Report No. 201-KE The Second Decade: iv A Basic Economic Report on Kenya Annex 1 The Macro-economic Model and Projections (Vol. 11 of Five Volumes) January 15, 1974 Not For Public Use Eastern Africa Regional Office Document of the International Bank for Reconstruction and Development International Development Association This report was prepared for official use only by the Bank Group. it may not be published, quoted or cited without Bank Group authorization. The Bank Group does not accept responsibility for the accuracy or completeness of the report. UNITS OF VALUATICN The official unit of currency in Kenya is the Kenya Shilling (Sh.) However, in accordance with the practice of the Kenya Government, most large values in the report are expressed in Kenya Pounds (L) , 1 = Sh. 20 Sh 1 = 100 cents Some values have been expressed in terms of constant US dollars for pur- poses of international comparison. CURRENCY EQUIVALENTS From Independence at the end of 1963 until March, 1973, the exchange rate between the Kenya Shilling and the US dollar was retained at $1 = Sh.7.143. This is the exchange rate used throughout the report. Since June 30, 1973, the Kenya Shilling, together with those of Tanzania and Uganda, has been set at a central rate of $1 = Sh.6.9, and all three countries have availed themselves of the wide margins of up to 2¼4 per cent. Exchange Rate Used in the Report Present Rate of Exchange US dollar Sh. 7.143 Sh. 6.9 Kenya Pound = $ 2.80 $2.8985 THE MISSION This report is based on the findings of an IBRD mission which visited Kenya in March/April, 1973. The main mission consisted of the following Bank staff: John Burrows - Chief of Mission Ramgopal Agarwala - Macro-Economist George Reier - General Economist (Project Planning and External Assistance) Ved Gandhi - Fiscal Economist Randolph Harris - General Economist (Public Services) Martin Wolf - General Economist (Private Sector) Sven Burmester (Education), Andrew Ilayman (Tourism) and Frank Stubenitskv (Health) also participated in the work of the Mission and have contributed to the report. Lyle Hansen was adviser to the Mission. A preliminary report was prepared in August and discussed with the Government of Kenya during October 1973. The present report incorporates the comments of Government, and where possible, includes more recent material. PREFACE 1. During the course of 1972, the World Bank decided to embark upon a series of "basic" economic reports on its major member countries. The nature of country economic work has been under review for some time in the Bank, and the decision to undertake these major reviews on a regular basis reflects the general desire to improve both the quality and the usefulness of this work. The basic economic report is intended to provide a periodic overview of the operations of an economy. From the Bank's noint of view, these reports are intended to provide a new perspective of the longer term structural developments in an economy, to assess the extent to which they can be shaped bv policy changes, and to identify the country's external assistance requirements. But more than this, a basic economic report is expected to provide a synoptic vriew of the many facets of the economy, and thus to bring into focus other work being undertaken by the Bank and else- where at a sectoral or project lavel. From the country's point of view, it is hoped that these policy-oriented reports will be valuable in giving ob- jective and possibly new insights into the dynamics of the economy and the options which may be open to the Government in the future management of the economy. 2. This is the first basic economic report on Kenya. The timing is particularly appropriate as Kenya prepares to enter the second decade of Independence and is about to publish the Third National Development Plan. We feel that it is therefore a suitable time to assess how far Kenya has come during the past.ten years, to review her major successes and failures, to assess what prospects lie ahead, and to identify major policy issues. This is the main purpose of the report. Of course, this is not the first Bank report to undertake this task, but the latest in a series. For exam- ple, the Bank published a report on the Kenya economy in 1963 which reviewed the development prospects of the country as it moved towards Independence. In 1967, a major Bank mission reviewed the revised development plan (1966- 70) and again in 1969 another mission reviewed the second (1970-74) plan. Each of these missions and the subsequent reports differed in composition and scope, but all served to make a critical review of Kenya's national plans and to offer constructive comments. At the request of the Government, both the 1969 mission and the recent 1973 mission visited Kenya while the new plan was still in draft form, so that the comments of the mission could be taken into account before the plans were published. 3. A report of this nature must essentially be the result of a com- promise between comprehensiveness and brevity. The Kenya economy is much too broad and its operations muclh too complex to allow for complete cover- age, even in a "basic" report. We have therefore deliberately circumscribed the scope of the report in a number of ways which it is important to make clear at the outset. First of all, the report is intended to be a review of the operations of the Kenya economy only, and makes no attempt to review progress or prospects of the wider region to which the Kenyan economy belongs, or even to assess in any comprehensive way how Kenya's development prospects are affected by her membership of the East African Community. Some of these relationships are referred to when they are of particular relevance, but the report has not tried to view the Kenyan economy from an integrated regional perspective. This limitation doeXs not in any sense mean that either the Mission or the Bank feels that regional economic considerations are unimportant. On the contrary, it is clear from its major financial commitment to the EAC - ii - corporations and the development bank that the Bank fully supports this unique development in regional cooperation which Kenya, Tanzania and Uganda have pioneered. The report focuses on Kenva and ignores the wider Community simply to keen the scone of the report within manageable bounds. This narrow focus becomes seriously mvopic only in those sections of the report (on trade policies for instance) where Kenya must clearly act in concert with her partners in the Community. Again, while we try to suggest what options might be best for Kenya, viewed in isolation, we are verv conscious that these options will have to be reviewed by all three Partner States, and that the decisions will ultimately be taken with the interests of the wihole region in mind. 4. The scope of the mission was circumscribed in a second major re- spect. Even in its focus on Kenya, the report will not undertake a detailed review of all sectors of the economv and of all economic problems. The eco- nomic literature on Kenya is prolific, and we have drawn heavily on this. In particular, the recent ILO/UNDP Report on Employment, Incomes and Equal- ity in Kenya has presented a very comprehensive and innovative analysis of unemployment and poverty, and we make no attempt to go over this ground again. Rather, we see this report, with its broader macro-economic focus, as being essentially complementary to the ILO/UNDP Report. We have not attempted to add in any significant way to the existing knowledge on the various sectors; instead, we have tried to consolidate and integrate this knowledge into our overall understanding of the operation of the economy. Similarly, we have not placed great emphasis on reviewing progress under the Second Plan or on describing the objectives of the Third Plan, because these tasks have been done very well by the Government itself. 5. Our report does not therefore try to deal with everything in great denth. On the contrary, it draws heavily from the wide range of studies al- ready available and tries to use this information to provide a synoptic view of the way in which the economy as a whole functions and perhaps some new in- sights into important relationships between variables. Thus, while the re- port tries to be as informative as Dossible and to present sufficient back- ground data on most aspects of the economy for the general reader, the de- tailed analvsis is highly selective and focuses mainly on a number of key issues which we see as critical to the future development of Kenya and the well-being of its people. 6. The report is divided into five volumes. The main report traces the major developments in Kenya's first decade of independence, identifies the emerging issues, and examines the major options open to the Government in the future, as the Mission sees them. The remaining four volumes contain the analytical annexes, which discuss the major issues in detail and extend the technical arguments. 7. An outline of the complete report is shown on the opposite page, and a select bibliography of some of the major sources of information on Kenya is given at the end of this volume. THE SECOND DECADE A BASIC EC'ONOMIC REPORT ON KENYA CONTENTS VOL. I THE MAIN REPORT: EMERGING ISSUES AND POLICY OPTIONS VOLS. I-V THE ANALYTICAL ANNEXES 1. The Macro-Economic Model and Projections 2. Fiscal Policy for Development . 3. Key Issues in the Private Sector 4. Domestic Savings and Financial Intermediation 5. Priorities for Planning and Project Design 6. Priorities for External Assistance ANNEX i THE MACRO-ECONOMIC MODEL AND PROJECTIONS CONTENTS Page No. PART I: THE MODEL AND THE RESILTS: A NON-TECHNICAL PRESENTATION Chapter 1 - On the Nature and UJses of the Model I Chapter 2 - The Structure of the Model 3 Chapter 3 - Statistical Analysis of the Past and Working Hypotheses for the Future 9 Developments over the Period 1964-71 9 Policy Variables in the Model 15 Basic Hypotheses Used in the Model 20 Chapter 4 - Projections and Policy Analysis 24 The Basic Scenario 24 Sensitivity of the Basic Scenario to the Critical Assumptions 26 Some Apparent Policy Alternatives 27 Some Real Policy Alternatives 29 Chapter 5 - On Some Development Policy Issues 32 PART II: THE TECHNICAL APPENDICES Appendix 1 - A Simplified Algebraic Description Appendix 2 - An Analysis of ICO:Rs in Kenya Appendix 3 - Analysis of Import Requirements Incorporating the Effects of GOF Composition Appendix 4 - Analysis of Savings of Household, Business and Government Sector Appendix 5 - Factor Prices, EmpLoyment and Growth Appendix 6 - Money, Prices and t:he Equalization of the Two Gaps Appendix 7 - Miscellaneous Equat:ions in the Model Appendix 8 - Some Thoughts on Further Research STATISTICAL TABLES AND CHARTS PART I THE MODEL AND THE RESULTS: A NON-TECHNICAL PRESENTATION CHAPTER 1. ON THE NATURE AND USES OF THE MODEL 1.01 In recent years, compulterized econometric models have become an im- portant tool of economic managerment in the hands of national economic managers as well as corporate management.. However, there is still a widespread misun- derstanding about the nature and use of such models (leading to either abuse or neglect of the tool), and it therefore seems worthwhile to begin by clari- fying our approach. 1.02 We regard a computerized model primarily as a computational device. In other words, given the best guesses about the basic relationships in an economy and working hypotheses about the magnitudes of various critical para- meters, the model works out their implications for different sectors of the economy and over a given time horizon. In formulating the "best guesses" and "working hypotheses", economfetric techniques play only a subsidiary role. These are based, apart from statistical regressions, on economic theory, on the study of economic history of the country and, above all, on the Judgments of the persons involved with the management of the economy. For example, t;hc model presented here makes some working hypotheses about the effects of rela- tive prices of capital and labor on employment prospects and capital require- ments within different sectors. It also makes some assumptions about the elasticity of imports with respect to import prices. These assumptions are not made by rigorous econometric investigations in the Kenyan economy (because relevant data are not available) but are presented as working hypotheses on the basis of experience in other countries and scattered evidence in Kenya. Quite often, parameters are put not at "most likely" levels, but at "optimistic" or 'pessimistic" levels, in order to work out the implications of alternative sets of conditions. Thus, the projections in the model are not to be regarded as 'predicting the future"; the purpose is not to foresee the future but to influence it in the desired directions by applying appropriate policy changes now. 1.03 It may be asked, "What is the use of a model, if it does not fore- see future and does not, on its own, test alternative hypotheses about the effects of policies?" Admittedly, this leaves for a model only a humble role; but, we believe, the correct roLe and a useful role. For example, it is ob- vious that if a country is running balance of payments deficits and borrowing abroad to fill the gap, it is accumulating a burden of debt servicing for the future. No model is necessary to tell us that. However, it is not intuitively obvious, but important to know, when the burden of debt will become intolerable (however defined) and how this could be ameliorated by changing the terms of borrowing. This is extremely laborious to work out by hand, and a computerized model can be helpful for this. 1.04 Similarly, if one accepts that abolition of "investment allowances" will improve employment prospects (as we in fact suggest in this Report), it is useful to know how much difference it would make to the overall prospects of employment, say, over the next 10 years. Again, if one suggests a list - 2 - Dt- policies for tackling, say, the unemployment problem, the question arises now serious the probiuem will remain if only parts of the package are accepted. The model by trying to quantify the broad orders of magnitudes helps in get- ting a feel for these figures. 1/ 1.05 Another important use of ali econometric model is to provide a frame- work, a skeleton, on which to hang the assumptiors about different sectors of a development plan and ensure tneir consistency. Even the purely clerical 4ob of keeping track of differernt sets of data of past and expected future - on national accounts, imports, exports, taxes, public expenditures, domestic credit, wages and prices--can be immensely simplified by computerized models. However, as emphasized before, these useful purposes can be served by the model only when it is used in an alert and flexible manner responding quickly to thne changing envirorLnent and changing policy issues, and the limitations of the data used in the model must always be kept in mind. 1/ A model, of course, does no more than work out the implications of the assumptions made by the user. However, quite often the process leads to some surprising conclusions. In th-s particular exercise, for example, we started with a supposition that factor price changes would prove to be a powerful instrument for influencing employment and poverty. Yet, on the assumptions used in the model, we found that the impact of factor price changes alone on employment and poverty was only marginal. Similarly, while we found that changes in the exchange rate could be a useful tool of policy for certain purposes, we were surprised by the inflationary implications shown in the model, and had to reduce the role of exchange rate policy in the final strategy. While some of these results were surprising at first sight, a closer examination of the interrelationships involved showed them to be creditable and helped to throw further light on the operations of the economy. For example, since factor price changes will primarily affect employment in the modern sector (which is a small proportion of the total labor force), it is not difficult to believe that their overall impact on employnaerit will be relatively small. Similarly, it could be exlpected that devaluation would be a very appropriate policy tool when che foreign exchange gap is much larger than the domestic resource gap. But in our projections, the difference between the two "gaps" was not very large, so that the usefulness of exchange rate manipulation in reducing the external gap was li-mi&ed by .he scarcity of domestic resources and the consequent inflationary ?-pessure on prices. - 3 - CHAPTER 2. THE STRUCTURE OF THE MODEL 2.01 The prime purpose of our model is to assess whether the past rate and pattern of growth in Kenya can be sustained in the future, within the resources which are likely to be available, and if not, whether a different set of policies could help in attaining the targets. Another important ob- jective is to analyze the employment prospects over the next decade and to examine policies that might be helpful in tackling unemployment. We hope that this kind of analysis can be of value in identifying the nature and sensi- tivity of the policy decisions which have to be made in Kenya during the Third Five-Year Plan and afterwards and in assessing Kenya's external aid requirements. 2.02 For assessing the resources required to achieve the target growth rates, we use a modified version of the celebrated "two-gap" approach which has become a standard tool of anLalysis in country plans as well as in inter- national aid agencies. Broadly speaking, the approach accepts that at an early stage of technological development, growth may depend critically on im- porting goods and services from abroad, and that domestic resources may not always be freuly convertible into foreign resources due to the limited poten- tial for expanding exports. In this situation, foreign capital inflows have to fill up the bigger of the two gaps - trade gap and saving gap. The problem of what happens to the nonbinding gap (the smaller of the two gaps) and how the two gaps are equalized at the end of the day (as they must be) are not resolved in the classical two-gap approach. 2.03 Because of this analytical problem in the usual two-gap model, we have tried to introduce a modification to it. We assume that the ex post equalization of the two gaps is brought about by changes in prices of consumer goods relative to the general price level. 1/ If this increase in price is un- acceptable, additional imports would be required to meet the increased pro- pensity to consume. In fact, in this procedure, there are not two gaps but two parts of the trade gap - one part due to basic imports requirements, an- other part to prevent an excessive rise in consumer prices. With this distinc- tion, one could examine various policy alternatives such as increasing foreign capital inflow, changes in exchange rate, relative prices of consumer goods, or change in pattern of growth, instead of being forced into the assumption that foreign capital inflow is always available to fill the bigger of the two gaps. This becomes particularly useful if part of the domestic consumption goods gap is for non-tradeables, for in this case, foreign capital inflow can- not fill the domestic goods gap in the current period. 2.04 The above gap analysis could, of course, be done in two ways: either through the "requirements approach" or the "availability approach". In a "requirements approach", we start with a target growth rate and try to 1/ For further discussion, see paras. 3.20 - 3.23. - 4 - compute the foreign resources required to achieve this growth rate. In the "availability approach", the model is run in reverse gear, so to say, and starting with the available resources, the model works out the growth rate that can be expected with these resources. The model presented in this annex may be termed a "modified requirements approach". In this approach, we start with the target growth rates and work out the net resource transfers required. However, instead of stopping at that point, we bring in our best guesses about the availability of resources from the usual sources and, after taking into account the pattern of disbursements expected to flow from these commitments and the debt servicing requirements of past debts, try to estimate the addi- tional resource requirements over and above those likely to be available. Moreover, even the residual resource gap is not taken as given; instead in- ternal policy options are also examined to see if the resource gap could be reduced without diminishing the objectives of growth, employment and the attack on poverty. In this fashion, the residual gap that we are left with represents the minimum of extra effort required on the part of the external donors. 2.05 Given the target growth rate, the first step in the analysis is to estimate the investment required. This implies that we assume that output is determined by the supply of fac'trs of production and not by demand. This is a reasonable assumption for medium and long-term analysis, but cannot be used to analyze year to year variations in output. The problem of demand becomes even more important when one is considering a disaggregated approach, as the Plan does (and as we are doing in this annex). In this context, it is impor- tant to ask whether the structure of output implied by different sectoral growth rates is consistent with the demand pattern likely to be generated by the structure of production. This is obviously a very difficult question, reqjuiring assumptions about input/output structure, final expenditure patterns and the influence of fiscal, monetary and other policies on demand pattern. In our analysis, we do not tackle this problem, beyond making some spot checks on consistency for some important sectors such as agriculture and manufacturing. 2.06 In estimating investment requirements for a given rate of growth, the parameter that we use is the incremental-capital-output ratio (ICOR). This does not imply that we assume capital as the only factor of production. We treat ICOR not as a technological parameter, but as an economic parameter influenced by the rate of growth, pattern of growth, technical change and relative factor prices. Gross ICOR in each sector is inversely related to its rate of growth, because with higher rates of growth, the depreciation component of gross in- vestment becomes a smaller part of total investment. Aggregate ICOR is also a function of the pattern of growth. For the same level of aggregate growth rate, ICOR (and thus investment) could be reduced if the pattern of investment is oriented towards sectors in which ICOR is lower than in others. Even in a particular sector, ICOR could be changing (as it is in Kenya), due to changes in degree of utilization and other sources of efficiency of resource use. In our model, we try to take into account the overall trends in ICORs in different sectors. Similarly, one could change factor proportions (and thus ICORs) even in a given sector for a given degree of efficiency, by changing factor prices. -5- 2.07 The next stage in the analysis is to compute the amount of foreign resources (imports of goods and services) required for target growth rates. Here again we try to take into account the effect of the pattern of growth and prices on import requirements. Imports are divided into five categories: consumer goods, raw materials, capital goods, government imports, and imports of non-factor services. Imports of raw materials are obviously dependent on the pattern of growth; because a. unit of value added in, say, manufacturing requires more imported raw materials than does a unit in agriculture. Imports of raw materials were therefore related to value added in different sectors, weighted by the ratios of imports of raw materials to gross value added ob- tained from the Input/Output Table for Kenya, 1967. 1/ For imports of con- sumer goods, the sectoral weights were derived from the ratio of taxable in- come to GDP in each sector, on the assumption that those whose incomes fall below the taxable minimum are not by and large rich enough to consume imported consumer goods. Lastly, for import of capital goods, it was assumed that ma- chines and vehicles are imported, but not structures or animal stocks, so that the weights used for investment in different sectors were derived from the ratio of machines and vehicles to total investment. Government imports have been taken as a function of GDP in the government sector, 2/ and imports of non-factor services as a function of time. Imports in different categories are also influenced by the ratio between domestic prices and import prices (adjusted by rate of exchange). 2.08 In order to assess the net foreign capital required to meet import requirements, we need to estimate exports. In the model, exports are largely exogenous. However, we introduce two influences to make them quasi-endogenous. In the first place, exports are sensitive (although to a limited extent) to the exchange rate. Second, they are sensitive to variation in the rate of growth of agricultural sector. Import and exports as discussed above are in constant prices. However, in order to assess the foreign capital requirements, we have to transform them into current prices in foreign currency (in our case US$) because this is how foreign capital inflows will be denominated. The model therefore uses import and export price indices to compute the net re- source balance at current US$. 'rhis gives the external imbalance expected from target growth rates. 2.09 To compute the interna:L resource gap, we have to estimate saving potential. This has been done art a disaggregated level: savings by house- holds, government, and the business sector, the latter separated into depre- ciation allowances and undistributed profits. 1/ It is possible that in the 'Light of structural changes going on in the economy, the 1967 table is a little dated. However, there was no other more recent source of infoniation on Input/Output structure. 2/ At a later stage, we hope to analyze the effect of composition of govern- ment expenditure on imports,. - 6 - 2.0 Household savings are computed as a residual after deducting con- sumption from disposable income. In computing household disposable income, account is taken of wages, property incomes, net transfers from abroad, income tax, and other taxes, etc. Government savings are calculated from figures of Wax arnd other .eceipts computed in the model and the govermuent consumption expenditure is a function of GDP in government. In estimating business sav- ings, we compute depreciation allowances as a function of past investment rates, and undistributed profits are then a residual. 'Tne ex ante consump- tion of the households sector is compared with the supply of consumer goods available after deducting investment and government consumption from total resources available. The gap between ex ante consumption and the supply of consumer goods shows the degree of inflationary pressure - or internal imbal- ance - associated with the target growth rates. Depending on whether such inflationary pressure is acceptable or not, an increase in foreign resources may be required. 2.11 If the domestic resource gap is filled through additional imports, the ex post imports will be higher than the ex ante imports. If, on the other hand, import controls are imposed, the ex post imports might be lower than ex ante imports. After adjusting for import prices and deducting export earn- ings, we get the current trade balance at current prices. In order to obtain the gross foreign capital inflows required, we have to make a number of other adjustments. First of all, net factor income payments and net unrequited transfers have to be added. By adding in the debt servicing payments from old loans, and deducting the disbursemerts from old connitments, we get the new gross transfers required. In order to estimate the transfers generated by new commitments, we have to assume the pattern of disbursement of aid from dif- ferent sources (such as IBRD, IDA, or bilateral sources) and on different terms of lending (relating to maturity, grace period and rates of interest). After deducting the disbursements from new loans, and adding the future debt ser- vicing from new disbursements anid additional foreign reserve requirements, we get the size of unfilled "gap" for which new resources will have to be found. In making these calculations about debt servicing, we have used a detailed computer program developed for the purpose in the Bank. 2.12 In addition to gap calculations, our model is geared to making esti- mates of employment in different sectors. Calculations are made for three types of employment: wage employment in the modern sector, high and middle level manpower requirements, and total employment, including wage employment in the informal sector, self-employed and family workers. The broad approach is to estimate the incremental employment in each sector, wnich depends upon the rate of growth of value added and of labor productivity in each sector. High and middle level manpower requirements in each sector are assumed to be fixed proportions of wage employment, the proportions varying for different sectors, as indicated by the 1972 MFanpower Survey. 1/ 1/ See "A Preliminary Report on the Kenya High and Middle Level Manpower Survey, 1972",' Kenya Statistical Digest, December, 1972. 2.13 Assumptions about labor productivity are made on the basis of past trends, but the model itself uses different assumptions about this parameter in order to analyze employment prospects under different assumptions. The in- cremental labor-output ratios are also influenced by factor price policies that change ICORs. 2.14 The above analysis is so far in terms of constant internal prices. However, to analyze government revenues and balance of payments, one needs an estimate of prices. The model does not, at present, integrate internal prices into the system, except for estimating the influence of internal prices on im- ports. Ideally, it would be desirable to link the domestic saving-investment gap to prices - either through its effects on money creation, or directly. However, at present, the model assumes the rate of domestic credit creation as a policy instrument. Domestic credit, plus foreign reserves, give money sup- ply. The total nominal value of resources (GDP, plus imports, minus exports) depends on current and lagged money supply. The price level is then deter- mined by dividing total resources at current prices by resources at constant prices. The consumption goods gap, however, determines the relative prices of consumer goods vis-a-vis general price level. 2.15 The aDovve section has given a brief non-technical description of the structure of the model. Further technical details are presented in Part II of this annex, and a flow diagram of the main lines of causation is shown in Chart I as a further aid to understanding the structure of the model. 2.16 How do we Use the Model? We start the model with the sectoral growth rates, which were the rates being used in the provisional plan projec- tions provided to the Mission. We also make somewhat optimistic assumptions about export prospects, terms of trade, ICORs and saving propensities. Under these assumptions, we use the model to project the basic economic scene up to 1985. We then examine whether the gap in resources is manageable over the next few years, and how employment position looks around 1985. 2.17 If there is a gap, the next question is: how to fill it? One pos- sibility is increased borrowing on hard terms such as suppliers' credit. The model then examines the consequences of this on the debt servicing capacity of the country around 1985. It also examines whether a softening of terms of lending by lending agencies can make a substantial difference to debt servic- ing position. A second possibility is to impose import controls to reduce foreign resources gap. The model works out some of the consequences of this on domestic prices and balance of payments situation. A third possibility is to reduce the gap by lowering growth targets. However, this is found to have unacceptable effects on employment and poverty prospects. Finally, we go on to examine more interesting possibilities of reducing the gap, by changing the pattern of growth and by changes in policy. 2.18 The model examines the effects of changes in the pattern of growth on the size of the gap and on employment and poverty. More specifically, it examines the effects of increasing growth rate in agriculture and reducing that in infrastructure investment. The second set of alternatives are exam- ined by changing successively the rate of exchange, investment allowances, and interest rates; and we examine how much difference is made by each strategy. - 8 - 2.19 Finally, the results of the above analysis are utilized to obtain an illustrative "preferred policy" package, and the model works out the scenario up to 1985 under these conditions. As emphasized in the preceding chapter, these results are to be regarded mainly as suggestions of the orders of magni- tudes involved, and these have to be supplemented by various judgmental analy- ses before a policy package could be decided. - 9 - CHAPTER 3. STATIST'ICAL ANALYSIS OF THE PAST AND WORKING HYPOTHESES FOR THE FUTURE 3.01 In this chapter, we present the main quantitative results of our analysis. As mentioned in the introduction, we shall only give the broad re- sults, and the methodological and analytical procedures are left aside for discussion in Part II. Developments over the Period 1964-1971 3.02 Our analysis is based on the developments in the Kenyan Economy since Independence, because of the limitations of data before that date. More specifically, the period covered is from 1964 to 1971, with the 1971 figures still provisional. 1/ Even for the conventional econometric analysis, this is a small number of observations, and the problem is further complicated by indi- cations that during the last two years, 1970 and 1971, the economy may have been going through a turning point. Because of these reasons, we have to exercise caution in using the equations estimated over this period for projec- tions through the next decade. 3.03 Productivity of Investment We begin our analysis with movements in incremental-capital-output ratios. In a very broad sense ICORs can be regard- ed as indicators of efficiency in resource use.2t In this respect, there are three observations to be made about the level and movements in ICORs in Kenya: a. The overall ICOR in Kenya was low by international standards at the beginning of the period. b. The overall ICOR (and some sectoral ICORs) have been rising rapidly over the years. c. By the end of the period, the overall ICOR was higher than the good performers among developing countries, but still better than poor performers, and not signi- ficantly out of line with the average for all develop- ing countries. 3.04 Using three-year moving averages for smoothing out investment and output figures, we get an ICOR (for fixed capital) of 2.1 for the total GDP and 2.34 for monetary GDP in Kenya in 1966. Adding 0.3 for inventories, we get on overall ICOR of 2.4. The estimates of ICORs computed by the Bank for a number of other developing countries are presented in Table 1. We notice 1/ Provisional data for 1972 have since become available but it has not been possible to take them Into account in our analysis. 2/ One very significant aspect of efficiency that ICORs ignore is the longevity of capital. - 10 - that even for countries that have experienced good growth recently, ICORs were not significantly lower than 2.4. For example, it was 2.46 for South Korea, 2.45 for Brazil and 2.11 for Indonesia. Kenya's performance, of course, was very creditable in comparison with ICOR's of countries such as Philippines (where it was 6.45), Colombia (3.93), Egypt (5.05) and India (3.64). 3.05 However, over the period since 1964, overall -COR increased rapidly and by 1970 it had increased by nearly 50 percent to 3.2. Since an analysis of the causes for this rise in ICORs would be useful for projections as well as policy analysis, we analyzed the disaggregated sectoral ICORs for 12 sec- tors. To obtain a better measure of efficiency of new capital, we subtracted depreciation allowances from gross investments and computed net ICORs, the results of regressiorns on which are presented in Table 34. From these sec- toral net ICORs we notice that there was a significant increase in net ICORs in five sectors: mining and quarrying, building and construction, transport, storage and communication, other services, and general government. However, there was a significant decline in net ICORs of manufacturing and repairs, so much so that by 1970, ICOR in manufacturing and repairs was lower than that in agriculture, forestry and fishing. 3.06 The explanation for these varying trends require more intensive study by sectors than we have been able to provide. However, we may suggest some plausible lines of argument. One possible explanation is that Kenya in- herited some excess capacity from the previous period, so that output could be increased without much input of capital in the earlier years. As the slack in the economy was taken up, more and more capital was required for increases in output. Another possible explanationi is that during this period factor prices have been distorted, so that capital was made excessively cheap, and this distortion encouraged an excessively capital intensive structure. This was perhaps particularly likely in building and construction, as well as in transport, storage and communications. 3.07 The declining trend in manufacturing and repairs is more puzzling. One possible explanation is that the system of protection -introduced over this period gave artificially high values to the output in manufacturing sec- tor. Ideally this price rise should have been captured in the price deflator used to estimate constant prices in manufacturing; but typically, these price indices use base-year weights, in which the new items of production are un- der-represernted. Since capital gGods are typically exempt from duties, and new manufacturing items were developed under protection, ICORs tended to have a downward bias. Moreover, as more and more new manufacturinig activities wear, started, the effective average rate o-` over valuation of manufacturing outpuit inccrea3ed, so that ICORs were not Ly low but falling over timre. Another possible line or explanation may be the lumpiness of investment in particular sectors such as petrolezLu and cement. Investment undertaken in these sectors did not lead to output until the later years, so that ICORs tended to decline as gestation lags in these investments were over. A similar reason may be that in the early period a considerable amount of industrial capacity was created (in contrast to the inherited excess capacity mentioned above) as a means of preempting a wider East Africa market, and ICORs there- fore declined over time as capacity utilization was increased. - 11 - 3.08 A more intensive study, industry by industry, would, of course, be required to understand the forces behind the efficiency of resource use. But even our preliminary analysis suggests some policy implications. As shown in Table 1, ICOR in Kenya was sign:Lficantly higher by 1970 than that of good per- formers such as Brazil, South Korea, or Indonesia. It is still better than poor performers, although no longer significantly better than the average in developing countries. If Kenya wants to be in the league of good performers, therefore, it must try to improve its efficiency or at least prevent further deterioration. 3.09 In our policy package analysis, we suggest two lines of action. In the first place, we consider a number of policy instruments, such as ex- change rate, interest rate, and investment allowances, which could change the price of capital relative to labor, and thus might influence ICORs. Second, we note that ICORs differ significantly from sector to sector, and that by changing the pattern of growth, therefore, the overall ICOR could be lowered. In particular, we examine the benefits to be derived from a relative shift towards agriculture. Some extra output (and value added) will be required in other sectors to supply the raw materials required by agriculture. But, even after these indirect resource requirements are considered, we find that overall ICOR can be reduced sign!ificantly by changing the sectoral pattern of growth. In particular, as we note below in this chapter, a strategy for promoting agricultural growth and slowing down on infrastructure will help in reducing ICORs. 3.10 Import Requirements. Except for the last two years of the period, nanmely, 1970 and 1971, imports have grown about the same rate as GDP. While this does not imply any significant reduction in import dependence (as might have been expected from the so-called import substituting protectionist strat- egy), this was better than the experience of some developing countries which have seen an increasing degree of import dependence while carrying out import substitution. The last two years have, however, witnessed a surge in imports (in constant prices), in spite of a rise in import prices. In part this rise in imports can be explained by the high level of economic activity and espec- ially investment, and possibly a build up in inventories in the expectation that import controls would be introduced. However, it may also be partly due to some loss of confidence in the future of expatriate business community in Kenya, and a consequent flight of capital. Whether, in the future, import re- quirements will increase more in line with GDP or not will probably depend on what kind of policies are followead regarding the expatriate business community and whether encouragement is given to agro-based manufacturing. 3.11 In our analysis, we break down total imports into five categories: consumer goods, raw materials, capital goods, government imports and imports of non-factor services. As shown in Table 3, imports of non-factor services have been almost static, while the percentage share of raw materials has been increasing fast. In our analysis of the demand for imports, we try to take into account the differences in :Lmport requirements arising from growth in different sectors. For example, the Input/Output table for 1967 suggests that the requirements of imported raw materials and intermediates in manufacturing - 12 - were as much as 77 percent of value added, wnile those in agriculture were only about 9 percent. One could also expect similar differences in the demand for imported consumer goods in different sectors. For example, it seems rea- sonable to assume that "poor" people do not consume any significant amount of imported consuner goods, and sectors such as the "non-monetary" and agricul- tural sectors are likely to generate a higher proportion of income going to the poor than are sectors such as manufacturing or banking. It is of course difficult to get any reliable set of data on patterns of income distribution generated by different sectors. 0ne approximation may be wage vs. non-wage income. But for sectors such as "non-monetary" or agricultural, where poorer people predominate, wage income is small because most of che people belong to the category of self-employed poor. 3.12 We therefore decided to use another set of data. The income tax data show the amount of income generated in differect sectors wnich is assessed for income tax purposes. Assuming that those people whose income is not high enough to be assessed for income tax purposes are not importers of consumer goods, we used the ratio of assessed income to GDP in each sector as a weight to apply to sectoral value-added in obtaining an equation for consumer goods imports. The highest weight was shown by banking, insurance and real estate sector, where 83 percent of the value-added was assessed for tax. The smallest weight (except for the non-monetary sector) was in agriculture, where it was 13.7 percent. The weights are of course not ideal but it seems to us that these differences might be closer to reality than the assumption of equal weights implied in a purely aggregative approach. 3.13 A similar approach was taken for imports of investment goods. It seems reasonable to argue that capital goods in mi-ning or manufacturing are likely to be more import-intensive than in agriculture. Again we could not get any systematic informiation on this and had to rely on a crude assumption that machinery and transport equipment are imported wnile breeding stock and structures are not, and used the information on type of capital goods required by different sectors for formulatirig weights in the equation for import of capital goods. 3.14 The weighting procedure described above enabled us to form some judgment as to how much of the increase in imports over the period was due to the increase in GDP and how much due to changes in the composition of GDP. On the basis of our analysis, we have estimated that 25 percent of the in- crease in imported consumer goods, and 28 percent o: the increase in raw materials, has been due to the chan;e in sectoral composition. Only 3 per cent of the increase in investment goods imports was due to sectoral changes. For the analysis of imports by goverrment and imports of non-factor services, we employed a fairly simple approach. Government imports were regressed on v value added in government, and the marginal propensity to import was found to be 7 per cent. Imports of non-factor services were growing at an annual rate of 2.3 oer cent, or distirnctly lower than t;nat of imports of goods or of GDP. 3.15 nlysis of Savings and Exrorzs. irn order to assess the domestic re- sources and foreign resources required f`or target growth rates, we have to de- duct savings from investzients requirements, and exports from import requirements. - 13 - Our analysis for domestic savings will be presented in Annex 4 and we do not propose to repeat it here. Broadly speaking, our conclusion is that Kenya's saving performance, until recently, has been very good. In 1970 and 1971, the saving rate, particularly of the household sector, has declined and there is some question in our minds whether the previously high saving propensity can be restored. However, for our model, we made the optimistic assumption that the savings performance during the period 1964-1971 will prevail in the future. The saving functions in different sectors - household, business and government - have been estimated in unpublished work undertaken by the Ministry of Finance and Planning, and in our model we utilized the equations formulated by them. Household savings are assumed to be 10 per cent of personal dispos- able income; depreciation allowamces are made a function of past investments, with weights obtained from tax :Laws. The undistributed profits of businesses are treated as residuals. Govermuent consumption is assumed to grow as a func- tion of value added in general governrment, and savings are estimated as the residual, after deducting tax receipts, which are analyzed at great length. The equations used in the model are presented in Part II of this annex. 3.16 Movements in EmploymenLt. The growing rate of unemployment in the economy is widely regarded as arn important shortcoming in Kenya's otherwise good performance. However, while the problem is regarded as crucial, adequate data are not available to make a. proper analysis of the problem. Data on wage employment in the modern sector are available for the period 1964-1971. However, there are no GDP figures to correspond to the modern wage employment only. On the basis of the results of the population census of 1969 and other related information, a set of data for total employment by sectors for 1969 and 1971 was prepared by the Mission and this is presented in Table 4. The formal wage employment in 1969 (627,000) accounted for only 12.4 percent of total employment (5 million). The urban informal sector accounting for 96,000 (2% of total) is certainly a significant sector, but could hardly be regarded as a likely mainspring of growth in employment. Similarly, the rural non-ag- ricu:Ltural informal sector, although much bigger than urban informal sector '3% of the total), is insignificant when compared with1 informal employment in .g-> culture, which accounts for some 83 percent of the total. The agricul- ttiral sector is assumed in the model to be the residual employer, and thus it Is by increasing productivity in this sector that the conditions of the "working poor", as ILO/UNDP ReporLt on Kenya calls them, can be really im- proved. 3.17 On the basis of these iLnadequate data, we could not estimate any meaningful employment equations. The simplified approach that we used was to make different assumptions about changes in labor productivity over time in different sectors and deduce theiLr employment implications. On an overall basis, over the period 1964-71, t:he rate of growth of GDP has been about 7 percent, and that of wage employment between 3 and 4 percent, thus implying (for sectors where the ratio of formal to informal employment did not change significantly) a productivity grcwth of 3 to 4 percent a year. As regards sectoral employment growth rates, we note that the figures in agriculture and trade are significantly influenced by settlement schemes and Kenyaniza- tion programs, so that wage employment figures may be misleading. Employment - 14 - figures in electricity and water and transport, storage and communication also showed some irregular patterns of movement at a generally low level of employment. We could not therefore infer any consistent sectoral differencies in rates of growth in labor productivity. For projection purposes we had to make judgmental estimates about the rates of growth in labor productivity in different sectors. We note, however, that the rate of growth in labor pro- ductivity is connected with the rate of growth of capital output ratios and, since we are assuming no further increase in ICORs, the rate of growth of labor productivity may be expected to be lower in the future than in the past. 3.18 For projections of total employment, we have used a weighted aver- age of the rate of growth of labor productivity in the formal and informal sectors, and have assumed that the agricultural sector absorbs the residual labor force. 1/ The income generated in the agricultural sector, divided by the labor force it has to absorb, determines the level of income of the "working poor" in agriculture. Thus, while the ceteris paribus gains in labor productivity in other sectors tend to reduce employment in these sectors and lower the standard of living of agriculturists, an increase in labor productivity in agriculture does not result in unemployment but in higher standard of liv- ing of the "poor". 3.19 In the Kenyan economy, skilled manpower may be regarded as a scarce factor of production, and in our model we make some rough projections for the future requirements of skilled manpower. Here again it is important to note the significant differences in high and middle level manpower required by dif- ferent sectors. Table 5 presents the figures of high and middle level man- power as a ratio of wage employment in different sectors: while agriculture requires less than 4 percent of its wage employment as skilled labor, the electricity and water sector requires 25 percent and trade 30 percent. Since the supply of skilled manpower will continue to be limited in Kenya for some years, changes in sectoral patterns of growth may become one of the most use- ful ways of reducing that constraint. 3.20 Money and Prices. In the usual two-gap analysis, money and prices are absent. This creates problems of estimation as well as analysis. Some of the functions such as tax functions should really be analyzed in terms of current prices. Similarly, the general rate of inflation as well as relative prices may have important implications for such variables as the capital-labor ratio, import requirenents, or income distribution, and these should ideally be analyzed within the model. Even the question of ex post equalization of the two ex ante gaps - the saving gap and the trade gap - can be tackled only by b-ir:king in price changes. However, while we have not been able to analyze the influence of prices at all adequately, we have attempted to bring prices into the model. In our system, price changes equate the two gaps, and also influence the level of imports. 1/ In this model, therefore, there is no unemployment. The actual scale of unemployment in the future will largely depend upon the "gap" between rural incomes and wages in the urban formal sector. - 15 - 3.21 Until recently, Kenya has experienced a remarkable degree of price stability. The average annual rate of increase in the GDP deflator was less than 2 percent, and this degree of price stability, associated with rapid growth, must be regarded as good performance in comparison with other develop- ing countries. The reasons for this price stability over most of the period, and the recent increase in prices, should be interesting to analyze. On the one hand, Kenya's experience could be regarded as a text book justification of the monetarist view of inflation. During this period, monetary policy was very conservative, and the rate of expansion of money supply was low. As shown in Annex 2, the deficits in government budget were not financed by bor- rowing from the Central Bank. In fact, until recently such borrowing was limited by legislation to E 12 million. However, in 1972 there was a change in legislation and the ceiling on government borrowing was raised to 25 per- cent of gross current revenues, or the equivalent of about t30 million in 1972/73. The Government has, in fact, borrowed more heavily from the Central Bank, money supply has expanded rapidly and the rate of inflation has picked up. 3.22 On the other hand, one could also make a plausible case for the non- monetarist view. There was probably some excess capacity in the economy in the early part o; the period. The import prices increased significantly in 1970 and 1971, and some agricultural prices were also increased by the Govern- ment in these years. One might therefore argue that the past price stability and recent inflation could be explained with reference to "eost-push" prices. In fact, the recent price rise in Kenya may only be a part of the present world-wide wave of inflation. 3.23 It is, of course, difficult to sort out the relative influences of these factors. However, it seems to us that for a long term analysis of in- flation a monetarist approach may have statistical as well as analytical ad- vantages. Using current and lagged money supply as independent variables, we found excellent fit for monetary expenditure. The elasticity of money expend- iture with respect to money supply was about .65, thus implying a declining velocity of money as has been found to be the case in many other countries. The fall in velocity could have been due partly to the declining proportion of the non-monetary sector in total GDP, and partly to the increasing demand for cash balances which is generally associated with growth. However, we could not separate these influences. Policy Variables in the Model 3.24 As mentioned before, we have tried to design our model so that we could make illustrative calculations about the effects of changes in policy instruments, such as exchange rate, interest rate, wage rate, or investment allowances. In general, it is difficult to estimate the effects of these policy instruments by using them directly in regression equations; it is even more difficult to do this in Kenya, because these policy instruments have not been changed significantly over the period. The general approach we have fol- lowed, therefore, is to analyze the effects of these instruments in a non- quantitative way outside the model and then use the model to get some quanti- tative feel by using different plausible assumptions about the critical -- 16 - parameters. Whenever appropriate, we have drawn upon the experience of other countries which have chosen to employ these policy instruments in the manage- ment of their economies. 3.25 Effects of Change in Exchange Rate. In Kenya, the exchange rate in terms of U.S. dollar has been stable since Independence. 1/ Over this period, however, it has been revalued in terms of sterling, the currency of Kenya's main overseas trading partner. Now that Kenya seems to be entering a phase of foreign exchange constraint, it may become desirable to consider the ex- change rate as a positive and flexible instrument of policy. At present, when the old dogma of fixed and inflexible exchange rates seems to be coming to an end, the rate of exchange should not be treated as a symbol of national prestige but as just another price. 2/ When viewed as a positive inistrument of policy, a reduction in the exchange rate might well be expected to confer significant benefits in terms of export promotion, import displacement, em- ployment promotion, stimulation of agriculture and agro-based manufacturing, and some redistribution of real income from the urban rich to the rural poor. In the light of these effects, one could in fact argue that it is preferable to err in the direction of an under-valuation rather than an over-valuation of the exchange rate. 3.26 In the usual analysis of the effects of devaluation, the impact effect is an immediate deterioration in the terms of trade of the devaluing country. If the increase in real exports and the decrease in real imports more than counterbalance the adverse terms of trade effect, its foreign ex- change position will improve over a period (of say 2 to 3 years). However, in Kenya (as in most other developing countries) the prices of most major ex- ports are determined in foreign markets, in terms of foreign currency. Thus, the most noticeable effect of a devaluation is not a worsening of the terms of trade, but the increased domestic profitability of export industries and the increased cost of imports. 3.27 The long-term effects of changes in exchange rates will, of course, depend on elasticities of supply of different commodities with respect to their domestic prices, and demand elasticities in foreign markets in terms of foreign prices. It would be desirable to make a commodity-by-commodity study of elasticities to assess the effects of devaluation. However, our prelimin- ary explorations suggest that the statistical data are not adequate for this 1/ We ignore the June 1973 revaluation against the dollar. 2/ Brazil's recent experience is yet another vindication of this point of view. Similarly, Pakistan's massive devaluation in May 1972, was fol- lowed by a dramatic improvement in her balance of payments. For the eleven months July 1972 - May 1973, exports increased by 39% (in US$) over the same period in 1972. Apart from merchandise exports, there was a considerable increase in recorded invisible receipts, especially remittances. - 17 - purpose. The scattered evidence 1/ that does exist suggests that Kenyan farm- ers who grow export commodities (such as tea, coffee, or sisal) are responsive to price changes. Moreover, since Kenya occupies a small share of world mar- kets, world prices are unlikely to be effected to any significant extent by changes in the level of Kenya's exports. 3.28 The import displace:nent effects vwork through a number of channels. Imports of capital goods may be reduced partly because of the increased cost of foreign capital goods vis-a-vis domestic caDital goods and partly because of the increased cost of capital vis-a-vis labor. Similarly, imports of con1- sumer goods may be reduced partly because of redistribution of real income from the urban rich to the rural poor, and partly because of higher prices of imported consumer goods vis-a-vis domestic goods. The effects on imports of raw materials are uncertain. They will increase to the extent that they are required for the production of consumer goods formerly imported but now domes- tically produced. However, to the extent that import-intensive luxury con- sumer goods are priced out of the market, the demand for imported raw materL.lsS will decline. The important point is that a higher price of foreign excharge would effect both the direct and indirect content of imported goods, and would bring about the desired effect:s on the domestic production system much more efficiently than the present ad hoc and arbitrary system of protection has done. 3.29 We have discussed at. length in Annex 3 how the exchange rate and a tariff-subsidy scheme could be used as complementary instruments in obtaining inter-sectoral efficiency as well as balance of payments equilibrium. In our model we have not been able to tackle the effects of exchange rate changes on inter-sectoral allocations, We have considered only its effects on exports, imports and capital/labor substitution. In general, we have taken conserva- tive estimates of elasticities. 3.30 Effects of exchange rate on income distribution - both inter- personal and inter-sectoral - can be quite significant. An over-valued 2/ exchange rate results in the taxation of exports and the subsidization of im- ports. Since poorer people consume a smaller part of imports than do the 1/ See for example, G.D. Gwyer, "Long and Short-Run Elasticities of Sisal Supply, "East Africa Economic Review," December 1971; D.J. Ford, "Long-Run Elasticities in the Supply of Kenyan Coffee: A Methodological Note," East African Economic Review, June 1971; and J.K. Maitha, "A Supply Function for Kenyan Coffee," East Africa Economic Review, June 1969. 2/ In the Mission's view the Kenya shilling is probably over-valued - not in the sense of short-run balance of payments eouilibrium, but in the sense that the value of the Kenya shilling (in terms of foreign exchange) is higher than it would be, were the exchange rate used as a more flexi- ble tool of development policy. ricner people, an increase in the price of imports will reduce the real in- come of tne rich more than that of the poor. Similarly, poorer people may derive a greater snare of income from export oriented growth (based on agri- culture 1/ and agro-industries) than they do from import substituting manu- facturing. 3.31 A reduction in the exchange rate by benefitting the agricultural sector will also benefit the rural areas, and thus reduce the pull of urban centers which is complicating the urban unemployment problem. Similarly, by increasing the price of capital vis-a-vis Labor, it may be expected to improve employment prospects for the rural as well as the urban labor force. 3.32 Apart from the above-mentioned longer run effects of adjusting the exchange rate, one should also note the expectational factors in the situation. Once a currency is generally regarded as over-valued (one indication is the rate prevailing on the open market) private capital inflow is discouraged, capital outflow is encouraged, the building up of imported capital goods capa- city ahead of demand appears attractive (leading to wastage in the form of excess capacity), and the situation becomes progressively more and more diffi- cult. 3.33 Devaluation essentially means increasing the cost of foreign ex- change, but the main problem created by such an increase is the possibility of a faster rate of inflation. If increased inflation could be used to re- duce the real wages of the urban labor force, it might not be unwelcome. However, if it leads to pressures for money wage increases which cannot be resisted, it may accentuate the inflationary process. 3.34 Interest Rate Policy. The interest rate structure in Kenya has been almost static for the last ten years. In the meantime, world money market rates have risen to unprecedented levels and inflation in Kenya seens to be gathering strength. This has had the serious effect of reducing the incentive of foreign private companies to borrow abroad, and has thus aggravated the resource situation. The increased rate of inflation is encouraging excessive capital intensity, by reducing the effective real cost of capital, and dis- couraging domestic savings, by reducing the effective real return from savings. Moreover, in Kenya there is no significant rentier class, so that low interest rates do not seem to have any significant benefits for income distribution. In general, it may now be appropriate for Kenya to reconsider its interest rate policy; on balance, a higher interest rate policy would seem to be de- sirable. 2/ 'J/ Eveni in co.mmercial crops such as tea, coffee, and pyrethrut, the share of productionr originatinig in small far-ms is quite significant. For ex- ample, smallholders produce 90 percent of Kenya's pyrethrum exports, nearly half the coffee, and a growing proportion of the tea. 2/ For further details, see Annex 4. - 19 - 3.35 Wage Rate Policy. 'Wage earners are generally identified with the poorer sections of a country, and higher wages may therefore seem desirable on income redistribution grounds. However, in Kenya wage earners, especially in the urban formal sector, represent a privileged minority of the labor force. Higher wage rates for this sector have distortionary effects on capital inten- sity, urban-rural migration, the education system, and urban social services. As the ILO/UND? Mission has emphasized, the real problem of poverty and unem- ployment lies with the "working poor"t, and a curb on the rate of wage increases in the formal sector may well help these poorer sections of the community. 3.36 As discussed above, all these policy instruments have far-reaching effects in the economy, and our model does not pretend to capture all these effects. In particular, it is unfortunate that income distribution effects which we regard as significant cannot be quantified. The model does try to capture the direct effects of exchange rate changes on imports and exports, and the indirect effects (via capital-labor substitution) on internal and ex- ternal imbalances, employment prospects, ar.d the prospects of the working poor in agriculture. 3.37 One critical element: in our analysis is the assumption that possi- bilities of capital-labor substitution are significant. A priori, it seems to us that the influence of factor prices on the capital/labor ratio works through so many channels - for example, changes in product mix, techniques, and the rate of utilization-that the total effect should be significant. On an empir- ical level,, it is worth nothing that Henry Bruton if, after an extensive sur- vey of the statistical results on elasticities of substitution, noted that "The most general conclusion is the most important: Factor substitutability is alive and well in developing countries. Policies and models that assume otherwise mislead." More specifically, his review of the results of statisti- cal studies in the United States, Philippines, Argentina, Chile, El Salvador, Korea, Paraguay, Peru, Portugal, Spain and Mexico found that elasticity varies from 0.5 to 1.6, and is generally significantly different from zero. In the Kenyan economy, data on output, labor and capital by industry are not available for fitting' an explicit production function. However, empirical analysis of the figures of labor productivity and wage earnings gives estimates of elasti- city of substitution which are not significantly different from unity (see Appendix 5). 3.38 Thus, in the light of empirical evidence on production functions in developed as well as developing countries, it seems reasonable to assume uni- tary elasticity of substitution in factor proportions with respect to factor prices. We also use the share of wages in gross output as an approximation to the elasticity of ourput with respect to labor. 3.39 Lastly, we have conservatively assumed that existing factor propor- tions are not malleable, and that these substitution effects apply to incre- mental capital and labor only. Under these assumptions, a change in relative 1/ See Appendix 5. - 20 - factor prices will change ICORs and the incremental labor-output ratio in the model. The exact equational form in which these factors are incorporated is explained in Part II. As an illustration, it might be noted that an invest- ment allowance of 20 percent may reduce the effective cost of capital by, say 8 percent (assuming a marginal tax rate of 40%), and this could increase ICOR by about 4 percent (assuming the wage share to be 50%). Similarly, an increase in the cost of foreign exchange by 20 percent may increase capital cost by 6 percent (assuming imported capital goods to be about 30% of total investment) and this may reduce ICOR by about 3 percent. These measures will also change the employment content of growth, pari passu. Basic Hypotheses Used in the Model 3.40 The projections in our model require a large number of assumptions about exogenous variables, and even about the future movements in parameters. Our approach to the assumptions made in the model has been as follows: a. We have used as far as possible exogenous assumptions close to those of the Ministry of Finance and Planning. These could be regarded as proxies for the assumptions of the forthcoming 1974-78 Development Plan. b. In general, we have taken optimistic assumptions with regard to the future. This affects our assumptions about the terms of trade, ICORs and foreign capital inflows. When we have introduced parameters such as elasticities of exports or imports with respect to exchange rate, we have taken conservative assumptions. c. Although there are indications that the Kenvan economy may be going through a turning point, and behavioral equations are shifting, we have made optimistic assump- tions that the past behavior pattern will hold for the future. 3.41 The starting point of our analysis therefore is the set of targets growth rates, for different sectors, which form the preliminary targets of the Plan. We had to adjust them slightly because of the differences in sectoral breakdown; we also use the same rates of growth up to 1985, whereas the Gov- ernment assumptions only go up to 1978. The actual figures are presented in Table 6. W4e note that the targets imply a healthy overall growth rate of 7.5 percent (which is even higher than the growth rate in the past decade) and a considerable acceleration in the manufacturing sector. 3.42 As regards ICORs, we make the optimistic assumption that, in spite of the past tendency to increase, tney would stabilize at the levels reached in 1972. These values are shown in Table 6. However, they are sensitive to changes in sectoral growth rates as well as to changes in relative factor prices. - 21 - 3.43 For imports, we use the equations estimated from the past data, but we assume them to be sensitive to relative prices (defined as the overall domes- tic price, divided by the import price index adjusted for the price of foreign exchange). The elasticity is assumed to be somewhat low (0.25) for all imports except for imports of consumer goods where it is assumed to be .50. Exports in constant prices are taken fromL the Government projections. These appear to be optimistic, especially with regard to exports of manufactures to the Community. Similarly, tea projections appear optimistic especially in the light of the general trends in the world tea market. However, in the absence of any better assumptions, we decided to use the government assumptions, which are presuma- bly based on discussions with the appropriate ministries. Since the government projections extend only to 1978, we bad to make some judgmental forecasts of our own to continue the series up to 1985. The elasticity of exports with respect to exchange rate adjustments is assumed to be 0.5 for agricultural products and minerals and 0.2 for other exports. 3.44 One important change we do introduce is to make projections of agri- cultural exports sensitive to changes in value-added in agriculture, from the level obtained in the basic projections. More specifically, we assume that if there is any increase in agricultural output over and above that assumed in the basic projections, the whole oj this increase will be exported. 3.45 As discussed before, the resource requirements for growth can be al- tered dramatically by changes in the terms of trade. In general, since both capital inflows and debt servic:ing are denominated in current foreign currency prices, it is desirable to estimate the resource gap in current prices (i.e., in US$). For this purpose, we start with projections of import prices. Im- port prices in Kenya increased very rapidly (at the rate of about 10 percent) over the years 1970 and 1971. Part of this rise was due to the inflation in developed countries (particularly U.K.) and exchange rate realignments (affect- ing particularly import prices from Japan and Germany, expressed in US$). Hlowever, as Table 7 indicates, import prices rose much more sharply in Kenya than in other countries in a comparable situation, such as Tanzania, Ethiopia, Nigeria and Malaysia. The other factor that might have contributed to the recorded rise in import prices is the over-invoicing of imports connected with the flight of capital. In recent years, the Government has taken severe measures to check over-invoicing. However, it is difficult to be sure of its effects on import prices, which seem to have increased by about 8 percent in 1972. For 1973 we assume a price rise of 5 percent and thereafter 3 percent per annum. 3.46 As regards export prices, we have used the information on projections provided by the Commodities and Export Projections Division of the IBRD, which recently concluded that "The immediate outlook for primary product prices, although clouded by monetary uncertainties, remains generally bright. However, the recovery should be viewed against the background of short price swings periodically experienced by prices of most primary products and the increased pace of world-wide inflation. In any case, buoyant market conditions are likely to be short lived and the weighted average of commodity prices (exclud- ing petroleum) in the seventies is expected to rise by less than half the - 22 - projected rate of inflation in the developed countries." 1/ More specifically, coffee prices are expected to remain relatively strong throughout the seven- ties; tea prices are expected to remain on their secular decline; sisal prices are likely to remain in buoyancy for a year or two and then decline to the trend values by 1980. 3.47 Coffee prices have gone through some well-defined cycles in the past. The recent experience of surplus and deficit phases in portrayed in Chart 2. The typical reaction to high prices leading to over production seems to have been modified by the restraint on investment imposed under the International Coffee Agreement and greater coordination, of national policies among the pro- ducing countries. It is expected that heavy surpluses would not occur in the eighties. The outlook of tea prices continues to be unfavorable. World con- sumption is expected to grow at about 3 percent per annum, and world produc- tion at 3.1 percent. The London average price for all teas is projected to continue its downward trend to a 1980 level of about 38 new pence per kilogram. As shown in Chart 3, Kenyan tea has improved in price relative to other teas in recent years. We assume that in future, Kenya's tea prices will move in line with the prices of other teas. For sisal, the present spurt in prices is expected to be short lived. Our assumption implies that by 1980, the prices will return to a level about 20 percent higher than those in 1970. For other exports - agricultural and non-agricultural - we assume a rate of inflation in terms of foreign currency at 3 percent per annum, which seems to be on the op- timistic side. The export projections in current US dollars are presented in Table 6. 3.48 The next set of assumptions relate to the capital inflows that could be expected to fill the gap in foreign exchange resources. In the first place, we already know the likely disbursements from existing loan commitments. These are prepared by the Debtor Reporting System in the IBRD and the projections are presented in Table 9. Next we have to add to these the disbursements from new commitments. This involves two sets of assumptions: one relating to the levels of commitments by different foreign agencies and the second relating to the disbursement pattern of commitments. They are also presented in Table 9. On the whole, we are expecting a substantial increase in commitment levels of foreign aid agencies, over the next five years. 3.49 From the disbursements from old and new commitments, we have to de- duct the debt servicing outflow. For the old loans, the interest and amorti- zation outflows are given from past agreements. These figures are presented in Table 9. For the new commitments, the debt servicing burden depends on terms of lending. In general, we have assumed that new commitments conform to the same terms as in the past. However, in our sensitivity analysis, we also change these terms and analyze the effect of such changes on debt servicing problems. The levels of foreign private investment and factor income payments 1/ Price Forecasts for Selected Primary Commodities for 1980, (Sec. M73-265), May 4, 1973, IBRD, p. 7. - 23 - from these investments also play an important role in an open economy like Kenya's. We have made rather optimistic projections of net foreign investment on the assumption that the cliLmate for foreign private investment will con- tinue to be favorable. (See T'able 10). 3.50 in;determining prices in our system, the exogenous variable is domes- tic credit creation. The asstmptions of monetary policy in the Plan period have not been stated in the documents on Plan we have received. However, the Central Bank has used 15 percent per annum expansion of domestic credit as a guideline to the banking sector. For our model, we assume that this direction is adhered to and domestic credit grows at 15 percent per annum over the pe- riod 1973-1985. 3.51 For employment projections, the critical assumption is about the rate of growth of labor productivity in different sectors. Our basic assump- tions about this are in Table 11. In general, we have assumed a slowdown in the rate of growth of labor productivity in the future compared with the past. - 24 - CHAPTER 4: PROJECTIONS AND POLICY ANALYSIS 4.01 On the basis of the parameters and assumptions discussed in the preceding chapter, we work out our initial set of projections of macrovariables over the period, 1973-1985. Since these assumptions are a close approximation to those of the Third Plan, the projections for 1974-1978 could-be regarded as approximations to the development in the Plan period. 1/ The projections uD to 1985 are obviously on shakier ground, but these are useful to analyze the long-term consequences of foreign capital inflow on debt servicing capacity and the long-run outlook on employment and poverty. As emphasized before, our interest is not in forecasting the future, but in anticipating the conse- quences of present policies so that the future could be changed, if necessary, by change in present policies. The Basic Scenario 4.02 External and Internal Imbalances. The basic scenario projected under the assumptions of the Plan is given in Table 14. 2/ We note that over the Plan period the annual target rate of growth of GDP at factor cost is about 7.7 percent which is higher than the Second Plan target (6.7%) as well as the actual growth rate over 1967-72 (which was 7.0%). This rate of growth will require gross investment over the period of $3,220 million, which is 26 percent of GDP at market price over this period. The ratio of investment to GDP at market price is not significantly higher than it was in the past. 4.03 However, over the past few years, import prices have been increasing at a rapid rate, and the import requirements in current prices expand at 11 percent per annum. Since exports even at the optimistic assumptions are ex- panding only at 10 percent per annum, this gives rise to a sizable trade gap of about $215 million per year. After allowing for disbursements from old and new commitments, debt servicing and other payments, we are still left with a gap in balance of payments of over $150 million per year. Since Kenya had until recently, a comfortable balance of payments situation, this would mark a turning point in Kenya's development. Kenya may be moving from an "absorptive capacity" constraint phase to "resource constraint" phase. 3/ 4.04 If the required amount of foreign resources could be obtained from abroad, the domestic savings would be adequate on the optimistic assumption that the household saving propensity increases from its low level in 1971 1/ Without making any formal consistency check on sectoral composition, we made some spot checks on output of agriculture and manufacturing. By using 1967 Input/Output Table, we found agricultural output marginally below and manufacturing output marginally above the "consistent" output. 2/ The main results are summarized in Table 12, where for convenience they are expressed in Kenya pounds. 31 Although, as we point out, Kenya has absorptive capacity problems in specific sectors. - 25 - (when it was 7Z) to its average rate of 10 percent. In this situation, there will be no sig,nificant consumption goods gap so that consumer goods prices will rise pari passu with the general price index. If however, the household saving propensity remains at 7 percent, there will be pressure on constuner goods, even after foreign resources are found to fill the trade gap. However, the trade gap is a serious problem. If one looks ahead to 1985, one finds (subject to the limitations of long range forecasting) that it widens ,to about L275 million. This however is only the implication of the target growth rates for the Third Plan period being assumed for the period up to 1985, which it rnay not be realistic to do. 4.05 Eloyment. Even if the external and internal resources are mobilized to attain the target growth rates, the emDlovment problem looks disturbing. The formal wage employment increases to 958,000 by 1978, thus absorbing abou7t 171,300 over the Plan period. However, over this period, labor force is likely to increase by 908,000. Thus the formal sector can absorb only aibouit 19 percent of the increase in the labor force. Even if we include the employment in icformal sector (outside agriculture) the situation is not mnaterially different. The total labor force employed outside agriculture increases from 816,000 in 1973 to 1.08 million in 1978, thus absorbing 24 percent of t'he increase in labor force. The remaining labor force therefore has to be absorbed by agriculture. where em'plovment in- creases fr-M nearly 5 million in 1973 to 5.8 million in 1978. The implica- tion of this for the standard of living of the agriculturalists are discussed below. 4.06 The model also works out the high and middle level manpower re- quirements. These increase from 113,000 in 1973 to 149,000 in 1978, thus implying increase of skilled manpower requirenent of 5,800 per year. It is worth examining (though we have not done it) whether these requirements could be mret from indigenous sources or whether it implies an increased dependence on expatriate staff. 4.07 The Povrerty Index. As discussed before, we take the average income of agriculturalists as an index of the standard of living of the poor, on the assumnption that it is in agriculture that the majority of the poor live and work. Taking the who:Le of non-monetary GDP and 86 percent of mone- tary agriculture 1 / as accruing to the "poor", we find that average income per worker in agriculture was 1543.4 in 1973 (at 1970 prices), and would rise to -b47.2 by 1978. The rate of increase in their standard of living would therefore be only 1.7 percent, which is significantly lower than the overall per capita increase of about 4 percent for total GDP and about 3.5 percent for household disposable income. Thus even if the growth targets are attained, the poor would get relatively poorer over the Plan period and even by 1985, the average income per worker would be only h3.8 in 1970 prices. 1/ See Table 35. Sensitivity of tzhe Basic Scenario to the Criti cal Assumptions 4.08 As imphasized before, the model is only working out the implications. of our various assumptions and it is worth exanining how far the prospects could change if our critical assumptions are changed. We have therefore made a sensitivity analysis of our projections. 4.09 In the first place, it is obvious that our gap projectiorLs are critically deperndent on our asstmptions about, import and export prices. If for example the import prices increase at 2 percent in 1973 (as they did in 1972) instead of at 5 percent (as assumed in the projections) the trade gap over the Plan period could increase by an additional $128 million, and would reach $108 million by the year 1985 (See Table 15). Similarly, if export prices or export earnings increase by one extra percentage point, the resource picture can change significantly. However, while noting the sensitivity of the trade gap to our assumptions about terms of trade, we have to emphasize that our assumptions are on the optimistic side. 4.10 Similarly as regards the internal resource gap, we note that there is considerable uncertainty about the saving-income ratio of the households. If this ratio were to remain at the 1971 level (7%), rather than increase to 10 percent (the average for 1964-71) as ass-med in the projection, there could emerge a significant internal imbalance, reflected in higher consumer prices.t/ 4.11 Our employment projections are dependent on our assumptions about the growth in labor productivitv. However, even on the extreme assumption that the productivity growth rate is zero, modern wage employment would, rise to only 1.2 million by 1978. (See Table 16). By 1985, modern wage employment absorbs about 43 percent of the increas.e in labor force. But, since pro- ductivitv is assumed to remain constant this. increase in employed would probably imply no improvement in the standard of living of the workers in the modern sector. Another possibility is to allow for a rate of growth of labor productivity in the informal sector at a lower level (say 50%) of tiat in the formal sector. This again 4m?roves the employment prospect. However in both cases, agricultu-re still has to absorb the major portion of the in- crease in labor force, and in neither case does the standard of living of the noor grow as fast as average per capita income. 2/ 1/ See Annex 4 for details. 2/ In this connecti7on, it is important to note the differences in the conse- quences of dif-ferent ways of reducing labor productivity in the formal sector. If this is brought about by "labor wasting' techniques, it does not reduce the capital resource requirements for growth. As discussed below, al iporovement in grmoth prospect (through reduced external and internal balance) could Ine brought about together with better employment prospects if changes in labor productivity are brought about by an active factor pricing policy. See A?nendix 5. - 27 - Some Apparent Policy Alternatives 4.12 On the whole it seems to us that our assumptions are quite opti- mistic and the problem of resource gap, employment and poverty has to be tackled by changing policy assumptions. Before we come to discuss some real policy options open to Kenya, however, it may be useful to discuss some policies which look attractive but are mere palliatives in the short run and may be harmful in the long run. 4.13 increase Commercial Borrowing. Faced with an increasing balance of payments gap, the first reaction may be to increase foreign borrowing. This may indeed be the unavoidable alternative in the short run when the country is threatened by a balance of payments crisis. However external borrowing on hard terms is no solution to the basic problem of external disequilibrium, and if resorted to year after year, would soon exhaust its potential and create a creditworthiness problem for the future. 4.14 In order to get a quantitative feel of the problem, we have analyzed a hypothetical situation in which the gap is closed by suppliers credits, which are on relatively hard terms: 9 percent interest and amorti- zation over 6 years. The amount borrowed is assumed to be $50 million in 1973, increasing by $10 million every year, to $180 million in 1985. In this case, the debt service ratio (defined as the ratio of debt servicing payments to export earnings) would rise to 16 percent by 1979, or more than double its present level. It is doubtful whether this increase in the external debt burden would be acceptable to the Kenya authorities. 4.15 It is of course difficult to define a precise point at which increasing external debt creates a creditworthiness problem. However in this case, there are other ways of looking at the problem which suggest its seriousness. On the terms under discussion, the extra cost of debt servicing catches up with the new inflows of suppliers credit by 1981, so that the extra net transfer is negative after that point (See Table 17). Another way of looking at this problem mnay be to compare the current account re- source balance, plus debt servicing payments, with the imports of investment goods. It seems reasonable to assume that except for some temporary short term loans, regular foreign cap:Ltal inflows are by and large used to finance imports of investment goods. Tlus the import of investment goods imposes an upper limit on the gross receipts of foreign capital. When borrowing requirements exceed this limit, it could be regarded as a danger signal. Thus whatever temporary respite it affords, increased foreign borrowing on hard terms cannot solve the kind of external disequilibrium which may be starting to appear in Kenya. 4.16 Impose Import Controls. Another apparently attractive palliative is to impose controls to reduce the demand for imports. However, while this may sometimes prove necessary in the short run, it should not be regarded as the solution of the problem. In our model, we work out the scenario under the assumption that import controls are used to reduce imports from the ex ante level by $50 million in 1973 (at i970 prices) and then increasing by $10 millio- every year up to 95. Thre resu-ks of this exercise are gIven in Table 18. Bef-ore examining rhe table, it is useful to analyze how import controls work in our model. To be,in with, import controls reduce the exter- nal resource gap by reducing imports. llowever this also reduces the total availability of resources, and thus accerntuates the internal resource gap. This leads to higher prices and thus increases ex ante import demand again. Tine extent to which prices rise depends, of- course, on the associated monetary and fiscal policies. I'- foreign borrowing is curtailed, and investment is financed through credit creation (rather tnan tax increases), the inflationary pressure is even higher. iH,owever even if we ignore the problem of domestic inflation, it is worth noting that part of the effec: of import controls is nullilied through the rise in internal.prices, leading to. a further increase in import demand. 1/ 4.17 It seems from Table 18 that, even with a high level of import controls, the problem of the resource gap is not resolved.. In fact, in tnis situation, the restriction imposed upon foreign trade.and domestic inflation (open or suppressed) can follow each other in an unhealthy sequence. This is obviously an undesirable situatLon,, but tnis phenomenon is-not altogether unknown in the developing countries. While Kenya has so far avoided serious problems in either internal and.external balance, the unfortunate experience of other countries does point to the possible dangers which lie. ahead. 4.18 Lower Growth Rates of GDP. If the resource gap cannot thus be reduced by foreign borrowing or i-,port controls, is it desirable to reduce the growth targets? To examine the consecuences of such an alternative, we calculated the effects of reducing all sectoral growth.rates by 3 percentage points. The results are as presented in Table 19. One is that the CDP growth rate is reduced to about 5 percent. This is still a reasonable growth rate, by comparison with many developing countries, and also reduces the resource gap to about $70 million per year over, the Plan period. However for a country with Kenya's poterntial- end past performance, this is not an acceptable altern- ative. In particular, it means that the rate- of growth of employment is slowed down, and it has an even more serious effect on the Doverty index. Whereas in the basic scenario, the poor were getting relatively poorer., in this alternative even their absolute level of living declines by about 1.5 percent per year. l/ In order to measure t-his "slippage" effect, we define an import control slippage coefficien-Lt (SC) as: SC I ( _ _ _E m where: Am = change in imports mex= change in import controls As we notice fLrom Table 14 and 13, about 16% of the effects of import controls are nullif:ied by ir.crease in internal imbalance. - 29 - 4.19 Lower Growth Rate of Labor Productivitv. It may seem that if lower growth rates of GDP are combined with labor-using policies (leading to lower productivity growth), the problem of external imbalance could be tackled without serious consequences for employment. However if these policies mean inefficient techniques (i.e. where more labor is used without any less capital), they do not tackle the problem. In Table 20, as an ex- treme case, we have assumed labor productivity growth rate to be zero. We notice that with slower overall growth, in spite of labor absorbing techniques, the prospects of employment and poverty are not materially imrproved (the income of the poor declines by 1% a year). In fact, it is reasonable to assume that with labor-hoarding, techniques, the propensity to save of the household sector would decline. If we assume a saving-income ratio of 7 per- cent (the level in 1971) the present policy package may give rise to internal imbalance reflected in higher prices. 4.20 It seems therefore that it is desirable to think of policy alter- natives which will increase labor requirements by reducing capital, and especially foreign capital requirement, so that a high GDP growth rate can be combined with steadily improving employment prospects and increases in the poverty index. This is what is discussed in the following section. Some Real Policy Alternatives 4.21 The results discussed in the preceding section show the dilemmas and trade-offs involved in development policy. The resource gap can be reduced by foreign borrowing in the short term, but that increases the debt servicing burden and thus reduces the prospects for continued growth in the future: a trade-off between present and future growth. Imoort controls can reduce the external imbalance but only at the cost of internal imbalance. A slower growth rate reduces the internal and external imbalance, but com- plicates the employment and poverty problem. Labor absorbing techniques could alleviate the unemployment problem, but they would merely accentuate the internal imbalance and poverty problem. We therefore consider whether these dilemmas could be avoidecd if factor prices were used as positive instruments of development policy and the pattern of growth changed. We feel that such a combination could help in reducing the resource gap while also helping the problem of employment and poverty. 4.22 Change in Factor Prices. We begin with an analysis of the effects of a change in the exchange rat:e. To test the possible consequences of a more flexible exchange rate policy, we have assumed a 20 percent increase in the price of foreign exchange. In our model this increases exports, reduces imports and also, by changing factor prices, reduces the amount of investment required for growth. The detailed results following from this change are presented in Table 21. We notice that an increase in the price of foreign exchange of this magnitude would reduce the resource gap over the Plan period to about $54 million per year. Moreover, it will provide extra wage employ- ment to 28,000 persons by 1985 and also improve the poverty index in 1985 bv 0.5 percent. On the minus side, however, it increases the internal imbalance, and thus prices, by reducing the resources available. - 30 - 4.23 A second policy alternative is to vary the rate of interest. More spEcificallv, we consider the effects of increasing typical lending rates by about a third, or from 9 to 12 percent. The effects of such a policy on the economic situation is shown in Table 22. Unlike a change in the exchange rate, this does not influence the external imbalance directly, but only trhrough a reduced demand for imported capital goods, due to change in factor proportions. Yowever its effects on employment are more powerful, and it does not increase internial imbalance to the extent that a change in excharnge rate does. rne resource gap over the Plan period is reduced to $153 million per year, wage employment in 1985 is increased bv 6n,000 persons, and the income of the poor in agriculture rises by slightly more than 1 percent. 4.24 A third policy affecting factor prices could be to abolish the present accelerated depreciation allowances. As discussed before, this would be the equivalent of an 8 percent increase in the cost of capital.. The effects of this measure are similar to those of change in interest rate (See Table 23). For the same growth rate, it reduces investment and import of capital goods, thus relaxing the pressure on the balance of payments, and reduces the resource gap during the Plan period to about $162 million a year. Wage employment in 1985 is increased by 32,000 persons, and the income of the poor in agriculture by 1 percent. 4.25 A fourth possibility is to combine the above three instruments as we have done in Table 24 (where the exchange rate is increased by only 10%). The resource gap over the Plan period is lowered to about $79 million a year, wage employment by 1985 is expanded by 120,000 persons, and the average income of poor in agriculture increases by about 2 percent. The internal imbalance is accentuated slightly, giving rise to an additional 2 percent increase in consumer prices by the end of the Plan. However, it is imoortant to note that while these policies would relax Kenya's resource constraint and help the employment and poverty policies, they do not solve these problems completely. Even under the corbined package, the resource gap over the Plan period is sizeable and wage employment by 1985 absorbs only 23 percent of the increase in labor force. Moreover, the improvement in the standard of living of the "working poor't in agriculture is still painfully slow. Clearly, there is need for additional policy changes. 4.26 Change in Pattern of Growth. The next important set of policy changes that we consider, therefore, are thiose relating to changes in the pattern of growth. hore specificaliy, we are concerned with the implications of an increase in the growth rate of the monetary agriculture sector (AG) as well as the non-monetary sector (NM) - associated with a decrease in growth rate of the principal infrastructure sectors, namely, building and construction (BLD) transport, storage and communication (TRAN) and electri- citY and water (ELEC). 4.27 To begin with, we consider the separate effects of changes in growth rates in each of the sectors mentioned above. The results are given in Tables 25 to 29. We also notice from Table 31 that dollar for dollar, the benefits of extra GDP through NM and AC are higher, in terms of employ- ment and the poverty index and the costs lower, in terms of internal and external resource imbalance. 4.28 The next exercise is to consider the effects of changing the pattern of growth as a whole. The resuLts, given in Table 30 are encouraging: although the resource gap arising over thee plan period is still considerable, wage employment in 1985 is increased by 10,000 persons and the Doverty index improves by 11 percent. This increases the standard of living of the "working poor" by 2.6 percent a year, which is close to the growth in per capita income for the country as a whole. 4.29 Preferred Strategy. On the basis of the above results, it seems that factor price changes and structural changes are complementary policy instruments. Changes in factor prices (particularly the exchange rate) are particularly valuable in reducing external imbalance, while employment and the income of the working poor are influenced more by structural changes. W4e therefore combine the two packages and the results of this exercise are given in Table 14 and the main indicators summarized in Table 13. Under the combined effect of all these policy changes, the external resource gap is reduced to $38 million per year, and additional wage employment is provided for 135,000 persons by 1985. The income of the poor in 1985 is 13 percent higher than under the basic scer,ario. The annual rate of increase in their income is now about 2.8 percent a year over the plan period. This would mean that the average level of the Door would be $170 (at 1970 prices) per worker by 1985. 4.30 To sum up: the combination of policy changes we have assumed in our "prferred" strategy looks very encouraging, except for the remaining resource gap which would not be filled under our assumptions. However, in the opinion of the Mission, Kenya would have a good prospect of attracting additional external aid (above that assumed in our projections), if she succeeded in achieving this high level of performance. 4.31 We realize that due to administrative as well as technological time lags involved, it may not be possible to change growth rate in different sectors by 1974 as assumed in the model. Our results are only illustrative of the effects of reformed strategy over a given five year period, and the Bank will be undertaking a more systematic assessment of the prospects under Kenya's 1974-78 development plan once the published document is available. - 32 - CHAPTER 5: ON SOM{E DEVELOPM^ENT POLICY ISSIES 5.01 In this Chapter, we comment briefly on the implications of the results of our model for some important issues in development policy, such as the role of foreign capital, the trade-off between employment and growth, the nature of poverty-focussed growth, the role of factor prices in develop- ment, and the possible trade-off between Kenvanization and growth. Needless to say, each of these issues has many complicated aspects - econonic as wel as socio-political - and our model is not geared to an intensive analysls of these issues. As emphasized before, models specifically redefsigned for these problems would be necessary, even to discuss the purely economic aspects of these issues. However, in spite of these shortcomings, our model does have some implications for these issues and especially may help to clarify a few miscorLceptions. 5.02 Foreign Capital and Growth. In recent years, the role of foreign capital in economic development has come under increasing scrutiny. In par- ticular, it has been recently argued, 1I that foreign capital inflow reduces domestic savings, and thus reduces the growth prospects of a country. The conclusion is generally based on a single equation regression analysis of domestic savings on GDP growth (or GDP level) and foreign capital inflow. Our analysis of the Kenyan economy suggests that, in the context of systems analysis, this approach could be misleading, Decause both savings and GDP growth rates are jointly determined within the system, with the exogenous variables such as foreign capital inflow and others given. In this context, if foreign capital inflow changes, savings, investment, and income all change, and the single equation regression, which examines the effects of foreign capital inflow on savings, given the income level, is therefore misleading. 5.03 For a more satisfactory arnalysis, one has to see how the system as a whole moves under alternative levels of foreign capital inflow. This is done in a heristic fashion in our model and the results are contrasted in Table 15 of Annex 4. We note that with a lower level of foreign capital iLn- flow and a given policy package, the level of savings, investment and income would have been lower, even though the partial effect of foreign capital on saving-income ratio were negative. 5.04 However, the above analysis does not take the effects of repayment of foreign capital on future growth into account. In order to examine this aspect, we have to compare two macro-econormic scenarios over the period during which a loan is disbursed and repaid, an/see if the discounted present value of the target variables is hig,her with an increased flow of foreign capital. However, this requires some redesigning of our model and we have not been able to do this so far, although we hope to analyze this later on. I/ For details, see the Appendix to Annex 4. - 33 - 5.05 The above discussion does not suggest that foreign capital inflow is always good. In fact, the results of our model suggest strongly that if foreign borrowing on hard terms: is used freely to fill the resource gap, it could create problems of creditworthiness in the future and thus reduce growth. Similarly, if the availability of foreign capital acts as a soft op- tion and diverts attention from rectifying fundamental problems of price dis- tortions and structural distortions, it could be harmful to the country. How- ever, these kinds of effects have to be discussed country by country by con- sidering institutional and political elements, and cannot be estimated by a single equation regression approach. 1/ 5.06 Employment and Growth. The results of our model suggest that the formulation of a two-dimensional trade-off relationship between employment and growth is misleading. This relationship may be valid in some idealized "production possibility frontier". However, most developing countries are well within this frontier, and it all depends on the type of policies followed whether employment and growth have trade-offs, or are neutral to each other, or could both increase at the same time. As discussed in Chapter 4, if the growth is constrained by fore exchange, and employment-creating techniques are introduced w'hich do not affect the foreign exchange gap, growth is neutral with respect to employment. If, on the other hand, the internal resource gap is the binding constrain, and employment-creating techniques reduce the saving propensity, the effort to increase employment could reduce growth. Hiowever, if employment-oriented policies work through factor price changes, which in- crease the demand for labor and reduce capital and import requirements, it may be possible to increase both employment and growth at the same time. Similarly, if the employment problem is tackled through changing the pattern of growth, such that it reduces demand for capital and imports, it is again possible to have more of both employment and growth. As Tables 12 and 13 suggest, and according to our analysis, Kenya's case fits in this category where, with corrrect policies, employment and growth can both be helped. 5.07 Poverty and Growth. Our conclusions on poverty and growth are rather similar to those on employment and growth. If incomes-oriented growth means transferring resources from the urban rich to the urban poor, there may be a conflict between growth and redistribution. However, the majority of the poor in Kenya (as in most other developing countries) live in the rural areas and make their living from agriculture. In general, the propensity to save of the agriculturalists - rich and poor - tends to be high and an increased emphasis on increasing the productivity and incomes of the rural poor (particularly the agricultura:Lists) should not reduce overall saving propensity. Equally important is cthe point that income generation in agri- culture requires less capital and :Less imports, and thus reduces the foreign exchange gap for the same rate of growth. As Tables 12 and 23 show, it is possible to improve conditions of t:he poor and have a higher growth if struc- tural changes in patterns of growth are achieved. Table 19 brings out the point that a lower growth rate would undoubtedly hurt the poor. However, a 1/ For a more detailed discussion, see the Appendix to Annex 4. - 34 - higher growth rate, while necessary for any attack on poverty, is not a suf- ficient condition for helping the poor. The answer lies in restructuring growth. The theme of our "preferred" strategy is therefore not to benefit the poor at the expense of development as a whole, but to "attack poverty through restructuring growth." 5.08 Role of Factor Prices. The role of prices - both product prices and factor prices - has been very much neglected in postwar planning dis- cussions. However, the postwar experience of many developing countries, as well as socialist economies, is beginning to emphasize the need for a "rehabil- itation of the invisible hand" as a useful allocative instrument, though not as a kingpin of economic policy. Our results suggest that, if used as positive instruments of economic policy, factor price changes could relax the con- straints on economic growth. Thus, instead of assuming that developing coun- tries are in a straitjacket and that the whole of the resource gap has to be filled by foreign aid, one could also look at internal policy instruments to see how far the gap can be narrowed befroe asking for more foreign aid. As Table 26 indicates, these instruments may not solve the problem of resource gap or employment; but they are steps in the right direction. 5.09 Kenyanization and Growth. We have suggested elsewhere in the report that there may be some trade-off between a rigid application of Kenyaniza- tion and an uninterrupted rate of growth, particularly in some segments of the private sector. It is, of course, primarily a political choice how much Kenya wishes to depend on foreign skills in the future, even if they can help to ensure a faster rate of development. However, from the viewpoint of econ- omic analysis, it is useful to know how the possible conflict between Kenyan- ization and growth could be minimized. Our results suggest (see Table 30) that restructuring growth, along the lines discussed, should reduce the need for high middle level manpower and thus help the process of Kenyanization. PART II THE TECHNICAL APPENDICES APPENDIX 1 A SIIWLIFIED ALGEBxtlTC DESCRIPTIO OF THE MODE- 1. In this appendix, we present a simplified algebraic description ot the model to highlight ^.ts structure and the method in which we use it for policy discussion. The details of each of the subsectors are presented in Lhe subsequent appendices. I. The Equations Output and fnvestment gi gi ki (ai + P'i) (1 + x) ai gi I =g*k*Y where gi =arget growth rate of G M for sector i k. ICOR for sector i x policy package affecting factor prices i=nvestment Y. = GDP in sector i Imports, Exports and Balance of Payments mr = ca1 +% ( iYii- 7(-j-) mc = a 2 + 62 ( i iYi )7< fc f 2 m mc = a + 3 ( WiIj)7(P) A3 o A t m = Ae m = mr + mc + mi + mg + m° + m PXx = (X°) (f) n GAP - p m - p X - Z APPENDIX 1 Page 2 r where m = import demand of raw materials m = import demand of consumer goods i m = import demand of investment goods mg = import demand by government m = import demand of other goods and non-factor services m = exogenous change in imports reflecting import controls (when mX is - ve) or imports to take care of local shortages (when m is + ve). The latter is equivalent to the saving con- strained case in the usual two-gap approach. m = total imports ex post Pm = import price index p = domestic price index pc = domestic consumer price index G = GDP in government t = time p X = exports in current US$ f = exchange rate X = exogenously given exports Z = capital invlows reflecting gross foreign lending minus debt servicing payments, etc. Credit, Money, Consumption and Prices M = DC0 + F E = f(M) p = E/(Y + m - X) HC = f(YD) YD = f(Y) APPENDIX 1 Page 3 GC = f (G) C = Y + m - X - I - GC PC = HC P C where 1i, = money supply DC = exogenously determined domestic credit F = foreign assets E = total expenditure in nominal terms p = price level HC = ex ante household consumption YD = disposable personal income GC = government consumption C = ex post household consumption pc = consumer prices Em_ loin-ent and Poverty AWEi = WEi (gi - PR.) (1 +- x) i -1 SE1 = Ai WEi ATE. = TE (gi - PR')(1 + x) i L = L (1 + .036) LA = L - ETE1 pov = YA WE. = wage employment in sector i SE. = skilled manpower required in sector i TEi = total employment (including informal) in sector i (except agriculture) APPENDIX 1 Page 4 PR, PR' = rate of growth of labor productivity in formal sector and total economy, respectively L = labor force LA = residual emplovment in agriculture YA = GDP in agriculture pov = per worker value added in agriculture taken as an index of poverty II. The Process of Adjustment of Internal and External Imbalances 2. In the general equilibrium system, both these imbalances are recti- fied by price changes--the external through exchange rate and the internal through domestic price. In the two-gap model, neither of these balances are rectified, both foreign resources are brought in to fill the larger of the two gaps, thus presumably creating excess supply for the market with the smaller gap. In our system, the situation is in between the above two ex- tremes. The internal imbalance is cured by price rises--although the price rise may prove to be too high in the light of social objectives. If so, foreign resources are brought in (positive m ) to relieve the pressure. However, the extent to which domestic inflation and m are acceptable is a policy decision, explicitly considered (in contrast to the general equi- librium approach, where m is zero, or the two-gap approach, where acceptable price rise is zero). If there is an external imbalance, it could be cured by an inflow of foreign capital, import controls, a change in the exchange rate, or other policies, each of which has different implications for the future. The choice is made outside the model, rather than hidden inside, as in the general equilibrium approach, where the exchange rate rises or falls to bring about equilibrium, or in the two-gap model where foreign capital inflow in- creases to fill the gap. 3. By bringing the instruments out into the open, our procedure helps in the discussions of various policy alternatives and their consequences. This is elaborated in the following section. III. The Method of Operation of the Model 4. In this system the instruments are: gi, x, f, m , DC APPENDIX I Page 5 The arguments in the objective function are: GAP, pc, p, poa, TEi Given a set of values of the instruments, the model works out the values of gap, inflationary potential and poverty levels for different years. 5. If gap is big, there are several policy options with associated consequences: A. increase Foreign Borrowing This solves the gap in the present but causes creditworthiness problems for tne future. B. Impose Import Controls (thus reduce mX) This, however,,increases inflationary pressure and also increases m so that the net effect on the 6-p is less than the reduction of m . C. Cnange Factor Prices Through x and f This reduces gap and improves employment pros- pects but is associalted with higher prices. D. Reduce Growth Rates This reduces gap, but also reduces employment prospects and worsens the poverty situation. E. Change the Structure of Growth Rates This also can r-educe gap depending on the type of structural change. Employment and poverty impli- cations could also be good depending on the pattern of change. 6. Similarly, if the rate of increase in employment and in the income of the poor is slow, various policy options could be examined, along the lines discussed above. It is important to note that in our scheme, the vari- ous problems and policy alternatives are discussed and the desired policy package is chosen in a judgmental fashion, rather than formal optimization through mathematical programming. APPENDIX 2 AN ANALYSIS OF ICORs IN KENYA 1. In spite of the numercous conceptual and statistical problems, 1/ the use of incremental capital output ratios (ICORs) remains popular in models depicting growth in developing countries. Among the reasons for its popularity are: limited data requirement (they do not require figures on capital or its structure, for example), simple interpretation, and ease of use in discussions of resource requirements for growth. In our model, we have used the ICORs but tried to refine them so as to take care of their more obvious weaknesses. More specifically, while using ICOPRs we have tried to assess the impact of depreciation allowances, sectoral composition, and factor prices on ICORs. 2. It is usual to compute ICORs by comparing changes in GDP with lagged gross investment figures. Iowever, as Chenery and Eckstein 2/ argue, this procedure ignores the fact that part of gross investmenc is ror re- placement purposes and does not contribute, therefore, to any increase in output. They also demonstrate that if this factor is taken into account, it is clear why "capital-output ratio . . . should be expected to be (and almost always is) lower at higher rates of growth." Following their pro- cedure, we can write the equationi for ICOR as: (1) k = k + 2 r where k = gross ICOR 1 k = net ICOR showing the effect of new investment on output Z = share of' current income devoted to replacenment 3/ r = rate of growth of GDP. This form of equation was also adopted in the model for Kenya presented in the IBRD Economic Report of 1969. 4/ 1/ See, for example, Paul Streeten, The Frontiers of Development Studies Chapter 6, John Wiley and Sons, New York, 1972. 2/ H.B. Chenery and Peter Eckstein, "Development Alternatives for Latin America," The Journal of Political Economy, July/August, 1970. 3/ In Chenery-Ecksteini formulation this also includes investment on social overhead investment. 4/ Economic Development Prospects in Kenya, IBRD Report No. AE-6a, October 22, 1969. AkPPENDIX 2 Page 2 3. As we found by Chenery and Eckstein, an unconstrained regression of equation (i), often leadi to implausible resultes for the parameters (e.g., negative values of k ). The procedure adopted by Chenery and Eckstein was to constrain k to 2.0 and obtain estimate of Z from regression. However, since we have some figures on depreciation allowances, we d2cided to use these estimates of Z. It seems to us thaY in most of the cases, - would account for a smaller part of k than does k and it is desirable to predetermine the smaller component. The values of Z for different sectors were computed from the Input-Output Table for Kenya, 1967 and these estimates are presented in Table 33. 4. Apart from the allowance for depreciation, we have a problem with changes in net ICORs over time. It is suggested that, in Kenya, productiv- ity of investment has been declining over the years, and our formulation of the problem siould be able to test this hypothesis. in order to test this we obtained k as k and then ran regressions of k on time. 5. As noted by Chenery and Eckstein, annual incremental capital-output ratios can not always be used as the units of observation in a time-series. Year-to-year movements reflect the effects of changes in degree of utilization of capital as well as in its amount. In order to tackle this problem, Chenery and Eckstein "identified . . . cycles and obtained single incremental ratios for each one by aggregating the total investment and change of GNP over the period." In the absence of any identified cycles, we used three-year moving averages as a smoothing device for obtaining smoothed values of GDP and investment. The equations thus obtained for different sectors are given in Table 34. The actual and estimated values of ICORs are presented in Table 2. The interpretation of the trends in ICORs and the analysis of their possible reasons are given in Part I (pp. 8-9). 6. The analysis of sectoral ICORs is useful for studying the trends in ICORs and for analyzing the impact of patterns of growth on overall ICOR. However, in using the sectoral ICORs for sensitivity analysis with respect to sectoral growth rates, a problem is created by the interdependence of different sectors. Output in one particular sector uses the services of other sectors, so that an increase in output in one sector will also require an increase in output of other sectors. For example, an increase in value added in the agricultural sector requires increases in output and investment in transport, construction, government services, and so on. In order to take this factor into account, we had to utilize the input-output table. Let b be the value added required in sector i for a unit of final value added to1i sector i. Then G percent of extra growth in sector j will also require G percent rate of growth in i sector where Gi is given by G ;Gij b a (i i) .n) APPENDIX 2 Page 3 For computation of bi we used t:he total input coefficient matrix given in Input-output Table f4 1967, and adjusted these coefficients for the ratio of value added to output in each sector. 7. In the main report we have emphasized the effects of an accelerated rate of growth of agricultural output, and a decrease in the rate of building construction and transport communications sector. Some illustrative calcu- lations of GDP and investment required in each sector to raise GDP by one percentage point are given in Table 31. 8. ICORs are generally treated as purely technological parameters. Actually they reflect the effects of various economic decisions relating to degree of utilization, factor Droportions, rate of technical progress and efficiency of allocation. With the limited amount of data available, we could not estimate the separate effects of the various factors mentioned above. However, on the basis of certain assumptions about e"asticity of substitution, we do obtain a link between ICORs and factors influencing relative factor prices such as wage rate, exchange rate, interest rate, and investnment allowances. Since this involves changes in employment generation also, it is discussed separately in Appendix 5 below. 9. The above analysis relates to fixed investment only. The rate of inventory accumulation is assumed to be 28.8 percent of the increase in monetary GDP. 1/ In practice there has been two inventory cycles in Kenya (the first one peaking around 1967 and the second around 1971). However, for long-run forecasting we did not try to analyze these cycles. The total investment required is given by the sum of fixed investment and inventory investment. 1/ Based on estimates of the Ministry of Finance and Planning. APPENDIX 3 AN ANALYSIS OF ThPORT REOUIREMENTS INCORPORATING THE EFFECTS OF GDP COtMPOSITION 1. In most developing countries growth is probably constrained by the availability of foreign exchange, and an analysis of import requirements thus becomes a critical element in assessing growth prospects. However, it is difficult to define a simple pa:rameter like ICOR for measuring import intensity of growth. The elasticil:y of imports with respect to GDP is sometimes used as such a simple parameter. However, this fails to capture the important point that in developing countries investment may have a higher import content than non-investment items in GDP, and an acceleration of growth through an increased investment-income ratio may increase the elasti- city of imports with respect to income. It is therefore necessary to analyze import data in a more disaggregated fashion. Fortunately, some disaggregated data on various types of imports are available in Kenya, and we estimated separate import equations for constnner goods, raw materials, investment goods, government imports, and imports of non-factor services. 2. In our macro model, we are particularly interested in working out the effects of changes in sectoral patt:erns of growth. Therefore, we tried to incomporate the differential import requirements for imported capital goods, raw materials, and so on for growth in different sectors. The main source of information for imported raw material requirements was the Inpt/ output Table for Kenya, 1967. Hotever, instead of using input/output (I/O) information as representative of the entire sample period, we tried to use both I/O information and the aggregate time series information. 3. In a straightforward use of I/O table, the imports would be esti- mated by equations such as (1) M = EmiYi where M = imports mi = import of content of sector i Yi = output of sector i On the other hand, purely aggregated analysis uses equations such as: (2) Mt = a + SY + 6p where Y = aggregate output P = price variable It is obvious, however, that equation (1) misses out on the effects of overall factors, such as price changes or technological changes over time, and equation (2) misses out on the e(ffects of sectoral changes. APPENDIX 3 Page 2 4. The equations we used are of the following type: (3) Mt = a, + m (miyi)+ tp I i t The time-series for Eim y was constructed bv using I/O values of mi and GDP time-series relat ng lo sectoral y s. Instead of assuming that there is no change over time in I/O coefficients, we allow the regression to capture any overall changes in these coefficients (though not in relative changes), as well as the influence of factors such as prices. Depending on data and regression results, equations (1) and (2) could be "particular cases" of equation (3). If = 1 and a' = 6 =0, we have the I/O case; if mi = m0 for all i's, we have the aggregate time-series case. 5. Apart from being more general, equation (3) also enables us to separate the effects of changes in sectoral composition from that of change in overall GDP. Thus if the equation (3) shows x percent increase in imports due to the variable Em yi, the contribution of sectoral change in GDP is y k = x (1 - z) where y is the percentage change in aggregate GDP Z is the percentage change in imiyi 6. In order to use the above procedure, we had to compute, from various sources, the relevant indicators of differential impact of sectors for different categories of imports. In the case of import of raw materials, we obtained the ratio of imported raw materials to GDP in each sector from I/O table, 1967. For capital goods, there was no breakdown by sector of imported v. non-imported capital goods. However, the Kenyan series of Economic Surveys give the figures of investment by type (construction, trans- port equipment machinery, breeding stock, etc.) for each sector separately. We assumed that transport equipment and machinery is mostly imported and used the ratio of these items to total investment as the indicator of import content of investment by different sectors. For imports of consumer goods, the problem of data was even more serious. However, income tax department publishes figures of income assessed by sectors. We assumed that those whose income is not assessed for income tax purposes are not significant users of directly imported consumer goods, and we therefore used the ratio of income assessed for tax purposes to GDP in each sector for construction composition variable. These ratios are given in Table 35. 7. In estimating import requirements, we hoped to capture the effects of prices of imported goods (and thus the exchange rate) on imports of dif- ferent categories. Unfortunately, in our regression equation, we failed to capture any such effect. The main problem was that prices - both imported APPENDIX 3 Page 3 goods and domestic - have been stable for the major part of the sample period. The only noticeable price rise was in 1970 and 1971, but this may have been partly associated with over-invoicing due to capital flight, and probably was not a wholly genuine price rise. Trhe estimated equations are given below. Import Functions 1. Import of Consumer Goods MC = 31.34 + .155EW Y (2.17) (3.4) -2 R = .66, D.W. 2.0, SEE = 6.88 W 's are the weights as given in column 2 of Table 5 i Yi = GDP :Ln sector i 2. Import of Raw Material 1 MR = 13.58 + .958EW Y (.66) (9.18) l l -2 R1 = .93, D.W. = 1.50, SEE = 10.3 W 's are the weights as given in column 3 of Table 5 3. Import of Capital Goods * MK = -3.61 + .62 SW II (-.4) (8.29)ii 2 R*= .92, D.W. = 1.90, SEE = 6.25 W 's are t:he weights as given in column 4 of Table 5 I= In vestment in sector i. 4. Import of Government MG = 3.79 + .07 GG (.54) (1.80) -2 R = .35, D.W. = 1.55, SEE = 3.99 5. Import of non-factor services log MO = 4.23 + .023t (101.0) (2.7) -2 R =.55, D.W. =1.76, SEE = .05 APPENDIX 3 Page 4 8. The estimated values of I miyi for different categories, as well as the actual and regression estimates of different categories, are presented in Table 36. The level of imports at constant prices in 1970 and 1971 was higher than estimated, even when we ignore the effect of prices, and this indicates a shift in the import equation. W4hether this is temporary or lasting has to be decided on the basis of onets judgment. 9. In spite of our inability to capture price effects from the data due to random influences, it seems inadvisable to ignore the price effects completely in projection. We have, therefore, introduced some purely sub- jective (probably conservative) estimates of elasticities of imports with respect to relative prices defined as (import price x exchange rate index)/ domestic GDP deflator. These estimates are 0.5 for consumer goods and 0.25 for other categories. APPENDIX 4 ANALYSIS OF SAVINGS OF HOUSEHOLDS, BTJSINESS AND GOVFERNMENT SECTOR 1. In the usual growth models for developing countries estimates of savings are based on an overall marginal propensity to save with respect to income. Ve disagree with this approach. In most countries - developed as well as developing - a significant part of gross savings comes from deprecia- tion allowvances, and these cannot be assumed to be simple function of income but are dependent on durability, age structure of capital, and tax laws. This part of savings should therefore be estimated separatelv. Similarly for the rest of the economy, government saving behavior cannot be analyzed in terms of the usual consumption function theories - relative income hypothesis, permanent income hypothesis or life cycle hypothesis. Government's recur- rent expenditure is influenced by its past capital expenditure in ways that cannot be measured by household type consumption functions. 2. Fortunately, we found that the Ministry of Finance and Planning in Kenya has been following the disaggregated approach and has already developed equations for savings of households, government and business. For most of these equations published data do not exist, and we therefore decided to use the equations developed by the Ministry as a starting point. ThMese equations do not incorporate any effects of sectoral composition on savings behavior and this type of effect is important for our analvsis. However, we could not get any data for this purpose, and we had to ignore this effect in our analysis. We hope that a later stage we may come back to these equations and reestimate them as necessarv. For the time being, we reproduce the equations used in the Ministry's model, after converting them to 1970 US dollars. 3. Aggregate savings is divided into four parts: household savings, undistributed profits of businesses, depreciation allowances and government savings: (1) TS = HS + SD + UP + GS where TS = total savings HS = household savings SD = depreciation alLowances UP = undistributed profits GS = government savings APPENDIX 4 Page 2 4. Household Sector - disposable income of the household sector is defined as follows: (2) YS = N + WB + WG + WH + FH - TH - VHR + VRH + VGH where YD = disposable personal income N = non-monetary value added WB = wages paid by business WG = wages paid by government WH = wages paid by households FH = interest, dividend and other residuals received by households TH = personal taxes VHR = transfer payments by households to rest of the world VRH = transfer payments by rest of the world to households VGH = transfer payments by government to households The equations for each of the above elements are given below: (3) WB = 1.42 * (FB + SD) (4) WG = (1 + YGGR) WGti1 (5) WU = 5.71 + .005 MGDP (6) FU = 27.27 + .18 MGDP + 1.96 t (7) TH = THO + TG (8) THO + 31.25 + .086 (MGDP - SD) (9) TG = 8.4 + .005 MGDP (10) VHR = 24.64 + 1.4 t (11) VRH = 20.44 + .84 t (12) VGH = 24.14 APPENDIX 4 Page 3 where FB = other factor incomes paid by businesses MGDP = monetary GDP t = time trend THO = income tax other than companies TG = graduated income tax SD = depreciation allowance YGGR = growth rate in GDP in government sector 5. Household consumption is expressed as a fixed proportion of dis- posable income in the light of the secular constancy of consumption-income ratio in the developed countries, for which data over long historical periods exist. (13) HC = .90 YD Household savings are then a residual: (14) HS = YD - HC 6. Business Sector - The undistributed profits of the business sector are determined by the following identity: (15) UP = MGDP - WB - WG - WH - FB - SD - TD - P - YG where UP = undistribul:ed profits MGDP = monetary GDP WB - wages paid by business WG = wages paid by government WH = wages paid by households FB = other factor incomes paid by business SD = depreciation allowances TD = direct taxes on businesses P = profits of government corporations and sales of government services YG = imputed government income APPENDIX 4 Page 4 Direct taxes on businesses are further broken down into three categories: (16) TD = TBY + TCD + TQD where TBY = income tax on companies TCD = other direct taxes of Central Government TQD = other direct taxes of municipal and county councils Profits of government corporations and sales of services are divided into two parts. (17) P = SF + PR where SF = school fees PR = other incomes of government corporations 7. The equations for WB, WG and ITH are given in the household sector as equations (3) - (5). Other equations are as follows: (18) FR = 31.61 + .19 MGDP (19) TBY = .0583 E (20) TCD = .98 + 0.56 t (21) TQD = 1.18 + .008 MGDP (22) SF = 7.6 SEN (23) PR = .84 + .049 MGDP (24) YG - 0.859 GDPG where E = net domestic product of enterprises defined as (monetary GDP) minus (value added by government) minus (depreciation allowances). SEN = school enrollment GDPG = GDP in general government 8. Figures for depreciation allowances do not exist in the published national accounts of Kenya. Estimates prepared by the Ministry of Finance and Planning are based on tax rates applied to published investment figures (with some heroic assumptions of depreciation before 1964). For forecasting the depreciation element due to investment before 1972, C(t) is estimated exogenously and for other subsequent years an equation is postulated on the basis of tax laws and past investment levels (I). The equation is: APPENDTX 4 Page 5 (25) SD(t) = C(t) + .038 Irt + .077 (-t+t- It-2 +It-3 + .061 I t4 + .042 1 5 + .039 t 6 9. Government Sector - Government savings (GS) are given by the following identify (26) GS = TD + TH + P + TI + VRG - FG - U - VGR - VGH - GC where TD = direct taxes on business TH = taxes on households P = profits of government corporations and sales of government services TI = indirect taxes on business VRG = transfer receipts, rest of world payments to government FG = government interest payments U = business subsidies VGR = transfer payments, government to rest of the world VGH = transfer paynments, government to households GC = government consumption 10. Indirect taxes (T1) are broken down in seven categories: (27) TI = TS + TM + TE + TP + TOT + TQF + TAO where TS = sales tax TM = import taxes TE = excise taxes TP = petrol and diesel taxes APPENDIX 4 Page 6 TOI = other indirect taxes and fees TQF = licenses, cesses, fees, property income and interest or municipal and county councils TAO = other revenues (retentions by East Africa Common Services Organization) The equations for each of these are given below. (28) TS = 38 for 1973 = 48 for 1974 = 1.125 TS(t-1) for 1975 on (29) TM = 33.49 + .129 (M - MR) (30) TE = -8.9 + .049 MGDP (31) TP = -2.8 + .008 MGDP (32) TOI = -.48 = .011 MGDP (33) TQF = .009 MGDP (34) TAO = 1.68 + .168 t (35) VGR = 10.61 - 1.288 t (36) FG = 19.6 + 1.96 t (37) VRG = 33.74 (38) GC = -.98 + 1.19 GDPG APPENDIX 5 FACTOR PRICFS, E!?LOYM-ENT AND GROWTh 1. In most growth models for developing coimtries, the factor propor- tions are taken as technological parameters, and it is implicitly assumed that factor prices are not important for determining emplovment and growth pros- pects. WIhat is often not realized is that apart from affecting choice of techniques at a micro level, fact:or prices are also important in influencing overall capital-labor ratio through change in produce mix, change in degree of utilization and change in type of labor and capital. 1/ As Gordon C. Winston 2/ has demonstrated. "in order for this to be true that there be no response in factor use to changes in factor prices - it is clearly necessary that simultaneously 1. Factor use is not affected by product mix, eith.er because there is anly one produce in the economy or because the consumotion of output is fixed and immu- table so its use of factors is not variable by varia- rions in product mix--this is effectively one com- posite product, AN D 2. Factor use is not affected by choice of techniques, because there is only a single kind of expost plant, that can produce fcor each of the industries that make up the sector or economy, AND 3. Utilization is not affected by factor prices, because expost all capital is utilized at its engineering maximum though that maximum is much greater than opti- mum--desired levels of utilization suggested by all empirical studies of capital utilization, AkND 4. The crew required to operate each niece of capital is everywhere inflexibly set so increasing the number of workers will never increase the rate of production and decreasing the number will always idle capital." 1/ The variation in type of labor and capital may be due to differences in degrees of "human capital" in labor and in degrees of "foreign exchange" in different types of capital. 2/ Gordon C. Winston, On the Inevitability of Factor Substitution; (mimeo) Research Memorandum No. 46, Center for Development Economics, Williams College, Mass., April 1972. APPENDIX 5 Page 2 These condlitions are clearly too stringent to be satisfied even by developing countries. 2. On an empirical level, the statistical studies in developed as well as developing countries give significantly positive values of the elas- ticities of substitution in factor proportions with respect to factor prices. Henry Bruton 1/ after an extensive survey of the statistical results on elas- ticities of substitution concludes: "The most general conclusion is the most important: Factor substitutability is alive and well in developing countries. Policies and models that assume otherwise mislead." More specifically, his review of the results of statistical studies in U.S., Philippines, Argentina, Chile, El Salvador, Korea, Paraguay, Peru, Portugal, Spain and Mexico found that the elasticity varies between .5 to 1.6 and is generally significantly different from zero. Similarly, after examining U.S. data on an industry level, P. Zarembka 2/ concludes" ... there is no significant evidence that the elasticity of substitution at the two-digit level departs from unity using either behavorial equation or direct estimation ... the estimates pre- sented here lie on both sides of unity." 3. In Kenya data on output, labor and capital are not available by industry for fitting an explicit production function. However, empirical analysis by the Mission of the figures of labor productivity and wage earn- ings gives estimates of elasticity of substitution which are not significant- ly different from unity. As shown in Table 37, the elasticity of substitu- tion (as measured by the elasticity of value added per employer to wages) was .988 (+ .19) for commerce, and .905 (+ .14) for other services. For the private modern sector as a whole, the elasticity was 1.19 (+ .12). 4. Thus in the light of empirical evidence on production function in developed countries as well as in developing countries, it seems reasonable to assume unitary elasticity of substitution i factor proportions will re- spect to factor prices. We also use the share of wages in gross output as an approximation to the elasticity of output with respect to labor (a ). Lastly, we assume that existing factor proportions are not malleable and these effects apply to new capital and labor only. Then if a policy package changes the cost of capital (r) relative to wage (w) by IOOX percent new ICORs and incre- mental employment-output ratios (e) will be: (1) ICOR = ICOR *( + x) i ~~~i i 1 1 (2) ei e° (1 + Xi) Where: subscript i's indicate sectors 1/ Henry J. Bruton, The Elasticity of Substitution in Developing Countries, Research Memorandum No. 45, Center for Development Economics, Williams College, Mass., April 1972. 2A Paul Zarembka, "On the Empirical Relevance of the CES Production Func- tion," The Review of Economics and Statistics, 1970. APPENDIX 5 Page 3 Superscript indicates t:rend values in the absence of policy changes as discussed in the earlier part of the paper. 5. The calculation of X.s will be done outside the model and our broad scheme is explained below: X is defined as (3) X = r - w r w where: = proportional change in cost of capital r = proportio:2al change in cost of labor w capital cost (r) is def:tned as follows: (4) r = pnv (1 - It) ( d + i _ pe) where: Pinv= price of investment goods t present value of tax benefits through accelerated depreciation allowances as a proportion of price of investment goods. 6 depreciation rate i = nominal rate of interest p expected rate of inflation 6. The influence of exchange rate change will be felt to the extent that imported capital goods enter into price of investment goods. Assuming no change in prices of domestic investment goods and rate of depreciation, X can be written as: (5) X = -w f X + ii - pe pe t w f I D pe D 1-t where AX = proportion of imported goods in total investment D = d + i _ pe f = exchange rate APPENDIX 5 Page 4 7. Ideally, it will be desiralble te ~lww1ulate X 53 for different sec- tors, allowing for differences with ˇtuspeut to intereGt rates 1s 5 depre- ciation rates, tax allowances, etct Utoic.e7te, in. the pr4salt exercise, we have made an aggregative. analyotis, mplJi:-g the t th X s to ell the sectors. 3. In the Kenyan ctise, -e finl that tihe irported investmet goods are abour 30 percent of total investment (1!,e,) = .3). Interest rates, of course, vary for different customers, but as an approximation 9 percent seems reasonable i.e., i - .09). For depreelation i.ate, we assume an average annual rate of depreciation of 15 percenit (i.ers, .15). ihe increase in prices has been slow historically but is gathEaz!ng 'moumetufi now-. It -seSs reasonable to assume that about 4 percent rare of guflation ay be expo-eted over the near future. As regards depreciat&o!± allo;aaces, a 20 perclit extra -write-off is allowed for tax purposes hind at thie p-rcesen-t comipany tax rate, this returns about 8 percent of capita-l vahua .- . t 5L7.6 595.0 646.6 702.9 761.3 eLj.6 1366.7 I Tndrect Taxes 8B.4 96.0 1014.2 113.3 123.8 1h6.7 225.6 - Subsidies 5.7 6.1 6.6 7.2 7.' 9.2 14.1 = GDP at Harket Prices 767.1 827.1 592.1 962.9 1039.9 4489.1 1215.2 1612.? Is+uorts .92 Coods and non-factor services 230.6 288.8 267.9 288.7 312.3 365.5 587. - Espozts of goods and non-factor services 207.8 221.6 237.0 252.3 276.2 321.0 ?2.1 -.neort Surplus 23.4 27.2 30.9 36.1 36.4 44.5 105.7 Total Resources rvailab-e 79C.5 855.43 923.0 999.3 1075 .0 1259.7 iSil. Cross Fixed Casital Porrcation 182.1 197.3 213.7 231.5 251 .0 295.1 887.2 increase in Stocks 11.8 13.6 14., 10.2 17.7 21.' 33.0 Gross Investment 193.9 210.9 22c.5 247.7 268.7 1149.7 316.2 8-75.2 Public Coronation 132.2 188.2 157.1 171.3 186c. 221.8 391.7 Private Cornsumntion 464.5 1,99.2 537.84 573. 6'0. 6 721., 1093.6 Total Consurption 596.7 643.4 69LI.5 751.6 907.3 943.2 1814.3 Balance of P6ayent (in mil_ions of 6k ) ix.ports of i,oods and Non-factor services 239.8 260.9 283.9 31,1.3 351.1 832.6 691.0 (10.1) (3.8) (8.8) (: r '0400'a0_0rw oo 0' 0 0 t ' .00 - 0 .0 K.4t c ., . .. ..... ~~~~~~.0.0 -.t........t....,..o.... ...... tt.t 00_,o 0e .t..0 0.0 oo00e _r~ otou .4 '..o6..0r, 'o * 0. ".70 0 0o 0 O cc c nt oou 0000 ooro o0 0000r 0. 000..00.t o o. K U @_ O O _ b N s,g,,__0 ..4.. 0 4 - - r0 0O '0 '0 CO0 0 .t t O '040*-0 .0.04r 0.004 004. _ 40 0 0 0 0 0 .00 0 .. . 0 0 00to000 00 0'fO ttK c 0- r0 K 00 0. o oo ru.0 eK~~~~~ o00 r 04oriouOOCs _u r rSKOK0'4000, O 0 K o0 * '000ro.0e_r._ ..0.o.w . v ~ ~ ~ ~ ~ ~ ~ ~ . tO 0 0r vo r00 <1 ~{ r_ to 0.-. 0 NO004 Ct K .to o o _0r _ ..4000 r0 vr r . r r o .n r oit c O.no S D i 0 0.00 ; t Ke o 00 .0 K K K _ 0.4qo:I 0_uwo .vCs3 . sssJa .KK,4o K.. - = _~ o r_ r D O C0 toOt_ o v0, r C '. 40 00 TABLE 17 THE DEBT SERVICING BURDEN OF INCREMENTAL BORROWING ON HARD TERMS (The Self-Defeating Character of Commercial Borrowing) (In millions of US $) 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 Debt Position I (showing the effect of the pattern of borrowing assumed in the basic projections) Amortization 16 19 24 29 27 50 36 42 46 57 58 64 70 Interest 19 21 24 28 32 35 39 43 47 51 53 57 60 Net Transfer 31 53 69 75 86 67 82 74 72 63 65 63 62 Debt Position II (showing the effect of additional borrowing on hard terms) Additional Suppliers' Credit on Hard Terms* 50 60 70 80 90 100 110 120 130 1140 150 160 170 Amortization 16 19 3L 51 63 102 106 122 136 157 168 184 200 Interest 19 26 3L 43 52 61 69 76 84 92 97 104 112 Net Transfer 81 108 119 118 119 90 92 81 75 62 61 55 50 *11ard terms defined as interest rate of 9%, grace period of 2 years and repayable in 6 years. Source: Mission calculations. 197? 19 73 1974 1975 1976 19 77 1978 19 79 H9BO 1981 1982 1983 19814 1 985 55.0. 42.49 52.79 96.33 47.39 73.35 90.02 173,94 074,73 741759 136.36 143.92 092.49 710281 183.45 F-oi- 1,L 317.433 741.626 4 73.3581 4 739.235 42(9 7 0a 556.857 764.375 0 673.7 78217 3 79.436 A30.078 3109 32 7 1.309 33.6 is1.1.isol7.3.29 329 349 393 292 472 .29 ., .65 .96 5 3.06 9 - 3.638 69gD 6.75 9.078I CiotooinsSSoins.srno 99.542 776.579 117.582 990.977 143.117 755.717 774.401I I 197.433C737.2972734.116 758.776 754,079 374.151 366.503 Ftott1- Oo~Ls 5.1799 9.75 5'.509 93.921 5 .05 3,.673 6C. 79 206.741 1 6.7736 6 a.45 .5 .0 .0 .0 o~~~olo 4 tIssil 30025 4~~~~~~~~~~~~~~~IL7.19 -a499 - 52.659 35,723 5.97q42 - 61.7 95.781 5 s9.520 4 773a91 746.25 0.433 84.80 -59617 3 9604.77 os 759.9793 2 .793.5 7578.5639 65 .332.775 315.3706 9 1 34.279 0 35.3214 4 52 2 3I. 9.349 136 .367 43729 455.7074 478.911 50.8 1 717.833 252.102 791.114 837.578 977.757 924.448 974.977 1028.166 1085.165 1166.967 1710.805 1280.062 1366,066 1633.198~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4 I 4I I OnissotOg .302 .919 .334 .371 .399 .389 .478 ..30 457 .475 .499 .525 .5537 .587 Il 1':11 2II 211:10I I1 : 5 1~1:46 S3 31 63 793 79i6 1.323 1.:03541 2 1674 1.435 -3 5 36I 1.96 1.467 7.917z 1.731 7,56 1.631 1.996 9.673 6.70 1.747 v`ti-1,- Coni17O:i39.22 5.4. 9.79 12II1 9.41,57 9.933~ 217.475 70.9031 11.4271 I71.963 217.532 13129: 175 1478 5.3 IsCt iI ie14.21? 7.3290 45.93 1.7 737 .7 7.6 2034 7642 7.17 24.05 25365 2.45 2. 59.467 53.27 95.9332 5.13 61.99 9576 s 7 7779 72I75 8007 86.256 41.S8i 3.6 9. 11- I IS'l~~~~~~~7782 00.55 77.3 13C2.31 359 11o3 III .3I17 I5.1 761.539I4 I7.3 8.5 199. 65 236.5 21.7 3.3 CSonitii_ g 351.7S 79.9 7,19 0755 2?83 4.. 4,71 7730 32. 5.1 973? 4073 47.1 7.7 Flos-yN - 5.099 9.192 5,79 5.4z .11 2.3 5.95I7 .S 5,77 592 .33 6.1I8 9,3 6.64 6.8 . T SL 11E~ I C-11- .IoIijii-io-6 187.937 796.99 .27.3 II 14.722 76 .42I 97. 256 45.525 97I.35? a374 7a 91.777 799.717 313.8 37.82 34.7 47170 54.7I3 572.7,0 557.97 ts. 54 11.57 21 ~ .361 75.6241 7 q. 35 83.696 6 8.196 8 9.98 99.A049 9 10,76 5 _1 7179.49 3 03.525I 8621.94I 05.3579 92437.4509 114 .77 4 7 9*4.I,774 2 176..3? 7239,726 5 9 738,0 1677.859 14930. 096 15727 69. to o4ti.-o,tooo 170.34? 1~~~~~~~~~~~~~~71.95? 12,49 19.9 777,907 I I I 11.13 13I79 19.0 73759 147,797: 12 1:43.10 II:14 146.0362169.098 152.8 - - ot c9.1210 15.23 77.82 73976 1771? 49.94 165.75 17733 75399 99.92 719,ts7I537 75.78 69.15252.7 n~ ~ ~ ~ ~ ~~~~~~~~~~C r. r4 ng , . 00 |4r-tnrn hry=niee umoOHi:wOOt O~~~~~~~~~C 0. O4 C- . c1 I _rr -ZoO 1 0 : _ 0 l,4 n 0o OCO O aOOCOr 30 nj>,Xr0:c .rn O C.- I -r_' , : .. -7.oef - CaN 11 D Cr 0 n 00 0o ~ ll, . . O 0g1C2_o o n JrCto o > n n J r ea -c 4 > - sr n Y r e Qn ic 1: < r 1 ;, ,> n iS ii Z n > C r 3 v > g e n a n ;G X n _ n ;D n o z~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0C O10 0'C 0 - 04 00- 04 - o. J ! o o. c o > & ' " )' 2n 00 _ rn OCC nO n l CCO 000 00 O400044n CMNs Nnr vn D On_sNr | r ' nN nw_w n 0 NO N n4 Sn 4O= _O 40CJ.0o X 0 o w s-^ oL o r_ _ N o e no LD _n o _nO. 00 0 0 40 0 M 0 .000 OCO _Ot 0001 W N rh -0 n00.00 n-- .o _nW_Zi, n|N0C n;°NoOt NO tO 00 >0 000 ,W o0 00-. 0CON._ .C_ n wN o o o o o o00v0 W__ 00 _oN r_ C _r 00 0._0 _ 0 .00 0. 0 r .00' ..0 NN~ ° X o oo - s oNi n rnD o N0 - Ir - N _o0, o1 N.00 r00 .4N oN .o o _ ~ ~ ~ ~~~ W, t . 0 _O O O O O O O o O CM _O _o oo r CWW >OW NV_Oo D W <, .n W o oo r - r .r _ W o 00o0 D - 00 010_.r0 N 0. NW s 0W.00° O 4° ° 1 0 r 0* - r000. T 01 00. CNWO N ON. NO _ W N CO '0 OO N o W W0 .J.0 U Vo .N OONDO'O_ WS .00 000--000 t0..0t W- Do D00 _ Ot 'w0. o'0 _t 0N4 No N Ir._ oOOr N N r N 0 OC0 t_-CIO0 D- 00 _0 NO _000.0 OtO 00 000 N0.00 _0 .4~~~ ~ ~ D '' -C - 0' N W - W r No C -. WI N.0 tt 0 O WN °NC NO _ N ~ r n W _ W N e0. n0 F W10 No N - VOtorN tNo_n_N O 0 O 0CM 00 _.F 0'O O r~~~. 0000 o DCC.004 F 0 0o. rO t C 00NO _ 0t N *-0 t oC__n_~°wO ans>sONN °non''0 0 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 4 r0. .00 -0. .00 o.00 .0 00. o.0_0.0Or oN.o W..oiNW WO .~~~~ ~ ~ ~ ~ ~ ~~ ~ ~ ~ ~ I N n _ N - N D09XW . n N ron X- W W 0 O I 0N 4 ro W 4 V-' N t CO 00 NO t 0114 W NO_.NOO t O0 N -0414 _0 W 0. .00 0 C P_=i W DN _rN O _0 .0o N 0 0 001 COt.0000.-Os_N NN 0114 N..w;oCW_ N0 t wwrooNNro Nrr0O 0 0.0 00000oJtM004 O1t NO ooorro r oo.WoWr No r r N4 _0000 NO0,0 0 W OoOO N0 O0M NO t N004'oOO 00 N.0 COoO0o rD .00.00.00 .0.1Not.Or _ 0000000.0 to, 0VN CoON 0 00 00.00 CM0N 0 rw wo>-w O~~~ ~ ~ 00 N N O 0qor N11 00 _O N CO . O0 oN OO W0 40 0 _. W0 WC o NM , V.'_ NN 0 tO W 00 Ot WO 0' n > oN nvw,o 00.0- °4j° rOW M00000 00.0000. °0 W C_ o ototto NWO 0C00 rrN 00 0 0 .0 . o w -N M 0140 00rN .0 0. 0. 'CCC '1 N N.0 0 00 WO O0 000 .00 .0 _00 aO n-W NOwr o00 F 0. N 00 COO .0001 COt _ HN O..00 0. _.0 WO0 r0 W04 tN t. 0$ 0000o l sOW0 _ N 00 0C 0 r0 o4400.0 ..0 00 0 NO .0 .t ' 0 1rt Nt u 4 nr nr ttX O0 0000 .010. -o 0on °O0r N 00aN .01.9a 0. -RUJE N~I iY I1.,.T 97'71 RL:GCEPCC TAB I: (OS7 I to-l tous 1115 197 PrI es Grrss D--s;ic Product at Fa -e Cast 166B.716 1787.926 1919. 519 2a7 0 .931 2235.4664 2614. 3 38 26 0.998 2820.622 3 05 1. 137 3302.233 3575.852 3874. 251 46199. 737 6554.977 Nn-lo-t-ry GDIt' at Factor C-st 356.730 1369.572 382.717 13 99.:192 I416.120 6430.684 '447.9112 6 465.512a 486.4 46 1 5I0 3.86 523.9931 3566.953 566.751 589.6 421 Slne- y ?a Factor Cost 1311.986 1419.352 1536. 639 1672.739 1821.344 193.654 2160.987 235o.734 25 6.676 279 8.3931 3051. 857' 3329.298 3632.996 3965.556 Total Indirect Toocs 169.645 224.911 248.113 249.5764 293.0186 28. 11 2 318.9459 3 0 3312.742 379.7.798 3414.275 1482.777 4 295.3711 542.03542 593.663 66630.7575 osiaos subsidie 13.96 14.74 05919 17.191 18.593 R.12 21791 23.6 12 25734 28.1 :110 3 0.:4 76 3319 36251. s39.622 Crs nsi trdo tMaktPic 186.25 199.51 25179 23331 50.5 271 3.:1 66 234.849 3174.309 339.678 3727.805 4 040.1 383.093 67.19 1612 lscst f Coed To.taco orte 523-. 3013 602.952 648.306 699.330 754 .20 3 813.345 992.519 952.448 13 32.427 1119.I7I 1 21 4.6453 1306.668 1425.726 15 51.696 tuanot c Goods C 2;on-Faoordorvies -539066 -542576 -584 283 -62.099 -67.6710-23.735 -26. 517 -57.99'72935.832_ 101.181769-109.199 -119.006 -128. 769 1302.73 T-ro IEtiod 1----oet 607.191 464.583 514. 267 957.454 604.446 655.594 711.284 771.938 839.819 90.37 9881. 5 58 1324.172 1 167 .5 79 1269.813 lovontery tonostssa'st 2~~~~~ ~~~~~~~~ ~~~7.471 30.63 3 34. 067 3 9.:197 62.79 8 46 .7 45 51072 55 . 51 7 61.02 66.735 73.006 79894 87.6642 9.8 Cros Rao..- I4ernn 3.652 495.217 549.334 596.651 647.244 702.339 762. 356 827.785 999. 061 976.272 1I1.5D15.0II28.4 1368.293 0000cr~~~~~~oot Coosoopttoo ~~~~~~~~~~311.395 339. 498 370. 150 4013.:5601 4 39.9289 479. :675 5 22.9'.6 57 0.:I112 621.524 677.564 738.649 835.236 877.812 96.925 i: ut l'se.ad Crstns 09 2.6473 122 3.:733 1297.3463 139 4.343 1501.084 1620.80I1 1233.574 1873.19 0 15.79 28.9 3180 28.7 8622 39. BA7.931CC IF P_AYHET7S2 USD) MillIIon) Expa: of Cor ds & 1ion-Fottor S-rior 56 1 .5 02 62a9.85s3 679.339 739.272 1809.456 1 83.205 1012.340 1123.058 1 26 1.8,138 14 06.33I1 15613.796 1707.663 1 866.768 2036.327 la fGoods Non-F-ctor Serices. 616. 45 1 745.825 825.967 917.6821 1019.119l 172.291 02621. 5842 1406.69 1570.635 1753. 7 50 1 96 0.22 2198.685 24666205 2736.650 ssoreBOlOn- -56.951 -1 35 .975 - 193.6 07 -1 79.4 08 _ -29.666 _239.2 84 -250.262 -293.635 -309.798 - 349.4 18 -396.229 -691.0221 -581.437 -7 00. 12 3 Coo laserest -~~~~~~~~~~~~~~~~ ~~~~17 . 1 08 -18.579 -28.471I -38.747 -5 0.69 9 -65. 816 -982. 731 - 103. 2 13 -12 7.833 -154.201 -188S. 15 7 -226.637 -275.511 -336.479 Interest oo Pobits OrDbo -1 7. 100 -18.l945 -21 2 12 -214.026 -217.944 -19643 -35. 371 - 38. 965 - 43.093 -466.860 -51.8 28 -53.221 -56. 793 -60.687 De i TiotIosen noe-3 5.6 00 -39a.000 -40.00a -42.000 -45.00:0 -48.0 00 -510.000 -5 4.0 03 -58.000 -62.00 -66.00 -7.00 - 26.000 144 -78.00 No IFacto irntc 1noo- -5 2 .7 00 - 56. 579 -68.489 -90.736 -95.685 _11I3.783 -133.692 -157. 15 -184.978 - 216 .6 30 -254. 070 - 296.6406 -369.62 -6 16. 43 0 Co ra rPaysseto 2 0.386 21114 21.842 2 2.570 23.298 24.0 26 24.754 25.4 82 26.210 26.938 2 7.6166 28396 29. 122 29.850 OaoooooCurret Aoc..o.t -8 7. 268 -1771.460 -1 97.2 26 _ -3 6.974 -2282.093 - 329.04 0 -359. 175 -415.3 39 -4 62.563 -539. 110 -62 2.6 33 -7649.036 _901. 787 -10864.703 Nat Diret F-rilE Isv-nnnt 35.700 41 . 81R 45. 10a0 489. 7 00 5 2 .6300 5 6. 8 00 61. 400 66 . 300 71I. 60 a 77 .30 0 0 3 .5 00 90 .20 0 97. 4 00 108.100o Total Dlb-rs-et of Poblits Debt 65979 66.61 93.4 l ll I 120 11.7 1 131.2 7006 044.:2 39 15.2 20 156.2 28 159.305 164.6 88 17 0. 267 176.379 13.833 19 2.10 4 Total Anortitatlon of PobIto Debt ~~~~~~~~-15.400 -16. 100 -19.350 -73.971 -25.700 - 26.5 80 -49.988 - 35.5 46 -642.06 9 -4 6.083 -56. 639 -88.02 2 -63.90 - 69.932 Ntie Po blic Debt 5 0. 579 50a. 31 4 73. 770 9 3 .8 06 10 3. 00 6 11 7 .655S 102. 812 12 0 . 679 117.236 118.604 113.628 118.387 1 19. 933 122.1L73 Cosisst f Pob iGo Debt 10 3 .7 0R 181t. 8 00 197.8600 140.900 14 7I.011 I 153.90 161.110D 161:2.250 174.7 17 182.561 190.754 199 .366 208S.6409 21 7. 9064 ltrtof PbDCo Debt -17.108II -18.945 -21. 212 - 2 4.426 - 27.944 - 3 1 43 KS. 321 - 38. 965 -43.09 3 -66. 860 -5 1. 028 -53.221 -86.793 -640.6457 P-:lio Icho ottod io,g&D Itobsos-d 462.4 79 512793 596.963 698.369 793.375 '.01.37 I 103.844 1124.5 23 1241.759 1 36 0 .363 1673.991 1592.365 1712.281 1836.653 lap Go Ga Ireo Payssnts -0.8'I62 122.01 106.796 12 4.3640 159.92 2 191.9 30 237.962 275.916 332.827 403.634 493.578 615.93 6 769.636 98 3.2182 Cba_c in Rse_no 6.47 -265 2.49 -072 -3 3.47 5 37347 -42. 996 -472.5857 -96. 101 -60.628 -69.0a70a - 75.689 -86.982 -955.781 Moder Sector W.oeCetmn Pgrtotiltore ~~~~~~~~ ~~~~~~~~ ~~~~21 7.6433 22 6 .13 0 235. 176 2464. 583 254.366 26 4 .5 40 2 75 .1 22 286127 29 7.8572 3 09.6475 321.886 336 .72 8 348. 117 362.0642 Mists,g C Qcarryin8 ~~~~~~ ~~~~~~ ~~~~~~~~3. 129 3.3041 3.689 3.685 381 409 4.:3 39 4.:5482 4.8639 8110 9.396 8698 601 6 .38 4 Mooatriog Eipairg 99547 10634 116.65 17.58 139.70 83.71 167.61 18363 201.48 220.321 26.2 2638 2950 3111 loitdto1 4 lonstrorrico 3~~~~~ ~~~~~~~~~~6.296 3785 3.99.. 40.16 40.369 42.611 43.089 49206 46.563 47.60 49.60 0 50.88l2 82.609 83.82 Electricity Water 5~~~~ ~ ~~~ ~ ~~~.258 5I7 .8 5.59 5.79 9.824 8.941 6.060 6.181 6305 6.632 6.61 6.693 6.2 Tronpor, Sorao &Cssnssotottos 4.66 50776 53.031 55.42 7.98 6.99 63344 66.22 9.28 7.37I75663 79.01 2.66I8.65 T-leab 4 otai Trade 47..-5165 6.7 5220 5493 57785 40.6295 63.689 66.08 70.290 73.2 77.5766 8169 55613 89.96 Serniocs 29.963 272.79 286.704 391.481 31.962 333.27 350.639 48.691 387680 4077491628.470660.5252673.30a698.29 Agrirolrssre 7.567 7.869 9.184 9.511 8.852 9.206 9.576 5.957 10.356 10.770 11.201 11.669 12.116 12.899~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~7 5 8 84 7?57 Miin 2 loRrytO d .307 .31 2.3379 4.q356 5 .376 .397 .613 9 6443 .67 .94.2 8.580 6 .88 .616 CoslsKa; C Canstrartion 4.~~~~~~~~~~~~~299q63 2 82079 8.86 857 8. 451 36 962 5.8637 3 60439 6.2366 8:4073 6. 599 4 6.797 4 57.00 47.12 7.49294 tilsolenatn & Ralsil Trade 14.212~~~~~~~~~~9 33 7 14.857 1.4 1630 721 1803 19.0692 2012 21.024 1 72 .256 1236.22 26.3765 25. 4607 4 2.01 Total 10~~~~~~~~~~~~~~ ~~~~~~~~ ~~~7.821 1 313.61 119.76 126.65 133.952 141.81 1574 13 99.05 I35 16.42 78.50220 89.36 20087 213.17 122.3 Rp0in8 0 ,_rrYing4.172 6201.52 493 5.8 . 5.28 .194 6.49 613 714 7572.2 .7 Itana:ctss,aC &tioptriog151.21 161586 23.98 193.257 212.33 232.57 294.744 279.02 325.62 43.36664 40061 4 338 9 .896a 681 .827 Ts:iisttss' 7. Ccli troctioO 80.06~~~~~~~~~~~~~~4:94 82.21 53.784 5352 87.861 5 867 60.37 62.30 64.22 66.529 67.13 70.18 72.L28 76.458 t_lruiotty 7. W__r 8.5"812I.7 531 .8 5.65 .71 92 5.944 6.063 6.5 .39 663.6 trrs1ssst, itoorago&t.Onrisstiostlonn ~~~~~~~~~188.602 16.634 25.536 214.83974 224.635 234.84 245.509 25.63 28.2 705339.27 30.7 3.571 335.068 Total~~~~~~~~~~~~~~~~~ ~ ~~~ ~~~ 776.472 860 8 862229 91178 96451 1820 .684 72 3 1In4I33 0212.91 129.3116 1362.52 i3144837 1533.04 1628.3I'93 tEttarngta I T-dngiutaa 1009.2.09 1324 2.5 129.036 3.2 3.4 137.27 1o.96 14 9 04771 184332 155.117 19.8 CnooteoousetsoldCusssonptlno 1157.~~~~~~~~~~~~~2123 122.7967 1389.9649 1401.62 1581.438 1r3.643 122. 481 186.693I 2007.823 216.745 2 33.28 2596 2271.86 293.701 TASL 27?NCECEON72111 ALOWR RTE E UEVIINI BUILDING AND CONSTRUCTION 197 2 197 3 197W 197 5 197 6 19K1 19796 197 9 1990 1981 1982 1993 1984 1 985 Vt:~~~~u Cross A~~t ;t,c Prodc.-t itFat-e Cost 1665.216 1257.924 1 914. 956 2 0 61I.I1 2 219).595 2 391.513 29 75.124 297 590. 7I7 3000.965 3240. 32 6 3600.666 3263.961 4090.399 4426.383 . .-..... torp CD? a Fc.or C-o 3 56. 7 30 36 957 2 3982,0e77 3 95.192 414.I1C0 43 1.6 84 4 47 .91 2 465 5.9258 94. 4 61 5 03 .5840 503.993 5 44 .953 565.7?50 5 9. 420I :itO..... EDP 1Cco Cos 1311.956 1 415.35 2 1 53 2.79 166 2.9 19 15 05.:469' I11 1963.29 21'30.12 2,314.949 2919.1503 2736.496 2976. 6 70 3239.007 3525.6468 36838.:961 Total IcdorRcl taxes ~~~~~~~~ ~~~~~~~169.448 224.4 03 247.167 267.99 2 907C 315.524 343722 374.260 4a0.06 445.263 486.302 531.4469 581.22 63.3 OtttODtR Salssislios ~~~~~~~ ~~~~~~~~13.956 14.79 15.919 21 I5.59 3 2011 2.9 3.62 9.I 8.0 3.76 319 3.21 39. 632 Gros -ss-sic Prod-c at M-rkl Price.1..7 1997.543 246.19 21.4 2a 1.42 267.17 200.094 313.2 33.306 35756 396.495 282.210 637.420 525.96 loads £ Dcvs-Faclor 10r52100s 5~~~~~~~2I.3C3 599.054 642.S7 61,.603 744.I95 S0.1 96.235 933,932 10 I.9 10 92 .92 2 18342 21.9 08794 55.1 CUOcot sE AoadsGUca.FodoRs -,1 -~~~~~~~~~5 09.06 _5942.576 -590.637 -6 230. 466 -66 3 .4 22 - 7 06 .2 34 -773.149 -9827.59 -9 _95. 376 -965. 373 -1038 . 064 -1100 .54 9 - 166 .7 96 -1237.930 cT-1 Cli-d lcsoslccsI 407.151 I455215 503.972 54 4.232 559.992:1 637.554 6910.:3511 1 797.26 9 10.1I4 I 283 10' 952.:3 26 1032:1.920 1 120.700I I 12.31 ItsrNa-y lracotaool ~~~~~~ ~~~~~~~~ ~~27.47 30.633 3274 37. 692 4 1.0594 46.744 4782 3.2 04 59 0 458 6 3.355 69.173 75.553 8.5 90234 Gro- ososcOesel434.652 S5.5.8a 535.9 26 I5.919 630.011t 692.297 739.063 a0.3 56932 94065 102149 10.7 2323 10.0 7iocROs5ooat Cerst,o,ptlRe ~~~~~~~~~ ~~~311.355 339.495 373. 140, 403.543 439.9 146 479.628, 522.881 1 570027 621.416 677.430 23 8. 484 805.-036 977.9 73 956.6404 Es Coo Ilcoobold ossoetlos 092.673 1228.696 1302.94 1392.5 29 I502.4589 169.987 1730II1 1866.4 31 2005.456 21 6605 23101 5425 27.79 0992 BALANRE OP P.lUCRS t, Ili. Exprt f Ood & No-facorSeo Rt 561.5 00 609.953 671.216 2 30.259I 7914.7791 871.450 992.641 1083.773t 1011.3:34 1340.720 1,686.926 1621.837 1770.71S 1983.626 Imrt f _GoA & osFacto 6escs616.451 7 41.007l 815.64 90.50 0.7011 114.991 1240.665 139347 1536.96 712.465 0909.910 2029.658 2376.381t 2656. 281 ESSRONRO Nalosoc ~~~~~~~~ ~~~~~~~~~-54.951 -131.155 -17432 -07251 -210.920 243.541 -298.024 -295 .5973 -325 .5462 -371.748 -424. 986 -502 .921 -605.662 -7200.657 O.t Istor.s. -12.00 -05.579 -28.038 - 39. 0 0 1 -49. 772 -64.9855 -52.0 286 10 3.031 -12 774 0 -156.7 43 -192. 077 - 23 2.881,, -26. -349.0619 IdeNext cc PoDlit Debt -17. 100 -18.945 -21.212 -24.426 -27.944 -31.943 -35. 371 -38.- 27 .94 3 1 9650-43.093I 446965494351.0286 -53. 22151:-56.79353-60.457 let Dirocs Iscostiseec Issoor ~~~~~~~~ ~~~- 3S. 6 30 -38.00 4.0 492.900 4 5 . 000 _4900 -51.D000 -54.000 58.5000 -6200 66.000 370.000 -7 .00 I7.0005 Net Foster Iseclco Iscti~~~~~~~~~~~~so ~-5 2.7 00 -56.579 65 .026 -7.9 9475 -12. 566 -33.006 -15.003 a II.77 -218.720 -5807 32.19 38.62 6593 Net Irossfor Po3asoots 20.386 21.114 2~~~~~~~~~~~~~~~~~~~~31942 1225701 23. 2981 24.026 _24.54 _25.4912 _26.210 _2,6.938, _266 28396 292 29S850 5oldsoRos C aCCost coosoet ~~~~~~~~~~~~~-57. 265 -66.619 -090.612 -34. 671 -292.3,92 -3 3 2 .352 -66.276 -22.094 -955.03 9 -63.509 S69534 -72.346 -934. 993 - 1116.760 Net Dic-t Po-ej, Io.ro..e.. 35.700 41 .800 45.100 4 9. 70 0 52. 6 00 5 6.580 0 61 .6400 66 .3 00 7 1.60 0 7 7 .3 00 8 3. 500 9 0 .20 0 97. 400 105.100 Tocal D-lb-rocost of Public Debt 65. 979 66.4014 93. 120 1 1 :7.750 13I1.061 1 4.2 35 152.720 156.2 25 1 599. 3 05 1646a88 1760. 267 176. 379 183. 833 199. 1034 Oc t01 AoatoCo f PublIc Debt -15.400 16.1I0 -19.353 -2 3.975 -29. 700 -26.993 -49. 908 - 35.546 -42. 069 -46.093 -56.639 -58. 022 -63. 01 -6 9.92 Sec CoblUc DONE, 50.579 50 .31 4 73.770 9 3 .906 103.a06 1t7.695 102.812 120.679 1 17 .2 36 119.605 113.628 118. 397 119. 933 13.7 Cosicoc f PoUls De~bt 103. 70 " 081.80 157.9001 140.:910I 142.000 153.501 10 160. 100 Il 167.250 74I717 1 82.5161 090.254 199 .366 208. 409 217 . 904 IcoUs f P.blic Dobc -17.100 -1. 945 -21.22 -249.926 -27.94 4 -31 943 -35. 371 -38.9 65 -4.93 - 46.540 -51. 0 25 -53 .221 -56. 793 -6 0.657 ItbIOc DeEc Ocor.sndia0 6Dtsb-etd 462 . 479 512793 586.563 6 50.369 783.375 9 0 1. 032 1003.549 1124.923 141.759 1360 .3 66 547 3.992 1592.349 17 12. 28 1 835.694 Cop Ic bIases oC Coysstceo ~~~~~~~ ~~~~~~-8.862 115.609 1a00.36 5 121.499 159.169 19 3.990 243. 537 285.88061 349I .107 425.549 52 3.376 646.241 799. 145 991 .195 Glargos to SORotoes ~~~~~~~~~~~~~~~~~66.47 -41. 104 -25.622 -29.334 -32. 393 -3 6. 0 66 -41 .4272 -45.76 -50.985 -5 7.:966 -65. 157 -72.651 81S. 486 -91.707 Stl9&Qoty 217.633 223.925 230.618 23 7. 5 07 244 .60 2 25 1. 909 259 .434 267.154 27 5 .166 29a3. 306 291.552 300a.520a 30 9. 549 318.796 MOIURO....... eoltog3.129 3.251 3.70 3.910 3.646, 3.78 3.93 4.090 4.249 415 4.5537 4.:766 4.952 5.165 C its ~~~~~~~9957 10.39 15965 126.744 135.21 15393 1 6 5.460 15.34 197635 216002 236.073 25.09 71.93 38.5 -. 7 & Wotor 3~~~~~ ~~~~~~~~~~~~~~6.296 37.494 35244 39.09 39759 40.589 41.397 2.25 4.6 4391 4.0 4.0946.11,9 47.552 IOAospcrt, Seorafo & Cctccccl cosjoco 5.295 ~~~~~~ ~~~~~5.372 5.474 5.577 5. 652 5.709 I.599 6.01 6.123 6.29 6.35 6.7 659 672 AUclcattla o RIttIl Grade 49.661 50.650~~~~~~~~~~~~~I 52934 55255 57747 6031 62.97 5.79 69.725 71.792 74.974 7839 81.79 85429 47.515 4 9.71 4 5.164 54734 57.431 60.261 63.30 66 . 345 69.615 73.045 76.644 80.420 84. 383 88.9641 local 2~~~~~~~~~~~~~~~ ~~~~~~~~~~59.963 27.79 256.11 30.256 36.65 32.81" 349.99 13 367.793 356.629 4 06. 43 9 477.273 44 9.:155 472. 231 496.7 717.833 749.922 755. 390 823.24 864. 075 906.591 952.235 1900.51I 1051.214 1 105. 238 1162.569 1203.641 1288.108e 1 381.61 UtcINto. 7.567 7.793 S.02Z5 9.265 8 .51 2 5 . 766 9. 029 9.7?98 9.576 9.962 1 0. 156 10 .460 10.270 1.996 'jIn Qsrr11lyg .30 .314 .2 .339 .32 .6 350 .395 .1 47.4 40.7 47 >scEaolCc &o--i-ls 2047 2.30356 2.04 2.0 117 3.52 7.16 40.74 449 4 5.:554 53098 58032 63. 424 '9Lit 4)fot.i .994 9.099 5.26 5 .365 9.475 .954 5 . 696 5.R10 9.:926 6.045 6.166 6.289 6. 415 6.663 Clol-iULS& Bao-1.2 0.31 I1.36 1.39q2 1.415, i1.45 04 72 1. 53 1.0525 1.55 7 1.587 1.617 1. 642 1.6786 loscepsor., It 'Easc 6 Cs'e,tcrrlo.sslar,s 8.72 .616 5999 9399 9.817 02.254 10.700 00.06 11653 02.203 02.746 1332 13.0 1.2 ooIip.i . iseosonicatiass 9.~~~~ ~~~~272 9619 9.007 9.412 9.839 13. 277 10.739 11.272 10.26 12.23 10.04 13.390 13.991 14.610 AbIacoaesil Frao, 14.212 14.979 Ie.622 06.402 17.221 09. 091 19. 994 19.982 22.929 71.973 213. 07 0 74.223 29.432 26.702 Sc-toO 5s 0.697 93. 204 99.949 591. 7 32 61.743 44.9'09 493.240 71.7143 79. 427 79.303 93.3s79 97.666 9)2.179 9.1 TeE 1 10~~~~~~~~~~~~~~~~~~~~~~~I7.921 113.222 119.904 126.249 133.623 1411 .0 62 149. 198 157.9467 16 7. 109q 176.395 192 .4982 199.709 210.700 273.9512 Isinl -srr-lo 4794.09 4999.369 9117. 364 9 29I3.93 6 94954.379 59 96173 3 9921.09 4 59914.700 61914.:319 9 378.497 69577.94 4 6791.749 6992. 3 69 70.9 7csslacc, rinG & Ropsiriog 4.1~~ ~ ~~~~~~ ~ ~~~~~~7 2 4.:36 9 4.9575 4.:7 91 9.01I9 5.29 9.03 .5 6.0 35 6.:3 20 46 19 6:93 1 7.2599 7.:6 01 Halldnog GGatuclo19.2791 061.478 174.7 46 113.4 64 2021 .7643 231.799 253. 7 17 2777. 7195 303.993 132.7 36 364.709 390.65 7 4 36.364 47.3 ilactoicicys lOrE-c 9~~~~~~~~~~~~~~~~~ro0.064 92.21 7 93.7983 93 9 7 57.09a 9. 7 72 609 3 3 47..5 46.77 541 44 6I 130 70.17 4 72.279 7 4 .449 Toaonpor C. fOosesoc 2. Tososo Ictions 9.099~~~~511 9.021 9.72 9.224 5.276 9379 9.:3 92 9.a34 5.94? 9,0:49 I I.0 9.656 9.7 13 9.7 ioslrcslo 5. teas) Fad's 199~~~~~~~~~~~~~ ~~.4I2 69.91 4 209.3 48 21497 9 724.729 234.28 74.939 259.5 7349 27943 290.921 30904a1.76 33.9 97 a0 94.46 9227 0.13 63.03 9.96 4196 3. 7.20 093 9.7 9979 93.2 34 97.990 379.I7 3I .8 3 95.714 37.9 394.19 41.El44f 79.296 El60.I0 I9*S 909.933 5 9304.1346 962 .71 6 91.662 622.10 Total 776.4 97 709.7u 147 91.9(8I 910.5I 76 942.73:I 2 i119.38 3 11I97 7.93 11-1:1.49 0'07.953 1279.559 1394.29 50 1437.977 19529.2467 11.4 icsa- olop-eo5srkiori AgrOr-lts- 122.04O7 12 1.601 1 2 3.310 09303 127.976 029.7 44 1 37.039 134.41 134.991 1 3 9u37 14 1.899' 1s4.9471 147.7 04 19.7 E.: doLe issbl Cas::pio '17.223 0229.979 13 07.908 1339.0 04 1499.994 16e01.185 L 17 14.964 I9?. 9 3 3 119~.094 2 74 1. 619 23.41 7991 26.97 9061 TABLE 30 PROJECTIONS WITH A ORRRPCAEIVLiCGAG. NPTERNS C) GRWT 1972 1973 1974 1975 1976 19_77 1978 1979 19190 1981 1982 1993 1994 1985 j:I>,'CE, 'L.2;1..r L(No- 11oo ooc.r 971 rcs Gros Heo xProduct or P~-ta Cost 1 66 8 .716 1782.974 1919. 422 2010. 435 2 23 4 .1I88 2411.831 2606. 67: 2913.919 3041.299 3258. 361 3586.969 3849. 065 4166. 908 81.9 o-oroyGOP or Footor Cost 3586.173 0 369 .5712 386.9573 4085. 90 1 42 6. 196 42 .950 6 4591.8981 493 . 325 91`9.044 543.946 571.144 5 99 .10:0 629 .6 86 64.1 Honoraryl CD01 otfato Cost 0 310. 9 56 146198.35W2 1532.H849 16 64 . 534 185017. 992 1 964 .3725 27134 .740 2320.964 2523.255 7744.415 2989.808 3249.364 3537.222 38.2 Trot Ind .IrCt rose 169.44 8 223.841. 2 46. 416 2617. 17 4 2789.9 01 31 4.865 3 42 . 630 313.0DI9 40.6963 s43.634 454 . 44 0 529 .21 0 57 S.6725 63.0 Dsxiocsulobsidies 1~~~~~ ~~~~~~~ ~~~~~~~~~~~~3.96 14.784 05.91 719 8593 20.171 71.11 73.617 25734 28.05 3046 33.19o 36.7851 39.:6 22 Erros IOoooorir Prudrot or IlorAr: I. lops 18~~~~~~~~~~24.278 896.981 249.919 230.419 2508491 16.575 2525460 363.342 3427.2 18 3 3.99R 010.91 3 4345.59 4109.282 106.483 Oxo - o God osFco N'083.303 594.172 631.152 684.918 16.220 19669 994.421 921.391 99577 106.214 16 4.26 2 128.42 1361.447 14.8 Export of Goods A Gao-Parrot 7errisas -809.0~~~~~66 542.574 _554.t105 -6,21.58 7 -578329 -73.715 -96. 959 -851.045 _-934.167 -1R09 .568 -10 91.618 -18737 -1926.991-10.4 Total flood Ocr-tsoo.t 407.1.051 45 3 .442 8 00.827 54 0. 7 83 58 4. 446 83 1.920 613. 235 7 39 . 064 9199.697 565. 57 4 93 7.0I24 0815 .02 0 I0899. 6 90 19.1 Tovoocory Iovestoooc ~~~~~~~ ~~~~~~~ ~~~~~~~~22.:4171 30.613 32.975 3 7.92 5 4 1.316 s5.024 49080 53.511 58315 463.:69s 4 69.852 0 I18.908 I 8 2.903 90.512 Crss," eciIlloyoo 44657 45405 5353 81.0 2.6 6 76.54 3231 12592 859072 929.268 1006.6 94 0090925 182.593 18.9 CGvor.urot Coostuotlse 311.385 339.498 323.139 403.531 439.940 4719.620 522.870 570.0 13 62 1.:398 617407 73 8.4857 505.001 817.9533 95.9 GoPstHsuhl _CossoispIie 102473 1275.595 1299.337 1395.713 1500.692 1048.9582 1129.963 1865.09 20318 213.690 384 225029 01357 04.8 BALANCE OF PAYMENTCS (US$ Milli o) Eopors of Goods S Non-Factor Ne-u-e 561.50 609.893 6175.225 139.0223 839.:044 892.404 01 1. 5 04 1 121.9146, 1260.402 1402.5151 1 861.5 32 0109.1L93 18S62.011 23.1 Ispoc of Goods F t-r-aco Hr-oe 61 6. 451 7135.61 4 911.136 898.841 995.010 11 07.1 15 1225.069 136 0.829 1514.799 1686.285 1819.972 2 09 1.:868 2331.013 59.1 Rosur Hl e-5 4. 951 -025.821 -136. 511 -1599.924 -186 .0 26 -709.711 -213.665 -239.863 -254.397 -283.769 -311.440 -386 .661 -468 .943 -6.9 NelOscrot -12.109 - 18.57 9 -27. 558 - 36. 994 -46. 84 1 -5 9.1017 - 73.7017 -90.14703 -109.607 -131. 431 -051.7:6 60 -1 87.061 -228.280 2104 Ost-oo 0Pobic DebL -17. 100 -08.94 5 - 21. 212 -24.:4 26 -21 .944 - 3 1 943 -35. 3 71 -38.965 -4 3.,0 93 -46 .840a -5 1.0792 -53.22 1 -56. 193 -6.8 ot DiIcctll- t OvtcocIos-35.600 -3500 -40.000 -42.000 -45.008 -48.000 91. 008 -I'. 000 -58.800 62.000 - 661.000 - 70. 000 -14.000I -78.000 Her Fa-tst S-oo I--ct -52.100 -6579 -67.548 -78.584 -91.8 30 -102.690 1246.615 -144.449 -67.567 -13.377 -223.55 -257 018 -2 99.224 -5.6 Not Tco..fo- Pey .ects 2 0.386 2 1.1 14 21.842 2 2.5170 2 3 .2 98 2 4.026 24.7 54 25.402 2 6.2 10 26.9 138 21. :6 66 28. 394 29. 122 2.0 HBancI o -uren _xoo 8 7.:265 -1 61.296 -192. 217 -21 5.839 -294 .5 50 -293.36E5 -3 13. 581 _3957. 8 31 -395.749 -450.209 -53.369 -615.291 -739. 048 8862 Noc Direct F-rigs Tosostmoot 35.700 41. 8 03 45. 100 49.700 5 2 .6 a01 96.9800 60 .400 66 . 300 7 1 .600 17 . 3 00 8 3 .5 00 90.200 97. 400 1510 Toru,l DIIb-r.e.eot of Pjbliv Dcbt 68.99 66.414 93.020 11. 1 81 131 .706 14 4.2 38 1 52.7 20 1956.2 25 15 9.3 05 164 .6 88 11 0. 267 116 .31 9 183. 833 1210 Trateuciatso Foblir Debt -18.:4 00 -16.130 09 3 50 -2 3.979 - 28.1700a - 26.5890 -49.988 - 395 546 - 42.06 9 -46.08 13 -536.63 9 -.022 -63. 901 -6.3 Ne APbictDeb -50.579 50.314 273.71 9 3.8 06 1 03.006 117.6 59 1 02.812 02 0.6279 1 17.2 36 11 8.604 1 13.6 28 018.' 31 1993 12.3 Csoit-otsoffoblIoDoSs 003. 7100 181.8a00a 1537 .800 0 14 0.900 1 47.0 00 1 52.5 00 160. 10a0 161.2 50 174.711 102.561 19 0.175 4 199 .366 2 0 8.4 09 21.0 lotoro-t of Pubplic Deb -1 7. 100 -18.945 "2.212 -2 4.426 - 21.94 4 -3 1.9 43 - 38. 371 - 38.9)65 - 43.393I -46.9 403 -5 1.0281 -93 221 -56. 793 -68.457 PebliI Dolt Cornra-diogbDisb--ud 462.47 5 12.19 3 5896. 563 650.36 9 1 83.375 9 01.03 2 10 03.844 1 124.5 23 1 24 1.759 1360.363 1473.991 1592.34 8 1112.281 13.5 Cop is Balos- of Poy-nets -8.67 008.a515 88.447 102.019 130. 7095 15 4.23 3 189. 98 3 2 15.6 12 251.729 310.994 37 9.829' 476.991 6 00.633 14.1 sogos is toserves 6.847 39.344 ~~~~~~~~~~~~~~I'-29 0 -874 -1.784 - 35.325 - 40.6 08 - 44.76 1 -50. 1 5.9 -63.50 -70.280 -18 . 921 -885. 726 EMPLCYOiENT (1,092) MoMr Hector Woos Eniplovycot Agri-totr 2117.4 33 226. 069 735.:0 49 24 4.:385 254.0 92 264.185 2 7 4.678 2985.5 88 296.93 2 308.12b 320.998 333.138i 3646.994 36.7 Sisig A Qeacying 31279 3.251 331 3.501 3.6,48 3.7911 3.99 4.093 Il I.293 4.420 45 93 4.172 4. 959 8.5 Slavforuio epiie -54 95991 19.7021 126.308 137.90 10.531 14.3 32 179.399 199845 213.900 2343.431 254.00 218.160 30.6 OxlldioCbfxustC- otlo- 36.79q6 31.492 39.740 39 . 017 39. 78a0 4 0. 5 73 4 1. 392 42 .2027 4 3 .049 4 3. 9017 4 4.19S3 45 .676 46 .5 86 4.1 Elcoc,lity JIrr9290 8.12 5.8 5.390 5.430a 5.470G 5.510a 5550 5.592 5.633 5675 577 5. 759 5.0 Troosport Orc GCsosiojo 966: 90212 51.962 93.173 55641 758N7 59594 61.612 6381 6.4 8348 8.3 13.196 1.4 Ibloe&Nccil Trad 41.515-i-9705 52.148 54.704 57.3989 60.206 63.161 6.261 69.14 12.926 16508 8026 8420 8.3 Scroices ~~~~~~~~ ~~~~~~~~ ~~~~~~~~ ~ ~~~25 9. 963 212.979 29 6.896 3 001. 225 316.6,07 332.782 349.792 362 . 690 3 86 .49 2 406.275 42 7. 0950 44 8.961 4 71 .91 3 496. 176 Toto1l 117.83 3 751.0058 788.-22 828.299 8 7 0.6483 9 15.125 962. 390 00 12.-4 51 1 065 .49 8 1 121.:7 33 1191.373 1 244. 654 1 3 11 .827 18.6 High 1 Middlo Level Eoplnrcxo_t Hieng l- arrto 7.567 7.967 9. 180 5.505 8.942 9.04 9.,599 q9.939 10.33I 3 19.744 11. 170 11.614 12. 018 1.5 Ining& _ying ~~~~~~~~307 .314 .2 .3 39 .352 .366 .381 395 .1 .4271 444 .41719 .9 MoostovtxCieg2R,opairg 23.92 2181 ?3.811 2.9 28.38 3099 33.I2 36.90 40I.30 44.000 40.04 52 . 438 57.24 6.9 Noildoc C -oorooio 4.95 5.059 5. 762 0.367 5. 474 5. 593 564 9.905 5.924 6.042 6.162 6.29 6.410 6.3 CILrctriity & Wacao 1.320 1.326 1 . 336 1. 345 1. 355 1. 365 1 .315 1.356 1.396 1.406 .0 .2 .3 .4 Tronojoorr, Strage &CG-ossi-xLios 5. 272 S.536 8.34.1 41 . 9.460 9.1790 10.131 IO.49s 10.95 128 1.0 2.024 12.443 1.7 Wholesale & lv. sil Trae 14.217 14.61 15.596 1.367 11.159 18.009 19.89 19.419520.92121.912 22893 24.006 5.184226Ss2 Terv ro 0.l 5.0 55 .8 58 559.109' 6 1.10 64.959 68.115 7oI.66o 7 5.32 79. 103 53.239 81.0 198 9.0 Torn ~~~~~~~~ ~~~~~~~~ ~~~~~~~~ ~ ~~~~1 07.91 11 3.04 119. 2072 17?5 .762 1 32.7 33 140I.1i44 14.26 15 . 411 165.33 7 11 4. 84 105 4.95 1 6 I95.787 2 01 .262 21 9.8535 Agri-ul G-c 4194.075 498304 5123.569 5293.298 5467.944 5641.596 5037.291 502 2.143 6 21 7.182 6417.464 6623.010 6833.90 9 7090. 1 18 71.5 HiIeIG& Qceig417 433 5 4. 505 4.651i 486 4 5.OSs 52 52 5.4 58 561 1 5893 6.14 6.363 6.612 6.1 ilavxfartsrieg I Wopairiog ~~~~~~~~ ~~~~~~~~ ~~~150 .7251 160 .0134 115 .7 97 091.914 2 09.5 09 718.7t7 249.6985 217.511 2917. 566 324. 8417 3-5 4. 639 3917.142 4 22 .635 461.382 tiIliog 0.l:vuro-tlao 50.06 5 1.71 3 52.2 44 53.1 96 5 4.069 55.063 7.0179 59.71 993159 60.562 61.769 63.01 64. 257 6.3 Elo-rritlly CIae 5.05 500 5.145 5.18 3 8.21 5.260 9.298 9.3 35 5.371 5.417 5457 5.497 9.537 5.7 Trvpc-i StoraGe& Cfo -x-lcaio- 189.602I 194.613 201.34 78.414 219 61 7223.196 737.926 239..02 42.5 Zs99 6.0 2418 23.9 9.8 uhol-uloI~ Hctoil Grade 52.100 84.50. 51.716 59.983 62.927 66.0195 59.756 71.44 7521 1,6 391 5.3 235 9.5 Noesirvo ~~~~~~~~ ~~~~~~~~ ~~~~~~~~ ~~375. 2723 341.509 359. 543 36.944 396.098 41I6 .324 4,37.4604 455 99 49.1 0.6 5421 516129.8 2.3 Totail 116.491 812.838 055.30o g00.014 949.5 1 908.og523) IOHN.lbl 1 11 32743 1175009 1,240.928 1311.044 1395.813 1465.516 15.4 TrIol-Trnio per bo-kr rio AgrIcolt- 12 0.:041 121.529 024.0725 1 21.5 25 13084.Ss 13 4.279 131.545 141.5ub 145.359 149.300 151. :5729 057.843 162.332 16.0 txavtr livuxebold Coxoosprlos ~~~~~~~~~ ~~~~~~~115 7. 223 0229.969 1311 .3 64 IsO5.l15 150.0 616.4 I 73661 187229 I01.a I 2176.177 2346.191 2531.913 2731.858 2948.961 TABLE 31 TOTIAL ITNVESTi4ENT REQUIRED FOR INCREASE IN GDP IN DIFFERENT SECTORS (in Millions of 1970 USS,) INVEST1,ENT SAVED BY INIVESTMEINT INVESTMENT REDUCTIONI TN GROWTH RATE IN IN REQUIRED FOR SECTORS AGRICULTIIRE- BUILDING ELEPTRICITY TRAN SPOR.T, ORIENTED AND AND STO1PAGE AND GROIWVTH COINSTRUCTION WATER COINTCATTONS 1. Non-monetary 0 0 0 0 2. Agriculture,forestry and fishing 2.4 -.03 -.003 -.032 3. Mining and quarrying 0.197 -.223 -.019 -.176 4. Manufacturing and repairs 0.158 -.454 -.029 -U6 5. Building and construction 0.005 -4.195 _ -.029 6. Electr:icity and water 0.012 -.194 -1.716 -.267 7. Transport, communication and storage 0.431 -.526 -.047 *-10.01 8. Trade 0.032 -.0A3 -.003 -.0'19 9. Banking 0.010 -.020 -.002 -.026 10. Dwellings 0 - - - 11. Other Services 0.237 -.173 -.023 -.162 12. Central Government 0.022 -.003 - -.025 Total investment required 3.506 -5.862 -1.84L -11.215 Change in GDP 3.607 -1.997 -.b6 -3.028 Total ICOR 0.97 2.94 4.01 3.70 Note: The figures in the table represent the sectoral investments requi-red to achieve a 1% increase (or decrease) in sector growth rates. In every case, the investment required comprises the investment within that particular sector together with the necessary suppor-ting investments in other sectors. The last two rows show the effect of a 1/ ch~:.xne in seCtor growth rates on GDF and the overall ICORL. "I 1A "Ea i9u sitlzdz 045 lots WTI 'riL I 461 I zzq I 902 IT 16T 6 92 1 9 0 91 III ILI 9 "I cri!" Lzallfill pl.j-H j.y g 11:11I 2s tal SI 1 3 L 0 21 -11-TAY .1 ..4..H ..d U, 'a S " III E t, I11 I S a 92 IT I ig 9 :6E 626 ZZC :!C; 150 9 9.1 10:9 0 J:11, 6:12 tzz ici 26i OEZ 1)56 tst I I'll., tic Ill 2 5. 60 S. 120 291 .9b 69 Ilf, 615. E99 OZE OTO Ls, zi 99c Zoe 14 6 , 9?:? 57 a :2, 99C - oz, "I 16GE Z20161C '16 :6SE 191 ?"s f?Z:SZE Lu FT 166:6, '94,0(lz So qt 19 act Zs -Z E09 2, %SJ.99Z 6?6 :qt9 TO* S90 es 0 16 IS t.1 iz 9 9 9: I e sce Zvi 5 112 Ss; Z09 OCS .9 _11CIz c ;1q, N9 OSL z ITI Ss..? - i 6"zS tc LST :S 3 I S99:, 9 9 't 543 :,tz G6?: !a ;bI 161 Ill IIS I'S n., I I:1 _ -3 -.0 ? 6 19 6ge Z;.-. 06? '661; S", i, icz 10 9 . I L: . I Zb 66- I Zo I r 'I Ila 21 6ii sib: Z99:15J6, Tt 6,5 esL 'S,I SSI. 9911LE-1 jog ST 11 iti-i "0121169 oil .9-Sts 21IT 9-1-1 I Sib "25 aaIIZ96 Olb'114*6 1lS'bLZS 162,611f, ?6i . IS61 SI 0 'IbIll I :6L.7 6 ! :%Z.., STI 07: Z40 '161, O6- 4g., C oql:oZI S' :t lzn 121 al.i -) - - ' Z' `1 9Sf:SfT *l::L?I TRZ.66 6i 1 . 11' f6f:6 6 9":09 269 4,61 .9s 90t.r, 199 a sZ - fi- i ,. 20 6? 161 911 9 19i 9 1, 2 2 9 I Z fwca SOOIC I 9; I L. 91 62 5% 9 6 _IZ VI P.,j 11 ... S 9 I..j U, S.; 9 aZ .16!:e 5 96 zjzl La. 21., 116 666 .9 31 I I L64:0 ?9t", 61 - ,S tS %?4 1 i III_; 6fic , lq," szc:T titI Ts. 6E,T I zc'. 021. EE9:9 96t 4 29E. it? f 0 1T 9 5. 6 SS9S IC 919 i 006 6 itt's I !. 13 L :91 6s:4 VZ9 1 toI 951:S5 I lbI -,I 90 1,99 :97 Zt I 136'67 129,LZ I 9 4 Ea I1 9 . .3 I 0 S 101 J-H Ic 15 69, S t L 1. 9is skrI 0it, 9 2zc L CZ 193-2i 2 I"T : 6 ZOL,I 9IiEI P?a Li 5 9's-1-w 4.99 -VI 2 ucl,b5zl Zl :1191 [11:02!1 6T I O,; -06 b6b:21 $61 :6169 529:21 291 Z,61 Q66'"_' El - 'Ill a 6-6 I .,):964 ss,:Si Sli.26, 110;.I: 2 TS 9b96 . 2 ol 26) -93 ZQS'it Tisix.Tffti2 Y_wy _-TPPI-w-V _4TTH I S6 OSS 19 I -SI :S?. S., 2 Ebi 20 Z- 0, IIZ"; LIg,!9 7,.,S 9,!.Si S[ :Z !io:a 0;6 5 CL Z 1 6:3 4 ""Z: 21 62Li 6 I ZS . UP, 06, ii 9 r, a IS ;162.L ag S. r, 9515 S!L.Sl S69.C, 50 35 955_, 20 l 64:52 llh:it f'r fi L Q2 IS I!Zf" 6 9ss 9f? -P 11 S Z T.E 9 Eva IT :91 ... I .t 9 NU 991. t Z 0le A S t 1.2 0 got .6 62 'I 6 aS I S I ",.I 991 Z69, 2 -196 t I 921: Z I Is, US; Z, - C as. tI as 9 I 3 it Z 97 51 21 Lz I VOE - CZ 9 Z SLG.0'I togleta I11 tI I 3 3 T11L a (ooo 1muold]43 6 S- Ug4lgl- 929'-- Seg'at- ull"It- fj115E_ Ifiolli!_ 69i QIc1VZ_ I-Ioli li:19 so? SI00 ; li9:Cq; 6 2. 96L 19 IIS9 zi SS91C6 Qtz,.. 0 '99 11191h, 661 1 a -1 000',Z EST lzllt 29 0 - 8 - Iq Z2 Ngi 10111, b'l tzs :112ii 4-11, IE911 2to:llfi sls:Eei 6q9:019 29S:995 E61 :Z J bil:2- P.-q.jo 9 tLIP.."A'o q.Q D 9 - , 9 Z  9i 7j,"Cl- izalts- 1w9- rbo:c.,- zR:,5V- ilb.le- %Sb,LZ- Zt Z- ziz 12- 1-16 SI- ons Z;- I1-qiII1qj 1. I .....r 06,ZTZ HIfioz 99i,ht'g 195 3 lit NI-I cszL9, gol 33Z oos EST 090 J.-IT 006-0-1; coells. 0 09 1 19 001-toll I ZES:691 ir:ol; Q0,E, 'u, a 09 9E2,11, 6 1 9 1 027 11912o; 2691113 900,10I 9091S6 011-Cl SIE'O bifilo. :qo zi6,69- IC62_ 2 25 pi- 690 1Z.- 911s:SE; 9 0 6 :62- 025 :92- 001:t!- GSS :61- 001:9-.- QDI- _, Iqa Ijlqd I. I t 'ebi Z 2 L I 6 4s LV 'UJI 1165; SZ? 95 3-t Z 7 2 ttl 9 LZ, 1*1 4.1 ol E6 91, btblql 50. , , 0, aT Iq__ 004't6 az-ab ouil._J4 oll-m 0 a 9 C09,26 Calla-, .101-St U09,1 DIL'4z I.K 009:59s- 2910:isq- 9149t- 99z:96g- 6Z91M_ T9Z :1.92- lb S:i;z_ 9-1 ti C61- If 51-06T- zt9.fi9T- ScilhEl_ I19 Ill - 102 I 80-1-- 592,  2- I.-D 0S& 6z 62 999 Z UTZ 92 22-12 S1112 92cl-li 96a,12 *Is-" 24911z Silliz 99LIO? S109:Tol. 6 5S3 Ea% El 5":STt- 9z, 0 ti- o:0L_ DOD: :Z01- 22:f log-'s- 109:S4- Z611,299: IIfi:Z9: bli gs: co I -S A.I- I.R 99 a I9- , 0 0 0 0 0 0 el: 04U.S.1- GG.,-, Qot.ot 000.9c 09 .52 122:cs: aza:ls- 0 -,IQ 194- Ebol 11 S961ST_ Tit'Sf- C16:lt -I-,& 19- I.,Z- 212 TZ- SI* 91- a I Al- iq.a I;1q.d zo Cil ICT LZS 6q lz Ij- lgills- 2i"61- ezt les- do 91 99- SROlarl- is 1 192 - pis bislel- co I , L 5Df:N9f- 999:Lrzi uss:491- 9,9'otl- i,!:6 T- Ewen- 9.,Z:iTz- flt:26- -15 I " obb III tgzl 56 rogi ag:-- :L 129 I12 9L LIL:ZI 166".j. 91 lobaz It , P-0 J. 610 gag.!I& SiT toz .2L, bit.761 fig Iso Ill 196,1III is 6bZI 22l'9SIT 99SI69EO41, ea, Lib Z96 to Z99 0 Gfi9 S6 26`9 oz 9 :p.,,z) 1. "d.a 'd Ao 2D4Yjw bSL:dTqLZ I'T 625? DIO.9TEZ 199:,IIZ 6b:52261 Z29:91Sl ZJX:fILT Ift-10,91 b-11:9911 EC-1:1RE3 SOS 21 Sqil.Izl EP,:2bgI 0 6E. C a 1:99 6 irl, la 9 iwqft is, iig 961 1 9 IIIL otu 22S OZ9 b I Nb L 5 L 4 bc 0 t GE,-, 69i fi6l 11 ojg:S6oj 1?9:5 11.,06o III 20 S 0101 61,'Ztb is,31099 0291t6l 199:1ti IJ66:119 966:9?9 661:61S 6jfi:5IS k1S:Eb1 C CZ It-, ?S9:%f% t691E? siclos If, ts 090 6I Izo 61 91i 11 126 Zt SL6 ZE C 9of TL% 12 296'96L I90019 7316'G29 scs'les Egglirs %S91WO 291,09-t 6 9 91 lei It" Ig, ill'IC6 16 z9a' P- j J.'., rI9z:TS9_ t29 69, 'Z 'l- ZTL-096- 19, 1- g11_:g1:- 6L1:t-,L- C61:1bg- J11:91q- 091:109- 92,1:bSS- 990 i- La S 0 E:bO 6;. ?14 ibli 699-to;l 19 tol 6t619I6 6S IVT 2 SEE ast -In 6bg E6 069 Tell 99 19i Lec 6E 1 s i 'IX ci-,9605 59S:or SC 9C. C Z, Zs, I 1,"`, 161:nq - Iifitt tfi:blb? 051:10il 061:0!fiZ 9TE TZ6.. _o -0 9T6:qCtt 16:EOOI SZ:IG9L Z L ' :g OZ qS; S 99i 6Z b12 '.11 I P q,, S 4 -'Z iS B.27 L Et oib at 92 Iiz 0 t T 22 1.1 11 ZET 91"" 11T ic Fcs I L Tn ZI i4- 9101tz 2 91 q; 9 litt I- 3 I. X-j-, 04, :E?5 I 52 t!i 4. I Zb6-1091 It 6 : r an 9 4 bE 9 cis SI[ la') lot bi. -;as .10 9612z43 las.0.3.. t 3;L:9s ..1-1 3. aao U. I-w-om 69416"s 19E.QBZE 66Z.,1St 616tTIZ 9 GI 2 S Is 76143 VW 2t37 3. J-Ij 'JI.-I ... I, T'd OZ61 WI 312YI SZOInOM .40 ISH 5 6-1 79-6-1 6-1 L-6 I TL a-, K-61 E!61 i Uff TABLT 33 PATOOF DE7 PRECTA1IO TO GROS VALUE ALOED IN DIFFERENT SECTORS SECTORS Non-moneta.ry 0 Agriculture and forestry .0858 Mining anid Quarrying .1116 Manuufacturing and RepairiLng ,0994 Building and construction .0520 Electricity and water .0673 Transport, storage and communication .1543 Wholesale and retail trade .0631 Banking, insurance and real estate .0822 Ownership of dwellings 0 Other servrices .0601 Government .0o64 Total Monetary .0711 TOYTAL .0530 Source: Calculated from In-ut-output Table for Kenya. 1967. '.ABL' 3L FTYTTUATEZ, EQUATIONS FCO NET ICOR.S Th DIFFERENT SECTORS _ECTOR CONSTANT COEFFICILNT R2 D.h. SEE ECTO CTER OF TIME 1. Non-n.onetary 0.23 .2057 .32 2.9 .38 (2.06) 01.70) 2. Agricutu-re, forestry and fishing 1.231 -.079 .1. 2., .35 (3-49) (-.71) 3. Mining rnd quarrying .363 .45 .5L 3.55 .60 C.589) (2.,4D) L. Manufactur-ng and repairing 1i.6 -.071 .67 2.1 .07 (19.9) (-3.0) B. uilding and construction -o.506 1.272 .80 1.49 .97 (.l5) (4.i6) 6. Electricity and Water 8.L2 -.506 -.03 1.38 1.69 (4.74) (--9) 7. Transport, Storage and .603 .962 .94 2.42 .38 Comimunication (1.53) (8.11) 8. .Wholesa]e and retail trade .118 .0h12 .093 2.65 .1097 (1.03) (1.19) 9. Banking Insurahce and Real -.83 .13 .32 2.94 .245 Estate (-3.2) (1.71) 10. Ownership of Dwellings 41.39 -5.78 .35 2.12 10.30 11. Other Services .354 .5.69 .89 2.14 .31 (1.08) (5.76) 12. Government 1.O04 .273 .96 2.72 .085 (Ii .58) (10.11 ) 1-3. TotL.' Yloneta-y GDP .155 .189 .785 3.39 .152 (7.26) (3.95) 1,. Tot,a CDP 1.05 .207 .71 3.45 0.199 (5 .02-) (3.29) Source: Mission calculations (Figures in ,arentheses are t values). TABLE 35 SECTORAL WEIGITS USED FOR ESTIMATIING RELATIVE IMPORT REQUIREBNT S OF CDP ^MJD Th4VESTYENT IN DIFFERENT SECTORS IIEIGHTS FOR IMPORTS CF SECTORS CONSUMTER EU CAPITTL _GOODS MATEPIALS GOODS 1. Non-monetary 0 .0045 0 2. Agriculture, forestry and fishing .137 .0875 .$97 3. Mi,ning and quarrying .h06 .i516 1.0 Ma. ranLfacturing and repairing .33 .769 .819 5. 'udldin- snd cornstraction .140 .4517 .828 6. Electricity and water .329 .131 .437 7. Transport, Storage and Conumunic2tion .329 .1617 .838 W. Nholesale and retail trade .258 .0787 .623 9. BankinEg Insurance and Real estate .828 .0159 .639 10. OJwrership of Dwellings .518 .0136 .036 11. Other Services .745 .2018 .,1h 12. Government .1h0 -- Sources: Input-Output Table for 1967, Economic Surveys and Report of East African Income Tax Department. TABLE 30 DATA USE FOR lMPORT EQUATIONS IPORT OF RAW MATHMIALS G D P G D P IMPORTS (CONSTANT PRICES) YEAR (EXCLUDING WEIGHrED GovIEwrN) (YWI) ACTUAL ESTIMATE 1964 869.2 148.7 14i.3 155.4 1965 870.6 155.2 173.0 161.6 1966 994.8 171.3 189.3 177.0 1967 1034.9 183.3 189.0 188.3 1968 1108.7 200.8 199.8 205.1 1969 1174.6 215.1 209q. 218.6 1970 1251.7 231.1 231.9 234.o 1971 1319.7 255.9 263.9 257.6 % Increase 1964-71 52 72 IMPORT OF CONSUMER GOODS G D P IMPCRTS (CONSTANT PRICES) YEAR G D P WEIGHTED (1W2) ACTUAL ESTIHATED 1964 869.2 237.9 69.6 68.3 1965 870.6 256.1 71.8 71.1 1966 994.8 281.9 84.1 75.1 1967 1034.9 294.3 70.h 77.0 1968 1108.7 320.2 80.0 81.0 1969 1174.6 343.0 75.1 84.6 1970 1251.7 375.1 87.8 89.5 1971 1319.7 402.7 1G1.7 93.8 % Increase 196h-71 52 69 IMPORT OF CAPITAL GOODS INVESTMENT WEIGHTED IMPORTS(CONSTANT PRICES) (Excl. Govt.) INVESTMENT ACTUAL ESTIMATED 1964 130.6 70.2 36.9 39.7 1965 125.0 65.3 35.9 36.7 1966 157.9 86.0 5h.1 49.4 1967 196.9 107.6 72.4 62.8 1968 209.1 112.5 60.0 65.8 1969 210.7 111.6 62.0 65.3 1970 254.2 141.8 77.3 83.9 1971 284.3 154.8 96.9 91.9 A Increase 1964-71 118 121 Source: Mission estimates TABU';' 3-7 OFDIIN~"ARY LZEAS SQUAfffS: PRITVATE i,D1ERN SECTOR? PRODUCTDITTY PEER Hi.AI DN) AVERAGE W1AGES2 1 96)4 - 1 970 log V a + b log W E ~ ~ ~ ~ ~ ~ ~ ~ ~ INDUSTRY CONSTAIT LOG W-AvUGES 71 D.W. (1) Manufacturing 7.766 . 988 . 93 1.51 (t ratio) (57.17) (8.9C) (2) Commerce 8.o60 1.627 0.92 1.36 (t ratio) (55.27) (8-65) (3) Cther Services 6.6&e, O X 5 C.86 1.27 (t ratio) (27.78) (6.25) (h) Total Private Modern Sector 8.286 1.151 .9 4 2.82 (t ratio) (38.58) (9.66) KEY: V = Gross Value Added (1,i COG) E = mnploypient W = Average W>Iages per year Source: MIission calculations. CHART I A Simplified Flow Diagram of the Model Exogenous Exogenous Variable Policy Instruments Variables Sectoral Growth Factor Price Exchange Rate Exogenous Imports Government Money Rates (gi) Policy (x) Policy (f) (incl. controls) Consumption Supply (mx ~~~~C) (M) ICOR;s (i5 1ISectoral GDP's Expenditure in | _, (Yi) | 1 |Income(YD) Current Prices| _Inve_stment _Household Consumption _ _ , r | ~Desired (HC)_ Imports~ _ _ Husehold ||Resure (m ) G onsumptionF Avial Import Prices Available (C) Foreign Exchange Foreign Exchange Export PriceE Available Required Labor Product- I IIIWage Employment ivity Change Kx|portF T (PRi) ,, l l § Z ~~~~~~~~~Income in I Residual 1 Foreign Capital Agriculture |Eployment in Labor Forc Inflows and Debt (YA) Agriculture(LA) L Servicing Needs GAP Consumer Prices Poverty Ob|ec(tC)i (POV r ces Objectives Chart 2 Million Bagos of 60-kg. WORLD SUPPLY AND DEMAND FOR COFFEE 90 80 - .. , **. YEAR-END STOCKS 70 SURPLUS * '_ 30 F F ,// 9 \ ia200Vk 0 tA 60 10 ~ ~~~ __ G 0 A P 40 V% 160 r | --I ..r l - * . -. - DEFICNICITE RED World Bank - 7505 - 30 /' 2 "~120 100 20 . ~~~~~~~~STOCKS AS % OF EXPORTS 8 "RIGHT SCALE" ~~~~~~~~~~~~~~~~~60 10 40 4*4 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~20 0 '47/48 '49/50 '51/52 '53/54 '55/56 '57/58 '59/60 '61/62 '63/64 '65/66 '67/68 '69/70 '71/72 '73/74 '75/76 '77/78 '79/80 '81/82 World Bank -7525 -m INDICATIVE TRENDS M ANNUAL AVERAGE TEA PRICES FOR SELECTED COUNTRIES AT LONDON AUCTIONS, 1955-1972 Chart 3 (NEW PENCE PER KILOGRAM) 60 - 0 SRI LANKA / s t _ _ i; (NORTH INDIA AVERAGE (ALL TEAS) 50;\~~~~~~~/ /K 40 _ UADA\ 30 _ _ ~~~~~~ ~ ~~~~~~~~MALA WIV 30 N 20 0 t I I I I I '55 '56 '57 '58 '59 '60 '61 '62'63'64 65'66'67'68'69'70'71'72 World Bank - 4605 (9R)