.00 239.90 Resource Gap (deficit a .) (341.16) (505.91) (546.29) (765.50) (769.60) (853.00) (949.00) Interest Payments (net) 30.60 23.30 29.90 38.30 41.00 65.00 55.00 Other Factor Payments (net) 2.99 77.15 79.88 174.05 176.30 171.37 186.00 MNt Private Transfers 62.12 233.38 249.89 230.00 235.00 226.80 293.70 Balance on Curr. Acet. excl. Not Official Transfers (310.99) (370.46) (396.70) (742.35) (744.30) (849.53) (892.30) Net Official Transfers 96.94 223.02 223.54 477.08 508.00 523.45 539.00 Balance of Curr. Acct. ncl. Met Official Transfers (214.05) (147.44) (173.16) (265.27) (236.30) (326.08) (353.00) Net NLT Borrouing 191.30 133.80 164.30 73.60 144.60 217.00 208.90 Disbursements 225.60 172.70 198.70 111.70 191.30 267.00 248.90 A4ortization 34.30 38.80 34.40 38.10 46.70 50.00 40.00 Subtotalt. . .... .. Other Capital (net) and Capital n.i.e. 63.57 5.83 72.06 91.07 136.90 141.08 160.40 Increase In Reserves (M) (40.82) 7.81 (63.21) 101.60 (45.90) (32.00) (39.00) Gross Reserves (end year) 26.87 16.00 61.09 31.80 77.71 54.20 93.00 SOURCE: IMF, SOT, and Mission Estimates on Accrual Basis. - 137 - Stt.istical Appendix Table 4 MERCHANDISE EXPORTS (AVERAGE 1985-90) Value (mitlions of USS) X of Total ......... . ......... ... .. .... .... ........ ................ ................ Coffee 120.27 32.14 Cotton 62.05 16.58 Tea 15.93 4.26 Tobacco 16.47 4.40 Cashuwnuts 13.52 3.61 Sisal 5.17 1.38 Manufactured Goods 57.68 15.40 Other Exports 83.11 22.21 Merchandise Exports 374.22 100.00 SOURCE: Bank of Tanzania and Mission Estimates. - 138 - StatisticaL APPenIx Table 5 TAZANIA-Debt Indlcators (NilLtons of USS) .............................................................................. .................. ............................................... .... External Debt, Deceber 31, 1989 Public Debt, nct. Guaranteed 4,760.00 Hon-Guaranteed Private Debt 8.50 Total Outstwnding & Disbursed 5.4S2.00 ............... ................................. 0............. Percenta Not Debt Service Ratio for 1989 Publfc Debt 1ncl. Guaranteed 21.40 Nan-Guaranted Private Debt 0.20 Total Outstnding & Disbursd 24.SO ........................................................................................................................ InRS IDA NSRD/IDA LediWn (12/31/89) autaandng & Disbursed 251.50 1,016.48 Undisbursed 0.00 306.37 Total Outstanding in. Undisbursed 251.S0 1.322.8S ..................................................... O.............. .......... WURCE: Debtor Reporting Syste", IECDI - 139 - Statitical Appendix Table 6 TANZANIA DATA SHEET .... ............. ................ .......................................... ................ 1989 .. ...... ........... .... .... .............................. General Area (thous. sq.km) 945.00 Population (millions) 25.63 Growth Rate (1978) 3.60 Density (per sq. km) 25.00 Social Indicators Population Characteristics Crude Birth Rate (per 1,000) 47.70 Crude Death Rate (per 1.000) 13.10 Health Infant Mortality (per 1,000 live births 102.70 Population per Physician 26,545.00 Population per Hospital Bed m7.00 income Distribution (Z of national income) Lowest Quintile(15-20 years ago) 6.00 Highest auintile(15-20 year qgo) 50.00 Distribution of Land Ownersh i Z Owmed by Top 101 of Owners Z Owmed by Smallest 10X Access to Safe Water Z of Urban Population 90.00 i of Rural Population 42.00 Access to Electricity X of Urban Population X of Rural Population Nutrition Calorie Intake as X of Requirements Per Capita Protein Intake (g/day) 52.00 Education Adult Literacy Rate (X)(1978) 46.30 Primary School Enrollment (X of relevant age group) 66.00 ...................................... .. ...................... SOURCE: Sociatl Indicetors of Development, 1990 IECSE. Statistical Appedix Table 7 GOP BY ItDUSTRIAL ORIGtN Pae 1 of 2 . _....... .............. _.............................. ..................................................................... .......... ....................................... .................................. ............ 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 ... ... .... _..... ......... .............................................................................. ................................................................................................................. ...... Official Accounts (millions of current Tsh) .__. ..... .. ............................ Agriculture, etc. 9,066 11,131 12.506 14,728 16,636 20,338 26.449 32.737 41,295 61,231 84,153 117,982 178.760 207,059 233,64 Ninin 214 243 228 284 329 299 266 249 337 251 474 64S 723 1,129 4.815 Uufactwiflg 2,811 3,287 3,859 3,868 4,097 4,501 4,361 4,869 5,932 6.665 8,551 14,7M2 24,453 30,353 20,436 Elec. & water 219 254 261 275 424 423 421 514 551 1,071 1,488 2,963 4,103 4,a31 5,068 Corstruction 884 1,111 1,052 1,229 1,498 1,614 1.863 1.252 1,661 2,061 3,131 3,543 4,800 5,904 12,650 Trade/hoteLt/rest. 2,839 3,407 3,889 4,344 4,713 5,479 6.814 8,148 10.447 14,195 19,476 25.963 43,800 53,572 57,717 Tr. sport/comm. 1,685 1,793 1,917 2,113 3,019 3,133 3,395 3,507 4,789 7,021 7,797 11,815 15,621 23,345 38,242 Finance, etc. 2,036 2.419 2,686 2,978 3,744 4.507 4,891 5,252 6,028 6,659 8,127 11,061 14,132 20.641 24,123 Pub adin,etc 2,342 2,596 2,873 3,342 3,959 4,732 5,446 7,372 8,614 10,735 10,213 13,291 17,163 22,437 31,968 Ilpted bank dhg. (424) (543) (689) (844) (965) (1,120) (1,360) (1,292) (1,511) (1,806) (2,544) (6,644)(12,U8)(18,043)(25,260) GOP at f.c. 21,652 25,698 28,582 32,317 37,454 43,906 52,546 62,608 78,143 108.083 140,866 195,611 290,667 351,228 401,583 GOP at inp. 24,419 28,868 32,169 36,285 42,118 49,102 58,226 70,509 88,892 120,621 159.668 221,678 327,686 404.155 541,008 Official: Percentage Share .. ..................... Agriculture, etc. 41.8 43.3 43.8 45.6 44.4 4.3 50.3 52.3 52.8 56.7 59.7 60.3 61.5 59.0 58.2 Nfning 1.0 0.9 0.8 0.9 0.9 0.7 0.5 0.4 0.4 0.2 0.3 0.3 0.2 0.3 1.2 Nmuwfacturin 13.0 12.8 13.5 12.0 10.9 10.3 8.3 7.8 7.6 6.2 6.1 7.6 8.4 8.6 5.1 Elec. & water 1.0 1.0 0.9 0.9 1.1 1.0 0.8 0.8 0.7 1.0 1.1 1.5 1.4 1.4 1.3 Construction 4.1 4.3 3.7 3.8 4.0 3.7 3.5 2.0 2.1 1.9 2.2 1.8 1.7 1.? 3.2 Trademhotels/rest. 13.1 13.3 13.6 13.4 12.6 12.5 13.0 13.0 13.4 13.1 13.8 13.3 15.1 15.3 14.4 Transport/comm. 7.8 7.0 6.7 6.5 8.1 7.1 6.5 5.6 6.1 6.5 5.5 6.0 5.4 6.6 9.5 Finance, etc. 9.4 9.4 9.4 9.2 10.0 10.3 9.3 8.4 7.7 6.2 5.8 5.7 4.9 5.9 6.0 Pub adain,etc 10.8 10.1 10.1 10.3 10.6 10.8 10.4 11.8 11.0 9.9 7.3 6.8 5.9 6.4 8.0 Iaputed bank chg. -2.0 -2.1 -2.4 -2.6 -2.6 -2.6 -2.6 -2.1 -1.9 -1.7 -1.8 -3.3 -4.4 -5.1 -6.3 6DP at f.c. 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 GDP br Industrial Original (millions of 1976 Tsh) ..~~ ~~ ............. ... . . . . . . . . . . . . . . Agriculture, etc. 9,046 9,150 8,998 9,066 9,418 9,511 9,639 9,914 10,312 10,931 11.557 12.066 12,606 13,183 13,567 Nining 214 231 189 200 189 193 193 174 186 174 154 149 138 139 165 Nwwfacturing 2,811 2,641 2,730 2,821 2,683 2,382 2,304 2,103 2,159 2,075 1,991 2,075 2,187 2,399 2,587 Elec. & water 219 244 286 318 400 417 420 413 439 461 544 584 574 506 512 Construction 884 915 783 879 932 890 930 549 660 601 705 73 780 858 937 Trade/hoteLs/rest. 2,839 2,782 2,797 2,839 2,839 2,725 2,668 2,612 2,640 2,662 2,958 3,112 3,225 3,549 3,747 Transpot/comun. 1,685 1,652 1,699 1,634 1,818 1,652 1,694 1,473 1,482 1,509 1,504 1,551 1,652 1,663 1,736 Finance, etc. 2,036 2,089 2,208 2,338 2,483 2,529 2,702 2,817 2,984 3,046 3,318 3,395 3,500 3,554 3,630 Pub admin,etc 2,342 2,497 2,997 3,255 3,188 3,551 3,556 3,543 3,549 3,616 3,225 3,243 3,343 3,475 3,553 Ilputed bank chg. (424) (462) (485) (501) (531) (549) (667) (716) (755) (797) (886) (862) (920) (950) (1,038) GDP at f.c. 21,652 21,739 22,202 22,849 23,419 23,301 23,439 22,882 23,656 24,278 25,070 26,049 27,085 28,376 29,396 GDP at m.p. 24,419 24,421 24,988 25,653 26,335 26,059 25,973 25,770 26,910 27,094 28,357 29,480 30,996 32,361 33.7Z0 Statistical Appendix Page 2 of 2 GDP BY INDUSTRIAL ORIGIN GRO4TH RATES Official Accounts .................. Agriculture, etc. 1.1 -1.7 0.8 3.9 1.0 1.3 2.9 4.0 6.0 5.7 4.4 4.5 i 2.9 Nining 7.9 -18.2 5.8 -5.5 2.1 0.0 -9.8 6.9 -6.5 -11.5 -3.2 -7.4 0.7 18.7 Manufacturing -6.0 3.4 3.3 -4.9 -11.2 -3.3 -8.7 2.7 -3.9 -4.0 4.2 5.4 9.7 7.8 Elec. & water 11.4 17.2 11.2 25.8 4.3 0.7 -1.7 6.3 5.0 18.0 7.4 -1.7 -11.8 1.2 Construction 3.5 -14.4 12.3 6.0 -4.5 4.5 -41.0 20.2 -8.9 17.3 4.4 6.0 10.0 9.2 Trade/hotelslrest. -2.0 0.5 1.5 0.0 -4.0 -2.1 -2.1 1.1 0.8 11.1 5.2 3.6 10.0 5.6 Transport/comnun. -2.0 2.8 -3.8 11.3 -9.1 2.5 -13.0 0.6 1.8 -0.3 3.1 6.5 0.7 4.4 Finance, etc. 2.6 5.7 5.9 6.2 1.9 6.8 4.3 5.9 2.1 8.9 2.3 3.1 1.5 2.1 Pub admin,etc 6.6 20.0 8.6 -2.1 11.4 0.1 -0.4 0.2 1.9 -10.8 0.6 3.1 3.9 2.2 Imputed bank chg. GDP at f.c. 0.4 2.1 2.9 2.5 -0.5 0.6 -2.4 3.4 2.6 3.3 3.9 4.0 4.8 3.6 GDP at m.p. 0.0 2.3 2.7 2.7 -1.0 -0.3 -0.8 4.4 0.7 4.7 4.0 5.1 4.4 4.2 DEFLATORS Official Accounts Agriculture, etc. 100 122 139 162 177 214 274 330 400 560 728 978 1418 1571 1723 Mining 100 105 121 142 174 155 138 143 181 144 308 433 524 812 2918 Manufacturing 100 124 141 137 153 189 189 232 275 321 429 713 1118 1265 790 Elec. & water 100 104 91 86 106 101 100 124 126 232 274 507 715 955 994 1 Construction 100 121 134 140 161 181 200 228 252 343 444 481 615 688 1350 ,- Trade/hotels/rest. 100 122 139 153 166 201 255 312 396 533 658 834 1358 1509 1540 41 Transport/coamum. 100 109 113 129 166 190 200 238 323 465 518 762 946 1404 2203 Finance, etc. 100 116 122 127 151 178 181 186 202 219 245 326 404 581 665 Pub admin,etc 100 104 96 103 124 133 153 208 243 297 317 410 513 646 900 Ioputed bank chg. 100 118 142 168 182 204 204 180 200 227 287 748 1401 1899 2434 GDP at f.c. 100 118 129 141 160 188 224 274 330 445 562 751 1073 1238 1366 GDP at m.p. 100 118 129 141 160 188 224 274 330 445 563 752 1057 1249 1604 Statistical Appendix Tabte 8 GDP BY EXPENDITURE ....... ............ . .......................................... ...................................................................................... ................... ................................................................. 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 OFFICIAL ESTIMATES GDP BY EXPENDITURE (miLLions of current Tsh) Private cansLuption 15.377 17,979 23,363 25,497 32,486 37,035 42.261 55,128 68,652 93,130 104.934 195,928 312,816 385,667 Gen'Lt gov't. cons. 3.989 4.308 5,585 5.956 5.494 6,105 8,046 9,443 13,844 18,555 23.621 25,433 32,663 42,189 Exports gnfs 5,297 5,627 4,692 5,131 5,540 5,994 4,546 5,111 6,321 7,453 14,580 29,943 49,882 74,545 lxports gnfs 5.841 6,570 9,565 9,759 11,087 10.162 8,862 8,761 13,543 17,480 37,026 79.592 132,551 197,998 Gross doa. inv. 5,597 7,524 8,094 9,458 9,685 10.130 12,235 9,588 13,618 18,963 31,166 49,966 64,876 99,752 Gross fixed imv. 5,159 6,663 7,330 8,592 8.630 8.632 10,825 7.752 11,973 16,872 28.679 46,281 60,754 94,270 Increase in stocks 438 861 764 866 1,055 1.498 1,410 1.836 1,645 2,091 2,487 3,685 4,122 5,482 GDP at mp. 24,419 28,868 32,169 36,285 42,118 49.102 58,226 70,509 88,892 120,621 159,648 221,678 327,686 404,155 Net factor income (100) (88) (45) (71) (112) (176) (231) (211) (173) (704) (3,281) (11,038) (18,542) (29,377) GNP at mp 24,319 28,780 32,124 36,214 42,006 48.926 57,995 70.298 88,719 119,917 156,367 210,640 309,164 374,778 SHARES OF GOP--OFFICIAL Private consumption 63.0 62.3 72.6 70.3 77.1 75.4 72.6 78.2 77.2 77.2 65.7 88.4 95.5 95.4 Gen't gov't. cons. 16.3 14.9 17.4 16.4 13.0 12.4 13.8 13.4 15.6 15.4 14.8 11.5 10.0 10.4 Exports gnfs 21.7 19.5 14.6 14.1 13.2 12.2 7.8 7.2 7.1 6.2 9.1 13.5 15.2 18.4 1 Imports gnfs 23.9 22.8 29.7 26.9 26.3 20.7 15.2 12.4 15.2 14.5 23.2 35.9 40.5 49.0 Gross doa. inv. 22.9 26.1 25.2 26.1 23.0 20.6 21.0 13.6 15.3 15.7 19.5 22.5 19.8 24.7 Gross fixed inv. 21.1 23.1 22.8 23.7 20.5 17.6 18.6 11.0 13.5 14.0 18.0 20.9 18.5 23.3 Increase in stocks 1.8 3.0 2.4 2.4 2.5 3.1 2.4 2.6 1.9 1.7 1.6 1.7 1.3 1.4 GDP at up. 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Net factor income -0.4 -0.3 -0.1 -0.2 -0.3 -0.4 -0.4 -0.3 -0.2 -0.6 -2.1 -5.0 -5.7 -7.3 GNP at up 99.6 99.7 99.9 99.8 99.7 99.6 99.S 99.7 99.8 99.4 97.9 95.0 94.3 92.7 BIBUIOGRAPH Azad, Vargha. 1990. "Tanzania: Balance of Payments and the Real Exchange Rate". Mimeo. Background paper for this report. Bagachwa, M.S.D., N.E. Luvanga, and G.D. Jema. 1990. "A Study on Non-Traditional Exports". Mimeo. University of Dar-es-Salaam. Background paper for this report. Bank of Tanzania. Annual Reports, various years. Bevan, D., A. Birgsten, P. Collier and J.W. Gunning. 1987. "East African Lessons on Economic Liberalization". Thames Essays, no. 48, Trade Policy Research Centre, London. Bevan, David, Paul Collier, and Jan W. Gunning. 1989. Peasants and Governments: An Economic Analysis. Oxford: Clarendon Press. Bevan, David, Paul Collier and Jan W. Gunning. 1990. "Price Controls and the Transition to Market Clearing: Theory and an Application to Tanzania". Central Bureau of Statistics. Foreign Trade Statistics. 1987. Government of Tanzania. Collier, Paul. 1988. "Aid and Economic Performance in Tanzania". Mimeo. Collier, Paul and Jan W. Gunning. 1990. "The Tanzania's Recovery, 1983-89", Background paper for this report. Collier, Paul, Samir Radwan and Samuel Wangwe. 1986. "Labor and Poverty in Rural Tanzania: Uiamaa and Rural Development in the United Republic of Tanzania". Clarendon Press, Oxford. Cooksey, B. 1989. "Incentive Goods in Rural Tanzania 1986-89". Background paper for this report. Coulson, Andrew. 1982. Tanzania: A Political Economy. Clarendon Press, Oxford. Economic Research Bureau: Tanzania Economic Trends, various issues. University of Dar es Salaam, in collaboration with the Planning Commission. Galabawa J.C.J. and A.G.M. Ishumi. 1990. "Self-Help Secondary Schools: A Case Study of the Kanyigo Development Association (KADEA) Initiative." Background paper for this report. Gordon, Henry. 1989. "An Overview of Basic Data on Grain Marketing in Tanzania, 1974/75 to the Present". Background paper for this report. 144 Hyuha, M. and B.J Ndulu. 1990 "Inflation and Economic Recovery in Tanzania: Further Theoretical and Empirical Analysis". Mimeo. Jones, Leroy P. and I. Sakong. 1980. Government. Business. and Entrepreneurship in Economic Development: The Korean Case. Cambridge, Mass.: Council on East Asian Studies, Harvard University: distributed by Harvard University Press, 1980. Kaufmann D., and S.A. O'Connell. 1990. "The Macroeconomic of the Unofficial Foreign Exchange Market in Tanzania". Mimeo. Kavishe F.P. 1987. "The Food and Nutritional Situation in Tanzania 1987". Tanzania Food and Nutrition Center Report No. 1215. Kavishe, F.P. 1990. "Access to Free Health Services in Tanzania. Tanzania Food and Nutrition Centre". Background paper for this report. Kilby, Peter. 1971. Entrepreneurship and Economic Development. Free Press, New York. Lipumba N.H.I. 1988. "Reflections on Long-Term Development Strategy in Tanzania". Department of Economics, University of Dar-es-Salaam. Little, Ian, David Malcolm, Dipak Mazumdar and John M. Page. 1987. Small Manufacturing Enterprises: A Comparative Analysis of India and other Economies. Oxford University Press, 1987. Maliyamkono, T. and M.S.D. Bagachwa. 1990. The Second Economy in Tanzania. Oxford, Oxford University Press. Mang'yena, Rose. 1990. "Issues in Public Sector Management: Institutional Reform". Background paper for this report. Mbelle, A.V.Y., and L. Karamagi. 1990. "Review of Foreign Exchange Management Issues in Tanzania". Background paper for this report. Mbilinyi, M. 1990. "A Review of Women in Development Issues in Tanzania". Background paper for this report. Mshomba R.E. 1989. "Price Elasticity of Supply of Tanzania's Major Export Crops". Graduate student working paper series, paper no. 13. Mtatifikolo, Fidelis P. 1989. "Issues Concerning Budgetary Subsidies to Parastatals and Budgetary Constraints on Social Sector Spending in Tanzania". Background paper for this report. Mushi, R. and S. Wangwe. 1990. "A Review of Investment in the Public and Private Sectors in Tanzania". Background paper for this report. 145 N'cho-Oguie, C. 1990. "The Basic RMSM-S Framework and its Implication for Tanzania". World Bank mimeo. Ndanshau, M.O. and M. Hyuha. 1990. "Economic Development and Adjustment: Policies After the First ERP". Department of Economics, University of Dar-es-Salaam. Ndulu, B. and M. Hyuha. 1989. "Inflation and Economic Recovery in Tanzania: Some Empirical Evidence", Uchumi, Volume 2, No. 1, University of Dar es Salaam. Ndulu, B., N.H.I. Lipumba, S. Horton and A. Plourde. 1987. "A Supply Constrained Macroeconometric Model of Tanzania". Mimeo. Ndulu B.J. and J.J. Somboja. 1990. "Trade and Industrialization in Tanzania: A Review of Experience and Issues". Mimeo. Nsekela, Amon J. 1987. "Organization of the Government: Report of the Salaries review Commission" Government of Tanzania. Rwegasira, D.G. 1987. "Balance of Payments Adjustment in Low-income Developing Countries: The experiences of Kenya and Tanzania in the 1970s". World Development, Vol 15, No. 10/11. pp. 1321 - 1335. Tanzania Audit Corporation, Annual Reports, Various issues. World Bank: Parastatals in Tanzania: "Toward a Reform Program, 1988 Report 7100-TA. World Bank: Social Indicators of Development. 1990. International Economics Department. World Bank: Tanzania: AnAgend.for Industrial Recovery. 1987. Report No. 6357-TA. World Bank: Tanzania: Public Expenditure Review. 1989. Report No. 7559-TA. World Bank: Tanzania: Women and Development. 1990. Report No. 9108-TA. World Bank: World Development Report. Various issues.