Toward greater food security for Tanzania : an overview of issues and prospects

Discussion Draft AGP-16 TOWARD GREATER FOOD SECURITY FOR TANZANIA An Overview of Issues and Prospects Country Case Study Report prepared as part of the Food Security Policy Work Program AGREP Division Working Paper No. 16 Economics & Policy Division Agriculture & Rural Development Department December 31, 1977   TABLE OF CONTENTS Page No. SUMMARY AND CONCLUSIONS ................................... . I. BACKGROUND ........................................ .. 1 A. Introduction .... I1 B. The Socio-Economic Environment.. ...... P...* 4 C. The Food Policy Context .................... .. ... ..... 5 II. OBJECTIVES OF THE STUDY ... .............................., 8 A. General Objectives and Definitions . . 8 B. Specific Objectives ................................... P. 10 III. THE METHODOLOGY ................................. ......... . , 11 A. General Statement ........................................ 11 B. The Tanzanian Case ...................................... 11 IV. PROJECTlONS OF-FOOD NEEDS . ....................... 12 V. DOMESTIC PRODUCTION PROSPECTS AND HYPOTHETICAL FOOD GAPS ................................................. 17 VI. CONCLUSIONS AND POLICY IMPLICATIONS ...... ................... 26 ANNEXES Abbreviations TGE = Tons of Grain Equivalent TSh - Tanzania Shillings ( 8 TSh - US$1)   LIST OF TABLES Table la Base Year Input Data Table lb Daily Per Capita Consumption 1968 by Income Groups Table lc Annual Consumption in Private Households Table 2 Population Projections by Age Group Table 3 Population in Each Income Class, 1969 Table 4 Human Energy Requirements in Calories Table 5 Fitted Regressions Table 6 Projected Changes in Total Daily Per Capita Calorie Consumption (Constant Prices) Table 7 Malnourished Population (Target Population) 1985,1995 Table 8 Total Food Needs, Market Demand and Nutritional Gap in Grain Equivalents Table 9 Hypothetical Food Gaps in Grain Equivalents Table 10 Grain Production 1970-76 Table 11 Food Crops 1970-76 Table 11 Other Production Table 12 Crop Production by Region Table 13 Land Use by Regions Table 14 Number of Holdings and Farm Size Distribution by Region   LIST OF FIGURES Figure 1 Projected Malnourished Population 1985 Figure 2 Projected Malnourished Population 1995 Figure 3 Projected Nutritional Gap in Grain Equivalents 1985 Figure 4 Projected Nutritional Gap in Grain Equivalents 1995 Figure 5 Projected Total Food Needs, Market Demand and Production Figure 6 Projected Per Capita Production, ImpQrts and Dqmestic Use of g.rains   SUMMARY AND CONCLUSIONS i. This is a preliminary study designed to assess the food security needs of Tanzania over a prospective period covering 1985 and 1995. The investigation is based on alternative simulations of demand and targets - total food needs, the nutritional gap and market demand for food - under varying assumptions about changes in per capita income growth rates, income distribution and changes in relative food prices. The results of the simulations are then compared with current projections of domestic food production in order to determine the changing size of hypothetical food gaps over the prospective period. ii. The basic finding is that malnourishment will be eliminated or substan- tially reduced in Tanzania if the nation is able to attain high per capita income growth rates, if it continues on the path toward greater income equality and provided that food is made available on a relatively cheap basis to the target (malnourished) population. Optimum results are likely to be achieved if the above are combined with sensible policies relating to population growth and nutrition education. iii. However, to the extent that past trends in domestic food production continue, there is no likelihood that the food gap will be closed. Under such circumstances there will be a continued reliance on food imports. Success in bridging the food gap will ultimately depend on the extent to which past performance in agricultural production can be improved upon and import capacity improved by means of strengthening and deepening the country's export base. iv. Thus, the results of the projections show quite clearly that Tanzania must now place greater emphasis on growth-oriented strategies. Since the overall national development is integrally linked with agricultural sector growth, this calls for greater efforts to raise agricultural output.   I. BACKGROUND A. Introduction 1.01 This paper reports some preliminary findings from an ongoing study of the food security problem facing selected developing countries. Within the Bank, the findings are expected to contribute to the World Development Report, which is its immediate concern. However, it is also expected to form the basis of future work on general issues related to food security. 1.02 The study recognizes the requests of the World Food Council stemming from its Communique on World Food Policy of June 1977. The Communique proposed that a catalytic role be assigned to international agencies in the areas of improving nutrition and increasing food production. In particular, such agencies were requested to assist in determining the resource, policy and program requirements on a country by country basis in order to achieve the desired increase in production and elimination of hunger and malnutrition. 1.03 An emerging consensus is that policies and strategies for promoting food security should be linked to policies and strategies for restructuring growth and reducing poverty. Since a large number of the poor are located in rural areas and engaged in agriculture and allied occupations, it is reasonable to suppose that pursuit of the food security goal demands that attention be focused on the importance of agricultural production as the major source of income for the population in most developing countries.  -2- 1.04 While there is obvious agreement that food is the principal wage good in low-income countries, there is now greater concern with the level of food intake, especially as it affects.the nutritional status of low-income households. Nutritional deprivation and the inter-related problem of ill-health are unques- tionably two of the most serious manifestations of poverty. Thus significant gains in socio-economic welfare are likely to accrue if action is directed towards the satisfaction of basic needs of target groups through qualitative in- creases in food intake and even the moderate provision of health services. 1.05 There is wide agreement that a major cause of undernourishment in developing countries lies in the inadequate intake of protein, vitamin or minerals. However, the FAQ/WHO Expert Group .1/ has observed that in the majority of cases these deficiencies result from an inadequate calorie intake, which is unavoidable reflected in energy deficiency. 1.06 It is known that for developing countries the current protein de- ficiency is very much higher than the calorie deficiency. From a longer term perspective the problem of how to meet the protein deficiency in a most cost effective manner deserves very urgent attention. This is especially urgent for children. However, in implementing a proper nutrition policy the first step is to remove the calorie gap since this would appear to be the most urgent. However, the evidence suggests that without improving the calorie intake, the protein deficiency cannot be tackled satisfactorily; this fact is of strategic importance in deciding an appropriate nutritional and food policy, and is one of the underlying assumptions of this study. 1.07 A point of departure of this study, therefore, is that policies and strategies for promoting greater food security should place special attention on the status of calorie intake and the nature of calorie deficits in developing countries. The evidence suggests that per capita income is one of the most 1/ FAO/WHO, Food and Nutrition Strategies in National Development. Report of the 9th Session, Joint FAO/WHO Expert Committee on Nutrition. FAO Nutrition Meetings Reprot Series No. 56. WHO Technical Reprot Series No. 584. Rome: FAQ; Geneva, WHO, 1976.  -3- important determinants of food and calorie intake, ana rneretore tnau ciose attention should be paid to the political and administrative feasibility of attaining income growth and the redistribution from higher income families to low income households. 1.08 However, the study recognizes that the calorie/income relationship based on the stratification of the population by homogeneous groups with respect to calorie intake should take into account factors such as regional location, urban or rural residence, occupational status, etc. Because of lack of suitable data, only the rural-urban dichotomy was employed. Further, it is noted that while energy supply is of crucial importance,nutritional status cannot be adequately defined in terms of the appropriateness of calorie consumption alone. The notion of a calorie deficit depends on the definition of a requirement level which may have no clear meaning for particular countries and in some cases may even carry unnecessary moral overtones. 1.09 An examination of the approach to the study of the food problem taken in most available studies reveals an emphasis on production aspects and/or an aggregated or regional focus. As such there have been relatively few attempts to identify and analyze the problems of target groups who are mal- nourished. A notable exception is the recent Bank staff study by Reutlinger and Selowsky 1/ who point out that the severity of the nutritional deficiency problem tends to be understated by the usual aggregative approach which neglects intra-country and inter-regional differences in incomes and food consumption. As a result it is argued that the problem of tndernourishment can be solved only by redistributing income to target groups. They assert that otherwise the needed increase in food intake would require a rate of growth of food production and consumption that could be achieved only if large subsidies were paid to 1/ Shlomo Reutlinger and Marcelo Selowsky, Malnutrition and Poverty: Magnitude and Policy Options, (Baltimore, The Johns Hopkins University Press, 1976).  -4- maintain high incentive prices to farmers and low food prices to consumers. Their conclusion is that the sloution muse lie in target-group oriented programs specifically designed to allow low income groups to achieve minimum standards of adequate nutrition. The present study is based on an extension of the methodological approach developed in the work mentioned above. B. The Socio-Economic Environment: An Overview 1.10 Tanzania is primarily an agricultural economy, with approximately 25% of its real GDP originating from the non-monetized sector and over 50% of agricultural production derived from subsistence farming. About 90% of its 15 million people are dependent on agriculture, and the country relies mainly on small farmers for its food production. 1.11 The understanding of the Tanzanian approach to food security must take account of the broader economic and social framework governing its development objectives. Since the Arusha Declaration of 1967, the country has embarked on a socialist path of development based on three-broad and interdependent strategies - increased state participation in economic activity, regional development,and rural reconstruction. 1.12 At the national level, the most important organizational changes which have been introduced include the nationalization of major industries, the establishment of state (parastatal) enterprises in a majority of economic activities (production, marketing, distribution, transport, commerce, financial services, etc.), and an increasing emphasis on planning and direct controls as tools of economic policy. The program of administrative decentralization launched in 1973 has resulted in the establishment of 20 regional governments with the responsibility of formulating and implementing regional policy. 1.13 Rural development in Tanzania is conceived both as the key to achieving agricultural production and spreading development to the masses of the people. The Arusha declaration made explicit reference to rural develop- ment and to the small farmer in the context of "agriculture as the basis of development".  -5- 1.14 The rural sector is being transformed through the establishment of ujamaa villages which bring together individual farmers previously located in scattered homesteads. The villages are conceived as compact agrarian communities, organized along cooperative lines, thereby enabling peasants to combine the instruments of production and utilize marketing, extension, credit and related facilities more efficiently. Major objectives of the villagization movement are to increase peasant participation in rural develop- ment, the reduction of income disparities and facilitating the provision of public goods. By the end of 1975 the majority of the rural population was living in newly formed villages. 1.15 These oolicies and strategies are supposed to lend -practical signifi- cance to the four basic principles of the Arusha Declaration, whicn are: (a) social and economic enualitv in conjunction with interpersonal and interregional balance; (h) self-reliance; (c) cooperative production as the ultimate objective of rural sector organization; and (d) social and economic transformation of people and production techniques. Thus the Tanzanian Government has an explicit policy tospread the benefits of development as widely as possible and to provide for the basic needs of its entire population. The Food Policy Context 1.16 The broad socio-economic framework outlined above sets the context for analyzing the approach taken in Tanzania towards meeting food security needs. On the demand side, the stated objective of the Tanzanian government is to Provide a nutritionally adequate diet. For example, in his introduction to the Second Development Plan, President Nverere states: "Although we emphasize cash crops in our agricultural policy ... our first aim is to provide sufficient and better food for everyone. Anyone travelling in this country sees poor and undernourished children, some of them sitting listlessly -  -6- lacking even the energy to play - and others easily exhausted ... The reason is not that our people are lazy. It is that they do not get enough meat, eggs, fish and milk, or other protein foods even to feed their bodies properly. Our present attitude to food is the result of ignorance, indifference and indolence. Many of our people do not realize what they can do, or why it is important that they should; and they find it easier to carry on in the old fashion than to make the effort to change the condition of their life - partly, of course, because they are themselves suffering from the effects of long-term bad feeding." 1/ 1.17 On the supply side, the fundamental objective is self-sufficiency which is to be achieved through two derived aims: (i) increased domestic food production by intensifying production of cereals (wheat, maize, rice, sorghum, millet) and to a lesser extent cassava, sugar and beans as well as increasing the productivity of the country's natural herds to satisfy the demand for meat and milk; (ii) to increase import capacity and therefore the overall purchasing power or effective demand for food by improving both the range and quantity of exports. Thus there is a recognition in Tanzania that on the supply side, achievement of self-sufficiency requires a bolstering of the nation's absorptive capacity to produce more food so that adequate reserves could be built-up and food shortages obviated. 1.18 This lesson was learnt during the early 1920's when the country witnessed a decrease in per capita agricultural and food production. The balance of payments came under severe strain partly because of a reduction in exports of main crops as well as a huge volume of food imports. A sharp contrast can be drawn between the above developments and the experience of the 1960's when the country experienced some measure of success in grain production and was able to more or less match domestic supply and demand requirements. 1/ Tanzania: Second Five-Year Plan for Economic and Social Development. Volume I, pp. xx, xxi.  -7- 1.19 The poor performance of agriculture led to the introduction of a new food policy in 1974-75. The policy was designed to boost domestic food production so to build food reserve stocks and cushion the effects of food shortages. 1.20 However, the swift adjustment to short term food insecurity masks certain problems related to agricultural investment. One of the major causes of decreased agricultural and food output was the structure of investment in Ujamaa settlements - mainly in social and economic infrastructure with long gestation periods. This resulted in low incremental productive investment in relation to rural employment requirements. Thus a combination of job unavailability for the rural labor force and lack of adequate incentives to farmers led to decreased agricultural and food output. While long term strategies for food and agriculture must be linked to rural income and employment creation, the evidence suggests that more attention should be paid in the future to achieving an optimal mix between directly and indirectly productive investment in agriculture, with relatively more emphasis placed on the former. 1.21 This fact was recognized by President Nyerere in a speech giving.a frank-assessment of the performance of the Tanzanian economy ten years after the Aru ha Declaration. While recognizing the progress made in instituting structural and institutional reforms, he placed special emphasis on the fol- lowing "mistakes and failures"- the-fact that the GDP growth rate was lower than in the period-immediately following independence; the poor production record--"the lack of sufficient agricultural growth" . . . "in food.production the increase in output has not kept.pace with the.increase in the population"; dispppointingly low rates in capacity utilization; the poor record-in managerial and worker efficiency; the failure of parastatals to generate surplus; and so on. 1 Julius K. Nyerere, "The Arusha Declaration: Ten Years After," reported ii International Development Review, Vol. 19, No. 1, 1977.  -8- II. OBJECTIVES OF THE STUDY A. General Objectives and Definitions 2.01 The fundamental objective of the study is to determine and assess the status of food security in Tanzania over two perspective periods covering 1985 and 1995. The term "food security" as used in this study relates to the country's capacity to meet its total food needs through some optimal mix of maximum feasible increases in domestic food production, a more efficient dis- tribution of available food supplies, and/or increased imports. Total food needs represents the aggregate calorie intake to sustain active working life. Thus total food needs is the same as the aggregate nutritional requirement which is the required per capita intake of calories necessary to satisfy FAO standards. This requirement depends on age, sex, body weight, and so on. 2.02 The satisfaction of food needs requires that simultaneous attention be paid to its two important elements-the actual expenditure on foods at a given point in -time as-well as any -nutritional deficiency in 'terms of calorie intake. Actual consumption expenditure on foods reflects the population's market demand for food. A well-known feature of such food demand is that it depends on a variety of factors, including per capita income growth, income. distribution and relative food prices, among others. To the extent that this basic demand for food is not met, it means that some proportion of the coun- try's population will be forced to live at starvation levels. However, a more interesting fact is that while most countries may be in a position to meet their population's demand for food from a combination of domestic and foreign sources, a large gap may still remain between actual food intake and-the popu- latons need for food. 2.03 The difference between total food needs .and actual food intake represents the nutritional gap which is the amount of food which  -9- should be made directly available to the population consuming at less than adequate levels. The identifiable feature of this group is its malnourished status. Therefore, any meaningful approach to the attainment of food security would require specific distribution and other policies designed to meet the food needs of this malnourished population which represents our target group. 2.04 The formulation of meaningful policy for minimizing the size of such target groups requires, first, a translation of their needs into foodgrain equivalents and, second, to compare this with domestic supply availability which is the amount of food available from domestic production and stocks. An optimal food security target is therefore achieved when some normative equilibrium is reached between total food needs and total supply availabilities, where the latter includes domestic production, stocks and imports. 2.05 By comparing total food needs with supply availabilities, it is possible to identify various types of food gaps. A total food gap can be said to exist when total food needs in foodgrain equivalents surpass amounts available from domestic production and net imports. This total food gap has at least two basic components. First it may represent the domestic food gap which is the difference between total food needs and amounts available from domestir- production. Alternatively, it is a domestic market gap, i.e., the difference between total market demand for food minus amounts available from domestic production. 2.06 There are two broad objectives of the study. The first is to identify and assess the relationship between total food needs and its component parts--the demand for food and the nutritional gap in calorie terms over the perspec- tive period 1985, 1995, and under alternative assumptions of changes in three economic parameters --per capita income growth, income distribution and changes in relative food prices. Second, by comparing the above estimates with estimates of aggregate production, to determine the relative sizes of the various food gaps as the economic parameters are allowed to vary.  B. Specific Objectives 2.07 In accordance with the general thrust defined above, the study has the following specific focus: (a) to determine the status of actual food demand in terms of calories consumed in the base period, as well as likely variations under alternative assumptions about per capita income growth, income dis- tribution and changes in relative food prices over the perspective period (1985, 1995); (b) to determine the size of total food needs (aggregated calorie require- ments as prescribed by the FAO) and its likely variations under assumptions stated above; (c) on the basis of (a) and (b) above to ascertain the status and changing nature of the nutritional gap; (d) on the basis of the nutritional gap, to identify the proportion of the population which is undernourished, their income profile and residential status; (e) to measure the relative sizes of alternative food gaps by comparing the results of the alternative simulations of total food needs, food demand and the nutritional gap with estimates derived from current production projections. 2.08 It should be mentioned that the objective of the study was not to make independent projections of production bvt rather to rely on previous estimates made by the World Bank and the International Food Policy Research Institute (IFPRI). 2.09 While the primary focus of the study is on estimating various demand para- meters, it is realized that the simulations under the alternative assumptions of per capita income growth, price and income distribution changes would themselves have an impact on aggregate supply. Under certain circumstances the objective of reduction or elimination of undernourishment may be attained more efficiently or at lower cost  -11- by altering the pattern of consumption rather than by increasing total supply avail- ability. Diets can be changed either by altering the distribution of income or by modifying consumers' selection of foods. However, the objective of the exercise is not to juxtapose "consumption-oriented" against "production-oriented" strategies. Our view is that even if redistribution is a realistic possibility, it may offer only a temporary respite from the need to expand food supplies to satisfy the growth of demand resulting from population increase and rising per capita incomes. III. THE METHODOLOGY A. General Statement 3.01 A description of the general methodology for all the country studies of which Tanzania forms a part, is given in Annex 1. B. The Tanzanian Case 3.02 The data sources used for the basic input data were the 1969 Household Budget Survey (HBS) and existing FAO and World Bank Reports. Because of problems of data reliability, the results should be treated with caution. Nevertheless, the conclusions of this study, however tentative, are not unrealistic. 3.03 Data for eight expenditure classes were available from the Household' Budget Survey. The eighth class, however, (.2% of the population) was dropped for the purpose of this study as the consumption data available were thought to be inconsistent with that for the other classes. The midpoint of the expenditure per class as reported by the HBS was assumed to represent the average annual income per household. Data on the average household size per class were not available and the average per capita income was derived as follows for urban sector (6%) the average household size (5.09) was multiplied by the ratio of the mean number of children to the weighted average of these means for all classes. The weights would reflect a rough pattern of distribution of children among classes. (For data and explanation of the procedure see Appendix A). Since the rural  -12- average household size was not available, the figure for each income class was estimated by adding two to the mean number of children in each class (from World Bank data). The national average household size per class was then calculated using these two sets of figures multiplied by the number of households per class per sector (from World Bank data). 3.04 The average daily calorie intake was derived by transforming the average expenditures on 14 aggregate commodity groups into calories using the FAO Food Con- version Table for Africa. The figures obtained are, however, very low. One explanation of this may be that the expenditure data reported by the HBS do not in- pute a value to the consumption of self-production by households. This view is shared by World Bank researchers on Tanzania. As explained below, the necessary adjustment was subsequently made. 3.05 The first regression of average calorie intake on income yielded a good fit for the data used. The average calorie intake was 1051 compared to the average daily requirement of 2009 derived from the World Bank's National Food Balance Sheet for 1972. A simple procedure was then used to raise the intercept by the average deficit (958) while maintaining the same calorie-income relation, i.e. the same slope. The adjusted results are shown in Tables la and 5. IV. PROJECTIONS OF FOOD NEEDS 4.01 In 1969 available food supplies in Tanzania provided an average of 2,010 calories per capita per day or roughly 10% less than the nutritional requirement of 2,230 calories as estimated by the FAO. This nutritional gap was concentrated in the bottom two deciles of the population who in the main live in rural areas. The malnourished population in this period represented 68% of the total population. The nutritional gap in grain equivalents amounted to 290,000 (TGE) in 1985. Food consumption and clinical surveys show that calorie  -13- deficiency is concentrated in Dodoma, Kilimanjaro, Morogoro, Tabora, and Karagwe It has also been observed that 20-30% of pre-school age children suffer some form of protein-calorie nutrition. As the data in table la show , the ratio of per capita cereal expenditure to total food expenditure is extremely low. Foods high in carbohydrates (cereals, starchy roots and tubers and sugars) provide about 77% of food energy at the national level. 4.02 The various calorie/income consumption relationships are shown by least squares regressions in table 5. While there is a close fit between income and food expenditure, the income elasticity of demand for calories is relatively low (0.22 at a poverty line of $80 per capita and 0.37% at 60% of the poverty line). This means a less than proportional response of calorie consumption to income increases for the malnourished population, with a high substitution between cereals and other starches in the food-basket of the typical low-income household. 4.03 The data.in table 7 (also figures 1 and 2) show variations in the size of the malnourished population over the prospective period under alternative assumptions about growth rates, income distribution and price changes. Starting with the low growth rate assumption and assuming constant prices and an unchanged income distribution, the proportional size of the malnourished (target population) reduces from 68% in 1969 to 59.8% in 1985 and further to 44% in 1995. There is very little change in the size of the target population even where there is a movement toward greater income equality (reducing the Gini co-efficient by 10%). 4.04 However, under the high growth rate assumption, with constant prices and an unchanged income distribution there is a drastic reduction in the size of the malnourished population - from 68% in the base period to 35% in 1985 and 17% in 1995. The Tanzanian economy experienced a highly variable growth performance in the 1964-75 period, so it is reasonable to assume that it has the potential of reaching much higher levels of income growth than postulated in this study. Under  -15- producer response. We do not attempt to provide answers to such difficult problems in this paper. What is clear, however, is that substantial efforts will have to be made in the Tanzanian case to achieve higher growth rates in both per capita income and agricultural production. 4.08 Analyzed sepatately, price reductions have very little effect on the size of the target population and the nutritional gap unless such reductions are combined with high income growth rates of 3.9% and a more even distribution of income. There are two probable explanations of this result. Because the subsistence sector predominates in Tanzania exchange transactions in food items are probably minimal and the size of the price effect is small. Second, even with increased monetization, the growing predominance of the state sector is likely to put a dampening effect on the operations of the price mechanism. 4.09 Reducing the size of the Gini coefficient by 10%, i.e., reducing inequality in the income distribution,does not cause significant variations in the size of 'the target 'population, even under alternative assumptions about the growth rate and changes in relative prices. This would be expected in an economy like Tanzania for the following reasons. First, the distribution of income is not as skewed as in other less developed countries. Second, the target population is concentrated in the bottom two income classes. Since poverty is more or less evenly distributed, the problem of policy is one of increasing incomes of the target group rather than redistribution from rich to poor. 4.10 The data in tables 8 and 9 show the relative variations of total food needs, market demand and the nutritional gap, all expressed in grain equivalents. The rates of growth of these target variables show a more or less constant growth trend over the prospective period,under both high and low growth rate assumptions. This probablyirepresents the effect of population growth. While there is no significant change in the nutritional gap between 1969 and 1995, under the low growth rate assumption, there is a marked difference when a high income growth rate is assumed. In this latter case, the nutritional gap shows a  -16- drastic reduction from 140,000 (TGE) in 1985 to 20,000 (TGE) by 1995. This supports our previous conclusion that the income growth rate has a positive effect on the size of the target population. Once again, the impact becomes even more dramatic when high per capita income growth is combined with price reductions and income inequality. This is vividly illustrated in both figures 3 and 4. 4.11 As shown in Figures 1-5, the target group is highly sensitive to variations in the minimum calorie requirement of 2230 4 and the -food needs are fully met in the high growth, more equal income distribution scenario using the calorie requirement reduced by 200 in 1995 (figure 2). Even by 1985, with the lower calorie intake, the size of the target population drops from over one half to around 10% (at constant prices). This sensitivity could indicate that the FAO yardstick for energy requirements is less suitable for Tanzania than for other countries. Secondly, the nutritional standards still might not be met for the target -populations who are borderline , as poor dietary practices (unbalanced family consumption, leaving children malnourished, taboo foods, or preference for lower nutrition sugars or limited availability of milk or high protein foods) persist. 1/ Derived from FAO/WHO guidelines for Tanzania, using 54 kg average bodyweight males, 46 kg average bodyweight females. _mpor L  -17- V. DOMESTIC 'PRODUCTION PROSPECTS AND HYPOTHETICAL FOOD GAPS 5.01 In table 9 comparisons are made between the target variables on the demand side (total food needs and market demand) with trend production projections made by the World Bank and the International Food Policy Research Institute (IFPRI) The objective is to identify various hypothetical food gaps which might emerge over the prospective period. Assuming low rates of growth in per capita income and agricultural production, the country will experience a domestic food gap of 2.1 million tons (TGE) by 1985 with a slightly smaller domestic market gap of 1.81 million tons (TGE). By 1995 the domestic food gap will increase by approximately 50% to 3.45 million tons (TGE), and the domestic market gap will increase by 6% to 3.2 million tons (TGE). 5.02 Assuming high growth rates in both per capita income and agricultural production, the domestic food gap will increase by 3% from 430,000 tons (TGE) in 1969 to 700,000 by 1985. Between 1985 and 1995, it is likely to increase by 550,000 tons or a 4% increase. The domestic market gap will increase from 560,000 tons (TGE) in 1985 to 1.23 million tons by 1995, a 9% increase. As illustrated in Figure 5 (also table 9), with high growth rates, the nutritional gap will be more or less eliminated around 1995 as market demand approximates total food needs. As expected at that point the market gap (1.23 million TGE) is more or less the same as the food gap (1.25 million TGE), which suggests the greater relative importance of income growth at this point. 5.03 However, when compared with the high growth rates in agricultural production as postulated by IFPRI (3.1%), there will be an increasing divergence between food needs, market demand and available supplies from domestic production. Since both the World Bank and IFPRI estimates are based on historical trends in production, a crucial question arises about the ability of the Tanzanian economy to reasonably improve its growth in agricultural production as well as its import  -18- capacity to purchase required food supplies from abroad when necessary. While questions of this nature cannot be fully answered in this report, a preliminary assessment of likely developments is presented below. 5.04 Taking a pessimistic view, to the extent that past trends continue Tanzania will be forced to rely on imports to meet its basic food needs. This--Ice is illustrated in the FAG projections found in figure 6. According to Bank estimates, total agricultural production in real terms grew at an average rate of only 2.8% per annum between 1964 and 1975. When adjustments are made for the effects of the 1974 drought and the partial recovery of 1975, the normalized growth rate, e.g., for the 1964-73 period is the same as that used by IFPRI - 3.1%. However, with an estimated rate of population growth around 2.8% per year for the 1964-75 period, the expansion of agricultural production was just sufficient to keep abreast of population growth. Range of Annual Growth Rates for the Agricultural Sector 1964-75 Growth Rate No. of Years - 4.0 to 0.0 3 0.1 to 1.0 3 1.1 to 3.0 0 3.1 to 5.0 2 5.1 to 7.0 1 7.1 to 9.0 1 9.1 and above 1 SOURCE: World Bank  -19- The rate of sectoral growth decelerated from an average level of 4.0% between 1964 and 1967 to 2.6% between 1970 and 1973 and 1.5% between 1973 and 1975. The overall output growth in agriculture is disaggregated below in terms of rates of growth of subsistence production and monetized output. While subsistence output expanded at a steady average rate of 3% Averag,e Annual Growth Rates of Agricultural Production. 1964-69 1969-73 1973-75 1964-75 Subsistenec Production 2.9 4.0 4.D 3.4 Mlonetary Production 3.8 2.6 -1.3 2.1 Total rroduction 3.3 2.9 1.5 2.8 per annum, monetized output grew at an average rate of 2.8%, with a much more rapid growth at the beginning of the period than in recent years. This resulted in a declining share of monetized output in total agricultural production from 52% in 1964 to 41.5% in 1975 (current prices) and from 47.5% ,to 44% in,constant 1966 prices. 5.05 The highly variable growth trends for the sector were also replicated in the foodgrains production subsector as shown below. Average Annual Growth Rates of Foodgrain Production Period Growth Rate 1961-65 2.24 1966-74 -3.45 1969-74 -2.03 1961-74 0.57 It should be noticed that the overall trend growth in foodgrain production was substantially lower than that for the agricultural sector as a whole.  -20- 5.06 Some general conclusions can be drawn from the above picture. If Tanzania is to meet its total food needs by the target period, ways would have to be found of preventing the trend rate of growth of overall agricultural production from fluctuating so sharply as it has done in the past. To the extent that the constraints to growth stability lie in natural factors such as droughts, this would call for supplementary measures for building grain stocks as well as for increasing the nation's import capacity. While the average growth rate of subsistence production is encouraging, policies will have to be directed towards reducing the variability and increasing the growth of monetized output. 5.07 The factors mentioned above are also important for the nation's import capacity since the poor performance of the sector to some extent reflects the output capacity for export crop production, and therefore the .foreign exchange earnings capabilftyw 5.08 Over the past several years, between 60 to 70% of Tanzania's export earnings were derived from agricultural products, consisting mainly of cotton, sisal, cashewnuts, tea and tobacco, with sugar being exported for the first time in 1975. Export volume has shown a disappointingly slow growth for several years. Between 1968 and 1972, its overall growth was 3% per annum, but in 1975 it fell back to the 1973 level. For most of the major crops, export volume was below the levels recorded for 1973. Tanzania's export volume represents a relatively small percentage of the total world trade in the crops mentioned., and is essentially a price taker for her exports on world markets. For purposes of analysis we take the years 1974 and 1975. Unit prices for most -exports moved very favorably in 1974, increasing-on--avexage- by some 55%. However, total export receipts increased by only 9%, since the drought affected cash crops, contributing to a drastic decline in the volume to be  -21- of exports of cotton, coffee and sisal. There was a further reduction in export supply in 1975 resulting in a 6% decline in export value. While coffee and clove production showed some recovery, cotton exports suffered sharply from the diversion of acreage to food crops, falling to roughly 60% of the 1973 level. The combined output of sisal and sisal manufactures declined in response to greatly reduced export prices. 5.09 The evidence suggests that the factors responsible for the generally disappointing export performance were the same as those which adversely affected food production. Most researchers agree that the list should include inadequate producer prices up to 1974, the incidence of adverse weather conditions in 1974-75, and until recently, priority investment allocation to the non-agricultural sector, coupled with inefficiences in the transport and distribution systems for marketed output. 5_10 Even-if the country is to rationalize its Aomestic agricultural production, the problem of increasing its foreign exchange earnings, i.e. import capacity would remain. Trade can make an optimum contribution to the solution of food problems when the exports of developing countries have a favorable access to the markets of their more developed counterparts, and on favorable price terms. Thus it can be argued that Tanzania's ability to meet its food needs is to some extent dependent upon some exogenous factors beyond her control. 5.11 The main thrust of the argument so far is that the food. security situation in Tanzania is closely tied to the prospects for national development in terms of both income and product. Achievement of this objective must be tied to strategies for increased domestic production as well as strengthening and deepening the export base. The Tanzanian experience shows at times the two strategies of import substitution in food and export promotion of cash crops to be  -22- competitive rather than complementary, and ways will have to be found of developing an optimal balance between these two strategies. 5.12 Tanzaniats total import bill rose from Tsh 1,400 M in 1969 to nearlyhe Tsh 5,700 in 1975. Between 1970 and 1974 total imports increased at an average annual rate of 15X with the ratio of agricultural imports to total imports rising from 7% to 20%. This was as a result of the higher percentage of grain imports following the drought years 1972/73 and 1974/75. Estimates of grain imports for 1974 vary, but most fall between 400,000 to 500,000 tons. This is very close to the estimated deficit of 536,000 by 1975. Thus a strategy based on import substitution in food combined with export promotion is likely to produce the requisite benefits in terms of income growth, increased domestic agricultural production and lessening the strain on the balance of payments. 5.13 These conclusions are by no means startling. In fact they come close to the findings and policy conclusions of regional Bank staff as well as other analysts. While there might be disagreement over the precise orders of maenitude, the results do give a general indication of policy options. To recapitulate, the projections show that in order for total food needs to be met in Tanzania, policy will have to be directed toward raising per capita income levels as well as the growth rates in the overall agricultural sector and its various subsectors. 5.14 Some statistics on trends in the domestic production of major food items are given in tables 10 to 14. This study does not duplicate the competent analysis of production trends which forms part of the ongoing sector work performed by Regional Staff. In the time available for preparing this study, it was not possible to make any competent independent judgements about likely prospects for food production on a crop-by-crop basis. It is envisaged that this aspect of the problem  -23- will receive more extensive treatment in the second phase of this food policy exercise. We therefore conclude with a brief reference to a few ways in which the organization of production in Tanzania can make a useful contribution to the nation's attainment of food security. 5.15 Food production in Tanzania is organized in three subsectors - the subsistence, ujamaa villages and "modern" subsectors. The greater proportion of production for the subsistence subsector comes from traditional agriculture. While there has been a steady growth in subsistence ouptut between 1964 and 1975, it is equally true that this subsector has the lowest levels of production, with the greater part of output entering auto-consumption. It is envisaged that this subsector will disappear as the process of villagization takes on added momentum. The 1972 Agricultural Census reported a total of 2.5 million farms holding 3 million hectares of land. Roughly 83% of the holdings are below 2 ha in size, and roughly 97% below 5 ha (see table 13). 5.16 The ujamaa village movement is designed to consolidate the agri- cultural base of the country. While it has led to a substantial disruption of production in the short run, it offers great potential for increased agricultural production if increased local participation is combined optimally with other services provided by the central government. These relate to irri- gation, extension, subsidized inputs, water and other public goods important to rural welfare. While the production of cash crops in these villages will provide an important source of earnings, there should be clear food policy objectives for each village. Accordingly, each village could work out its own food requirements and plan its food production to meet those needs. This should form part of an overall strategy for increasing directly productive investment in the agricultural sector. 5.17 The modern sector includes those farms with high investments for  -24- mechanization, crop protection, irrigation, etc. These concentrate in the production of wheat, rice, sugar, livestock, coffee and tobacco. The sector includes the state farms as well as those expatriate farms which were nationalized. 5.18 The organization of food production can also be looked at in spatial terms, i.e. on a District basis, where the latter comprises an aggregation of villages. This regional dimension is important for food policy. While Tanzania has a national policy for self-sufficiency in food production, there is no clear indication that this is matched by a policy for attaining an optimal inter-regional balance in food security. Such a regional focus should emphasize inter-regional specialization. Under such a system there would be a concentration of development efforts on raising productivity in those regions with a high production potential, and then redistributing to deficit regions. In this context, a careful balance will have to be maintained between export commodities and those food items to be distributed on an inter-regional basis. This means that the regional food policy should have as an objective the feeding of those regions which specialize in export crops. Regional surpluses and deficits should be balanced out either by transferring surpluses to deficit regions and/or to the national level. 5.19 However, plans for food security on a village, district or regional basis can only bear fruit if these are properly linked to national plans, especially where such efforts call for finance or the provision of other services from the center. Such national coordinating efforts will be particularly crucial in areas such as marketing and distribution; the coordination and implementation of village production plans; the conducting of applied research and the provision of extension staff to dis'seminate research results; providing adequate infra- structure, transportation, storage, processing etc. for priority food items; increasing the efficiency of those state farms and para-statal organizations, which  -25- provide services at the rural and regional levels; and so on. Such technical, institutional and economic constraints continue to pose problems for Tanzanian agriculture, as they do in other less developed countries. 5.20 In conclusion, it seems quite clear that in order to achieve food security Tanzania should now pursue what is known as a "unimodal" agricultural strategy based on a progressive modernization of its subsistence sector.-/ Such a strategy has paid rich dividends in countries like Japan, Taiwan and South Korea. Given that Tanzania is making some progress in reforming its basic political, social and economic institutions, it seems logical that it should build on the foundations already laid for providing equity and basic needs, by now instituting strong programs for growth. 1/ For theoretical and empirical support of the "uniiiodal strategy" see B.F. Johnston, "Food, Health, and Population in Development," Journal of Economic Literature, (Sept. 1977); B.F. Johnston and P. Kilby, Agriculture and Structural Transformation" Economic Strategies in Late Developing Countries (Oxford University Press, 1975); Y. Hayami and V.W. Ruttan, Agricultural Development: An International Perspective (John Hopkins Press, 1971).  -26- VI. CONCLUSIONS AND POLICY IMPLICATIONS 6.01 To meet the projected demand in 1985 or 1995, whether at current levels of calorie consumption or at nutritionally adequate levels of daily calorie intake, the Tanzanian government must rely on expanded agricultural production or an import policy on major commodities, or both. Historically, the domestic agricultural policy shaped by the Arusha declaration has been based on ujaAma (villagization), parastatals' procurement and processing, subsidies of inputs such as fertilizer and irrigation development. However, the heavy investments in social and economic infrastructure have severely restricted the funds available for directly productive investment in agriculture, and consequently there has been a decrease in per capita agricultural and food production. Future policy therefore calls for a more optimal mix between direct-ly and indirectly productive investments-iii agriculture. -Such increased investments will help to promote the interdependent goals of increased agri- cultural and food output, employment creation and rural reconstruction. 6.02 During the past decade agricultural investment per additional worker in Tanzania amounted to Tsh 1430 (US$200). The FAO Perspective Study estimates that in order to achieve a growth rate in agricultural production of 3.5% per annum, investment would have to increase to approxi- mately Tsh. 1530 per additional worker (or $30 million per annum, on average). Since it is reasonable to assume that such investment requirements cannot be adequately met-from domestic resources, massive increases in aid will be necessary. 6.03 Policy will also have to be directed toward preventing the trend rate of growth of agricultural output from fluctuating so sharply as it has done in the past. Because this is caused by natural factors such as the  -27- drought of 1972-74, supplementary measures could include building grain stocks whether at the Ujamaa level or in specific vulnerable regions, e.g. Kigoma. Tanzania is taking steps to establish a Natural Strategic Foodgrain Reserve and a Foodgrain Reporting and Early Warning System.-/ 6.04 Ideally, the foodgrain reserve' corresponds to the minimum stock level which must be held at any given time so as to deal effectively with unpredictable events, emergencies that disrupt normal supply/demand relation- ships, or at least to provide immediate relief. These reserves would have to be in addition to the operational and seasonal stocks which the National Marketing Corporation (NMC) holds to satisfy normal supply commitments. 6.05 Tanzania lacks an efficient system of foodgrain reporting or crop forecasting. This has been.one,of the most important contributory factors to the foodgrain crisis-of 1972-74. The evidence suggests that the crisis would have been foreseen earlier and action taken to meet the food deficit through imports, had the necessary information been available. The organiza- tion of both an efficient grain reserve and a reporting and early warning system would require projects which are externally financed in the main. 6.06 Policy to raise agricultural production would have to take into account producer prices. The primary focus of this study has been on consumer prices. However, since the attainment of food security depends on food availability at relatively low prices, or alternatively higher incomes to purchase food at existing prices, the overall food pricing system would 1/ See FAO, A Policy and Action Plan for Strengthening National Food Security in Tanzania, World Food Security - Country Reports ESC/FSAS/TAN (Rome: FAO, August 1976).  -28- have to be rationalized so that farm incomes could be increased. Government policy through parastatals such as the Marketing Development Bureau and National Milling Corporation includes foodgrain procurement and marketing at the prices set by the Economic Committee of the Cabinet. Until 1974 maize and rice prices were kept low, which exacerbated the already low production due to drought. Government policy through parastatal pricing must take into account both consumer pricing and producer prices; the supply response to producer price increases in 1975, for example,increased total production of maize and rice by 50% in the two year period after the price increase. 6.07 The continuation of a realistic producer price system can be combined with a system of subsidized inputs currently being carried out at both the regional and national levels. Fertilizer was fully subsidized until 1975 when the subsidy was cut to 75% in selected regions and further to 50% by 1976. Fertilizer distribution channels are good in more developed agricultural areas (Mwanza, Mbeya, Iringa, etc.); outside these areas farmers have trouble getting fertilizers, and credit channels are also poor. Both the choice of subsidy and the input distribution must be incorporated into a national agricultural strategy to assure increased agricultural production of foods (as well as export production). 6.08 A clear and steady national food security strategy should pay attention to marketing and distribution, co-ordination of village production plans, applied research and the provision of extension staff, infrastructure, storage and increasing the efficiency of state farms and parastatals. However, reorientation at the national level should be accompanied by  -29- policies to meet regional and village level meeds. Such a policy could be based on inter-regional specialization, with some regions specializing in cash crops and having their food needs met by regions specializing in food production. Regional surpluses and deficits can then be balanced by either regional or national transfers of the surpluses. The inter-regional approach to food security and overall development should offer great scope for increased Bank lending. 6.09 Under such an arrangement, special attention would have to be paid to the needs of the Ujamaa villages. Ujamaa should lead to more equal income distribution and has been used for social and education programs for the rural population. In addition, the village grouping has been important in government agricultural policy, as it is a locus for both export crop and food crop development. Ideally, Ujamaa could be a backbone for extension services and the basis. for viable cooperatives in Tanzania. However, food crops continue to be principally the smallholder's domain or are grown in small plots for autoconsumption, so may not be affected by extension work or cooperative development. In addition, the Ujamaa's potential has not been reached as technical assistance programs and extension development have lagged. While indirect, these are investments that the government must consider as part of an agriculture strategy. 6-.10 Increasing growth in per capita income and output would also require that a sensible population policy be designed. There are two aspects of such a concern - one relates to rural-urban migration, the other to population growth itself. The recent rural-urban migration has increased by 9% annually,(1970-75). This rural exodus exceeds the urban wage employ- ment which has grown at 5.6%, resulting in higher food demand in the (monetized) marketplace but not increased per capita income due to unemploy-  -30- ment. Secondly, the migration signals a decrease in subsistence farming. In sum, this implies that more output is necessary for an urban population at the same time that autoconsumption decreases. 6.11 Demographic projections show that some 2.3 million persons (approximately 150,000 a year) will join the agricultural labor force by 1990. Although the rural development efforts are likely to provide employ- ment for some portion of this increase, there is a likelihood that wage employment will continue to predominate. Therefore substantial underemploy- ment will arise unless the rate of agricultural investment is stepped up. In any event, the situation would seem to call for a sensible policy on population and migration. 6.12 Policies for increasing domestic production must be integrally linked with import substitution and export promotional efforts. The country's ability to close the total food gap depends not only on its ability to increase domestic production, but also on its future import capacity which in turn depends on its export capability. The earnings from six agricultural exports are responsible for more than 50% of the country's total foreign exchange earnings with unprocessed agricultural products providing over 70% of all exports. This reliance on primary products is not different from the experience of other developing countries but it does highlight the fact that the Tanzanian government must try harder to diversify its export base.  ANNEXES  Annex 1 Page 1 The Methodology Used The methodology has two parts. The first part is the estimation from household consumption expenditure studies, the relationships between calorie consumption, food expenditure, cereal expenditure and income levels. These relationships allow the estimation of the number of malnourished people, the size of the nutritional gap and the minimum income necessary to purchase an adequate food bundle. The second part is the projection of calorie consumption, the nutritional gap, and the size of the malnourished population under alternative development scenarios and policy options. These include alternative growth rates in income per capita, redistribution of income, changes of the price of cereals, and other food policy options such as sub- sidies. While the main emphasis of the study at this point is on these two aspects of consumption and food distribution policies, the study also examines the prospects and issues on the supply and production side of food policy. The basis of the first part of the exercise is the estimation of the calorie consumption income function from household con- sumption expenditure studies. This function shows how calorie consumption varies by income levels. It is estimated by using simple econometric techniques with data from the consumption and expenditure studies. From this relationship between calorie intake and income, the effect on calorie consumption of changes in per capita income growth rates and income distri- bution can be estimated. For a particular income per capita level, calorie intake can be found and compared to that calorie intake that is required to maintain adequate levels of activity. Differences between actual con- sumption of calories and the requirement determines the nutritional gap Since the number of people that have income below that adequate to meet  Annex 1 Page 2 this calorie requirement is known from the income distribution of the country, the number of malnourished people can be estimated. This function besides allowing the measurement of the number of people malnourished and the extent of the deficit in terms of calorie intake also shows the influence of dietary habits on the nutrition problem. For some countries the relation- ship between calorie intake and income levels is very steep indicating that income increases allow foods that are high in calories to be purchased - for instance fats and oils as in the case of Morocco. In such countries income increases or redistributions would have a major impact on calorie consumption. In other countries the relationship between calorie intake and income is relatively flat which indicates that the impact on calorie intake of rising income is generally low because either other high calorie items are not available or are not within traditional consumption patterns, as in Pakistan, or additional income is going to meet other basic needs (housing, clothing or possibly education as might be the case in Brazil). For the second part of the projection exercise, a cereal price term that reflects how calorie consumption would vary with changes in prices is added to the calorie consumption income relationship. This price term in weighted by the ratio of the expenditure or cereals to total expenditures so as to capture the differential impact that a price change might have in influencing calorie intakes as budgetary shares vary. Since this cereal to total expenditure ratio varies by income class the impact of the price change will also depend on income levels. The calorie consumption function used in the projections therefore has two terms: an income term, which measures the effect on calorie intake of changes in income, and a price of cereals term, which measures the impact on calorie intake of changes in the price of cereals. Using this relationship between calorie consumption  Annex 1 Page 3 and income and price of cereals, the potential food deficits, the malnourished population and the market demand for food can be determined in 1985 and 1995. The projections to 1985 and 1995 of the potential food deficits and malnourished population have two distinct components. The first component the consumption projections involves determining the market demand and nutritional gap originating independently of any direct production or supply scenarios. Supply enters indirectly into these supply projections in that there would exist a supply level along with a government pricing or distribution scheme consistent with the consumption level. For example, a consumption scenario that has a 50% price rise implicitly assumes, in a market economy, that supply (production plus net trade) and demand are such.-,as to have- yielded this price i9crease or, as-Ln-the scase of a non- market economy, where government rationing R;ight exist along with some market transactions, that market price and a ration price have risen by 50% and the supply consistent with consumption has been made available either through production or trade. If such a supply has not been made -available then market and ration shop demand would not have been fulfilled at the given prices and future price rises or further rationing would be necessary. In summary, the consumption projections examine the effect of per capita income, income distribution, and price changes on market demand by income class and the resulting deficits in consumption that would exist for those groups having insufficient funds to purchase required amounts of food. The second component of the projections dealing with production attempts to measure the levels of production that would exist 'n 1985 and  Annex 1 Page 4 1995 if current trends and policies continue; also the levels of production that would exist in the same years under alternative policy options or if additional resources are transferred to food production. The difference between the total food needs of any particular consumption scenario and the production projections yields the food deficit or surplus that would need to be imported or exported on the condition that only market demand is being met and the condition that both market demand and direct distri- butional needs of the poor are being satisfied. The projections of consumption and production hence measure the effects of several different components of the food policy. They yield the level of supply that would be necessary to fulfill market demand and total food needs under several different income, price and income distri- butional changes. They also show the amount of food that needs to be made available through direct distributional channels in order to eliminate malnutrition. Comparison of these food needs with-the supply availability as presented in the production scenarios determines the food gap which could be closed through a combination of policies for increasing domestic production, increasing imports and/or delivering food directly to the mal- nourished population.  Appendix A TANZANIA Derivation of the Urban Average Household Size per Class (Urban average household size: 5.09) Mean Number of Expenditure Children Mean Average Class Urban Rural Percent Percent Household Size (1) (2) (3) (4) (5) 0 - 999 1.4 1.62 0.08 0.112 3.0 1,000 - 1,999 2.2 2.38 0.12 0.264 4.0 2,000 - 3,999 1.8 3.28 0.10 0.18 3.7 4,000 - 5,999 3.1 3.37 0.17 0.53 6.5 6,000 - 7,999 2.8 3.87 0.15 0.42 5.9 8,000 - .9,999 33.2 4-29 0.18 O;o576 6.7 10,000 - 24,999 1.6 5.70 0.09 0.144 3.4 25,000 2.1 1.00 0.11 0.231 4.4 Total 18.2 25.51 1.00 2.4 38.2 Source: Tanzania: 1969 Household Budget Survey. The mean number of children is presented in column 1, its percent equivalent in column 3 and the weighted average number of children (2.4) using the percent distribution per class as weights is obtained in column 4. The average household size (column 5) was obtained by multiplying the urban average household size, 5.09, by the ratio of the mean number of children per class and the weighted average number of children, 2.4.  OOZ- TaAal luatuaiTnbea TaOTEoe OOZ -doz TaA&aT juawaaTnbaa aaoTsa pauxnsse V ' tUJIl1InlaxusIa 3!!03NI lunzijJ '1U I oH1I' -+.- -- -N- N~)ifljIu.~gj tA3NIO~5J ]vf ]j4 I;nJWL - - - -- - -- UOIN231f ] !33O niNUHO I Id 0*0 ot OlOZ 00 003 - 04 07 - - ~~. .-. ~~~~- - -~~ r-r- T-rr- o cn -1 - 0- 01 -~~~~F+ - - _ _-- - CD, ooz- D -'do, olz n ooz- as _- ~ ~ ~ ~ ~ ~