Agricultural pricing and marketing policies in Malawi : a multi-market analysis

DRAFT FOR STAFF USE ONLY AGRICULTURAL PRICING AND MARKETING POLICIES IN MALAWI: A MULTI-MARKET ANALYSIS J. Kirchner (Consultant) I. Singh L. Squire CPD Discussion Paper No. 1985-17 May 1985 CPD Discussion Papers report on work in progress and are circulated for Bank staff use to stimulate discussion and comment. The views and interpretations are those of the authors. PRELIMINARY DRAFT (For Discussion Only) AGRICULTURAL PRICING AND MARKETING POLICIES IN MALAWI: A MULTI-MARKET ANALYSIS by J. Kirchner, I. Singh, and L. Squire* May 1985 * The authors are Consultant; Senior Economist, Economic Policy and Management Unit, Eastern Africa Region; and Division Chief, Resource Mobilization and Public Management Division, Country Policy Department, at the World Bank, respectively. The views expressed in this paper should not be interpreted as the official view of the World Bank Group. Disk8504/alagrpol/5-01-85/ae Agricultural Pricing & Marketing Policies In Malawi: A Multi-Market Analysis ABSTRACT This is the second in a series of papers describing analytical tools for examining agricultural pricing in East Africa. It documents a multi-market model for the analysis of agricultural pricing and market- ing policy in Malawi. This model was used to explore the implications of a range of alternative policies in preparation for price policy dis- cussions for the 1984/5 season. Because many of the policy issues exam- ined here are also shared by other African nations, this analysis serves as a case study demonstrating how similar approaches could be applied elsewhere in the region. The model is designed to illustrate the impact of a variety of policy instruments on diverse and often conflicting policy objectives including food self-sufficiency, parastatal revenue, foreign exchange earnings, and smallholder and urban real incomes. The primary policy instruments examined in this analysis are producer prices for the five major crops and fertilizer, consumer prices for the major food crops, and export taxes. The supply and demand equations of the model are linked together, allowing the user to trace the effect of a change in one market on other markets as well. For example, a change in the pro- ducer price for maize does not simply alter maize prodoction; it also affects production of other crops, demand for farm inputs, and the farm household's pattern of own-consumption. The results suggest that one can improve on the present pricing policies in terms of almost all goals. In particular, adjustments in maize pricing that reduce unsalable maize surpluses are shown to sub- stantially improve both parastatal cash flow and balance of payments. Movements of fertilizer and export crop prices toward their parity values are also demonstrated to result in improvement in most of the policy objectives. AGRICULTURAL PRICING AND MARKETING POLICIES IN MALAWI A MULTI-MARKET ANALYSIS PAGE I. INTRODUCTION II. THE AGRICULTURAL SECTOR IN MALAWI 2.1 The Smallholder Sector 2.2 Estate Sector 2.3 ADMARC III. MODEL STRUCTURE AND CALIBRATION 3.1 Smallholder Agriculture 3.2 Urban Consumption 3.3 Estate Agriculture 3.4 Market Structure 3.5 Calibration IV. POLICY ANALYSIS: ALTERNATIVE SCENARIOS AND THEIR IMPLICATIONS 4.1 Government Objectives 4.2 Alternative Policy Scenarios 4.3 Models Results 4.4 Evaluating the Policy Alternatives ANNEXES A The Model Specification B The Detailed Results Disk8504lagrpol/5-01-85/ae I. INTRODUCTION Agricultural pricing issues have become increasingly important in all African countries in the past decade. Following the oil shocks in the seventies, most Eastern African countries have been faced with a variety of external shocks including rapid increases in the price of fuel and capital goods, a decline in the price of primary exports, an increase in the cost of borrowed capital and declining official and private capital transfers. These external factors have created enormous sti9in on not only the external balance of payments -- in some cases of crisis proport3ons -- but also 'structural' imbalances in the domestic economy. These have been reflected in rising budget deficits, domestic inflation, declining real investments, excess capacity and in many cases, declining output and export earnings. While some economies have attempted to adjust to these adverse conditions by changes in domestic fiscal and pricing policies, in most these adjustments have only recent- ly been undertaken. A major component in this structural adjustment process has been the heightened concern for agricultural pricing poli- cies and exchange rate regimes. Agricultural prices have been of central concern both because of the need to provide producers with "adequate incentives" so as to reverse trends in declining output of both fo.odgrains and export crops, as well as to reduce government deficits due to implicit as well as explicit subsidies provided to both producers and consumers, subsidies which government in their presently strapped circumstances cannot Disk8504/alagrpol/5-01-85/ae - 2 - afford. There are other concerns on which agricultural pricing policies also impinge -- the concern for adequate government revenue, for inter- sectoral and intrasectoral or regional equity, for "food security" often seen to be the same as "self sufficiency" and for the real welfare of consumers and producers. While governments are often clear about the objectives of agricultural pricing policies, they are reluctant to accept donor advice on what to do about agricultural pricing policies in any given situa- tion. This is partly due to the perception that donors do not share the same concerns or attach different weights to alternative objectives. But part of the reluctance also stems from the inability of donors to provide policymakers with a credible analysis of the likely impact of alternative policy scenarios or recommendations on what they perceive to be complex, multiple and interdependent objectives. Unless such an analysis is available, policymakers will continue to express reluctance about the recommendations being made to them. Recommendations of one kind or the other relating to agricultural pricing are increasingly becoming a part, not only of the economic and sector work, but also of the project and structural adjustment loan documents at the World Bank. Agricultural pricing issues are also central to the dialogue between the Government of Malawi and The World Bank. Proper advice, however, needs to be based on effective analysis of the role and impact of agricultural prices. Malawi shares with many African countries certain features in its institutional and pricing setting that make a case study of Disk8504/alagrpol/5-01-85/ae -3- agricultural pricing policies in Malawi of greater general relevance. In its institutional framework, it shares at least the following features with many African economies: (a) the extensive government administration of prices, where government parastatals, market- ing boards or other agencies set not only producer prices for most inputs and outputs, but also set prices for urban and rural con- sumers as well as the margin of private trad- ers and processors; sometimes these price controls are combined with the control of marketing and distribution functions, with a legal monopsony/monopoly position granted to the state agencies; (b) the possibility of "unofficial" markets, some- times tolerated, but most often castigated, alongside the "official" markets; these "un- official" markets arise both where state con- trol cannot be easily enforced (as in the case of millions of smallholders producing food- crops) or in conditions where prices offered by state agencies (agents) do not reflect the true scarcity value of commodities in "free" markets; and Disk85O4/Mlagrpol/5-01-85/ae -4- (c) the existence of extreme dualism, with a vast majority of the rural households in the small- holder sector farming small plots and produc- ing mainly for subsistence but also for domes- tic and export markets, alongside a highly commercial, more capital- and skill-intensive estate sector -- either government or private- ly run -- with a relatively small proportion of rural households farming large areas and producing mainly for domestic and export mar- kets and consisting often of expatriates or local elite. Apart from these institutional features, Malawi also shares with many African economies the concerns over similar types of policy issues relating to the agricultural sector. These include a concern for: (a) the prices of the major non-tradeable, mainly food grain commodities; on one hand, the prices paid to producers to provide "adequate incentives" so as to encourage adequate domes- tic production to meet "food security" con- cerns and provide adequate incomes to produ- cers and on the other, the prices charged to Disk8504/Mlagrpol/5-01-85/ae -5- the politically sensitive mainly urban consum- ers protect their "welfare"; (b) the taxing of export crops to provide govern- ment revenue; partly because export crops are easier to tax and because they have always been the main source of government revenue in most African countries (marketing boards were created for this purpose); (c) the subsidizing of imports, mainly imported inputs like fertilizer and fuel; these imports are becoming more and more costly, and at- .tempts to keep prices low for producers often lead to large budgetary deficits; and (d) pan-territorial pricing, where due to con- siderations of equity, government agencies administering prices, charge all consumers the same prices for goods and pay or charge same prices to all producers irrespective of their location and the transportation or other costs. Such pricing policies imply an impli- cit subsidy or tax depending on the location and can at times become a major deficit item in government budgets. (Malawi is a relative- ly small country but, nevertheless, internal Disk8504/Malagrpol/5-01-85/ae -6- transportation costs from one region to anoth- er are not insignificant.) Given these common institutional features and policy concerns, a study of agricutural pricing issues in Malawi has a greater general relevance to other African economies. This report explores these issues in the specific context of Malawi. The remainder of this report is organized as follows. Section II provides the descriptive details on Malawi's agricultural sector and its institutional setting, Section III describes the framework and the application of a multi-market model developed to analyse the impact of alternative agricultural pricing and marketing policies on selected government objectives; it also describes how the model was calibrated against observed data to "test" it before using it for policy analysis; Section IV provides an analysis of the impact of alternative policy scenarios based on model results and discusses their implications. Annexes on the algebraic specifications of the model and on the detailed model results conclude the report. II. THE AGRICULTURAL SECTOR IN MALAWI An Overview The purpose of this section is to present the salient features of Malewian agriculture that are relevant to the present study. This Disk8504/Malagrpol/5-Ol-85/ae -7- section is divided into three parts. The first part examines Malawian smallholders, the second reviews the estate sector, and the third looks at ADMARC's role in agricultural marketing. The treatment here is not intended to be exhaustive; it is meant instead to highlight the most significant and fundamental of the available data. Unless noted other- wise, all information is for the calibration reference year of the model, 1981. Table 2.1. Malawi: Land Use, 1978 (millions of acres) North Central South Malawi Total acreage 6.64 8.79 7.81 23.24 Arable land 2.22 3.78 2.87 8.88 Smallholder land 1.22 3.01 2.88 7.10 Estates .24 .59 .33 1.16 Source: Ministry of Agriculture and Natural Resources, cited in IBRD Report £ 3459-MAI, p.2. As Table 2.1 indicated, 80 percent of Malawi's cropland is farmed by smallholders. Note that except in the north, there is little or no unused arable land available for agricultural expansion. In fact, when one considers that some arable land is occupied by villages, roads, and the like, it seems clear that, particularly in the South, more land is being farmed than is arable on a long-term basis. Disk8504/Malagrpol/5-01-85/ae - 8 Past agricultural production increases were achieved primarily by expanding the land area under cultivation. However, almost all of the arable land is now being cultivated, so further increases in produc- tion will have to come through improving yields. Estates and smallholdings are defined solely by land tenure, rather than by size or management system. Any agricultural enterprise on freehold or leasehold land is by definition estate agriculture. Smallholder agriculture takes place on "customary land", that legally belongs to the head of state and is administered by village chiefs. Alienation of land from customary tenure through leases is increasingly common where estate agriculture's advantages are evident. Estates and smallholders each may grow only certain crops for cash sale. The major estate crops are: burley tobacco, flue-cured to- bacco, tea, and sugar. Each estate generally grows only one crop for sale. In addition each estate grows subsistence crops to meet the food requirements of its workers, but not for market. In March 1979, estate agriculture occupied 13 percent of cultivated land, and was growing at roughly 10-15 percent per year, largely through alienation of customary land. Estates market their crops directly at auction or through private contracts. 2.1 Smallholder Sector Smallholders make up the vast majority of Malawi's farmers, and are responsible for most of Malawi's foodcrop production. The major Disk8504/Malagrpol/5-01-85/ae -9- smallholder crops are: maize, groundnuts, fire-cured tobacco, sun-cured tobacco, cotton, and rice. In addition, smallholders grow beans, wheat, sunflower seeds, cassava, coffee, tea, millet, and sorghum. Smallholder agriculture has been growing relatively slowly (e.g. 3 or 4 percent per year), largely due to alienation of customary land for estate agricul- ture. The overwhelming majority of smallholder foodcrop production (over 85 percent) is consumed locally, without ever passing through official markets. Malawi's smallholder population and primary crop production is detailed in Table 2.2. A glance at Table 2.2 conveys a number of important points. First, Malawi's smallholder population and production have uneven geo- graphic distributions. The northern region claims only about one-tenth of smallholder population and crop production. Of the remainder, the southern region has the majority of population and the central region has the clear majority of production. Second, maize is obviously the dominant smallholder crop. Suffice it to say that Malawi's smallholders produce nearly six times as much maize as all other crops combined. Disk85041alagrpol/5-01-85/ae - 10 - Table 2.2. Malawi: Smallholder Population and Cop Production by Region, 1981 North Central South Malawi Population 580,000 2,255,000 2,565,00 5,400,000 Number of Households 118,800 439,800 577,100 1,135,600 Average household size 4.88 5.13 4.45 4.76 Production (MT/yr) Maize: Unfertilized Local 66,000 386,000 283,000 735,000 Fertilized Local 25,000 144,000 106,000 275,000 Fertilized Improved 43,000 171,000 12,000 226,000 All Maize 134,000 701,000 401,000 1,236,000 Groundnuts 4,900 48,200 9,300 62,400 Rice 21,900 7,300, 18,800 48,100 Cotton (sed cotton) -a/ 400 6,600 14,700 21,700 Tobacco a 600 11,500 700 12,800 Other b/ 9,900 12,200 44,100 66,800 al ADMARC crop purchases, assumed equal to production. Includes pulses, millet, and sorghum. Source: 1980/81 National Sample Survey of Agriculture, ADMARC crop purchase tables, Malawi Smallholder Fertilizer Project Staff Appraisal Report, p. 73. Table 2.3 presents a summary of data on smallholder cropping areas. This table illustrates, as does Table 2.2, the strong regional differences in emphasis on production of different crops. Disk8504/Malagrpol/5-01-85/ae - 11 - Table 2.3. Malawi: Smallholder Cropping Areas by Region, 1981 (Hectares) North Central South Malawi Maize: Unfertilized Local 61,000 326,000 299,000 686,000 Fertilized Local 18,000 94,000 86,000 198,000 Fertilized Improved 18,000 61,000 8,300 87,300 All Maize 97,000 481,000 393,000 971,000 Groundnuts 13,300 128,400 24,200 165,000 Rice 8,500 4,500 12,000 25,000 Cotton 200 11,500 27,000 38,700 Tobacco 700 36,800 1,700 39,200 Source: 1980/81 National Sample Survey of Agriculture, Malawi Smallholder Fertilizer Project Staff Appraisal Report, p. 69. For maize, pure stand and mixed stand areas from NSSA data have been added. For other crops, pure stand and12 mixed stand areas have been added, to reflect the lower yields of these crops in mixed stands. The areas are, therefore, approximate pure-stand equivalents. It is clear f-om Table 2.4 that some of Malawi's crops are significantly more land-intensive, labor-intensive, or fertilizer- intensive than others. Thus, production of these crops will respond differently to changes in the availability and cost of these inputs. It is worth noting that improved maize varieties show much higher yields than local strains, but at the cost of much greater fertilizer inputs. Disk8504/Mlagrpol/5-01-85/ae - 12 - Table 2.4. Malawi: Smallholder Ccop Yields, Labor Use, and Fertilizer Use, 1981 Yield Labor Fertilizer (MT/ha) Use Use (Man-days/ha) (Kg/ha) Maize: Unfertilized Local 1,070 85 -- Fertilized Local 1,390 110 150 Fertilized Improved 2,590 120 350 All Maize 1,270 95 60 Groundnuts 380 170 25* Rice 1,900 270 100 Cotton 560 250 25* Tobacco 330 400 250 * Assuming fertilizer is used only occasionally on groundnuts and cotton. Source: 1980/81 National Sample Survey of Agriculture (yields); GOM, "Mzimba Rukuru-Kasitu RDP Proposals", Annex 12, p. 1, 1977 (labor); GOM, for the Agricultural Price and Intervention Policy 1983/84, pp. 79-98, 1983 (labor); Malawi Smallholder Fertilizer Project Staff Appraisal Report, 1983, p. 71 (fertilizer); authors' e3timates. 2.2 Estate Sector There are four major estate crops in Malawi: flue-cured tobac- co, burley tobacco, sugar, and tea. These crops are grown primarily for the export market, although some sugar is sold domestically. In addi- tion, estates grow subsistence crops to feed their workers, although they cannot market these crops in large quantities. Estates sell their export crops by contract or auction, with little government Disk8504/Mlagrpol/5-01-85/ae - 13 - intervention. Estate profits are, however, subject to Malawi' s corporate profits tax. Table 2.5. Malawi: Estate Aggregate Statistics, 1980/81 Tobacco Flue-cured Burley Tea S.gar Area (Ha) 14,892 15,989 16,947 -a/ 14,782 Production (MT/ha/yr) 19,714 18,804 31,965 165,100A/ Yield (MT/he/yr) 1.324 1.176 1.849 a/ 11.170 _ Average Price. (Kw/Mff) 1793 2316 895 545 (export) . Estates 4L0 1,068 26 a/ 2 Z Tenants/lEmployees n/a 19,985 35,362 -7 11,000 Fertilizer purchases fra Optichen (MT/yr) - 22,594 - 8,458 a/ 9,700 a/ Tea figures are for 1979/80 except price and production, which are 1980/81 (Rnthly Stat. Bulletin 3/83). b/ Estimated frcn data in Devres, 1979 "Malai: The Agricultural Estate Sector'. Source: Tobacco Control Comission Ref. no. 78/81 Table, "Tobacco Statis- tics 1980/81" Optichen sales and price data fron RD Kirk, ( Price Waterhouse, The InDustry in Velawi Monthly Statistical Bulletin, GEM, March 1983. 2.3 ADMARC and Smallholder Marketing ADMARC has a legal monopoly on the marketing of all small- holder export crops. ADMARC markets smallholder food crops, but there are also legal, informal markets for these crops. Smallholders must also sell produce to ADMARC to pay off any credit they have received during the growing season, usually in the form of agricultural inputs Disk8504/Malagrpol/5-01-85/ae - 14 - that ADMARC is legally charged with providing. ADMARC's purchasing prices are uniform throughout the country and are set in consultation with the government. Prices are set at the start of the planting season, so farmers know the floor price for each crop before they plant. The only excep- tion is smallholder tobacco, where ADMARC may make a small supplementary payment to farmers as a result of high auction prices. Table 2.6. /XNARC Purchases and Sales (to anallholders, urban, e>port markets) 1981, crop (MT/yr) North Central South All Malawi Maize Bought 35,387 94,543 6,660 136,591 Maize Sold 6,522 13,892 29,459 49,873 Groundnuts Bought 600 18,036 859 19494 Groundnuts Sold 206 2,369 13,869 16,444 Seed Cotton Bought 447 4,132 17,160 21,740 Seed Cotton Sold - 260 21,480 21,740 Tobacco Bought 5582/ 11,46562/ 733-2/ 12,756 Rice Bought 7,088 2337 5,254 14,629 Rice Sold 78 3,054 10,939 14,071 Other Bought 721 527 9,339 10,586 Other Sold 655 428 10,990 12,073 -2/ Distribution data for Tobacco from Dept. of Agriculture Annual Report 1981 draft tables. Source: ADMARC draft data sheets. Disk8504/Nalagrpol/5-01-85/ae - 15 - ADMARC plays a substantial role in interregional distribution of crops, as Table 2.6 illustrates. In the northern and central re- gions, ADMARC procurements generally exceed sales. In the south, ADMARC sales exceed procurements, for at least three reasons: farms are small- er, farm households are larger, and major urban and export markets in the south. Table 2.7. PCMARC Producer Prices and Procurenents, 1978-1983 (Average prices in Kwacha per Vetric Ton, Procurnents in %tric Tons) 1978 1979 1980 1981 1982 1983 Maize price 52 52 66 66 111 122 procurements 120,617 82,171 91,888 136,647 .246,062 244,899 Groundnuts price 200 302 308 307 327 497 procurenents 11,130 24,296 31,418 19,494 10,432 10,218 Rice price 98 99 101 100 109 125 procurenents 30,816 20,487 17,498 14,682 12,543 7,933 Cotton price 213 218 218 214 269 363 procurenents 24,218 22,411 23,114 21,739 14,800 13,368 Tobacco price 475 412 423 481 446 741 procurenents 23,742 19,541 11,341 12,755 8,794 9,279 Source: Malawi: Monthly Statistical Bulletin, ACMARC Annual Report. Table 2.7 shows ADMARC's recent procurement experience. This data demonstrates some general production responsiveness to producer prices, as one would expect. However, the correlation between relative prices and procurement is far from perfect. The correlation improves, Disk8504Mblagrpol/5-01-85/ae Table 2.8. ADMARC Purchases and Mbrketing Expenses, 1981 ceneral Total Cotton Groundnuts Tobacco M3iz Rice Prxdjce 1981 Net sales value of crq purchased 11,700,639 12,258,426 20,000,223 13,579,53 3,788,531 1,870,662 63,198,422 Purchase of crops 4,657,262 6,199,963 6,141,233 9,049,248 1,614,549 1,1%,387 28,858,742 Buying and direct expenses Marketing 617,500 613,749 1,437,091 1,274,971 313,895 274,694 4,531,900 Depot and Storage 243,123 80,985 60,738 972,152 60,738 283,615 1,701,351 Bags, Twine and Hessian 197,834 129,960 110,114 705,893 97,526 71,310 1,312,637 Grading, Ginning and Baling 1,244,476 389,556 1,095,419 - 1,035,563 - 3,765,014 Transport of Crops 773,872 805,974 665,872 4,031,156 631,327 224,731 7,132,932 Famigation 25,551 45,440 10,658 220,848 14,525 39,693 356,715 Insurance 21,833 16,928 17,270 30,769 (29,150) 3,333 60,983 Seed Distribution 97,658 1,585 139,480 11,939 1,329 161 252,149 Total 3,221,847 2,084,177 3,536,642 7,247,725 2,125,753 897,537 19,113,681 Total Purdse and Buying EVpnss 7,897,109 8,284,140 9,677,875 16,297,073 3,740,302 2,093,924 47,972,423 Gross Profit 3,821,530 3,974,286 10,322,348 (2,717,132) 48,229 (223,262) 15,225,999 Net Administrative Expenditure 1,072,814 694,211 1,177,805 2,412,878 708,210 297,155 6,363,073 Net profit on crop trading 2,748,716 3,280,075 9,144,543 (5,130,010) (656,981) (520,417) 8,862,926 Source: AEMARC Annual Report, 1982. - 17 - however, if one allows for some adjustment lag between price changes and production response. One must also allow for the effects of fluctuation in weather and, particularly in the case of tobacco, the conversion of smallholdings to burley estates. These factors must also be taken into account in interpreting the model results in the sections to follow. Table 2.8 describes ADMARC's costs and revenues from crop trading. Note that for riceg groundnuts, and particularly maize, trans- port represents a substanti"tal part of marketing costs. ADMARC's balance sheets value most crops at the official selling prive. However, the export price (for crops which are both sold domestically and for export) may be quite different. Similarly, if maize is stored rather than sold, its value depends on storage costs and future maize needs and prices. These factors must be kept in mind when interpreting ADMARC's cost and revenue data. III. MODEL STRUCTURE AND CALIBRATION The analytical structure developed for Malawian agriculture and the steps taken to calibrate the model are described in this sec- tion. The description presented here is supplemented by a complete algebraic statement of the model in Annex A. The institutional setting for the model is taken to be that prevailing in the years 1980 to 1981, this being the period used to calibrate the model. In Section I', however, the consequences of alternative institutional arrangement ara explored. Disk8504/Malagrpol/5-01-85/ae - 18 - 3.1 Smaliholder Agiculture Smallholder agriculture is the dominant mode of production in Malawi (see Section II). Although the underlying behavior -- profit maximization -- is assumed to be the same for all smallholders, differ- ences in local conditions imply substantial differences in cropping patterns. The model, therefore, identifies eight smallholder types differentiated by region and cropping pattern: Region Cropping Pattern Farm Size (Ha) E of Farms North Maize, groundnuts 1.248 66,700 North Maize, rice 1.248 52,100 Central Maize, groundnuts 1.542 219,600 Central Maize, groundnuts, tobacco 1.542 191,900 Central Maize,cotton 1.542 28,300 South Maize, groundnuts 0.876 211,600 South Maize, rice 0.876 128,600 South Maize, cotton 0.876 236,900 Total 1,135,700 Source: National Sarple Survey of Agriculture 1980/81, Set A2, Table 3. Although data are not available for this disaggregation of smallholders, the aggregates arc consistent with observed data. More- over, this disaggregation is important from a policy perspective for at lecc two reasons. First, average farm size differs significantly among regions. And second, changes in the producer price of, say cotton are not likely to affect the incomes of smallholdera in the north. Neither Disk85O4/Mlagrpol/5-01-85/ae - 19 - of these points would be captured if the model structure comprised only a single, representative smallholder. Note, however, that the disaggre- gation used here is not totally innocuous. Thus, while smallholders can switch land from maize production to rice production or from maize production to cotton production, it is not possible to switch land from rice production to cotton production because this choice does not lie within the feasible set for any of the smallholder types. This may not be a serious error, however, since the important substitution possibil- ity is undoubtedly that between maize, on the one hand, and all other crops, on the other hand, and this possibility is fully captured by the model. At the time of planting, smallholders are assumed to allocate land among crops in order to maximize profits. Their ability to do so is constrained by the availability of land, the technological possibili- ties for substitution, and the prices of outputs and inputs. Smallhold- ers are assumed to be price-takers in both output and input markets. In the particular case of Malawi, most prices are in fact set by ADMARC and, with some exceptions, farmers apparently trade at these prices. Given prices, technology, and the availability of land, the model simu- lates the behavior of a profit-maximizing farmer. This should not be interpreted too strictly. The presumption is that farmers on average tend towards the profit-maximizing solution. In addition, the model contains two features which may lead to departures from the profit-maximizing outcome. First, the evidence of field visits suggests strongly that farmers grow local maize primarily Disk8504/alagrpol/5-01-85/ae - 20 - for own-consumption and grow improved maize mainly for sale. In other words, the profit-maximizing motive only seems to apply to the produc- tion of improved maize. This may be for one or two reasons. Improved maize may be more risky than local maize in which case a risk-averse smallholder would prefer to guarantee his subsistence requirements before turning to production of more profitable crops for the market. Alternatively, smallholders may prefer to consume local maize and are aware that ADMARC sells mainly improved maize. To avoid being affected by possible ADMARC supply shortages, farmers may choose to ensure their own supply of their preferred variety even if improved maize has higher returns. Either one or both of these reasons may be compounded by a technical factor. Local maize can be stored and pounded successfully on the farm. This is not so for improved maize. The cost to the farmer of consuming one unit of maize may, therefore, be lower for local maize -- that can be processed on the farm -- than for improved maize -- that has to be purchased from ADMARC. For these reasons, the model contains a subsistence constraint -- sufficient land has to be allocated to the production of local maize to satisfy the household's consumption needs before attention is turned to profit maximization. Second, the evidence of field visits in 1983 suggests that farmers, especially in the northern and central regions, are anxious to use more fertilizer but are unable to do so because national supplies are limited. In other words, farmers are not able to attain the profit- maximizing solution because of insufficient fertilizer at the (subsidiz- ed) ADMARC price. This is captured in the model by introducing a Disk3504/alagrpol/5-01-85/ae - 21 - constraint on the availability of fertilizer to the farmer which is, in turn, related to the total availability of imported fertilizer at the national level. Finally, the assumption that farmers trade at ADMARC prices is relaxed in two instances. These arise with respect to fertilizer use and tobacco sales. There is evidence to suggest that smallholders are "leaking" ADMARC-purchased fertilizer to the estates and are smuggling tobacco abroad. These illegal activities are a direct response to the large differential between the ADMARC selling price for fertilizer and the estate buying price and between the ADMARC buying price for tobacco and the price ruling in Zambia and Zimbabwe. Accordingly, leakage functions are introduced to capture this behavior. This specification of modified profit-maximizing behavior yields the farmer's desired combination of outputs and inputs at the time of planting. Realized outputs and inputs, however, may be diffe- rent because of the intervention of unusual weather conditions. Obvi- ously, no attempt is made to model the behavior of weather. Neverthe- less, exogenous changes in weather conditions can affect realized out- puts by means of a shift factor in the production technology of each crop. The consequences of unusually favorable or unfavorable weather conditions can, therefore, be analyzed within the model. Once weather conditions have been allowed for, the realized inputs and outputs can be aggregated for all smallholders to arrive at national, smallholder demand for inputs (such as fertilizer) and nation- al smallholder supply of outputs (such as maize). In addition, given Disk85O4Alagrpol/5-O1-85/ae - 22 - f arm prices, the reflized inputs and outputs can be expressed as a realized profit or loss to the smallholder's farming operations. For some crops -- tobacco -- production and marketed surplus may be the same; for others -- maize, for example -- it is very important to allow for own-consumption to obtain marketed surplus. Accordingly, we des- cribe next the behavior of smallholders as consumers. Given realized farm profit and any other income, and given consumer prices, smallholders are assumed to allocate their expenditures among commodities such that they maximize their satisfaction. Behavior of this kind results in demand curves which depend on total income (in- cluding farm profits) and consumer prices. This formulation, therefore, allows explicitly for the effect on changes in farm profits on demand. It also allows the derivation of the household's pattern of demand. Aggregating these demands over all smallholders yields smallholder con- sumption at the national level. For those crops that are both produced and consumed, the production and consumption figures can be used to arrive at total marketed surplus from the smallholder sector. 3.2 Urban Consumetion Consumption behavior of urban households is modeled in a similar manner to that of rural households. For the present analysis, urban households are disaggregated by region: Region Household Income (K) E of Households North 1,124 3,660 Central 1,487 41,480 South 1,924 76,320 Source: Urban Household Expenditure Survey, 1979/80. - 23 - Given household income and prices, the model determines the household's pattern of consumption. Unlike rural households, income for urban households does not change as a result of changes in producer prices. The sole source of impact on urban consumption, therefore, is through changes in consumer prices or the availability of consumer prices or goods. Subtracting urban consumption from rural marketed surplus, one arrives at a measure of national exports (or imports) or of additions to (or reduction in) national stocks. 3.3 Estate Agriculture The short-run production response of the estate sector is also modeled. Four crops are included: Crop Area (Ha) Production (MT) Tobacco (flue-cur d)-/ 13,921 18,800 Tobaco (burleya7 23,309 19,700 Tea-/ 17,631 31,500 Sugarc/ 9,211 165,100 Tobacco Control Commission,(data are for 1981/82). The Tea Association, Blantyre (1981/82 estimate). LONRHO Sugar Corporation, Sucoma (1982/83 estimate). Since the estate sector ranges from large corporations such as LONRHO's SUCOMA sugar estates to relatively small tobacco farms, it is difficult to conceive of a representative estate. We have chosen, Disk8504/Malagrpol/5-01-85/ae - 24 - therefore, to model output of each crop for the estate sector as a whole. That is we have worked directly with market-level, supply func- tions rather than with estate-specific production functions. Thi approach allows us to explore the consequences of changes in output and input prices on production and exports. In addition, the model computes the effect of these changes on total profits for the estate sector and government revenue from profits tax. With respect to the consumption behavior of estates, it is assumed that consumption of items such as maize is supplied by own production. That is, the estates are assumed to be neither net demanders nor net suppliers of food crops. This as- sumption allows us to ignore the consumption and production of food crops within the estate sector. As noted in Section II, most of the government's efforts to stimulate agriculture through pricing policy are directed towards the smallholder sector. By and large, therefore, the estates face market- determined prices for their outputs and main inputs. There are, how- ever, two important interactions between the estates and smallholders. One -- the illegal sale of fertilizer to the estates -- has already been mentioned. The other arises from the government-enforced constraints on smallholder production of tobacco. Smallholders are subject to produc- tion quotas for tobacco and receive prices from ADMARC which are sub- stantially lower than f.o.b. prices (Section II). It follows that land under tobacco, yields a higher financial return if it is within the estate sector than if it is within the smallholder sector. Moreover, it appears relatively easy to change tenurial status and tobacco land has, Disk8504/Malagrpol/5-01-85/ae - 25 - therefore, moved from the smallholder sector to the estate sector. Accordingly, we incorporate a land transfer function in the model which allows for the intersectoral movement of tobacco land depending on the relative profitability of smallholder and estate production. 3.4 Market Structure The various producer and consumer units of the model combine to determine total domestic supply and demand. Markets clear in a variety of ways, the most important distinction being between ADMARC's control of most prices facing smallholders and the exposure of the estate to prices determined in international markets: Market Structure M3rket Clearing Cp Producer Producier Price Consaner Price chanian Maize Smallholder MARC AMARC International trade, domestic stocks Groundnuts Smallholder AEMARC ADARC Exports Rice Smallholder ACMARC ACMARC Exports Cotton Smallholder AD4ARC ? Exports Tobacco Exports (flue-cured) Snallholder ADARC - (smuggling) Tea Estates World market - Exports Sugar Estates World market - Exports Tobacco (flue-cured) Estates WDrld market - Exports Tcbacco (burley) Estates Wrld market - Exports Mniwting Clearing DMLsUer _ _ __tePrc Fertilzer Smallholders AEMARC Imports or quotos leakage) Fertilizer Estates WDrld price Imports Labor Sallholders Free market Elastic supply Land Sallholders Free market Fix supply - 26 - Within the smallholder sector, market-clearing for maize dif- fers from that for the other crops. Given ADMARC prices, the production surpluses of groundnuts, rice, cotton and tobacco are exported. (Note that the possibility of smuggled tobacco is modeled explicitly.) Maize, however, is essentially a non-tradeable, that is, given the large dif- ference between estimates of the import-parity price (K 367/MT in 1984) and the export-parity price (K 72/MT in 1984), average maize should be neither imported (domestic production is cheaper) nor exported (domestic production is too costly). Variations in weather have resulted in ex- ceptionally large or small crops and in turn have led to exports or emergency imports. In the model, it is assumed that policymakers will handle shortfalls by calling on reserves to the extent possible and only then resorting to imports. Similarly, surplus production is assumed to result in stock accumulation, although the consequences of exporting can, of course, be explored in the model. The estates, however, operate differently. They face prices determined in the world market and most of their production is exported. They also import fertilizer through OPTICHEM and pay prices determined in the world market. Inputs for the smallholders comprise fertilizer, labor and land. Fertilizer is provided by ADMARC at a fixed price. Unless ADMARC predicts demand for fertilizer accurately, smallholders may suffer from insufficient supplies or ADMARC may be left with unused stocks. Accord- ingly, the model incorporates the possibility of quantity rationing depending on the amount of ferilizer imported by ADMARC. Labor is Disk85O4/Malagrpol/5-01-85/ae - 27 - assumed to be in perfectly elastic supply so that the wage rate remains at its 1981 level. Land, on the other hand, is considered to be in fixed supply. The model allows explicitly, therefore, for the realloca- tion of land among crops according to the dictates of profit maximiza- tion. Policy Results Policy results are obtained at three levels. At the household level, the model, reports the effect of policy changes on production, input use, farm profits, consLimption and real income. At the market level, the effects of policy change on fertilizer usage, total produc- tion, rural consumption, marketed surplus and urban consumption are documented for each crop. And finally, at the government level, the consequences for net revenue -- taxation less subsidies -- and for the balance of payments -- export earnings less import bill -- are assessed. The model thus provides information on most of the objectives of con- cern. For example, the analyst can compute a self-sufficiency index for maize by calculating the proportion of urban demand that can be met from domestic procurement. Another policy result of interest is the effect on income distribution -- both regional and urban-rural. Thus, the model provides information, not only on the immediate agricultural magnitudes such as consumption and production, but also on more policy- relevant matters such as equity, government revenue and the balance of payments. Disk8504/Malagrpol/5-1-85/ae - 28 - 3.5 Calibration To be a reliable guide for policy formulation, it is essential that the model be well based in fact. Unfortunately, data in Malawi, as in most African countries, are scarce. To maximize the value of avail- able information, a calibration technique was used. Given the structure of the model (the production and consumption functions at the household level and the market clearing mechanisms at the aggregate level), the model was calibrated to data for the 1980/81 crop year. This entails minimizing deviations from known or estimated values for parameters of the model -- demand elasticities, for example -- and for outputs of the model -- total production of maize, for example. In the minimization process, parameters for which values are not known are determined such that model departures from known parameters and values are as small as possible and all values are consistent with the basic structure of the model. Given the profit-maximizing behavior of producers, the demand patterns of consumers, the regional disaggregation of the model and the assumed market-clearing mechanisms, the values of production derived from the model are compared in Table 3.1 with those from the National Sample Survey of Agriculture for crop year 1980/81. Because departures from known values are penalised very heavily in the calibration tech- nique, the calibrated values for production are very close -- within + 8 percent - of the actual values. Disk8504hlagrpol/5-01-85/ae - 29 - Table 3.1. Calibrated & Actual Production for Smallholder Production; Crop Year 1980/81 (MT) North Central South All Malawi Local Maize Calibrated 95,173 538,822 391,571 1,025,566 Actual 91,000 530,000 389,400 1,010,400 Improved Maize Calibrated 43,447 166,461 12,671 22,579 Actual 43,100 171,500 11,800 226,400 Groundnuts Calibrated 4,979 49,176 9,042 63,198 Actual 4,9000 48,200 9,300 62,400 Rice Calibrated 23,919 21,497 45,417 Actual 26,200 21,900 48,100 Cotton Calibrated 6,865 15,053 21,918 Actual 6,700 15,000 21,700 Tobacco Calibrated 13,000 13,000 Actual 12,800 12,800 Source: National Sample Survey of Agriculture, 1980-81, and Model Results. The calibration simultaneously determines inputs. Table 3.2 shows the yields and input intensities produced by the model. Although "hard" data are not available for purposes of comparison, the estimated values agree with the general data that are available. As expected, rice and tobacco are the most labor-intensive crops, while improved maize and tobacco are the most fertilizer-intensive. Disk8504lagrpol/5-01-85/ae - 30 - Table 3.2. Yield and Input Intensities, 1980/81 Yield Labor Fertilizer (kg/ha) (Man-days/ha) (kg/ha) Unfertilized Local Maize 1,111 88 0 Fertilized Local Maize 1,522 116 79 Improved Maize 2,575 130 232 All Maize 1,318 97 36 Groundnuts 359 254 103 Rice 1,722 233 103 Seed Cotton 458 86 32 Tobacco 403 225 147 More importantly, the calibration determines the values of various parameters which in turn determine the responsiveness of output and input usage to prices. These elasticities are reported in Table 3.3. The price of maize is by far the most important price in Malawian agriculture. According to the model, a 10 percent increase in the price of maize increases maize output by less than 4 percent. This is consis- tent with the view that farmers in Malawi respond positively but not dramatically to changes in the price of maize. The recent record sur- pluses in ADMARC storage bears witness to the responsiveness of produc- ers. Unfortunately, firm empirical estimates are not available. This is a critical gap in information since a small percentage change in the production of maize can imply a large percentage change in the produc- tion of other crops as land. is reallocated among groups (Table 3.3). The particular value of the supply elasticity used in this report is subjected to further examination below. Disk8504/0alagrpol/5-01-85/ae - 31 - Table 3.3. Supply Response Price Changes In Maize Gourdruts Rice Cotton Tobacco Fertilizer S.pply ResMnse of Meize Unfertilized Local 0.31 -0.06 -0.01 -0.01 -0.01 - Fertilized Local 0.44 -0.07 -0.02 -0.02 -0.03 -0.12 Improved 0.42 -0.07 -0.01 -0.01 -0.05 -0.19 Maize 0.36 -0.06 -0.01 -0.01 -0.02 -0.05 Groundnuts -0.25 0.69 - - 0.02 -0.01 Rice -0.22 - - - - -0.01 Cotton -0.22 - - 0.38 - -0.01 Tobacco -0.26 -0.05 - - 0.63 - Empirical estimates of response elasticities are available for most of the other crops. Supply response elasticities are estimated to be 0.48 for tobacco by Dean (1966), 2.30 for marketed surplus of ground- nuts by Gordon (1971) and 2.36 for cotton by Colman and Garlett (1975). Estimates for other African countries are reported in Table 3.4. The elasticities used in the model (Table 3.3) are within the range observed in other countries. For cotton, the value used in the model (0.38) is below that of most other estimates, while, for ground- nuts (0.69) and tobacco (0.65), the model values are consistent with the limited number of available estimates. Disk8504/Malagrpol/5-01-85/ae - 32 - Table 3.4. Supply Response in Africa Long-Run Dependent Variable Crop Country Elasticity Elasticity Output Cotton Nigeria 0.67 0.67 Output Cotton Sudan 0.39 0.50 Output Cotton Uganda 0.25 0.25 Gitput Cotton Uganda Bganda 0.67-0.73 0.67-0.73 Output Groundnuts Nigeria 0.24-0.79 0.24-0.79 Acreage Cotton Nigeria 0.23-0.38 0.28 Acreage Cotton Uganda Buganda 0.50 0.63 Acreage Cotton Uganda (East) 0.23 0.44 Acreage Cotton Uganda (Vst) 0.26 0.62 Acreage Cotton Uganda (North) 0.02 0.07 Acreage Tobacco Nigeria 0.60 0.82 Source: The consumption side of the model is calibrated in a similar fashion. In this case, however, estimates of rural consumption by region are derived by subtracting net ADMARC procurement from regional production. Model estimates (shown in Table 3.5) calibrate exactly with consumption data derived in this fashion. The consumer response elasticities obtained from the model are shown in Table 3.6. Own-price elasticities lie in the range 0.2 to 0.6. That is, consumers respond negatively, but not dramatically, to price changes. Cross-price elasticities, on the other hand, are gener- ally low. The income elasticity for maize, a subsistence crop, is relatively high in rural areas (0.53) but it falls rapidly as incomes increase and is negative in urban areas. The income elasticities for the other food crops are less than 0.5. A slightly higher income Disk8504/Malagrpol/5-01-85/ae - 33- Table 3.5. Consumption Calibration Data Rural and Urban Consumption 1981 (1000 MT) Crop Rural Urban Maize 1,140 60.4 Groundnuts 41.9 .56 Rice 30.1 5.7 Cloth (seedcloth equivalent) 9.1 11.3 Sugar 19.3 22.5 Table 3.6. ConsuTer Response Price Change in Change in Maize Groundruts Rice Cloth Sugar Income Demiand FsEMnse in: Maize - Rural -0.21 - - 0.02 -0.01 0.53 - Urban -0.22 0.01 0.02 0.08 0.02 -0.25 Groxunuts - Rural 0.10 -0.29 - -0.04 - 0.19 - Urban 1.10 -0.30 - 0.04 - 0.23 Rice - Rural 0.06 0.01 -0.44 0.04 - 0.20 - Urban 0.05 - -0.61 0.02 - 0.49 Cotton - Rural - - - -0.49 -0.01 0.72 - Urban - - - 0.49 -0.01 0.71 Sugar - Rural - - - - -0.30 0.32 - Urban - - - -0.30 0.34 Disk8504/Malagrpol/5-01-85/ae -34- elasticity -- 0.7 -- is used for the single non-food item in the study -- cloth, a proxy for cotton. While the expenditure elasticities for urban households were estimated from the Urban Household Expenditure Survey of 1979/80, the other response elasticities are based on estimates for other countries. Because the estate sector is not modeled at the microeconomic level, it was not necessary to perform a full calibration. The produc- tion levels for 1981 and the response elasticities used in the model are shown in Table 3.7. Table 3.7. Predicting AEMARC Procurenent (k/mt) ACMARC FRODUCER FRICES 1980/81 1981/82 1982/83 Maize 66 mi LU Gcouirduts 307 327 497 Rice 100 109 125 Cotton 214 269 363 Tobacco 481 446 741 Fertilizer 144 164 232 Marketed Surplus/A[MARC Procurenent (metric tons) 1980/81 1981/82 1982/83 ObservecPredicted Predicted Observed Predicted Observed Maize 137,000 336,000 246,000 248,000 245,000 Groundruts 19,500 9,600 10,400 31,900 10,200 Rice 14,700 8,800 12,500 13,100 7,900 Cotton 21,700 18,400 14,800 20,800 13,400 Tobacco 12,800 9,300 8,800 13,400 9,300 Disk8504/Malagrpol/5-01-85/ae -35- The supply elasticities show in Table 3.8 were estimated for this report from historical data. To test further the response elasticities embedded in the model, "predictions" for crop years 1981/82 and 1982/83 were compared with actual values. Since the only information available for these years is for ADMARC procurement, the test involves an assessment of the accuracy of both the production and comsumption sides of the model as revealed by predictions of marketed surplus. There are several reasons why this is an especially difficult test. First, marketed surplus is often a small proportion of both production or rural consumption. It follows that relatively small percentage errors in predicting either of these totals can lead to large percentage errors in their difference. Second, the model assumes that producers and consumers adjust instantan- eously to price changes, whereas available empirical evidence suggests a lagged response. (Table 3.4 shows the difference between short-run and long-run responses for some crops.) This is especially important in the present context because major changes in prices occurred between crop years 1980/81 and 1981/82 (Table 3.7). Third, variables exogenous to the model may have changed between years. Of these, the most important is weather. Thus, while 1980/81 and 1981/82 enjoyed relatively favor- able weather conditions, 1982/83 suffered a drought in the southern region. These changes in weather have not been incorporated in the predictions presented below, although in principle changes in weather can be accomodated within the model. Disk8504/Malagrpol/5-01-85/ae - 36 - The comparison of predicted and actual values for ADMARC procurement are shown in Table 3.7. Following the increase in the price of maize by two-thirds in 1981/82, the model predicts a large increase in marketed surplus. Table 5.8. Production Data for the Estate ector Flue-Cured Birley Crop Sugar Tea Tobaco Tobacco Production (1000 MT) 165 31.5 18.8 19.7 Prcfits (1000 K) 53,000 1,060 5,310 12,000 Fertiliser Use (1000 MT) 13.1 11.4 18.4 13.6 Nerlovian supply elasticities: Short-run .9 .2 .7 .2 Long-run 1.8 1 3.5 2 Source: Tobacco Control Comssion; The Tea Association, Blantyre; LOWHO Sugar Corporation, Sucoma. The predicted surplus exceeds the observed surplus by almost 40 per- cent. Nevertheless, if this difference is attributed equally to diffe- rences between observed and actual production and between observed and actual consumption, the overprediction of production is only 6 percent and the underprediction of consumption is only 8 percent. Since the model assumes instantaneous and complete adjustment, the difference between predicted and observed may reflect the lag in producer and consumer response. The errors for the other crops are -8 percent for Disk8504/Mlagrpol/5-01-85/ae - 37 - groundnuts, -30 percent for rice, 24 percent for cotton and 6 percent for tobacco. Between 1981/82 and 1982/83 the prices of all crops increased relative to that of maize. As a result, the model predicts increases in the marketed surplus of groundnuts, rice, cotton, tobacco, and a decline in procurement of maize. In this year, maize procurement is predicted almost exactly. The marketed surplus of all other crops, however, is overpredicted by between 4 percent (tobacco) and 213 percent (ground- nuts). This may reflect unfavorable weather conditions especially in the south. Despite these errors on the prediction of marketed surplus, for the reasons cited, the model is judged to represent reasonably accu- rately the underlying structure of agriculture in Malawi. IV. POLICY ANALYSIS: ALTERNATIVE SCENARIOS AND THEIR IMPLICATIONS 4.1 Government Objectives Smallholder Incomes. The impact of various policy changes on household incomes in the smallholder sector is a major concern. The model provides data on household incomes for eight smallholder farm types classified by cropping pattern and region. In this report, how- ever, we only record changes in smallholder incomes averaged across all cripping patterns and all regions. Income is calculated on a net basis by subtracting the cost of any purchased inputs. Changes in smallholder incomes result from changes in the prices and production of the main Disk8504/alagrpol/5-Ol-85/ae - 38 - smallholder crops -- maize, groundnuts, rice cotton and tobacco, and from changes in the price and use of fertilizer. ADMARC Cash Flows. As ADMARC is the main channel for market intermediation for all smallholder crops, any changes in smallholder production and marketing affect ADMARC's operations and its cash flows. These cash flows are calculated to include all ADMARC operating revenues and expenses associated with the purchase as well as sale of smallholder crops -- maize, groundnuts, rice, cotton and tobacco -- and fertili- zers. On the cost side, we have included the prices paid to producers, the costs of transportation of crops to depots, the costs of grading, bagging and other marketing and distributive functions as well as admin- istrative and other overhead costs. On the revenue side, we nave in- cluded the prices at which crops are sold to consumers in urban areas, sold at auction or exported and the prices charged smallholders for fer- tilizers. Note that we do not include the investment value of surpluses that have been stored, or the costs of building and operating storage facilities for them. Food Self-Sufficiency. The effectiveness of policy changes in maintaining food self-sufficiency is measured by the amount of maize added to inventory. This amount is the net result of ADMARC maize pur- chases, sales, exports, and imports. Thus, if the surplus production is insufficient to cover exogenously-determined maize exports, maize will be shown as being reduced even though Malawi itself may not be in defi- cit. Note, however, that only changes in maize stocks are reported. The levels of stocks are not calculated and the model assumes that there Disk8504/Malagrpol/5-01-85/ae - 39 - will be sufficient stocks to cover any deficit and sufficient storage capacity to absorb any surplus. Persistent large additions to maize stocks may not constitute "success". Additions to stocks saddle ADMARC with purchase and storage costs, but do not generate current revenue. Balance of Payments. The impact of policy changes on the bal- ance of payments is measured via the impact on exports and imports. Any production of groundnuts, rice and cotton in excess of domestic demand at prevailing prices is assumed to be exported. Smallholder tobacco production is also assumed to be exported either directly through ADMARC or indirectly through auction markets. Similarly, all estate production of teas, flue-cured tobacco and burley tobacco is assumed to be exported as is any production of sugar in excess of domestic demand. The unit value of these exports is the relevant f.o.b. price actually received by ADMARC or the auction floor price. Any procurement of maize by ADMARC in excess of domestic sales, however, is assumed initially to be added to inventory. The decision to export any of the maize is treated explicitly as a policy decision. Turning to imports, all fertilizers, whether marketed via ADMARC for sale to smallholder or via OPTICHEM for sale to estates are assumed to be imported at the relevant c.i.f. prices. Maize can also be imported but, as with exports, this is treated as an explicit policy decision to be made after stocks have been fully drawn down. The net foreign exchange revenue from agriculture is then the sum of all these items expressed in Kwacha after conversion at the official exchange rate. All f.o.b. and c.i.f. prices are treated as exogenously determined with Malawi acting as a price taker. Disk8504/%lagrpol/5-01-85/ae - 40 - Government Revenues. The measurement of the impact on govern- ment revenue from the agricultural sector is confined to the sum of ADMARC's net cash flow from smallholder marketing activities (as des- cribed above) and profit taxes on estate production of sugar, tea and tobacco. This would also include any revenue from export or other taxes on smallholder crops that might be introduced. This does not capture the multiplicity of other government activities that take place in the agricultural sector, or other indirect consequences on government reven- ue that result from policy changes. The measured impact is therefore limited, although it does cover the main direct implications of altern- ative policy scenarios on the government budgets. 4.2 Alternative Policy Scenarios The Malawi pricing model can be used to explore a large number of policy scenarios. After discussion with our Bank colleagues and in light of the earlier discussions with the government we have initially focused our attention on three broad set of policy issues. These issues and their related policy scenarios, are outlined below. Elimination of Fertilizer Subsidies. Although the impact of a variety of input price changes can be examined, the most important is the removal of fertilizer subsidies for smallholders. ADMARC provides fertilizers to smallholders at less than "full cost", while they are sold to estates by a private firm (OPTICHEM) at a higher cost. This leads to "leakages" of fertilizers from the smallholder sector to the Disk8504/alagrpol/5-01-85/ae - 41 - estate sector. The Bank has been recommending as part of the SAL nego- tiations that these subsidies be phased out and fertilizers be priced at "full cost" to cover ADMARC costs, so as to allow efficient allocation of this scarce input. The Government of Malawi (GOM) has agreed to phase out fertilizer subsidies over a period of time, and fertilizer prices to smallholders have been increased from a low of K 144/MT in 1980-81 to K 164/MT in 1980-81 and K 232/MT in 1982/83. In the 1983/84 crop year prices were again raised to K 282/MT I/ (Table d.1). GOM has been concerned with the impact this may be having on smallholder produc- tion and incomes. There is particular sensitivity about the impact on smallholder production of maize which is a relatively heavy user of fertilizers. Although GOM feels that it has gone a considerable way in eli- minating subsidies on this critical input, the Bank's concern is that ADMARC is still not covering its "full costs" on smallholder fertilizer distribution. The differences in perception arise from the definition of "full cost", and this in turn depends on the source of imports. Two years ago GOM had already bought some 52,000 MT of fertilizers but these had been stranded in Beira in Mozambique due to transportation prob- lems. As these problems persisted, a decision had to be made to import these via the more costly routes through Zambia or the Republic of South Africa and the port of Durban. Any additional supplies that become 1/ This is the weighted average price of SA and 20:20:0 with the weights being the amounts sold in 1981/82 -- that is 80 percent SA and 20 percent 20:20:0. Disk8504/Malagrpol/5-01-85/ae - 42 - available (some 40,000 MT were available via an IFAD loan) has to be imported via Durban. It is highly probable that future supplies will also have to come via Durban. Although the increase of fertilizer prices from K 144/MT in 1980/81 to K 282/MT in 1983/84 (an increase of Table 4.1: Malawi: Smallholder Producer Prices Paid by ADMARC 1981-84 (Kwacha/MT) 1983/84 Estimated Crop 1980/81 1981/82 1982/83 1984/85 1984/85 Parity Prices Maize 66.3 ill ill 122 (67.4)b/ ( 0.0) (9.9) Groundnuts 307 327 497 547 658 (6.5) (52.0) (10.1) (20.3) Rice 100 109 125 256 251 (9.0) (14.7) (24.8) ( 60.9) Cotton 214 209 363 400 542 (25.7) (34.9) (10.2) (35.5) Tobacco 481 446 741 814 1992 (-7.3) (66.1) (9.9) (144.7) Fertilizeraa/ 144 164 232 282 421 (13.9) (41.5) (21.6) (49.3) ,/ These are the prices charged by ADMARC to smallholders. b/ Figures in brackets give the percentage increase over previous year and column. Source: ADMARC. Disk8504/Nalagrpol/5-01-85/ae -43 - almost 100 percent) was adequate to remove the "subsidy" on any ferti- lizers shipped via Beira and bought at lower prices, they do not pre- sently cover the costs of importing fertilizers via Durban. In order to cover these high import parity prices as well as ADMARC's distribution costs, we estimate that smallholder's need to be charged a cost closer to K 420/MT for fertilizers in order to eliminate any losses on ADMARC's fertilizer marketing account. We can verify whether this import parity price is an appropriate "full cost" price by comparing it with the prices prevailing in the free maket for the estate sector. The prices charged by OPTICHEM to private estates work out at K 426/MT in 1984. Thus the first scenario can be described as follows: FI: Elimination of Fertilizer Subsidy This requires the removal of subsidies on small- holder fertilizers by charging a price close to the ex-Durban import parity price plus any distribution and other costs to eliminate any losses in ADMARC's fertilizer account. This requires an increase in average fertilizer prices from the present K 282/MT to K 421/MT -- an increase of almost 50 percent. Elimination of Maize Subsidy. Apart from the losses incurred by ADMARC on its fertilizer account, it also incurs large losses on its maize marketing account. ADMARC does not cover its "full cost" (pur- chase price plus distribution margins) on its maize account. The Bank Disk8504/alagrpol/5-01-85/ae -44- has been encouraging GOM to remove this subsidy. Although subsidized maize prices are available to both rural and urban consumers, GOM has quite correctly been particularly sensitive to the impact of increased maize prices on low income urban consumers. There was a substantial increase in maize consumer prices from K 100/MT in 1980/81 to K 147/MT in 1981/82. These remained unchanged essentially for three years and were recently raised again in 1984/85 to K 159/MT (Table 4.2). But these still do not cover ADMARC's full costs in 1984. There are three ways of eliminating maize subsidies: raise consumer prices by the amount of the unit loss in ADMARC's maize account in 1984, lower producer prices by the same amount or some combination of the two. We evaluate the two extremes because they have dramatically different implications. The first would affect only consumers, mainly in urban areas (as rural producers are able to grow their own consump- tion requirements), while the latter would affect maize producers in the smallholder sector. We also want, in the latter case, to look at the joint effect of both reducing maize producer prices to eliminate losses on ADMARC's maize account as well as increasing fertilizer prices to eliminate any losses on the fertilizer account. Disk8504/balagrpol/5-01-85/ae -45- Table 4.2. Malawi: Average Consumer Prices for Agricultural Commodities (1980-81 to 1984-85) ((wacha/MT) Commodity 1980/81 1981/82 1982/83 1983/84 19184/85 (Base Year)-/ Maize 100 147 147 147 159 (47.0) (0.0) (0.0) (6.8) Groundnuts 629 649 819 869 869 (3.2) (26.2) (6.1) (0.0) Rice 259 268 284 315 315 (3.5) (6.0) (10.9) (0.0) Cloth 1700 1900 2100 2400 2400 (11.8) (10.5) (14.3) (0.0) Sugar 360 400 400 400 400 (11.1) (11:1) (0.0) (0.0) a/ Assumed to be the same as 1983/84 unless otherwise announced. Note: Figures in brackets give the percentage change over the previous year. Source: ADMARC. (We assume that ADMARC domestic sales prices have been moving parallel to producer prices as in the past). Therefore we define three related policy scenarios as follows: M.: Eliminate Maize Subsidies by Raising Consumer Prices This requires raising ADMARC's selling price; for maize by the unit amount of losses in ADMARC's maize marketing account in the base year. (This account excludes the value as well as the storage Disk85O4/Malagrpol/5-01-85/ae - 46 - costs of the maize stored in ADMARC.) This would increase the prices charged by ADMARC to consumers from K 159/MT of K 189/MT -- an increase of almost 20 percent. M2: Eliminate Maize Subsidies by Lowering Producer Prices This requires lowering ADMARC's purchasing price for maize by the unit amount of losses in ADMARC's maize trading account in the base year (again excluding the value and storage costs of the stored maize). This would reduce the producer price from K 122/MT or K -92/MT - a 25 percent decline. M3: Eliminate Maize Subsidies by Lowering Producer Prices and Eliminate Fertilizer Subsidies (M2 + Fl) This is a combination of the M2 and F1 scenarios and requires raising fertilizer prices fromK 282/MT to K 421/MT as well as reducing producer prices for maize from K 122/MT to K 92/MT. Export Parity Pricing. The extreme sensitivity on food secur- ity (due to real shortages felt in 1980) has led GOM to set producer prices of maize high relative to the prices of export crops such as cotton, groundnuts and tobacco. This assured self sufficiency but has Disk8504/0alagrpol/5-O1-85/ae - 47 - had a negative effect on production of export crops and earnings. Between 1980/81 and 1981/82, GOM further increased maize prices by a dramatic 66 percent. This further depressed the production of export crops and earnings. The Bank has been concerned with the movement in the relative domestic price trends against export crops and with the consequent impact on export earnings, especially during years of in- creasing balance of payment difficulties. It has suggested that these relative price trends be reversed in order to encourage higher produc- tion of export crops and higher export earnings to ease the balance of payments situation. The domestic producer price ratios can be changed by either reducing the price of maize or increasing the prices for export crops. The former may be preferred if domestic prices of export crops are already close to parity and increasing them further would 'lock' the country into further distortions. The Bank has been advising GOM to r-iduce maize prices relative to export crops, but the government has been concerned that production and procurements, especially from the smallholder sector will respond adversely to jeopardize "food secur- ity". There is also concern that reduced prices offered by ADMARC will reduce its procurements and, given fixed overhead costs, affect its budgetary position. The alternative is to increase the smallholder producer prices of export crops relative to maize prices. This has been the preferred alternative for the GOM. In 1982/83, the prices of all export crops were increased substantially while the price of maize was held Disk8504/Malagrpol/5-01-85/ae - 48 - constant. In the following year, although the maize price was increased by 10 percent, the relative price ratios were maintained except for rice whose price was increased relative to other crops (Table 4.1). These decisions have been in the right direction but in the current crop season, if producer prices are not changed for export crops, they will remain substantially below their parity equivalents. We need to evalu- ate the implications of setting export crop prices at their parity levels. Since one of the purposes of setting domestic prices at their parity levels is to set them at the true 'opportunity cost' for all traded goods to encourage efficient resource allocation, it would be equally useful to also set traded input prices - in this case fertiliz- er prices -- at their import parity. In this case both traded outputs and inputs would be set at their parity prices. However, if this is done in order to meet the efficiency and proper resource allocation criteria, how is the government to obtain revenue? One of the major sources of government revenue from the agri- cultural sector are commodity export taxes. Although some profit taxes are levied on estate crop production at present, no direct taxes are levied on the smallholder sector. Indirectly, however, the government can obtain revenue by paying farmers a price lower than the export parity price and pocketing the differences in this "trading account" as revenue. Indeed one of the major sources of such indirect government revenue are the extremely low prices -- less than half the export parity price in 1984 -- paid to smallholders for Lobacco by ADMARC. This Disk8504/Malagrpol/5-01-85/ae - 49 - indirect method of taxation, however, introduces some distortions if the ratios of domestic to parity prices differ by crop. An alternative way of obtaining revenue is to place a flat commodity tax on all export crops. In addition, net revenue can be increased by eliminating the subsidy on maize. These export parity pricing and related scenarios can then be summarized as follows: El: Set Export Crop Prices at their Parity Levels This requires increases in the ADMARC producer prices for groundnuts (20 percent), rice (61 percent), cotton (36 percent), and tobacco (145 percent). E2: Set Export Crop and Fertiliser Prices at Parity This is a combination of scenarios El and Fl and would increase export crop prices as in El above and fertilizer prices almost 50 percent as in Fl. E3: Set Prices of Traded Injputs and Outputs at Parity and Levy a 10 percent Tax on Export Crops This is a combination of E2 and an export tax. This is done by lowering ADMARC's selling price for export crops to allow it to recover 10 percent of export value, net of its costs, as tax reven- ues. This implies an increase in producer prices for groundnuts (only 6 percent), rice Disk8504/Malagrpol/5-01-85/ae - 50 - (28 percent), cotton (21 percent), and tobacco (111 percent), as well as fertilizers (491 percent). E4: Set Prices at Parity, Levy a 10 percent Export Tax and Remove Subsidy on Maize This requires setting the prices of export crops and fertilizers as in E3, and removing the maize subsidy by both reducing the price paid to produc- ers and raising the price paid by consumers such that the unit losses on ADMARC's maize account are eliminated. This would entail a decrease in producer prices for maize from K 122/MT to K 107/MT (a 13 percent decrease) and an increase in consumer prices from K 159/MT to K 174/MT (a 9 percent increase). The results of the eight policy scenarios are described below. 4.3 Model Results The implications of the policy scenarios and their impact on various government objectives in summary form are shown in Table 4.3 (The detailed model results are shown in Annex B.) The table shows the model predictions for the base run for the crop year 1984/85 on the assumption that existing pricing policies are continued. The results Disk8504/Malagrpol/5-01-85/ae - 51 - for the policy scenarios are then shown as deviations (in percentages) measured from the base run. Since all the policy scenarios under review are aimed only at the smallholder sector, and do not in anyway change the production responses in the estate sectors, no information on this sector is reported in Table 4.3. Removing Fertilizer Subsides (Fl). An increase in fertilizer prices from K 282 to K 240 per metric ton, ceteris paribus, has the following implications (Table 4.3): (i) Production of all smallholder crops declines, but at different rates. The most important are the declines in fertilized local maize (5 percent) and hybrid maize (9.4 percent) (local unfertiliz- ed maize production remains unchanged) which jointly account for an approximately 2.4 percent reduction in maize production from a predicted 1.5 million metric tons to around 1.46 million metric tons; (ii) This inevitably reduces smallholder incomes, but only marginally by 2.3 percent from an average of K 174 per household to around K 170 per house- hold; (iii) ADMARC's cash flows, however, are now considerab- ly improved, as losses on its fertilizer account are eliminated. We estimate these losses at Disk8504/Malagrpol/5-41-85/ae Table 4.3. MAlawi: MAdel ResAts for Alternative Policy Scenarios (Deviation Fron m98 Rn in Percentages) F1 MO M2 M3 El E2 E3 [4 E>xrt & Fertilizer Eliatinate Elminate Elinete Export & Parity, 10% Snillholder Maize Mize Maize Sbsidy Export Fertilizer Export Crop Tax S"eidy by S"sidy by by lowerig Export Cp and Parity 10% Eliminate Vhize xMidy Fertilizer Raising Lowering Prodcer Price, Crop Fertilizer Sallheolder by Lowering Import (bnsaner Producer with Fertilizer Parity Parity Export Prodkuer Price & Raising Base Rn Parity Price Price Inprt Parity Pricirg Pricing Crop Tax ODnsfner Price Saiahlder Production Maize: Lhfertilized Local 915,000 MT/yr 0.5 0.0 -9.6 -8.9 -2.2 -1.9 -0.5 -4.9 Fertilized Local 330,000 MT/yr -5.1 0.0 -14.4 -20.4 -2.9 -6.9 -5.8 -13.2 Fertilized Inproved 246,000 MT/yr -9.4 0.0 -14.6 -25.5 -3.1 -10.6 -9.8 -17.7 All Maize 1,491,000 MT/yr -2.4 0.0 -11.4 -14.2 -2.5 -4.4 -3.2 -8.9 Grojxhuts 78,000 MT/yr -1.4 0.0 7.0 5.1 13.9 12.5 2.5 5.6 Rice 49,000 T/yr -2.8 0.0 5.9 1.5 34.4 33.5 15.2 17.5 Cotton 21,000 MT/yr -1.5 0.0 6.4 4.2 12.9 11.4 6.3 9.0 Tobacco 16,000 MT/yr -2.6 0.0 6.3 1.2 5.3 5.1 5.1 5.3 nallholder Incones 174 Kw41-Vyr -2.4 0.0 -11.11 -13.2 8.91 6.4 3.0 -2.3 c -4 AD wRC Cash Flow -22,263,000 Kw/yr 47.7 3.9 165.8 20B.7 -0.7 45.2 39.2 124.0 Paize added to Irwentory 196,000 MT/yr -7.9. 2.7 -85.6 -96.0 -65.2 -70.0 -42.2 -84.3 Smallholder Balance of Payments 42,866,000 Kw/yr 5.0 0.0 38.1 40.1 51.1 57.4 29.9 45.3 Urb inces 1,252 KWW'yr 0.0 -0.4 0.0 0.0 0.0 0.0 0.0 -0.2 NUTES ON 9N4RI0S F] - Fertilizer import parity: AEMC's salling price for fertilizera is raised by the <nit aount of loses in /oMAR's base year fertilizer accont (49%). Ml - Eliminate maize ssidy by raising consaner price: AEMR's selling price for mai2r is raised by the .nit enont of 10ses in AD44RC's base year maize acnt (18%), excluding valoe and costs of stored maize. M2 - Eliminate maize sbsidy by loeing prockner price: !"NRC's buying price for maize is loered by the Lnit mant of les in AC's base year maize accoLnt (-24%), exclutdir value and costs of stored Mize. m3 - As in 2, with fertilizer pricing as in Fl. El - Export parity: AMAFC's buying prices for )prt crape (gTudts, rice, cotton, snallhlcde tdbsoco) are raised by tife nit awnt of ADMRC's base year profits fron each crop (grorxhks tp 47%, rice p 49%, cotton Lp 42%, tobacco p 368%). E2 - Export and fertilizer parity: As in El, plLs XMiARC's selling price for fertilizers is raised by the nit anant of losses in ADM4RC's base year fertilizer account (49%). E3 - Export and fertilizer parity/10% asnallholder eport tax: As in E2, except ArHMRC buying prices for export crops are loered to allcw ADM4RC to recover 10% of eqort value, above costs, as tax (grourhit price 29% above bese year, rice Lp 19%, cotton up 24%, tdbecco Lp 315%, fertilizer tp 49%). E4 - As in E3, with IEARC's maize buying price lowered by 1% and IDiADC's maize sellirg price raised .9%. -53- around K 8.3 million. Although pricing fertilizers at "full cost" alone does not put ADMARC's cash flows in the black, this policy option does reduce the operafing losses by almost half from a loss of around K 22.3 million. The additional K 2.3 million gain' comes from reduced losses on the maize account, because maize procurements also drop; (iv) The maize surpluses added to ADMARC's inventories decline somewhat (by 8 percent from the base run), but remain positive and large. Although maize production declines by around 37,000 tons, declining smallholder incomes reduce own-consump- tion also by about 21,000 tons (a mere 2 percent decline in maize consumption among rural house- holds), so that marketed surpluses brought to ADMARC change by only 16,000 tons. With domestic demand remaining unchanged ADMARC continues toadd nearly 180,000 MT to its maize inventories after meeting doemestic demand. This pricing policy scenario does not threaten domestic "food security"; (v) The net balance of payment attributed to the smallholder sector improve marginally mainly through a reduction in fertilizer imports (a Disk8504/alagrpol/5-01-85/ae -54 reduction of K 4.8 million). These gains are partly offset by a marginal reduction in the exports of groundnuts and an increase in rice imports; (vi) The amount of fertilizers required to meet small- holder demand decline substantially. When ferti- lizers were subsidized, all the 60,000 MTs were purchased by smallholders, although some 10,000 tons were leaked to the estates. With full cost pricing, total fertilizer demand and use drops by about a quarter -- from 50,000 tons to no more than 38,000 tons. As this price is close to the free market price paid by the estate sector, there are no leakage to the estates. This drop in fertilizer use translates into a proportional decline in per acre use for local maize, ground- nuts and cotton (about 25 percent), a higher de- cline on improved maize (a 30 percent drop),lower decline on rice (20 percent) and the least decline of tobacco (12 percent). As very small amounts of fertilizers are being used on a per acre basis, per acre yield declines are very small -- less than 2 percent for most crops. Disk8504/Malagrpol/5-01-85/ae - 55 - Reducing Maize Subsidies (Ml - M3). Eliminating consumer sub- sidies on maize to cover ADMARC's transportation and marketing costs on its maize accounts can be accomplished in two ways: a) By Increasing Consumer Prices by 19 percent (MI) This policy scenario leaves the production side of the model unchanged and its major impact is on urban consumers and on ADMARC's accounts. The main implications are: (i) Average urban household adjust their consumption preferences by reducing their consumption of maize (by some 8 percent) and increasing, only marginally, consumption of groundnuts (1.4 per- cent) and rice (1 percent); (ii) Real average urban household incomes decline but by a mere half percent. This is primarily due to the fact that only a small fraction of urban ex- penditure is devoted to the consumption of maize and maize-flour products. In 1979/80 the average urban households allocated no more than 3.1 per- cent of its total expenditure to these products in Lilongwe, 2.6 percent in Blantyre and 7 per- cent in Mzuzu. 2 2/ See Urban Household Expenditure Survey, 1979-80, National Statisti- cal Office, Zomba, Malawi, September 1983. Using these data the income elasticity of demand for maize in urban households was be- Disk85O4/Malagrpol/5-O1-85/ae - 56 - The real sensitivity is about the impact of increased maize prices on low income urban groups. The Urban Household Expenditure Survey defines urban "low income groups" as those with average total household incomes of less than K 100 per month. Of a total of some 113,000 households residing in the three major urban centers -- Lilongwe, Blantyre and Mzuzu -- some 71,000 or 62 percent belong to this low income category. What is the likely impact of removing maize subsi- dies on two-thirds of the urban population that falls into these low income groups? These low income urban households devote a larger share of their total expenditure to maize products than the average urban house- hold. The 1979/80 Urban Ho.usehold Expenditure Survey shows the maize expenditure share to be 6 percent, 9 percent and 11 percent in Lilongwe, Blantyre and Mzuzu, respectively. A 19 percent increase in the price of maize implies a group in real incomes of no more than 1 percent, 1.7 percent and 2.1 percent for households in these three urban centers, respectively. Again no major loss of real welfare will occur from removing the maize subsidies. There is finally a group of urban households with average incomes of less than K 20 per month. These are the poorest urban group and they comprise only 4 percent of all urban households in these cities. Their maize expenditure share varied from a low of 8 percent in Lilongwe to a high of 19 percent in Mzuzu. For these households, the decline in real income as a result of the maize price hike would very Disk8504/Ialagrpol/5-01-85/ae - 57 - between 1.5 percent and 3.6 percent. Again this is not a major change in their welfare. From these results, it seems that the concern over removing maize subsidies and their adverse effects on the welfare of urban, specially low income households, are highly exaggerated. Even the poorest urban households stand to experience a loss of real income of no more than 4 percent and these households constitute less than 4 percent of the all urban households. (iii) ADMARC losses on its maize account are by defini- tion eliminated, and its cash flow position on maize marketing is improved considerably. We estimate that with an implicit subsidy in 1984/85, ADMARC will stand to lose K 39.2 million on its cash flow from maize marketing. Once the subsidy is removed, the maize account balances and the negative cash flows are decreased by some K 860,000. This is a. very small improvement because all additions to maize inventory -- that is, maize purchased and paid for but not yet sold -- are a drain on the current cash account. The maize added to inventory actually increases be- cause of a decline in urban consumption, and some 201,000 MTs are now being added to ADMARC stocks; Disk85O4/%lagrpol/5-O1-85/ae - 58 - b) By Decreasing Producer Prices by 25 percent (M2) An alternative to increasing consumer prices is to reduce the producer prices paid by ADMARC. This would reduce maize producer prices relative to other crops while keeping urban prices at their present levels. The main implications of this policy scenario are: (i) Smallholder maize production declines by about 12 percent, while the production of other crops increases by about 6-7 percent; (ii) Smallholder incomes decline on the average 10-12 percent, depending on the farm type; this also decreases rural own-consumption of maize, ground- nuts, rice and cloth, but by smaller amounts; (iii) Maize surpluses added to ADMARC inventory de- crease dramatically from a high of 196,000 tons in the base run to only 28,000 tons; (iv) ADMARC's cash flows from maize marketing show a dramatic improvement dropping from a loss ofK 39.3 million to a loss of only K 0.5 million; (v) the net foreign exchange earnings attributed to smallholder crops increase from K 46.5 million to K 62.8 million -- a 35 percent increase attribut- ed mainly to an increase in the production of export crops and a slight decline in fertilizer and rice imports. Disk8504/M%lagrpol/5-01-85/ae - 59 - From the point of view of some government objectives eliminat- ing maize subsidies by reducing producer prices may seem preferable to eliminating them by increasing consumer prices. The improvements in ADMARC's cash flows and maize account are much larger, the improvements in net foreign exchange earnings are also substantial and maize added to ADMARC's already large inventories (which have a large storage cost and carrying charge equal to the interest rate on the value of the maize lying in idl, storage) is far smaller. This is of course achieved at a price -- lower incomes for smallholders. Smallholder incomes are on the average between K 150-K 220 per year and are as low as the lowest urban income groups. They, however, constitute the vast majority of Malawi's population. If the government is sensitive about the impact of its policies on the poor, then it may prefer not to eliminate maize subsi- dies by reducing producer prices, preferring instead to increase prices to consumers as in Ml. c) Eliminate Maize as well as Fertilizer Subsidies (M3) If GOM chooses to eliminate maize subsidies by reducing produ- cer prices, as well as fertilizer subsidies by increasing fertilizer prices the impact on the smallholder sector incomes is even larger. The main implication of this scenario compared to the one in which only maize producer prices are reduced (i.e. M3 compared to M2) are: (1) Smallholder incomes decline even further -- an average decline of 13 percent, compared to 11 percent; Disk8504/Malagrpol/5-01-85/ae - 60 - (ii) Maize production decllines even further from 1.32 million tons to 1.28 million tons, while in- creases in the production of export crops is smaller between 1.5 percent. (The major differ- ence is in the production of smallholder tobacco which barely increase by 1 percent from the base run output of 15.6 million tons, as distinct from an increase of over 6 percent under M2); (iii) ADMARC's cash flows are now improved even further as both the maize and fertilizer marketing losses are reduced. Total ADMARC cash flows from mar- keting activities are now in the black with a net positive gain of K 24 million. (In the base run, ADMARC cash flows were running at a loss of around K 22 million); (iv) Maize added to ADMARC inventory declines even further to a low of 6,000 MTs; (v) The net foreign exchange earnings are also im- proved, but only marginally increasing from K 63 million (under M2) to K 64 million. The net addition to maize inventory in this scenario warrants further discussion. At present income levels in rural and urban areas and at prevailing prices, the model predicts a total domestic demand for maize of around 1.3 million metric tons in 1984/85. (That is in the Disk8504/Malagrpol/5-01-85/ae - 61 - base run scenario.) As ADMARC already has considerable maize stocks at hand, any policy option that adds little or nothing to these stocks would be an acceptable long-run scenario, from the point of view of "food security" considerations. Net additions to the inventory in any scenario imply that under price and income conditions prevailing in that scenario, domestic demand just equals domestic supply. Even if GOM is risk averse and wishes to hold some emergency stocks in excess of domes- tic demand -- say 20 percent of a 1.5 million ton demand or around 30,000 tons -- it should aim at reducing its presently held stocks to that level and assuring that no net additions are made to the inventory in the long run. A combination of eliminating maize and fertilizer subsidies by reducing producer prices for maize and increasing prices for fertilizers seems to come closest to this long-run desired equilib- rium. Holding larger stocks may not be desirable at least not merely for normal "food security" considerations because large unused food stocks represent a high opportunity cost because of the associated storage costs and because of the capitalized value tied up in the stocks. We do not know exactly what ADMARC's storage costs per ton are, but at an estimated Zimbabwe f.o.b. price of K 180/MT 3/ and a 10 per- cent annual interest charge, each ton of maize in inventory costs K 18 3/ Most recently in March 1984 Malawi sold 40,000 tons of maize from ADMARC stocks to Zimbabwe at approximately K 242 (US $220)/T. Assuming a transportation cost of K 62/T we get an f.o.b. price of K 180/T. Disk8504/Malagrpol/5-01-85/ae -62- plus storage costs. If we assume storage costs of K 12/T, then the ann- ual opportunity cost of keeping a 30,000 MT reserve would be around K 1 million per year. 4/ The main losers from this policy scenario is Malawi's small- holder sector -- its rural poor majority. Equity and welfare considera- tions might alone make this policy scenario undesirable in spite of its positive impact on other government objectives. Export Parity Pricing (E1-E4). Allowing border parity prices to prevail is a common policy prescription because it removes relative price distortions in the domestic economy by bringing prices closer to their international opportunity costs. In Malawi the prices of most export crops remain below their export parity level in spite of recent increases in export prices. Maize is a special case in that due to its bulk and the high transportation costs -- at least via Durban -- its import parity price is higher, and its export parity price is lower than the domestic produ- ction costs. Thus, in a normal year, maize should not be traded inter- nationally. It is true that in the last two years Malawi has been able to sell its accummulated stocks of maize to some neighbouring countries -- to Zambia, Zimbabwe and Tanzania -- and probably at a price high enough to cover ADMARC's purchase, transportation and storage costs. This has been due to the special drought conditions in these countries -4/ These figures are guestimates and can be improved once the actual storage costs are known. Disk8504/lagrpol/5-0l-85/ae -63- that have dramatically affected their maize production. These export outlets do not provide long-term unavailability of hard currencies in these countries which have enormous balance-of-payments problems of their own. Exports to these countries have to be financed via donor aid, as in the case of Malawi's exports to Zimbabwe, a share of which were financed by aid from UK. Any maize exports from Malawi have to come out of ADMARC's inventories and each deal requires a one time explicit policy decision. The producer prices of other smallholder crops, however, can be increased to their export parity levels. We examine several related scenarios. (a) Increase Export Crop Prices to Parity (El) This involves increased smallholder producer prices for cotton (36 percent), groundnuts (20 percent), rice (31 percent), and tobacco (145 percent). The major implications of this policy scenario are: (i) The production of all export crops increases, and the production of maize declines. Although there are quite large increases for rice, groundnuts and cotton, tobacco production increases only by 5 percent because of the imposition of quotas on tobacco marketing by ADMARC; (ii) Smallholder incomes improve, dramatically for those households which grow groundnuts and tobac- co in addition to maize and where their Disk8504/Alagrpol/3-01-85/ae -64- production substitution possibilities are high; on average smallholder income increases by around 9 percent; (iii) Since ADMARC pays higher prices, its cash flows and operating losses increase and its total accounting balances specially for export crops but also for maize, decline. This is because as less maize is produced, less is procured and traded by ADMARC. Its operating cash flow losses on the maize account also decline because of the lowered procurements; (iv) Maize surpluses added to inventories decline considerably -- dow from 196,000 tons in the base run to 68,000 tons. Domestic food demand (some- what increased because of an increase in small- holder incomes) however is more than amply met and "food security" is not in anyway jeopardized; instead ADMARC's maize inventories continue to accumulate although at a lower rate; (v) The increased production of export crops implies higher exports, and net foreign exchange earnings from the smallholder sector increase by 47 per- cent -- an increase from K 46.5 million to K 68 million. (Not all crop exports increase. In- creased incomes generate a higher demand for Disk8504/ alagrpol/5-0l-85/ae - 65 - cloth and hence cotton, such that the increased output of seed cotton (13 percent) is absorbed by increased domestic demand.) (b) Export Crop and Fertilizer Prices at Parity (E2) In addition to increasing producer prices this requires in- creasing fertilizer prices. The main merit of this scenario is that it removes domestic price distortions on both traded inputs and outputs. Governments often pay producers low prices for traded outputs and then try to ffset this distortion by providing fertilizers at subsidized prices and introducing an additional distortion. Of course fertilizer subsidies may be designed to keep the output of non-traded foodgrains high, and lower prices may be paid to producers to provide government revenue via an implicit tax on export crops. This set of motivations may be close to what has been happening in Malawi. It is useful to see what the implications are of removing both sets of distortions -- in output and input markets. (This is a combination of El and Fl.) The main implications of this scenario as compared to the previous one (E2 compared to El) are "predictable" (i) The production of export crops increases but by soalier amounts; as maize is a heavy user of fertilizer its output declines even further; (ii) Smallholder incomes improve but by a smaller amount; Disk8504/Malagrpol/5-01-85/ae -66- (iii) Maize surpluses added to inventory decline to a lower level; (iv) ADMARC's cash flows improve considerably mainly because of the reduction in losses in its fertil- izer account; total ADMARC cash flows remain in the red in this as in the previous scenario because of the large losses due to subsidies from the maize marketing account; (v) Net foreign exchange earnings increase even more both because of the increased exports and because of some decline in fertilizer imports. Fertiliz- er leakages to estates disappear; (c) Parity Pricing and Export Taxes (E3) Although the previous scenario eliminates price distortions for traded inputs and outputs it does not address the issues of how to generate government revenue. At present, although ADMARC's accounting balances from its total operations are in the black, this occurs mainly because its losses on maize fertilizers are offset by its grains from groundnut, cotton and tobacco trading. The operating cash flow ac- counts, however, continue in the red because of large losses on tghe maize and fertilizer marketing accounts. There are no direct taxes levied on smallholder production -- all such taxing is implicit in the lower producer prices. The "implicit tax" as measured by the ratio of current producer to parity prices works out at 17 percent for Disk8504/Malagrpol/5-01-85/ae - 67 - groundnuts, 38 percent for rice, 26 percent for cotton and over 60 percent for tobacco. The alternative is to levy a flat 10 percent tax on smallholder export crops after removing price distortions. The implication of this scenario compared to the previous one (E3 compared to E2) are: (i) Production of maize is relatively higher while that of export crops somewhat lower; (export crop production increases compared to the base run with -present policies, however); (ii) tobacco production which should be most affected, remains unchanged because of the imposition of the marketing quota; (iii) increases in smallholder incomes are much lower as are the net foreign exchange earnings from the smallholder sector; (iv) the amount of maize added to inventories is somewhat larger; (v) ADMARC's accounting balance is improved somewhat and so is its negative cash flow. The reason for these less than dramatic results on government revenue is the small amount of the export tax. A 10 percent tax on all export crops does not begin to replace the high "implicit taxes" currently in place. At present, what government gains via Disk85O4/Mhlagrpol/5-01-85/ae - 68 - export taxes are still being lost via the subsi- dies on the maize account. Even higher export taxes would bring in more revenues but also more maize and as long as this is not sold, ADMARC's cash flow situation does not improve dramatical- ly, and unless we assume that the maize can be exported it does not improve government revenues either. For this we need to remove the maize subsidy also. (d) Parity Pricing, Export Taxes and Removal of 1Iize Subsidies (E4) This scenario removes all price distortions by pricing inputs and outputs at parity and removes the maize subsidy (by both reducing producer prices and increasing consumer prices) and imposes a 10 percent tax on smallholder export crops. It is useful to compare the results from this scenario to the results from continuing the present poli- cies. Such a comparison shows that under this scenario: (i) Maize production declines -- by 9 percent -- while production of export crops increase, the largest increases occurring in rice and cotton, because marketing quotas still restrain tobacco production; (ii) smallholder incomes fall, but only marginally, because although export crop prices increase Disk8504/Nalagrpol/5-01-85/ae - 69 - somewhat, maize prices are reduced; maize being the major crop, farm incomes are reduced; (iii) ADMARC's cash flows improve dramatically and are in the black (for the first time) because both fertilizer and maize subsidies are eliminated. The maize cash flow account alone continues to show some, though much reduced losses, because ADMARC is still buying maize that it adds to its inventories; (iv) the maize surpluses added to inventory are much lower -- around 30,000 tons compared to the current 196,000 tons; (v) the balance of payments improve considerably, due to higher exports and some reduction in fertiliz- er imports; (vi) urban consumers experience a very small decline in real incomes -- only 0.2 percent. (Recall that maize subsidies are reduced by increasing consumer prices by less than in the earlier scenario (MI) in which all the price changes were borne by consumers. Therefore the impact on the poor and the poorest urban consumers will be lower than in that run.) Disk8504/Malagrpol/5-01-85/ae - 70 - As compared with other policy options this scenario offers a number of advantages -- it reduces costly maize inventory accumulation, it improves ADMARC's cash flows bringing them into the black, and it improves the balance of payments. It accomplishes this at a small loss of real incomes to smallholder producers and urban consumers. 4.5 Evaluating the Policy Alternatives The central purpose of this modelling exercise is to make the trade-offs among objectives as explicit as possible. Each policy can be evaluated in terms of how well it meets various goals by assigning ordinal rankings. By examining these ordinal rankings it may be pos- sible to find out which policies dominate others in terms of the policy goals being considered. Such rank-orderings are presented in Table 4.4. Several interesting points emerge: (i) One can certainly improve on the present pricing policies in terms of almost all goals; the excep- tion is the real incomes of urban consumers, but in this regard many other policy scenarios do just as well (because consumer prices are left unchanged); (ii) In terms of nearly all objectives except small- holder incomes and a small difference in net Disk8504lagrpo/5-01-85/ae - 71 - foreign exchange earnings, M3 policy scenario dominates other policy scenarios. If the govern- ment does not assign a very high priority to the incomes of small producers, it shoild choose this policy package -- that is eliminate maize and fertilizer subsidies, leavings other policies essentially unchanged; (iii) considering that the impact on urban consumers is very small, and that on small producer incomes far less than M3, the E4 policy scenario also offers a very good alternative for meeting both equity and efficiency considerations (in terms of removing distortions), as well as improving ADMARC's cash flows, while maintaining a rea- sonable level of net additions to maize inven- tories. Disk8504lagrpol/5-01-85/ae - 72 - Table 4.4. Malawi: Alternative Policy Scenarios and Their Rank Ordering in Terms of Alternative Policy hjectives Policy Scenarios Base Cbjectives RLn Fl Ml M2 M3 F1 E2 - E3 E4 1. Real Smallholder Incorres 4 7 4 8 9 1 2 3 6 2. ACMARC Cash Flows 8 4 7 2 1 9 5 6 3 3. Reduce Maize Invntory 8 7 9 2 1 5 4 6 3 4. Balace of Payments 8 7 8 5 4 2 1 6 3 5. Real Uban Incomes- 1 1 9 1 1 1 1 1 8 Since all scenarios except Ml and E4 leave consuner prices unchanges, other policies rank equally with the base run, with regard to their impact on real urban incomes. Notes on -cenarios Fl-Price Fertilizers at "full cost" Ml-Eliminate maize subsidy by raising consaner price M2-Eliminate maize asidy by lowering producer price M3-M2 + Fl El-Price export crops at parity E2-El + Fl E3-E2 With 10 percent tax on export crops E4-E3 wth elimination of maize sbsidy partly by raising consLner prices and partly by lowering producer prices. Disk8504/Mlagrpol/5-01-85/ae - 73 - ANNEX A Algebraic Model Specification The purpose of this appendix is to explain how the model described in this paper is specified in mathematical terms. For the sake of clarity, this discussion is simplified somewhat. An exhaustive listing of each equation and procedure is not given. Readers interested in a comprehensive specification should contact the authors directly. Smallholder Production Smallholder production in the model is based on the Cobb- Douglas (or C-D) production function: al a2 Q, aA. L. F. 1 0 1 1 1 where Q is farm production A is farm area L is labor use F is fertilizer use a , al, a2, and a3 are coefficients i is the crop type subscript. Disk8504/%lagrpol/5-01-85/ae - 74 - Smallholders are assumed to mazimize profits which are speci- fied as the value of production minus the value of inputs, as follows: i i = Ep Q - wL -rF i ii i i where pi is farmgate price w is the wage rate r is the price of fertilizer Two limits act as constraints on smallholder profits. Both land area and fertilizer supply are limited. Therefore, the fol- lowing constraints are specified: A A limit i limit In tobacco-growing households, tobacco production is also constrained by quotas. For policy simulations, the profit-maximizing problem specified above is solved directly using non-linear optimization algorithms. However, for model calibration a different approach is used. The fol- lowing paragraphs explain this procedure. Disk8504/Mlagrpol.5-01-85/ae - 75 - For each production system as specified above, there are profit-maximizing levels of production and inputs for any given combina- tion of prices. The equations specified above can be rewritten to express the profit maximizing behavior directly as a function of prices, as follows: 1 1-a -ca -a 1 2 3 Pa 1 P a 2 P a 3 *Q {a (i2 __3 Q.= {2( ) ( ) i 3 1 o 6 w r+4 1-a -a -a 1-a -a a 1 2 3 a 2 3 a 2 a a * 1 2 3 3 A ={(Pa) () - (-) ( ) 1 1 io 6 w r+ 1 1-a -a -a a a a 1-a -a a a 1 2 3 * 1 1 2 1 3 3 3 L {(P a) (-) (-) (-) I i o 6 w r+ 1 i-a -a -a a a a a a 1-a -a 1 2 3 * 1 1 2 2 3 1 2 F { (P a) - - - i io 6 w r+ where Q is the profit-maximizing level of production A, is the profit-maximizing level of land area 1 L is the profit-maximizing level of labor use L. is the profit-maximizing level of fertilizer use Disk8504/Malagrpol/5-01-85/ae - 76 - 6 is the lagrangian for the land constraint 4 is the lagrangian for the fertilizer constraint In the calibration, these equations serve as constraints. The calibration, taking these profit maximizing conditions as given, seeks to minimize the divergence of the calibrated behavioral variables from observed behavior, by the following function: Q -Q A -A L -L F -F Z = En ( +n ( + n ( +n ( )+ 1 Q - A - L - F - Q A L F i i i i where Qi is the observed level of production Ai is the observed cropping area Li Is the observed labor use Fi is the observed fertilizer use r, nA, n[, nF are weighting factors reflecting the relative confidence in estimates of each observed quantity. The calibration routine minimizes Z, once again using non-linear optimization algor- ithms. Disk8504alagrpol/5-01-85/ae - 77 - Estate Production Estate production is modeled using Nervolian production equa- tions of the form, Qt = a oQt1 al Pt-1 a2 where Qt is current production Qt-1 is last year's production Pt-1 is last year's price ao, al, a2 are coefficients. In other words, estate production is a function of lagged production and price. The coefficient values are estimated using econometric regres- sion techniques. Household Income Smallholder household income is defined as Y = Tr + Z7 + wL o ii hh where Disk8504hlagrpol/5-01-85/ae - 78 - Y is household income 7o is non-farm income 7ff is farm profit Lhh is household labor used in farming. In other words, household income is the value of production minus the cost of purchased inputs. Household labor is added back in here because it is not purchased, but was subtracted in computing farm profits. Urban income is simply a constant. Demand The model uses the Almost Ideal Demand System, or AIDS and is expressed as follows: D i1 P. i S. =a. + 8ln + EY..lnp I I 1 PI i 3 ln(PI) = a + Ea lnp + 1/EE y lnp lnp o k k k 2jk kj k j where D. is demand p is consumer price as is the share of household expenditures Y is household income PI is a price index Disk8504/aagrpol/5-01-85/ae - 79 - ct, , y are coefficients i, j, k are cnmmodity subscripts. It is important to note that the AIDS uses an unconventional price index. This is not the price index used in the model to calculate real incomes. Disk85041alagrpol/5-0l-85/ae 」 긔 б г+1�.•А.ьс Ад �цcsN1 мв�Е� . � ' а в� м�sгt s�zu:.� 6сг-�ги�гг�у _ �,..__._, С - .'-- -- •. С ,иг мz н t .. � � � лгвсit'с.ьи1lr.+сп r.гг.t ,•1в1ггС vа.лиссс г.�Е А5 iы м L, �А[.са ro с.+иег 7г.+иУ: Lo.+r,г66 ��рrьп 7АМК. WLTM fЕАт2сг+Е� , . ' т иГ,.ги� Саtп 4 n}кГ,JS Gдtli РRгсЕ Ar Рп1г.кТ`{ �, ;,;А : Е ,(3UN rс5иь15 гWгл..ьl2ё� XfS✓tд NadнM[ziD FEs�cп h��wso С 4 „ос[наиt .11 .! tьисtмии( l3иыМ,7и1 ью пиьlСг Гвlвwtl[.I +� у I1асвllС[[ ,I1 •3 1lУОГИ1С[3 г+0ииССв гв1Сl1 1ви /С! мг1 rем2 t33,/s ,е1 1:цl1 1!!.!в !t,11 7t.11 .[.1е гf,ы +..,t .lоаиее! ггlе[ rn. пьlг[ rl.сн 11г.1в ,11 fl1,11 11l,1! 71}.1! I!!,!! 1ь1,[1 111,а! гг.си rlииис[в м1с[ rur сиСиномик! С г.мt 1sь.1в .te 11ь,1в 1е1.1е 1tl,1! 111,и 11,,1в 1в1,l1 .в.иг rьаоис[в гllсс rин !tc[ яв•с1 ьаl,1! .1е ее1,!! 1l1.!! и1.1а 111.е1 ва! !! 1е[,1! , ...с3 r.uuuctн г.1с( rп! сиtlаи и.ео llь,ев ,[t l1в,11 t11.!! !1l,11 111,1! 111,1в 111.l1 ...га .wuucco .г1сс гие 7ны(.�цаг! гиllгсо и.г[!г 311.!! ,1t 3и,!! 1lв,1! 33[.11 111,1е ы1,1! 1l1,N rr.rlмг .гиаисtи в!!с[ rue [пыьппьп[в rt.rlLtte[ � ;' rя.trllt ыеь11 ;i1 , ь7о.1!_ 111ьА! ^ _. _ . . вС1,!! !1!!(в _. . !к!ьв! t!а1!! , . .в.ц[.г r,оьи[ы ..te[ гии ([ка1( гевкl(г1[в ь• С' еомvвг[и ,и ч еиМl.мtеи синlин(. r.teя и. +е! иг1 Сl.и1 1:1в,ев .11 10в,11 I10r11 119�:0 (l1,1! 111,1в 11l,1! И, [л.си ьь.ьlв ,и аь,п 1н.н иl,е1 и[,1е аз,и т,1а �i.a Сииlи.еа ..1с[ гС■ ыв��инонигi Сl.ы сипlw 1С[ rиы с. t 111.1! ,11 71s,lе t/1.1! 11t.1! 1l1,[а 111,1! 1!l,11 Сr.+г симwпс. .•1ct ru■ .tc[ • Се.Сlагм 3и1,Ч ,i1 3(Са,1! 11l.1[ 1(!!,1! 111ы ! ав!lгlа 111,11 св•пигн еи,lи.ев rllсе rua сиlгим е[агн Сл1и l7yn ,11 иг,l1 1Ч,1[ !1l.11 и1lи иС,!! 11а,р ег.lУ еииlимев relet ru! 1иС3■ .. (� :I1 /С1в![С! 11 .) lавегlС[ Ги! вьiС(! 1е. г!в nI1 1ов;+t tн.lв ,tl 1l1,1! 1!!.!! tЧ,1! t11ь1! 1.а,1! 11e,!! ra..nt .ае гМ(с[ Аае n.ltc ' Аов.а, lеа.ll .tt ш,11 1и.и и1,1е 1и,н ие,а! 1и,н го..еи гав rмееС rпи ерииномил . rol.lt 1е(,!1 .11 f!1 1! 111,lе !еl,1! 111,1е 11l,1в 11l,ав rи..н1 гив лысt rиг .tc[ гаl.еl !l1,е! ,t1 ы1;е1 11l,ев ![1,г1 1!l,11 вl1,1[ 1lе,!! rw.e1 еов r.te[ ro. кееоеиггии гаиго 1нв.1а ,11 31!l,11 1еl,11 3пе,и tu,l1 :3и,1! и1.!! го.ло ги1 ввге[ rив [и.ьепи(оеи гаевеса 14l.1и 11l,1! .11 1lв,11 111,f! +�3e.11 t11,1e 1fв,![ 101,1! 1ив.ви rw лыt[[ 1! иаиоискм r.tetl rпи efrы/ (иеи . гаl./l 1ь1в,!! в11 1l1/111 tl1,11 1111,1! 111,(! 1101,1{ 111,0! 1tlв.11 1W в41С[ ft .ииоtlсlн ге1С[1 /oi С1/в1t г[1 1и1.г1 е!!!ь1• ,t1 1!!l,11 t11.11 1111,11 t11,1F 1!l1.1в 11е,1! 1ов.в1 1W •ь1е[ l3 rrnouetн ггlеll ra, [1г,г1 ILUL.eu,ьo 1п•ы[и еае•11 111l1lt ,tt 11l1ь1! t11,!! 11/1.1i 1t11/{ 11il,10 110,1! . 1иь.аг 1ив внlС[ !! гьиоиСtи /etC[1 гиа t[11tt lwecl гиlв[еи Clrrne! ;n .r Cur.lcA еи lвкра лм .е. »11 l1г:.1 71l.!! .te 331 11 11l.lе 1!!;!• 1и,1! f!!,!в 11а,11 е1г.,1 (1v rМ1ег run ,ь17е CIA.lи 11l.1! ,1! 111;11 1е1.ев 1ы ,!! 11l,1! 11l,1! 1!l,11 С[r.си ctr .е1ее п7в екииноИиlа С1е.е1 !1[,1в ,11 1!!,о! 11l.!! 11l.11 t11,1в +l1,11 1еl.ее с1r.и( (1г rысг ron .1ct C11.C1 11l1111 •11 11!l,t1 111,11 Il1в,lе 11l,1! (1С1,l1 111,ав . Ct1.c1 C!I ги1Gг 1Uи CQIIUИ • С1�.1lв1 11l1l1 ,11 111,11 11l.е• ц1.1! 111,1! 13l,!! 1lв,!! C11.rta1 С11 г,[с[ гив r[,t1Lt7[s [ ,11 i!!!ы! !l,11 'ii ы,!! 1еl.lв ы:е! л!}и н,и ы!,Ч геиг.ь! [ивиСпwе(л rееllцСе+ 1и.ги еинlrгьеиl иевв и1 .ьа ,.1 �овУУокв 11ьl1 �!1 11,1! 11l11! 11i1! 1lв,lв 11ь1! 111,!! IOl9иУ1ь lивllnalo[■ 1иввССU •ьввlгlне аииtь (111в и1 гtи .в1 ' [ьааl!(в .11 •1l ввУ.,lь1 ьопввС nьtl[ [глиеТ У1Ч1tь 111i! wr в!■ 1+1 , еьаиllв► ,11 ььУм3l.. ьаивнС иь12[ 1•ваll УЧ1tв 111!! и1 в1. 1в1 � ,11 - ,11 tнOU[IьW ,11 .l IмYUYLMYY! nua1L Yl1YlI[ вiааСон[ ;11 - - - ' ' у .в ввЧиGин1 маоиС[tон Сан[и•rtlVн .,0 .1,е[Т Сс[веtм! ltвее ик в111 ,! ина(11 ьslLlrl[a1 s 2 ь1t вв11[ ,1t � _ .> авА3[ •/1t1 11 ,11 NвiaCM ОЛ1в1l •1l 1e1,11 11l,16 /11i![ 1!,!! 11)в11 •1,1е . ин.воС. Yn/lд1l177[a пв111 1и1LLMULO[л lеtlацСllан дн.ltlСо е3lв,ге .11 !3lв,11 11.lв 1l1е;1! 1l,11 11•г.ь1 13,11 • и.м.lци Y,rt.llLer[о иь17[ 1•.(ииьаt. .г((! t!е re! п. rtл зв1 �и.ь![! t11}[1 .11 111,1! �!!•е[ в11е1! 1l,1! ef1,1/ l1,11 ип.ь.[в ииltlllLttCO wь11( вь[в 1lкlв •в1 У11вьl1в l1l,1f ,11 1аl,f1 1!l.01 ll1.1! 11ь7! 11е,!! 11,в1 иw.Увlи инг(л11Сtt(о ньl3[ Lв1ов Уц Сииоь•с вtл иа v(и 1!1 I] . /л.вОСе 31l.l► ;11 l31,l1 I{.1,fв 111,1[ t1,li lв3,11 1!,!! /и.ваСи г(А11(17[0 LПСвI nьl[! 1^вымоLаее вво0иСТlои rи.r1[ьо 11и.г1 .11 1г!(,ев 1!!.1! 1111,го 1lв,lf 1!1l,,е 11t,!! ги.г1[w Г[.ete13(о ше,1 ,ь1(г 1пн(ниьоt■ .гесо 1.[ .еи пв •и .+1 i�o!! i и1.и .1е 1н,1! 1eе.ое 11[,1е 1t,1t 1,.,rt n.31 . ги..ы. ги11Ч7tо епеь[ ,ыге lчыrи4о[+ ве(в иl.У пь1 ьв 1ье,вв ,11 3вl,1! 1l1.1! 1!! 1в rf.lв 1и,1t 1в,Ч rн.У... rеи11ь13[о (иец пы7е кгьымиеоlе (•аов иы гп..о.г1 вг1 н. .г, r!1 1..!l![ 11,1в .t1 13,Ч 1вl,1е н:[е !1,lе 11.и Ч1е! r..г(нг re.11u3[о и1и( ньt7[ 1пы ьииь(е■ в[erltlte! Уке !.а .е. мь гг■ гв1 1л.е '11 оси tlf.ы .11 3•f,11 I1l.1в 1![�10 ,е.lв 113,в! .1,1е tи.воси 7•г•о.tо ,в1tt lп.им:ьо[в r.uouctln. _ 1r.re3lo 1вгв.3! .1! 1!lе,11 1l1.1! 3l1е 11 1l1ь([ 1ввs.lв 1е.1! Iи.rltьu 1•••п.го иь17[ (пвипаьо[! 11eto tra �t! и• rt! гг1 tьвьвt! lь.1в .1l 1ь,1в .1l.11 вl;71 lь,11 ье,1е /f,l1 1м.•.l• lпвмиыD .в11( f,•tLmno[ь ьвtв 111е1 Иь1 � [e1,ef .!1 ег.,lf 111,1! 11! [1 rl,1! г11.11 1.,1А гн.ь•.и 1..+u.to ,.11[ (,пмиьо(е евlа. и!. 1и.ьовг! .е, и. ,[! r.1 G i�иl[!1 i1t113 ,ii 7l1,l3 11l.lв 3t1:11 11в1! 1ье,l1 [1,1! tи-sьиг 1,r•u.to и.1[[ lицьпиеоев r[ег1(17г■ ие1 1ве re! иь вг! r!I п1•.аси 1.•е.3! ,11 1в•1,11 1аl,!! fltl;ы le,f[ t1гв,[, !l.13 ,[•.ос, ..1 и 1..ьи,1».оС! .гУоис7lин _ .t.11[Со 111в.11 с11 tf3l,lь 1ы,1в 1111,1в г1,е1 1ввl.3! 1ь,•ь пг.г1[ьи иьlt[ е..,(..иьоt! .tеьа 1.и вев п.(и гИ1 иtвьь[ь 1гl,ье 11 .I0,1� (11,11 1в3,в{ t1,11 1в1,ьв 1в,1! ,е•ь.ь• гвlt[ 1w,LLnulo[в вь!• 1.Ои! nв1 в +1.ььи гlr,r• �11 3.1.l1 е1l,11 кf1,3! tt.3в 1fв.г! гг,ье ,е•иь.• ны п е,ьцмУеекв е.еи. 1с г,г.о..! ге. .. •ен 1.1 С +t.rtll 1ь.l. !71 1.,3е 111,[1 1l,1[ tl,е1 1в,13 11,l1 и7.rьнг ..trt lмьыпиьои геигпt �+ и![ t.e ег! ив ве! 1!1 +е.lеаи! 1н1.п ы1 и1l,17 n1л! 13и,l1 l1,ее 1tl3,l1 /в,3з nt•неи•! .ыее 3,выпи(ое+ еомlии.l7ан , гl.оеанl rl,n ви в+,е3 .г.1[ п,11 w.u 11,и w.u пе-иwн! ,ыг[ ини., еипlи•мlаи п1в1Ll1[Уа Ifl,11 All 317i1[ (1[.1, [1,71 1в,3! в,[l 3,1в •I.Iиивеи! пькгL 1YвILY1 ьоа(0 1П [7а.ьС[ n[нttrlu/ 11l,13 ь11 ь1в,lв 11l.[! i3з.13 !11l1 �!1,в► 1е11! _ ,3•[еы 1иг и.t7[ rамв� иье.[1 1еы 3urrrelCнe. •.t1a 11Lв:Сиt1 . ! ,11 Сии/1 t1 .s ьииl! сииини,иг[ i ан.Аась 11,3! y1t г1,й1 111,!! 1t;1[ 111,1! l1,.1 1l1,01 FИ.во[и сииино,игв f•ыlnntot! rвоииСllон аи.11еLС 1l1,11 11 ьв1,(1 110ы1 7t1,�! iв,ь1 31в.а1 11,11 eи.rlelu едииМи.иl1 (•.(ениеоlв rr[La 1ве г[в ив !(■ rnt � Сь.ь1Аi 1е7,31 1t1 11з}е1 11t.1! 1,1�1! 171,1e 21l,11 1ь1,13 :и.в.l. С.иинониlа 1м•ынпьl,t! ьвtь (ееое иы би.г в 177,71 11! 311,11 111,01 1/1�1а 11,lr 1[1,lе 1(,17 � сн.ьввв Сгииниииге 3.выпиьУtв tвеив и3[ гив.о.г1 в1в .. в[в гв1 1 г [в,в! ,11 11,.! 1е1.!! 1з,м 13.в1 1в,а1 f1,e1 снлtгг саиино,игв 3-вььпиьое! rгвгlь7t[■ и![ с.е rев и. rte г■1 С [,•1LСвв !!,!1 !11 1lьl! 1!!,[1 71;11 1311в 3з,[1 ы,1! ск.всииl сьиининиг[ fнььLн�иtы гин7игrьlоМ . аи.4ССнв ,11 11 1i 111,10 ll 11е,1! ,71 1аl,11 ,� еw.н<УМ! еlUУвоиУГа инllьи СанlУивгеи. О аи.г3� и}I1 (ii 1l;3е !1.l1 е1:А[ 11l,[1 11.(! 131ь[1 , си.(ь• crYUw,иr! [ьги!(! ! , 11Cв �11 вlьв! 1а1,l1 � •! нiС[ •tCl С 1. оа ь1.и 11 ь1,lь 11в.а t1,и tн.ь1 у ,t•воСи .гс[ l..и�.ии eи,wuclto, е1.ггlСС 11tl,ы 111 [111.t! 111.!! 11[(,ьв 11,1t 1в3г,lв г1ь![ f •е•rlееи в1е1 3•ьеепи(п[е .1ew 1.е rев Мь .гв rв1 , !},•!1в 71ь1! ,11 eI,11 111,1! 71 11 1+1 вв 71,11 1!!,1! � � не•в•lь .1C1 1.всемцоtв ььlь 11вl1 n•1 !1.(ь1■ 11a,17 11 ь11,11 111ь1! 711�ь+ 11,11 7lв,!! Iв,1! .1•свlи л С А1вг[!1 а,е! 11 1.,1! 1l1,1в е1,31 l1,1! 1г,н 1в,р 1се и.(ии.аи иlии ы 1и.Мо..! .ав .в .(! г.1 , ы.ыаиl н,[! ;11 11,н tl1,lв 31,и н 1г 3! гг ..,[[ � �lльиг ne[ !и<сиlи.о[н rаПй7ев Уе( ие r(г и. ее! г!1 вtвУсои! вJ! � � 1•.w,s ысе 1•ьи•УСаи сонси,nгиМ („' ljвe/f •1t l,11 111,1к l,11 11l,1в 1,1! 11t,t1 и1.иСин! •1Ct ивевw [инки.вllУи , 13,1! е11 17,1[ 1l,f7 I,.1! 11iьf1 1еь11 111,1[ •1.(гв в1С( tгвпы7[ ' 11[аСС1Iаи i11 •) s[aoCU1Ia, 1([о Go1IW " О [1.ьоСи 1I,1t ,1i 31;13 111.lе 1l�/t 1lе,1[ 31.lf l1l,11 C1.roCw eLLO eJltuи 3wь4ьпиьи[и вlиоиСг(аМ. . • •� • • у� tt.llпa s71.71 .11 11l,1t 111,а! i11,/e /f,1! 311.1! ts.ьr е1•.кии lьtо са7rин l.вьСпа(ое! rr[[о г!е .е! ,• rеи r!1 t [1.ыlь' ы,ь1 .1t 1!,{[ 1в[.1в 11,33 1l1,1f 11,1! 1l1,ге + О Сг.А,1! 111,3! 11 1l1,3г 111.lв 111,ы 77,lе 131.1! 13,ь3 С7.ь.е• 1е[о еаггии lив4(пи(оt! ь!t• г1!!в н•1 , C1.rl, Ct•:.lе иео со11и, [Иьеьпиьоен ь,1пг и3е г.ноь.3 •еа п. гев r.} Т 1111i 311 71,11 11l,10 1l,11 i1,/! (l,11 11,l1 С1•Iен1 1[ео CotfU, 1+llcnulY[е /lи11ll![в 111С 1ве et! nь вtв I!{ С7.lСои! 11,1! ,11 11,вь 111в01 !13[ lа,1в 1.11 1в,11 С1•ыУп, [о11ип Ски1п г,в((ниьо[а сиМ3и•вкlпr lfeto [аТlаи (аиlrь(1н11 � G С1.ыаи1 1[,3! ,11 1А,1! 111.ы к[,1! 1t1,t! 13,[! t1l.Ч е!-ыин, еиlгин Сеиги иивьн Сииьипвllом гыео еогlан tеиlигееМll . с7.!(в .l,11 .11 .1,и tu,1e ь!1 .п.и ь31 .1113t [г•ur utu canw игилп А' ' О 1иlь! А!1 .! lи.ви wGaи , 1УвваСМ 1(l,е1 •11 13t,11 11в,ь! 11l,l0 1F1,1! 13l,вf 1l1,1в , 3и.виСи tuelв [ltвг[ iви0иС1lПн [имlСоиl 1e,1! ,tt з!;(1 1е[,11 tl,н lr,fe t!,!в .r,1! еи•меw[ wси мв,ьпи(оt! Cn,�wrllaч С 1У.иС4м! 3в fs .t1 г1,ы 1е1,!! 3l,f/ 1l1.1в [в.1! 11е,11 lи.исУ.а 3ис.. У.сви сои3ипв11ои . 1u.11r 11�l1 ;11 11,10 1l.з! 1l.11 1ве11е 1t,33 llв,fe 1 [и.еьг lиси [вWв1! и� G t1! 1[ввеLи 31ь11 ;ii 71,n 11е,1! 1s:1i 1из1! 11ь11 1Ч1Ч �� •! j�.lисн tt. t31ы[ вlиоисгtаи амо Leruera о' G ТавьС[0 •1t • I .в IапвССа 1аlьССо , 1о.lо(и !! 3s ,11 1f,11 111а1! 1l.13 11l,1! 1f,1[ 1в1.31 , 1о•+иси ьиьь,пхоеы [пвысП иааисгlпн tонеl[ta [(в�1в ,l1 1l1,1А l1l.11 [11,1! 11,7! 11l,!! п,1е I' 1а•.1elu 1и.ииае,оlи п,ивеео rftto (ьС .(в ив ген .r1 t:. 1U.ве1ь' 11,1в ,11 31,1! 1//,t1 ь1 11 117,!! ь!,!/ 11i,1! 1о•a,l! lмвььnУ(оfУ гпивССа lьlь (11ви нь1 ta.lьls в!lьгl ,11 111,17 11l,lв )11:1• 14,if 1в[,31 1!.!! 1а.tьlв lнвььnйо(в гиlвеео еьlии и!t (иьневгв вев п• rьи гв1 1 1a.rl[! 1г1,1! .11 113,11 11l,lе 1ев111 1t,11 . 111,!! ьl,1е � Ia•гlиг lицемаео(л rue.e[а l[е11е[г[. и1[ 1ье в[■ Мь vl! 1м1 t; ru.Ct+e f f1 .11 !,!t 111.1! 1 11 tl1.1! 1.f1 311.11 1а.([и (п,ььмаr,осв lпвысо ыь..ее 1и [ь.мг ..вь(г1 1а.вlаС I!�!! ,1[ 11,/e 111,в1 1[;1! 11вг1е 1l,11 111г11 1о•в,и[ 1nв4lnYLaL! 1UlвССU ГlUСУв[а !в виwь![ С J7�l7г 11111 •,1 1iet1 111•t1 1l,1[ 11l,во fe,t[ 11Ir11 '� 1а•Сьг lдвььниlоtв IWв[Са t1luиtto вr ьоиввС 1rlt! GucclhC гг.lоси 11.13 ;11 (1•.1f 1t/./e 1з,1f 1!!•!! 3f,11 111.1! rг.•ись гtо(.саиlи (А1ьгt w.всси .■nоисг7ии ..о [ьrагl! и.еаеи !!,3! ,n н,1! и[.и .l.1! tиьо! 11,1а 11е,1! ' ее.rаеи wныТ Сии[ кииееи rwoue7taн .ьо [ьnиг! ( А Iсв11еl!(л ь11 .е ггигlеlи! rtrltьtl[в " г!llвв ыьlв j11 в1,1! 1![.11 t1 31 +[,!1 3е,![ 1г,1• � l[•1-. 1м.ымаьоеи rеигlСl7ен rине.ьае! t� г[вlвfв 111Is •1) 11,1е 1!l,11 11�11 11,11 1,11 ьl,а1 1ги1.в• lnвllnULOlв 11Yt1[lкlв 471Y l1и /loItLllta lUCвl ,11е1 , г[l1.Iи е1,вв ,i1 t1,1ь 111,!! 1[;е1 1!,!! 1,11 ь1,l3 11иt•1и 1ивыИоьпtм г[•11(I1l! иlia пи гИ.+мьа ,в171 /ll1..[ 11,11 .11 11,11 11[,!1 l1,te }l.t3 tl,ва /f,/T , 1ев1.н[ 1,ыкписрtв iteilLlttв и3ка иМ .ь1tС � � Illt.си 3,lв ,11 [,!1 111,!! 11 111,13 е,1ь 11,!! rt•r.в, lпвиниtоtв /tн11ь1С1! и11о 17и сеиинониll ! г[лиlг !}13 нtе [,11 1и.!! �;ц nr,l1 1,п и,1. „ гиl.ьг 1п•ымаепи lurturu ипо ои вlте 1(в1•е1 1.31 ,11 1,1! t1[,![ 1 f3 t11./s 1,1! ы,.в ы,1.е7 1•ьеLмиlп(! relrttl7(! u3to аи еокlпн (, rtвr.la [,1е .t1 !,3! tlе.!! ь:3f 111,г( ь,11 11,1t г(+1.1и lnвыниьаtв ге•г1(1ц! и7еа ии tиlьгеи rtltСив 11г11 .и и,п 1Чги 13.и Ч1.и .и [.и г[и1и.. ruпYUV (иыо гиии 1,вУьниtо[гв tu tпlгев А9 .1t t/elt.o l1,в1 ы 1 ев,11 1lе,11 1l;11 t!!,!! аl,11 11а,1! ег(l1.о 11fь1t г1и11(7е[в оlиьнУ qr 1rL�r.eu и,и .n 1l,и 1н,w n 11 !1 [! ,1l,lа 111.а1 иев1.[и иегl[ гг•r1e1[[н тео оИ lиеьг ' .. ertlr.lt 11.1! •11 11,1! 1l1.!! 11.1! �•• �!!:1! '� t3,1в ` 101,11 ' [1ьм1•tc еfгьы /tиrttl[(в и1ьD ин гСв � [r1At.oI 3l,31 .11 3l.33 111.1в г1,3! 1!в 1в 3l,t7 t11,1! егемl.rг t31.1L г(мг1С7е[! иlео ии гьиt.еи.(п lмвеео 111вI.а� 11,11 ," 1Ъ,11 l1l,/1 1\,11 1[!�!1 73,а1 1!l,11 (Геиl.вl ([1ьг[ гlЧ11({23! У1еа U, гив(LТ IalьССд а" (1еь(пв ~ ! •11 .! lвв.Iв• е11Уи13 1••ие1! вио /URCICM [гСпЧс[ 11001 вн /ьв гМ1 •1.сьвrьУ ,11 .[•[ьвьь. •ь1tС [ввииt вциt С :M.t,evц [г!1•,f1 .11 31а+l�1в lв.n [ьвlв,Ч 1lв.n 111вг.11 131,11 ен•е.г.ц еииининиtа е,.ии1 r,wL � !е•Ltrвll ь13.в31 �11 !!lи,1! 11,![ 1!1 ,11 131,15 1(.1,l1 111,11 •(.lввьвL /1С( tьвПнl вьlи! 1 ег.еьr.ц .1п1.31 .1е .1н(,е1 111.1! 1[� 1. .1/,r[ 1Ч,еа .1l,13 е1•иnьь сиt7ын еььсн7 r•ьи[ Г 1и.[гlвьl, Jf11в,7! ,11 371ь1,1! .в.r! 3l111�11 1!l,1в .flв3.3! 111.(• 3и.lьr,вь 1ис1! (гв11аТ ыlи[ к ([вlьввв4 вг1ll,г! ,1F 11ь11,7в 110,/1 l1111�г! 111г11 /Tiil,lв 1i/,1! 1t•Lьвввl 1[ь (ьгивl •в(Уl .r го.а.гьь 33н(.и .Ч ипв.и 1!!.и 1вн1,1! 1и,ц и.н,ц 1л1,13 го.иr.ь( 1и.ьмоьее. rue.eeo [,пгиг ..ьие � rr.ев...ь 1tг33.t3 .71 ьf311,13 1ьl.!! в[1!!,3! 111.1! ь1731.33 11а,ге .1.к....( г(.ь•ьимсо ги..сс.. ь.rииr .ь.и( 11.1.re.� ь!еl1.11 .11 1ьеl1,l3 1l1.ь1 !1lь1.11 1l1,!! 1(!!!,1( 101.11 1г.lьг.ьl lи.ь(1 ге.в[Си L.ru+t г.ьvа г[г1.1,. ь!ьl1,l1 lt и.11,i! t1l,11 ы1гl;и 13,!! 3с131,1! 13,tr гtн1.1.+ r(и11,1ггм 1..и.г иьwе „ •ее..и!! 31lь3в,и t11 г[l111,l1 .l,1в 1111ь3е1t 71г(3е 3+1t1l,t1 111ьl1 lьепвиС( .[г гиеьlен сгепвисt lьииlне! rвw ьвиlеиьlи.г 1 .11 ! ь.�1[ .ыь•,е1 •11 ! . в7l,11 1!. � •г .ви.ы1 ьо•ьне вы.ые 1пь(г 1r ьопьсе гсеnинltма гlие[ииг(! 111l1 .• rC� гг1 1ьl.Св r/1 .7lг,1Т 1,1! •1111,31 1г, 1 f 1в ввС.w1 ьоиввС ьССУиМ/1Ис 1гL,нС[ ГУв иь17[ !.!1 11 •11 в[в1,11 11/,t1 (311,1! 1l3,l1 1!1!,!1 110,г3 1 1в4.w •УмвиС ьС[UцнгlИс lвlвМС[ 1Пв GдW,nMU1/ !и.l1 .1l1.t1 ,11 .1ь1,11 /1,в! ь11t1! 1(1.3е в11,1t 11l,11 +вс.м( ьоиьыс ысиииllис l.ьисе rn• ы1Сt iiti3 7..1.г1 .11 3..3.11 11l,о1 1111,ьг 11l.1f 111f.1e 11(,ы в.и.с1 ьУ•.ы ас[иинllиа [•ььнс[ r„в cotlu« J[31ь,г1 ь11 1re11,1f (!1,lе 11111 11 1!l.1е t3131.fe 1/t,t1 вьь.ги ,и..вс .Ссиипl4иУ lьььисt rов lalreto 1lиыив![ 171е1,11 ,1t 11111,11 11l,t1 11111�11 111111 111гl,т'1 13з,1! � r•C.sw..e ьап.и[ •ееиинllме [ьевиеt го! •(с 3нвь[пи[ае■ с•ив1 1г ,t1 Сьепlьо. ,11 г .г ев[n.w, е�пlМпеИ1 Сьlи 1еи. (нfгl ггин .ев(eueu.t (1l1ь .н в[! гв1 1t•.п! .С1331,11 .11 .1[1l1,1t [1г1! .lУ1з;11 11,1[ .l113,l1 ь,l1 ыь.,.ь curr еь7и г(и. rипь •ьгке •ьи.сlгИ! •[r.e, 3113,1! ,11 11гl,д П1,1t ч1l,lt 133г11 � и11,11 и1.fв � у ,г..ии еигl еьди reu. пии wаинониll ..гвгllае ![..lt 1l17,1! ,11 17ь1,11 1ь.11 1eI[,1! 11е,71 77.f,11 111,11 .ев•и( CW1 Св1и reU, 1иг1и xIC[ •вг.tllм: 1l1.С1 l111,tA •11 11[1,11 (11г11 7111,11 1071в1 )[1[,е1 103,1i •(в.С1 .YeI Сь1м /141в гигN C411'1И гJ,.[г1У4 � [С1.17 1111t,f! ,11 t1111,1! 111ге1 fГ11ь,11 111,А! 1вА(1,11 11t.11 г не••ги .и.1 Сь1м геи, гиии 1.etLнuLO[+ гевьССо иь+•екlне цг.r[!t 1и3,1! 1t е11г,1е 1f1,1e tt11,11 r11t1 .11! ,1fк3.1t t иС.••еиг • 11 Сь3и г(.1м rrпи 3ньип.и es! ггпrlекц[ -ьа,ьlеье r1tг.ьаьеlе .i[3l:,l1 ;11 •i1ьк3,11 !l,1t �tt[(:[[ .ьf,t1 113l1,11 .11[ь1t � ие..ьо.ьие 'roelt ьон.lе Се1и rкии lнум 1•ч tиП,егв .ьн.lгlче eCilrt11C7 1t1.1;J 7�1.С7 t 31�( 1 77.:7 11 17.31 !31{rC1 1 11 1 �4,7� :u.t Сь1и гlU. гв:и ty^!tt �11(1 ".М t1t�1( 7-С1н E{ .)' �1 .'�1jг{ 11 �Э 71.• �: г1.i1 (1lt, ,е1 1 [{ .1 :.r' :I �, i;, а••� �.•�г 1 t1rc3 '1�е !1 �, `, i. . �!, ` '1'в1� ��1 „!� '1 �� �i� ��ft .�i: 1lеСв'�t �••it � +i1 �� ':1� I7. �:' �мы� �1�1 11 -i .�. ... гь1ГГ1 м �•1.1ьь •!e►i,1o ,t1 -1!tl.It 11ii1 11(111� {�.I� {t171.l1 •11{11j ,t�.i..,, га}t �13.� 11�• гп.:л i-,�..,..1..iь ii.1., зS 1..�- �., 1l1г.,. �мсипtl /1t •. ь..иИ•ь+ lеьь ь•w��ь1 ьгь. и�, кtrw • г.м 1,ь1,пм.,унь . м...п1 ' tме..1.., 1г1,в3 ,11 tr1,u 111.е1 Ч,1! 11.и 131,![ 11,11 1ме.,t.., ииС Им п е..ц с.г ,ьi7[.сl.,.И�ми1 1.ы (�иьеспi,и 1мC.nt.At 131.1в ,11 111,11 111.11 �1ь,1[ l1,11 113,31 1t,13 tие-,е.иt ыы пи ti•Cn•[ и1 +117t•!1е[ 1иь..ппьаеиl IN[ви2в[I I11,l1 ,11 111,11 111,1[ 11i,11 [1,1А 111ь11 1е,l3 1MC.wt•С) ll1L пм (ИC4rI а1 2ьI2l.S8I1uM 1и1L,nhlulв! Uе.иг.га 337111 ,11 3�),11 1[1.(в 1!! 11 11,l1 11l,11 1l,е! 1MC.,t.1u .сь( пи (Ис.зм( иг м,1t(.еlУинониг.епь.ееа lnвипиьоь!! . 1ие.иьlеи 1дl1.!! ,[7 13lt,13 1!,!! 1171�[! I11ь1! 13[1,1! 11l,е! 1ис.w..п ы.У и+иьw пп 1ись.! . 1мС.аи e4t1.f1 ,^t н!•,1А !. ![ 11(7 f1 1lе,l1 .г31 11 11l,1! tMC.lи ы.[ tыо,е ur 3ьи•! еllьге[ 11е1! ..е 1иС.1[ (1lв3,I1 .11 13l1l,w ев;lе lta11;11 1!!,в! 11fв7�11 111,1! 1MC.1t ltь4 [Испн nr 1t. lsгьlг! <1!!1 . 1 (:, tнC.r1 1е71,[t .13 1013,А1 11,ь1 1!1!,!1 1!!,1! 111в,11 11в,1! 1,Сиt +1•ь lиспп( ar геие гwео 1o1.ero i11в1е1 сlаа! ..1 yu[./t ц}Гt .11 и,13 1[,![ 1[.tC иl,и l3ь1[ 1ы,1[ 1иг..1 !ие lиеn•е w w•ttr lпеыеп 1lл1(! 11!!в ..1 у Г ы[�!у lв11еl.1! .�i а!lв3А,{1 1е1.l1 !1вlotji/ t!1!!в ьеаil►,Ав tltьвl •r .еиьи( .!г пеАгн[ r,.r.1г7' �лrгsty ♦;гtсVЧ4 �1 � � г..г:а.! пг:.• �.�.п rr.r в.вС I.мrir F - е г 10�Ь Г.(а л• ,Од Г•. .м (+hи г' АГ j.Iв<Г i.L1�TY ГtLi:;,.�le+ CL.,S, гIl13LS1.1 21ц�е rEfГ:+�Tt � i'�Сс:а w7 rwaery РАе�ту Мтtе 5+ts^t a;ai^.ги..г[� и0 о - Е^с-Е ,_� 3АыгьУ1 Ас[.,ий ,, RJ�J ci>,.п w�,.�ro еLеУ.п жм.г+вn 1iь~п гил++.та1 е1+.+иС nва.,..гтд сеиьiьwь .11 •) tгиычии♦ tви4(»пиь •»У sи(Св вввв.:(Е-[ ь( .11 � •ви»виtьlв .11 .г ваиУ•и1[е3 ввиииС[в retcl7 1•. в[в м11 -^1 tle.11 .1! 11t,1• 1ьС,(ь t11,11 1ь1,ьв 11г,е1 11s,i1 117,1в 17,[о , rr..L гмио„С!в ..1[t (иа ••1!е iг.ил 5в1,с1 .11 1е,ее 1((.27 .11,et 111.1. 1[1,1ь 1е7.•1 fL/,ot 11f.ь7 rr.c, rиии4с[е r.1Ct 1w 4aouw,ull а: r...1 17е,1ь ,11 l71,[1 Iве.гь ()1.1ь 11а,в1 7/1,11 1ге,17 111,1ь 11е,г1 rr.e1 .ниииг(С .в1СL 1ив вlсе гв•CI в1е.г1 в11 1,7,1! t11.f1 5вlв11 171,71 ,17,11 (lв,11 вьг,lь 171,11 „ .в•С1 ввJи4Сlь гв1[( /Ua CurlVч вв.1и 11в,7i .11 1••t,ьь t•У.1г 1,в[,ее 1ае,1[ 1111,11 111.13 1Itl,eь !11•1ь ..1и Ииии[lи relc[ (ии lмвltииьп[в г.;в.сСУ v .вl.( lit,11 ,LI 1.7,в1 1е1,1г •11,1ь 111,7е 1l1,е1 (11,17 вг{,1ь I11,11 вв•1t,I lнириС(И вУ1[е 1Uв 3ивсьпlНСгм •е•rtC1LE, вг.ыLа1 (!в,]0 .11 •7•,11 t1e,1C ,7ев11 111.е1 1гь,11 111,11 •г1,ь( 111,11 „ rr.l/Lг1 в•иииСlи вв1Сl 1ии (]1.11 1lьг1Е11(ь ,1f Cuw]в4!(е] ,11 ь. CUч1.N1G[е [п»+ич(я ве1Сl1 1•4 'еи иг1 +. Сг..( 15..1в ,11 11•.11 1[1.(1 111,11 1е1,fь 171,1( 111,11 11е,10 1ь1,11 Cr..l Синаии(Ч вв1Сt (ии w 1(! Сг•4н 1•е,ье .11 ье(,10 101.Оь ьа,,11 гь1,1ь 1•1,lь 11о.11 1вв,l1 101,11 'r.w [ичfииtв r.1Ct (Ув ССОии04иг1 er•»1 115,ь1 .11 115,11 1ь1,еь [1f.1/ 111.ов 1tt,11 t11.1ь 31t./3 111,1г iлмl lинСи-еИ ra1CI /w вlее � а.Сwгм 1е.а„в .{1 г.и,n 101,01 l.н.1ь ив.От 1и1.а1 1а3,11 tаlг,еl Ч1,1а _ cмcwrn сии,Ум[и ralee 1и. српп» сааи Св•1и .Уа.1ь .11 .1в,11 1о1,[1 .11,1е 11ь,10 (ь1вь1 tее,Ч (и,11 111,Е1 cr.[и сииьииtи rиlct (ии [ие1• • 1и...1с[ .1i .л roarнtcE rиь v.1c(] {.. а(• мt1 (и..м1 1.1.а. ,11 1.а,11 111,аь /11.ьв (11.1о tC1,1o 1еь,1ь 1еь,ь1 10ь,11 (о..мt ruo .и1сt rue ..1ге ьиие» 11в.еь ,1t 1и,1ь 1аввОв 111,ь1 101,1е 11е,оь 111.а1 11г,11 1Е1,11 ep..4r 7ие ввСС[ rua еь0ии0»urt (Уе•.( 1е0,1о ,11 11f,ь1 1Y1,i/ 511,ьь 1ь(,1в 111,11 111,1е 111,1ь 1ве,11 1ивь.[ 1w ви1[[ 1ив а1с[ /Ue.c1 •1е,1е ,11 01,01 1a1,i0 ььь,Оь 111.11 ь11,[1 11в,11 е10,01 l11,11 1ов^С( 1ие ввlс( J11в ]llDC11ITU, гив•tJ ((ООв0е ,11 l1ь1,t1 10l,00 2/10•1в 101,1ь t111,1t 111,10 l111,ве 111,11 1Ув•IU 1ие в41С( 1Uи 1чlСlииt0(И 1иввССJ 1С,••еи 11о,1а ,11 711,е1 1ее,оь 1fe,eo 11о,1ь 1f0,oe t11,o1 ffl,lь 1ье,1в 1ив•1У 1ив г»1([ {ь .•,i0uee» .У1[е1 In• L7,•tC (ие.Ч а (и..1Е 1вУь,1в ,1t 1вао,lf 1оь,ье 1и1(,11 1ьв,11 1111,1( 1ьь,(1 1•11,1г 101,1Е 1пв.гЕ (иа .•1с[ 1ь •+cpute+ .е1п1 roe (]гь11 г(. т (и.•1( 1еи0,]1 .t1 1в11,1ь 11ь.11 1ьев,гl 111.ь1 1еы ,Ч 11(,ьг 11ь1,ьо 1ь1,1ь ru.•Lt rue ваlс( 1ь rииоисtи .в1Се1 1ии С1(в1( rcu[.си.ао гп.вии 3Ув•0( 1t1e.1o .11 1воь.10 111,11 (е1Е,ьь 11е,01 1о11,Сг 1ев.10 111ь,1( 11ь,1� (да.в1 ьив IвIC[ 1г гаи0иСси вм1СС1 1пв L37aft Еигьl( гиьвССУ а С1r.в1г[ ,11 •• С1rг• [[1.иt 11е,Уе 1[Е C3I вв1С[] 1в. в1п »f1 ,1l 1ав,ьв ft1,/1 111.ав 1аС,1в )ео,(1 11в,е1 41в,1е 111,1! С11.м1 C1I в,1[е 1оа .в1(! [11-е, г10,ьь ,11 1а1,гв 11ь,11 10е,11 111.ь1 /11.11 1ь1,(1 Че,ь1 11о,ег <rr.i» [1r вв1Сt rua ееиин0»игг _ С71.•1 ееl,гв .11 аее,ьl 1о1,ьь •11.11 1ьо,С1 иь.ьь 1вь,оь е1е,1ь 1ь1,ьь егr.м[ c1r веlее гии .teE Сtг•с1 1ьае,гв .tL 1ьвв,аь 111,00 (1ье,е1 111.ьв (111,1ь 1е1,еь 1Ч1,Ее 11ь,1ь С11.сг c1r ••1с( (п. спг]и» С17•гсиг 7]в,Ч .11 н1,ьо 101в1г 77ьвьг 1о1,оь 11е,1• 1ьь,аь 1f1,/1 иг11в Си •гси( Clr гиlс[ гие 1(и1[Е1:[■ 0 ,11 , ьlиlвва .ьл1ь ,11 ев,11 111вОь •ев11 11(ь1о \1,/] 11ь,10 ♦(,ье 1е1,11 1lИlвьа lдьlьира0с. (ЕМ11с1[CR ]иГис1 С11и1I•вlиг I10Y1 и( 1еи (в1 е ,[1 [ии.wгl ltвеь ,11 13,вь 1ьо.1( 17.Сь 1оь.1ь tf,1C 1ь1,1ь ts.гг 111.1ь 1о•аwгь lниемаери гое«ео и...Е11не Оио[е 1lьоь и( геа Тьf г ,11 •виw(ав• ,11 в{Уи1{ви в0ивьС ,в17[ (.ОиаС ЧиПгь 111Y0 »Г в[и Ie1 авУи2l,У .11 ввУмСlми в0»виС Ив1!! lи/иа7 OUO[ь (1000 мi в(Ч ТМ1 с .п L»рисtмwв ,11 .) lЧиУс[нУУа nUOac ьС]4с13 IauuCWe в11 •) нии0СУ»ь /ИУУУС110» Сон7иnг((и» вл0 и.а.(.Т СlСаьlЧс {101е мI Гtе ,и УьlЕ17 11(G1/1101 о 1 .и alll .11 •s ив11с и•![[ 11 ,11 • и».мрси .15,11 .1( .и,.1 ,7.Ct и1.\1 ы,1� ,11,ь1 и,s3 е7а,ее н,и .+исы w.еь[1e1t[о •.1С[ м.ььмоепеи renm,ti[аИ м-Сtиа 1l7е,(• .11 17•е.[t 1а[,7о (110,11 11е,1г 1 и1.71 (ь1..1 Li17,le 10t,ы u,.rlLtu и»геиllЕlt[о ,.1« ]ивttмее0(. гlСсп (.е гtы и. rEa .е7 Уи.•всв 70•.to .1t еЧ,1f е1,11 е5е.11 ,г,•е Ч7,г• е7,,1 е•t.fa 11.е0 в»е• uи11M1llttto »altl (нев 11о11 ив! О Униввв lьвьff .11 гвl,lв 1or,1. г1е,г1 1ьь.11 1о1,1г 1и1.fв 1е(,и и,(t v.:u., и,(сиlсьlге0 •.1tC tвеое иг[ г»в»да,е .и и. .[И ,и1 i•••pe» 17в.17 .ii t)в,,, ef.11 1•t.ft в1.1( [1(о13 .e,ts гье,11 ье,7У rи.rper (Си11ь1t[о wue •.1tC ]ч.аЕми.о(. .a�puertoa Q r,.rLalY 11и.7в .1i I,(L,a/ 111.1• 1ги,е1 11\,1• 1111.71 1ь1,19 1ггf,еl 1аг,е[ 1^.,1elu r[иLtll7LO еиевг н.1« f•в[tмцо[и ,1еер г•е •еи ма нв ,и3 /w.•вlа 1в7.аа ,1t {е(,1ь .1.1f 1е7.г+ г1.\7 1г1,1• ег,г1 111,аа ьf,1в r,.в.ев 1е»1гЕСг[р weвl ив1« ]ив[ениеоев (е[ь 1110о ,.1 ьw.lв.и 1ее,и ,11 г•С,11 1е7.13 дьS.ьУ ы ь.г. г11,1в lг1,Ч гг1.77 17,7• (ивьвае 1[н1IЕ11[о Еигвь иьttе (•веит 01и ,,.пи wc 1»виовгь r[в ,•(в r.1 С 7-в1Си1 ьlвье .11 .5,33 11г.С1 еlве• 1s.1ь гв,1• 71,f/ 7ь,s7 1(.1г 1м•ьеи( rеиг[ь1е[о еие.ь и11![ 1•весиисо(и а[нгll(« а wC 1.е r!в н: rCe гь1 1:..оеи 7.7,в( ;i1 г31,.в ,•.и l1е,в1 11,1t гг1,1t ,о,17 7ы,1е 1г,ft 1м•аоеи 1м••р•ер ,•t« ]ивеv.цреЧ гиопиеllпи О 1w.rlelp 1вl•,!в ,11 12•в,а• 111,[• 1J7ь,1t {1•.е• 11/е.1е 111,(о 11ье,1У 111,73 1и.r1lLY lиr,.lreo w.1tC 1»ве(.миЕ0tи r1tCp 1вг. r[м мв rlи ,•1 1м.ввt• 11,16 .1l (С,11 ее.в1 еьв13 17.1t (г,10 1в,1а в7,13 11,1ь 1в•и!а 1wr»or[0 мв1« 1•,lL^UlOla lвtь {1в.в »в1 (,•{веи 7ге,,7 ,11 1ве,11 111.ав 7Се.(е 11•,.1 113,11 t11.•1 11.,,• 1,,ге 1и..•в• 1ив.и,l0 гвt!! lивегr,цОlа ыыlя ь/ i»вчОвгl в(в w• rle ,в( О 1».rtи1 711,в1 .11 г••,С1 1ь7.а1 (е(,1[ гг.ег t•е,ьг 11,17 17е,ь• 17,е( 1и•геи( 1м•ы-•3р ,.11е (Ивl,емцоеи rеигlсгlга wC 1.е гев и. гtи re1 ;� 7..uGw 1в,1,7в ,11 1в5в,ав в7.lt 11г7.в1 в7,1ь 1в1),1е ••,1е 1371,г7 11,11 ,г•гиСИ .вltC fи,LсмаЕсаи гrиоисгlии о н(.(«еУ t51в,[е ,11 1•.в,е3 17(,11 1в17,11 117.tr tf1•,1ь 11(,ь! 15fe,l1 101,11 .t•(lеьо и.l2С ]и,Еемрlое► ь1lео 1ве в(. ив в[и r■1 »1••»!♦ ,Iв,,t .11 ввl,lе .1.,1 1(1,71 ге,г1 771,{• Со,1ь ге1,3! 71,1f мг•в.ав мв1![ ]ивьеnКОСь (+ев ct0a1 ив1 . ,ав..е. г11,ьь .1! гг1,1. 107,3в (11.ь3 117.t7 «e,tt 11(,17 LIl.1S 11,1, .г•авае .в1:е ]мвьм�и4(ы еим и1[ 1»1^сl,7 r[• иь ге. (в1 1р и[-.Св7 1е... .11 f7,.в IYe.1f (е.ь1 1С.1• 17,Сь Г1,ее 33,е[ \1,1• в(ь1(иг ..1г[ 1^аСениесlУ ь[+rtule. ие[ 1r0 •[А иь rea rи1 И(••еаь 17г.в11 .11 1е11,11 117,ы 1lеf,гв 111.г. 1u1,t3 1ьг.1е 1г•е,11 11l.и .г•ьlи. •в11( ]ивььмиеиеь ео,]и.в[tоЧ иС•ucw• г1,ь1 .11 ]1.11 1о1,0е 11.1! 111,Оь 71,и 1ьг,tо ьв,1г С\,(1 »7.ulwt ..((с нев, сwьигв7lпн О иС-вив.Уиь 11е.ь1 .11 a7,lS 1в.11 7ь,7е гв,,• 111,г1 11,17 1l,73 11,1J .1-[навьиа гвl/l 7uarluf в0а[D 1о гtовlС[ и7-ееь[Си1 17..11 .11 17в,1а 1(.1в 1ва,гг .г.1( г37.ft ов,11 [е1,1• 7ь.3ь •г.«v пь ,.1 «/ивмве ,.в.[г ](ь( 3иьг[etL•e. •(г[о {.[исеигl ! .п О w1в ,11 .3 .»иг1 сшичо4иге е...аеи (1.1. ,71 ее,г. 111.1• 11.(. 1t3.•e 11,и 1вl,fе г1.1. 1н,\1 ..аеь Gиии»o,u1f lивlемь,оLи .ЧОоиеltон ен.(1tеУ ....ег .11 1u,se е.,а. 1ее,11 17.t7 е1,,01 н,а 1\1в1е 17в13 .г«Еи ььиинрииlf ]м.Сь�и.Оьи ,1lе0 !.е .ее мь .(в (•1 р е^в.аь1 1(1.11 .11 !г•,/I 127,1ь 1гl,71 1г1,11 111,•7 1е7,1ь l71,1е 11е,1! в.tв еьпимо,иlг ]»цЕниеоеЧ в.1• 1101а »•t 6».ь.•е 131.t1 .t1 77,,l1 101.11 17ь,1ь 111,l1 71•.е( /вь,.7 t.•,t1 1ь,(С си+иввв сиии.о,и17 3•.имиоtь l.мии ]( 1Ч,»Овг] в(• »в rl• ,и1 е,.,г»1 1в.4♦ .1f 7•.17 101,11 г1,11 11.13 (1.ве 71.1е 1v.ьв •1,11 .,-(tи1 сУJичОнУ1] ]ивымпсСtи 1lиrteltte и1( l•е гtи нв r[а (е1 ! с,..(р»е в1,51 ,1l в1,е1 •С.l1 (1в1, 11,,в вгв11 е.,11 (1в11 Се,11 :4••CU41 сиии,оwиll 7ч,ььмУlоlв CYчl�nrflU» G,.иСиие .)l ,1{ ,1! 100,(1 ,11 I1e 1в ,1l 111,11 ,11 1г1,I1 :»-иСиЧв LNUU,OнUr] иаивь CGЧ4мO1lW G».all 1в,1в ,11 в5,1ь 111,1[ (1,1• 111вI9 1е,3в 10•,1ь .в,е• 111,11 Gx.fв+ GYUU»Овиt] [fIUMr] о . .и slta ,{1 ' •3 .1ьс ;tct аl•вре» ..... ,11 и,11 11.,!\ и.71 113вц 7ь..1 tff,l. s1,.• 117.ь. е1•в0и 1ег ]•виниаев r,cuuerlw О ■1•rLaw 731е,вг .1t 1.1[.11 .,.11 1а0г.1ь и,l• 1ь•1.1ь [о,и 1•t3.11 и,•г .1•rlecu .1еС ]+цtмхееи r1[ь0 1+е r!и м.(0 .и7 е1•в•1в г1.1ь ,11 .У,11 1,г,•1 .1.11 1ег,•• 11,1о 1аг,1. 7s,в! 1e1,1a 11•виеа Ntee ],пеииt0(ь в.tа ltaeo ив1 • е1-.вав • .и,11 .11 1•,,1о 11t.\1 f.1.s1 11г.1s 5вt,1ь 1е1,е1 еев,fl е1,г1 +1•еввь и1(е ]nвеtнlЧ.о[и гвеw w( ги•ио.вf rев на rеь ги1 ,' ■1-.(м7 .,,7ь .11 ..t .(,0. ♦г,г1 е[.1r о7,и 1.,11 rв,11 11,\f в1•(еа( .1et ]•аtьнце[а re.rtE1« . w[ г.е re11 н1 rеи ra1 в1•,ео»• 71.,е ,11 t/:71 ,1.1► 77.71 7г,11 t1,в1 е4.11 11.,1 е7,11 У1..еине •1[t lмвеьниьоеи еои]ииrгlии ,1.иfинl .,1, ,11 •,7, 1е1,7в ь,1• 111,Оь ь,1в 1еь,[0 в,вг 111,1J r1•исинг r1[t иаев» сУнfиивгlим , � и1•Gas 1l,t] ,11 71,ьь 711.11 7о,tь 1ft.1C СО,оь {1tв!} tl,71 1(7,{! в1•е+г и1Сt llгoMrE L .1t 3atuCptlW •11 .ь 11t0[UIIUи Jl[0 COI1n» 1 е1..иеИ г1.17 .11 !],7. 11t,,0 г7.1е 111.•в гг.С1 1оь,7г 1г.7• 11г,ьt С(•.ои «Еи eorlw ]неееиоьиен r.uoue7to» CALIEW ч1в,7t .11 .(1,11 L.,7s че,г1 1о и ....tь e1.7f 3ео,Ч 11,!( [t•nnY 1ет сегтИ fnвеtииьоеи rueo ае .Си ». г(м гм1 � Ct.в.le ,аг .11 77,1ь 1(1,31 1(.9е 1С1:l1 f1,7e 117,,. fв,1Р 1а1,ь1 Сt.в,ев ]СЕУ ealtu, ]»вгеиае0еи вьеl 11nОв .а1 .7 С1-.авУ te7,1, •11 1в/,11 10ь,г/ 1.е,11 11ь,lf 11г,с1 1о7,1ь 1в1ве1 С1,77 CI•lввв ]tL0 Сд11ии lивLLNItOIЧ lвьов и]С (»ьвОв/1 1/1 ив в(в l■1 С1•.!и( 11.1г ,t1 12,e1 10.,гь 1•.N 1t.51 гs.ч 1е,7ь tt.3в (1,1• С1•rtмr 1е[0 ealtuN vцlмиьеtв ([М(п[t[в u3t t.e гев ив в(• rи1 [1•.CUwi 1У,еа ,71 11,(1 111в1! 11,71 117,7в 1i,ьУ 1в7,.1 1t,73 е7,,, С1•+[и»в Со1Год ClUrн ]ивгьмсь0[» Си•]uwrtr0» 1][[0 Си71и4 l0иlвгеlы j J [t.ucu»ь 1У,7в .,11 17,7в 11ь,11 ti.1e t1С,Ов 1е,5в 111,1е 1г.1ь 111,11 С1.иСии► Сиггон сlиlы и•ев4 Сиьfииrгlрч 17t[n СОi1оы 10и1.в11и11 С1-е(в •1,,1 ,11 -,1• tв1• •,г1 1ь,ь1 ь,1ь 11,1ь ,Ii •в,11 � CI•С1/ 1С[У Go11U» [1гии71 ,11 О 3ис.Ч .11 .в 1и..И еисви 7u..OCw lгеве5 ,11 1lУ,в4 111,ьь 1гв.в1 111в1в 1гь,е3 11e.1t !l1,ь7 111,11 еи.•оС, 3иС1• С]гьгl ввиоиСгlои ги••СУ»• г1,в♦ ,1I 7.,71 ,е,1в 11,11 1е,lг гь,1ь еС,11 гь,г7 11,'i 1р•иСUиг 0Uсве ]д.ььмОtсси СО»}Умв11о» . О [и.ос.1н• гь.5. ,1t гв,s. ьое.вг lд.fе 111,11 7г,fе 111.11 г1,fs 11ь,17 1u.ueuw ]иеев и.м, совlии.7[ии 1У.свг 11.•t в{1 Т1,вl 1е1,г1 7(,ь1 11ь,гг 71,77 111в(• 11,1• 11е,1С 1u.ler ]иеед евеиигl О iE. ;j� •г 1!. г[. «•.и[, 71,г1 .11 31,11 1е1,1е f3,11 1го.1ь 73,7[ I3ь.1е 11,11 1ег.1г 1е•.иеИ п3 С(1е(С r.pouerlo» ио elru•r] р .11 f4авССо .11 •, 1ое«си 1и.•сс0 (и..ае, t3.31 ,11 1,.1. 10,.гs 1..[1 117.г• 1\.71 1еС,1г 1..1в 117,11 го.гОеИ 1мц1,Wе0tи гоlвеео мвиоие110, 1и.,(е1.и 1а,.1в ,1( ..1,11 ...11 1ц.11 е\,r7 а•е,7( ,г,11 .71.е3 (г,ь1 1о•,1еЕи 1•..иаЕ0еи 1пе.еео (tеьр 1.е иеи ,в г[и rи1 О 1и.в»ел 1е.5в ,1l 73,11 1ее,11 17,е( 111,(1 1ь,S3 11в,о5 11.ь1 171.l7 ги,в•ав ]мвееноlОt• (пьвееи вЧев (110У и•1 1и.Св.. .ал.15 ,11 11L,S1 107,s1 7ее.17 1С7,77 С,1.11 1егв11 в11,11 ев,1] tu•ав•и ьивьЕмцо[а гиваг[о ь иnв и]l 1и(.овrf re. ив .еи ги1 1и-вьиl 111.17 .11 111,еь 100,г7 1•ь,ь1 11,71 1]е,1е ,7,(а t1/,Jг 1L,1f 1о.(еи1 Ч.[снц0(и 111ввССU 1(и1Iс17(+ и]е 1ве вее ив гlл (а} а 1и-еtв. 1.11 .11 t,и •e.te 1.1о e1.1s 1.1. 11,.• (.3е .1,1е 1ииl.в м.Еииеоtа lрьвеер Ееа..ее ги tв.п.г м..■е11 tu.rвa 1е,Уе ,11 1s,ев 11ь.1. 17.11 11\,ьв 15.11 11•,ь. 17,1ь 11ь.1е 1о•.ИУе 1и.ьсма0(в гиь.еlо ввпеиыtо Уг .р„ее 1u.w 11.n ,и 1l,п 1У•,е• u,n и•.ье и,(t 1л,л и,н u.,w 1а•е.г мвеено\оtв 1ов.еегl египип0 1г в0»(ве .rпи еиенне ® ii-ьoc, t5./s :ii Н,13 1•г.1в ef.11 1вв,11 1f.tf 1в1,11 17,13 111,11 1(-гое, 7сиа-[иИ(0 [гг.1е rw.eeo г.поисгtан .»о ((rа7в е1-г0[н .•.sь .11 1е,3ь 1•1.1ь ев.11 tг(.1е н,7• 11ь.1ь 1.,3е 111,еь ь1..Оеи еиеееl e3r.1t га3вееа ььиоиегюИ (»е еrгпьl] � е . /Еи1/е1{еи ,11 •+ 1С»rtel/lи lСигtе[1[в /1и1.• .в.ье .11 •1в1ь 111в11 вl.17 11.ев Iг,() ••,1ь 1в,71 lг,гС (lи1.в 1»вы rцele ьсиг(L(ttа вивСмв]lt ® reиl.r« 17.3е .11 1.,гв .г.Св 11.г( (3.•ь t1,1f 13,1, 1.гг а,ее 1[а1•r, 1,.емиеои ([и(1е[геи и][р пн r1и[(ь17'р еоец иь1« 1l+гвiи 7евв, •11 1а,в• 11,1в 1i,11 \1,1в I1,7[ (1,17 11ве• 11,гь 11и1.1и Сnьььна.ОСа 1IMLIl1Llи ulLU Пи 1ввв�,'.: ив13L 1А,1.м( 1f.1. ,11 11,1. С[.С7 21..1 11,•е гf.•3 11,ь• 7в,1• 1[,1ь r[г1.иt lивtсицОl• rtвtlElLea р](0 пи •в1(( ® 1lиt.ь. 7,ь• ,1( •,1ь 111.1) 1,71 117.ь1 ..(! 1г.11 1,1t С1вС• 1[+t.Ом гивеьииьпlll /t•r1Lltlи uffU пи еаои4о»иг1 11и1•в1 lвlв ,11 1,а1 11\.f• 1.1е 11е,17 гв10 1l7,1'7 г,•С 1lь,ьь , 1(н1.м1 lмвlьиц0lи I[•/1L11tи и]l0 ои в1С[ 11и1.С1 1вlв ,11 1в1е 1е1в1S 1а11 11в,г• 1,eJ 111,1• 1,11 11г,11 1(и1.Сг 1•вььнОьоее 1Ca111111M и]t0 ои СОТfпн д ьtиl-(р 5.!• ,11 5,7r 111.ь1 3,fe 11f,•1 l.17 113.11 1,1f 111,е( Ie•1•to ],аы иц01и 1[иг1Е1[(в и]Со пн (11ав[Си 1lвlttьa 1o,f1 ,11 e,Lf 1е,1С .!7 i,CI ,37 1,е1 ,3f 1,1г . 1Сиtt(,м lLи11e1t[и Еfьв(0 lиии f»веtнц0еьь !п е]7а1tг Ф [i(и1.о •,С1 ,11 1•,е1 1ее.01 11.01 111.ьь 1е,11 10в,1о 10,г1 1г1,11 t1аиl.а е]tCrE /laLllttlи ol»ьчо ' ь[(и1.еи iе,.в ,11 1е,•в геl,вь 11.еь 101,10 l/.ae 111,11 1в,в1 (а1,1( „ Чtиг•1и Eftere Lt•(1Е17[1( ицо 0» (иtаи lLlи/.tt 1г.1в .I1 17,10 111,Ов 17,11 1о1,е1 1i,1ь 111,11 1г.1о 1е1,11 l1ьи1.1t t]tall 1lи11Еt11в и]t0 ии г(в , 7i Е1tи1•I1 7а,tг ,11 1•.!! 1•1.1е ге,гг 111,оь 10.г( 111,1ь г.,l1 11(,и 11еиI••г [flьге ree[tLltla и1(д он 1tuL.CVи[п гпевСеп 1/lв1•►1 11,i1 ,11 I1,11 111в1е 13,г1 1/1,11 J1вгl 111,1E S7,ь[ 111,11 l/lЧ1•11 C3t11t 1lнtlLl![а У1l0 иИ ьиььlТ Io1,Cto i' ,I1 ® � .11 Е(в. � 1•• .1! .г l.r.lни lan„t1 t•luali а»о оИ[1еи С.епв,е( 11в1ь ., ве ( ' »(.а1•, ц ,11 •(•llrr.и мв1« (..и.( ,.еи! У iw•ав...l 1/в1•,Т7 в11 14в(в,,в 117,а1 3е11ь,11 111.1в 1s110в11 1Dе,{ь 1геlь,ьв 11•,(С .lвrvlc еи11и»Очиг1 llгпиl УвlиС { в(.е.в,вt \17в,1r ,1l 1l5/t,L1 l11,11 151.гв11 111,и 11.1,,11 1(1,l7 111•!,1• 1lь,fг •1.1..,вс ы[е lвsриг .вtи( е1•e.rr.l •(,l.,1( ,11 •1\•.11 (.fв .1е1,1а 1ь,1е .ее7,ь{ 1,,11 111,// .l,11 Ct•lгг,ве еи11оИ L.ru•( ,1ьиl ® ги^(УР,вI 1]Ува,1i ,11 7,I11,вГ IY1,11 l511{,7, 1l1вг0 Jl111,10 (ее,1• 111в1,71 (11,[е 14вLвв•1с 1Уьви 1/вииl ввеУl ra.fв.sвt •111е,1о .11 а1lви,lв 1У1,1. 11ьв,11 1о1,11 (111в,7• 11в,Ф1 1(1ь1,1• 111,11 11•lвг,аL !!( Евгивl .ваиl tu.e.rT.e 1111в.в. ,11 9ев7f,Се 11в.1• 1еег7,11 11ьв1• 11е7l.1ь 1е•,г• 1eelf,ll 11\,г• 1e.t•вы е 1•веьиwп(• lиьвеnl l.ru•1 в uut ® i1•агв,в4 .f111,77 ,11 a5l11,tJ 1ь1,ь7 вь731,гг 11ь.11 (fе11в1! 11в,11 1f711,7t 111,11 l1•l+е,аи 1tua.cu•(р lиевсси [[ви»[ rLUUI ы•tдв,ц ,н.1.ь! ,11 11вл1,ь! 1ае,ьв ь1111.17 1ее.11 [.111.1г 1еИ,1е г1ь1[,17 11г,11 .1.ев.,.е аиЧее7 1пе«еи (вгп.1 ч.еиt �' [Еи1.1»r .5•а5,,1 ,[( .•ll5,le 101,1ь егв•1,11 ,l,гв г1г1,11 el,t1 L11t1,31 е1,(! 11и1.1•г (си11t1г[и 1,.иег ив.ие й tьеммее 7иы.,1г .и !.ь].в,71 1н,7• l..7vл( пе.л 7йи7,г1 w/,u !виl7,13 11(,гв иемв4са н(л 1аС1еИ е]сн.чсе имlнw r.p, ,оЧеиУгиие iLU»се ;ii • ввьиь! ш»„е иЕаисt ь^[и в( •ом.ее {eenu»rh� маеwивсь ивеl еи .еи (и1 . ьаt.»1 .ьовв.ll ,1{ .вl11,11 5в,1в .•111,11 f1,72 .Ив1,7в 13,11 .711вв• 1,13 ввЕ.мС 1Уивас 1ССОw(1ыс ь1ь(»С( !Uв ч.1lЕ ♦ евевеw ♦!lf,11 ,1I 71в1,S] 51в7в 11в,,11 (С,еь 5ееl,lе 11,11 а171,(г 1\,11 в1\ь4» в0ивиС /CCYU,1IиG ьaYawCt 1ои СУиинОwиГ] ьв4.•1 .1i1,1I ,11 .,.7в.1ь 11l7,(1 .ое11.1в I11в,11 •11а•,!1 С11,31 •1в7г,,1 (l0,1ь авь.вf (ОwвЧС вССUинllчс в.ьв»<С 1р1 вIС( , л. 1в[.С1 [вв1,1ь ,11 1,ОС .•ь 1,1! ,11 111авв1 (\,.t 1111,Ъ3 в•,11 ваь.Сl ьОwв.С асСои»г1мС вьььчС[ 1ои соflои в•ь• r гг!l.,7. ,11 .ое1,11 t1,1• wг1,ь1 13.1е (1131,ь1 еь,lь (113i,1[ в1,1г гва.1У во.внс .Ссиинtlис в.ьв»ск 1оь 7пеысо 1ы•.и.вие 73ь11,С. ,71 1.ь,\1 1.1t +1l,11 г.7f [t17,гs (1.г1 1f1f1.71 1а,[1 .вь.еии.ие .Ои..е вееоиигtие овьие[ 1Уя (ег 1•1EEnaeotr евиrь s; ]У ,11 еимrw, .11 •, С1еч(w• ео.еи.е»г е.1» [ь0, 1ле11 (.ии виlСи.ни1 110ое ви гЕе гв3 ье,..< •1.г1,,15 .11 -1]117..f 7•,гв .1и7е,17 f1,37 •г3и1,11 ь0,.о .ffee.•1 11,11 леГ-•1 оо.1 евlи по. 111.» мв71[ •(ь.С(1не �° О всв•е, /ьи1,1• ,11 .71 ,вввl•1г r/7 ,ь1 (a11,S1 1е,ь( 1г1е,7• 13,1• в[..ЧИ сиУ1 Свги 1lUв (впи Св�иипиигl мlев[1{»е ин..( 1иь.н ,11 1trv.п .ь..t 17е5,г• н,гг lвtь,77 1e7,n н11.17 1t,,ь• •(.••1 си,1 С.ы пи• лпн И1и •в.•иис .c..et 1и1,71 ,11 s,Y/ .1е 1.11 ,ьа 1111,ь1 e/,1t (stь.гг (],е( в(,•е1 еи.1 е{1и reu. (ы и, [п11иИ ,ви.еllнС о .i.'1а 1.s.1.1. ,11 .,11 ,г1 1.[1 ,ь. 105е,ь1 га,5г (11е,ы 11,1г (!1•(У еи.1 гвьи Lw. (.i,и lиваьииаое. гпввсlд и.ь.([1,и (евг ьа17l,ьв ,11 -ь1аl,о1 1е1,11 ьl1,11 !7 •11,11 7г 17.(7 11• .E.•rGN1 еи,1 е.и (ьо. lьии lивtlrueotв re.tlaltt. н..а(1не и.•ви•в,е .игн,.• ,11 -7гвlе,ье 111,•ь .иlе.лl и:е! .1а3У,гь .е;и 3iц,t: чи,1г ,е...ам.Чь lоие .Омвее аьи пи. rwм 1ицЕ.о,0и .вя.ииь (еnьlиlь � �� ,11 ••вУ 71в1,f1 ,11 1111,07 I00,1в 51ь1,в1 111,11 11в1ве{ 1е1,11 11е1,11 /1в,/1 .1.•ви W.1 Lвfw lьив Iы1w ив.�/1f tв.fь ии t]1в1! fисви мl,ь{t а•1tв+• в11 •41г,ве 1а0.Ов •17вь• 1ь(вСв ввlе,ь\ {Оь,ье 1•1г,еь 11е111 •Lt.1C 'Ut1 C1lw 1ЕUв 1И1iи rв,.11( 1ь1(е Ои (ь1в1[ f!1 4 a1r./1 14,в,(! .11 1в•1,1• 1.а,01 1вв1,1t 111,1ь lваь,ll 100вг1 {1в1,г1 I/0,i1 .L..r1 сУ,1 С.1, 1(Ов (•11w гиt117( гв.(! Uи (ьиь.l4и(0 гиьаССU иf.r.11 в5в(0 .II .l,10 101,01 ,7,1/ l01.е1 ,1,1ь 111в1е в7,11 101,11 .(,.в1 Сив1 eI]w 1Е11в Lыlw �ЧU/lL fвгl3 Ои вивlеt 1c0.CCtl ме..1и1.\ •711в,г0 в(1 ивь(,ь0 1аь.ье 1f7f.f1 ^1г.1f 111в7г •г,73 1.1•7,1ь .lг7,1г .С..гог.Е wr1 е.]и rw. 1ерИ 1еИlеиьtимвl atcrfl. сиаЕмии0еи 1н0 (егвгС3 . Ч 1Miuwtl ,11 ,. •> 1н[�•Е1 •[fu 1чсД.(] 1.[и нииlСние0 /Uв 1,ваьппL0(и1 вмУ Ув1в,1 1ыС.»1.ьы 1Lf,.] ,11 1ви,7г 1аг,.( 1/е.1[ t11.73 1)г.в1 ь1,г6 1ьl,1ь tC,1I IнС.•1.4и иtвь мм lпеп•[ ur мв17[.�аиинр»и1 !»веегиеп[и1 С 1н[.,7в.1 1f1.аь .11 ill,ts 1Lr,3f 1ьг.11 111.е7 11•,r1 1е1,11 1е7,l1 (г1l! 1нС.иt•м .Еаа пи иеи^С Пr и.( «.а 1»С.• ( 1 1Ct 1w.lt•пс0(м1 1.GI 1и1,г1 ,11 14е,1е 1ь1,11 171вв1 107r1e 1.1.1. 1t(,,в 1e1,1i 1•,11 � 1мС.,1•С1 игвб мИ 14СUи[ U1 wв1([•Ср11ии fиactn�lO/яв 1»Сь^(.1и 7l7,11 ,11 ш,н 1гь,7: 11еы ♦ иг.в1 иlдl 1п,1г га,ю u1,n 1че•,1•го Оие ии 1чеУ^е nr •ип•еаиино»ut•гм•Сео л,вес^цаw О 1»е.и..вм 1г71.1. ,11 1п+,70 1аь,ьв t151.ц 101,ее 1гft,fг 1ь1.1е 17и,71 ,,,1г 1ис•wиси +свв wевн ми 1ис.iне ' l•С.ьи ,1t1.30 ,11 .711,51 101,(ь ,7t/,7, 11о,01 11г1r1ь 111вСi ,7ь[,г1 1,,11 14С•]У •!вь 1нСои( П( 7иСьв L]1/Ll1 (1001 в.i � (»с•1е {1eьt.t7 .11 11.ь1.1. 101,1ь 11111,t• 11ь.еЕ 117е1,1г 1г1,Оь 111и.3( [е,г1 г4с•« .1ьt 1нn1^[ ur 1еь е]гвг(3 11У01 .4, О 1н[в(1 {ьlв.вь .11 7в1•,.1 1оь,01 1а11,(1 1лг,(( 711(,е1 11о,11 [1ь7,гг 1r,l1 1•с•.7 ме.1. [исичг ur reue tw[о го1«ео t11вг[1 11ьОв ..1 1и[.е1 и,(7 ,и и,11 1о(,еь 1г.7f п1.11 u.r1 л1,1ь 1г,3• е7,[1 1иСвег .ul 1,Сриг аг 1we[г (Уиеео []1ц(1 г100г ..1 � : ьи.,( ..иил. :ii ипе.,н 1аС,1е (г\17lв77 иь.и +1ии.\1 1п,•• ц1,1.,н и7111 ' •. ,ген.е ие1 •еи{ь[ � ,11 . ,11 t 1 .ii .11 ,f1 ,I1 1i1 • . 1 ц � J J