__ .__‘_o.—._. LIBRARY Michigan State Unlverslty PLACE IN RETURN BOX to removo mic checkout from your record. TO AVOID FINES return on or Moro dd. duo. DATE DUE DATE DUE DATE DUE 4,1 ”—1 L’T‘W MSU It An Affirmative ActiorVEqunl Opponmlty Instirmion cha-QI AN ECONOMIC ANALYSIS OF FERTILIZER ALLOCATION AND IMPORT POLICIES IN SYRIA Volume I By Maurice Emile Saade A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Economics 1991 ABSTRACT AN ECONOMIC ANALYSIS OF FERTILIZER ALLOCATION AND IMPORT POLICIES IN SYRIA BY Maurice Emile Saade The main objective of this dissertation is to develop a national economic decision model to be used as a tool by Syrian policy makers to plan more economically efficient allocations of the limited fertilizer supplies in Syria. The model also serves as the main framework for analyzing the economic implications of the current constraints on fertilizer supplies. These constraints are mainly the result of government restrictions on fertilizer imports in an attempt to reduce foreign exchange expenditures. The model was formulated in terms of a separable linear programming model, based on the results of fertilizer experiments undertaken between 1965 and 1989 on the most important crops in Syria. Compared to the current government fertilizer allocation strategy, the results indicated that a strategy based on the proposed model would significantly increase ‘national and. farm incomes and. aggregate crop output, and would reduce the government's foreign exchange and general budget deficits. The main effect of the current constraints on fertilizer supplies is a substantial reduction in fertilizer use on rainfed cereals, particularly barley. Thus, a policy of importing all the fertilizer needed to satisfy total requirements would substantially increase aggregate crop output. Compared to the current situation, this policy would increase national income by 2.4 billion Syrian Liras (SL), a 4.2% increase in the agricultural Gross Domestic Product (GDP). This would also result in a decline of 54 million $US in net crop imports, which would more than offset the Al million $US needed to import the additional fertilizer. Given the current heavily subsidized fertilizer prices, such a policy would require up to 800 million SL in additional fertilizer subsidies. The results showed that a 71 to 151 increase in the price of nitrogen and 451 to 601 in that of phosphorus would allow the government to satisfy total fertilizer requirements without increasing its current expenditures on subsidies. Farmers' unlimited access to fertilizers would allow them to increase their output, which would lead to higher revenues that would offset any increase in fertilizer costs. Compared to the current situation of subsidized but limited fertilizer supplies, farmers' income would increase by an average of 1.5 billion SL, in spite of higher fertilizer costs. Copyright by Maurice Emile Saade 1991 to Lamla ACKNOWLEDGMENTS This dissertation is based on a collaborative research project be tween the Soils Directorate of the Syrian Ministry of Agriculture and Agrarian Reform, and the International Center for Agricultural Research in the Dry Areas (ICARDA). Many people from both institutions contributed to the success of this project, to whom I wish to express my app reciation and gratitude. From the Soils Directorate, I am especially grateful to Mr. Kl'laz 2aa El-Hajj for sharing with me his valuable time, in-depth l|(I"I<>“.r'.‘ledge, and long experience, which were instrumental in the writing of this dissertation. Ms. Lina Meda helped with the tedious task of data entry and analysis. Mr. Adeeb Safi, Ms. Bdour Al-Bunni, and Ms. widad Al-Shahadeh helped in compiling the results of fertilizer ext) 6 riments . From ICARDA, my deepest thanks are to Dr. Thomas Nordblom for his Cruc 1431 support, encouragement and contribution to this research, and for his, and his family's, warm friendship. Mr. Ahmad Mazid and Drs. Peter Cooper, Abdallah Matar, Elizabeth Bailey and Michael Jones pl-o"':l>~‘=1ed insightful comments at many points throughout this project. I am grateful to the Ford Foundation who provided funding for this dissertation, and the efforts of Dr. David Nygaard in supporting it are r g eatly appreciated. vi vii I am especially thankful to my advisor, Dr. Stephen Harsh, for tak ing the time to visit me in Syria during my stay there, for his ins :lghtful comments, and for the long hours he spent with me before this dis sertation was completed. My deepest appreciation is to my graduate program supervisor and thesis committee member, Dr. John Staatz, for his numerous valuable comments on the dissertation, and for his encouragement and practical wisdom which were of great help throughout my graduate program at Michigan State University. I am also thankful to Drs - Eric Crawford and Roy Black, members of the thesis committee, for the ir useful comments. Finally, I wish to thank my mother, Béatrice Dagher, for her love and patience throughout my graduate studies, and for her faith that, Somehow, one day I will finally end up getting a “real job" as a result of this long and tedious endeavor. I dedicate this dissertation to my wife and best friend, Lamia El-Fattal, for her emotional and moral 8ut>r><>rt, and for her faith in me, which provided me with all the Strength and patience in writing this dissertation. TABLE OF CONTENTS L1-t: of Tables List of Figures (ring-L1F>ter 1. (Ital-rlppter 2. INTRODUCTION 1.1 Background and Problem Definition . 1.2 Objectives 1 3 Dissertation Outline THE PRESENT FERTILIZER SITUATION 2.1 Land Use and Crop Mix . . . 2.2 Evolution of Fertilizer Consumption . 2 2 1 Consumption Trends . . . 2.2.2 Estimating Future Consumption 2.3 Domestic Fertilizer Production 2.4 Fertilizer Imports 2.5 Fertilizer Prices . . 2.6 Planning Aggregate Fertilizer Requirements 2.6.1 Institutional Setting . . 2 6 2 Estimation of Aggregate Requirements . 2.7 Fertilizer Rationing and Allocation Strategies Fertilizer Availability 7.1 7. 2 Fertilizer Allocation 2. 2. 2.8 Fertilizer Marketing 2. 8.1 Official Channels . 2. 8. 2 The Parallel Market 2.9 Farmers' Fertilizer Strategies viii Page ON!“ 11 11 15 15 21 24 26 27 29 29 31 36 36 37 39 39 40 43 (:lbaaaleipter 3. ix THE RESEARCH APPROACH . 3.1 Optimum Fertilizer Rates at the Farm Level 3.1. Unconstrained Optimization . .1 NH Fertilizer Rates . 3.1.3 Risk and Uncertainty . 3.1.4 Constrained Optimization . 3.2 Fertilizer Allocation at the National Level . 3 2 1 Conceptual Approach . . . 3.2.2 Financial vs Economic Prices . 3 2 3 Incorporating Policy Concerns 3.3 The Fertilizer Allocation Model . 3.3.1 Linear Programming . 3.3.2 The Basic Model 3 3 3 Specification of the Production Functions 3.1 Functional Form . .3.2 Explanatory Variables 3 3 Linear Approximation . WWW UUU 3.3.4 The Expanded Model . 3.3.4.1 Data Sources and Crops Covered . 3.3.4.2 Model Specification . 3.3.4.3 Algebraic Formulation of the Fertilizer Allocation Model ESTIMATION OF FERTILIZER AND CROP PRICES Financial Prices Economic Prices . fit§¢ wNH General Budgets . 4»3.l Foreign Exchange Earnings and Expenditures . . . . 4. 3. 2 Taxes and subsidies 4.4 Fertilizer Prices . Fertilizer Financial Prices Fertilizer Economic Prices . 93>§ §b§ UNH Fertilizer Imports . . . . 4.4.4 Net Subsidies on Fertilizers . Assessing the Feasibility of Optimum Impact on the Government s Foreign Exchange and Foreign Exchange Expenditures on 47 47 47 50 53 S9 62 62 64 66 71 71 74 75 76 78 82 84 84 87 95 102 103 105 112 112 113 114 114 115 115 118 4.6 4.7 4.8 4.9 4.10 Wheat Prices 4.5.1 Financial Prices of Wheat Grain 4.5.2 Economic Prices of Wheat Grain . 4.5.3 Financial Prices of Wheat Straw . . 4.5.4 Financial and Economic Prices of Total Wheat Output . . . 4.5.5 Net Savings in Foreign Exchange on Wheat Grain . 4.5.6 Net Taxes on Wheat Grain . Barley Prices . 4.6.1 Financial Prices of Barley Grain . 4.6.2 Economic Prices of Barley Grain 4.6.3 Financial Prices of Barley Straw . . 4.6.4 Financial and Economic Prices of Total Barley Output . 4.6.5 Net Savings in Foreign Exchange on Barley Grain . 4.6.6 Net Taxes on Barley Grain Cotton Prices . 4.7.1 Financial Price of Raw Cotton 4.7.2 Economic Price of Raw Cotton . . 4.7.3 Foreign Exchange Earnings from Cotton 4.7.4 Net Taxes on Raw Cotton Corn Prices . .1 Financial Price of Corn .2 Economic Price of Corn . . . .3 Foreign Exchange Savings from Corn . .4 Net Taxes on Corn . . bbb? ooIooZoooo Sugar Beet Prices . Financial Price of Sugar Beets . Economic Price of Sugar Beets ##4‘ 000 UNH Beets . Net Taxes on Sugar Beets . b ‘0 & Potato Prices . 4.10.1 Financial Price of Potatoes 4.10.2 Economic Price of Potatoes . . . . . . . 4 . 10 . 3 Contribution of Potatoes to the Government's Foreign Exchange and General Budgets Foreign Exchange Savings from Sugar 118 118 123 123 127 127 129 130 130 132 134 134 135 137 138 138 138 140 141 141 141 141 141 143 143 143 145 146 146 146 146 148 148 (Itmualmcjgmter 5. xi 4.11 Summary of Estimated Financial and Economic Prices FERTILIZER. RECOMMENDATIONS UNDER. NO CONSTRAINTS ON FERTILIZER SUPPLIES . . . . . . 5.1 Summary of the Estimated Production Functions . 5.1.1 Statistical Performance of Estimated Functions 5.1.2 Zone-Specific Production Functions for Rainfed Crops Estimation of Ideal Fertilizer Recommendations Feasibility of Proposed Fertilizer Recommendations . . U‘U‘ WM 5 3 1 Profitability of Proposed Rates 5.3.2 Sensitivity Analysis . 5.4 Financial vs Economic Optimum Fertilizer Rates 5.4.1 Impact of Economic and Financial Optimum Rates on Aggregate Crop Production . 5.4.2 Impact of Economic and Financial Optimum Fertilizer Rates on National Income 5.4.3 Impact of Economic and Financial Optimum Fertilizer Rates on Farm Income 5.4.4 Impact of Economic and Financial Optimum Fertilizer Rates on Foreign Exchange Earnings . 5.4.5 Impact of Economic and Financial Optimum Fertilizer Rates on the Government Budget . . . . . . . 5.4.6 Financial vs Economic optima: Summary 5.5 Current vs PrOposed Fertilizer Recommendations 5.5.1 Impact of Current and Preposed Fertilizer Rates on Aggregate Crop Production . 5.5.2 Impact of Current and Proposed Rates on National Income 5.5.3 Impact of Current and Proposed Rates on Farm Income 5.5.4 Impact of Current and Proposed Rates on Foreign Exchange Earnings 5.5.5 Impact of Current and Proposed Rates on the Government Budget 5.5.6 Current vs Proposed Fertilizer Recommendations: Summary . 5.6 Chapter Summary . 148 150 151 151 155 157 159 159 163 170 172 175 177 178 178 181 182 182 185 188 189 191 193 194 (TlnLazlelpter 6. xii FERTILIZER ALLOCATION STRATEGIES UNDER LIMITED O‘O‘ NH 6.3 6.4 6.5 FERTILIZER SUPPLIES . Constraints on Fertilizer Supplies A Test of Accuracy of the Fertilizer Allocation Model . 6.2.1 LP vs Calculus Solutions of the Unconstrained Problem . . 6.2.2 LP Solution vs Farmers' Optima . Implications of the Constraints on Fertilizer Supplies 6.3.1 Optimum Fertilizer Rates Under Limited Supplies . 6.3.2 Impact of Limited Fertilizer Supplies on Aggregate Crop Production 6.3.3 Impact of Limited Fertilizer Supplies on National Income 6.3.4 Impact of Limited Fertilizer Supplies on Foreign Exchange Earnings 6.3.5 Impact of Limited Fertilizer Supplies on Farm Income 6.3.6 Impact of Limited Fertilizer Supplies on Government Expenditures 6.3.7 Impact of Limited Fertilizer Supplies. Summary and Policy Implications Alternative Fertilizer Pricing Strategies . Fertilizer Normative Demand Functions 6.4.1 6.4.2 Definition of Fertilizer Price Scenarios . 6 4 3 Optimum Rates under Alternative Fertilizer Price Scenarios . 6.4.4 Impact of Higher Fertilizer Prices on Aggregate Crop Output 6.4.5 Impact of’ Higher Fertilizer Prices on National Income 6.4.6 Impact of' Higher Fertilizer Prices on Farm Income 6.4.7 Impact of Higher Fertilizer Prices on the Government Budget 6.4.8 Impact of’ Higher Fertilizer Prices on Foreign Exchange Earnings 6.4.9 Impact of'.Alternative Fertilizer Price Scenarios: Summary and Policy Implications . Comparison of Alternative Fertilizer Allocation Strategies for the Winter Season 197 198 200 200 202 203 204 206 208 209 213 215 217 218 219 224 228 230 230 233 235 237 239 241 Gupta: 7 . xiii 6.5.1 Constraints on Fertilizer Supplies for the Winter Season 6.5.2 Fertilizer Allocation S trategies for the Winter Season 6.5.3 Impact of Alternative Fertilizer Allocation Strategies on Aggregate Crop Output . 6.5.4 Impact. of' Allocation Strategies on National Income 6.5.5 Impact of Allocation Strategies on Farm Income . 6.5.6 Impact of’ Al location Strategies on Foreign Exchange Earnings 6.5.7 Impact of Allocation Strategies on the Government Budget 6.5.8 Alternative Fertilizer' Allocation Strategies for the Winter Season: Summary and Policy Implications 6.6 Chapter Summary . SUMMARY, POLICY IMPLICATIONS AND FUTURE RESEARCH 7.1 The Research Objectives and Approach: A Recapitulation . . . . . . . . 7.2 Summary of the Major Findings . 7.2{1 Fertilizer Recommendations under Unconstrained Supplies . 7.2.2 Alternative Fertilizer' Allocation Strategies . 7.2.3 Implications of the Constraints on Fertilizer Supplies 7.3 Policy Implications . \JV WU) NH Current Fertilizer Allocation Strategy Based on the Proposed Fertilizer Recommendations 7.3.3 Equimarginal Allocation of Limited Fertilizer Supplies 7.3.4 Unrestricted Fertilizer Imports at Current Official Prices 7.3.5 Unrestricted Fertilizer Imports with Higher Official Prices . 7.3.6 Other Policy Issues 7.3.6.1 Role of Fertilizers in Increasing Wheat Self- Sufficiency 7.3.6.2 Barley Fertilization in the Drier Zones Fertilizer Policy Objectives and Options . 241 244 248 251 253 255 255 258 259 263 263 266 266 267 269 271 271 273 274 275 278 283 283 284 ..... xiv 7.4 Implications for Future Research APPENDIX A: LAND USE AND CROP MIX . APPENDIX B: THE FERTILIZER REQUIREMENT SCHEDULE, 1989/1990 APPENDIX C: COMPUTER INPUT FILE FOR THE FERTILIZER ALIDCATION LINEARFROCWINGHODEL BIBLIOGRAPHY 287 296 300 305 322 .h5 .u. a II o - 4 ‘.~.E pl» h ‘ . .4a 0 . IITaaLIE>ile TII£11t>Zle fireaJt>ile ‘1?£11b>21e 1Catt3filxs frillabile Trailaije 'réilavlLe Th313C1_e 'raflbb]_e 133t>3l¢e 1R3t>JL¢3 Tater‘a TabJL‘at T31>JL¢3 LIST OF TABLES PLANNED CROPPINC PROGRAM Syria, 1989/90 Season . EVOLUTION OF FERTILIZER USE (TONS) Syria, 1954/55 to 1989/90 . ACTUAL AND PREDICTED FERTILIZER CONSUMPTION Syria, 1980 to 2000. FERTILIZER PRODUCTION CAPACITY AND ACTUAL PRODUCTION LEVELS: Syria, 1981/82 to 1988/89. FERTILIZER IMPORTS: Syria, 1984/85 to 1988/89 FERTILIZER OFFICIAL PRICES Syria, 1984/85 to 1989/90 . THE OFFICIAL FERTILIZER REQUIREMENT PLAN Syria, 1989/9O ESTIMATION OF PLANNED FERTILIZER REQUIREMENTS Syria, 1989/9O . . . . . . PLANNED AND ACTUAL FERTILIZER REQUIREMENTS Syria, 1984/85 to 1989/90. . . . DEFINITIONS OF RAINFALL SCENARIOS Syria, Agricultural Stability Zones 1b, 2, and 3. OFFICIAL FERTILIZER SALES PRICES Syria, 1989/1990 FINANCIAL PRICES OF N AND P205 FERTILIZERS Syria, October 1989.. . . ECONOMIC PRICES OF N AND P505 FERTILIZERS Syria, October 1989 . . . . . . . FINANCIAL PRICES OF HYV WHEAT GRAIN Syria, October 1989. FINANCIAL PRICES OF LIV WHEAT GRAIN Syria, October 1989. Page 16 17 23 24 26 28 31 34 35 56 115 116 117 121 122 CITaaible TITaaible TITJSEble TITaaEble TITzanle 'Ifzajble 'Ifealble Ttfiallole 1EHELIOle 'Iflallale TTanlble jtzilale t1.411316 tré‘lafilcl fir‘ila'luo 1t£ilaflue 4.6: 4.7: 4.8: 4.9: 5.1: 5.2: 5.3: xvi ECONOMIC PRICES OF HYV WHEAT GRAIN Syria, October 1989. ECONOMIC PRICES OF LYV WHEAT GRAIN Syria, October 1989. FINANCIAL PRICES OF WHEAT STRAW Syria, October 1989. PRICES OF TOTAL WHEAT OUTPUT Syria, October 1989. : FINANCIAL PRICES OF BARLEY GRAIN Syria, October 1989. : ECONOMIC PRICES OF HARLEY GRAIN Syria, October 1989. : FINANCIAL PRICES OF BARLEY STRAW Syria, October 1989. : PRICES OF TOTAL BARLEY OUTPUT Syria, October 1989. : FINANCIAL AND ECONOMIC PRICES OF RAN COTTON Syria, January 1990. : FINANCIAL AND ECONOMIC PRICES OF CORN Syria, March 1990. : FINANCIAL AND ECONOMIC PRICES OF SUGAR BEET Syria, March 1990. : FINANCIAL AND ECONOMIC PRICES OF POTATOES Syria, March 1990. : SUMMARY OF FINANCIAL AND ECONOMIC PRICES, NET TAXES, AND FOREIGN EXCHANGE EARNINGS OF FERTILIZERS AND MAIN CROPS Syria, 1989/1990. ESTIMATED PRODUCTION FUNCTIONS FOR THE CROPS INCLUDED IN THE FERTILIZER ALLOCATION MODEL IN SYRIA . OPTIMUM AND CURRENT FERTILIZER RATES (KG/HA) Syria, 1989/1990 . . . . . . . . VALUE—COST-RATIOS OF FERTILIZER USE BASED ON PROPOSED RATES Syria, 1989/1990. 124 125 126 128 131 133 135 136 139 142 144 147 149 152 158 161 TITuEaIDIe TITaallale TITaallale 'ITaallale ‘Ifaallale 'Ifiallale 'Ifiaalade Tiaatade 17£alade 7r£113142 Trait):l‘3 't€(E>]L‘, .10: .11: .12: .13: .14: .15: .16: xvii COMPARISON OF PROPOSED FERTILIZER RATES ON RAINFED CROPS WITH MAXIMUM RATES IN DRY AND VERY DRY YEARS Syria, 1989/1990 COEFFICIENT OF VARIATION IN INTERNATIONAL FERTILIZER AND CROP PRICES ESTIMATED RANGE OF ECONOMIC VALUE-COST-RATIOS OF PROPOSED FERTILIZER RATES DUE TO CHANGES IN INTERNATIONAL FERTILIZER AND CROP PRICES FINANCIAL. VALUE-COST-RATIOS OF FERTILIZER. USE BASED ON ECONOMIC AND FINANCIAL OPTIMUM RATES Syria, 1989/1990. IMPACT OF ECONOMIC AND FINANCIAL OPTIMUM FERTILIZER RATES ON AGGREGATE CROP PRODUCTION Syria, 1989/1990 ACCRECATE ECONOMIC IMPACT OF ECONOMIC AND FINANCIAL OPTIMUM FERTILIZER RATES Syria, 1989/1990 IMPACT OF ECONOMIC AND FINANCIAL OPTIMUM FERTILIZER RATES ON FOREIGN EXCHANGE EARNINGS Syria, 1989/1990 IMPACT OF ECONOMIC AND FINANCIAL OPTIMUM FERTILIZER RATES ON THE GOVERNMENT BUDGET Syria, 1989/1990 IMPACT OF CURRENT AND PROPOSED FERTILIZER RATES ON AGGREGATE CROP PRODUCTION Syria, 1989/1990 AGGREGATE FERTILIZER REQUIREMENTS BASED ON CURRENT AND PROPOSED RATES Syria, 1989/1990 AGGREGATE ECONOMIC IMPACT OF PROPOSED AND CURRENT FERTILIZER RATES Syria, 1989/1990 IMPACT OF PROPOSED AND CURRENT FERTILIZER RATES ON FOREICN EXCHANGE EARNINGS Syria, 1989/1990 IMPACT OF PROPOSED AND CURRENT FERTILIZER RATES ON THE GOVERNMENT BUDGET Syria, 1989/1990. 162 165 168 171 174 176 179 180 184 186 187 190 192 I. . v v I I .a:. .1. a Q “"'L‘ .534 . ' Q ~aa‘( in“ 133‘ ‘Jrhealale TITuEIJble TITaatlale Table Triéilale 'IIEIIDIe TIVatlale 'Ifiallale 13:11:19 1P£1131e 7r£1t>141 fir‘iID'IAa TIalbla .10: .11: .12: .13: xviii OPTIMUM FERTILIZER RATES BASED ON CALCULUS AND LP SOLUTIONS. OPTIMUM FERTILIZER RATES UNDER VARYING LEVELS OF FERTILIZER AVAILABILITY Syria, 1989/1990 AVERAGE CROP PRODUCTION INCREASE UNDER VARYING LEVELS OF FERTILIZER AVAILABILITY Syria, 1989/1990 AGGREGATE ECONOMIC IMPACT OF FERTILIZER USE UNDER VARYING LEVELS OF FERTILIZER AVAILABILITY Syria, 1989/1990 IMPACT OF VARYING LEVELS OF FERTILIZER AVAILABILITY ON FOREIGN EXCHANGE EARNINGS Syria, 1989/1990 IMPACT OF VARYING LEVELS OF FERTILIZER AVAILABILITY ON FARMERS AGGREGATE NET RETURNS TO FERTILIZER USE Syria, 1989/1990 IMPACT OF VARYING LEVELS OF FERTILIZER AVAILABILITY ON THE GOVERNMENT BUDGET Syria, 1989/1990. FERTILIZER NORMATIVE DEMAND FUNCTIONS AND PRICE ElASTICITIES Syria, 1989/1990 DEFINITIONS OF FERTILIZER PRICE SCENARIOS . OPTIMUM FERTILIZER RATES UNDER VARYING LEVELS OF OFFICIAL FERTILIZER PRICES Syria, 1989/1990 AVERAGE CROP PRODUCTION INCREASE UNDER VARYING LEVELS OF OFFICIAL FERTILIZER PRICES Syria, 1989/1990 AGGREGATE ECONOMIC IMPACT OF FERTILIZER USE UNDER VARYING LEVELS OF FERTILIZER OFFICIAL PRICES Syria, 1989/1990 IMPACT OF VARYING LEVELS OF FERTILIZER OFFICIAL PRICES ON FARMERS AGGREGATE NET RETURNS TO FERTILIZER USE Syria, 1989/1990 201 205 207 210 212 214 216 223 225 229 231 232 234 Table 6 I ‘ V .&:;€ 5 p A! A. m ‘3.» ‘ A:~‘e ‘ :a‘Zle f 13‘- 'hl“,v ~K Table A rable A ‘1lfaafib1e TITaaEble TITaaEble 'ITaaflbIe 'ITzanIe frfiallale Triallale TPéilale .14: .15: .16: .17: .18: .19: .20: .21: xix IMPACT OF VARYING LEVELS OF OFFICIAL FERTILIZER PRICES ON THE GOVERNMENT BUDGET Syria, 1989/1990 IMPACT OF VARYING LEVELS OF OFFICIAL FERTILIZER PRICES ON FOREIGN EXCHANGE EARNINGS Syria, 1989/1990 FERTILIZER. RATES UNDER. ALTERNATIVE ALLOCATION STRATEGIES FOR THE WINTER SEASON Syria, 1989/1990 IMPACT OF ALTERNATIVE FERTILIZER ALLOCATION STRATEGIES ON AGGREGATE PRODUCTION OF FALL- PIANTED CROPS Syria, 1989/1990 AGGREGATE ECONOMIC IMPACT OF ALTERNATIVE FERTILIZER ALLOCATION STRATEGIES FOR THE WINTER SEASON Syria, 1989/1990 IMPACT OF ALTERNATIVE FERTILIZER ALLOCATION STRATEGIES FOR THE WINTER SEASON ON FARMERS NET RETURNS TO FERTILIZER USE Syria, 1989/1990 IMPACT OF ALTERNATIVE FERTILIZER ALLOCATION STRATEGIES FOR THE WINTER SEASON ON FOREIGN EXCHANGE EARNINGS Syria, 1989/1990 IMPACT OF ALTERNATIVE FERTILIZER ALLOCATION STRATEGIES FOR THE WINTER SEASON ON THE GOVERNMENT BUDGET Syria, 1989/1990. ALI’PFEJNDIXTABLES: it‘lIDJLe A.l: .r£.]:’]u0 AS2: It‘lla’lue B.1: T£it>3Le.B.2: EVOLUTION OF TOTAL CULTIVATED AREAS IN SYRIA, 1984/85 to 1989/90 (ha). PLANNED CROPPINC PROGRAM: Syria, 1989/90 Season (ha), FERTILIZER REQUIREMENTS OF WINTER CROPS Syria, 1989/1990 . . . . . . . FERTILIZER REQUIREMENTS OF SUMMER CROPS Syria, 1989/1990 . . . . . . . 236 238 246 249 252 254 256 257 296 297 300 301 'JTIaEble 8.3: TITanble 8.4: Table 8.5: XX FERTILIZER REQUIREMENTS OF IRRIGATED FRUIT TREES Syria, 1989/1990 . . . . . . . . . . . . . . . . . . . 302 FERTILIZER REQUIREMENTS OF RAINFED FRUIT TREES Syria, 1989/1990 . . . . . . . . . . . . . . . . . . . 303 AGGREGATE FERTILIZER REQUIREMENTS Syria, 1989/1990 . . . . . . . . . . . . . . . . . . . 304 LIST OF FIGURES Page Figure 1: Agricultural Stability Zones in Syria . . . . . . . . . 13 Figure 2: Fertilizer Consumption: Syria, 1955 to 1990 . . . . . . 19 xxi CHAPTER 1 INTRODUCTION 1-1 W The access of farmers to fertilizers is one of the most important factors affecting the productivity of Syrian agriculture and its ability to meet current and future food and fiber needs for domestic consumption and for exports. Chemical fertilizers were first commercially used in Syria in the early 1950's, with the rapid expansion in cotton cul tivation. Since then, fertilizer use has grown dramatically, especially during the past two decades. Fertilizer use per hectare of arable land increased from 6.7 kg of plant nutrients in 1970 to 32 kg in 1984 (World Bank, 1987, p. 213). Between 1974 and 1978 alone, the use of nitrogen fertilizer doubled and that of phosphate fertilizer quadrupled (Nixon, 1979, p. 17). Faced with the rapid growth in fertilizer consumption and imports, the Syrian government has invested heavily in expanding production capacity, These investments were encouraged by the discovery in the late 1960's of substantial deposits of rock phosphate, oil, and natural gas that can be used as raw materials for fertilizer production. The “J Or expansion in production capacity in the early 1980's had the goal of Making Syria self-sufficient in fertilizer production and being able t° Sell fertilizer on the export market. This expansion was also part 2 of a broader fertilizer policy that included heavy subsidies to encourage farmers to rapidly increase their fertilizer use. Throughout the 1980's fertilizer production levels remained much lower than the planned capacity due to a multitude of technical and Ianagerial problems. These problems, in conjunction with the rapid 1“filtease in consumption, resulted in continued importation of SUbSCantial quantities of fertilizer. These import levels could not be sustained for a long time, especially in light of the severe foreign exchange crisis facing Syria since the mid 1980's. In response to this crisis, the government has resorted to reducing all imports, including that of fertilizers. As a result, the quantities of fertilizer demanded by farmers, given current subsidized prices, have often exceeded the total available supplies, leading to a situation of chronic shortages. In an attempt to address the problem of frequent shortages, the g<>\’ernment introduced a fertilizer rationing system. According to this system, each farmer is allocated a quantity of fertilizer based on the type of crops grown and the area planted to each crop. There are 8e"eral potential problems associated with such a rationing policy. These include: (1) the economic rationale of fertilizer allocation Itrategies, (2) the accuracy of information in a centrally planned all<>cation system, and (3) the economic implications of restrictions on fe 1‘ t ilizer imports . 1. o t e a oc t on t e e ' The fertilizer rationing and allocation system is a centrally p1"'Eu'lned and managed system. In other words, government planners decide 1‘. he fertilizer rates to be allocated to each crop in each agro-climatic 3 zone , given the limited supplies available for distribution to farmers. Thus , an important potential problem related to this system is the underlying assumptions upon which the allocation decisions are made. The main purpose of the rationing system is to allocate the limited supplies to various crops so as to maximize net returns from fertilizer use. At the same time, this system attempts to incorporate P011cy concerns about food self-sufficiency and balance of trade defic it. Thus, every effort is made to ensure that “strategic crops” rece lye their ideal rates, i.e., those maximizing net returns to fer't-Zilizer application. These crops are defined as those providing suI>Stantial export earnings (e.g. , cotton) and crops that substitute for the main food imports (e.g. , wheat and sugar). Another important factor influencing allocation decisions is whether craps are irrigated or rainfed. The general rule is that irrigated crops should receive their ideal fertilizer rates. This is because net economic returns to fertilizer use on rainfed crops, e'313ecially in the drier areas, are lower and more risky than on it’1‘1gated crops. Given the existing constraints on fertilizer supplies, the practical implications of the above allocation strategies are that itIrigated crops and rainfed wheat in the high rainfall zones usually receive most if not all of their ideal rates, with very limited fel'l‘tilizer left for rainfed barley. In fact, the first time that Eel:‘tilizer was allocated to barley was in 1986/87, but even then only in the wetter areas (above 250 mm average annual rainfall). The above policy priorities may conflict with the initial obj aetive of maximizing net returns from the limited quantities of f artilizer available. Maximizing net returns from fertilizer use under 4 rationing conditions requires that marginal revenues be equal across all creps (equimarginal rule). Thus, if aggregate fertilizer supplies decline, the rates on all crops would need to be reduced in such a way as to maintain the equality between marginal revenues. The above priority system implies that the fertilizer rates on high-priority crops are kept constant, or slightly reduced, while the rates on low-priority creps are reduced drastically. This would lead to a situation where marginal revenues from fertilizer application on low-priority crops are larger than those on high-priority crops resulting in an economically inefficient allocation of the limited fertilizer resources. Therefore, an allocation strategy based on the above priorities may result in a Clec-‘—:Line in aggregate net returns to fertilizer use, compared to net re t\arns potentially obtained if the equimarginal rule was used as the has is for allocation. Concerns about potential inefficiencies in the current allocation stll‘ategy have been reinforced by results from recent fertilizer e)‘I>eriments in northern Syria.~ These experiments suggest that barley fertilization in the drier areas might be much more profitable and less r1isll:ts is to reduce both the government's foreign exchange deficit and the balance of trade deficit in general. Such an objective is attadinable only if domestic fertilizer production is increased to s“E’stitute for any decline in imports. This could be achieved by sol‘Wing some of the technical and managerial problems facing fertilizer prodUction, and/or by a further expansion in production capacity. lIlo‘we‘rer, the recent downward trends in fertilizer production clearly suggest that production problems will probably continue to hamper the Syrian fertilizer industry for the next few years. Also, any planned expaFrisian in production capacity would require several years to Lllplfi‘daent and would still be subject to the same foreign exchange constraints facing fertilizer imports. Therefore, the option of increasing fertilizer production levels w Mild, be feasible only in the medium or long run. In the meantime, any 7 reduction in fertilizer imports would lead to lower use by farmers. This my result in reduced aggregate crop production, particularly of export crops and/or crops that substitute for major food imports. Such potential declines in net crop exports may offset any savings in foreign exchange from lower fertilizer imports. Therefore, the net result of a Policy of reduced fertilizer imports may be to further exacerbate the balance of trade deficit and the government's foreign exchange deficit. 1'2 912.15.95.12: The potential problems associated with the policy of fertilizer rati<>t1ing in Syria indicate the need for a systematic analysis of two basic aspects of this policy. First, problems associated with the centrally planned implementation of fertilizer rationing need to be ad“(ll‘essem This includes an evaluation of the current fertilizer allocation strategies, in contrast to a strategy based on equating marginal revenues across all crops. Furthermore, in order to implement any fertilizer allocation strategy properly, a systematic review of current fertilizer recommendations needs to be undertaken for the most 1.150 Itant crops grown in Syria. Second, the economic rationale of the policy of limited fertilizer ilnt><>1I:'ts needs to be analyzed, especially in light of the short- to “dim-term constraints on any increase in fertilizer production. This re‘1\-‘l»1res an assessment of the economic implications of this policy in c0“I’larison to alternative policies based on importing all the fertilizer n eeded to fill the gap between domestic production and aggregate re q“ 1 rements . 8 Consequently, the main goal of this dissertation is to develop an analytical framework for estimating the efficient allocation of limited fertilizer supplies in Syria. This framework will be formulated in terms of a fertilizer allocation decision model based on the equimarginality principle. Such a model would allow to achieve three basic objectives: (1) it can serve as a tool to be used by Syrian Planners to assist them in formulating economically efficient annual fertilizer allocation plans; (2) it will allow the evaluation of altelf‘li'uative fertilizer allocation strategies; and (3) it will provide the basis for estimating the economic implications of the current policy of 1-i_u|ited fertilizer imports. An important prerequisite for the proper formulation and i“nt’lementation of this proposed framework is a systematic review of the current fertilizer recommendations for the major crops in Syria. The‘L‘efore, the specific objectives of this dissertation can be u"“"lllarized as follows: 1 - To review current fertilizer recommendations for the major crops in Syria. 2 - To develop a decision model to be used by Syrian government planners in formulating annual plans for the allocation of fertilizer among alternative uses. 3 F To analyze the economic impact of alternative fertilizer allocation strategies under different fertilizer availability constraints, and to make recommendations regarding more appropriate strategies. 9 la» - To analyze the economic implications of the current policy of limited fertilizer imports, in comparison with alternative policy options. 5 - To identify areas in which future research would be useful. 1- 3 W The dissertation includes seven chapters, including this introductory chapter. Chapter 2 presents an overview of the current fertilizer situation in Syria. This includes recent trends in fel-"‘t:llizer consumption, production, imports, prices and marketing, a diSczussion of the institutional setting for planning fertilizer all<>cation policies, and a brief discussion of the fertilizer strategies used. by farmers in Syria. In chapter 3, the research approach used in this study is presented. This includes a discussion of the conceptual approach and its underlying assumptions, and an outline of the main steps in the specification of the fertilizer allocation model. Chapter 4 presents the procedures and assumptions for estimating the financial and economic prices of the crops and fertilizers included in this study. Also inc luded in this chapter are estimates of net taxes (or subsidies) and net foreign exchange earnings (or expenditures) associated with each uni; t of crop produced or fertilizer used. In chapter 5, the main results related to fertilizer recommendations for the main crops in Syria are presented. These res\-l-:l.t:s are based on the ideal situation of unlimited fertilizer supplies, iven the prevailing prices. The chapter includes a summa 8 1')’ of the estimated production functions, which constitute the basis for lO calculating the economically optimum fertilizer rates. The economic feasibility of these proposed rates and their aggregate economic impact are: then compared to the current recommended rates. Chapter 6 addresses the issues of fertilizer allocation under lit-11 ted fertilizer supplies. Given the constrained optimization nature of the problem, most of the analyses in this chapter are based on the fertilizer allocation model discussed in chapter 3. The results of the model are used to determine the economic impact of the current Cons traints on fertilizer supplies and to compare alternative fertilizer 8]. 1 ocation strategies . The concluding chapter 7 includes a summary of the main findings and. their policy implications, followed by recommendations for future res e arch. CHAPTER 2 THE PRESENT FERTILIZER SITUATION 1 This chapter gives an overview of the fertilizer situation in Syria. Its purpose is to put the specific research objectives of this study into the broader context of recent developments in the fertilizer s\-1t>-sector in Syria. The chapter begins with a brief description of current cropping patterns in the various agroclimatic regions of Syria. “118 is followed by an overview of recent trends in fertilizer cotisumption, production, imports, and prices. Next, the institutional setting for planning fertilizer allocation policies is discussed. This 13 followed by a section on fertilizer marketing, including a brief discussion of the role of the parallel market. The chapter concludes with a discussion of the available information on some of the fertilizer st:I‘ditmgies used by farmers in Syria. 2-1 W1; Syria occupies an area of 18.5 million ha, the majority of which is 8 ituated in the dry and semi-dry regions. The area of arable land is 6' 1 trillion ha, representing 331 of total area. Only 5.6 million ha are \ 1 Unless otherwise mentioned, all the information referred to in this it chapter is based either on personal communications with officials Geo“ the Soils Directorate, the Agricultural Cooperative Bank, and the thze ral Fertilizer Company, or on unpublished internal documents from above institutions. See also El-Hajj (1985b). ll 12 actually cultivated, representing 921 of total arable land. Water re:as=«¢:»urces are very limited in Syria, with 695 thousand ha irrigated land rep resenting 11.42 of total arable land (Annual Agricultural Statistical Abs tracts, 1989). Hence, rainfed agriculture is the predominant type in Syr 1a and it is characterized by great fluctuations in annual yields due to the variability in rainfall. Syria is subdivided into five agroclimatic regions or Agricultural Stability Zones, based on average seasonal rainfall (see Figure 1). The first region is further divided into two sub-zones (ICARDA/FSR, 1979, I>-‘¢3~]L): Zone la: Average rainfall over 600 mm. A wide range of crops may be grown here. Fallowing is not necessary. Zone lb: Average rainfall between 350 and 600 mm, with at least 300 mm in two thirds of the years surveyed. The main crops are wheat, food legumes, and summer crops. Zone 2: Average rainfall between 250 and 350 mm, with at least 250 mm in two thirds of the years surveyed. Barley, wheat, food legumes, and summer crops are grown. Zone 3: Average rainfall over 250 mm, with this level achieved in at least half the years surveyed. Barley is the principal crop but some food legumes also produced. Zone 4: Average rainfall between 200 and 250 mm, and at least 200 mm during half the years surveyed. Barley is the predominant crop. This area is also used for grazing. Zone 5: Covers the rest of the country. This steppe land is not suitable for rainfed agriculture but parts of it can be used for winter pasturing. The above system of Agricultural Stability Zones was introduced in 197 5 . This is part of the overall effort by the government to regulate ‘gricultural production better through medium- and long-term planning of the agricultural sector. This central planning effort involves a H me .o .mnoa .Mmmu omo.on 5.0 mnm.am -- ooo.e onu.na omm.~a N.o «an.nq ”amouu Ado oom.oH~ m.o 0mm.HH -- -- -- 0mm.HH H.- cma.mo~ "moose aefiuumscsH mmH.moH m.H omc.om How mmn.m Ome.mm mnm.mH «.0 man.mq H«mono owuuom ~m~.~q o.H mmH.eq -- mum.o oom.mH coo.oa q.o noo.n "moadwed cacao poem who.0mu o.m ooo.mm~ -- mmo.n mm~.m~ nmw.mma o.H who.~H "coasmoa :«euo cook so~.-~.e n.mm oeH.mmn.m ~H¢.¢~e Hn~.onm o¢~.m~u muo.~m~.a m~n.moe m.ee eom.~ected to reach 400 thousand tons and that of P205 would exceed 300 t1'"xotasand tons, or 1601 and 2301 increase over 1990 levels, respectively. Tab 18 2.3: ACTUAL AND PREDICTED FERTILIZER CONSUMPTION Syria, 1980 to 2000. \ Year 1980 1985 1990 1995 2000 \ (Thousand tons) N Actual 79.2 126.3 153.6 na na Predicted1 77.1 116.5 176.0 266.0 401.8 P20 5 Actual 44 . 9 74 . 3 91 . 6 na na Predicted1 43.1 70.3 114.5 186.5 303.9 K20 Actual 3.5 5.7 4.4 na na \ Predicted1 4.0 5.7 8.1 11.5 16.4 1/ Based on equations 2.5, 2.6, and 2.7 24 2 - 3 WM Faced with the rapid growth in fertilizer consumption the Syrian government has invested heavily into expanding fertilizer production capacity. These investments were encouraged by the discovery in the late 1960's of substantial deposits of rock phosphate, oil, and natural gas that can be used as raw materials for fertilizer production. Currently, three fertilizer plants are operating at a complex near the (:1 ty of Roma in central Syria: 1.- A plant (originally built in 1972) producing Calcium Ammonium Nitrate (CAN) with 261 N, was upgraded in 1984 to produce 301 N. The max imum capacity of the CAN plant is 36,000 tons of N per year. 2- A urea plant, operating since 1981, has a maximum capacity of 267 - 750 tons urea or 123,165 tons N per year. 3- A Triple Superphosphate (TSP) plant, operating since 1981, has a m"=13'Ctlmum capacity of 315,000 tons TSP or 144,900 tons P205 per year. Tab 16 2.4: FERTILIZER paonucnon CAPACITY AND ACTUAL PRODUCTION LEVELS Syria, 1981/82 to 1988/89. CAN Urea TSP Y 1 of 1 of 1 of earl Tons N Capacity Tons N Capacity Tons P205 Capacity .___“______ _________ _______ ________ 38 1/82 22,535 61.7 14,443 11.7 59,511 41.1 1982/33 27,224 74.6 51,779 42.0 45,752 31.6 1983/84 30,572 83.8 73,725 59.9 69,515 48.0 1984/35 30,561 83.7 88,366 71.7 70,997 49.0 198 5135 31,031 85.0 88,406 71.8 86,477 59.7 198 6137 32,883 90.1 71,171 57.8 79,019 54.5 1987/88 22,376 61.3 12,122 9.8 57,779 39.9 38/39 34,690 95.0 46,573 37.8 30,005 20.7 Cap Q 0 \city. 36,500 123,165 144,900 El, Fertilizer year: from July lst to June 30th. Sources: Soils Directorate internal documents. _+— 25 These plants have the potential of covering all the domestic N and P205 needs and to be able to export P205. However, a multitude of design, technical, and managerial problems and shortages in spare parts have lead to total production much lower than maximum capacity. As shown in Table 2.4, only the CAN plant has maintained relatively high production levels, with actual production representing more than 801 of capacity after the upgrading of the plant in 1984. The one exception was 1987/88, when the ammonia/urea unit was being modified to operate on natural gas instead of naphtha. After a relatively slow start, production at the urea plant rap idly increased to reach a peak of 721 of production capacity in 1984/85. Technical problems prevented higher production levels and the shift from naphtha to natural gas in 1987/88 caused a further, though tearE>orary, decline in production. Hence, total N production (CAN plus urea) declined from a peak level of around 751 of production capacity in 1985/86 to 21.71 in 1987/88. In 1988/89, after the modifications in the ammonia/urea unit were completed, total N production recovered somewhat but total production figures were still very low, 51.11 of total N capac ity. Technical problems at the TSP plant emerged shortly after it began operating in 1982. These serious problems prevented actual production levels from exceeding 601 of production capacity. They have become more serious in recent years leading to frequent shutdowns of the TSP plant for Prolonged periods of time. This is clearly implied by the gradually declining production figures presented in Table 2.4, with actual prodUCtion declining to a mere 30 thousand tons of P205 (201 of °apa¢ity) in 1988/89. 26 2-4 W Prior to the dramatic expansion in domestic fertilizer production capacity in the early 1980's, Syria depended on imports for a large proportion of its fertilizer needs. The construction of the urea and TSP plants had the goal of making Syria self-sufficient in fertilizer production and being able to sell fertilizer on the export market. In fact, in 1983 Syria exported 6,000 tons of urea to Burma, 5,700 tons of (Egzgilize; The following year urea to Jordan, and 15,500 tons of TSP to Iran W, No. 220, 13 Feb. 1986, pp. 13 to 17). urea exports increased, with 6,000 tons shipped to Burma and India and 19 , 200 tons to Iran, while TSP exports were negligible. Tab 1e 2.5: FERTILIZER IMPORTS: Syria, 1984/85 to 1988/89 N P205 Use Imports Imports as Use Imports Imports as Year‘ (tons) (tons) 1 of use (tons) (tons) 1 of use \— 19 84/85 126,279 16,415 13.0 74,263 4,324 5.8 98 5/86 136,994 15,235 11.1 85,588 13,064 15.3 1986/87 143,911 29,912 20.8 95,530 -- -- 1987/88 158,391 154,409 97.5 99,774 51,036 51.2 388/89 160,604 71,161 44.3 109,271 92,678 84.8 %/ Fertilizer year: from July 1st to June 30th. / Differences between use and imports do not necessarily correspond 3 to production figures because of inventories. / Sources: Soils Directorate, internal documents. These export figures were far below Syria's ambitious plans to exI:- trt 60,000 tons of fertilizer in 1984, because of technical D eruction problems and the growth in Syrian fertilizer consumption. By 1 985 Syria had reverted to importing substantial amounts of fertilizer. 27 In 1984/85, fertilizer imports accounted for 131 of total N and 5.81 of total P20, use. By 1988/89, these percentages had increased to 44.31 for N and 84.81 for P205 (Table 2.5). 2-5 humanities: Fertilizer producer prices for the entire country are set by the government. In spite of high annual rates of inflation (approximately 602 in 19871), fertilizer official (nominal) prices remained unchanged throughout the period 1984/85 to 1986/87 (see Table 2.6). Thus, the SOVernment covered the growing gap between the cost of production (or 1lllt>ort costs) and sales prices. Starting in 1987/88 the government dec ided to gradually increase fertilizer prices with the aim of reaching the true costs of production in the near future. Official prices of crops were also increased to offset the higher fertilizer prices. The data in Table 2.6 show that, between 1986/87 and 1989/90, the no“ inal price of calcium ammonium nitrate (301 N) increased by 4051 and a“I'llonium nitrate (33.51 N) increased by 3981. Also, the nominal prices of urea, TSP and potassium sulfate increased by 5161, 5201 and 8101, respectively. However, when measured in real terms, all fertilizer prices in 1988/89 were still below their 1984/85 levels. In 1989/90 the Twatantial increases in nominal prices and the decline in inflation rates resulted in real fertilizer prices being above their 1984/85 19"els. In 1989/90, the real prices of ammonium nitrates were about 51 higl‘aer than their 1984/85 levels, while the real prices of urea and TSP thet‘eased by about 351 during that period. The largest increase was in \ (C Unofficial sources claimed a 1001 rate was more realistic owzltt, 1991, p. 769). 28 §-al cropped area, changes in the cropping plan, and adjustments in f e ‘7 tilizer recommendations . 33 In 1984/1985, the SD estimated ideal aggregate fertilizer requirements at 240 thousand tons N, 200 thousand tons P205, and 60 thousand tons K20 (El-Hajj, 1985, p. 3). Actual fertilizer used during that year represented 531 of the ideal requirements for N, 371 for P205, and only 101 for K20. Based on these estimates of ideal requirements, the SD recommended a target of 201 annual growth rate in fertilizer use for the 1985-1990 five-year plan. If this target was achieved, ideal N levels were expected to be reached by 1988/89, and ideal P205 levels by 1990/91 (ibid., p. 18). Based on the above goals, estimation of planned fertilizer requirements for the following year were based on actual fertilizer Ba lea during the current year plus 20 percent. This figure is further adj usted by adding 151 to allow for enough contingency stocks to handle the possibility for increased demand during good seasons and to address Possible shortfalls in production or delays in imports. Hence, the following general rule was used: Planned Requirements including contingency stocks - (Current season's sales + 201 annual growth rate) * 1.15 - 1381 of current season's sales According to the above rule, given 126,297 tons of actual N sales “I‘d. 74,262 tons of P205 sales in 1984/85, the planned N requirements for 1985/86 were estimated at 174,200 tons (i.e., 126,297 x 1.38) and that Of P20, at 102,500 tons. Starting in 1987/88, the above rule was nod ified. Instead of basing the estimates on previous season's sales, thfi current approach is based on satisfying the ideal N and P205 taQuirements of all field crops, 331 of ideal K20 requirements, and 751 Oe N and P20, requirements for fruit trees. Thus, for the 1989/90 8 Q“-.-on, the ideal requirements were estimated at 328 thousand tons of N, 34 222 thousand tons of P205, and 70 thousand tons of K20. In contrast, the planned requirements were 306 thousand tons of N, 210 thousand tons of P205, and 23 thousand tons of K20 (Table 2.8). Table 2.8: ESTIMATION OF PLANNED FERTILIZER REQUIREMENTS Syria, 1989/90 N P205 K20 W1 (tons): Crops: 237,900 172,927 17,358 Fruit trees: 90,407 48,713 52,484 To tal: 328 , 307 221,641 69,843 a ned ements (tons): Crops: 237,900 172,927 5,728 Fruit trees: 67,805 36,535 17,320 To tal: 305,705 209,462 23,048 14" See Appendix B for detailed estimation of ideal fertilizer requirements for 1989/90. The shift to this new system, in conjunction with the reduced 1e\rels of domestic production and constraints on imports, has led to a growth in planned requirements while at the same time actual use has (lee lined. Discrepancies between planned and actual fertilizer use have become more accentuated in recent years. The percentage of the planned N requirements actually used declined steadily from 911 in 1984/85 to 50x in 1989/90 (Table 2.9). A similar trend was also observed for P205 (89 1 in 1984/85 as compared to 441 in 1989/90). In the case of K20, the arch]; was not as sharp. The percentage of planned requirements actually gel-1 ieved declined from 691 in 1984/85 to 471 in 1988/89. In 1989/90, t hfi cancellation of all potassium import contracts drastically limited lQ distribution of potassium fertilizers, with supplies coming from e x 1 sting stocks . t 35 Table 2.9: PIANNED AND ACTUAL FERTILIZER REQUIREMENTS Syria, 1986/85 to 1989/90. N REQUIREMENTS Planned 1 Actual 1 Actual as Year1 (tons) Increase (tons) Increase 1 of Planned 1 984/85 138,500 -- 126,297 -- 91.2 1 9 85/86 174,200 25.8 136,996 8.5 78.6 1 9 86/87 187,900 7.9 143,911 5.0 76.6 1 9 87/88 260,000 27.7 158,391 10.1 66.0 1 9 88/89 290,000 20.8 160,606 6.6 55.6 1 9 89/90 306,000 5.5 153,565 ~4.4 50.2 p20, REQUIREMENTS Planned 1 Actual 1 Actual as Ye ar (tons) Increase (tons) Increase 1 of Planned 19 84/85 83,500 -- 76,262 -- 88.9 19 85/86 102,500 22.7 85,588 15.2 83.5 19 86/87 118,100 15.2 95,530 11.6 80.9 19 87/88 147,000 26.5 99,776 4.4 67.9 19 88/89 205,000 39.5 109,271 9.5 53.3 19 8 9/90 210,000 2.6 91,593 -16.2 63.6 K20 REQUIREMENTS Planned 1 Actual 1 Actual as Year (tons) Increase (tons) Increase 1 of Planned \_ __ __ 19 86/85 8,200 -- 5,678 -- 69.6 198 5/86 8,200 0.0 6,182 8.9 75.6 1986/87 9,250 12.8 6,806 10.1 73.6 198 7/88 18,000 96.6 9,605 38.2 52.2 1988/89 22,250 23.6 10,567 12.1 67.6 889/90 23,000 3.6 6,358 -58.7 18.9 1/ . Fertilizer year. from July lst to June 30th. figc"4l:rces: Soils Directorate internal documents. 36 It should 'be noted that the above mentioned drastic decline occurred in spite of a gradual increase in actual total fertilizer use (see Table 2.9). Thus, the widening gap between planned and actual use was partly due to the rapid increase in planned requirements. This is especially true in 1987/88 and 1988/89 because of the change in procedures to estimate fertilizer requirements. 2.7 W 2.7-1 Wills): The Soils Directorate defines fertilizer availability as the amount of fertilizer, domestically produced or imported, which is available for distribution to farmers. Availability is usually expressed in terms of the percentage of planned requirements actually available for distribution. In general the figures on fertilizer availability overestimate the true capabilities of the .Agricultural Cooperative Bank (ACB) to satisfy the quantities demanded by farmers. This is because production and imports are not always available to meet the demand of peak periods. Reduced fertilizer availability has become a chronic problem that Syrian policy makers are faced with almost every year. These problems have tended to be much more serious for fertilizer use on winter crops (primarily wheat, barley, food legumes, and fall-planted sugar beets and potatoes). The peak period of fertilizer demand for winter crops, especially phosphate, occurs very early in the season (October- December). Thus, any delays in importing fertilizers and/or any early disruptions in domestic production would cause serious reductions in fertilizer use for the fall-planted winter crops. 37 In fact, for the past few years the impact of fertilizer shortages was mostly felt during the winter season, whereas summer crops usually receive all their fertilizer requirements. This situation is partly caused by the delays in the development of the annual agricultural plan, which is rarely finalized before the end of August. Thus, officials are usually left with only one month to plan all the details for fertilizer distribution to fall-planted crops. For instance, in the 1989/90 season the total amounts of N fertilizer expected to be available for the winter season was estimated at 127 thousand tons. This is only 731 of the 175 thousand tons planned for the winter season. The situation for P205 was even more serious, with only 97.5 thousand tons available, representing 611 of the planned needs for the winter season. 2.7.2 W Given the centrally planned nature of fertilizer marketing and the highly subsidized official prices, the government has relied on fertilizer rationing to address chronic fertilizer shortages. Each farmer is issued a quantity of fertilizers based on the type of crops grown and total area planted to each crop. The per hectare allocation for each crop is determined by the SD, and it varies from year to year depending on the total amounts of fertilizer available at the beginning of the season. Estimations of fertilizer requirements for the coming season are usually finalized in July of every year. Actual fertilizer applications on fall-planted crops usually begin with the sowing of rainfed cereals starting in October. Thus, to ensure that farmers have timely access to fertilizers, allocation and distribution decisions for fall-planted 38 crops have to be finalized by early October. At that time the officials at the SD have acquired sufficient information to make realistic predictions about the actual amounts of fertilizer likely to be available for the winter season. These predictions are based on existing stocks, on the most recent production figures from the fertilizer plants, and on procurement contracts signed with international fertilizer suppliers. Having relatively accurate estimates of' the actual fertilizer available at the beginning of the winter season, the next issue is how to 1allocate the limited. fertilizer resources to the ‘various ‘winter crops. Fertilizer allocation strategies adopted by the SD tend to vary from year to year. However, one principle that seems to be common in these strategies is to meet the fertilizer requirements for the ”strategic crops“, i.e., either those that provide substantial export earnings (such as cotton) or those substituting for the major food imports (such as wheat). Whether crops are irrigated or rainfed is another important factor influencing fertilizer allocation decisions. Generally irrigated crops have their fertilizer requirements met because the economic returns to fertilizer use by rainfed crops are lower and more risky than for irrigated crops. The impact of the above strategies for fertilizer allocation to. winter crops is that irrigated crops, including irrigated wheat, usually receive all their fertilizer requirements. As fer rainfed crops, the first priority goes to high-yielding (HYV) wheat varieties. Of lower priority are local (LYV) wheat varieties and food legumes, with barley having the lowest priority. Given the general fertilizer shortfall for the winter crops, fertilizer is rarely allocated to barley. In fact, 39 the first time fertilizer was allocated to barley was in the 1986/87 season. However, this allocation excluded barley in Zone 3. 2.8 Wins 2.8.1 W021; The Agricultural Cooperative Bank (ACB) is the sole legal distributor of chemical fertilizers in Syria. Based on instructions from the SD, the ACB issues individual licenses specifying the amounts of fertilizer allocated to each farmer based on the type of crops grown, area planted to each crop, whether these crops are irrigated or rainfed, and the Agricultural Stability Zone where the farm is located. All fertilizer purchases are made based on interest-free short-term loans by the ACB, with re-payment after harvest. The ACB operates a network of 68 local branches covering most of the agricultural regions in Syria. The government provides the required facilities to open up needed branches. Fertilizer and other input stocks at the ACB branches are monitored on a weekly basis and sales on a monthly basis. The data are entered in a central computer at the ACB headquarters in Damascus. This allows the ACB to direct the flow of fertilizers towards the branches with the highest demand. The ACB has also expanded its network of warehouses. The total storage space is 93.2 thousand square meters, with a capacity of 233 thousand tons. They are used for various agricultural inputs distributed by the AGB (fertilizers, seeds, pesticides, bags, and so forth), with fertilizers being the largest user of space (196,535 tons in October 1989). The ACB can lease extra storage space if the need for storage arises. 4O 2.8-2 W Private trade in fertilizers is illegal in Syrian However, discussions with farmers clearly indicate that a large number of farmers purchase part of their fertilizer needs from the local or regional parallel market. Very little information is available regarding the parallel market sources of supply. However, based on informal interviews with farmers and local traders, it was possible to identify the following possible sources of supply: 1. fitnint_A§fl_Q£fitinl§, especially in the provincial branches, have enough discretionary power to divert some of the fertilizer under their authority to be sold in the parallel market. A typical arrangement would involve a provincial ACB senior official selling the fertilizer to a local influential notable or politician. The latter would then use his influence to protect the corrupt official. Another variant is the practice of local traders who bribe officials to obtain licenses allowing them to receive very large quantities of fertilizer at the official price. 2. lgnigt_AQfl_gmnlgygg§ are able to issue farmers only a portion of their fertilizer entitlements and sell the rest on the parallel market. This practice is facilitated by the illiteracy of many farmers and their inability to understand the complex fertilizer distribution system. 3. u o t v : Another reportedly major “leakage“ in official distribution channels takes place at the agricultural cooperative level. Cooperatives are treated as a single unit by the ACB. Thus, the entire fertilizer ration is received by the head of the cooperative who would then allocate it among the other 41 members and in the process be able to sell some fertilizer on the parallel market. 4. Egtmgtfi: There is enough evidence to indicate that at least some farmers also sell part of their fertilizer rations on the parallel market. Given the inaccuracy of fertilizer allocation procedures, it is very likely that many farmers would receive rations in excess of the amounts they had intended to apply. Thus, they may sell the rest on the market or to neighboring farmers. However, interviews with farmers suggest that they would prefer to store any excess fertilizer for future use rather than selling it. Moreover, the use of the parallel market prices as the base of fertilizer transactions between farmers is viewed by most farmers as highly unethical. Hence, most of these transactions tend to be in the form of barter exchange or short-term borrowing. Therefore, the available evidence suggests that fertilizer sales by farmers represent only a minor source of supplies to the parallel market . In summary, two factors seem to be the most important determinants of fertilizer supply in the parallel market. The first factor is the total quantity of fertilizer available for distribution by the ACB. Since the majority of market supplies are suspected to originate from the ACB, it can be assumed that only a certain proportion finds its way to the market. It is virtually impossible to estimate the magnitude of this 'leakage". However one can assume there is a positive relationship between the two. Thus, for a given level of quantity demanded, an increase in fertilizer availability would translate into an increase in parallel market supplies and, hence, would result in lower market 42 prices. The second important factor is the degree of government monitoring and enforcement of the distribution activities of the ACB. Therefore, if the government decides to increase its vigilance, the proportion of total fertilizer sold on the parallel market would decline. Information on fertilizer prices of the parallel market is limited and sketchy, with figures ranging from 151 to 351 above official prices. One of the most important factors affecting market prices is rainfall. With higher rainfall farmers' demand for fertilizer increases. During exceptionally good years, such as in 1987/88, market prices were 50 to 1001 higher than official prices. In Contrast, the margins between market and official prices were less than 101 in dry years (e.g., 1988/89). During the 1989/90 season, after two consecutive dry years, farmers reduced their fertilizer use to such an extent that very few of them had to buy additional fertilizer from the parallel market. Thus, the parallel market was virtually non-existent during that year. Another important determinant of the quantity of fertilizer demanded from the parallel market is fertilizer availability from the official channels. Since market demand represents aggregate excess demand, any increase in fertilizer availability would reduce fertilizer shortages at the farm level and, hence, would reduce market demand Finally, it is important to mention that any increase in agricultural product prices (official or market prices) would also increase farmers' demand for fertilizers. This would be reflected into increased market demand and higher market prices. 43 2.9 ' e e Very little information is available on farmers' decision making in relation to fertilizer use. Some information is available from farmers' surveys in northern Syria undertaken by ICARDA since the late 1970's. These surveys were mostly designed to analyze farmers' cultural practices in general, with some questions directed towards fertilizer use. More recently, a detailed study by Meri Whitaker (1990) focused on farmers' fertilizer strategies in northern Syria. This case study focused on nitrogen fertilization on rainfed wheat. The following discussion will rely heavily on Whitaker's results, in addition to personal discussions with farmers and officials from the Soils Directorate. First, it is important to note that although fertilizer allocations are based on the assumption that farmers will actually apply the SD recommended rates for each crop, the majority of farmers make their fertilization decisions independently of the SD or the local extension agent. Some policy makers have seriously considered making it compulsory for farmers to apply the recommended rates. Such a proposal was briefly discussed in a meeting by the Higher Council on Agriculture (HCA) in 1985. However, it was rejected due to opposition by the General Peasant Union representative (El-Hajj, 1985, pp. 19,20). Moreover, although most wheat farmers surveyed have used fertilizer for 10 to 15 years, most of them had little information about the recommended rates or the official fertilizer allocations for each crop. They take whatever is allocated to them and rely on their own experiences, and others, to decide on what fertilizer strategies to employ. 44 When farmers have to make decisions about fertilizer use they need to address the allocation of fertilizer between crops, the number of applications per crop, in addition to the rates, timing, and method of applications. All of these decisions are made under a highly uncertain environment characterized by wide year-to-year variations in rainfall levels and in seasonal distribution. Whitaker identified several fertilizer strategies adopted by rainfed wheat farmers in northern Syria. First, practically all farmers surveyed have indicated that they applied.P505 only once, at the time of planting (mid-November to early December). Average rates used in the wetter areas (Zone 1) are almost twice as large as the rates applied in the drier regions (Zones 2 and 3). These rates vary very little from year-to-year given that.1505 is applied at the beginning of the growing season, when future rainfall is unknown. Unlike with P305 application, farmers have greater flexibility with N application. This allows them to modify their strategies based upon rainfall levels. Wheat farmers in Zone 1 generally apply two N applications. The first application is done at planting, while the second one is applied around tillering time (end of February). As with 1505 rates, N rates at planting time show relatively little variation. The rates of N for the second application depend essentially on rainfall levels during the first half of the growing season (October to February). If prior rainfall is considered normal, then farmers usually apply a rate twice as large as the first N application. This rate may be cut by one third if rainfall is below average, or increased by up to 501 in a wet year. This depends essentially on the level of previous 45 rains and on farmers' expectations about rainfall during the second half of the growing season (early March to early May). In the drier zones, farmers are less likely to apply N at planting than in Zone 1, and they tend to use much lower rates. This is a sensible strategy considering uncertain weather. During the growing season, farmers usually apply one to two additional N applications. The number of N applications during the growing season and the rates used per application represent the most important variables that farmers can manipulate to adjust to the amount of rain received. In a normal year, most farmers reported applying at least one additional N application around the end of January. But if the year was dry, two thirds of the farmers would not apply additional N during the growing season. In a wet year, more than 801 of survey farmers reported applying the same rate of N applied in normal years. However, more than half added a third application, about a month later. Therefore, in the case of rainfed wheat (and rainfed crops in general), rainfall levels and seasonal distribution constitute the main determinant of fertilizer strategies adopted by farmers. In addition to weather uncertainty, uncertainty about government fertilizer policy also plays an important role in shaping farmers' strategies. Delays in fertilizer distribution and the size of the allocations are two important factors that influenced farmers' strategies. One of the most frequently stated complaint is that they often have to delay sowing their cereals or to plant without fertilization due to delays in fertilizer distribution. Delays also occur during the growing season. In fact many farmers have reported that they would increase the rates and/or the number of N applications during the growing season if the 46 second distribution of fertilizer (allocations for fruit trees and summer crops) was done earlier. A number of farmers also mentioned that they seldom receive their full allocation. Thus, they often resort to diverting fertilizer from other crops (e.g., olives) or to buying from the parallel market. CHAPTER 3 THE RESEARCH APPROACH The main objective of this research is the development of a model for determining economically optimum allocations of limited fertilizer supplies in Syria. The model also will be used to compare alternative fertilizer allocation strategies and to assess the economic impact of limited supplies on national and farm incomes. The purpose of this chapter is to present the conceptual framework for this model and to outline the main steps in the research approach. The chapter starts with a presentation of the basic microeconomic models of unconstrained and constrained optimization at the farm level. This is followed by a discussion of the main issues related to the extension. of the farm-level approach to the problem. of fertilizer allocation at the national level. Finally, the last section outlines the main steps in the specification of the fertilizer allocation model and its underlying assumptions. 3.1 8 ve 3-1.1 W The basic problem of how much fertilizer a farmer should apply to a given crop can be presented based on the standard neoclassical static model of profit maximization. This model assumes that the only criterion guiding the farmer's decision is that of maximizing profits, 47 48 or net returns, from the use of fertilizers. The model also assumes that the farmer is risk—neutral and has perfect knowledge about input and output prices and about the relationship between the level of fertilizer applied and yield. Such a relationship describes the rate at which fertilizers are transformed into crop output, and it is often referred to as a yield response function or production function (Doll and Orazem, 1984, p. 20). Assuming a yield response function to nitrogen (N) and phosphorus (P) fertilizer applications’, this function can be written in the following general form: Y - F(N.P. X. 2) (3.1) where Y is the per ha yield; N and P are the per ha fertilizer application rates; X refers to other variable inputs; 2 refers to environmental factors; Based on this production function, the farmer's net returns to fertilizer application can be expressed as NR-P,*(Y-Y°) - (Wni'N) - (Wp'kP) - (Tvc -TVC°) where NR are net returns to fertilizer use per ha; Y0 is the per ha yield of the unfertilized treatment; P, is output price; W5 and Wg are fertilizer prices; TVC are total variable costs other than fertilizer costs; TVC° are total variable costs in the unfertilized treatment; Assuming that TVC and TVC° are equal, these two terms can then be dropped from the net return equation to give NR-P,*(Y-Y°)-(WD*N)-(WP*P) (3.2) 1 Potassium is not included in the analysis given its limited use in Syria. 49 Such an assumption may not be realistic given that fertilizer use is often accompanied ‘by an increase in other variable costs such harvesting and transfer costs associated with the additional output due to fertilizer application. Such variable costs that are proportional to yield can be implicitly incorporated into the net returns equation by adjusting output prices to reflect these costs. A detailed discussion of these cost adjustments is presented in chapter 4. Although eq. 3.2 is the simplest form for’ modeling farmers' decision making, it is not the most accurate. This is particularly true if' the carry-over' effect: of' the applied. fertilizers is significant enough to affect the decision making process. A more accurate modeling of the net return equation needs to incorporate this carry-over effect, particularly in the case of phosphate and potassium fertilizers. Such fertilizer carry-over models were suggested by Kennedy et a1. (1973), Stauber, Burt, and Linse (1975), Dillon (1977), Taylor (1983), Smith and Umali (1984), Kennedy (1986 and 1988), Lanzer, Paris, and Williams (1987), and Segara et al. (1989). Although these models would greatly improve the accuracy of the results, they require more detailed biological data such as the content and availability of soil nutrients before planting and the patterns of nutrient uptake by the crop during the growing season. Such data are very limited in Syria and their quality is questionable given the inaccuracy of the soil testing procedures. Therefore, the analysis throughout this dissertation will be essentially based on the simple profit maximization model given by eq. 3.2. To calculate the fertilizer rates that would maximize net returns from fertilizer use (NR), the first partial derivatives of the profit 50 function with respect to N and P, aNR/aN and aNR/aP, are set equal to zero (first order or necessary conditions), as follows: aNR/aN - (P, * aY/aN) - w, - 0 (3.3) aNR/aP - (P, * aY/ap) - u - 0 (3.4) P Defining the Value of the Marginal Product (VMP) as the value of the increase in output due to the addition of one unit of fertilizer, eq. 3.3 and 3.4 can be rewritten as follows: VMPB - Wn (3.5) and VMPp - W (3.6) 9 Equations 3.5 and 3.6 represent the necessary conditions for profit maximization, i.e., the cost of the last unit of fertilizer added should be equal to the returns from the yield increase due to the addition of that unit (Doll and Orazem, 1984, p. 183). The optimum fertilizer rates, N' and P', are calculated by solving simultaneously a system of two equations (eq. 3.3 and 3.4) with two unknowns (N and P). Total fertilizer requirements can then be computed by multiplying the total area to be fertilized by the calculated optimum rates. If the farmer is growing several crops, then the above procedure is repeated for all crops, assuming the farmer has perfect knowledge of the yield response function for each individual crop. 3.1.2 WWW As mentioned earlier, the above optimization model is based on the assumption that the only criterion guiding the farmer's decision is that of maximizing net returns from the use of fertilizers. In reality, however, farmers' decision criteria are much more complex. These 51 criteria are usually affected by many factors, including (FAO, 1984, p. 132) the anticipated yield increase, expected crop prices, cost and availability of fertilizers, level of financial resources and credit availability, land tenure considerations, the degree of risk and uncertainty and the farmer's ability to bear them. Given the above factors, farmers are expected to be cautious when deciding the fertilizer rates to apply, by building in a fair safety margin in their profitability calculations. Two measures, or indicators, of profitability are commonly used. The first one is the marginal rate of return (MRR), which is defined as (CIMMYT, 1988, p. 32) ... the marginal net benefit (i.e., the change in net benefits) divided by the marginal cost (i.e., the change in costs), expressed as a percentage. CIMMYT (1988, p. 35) suggests, as a general rule, that farmers will not use fertilizer beyond the point at which the MRR is at least 501 in one crop season. A similar rule of thumb is suggested by FAO (1981, p. 41) whereby the minimum acceptable value for MRR is set at 401. Such minimum MRR rules are frequently used in practice to set recommendations that are considered to reduce the risk of not obtaining a yield response to the last increments of fertilizer (see, for example, Josephson and Zbeetnoff, 1988). However, in assessing the profitability of fertilizer use in comparison to no fertilizer application, increases in yield, returns, and costs have to be viewed as non-marginal changes. Such economic evaluations of fertilizer use frequently rely on the Value-Cost-Ratio (VCR) as an indicator of profitability. The VCR is defined as the value of the yield increase due to fertilizer use (i.e., over the unfertilized treatment), divided by total 52 fertilizer costs (FAO, 1981, p. 42). The VCR value associated with the calculated optimum N-P combination can be calculated as follows: P,*(Y‘-Y°) VCR - (W. * N‘) + (W, * I") Where Y. is the yield obtained as a result of applying the optimum fertilizer rates (N' and P'). The VCR is an indicator of nvetngg returns to the investment in fertilizers. Thus its use as a basis for assessing profitability seems to be a theoretically weak approach since it is based on average rather than marginal comparisons of costs and returns. If the magnitude of the VCR. value is used. as a basis for comparing the profitability of alternative fertilizer rates, this could result in 'very’ misleading conclusions. For instance, the VCR value associated with very low fertilizer rates would be much larger than the VCR associated with higher rates, since fertilizer costs would be close to zero if the very low rates are applied. The most relevant conclusion that can be obtained from the VCR is whether its value is greater or smaller than one, with a value greater than one indicating that the investment in fertilizer use is profitable. However, in order to build in a safety margin, a VCR of 2.0 is typically used as the critical value for the profitability of fertilizer use in deve10ping countries (FAO, 1984, p. 132; Lele, Christiansen, and Kadiresan, 1989, p. 41). There seems to be no valid justification for using this critical value except for its widespread use in many international organizations, particularly the United Nations Food and Agriculture Organization (FAQ). This, in turn, has led to its adoption by national research centers in several developing countries. 53 In Syria, the Soils Directorate (SD) relies on the VCR as the predominant indicator of profitability in its economic analysis of fertilizer experiments. Therefore, in order to facilitate the communication of results to SD officials, the VCR will also be used in this study as the indicator of profitability of fertilizer use. The above general requirement of a VCR of at least 2.0 will be relaxed to allow for a minimum VCR of 1.5. This is justified by the fact that most VCR calculations do not explicitly incorporate additional transport and labor costs associated with increased fertilizer use or ‘91 th the resulting increased output (Lele, Christiansen, and Kadiresan, 1989, p. 41). If these additional costs are accounted for, then the Value of the yield increase would decline and fertilizer costs would inc rease. This would give a lower value for the minimum VCR than the 8““ggested value of 2.0. In this study most additional transport and labor costs will be explicitly included in the fertilizer and crop prices based on which VCR calculations are madel. Therefore, it would be reasonable to assume that the critical value for minimum VCR would be c]-<>ser to 1.5 rather than 2.0. 3.1-3 Win31 The wide variation in rainfall patterns prevailing in Syria 11introduces a large element of uncertainty into the farmer's decision l‘43‘ltjeng about fertilizer use. Although the sources of uncertainty 1 tI§1ude the variability in input and output prices, the most important \ 1 ea Refer to chapter 4 for details about the methods used in §1mat ing fertilizer and crop prices. ‘ 54 source of uncertainty in Syrian agriculture tends to be yield uncertainty due primarily to variability in rainfall. The farm management literature includes a broad range of methods wi th varying degrees of complexity that attempt to incorporate risk and uncertainty considerations into agricultural production analysis (see, for example, Anderson, Dillon and Hardaker, 1977; Antle, 1983; and Taha, 1987 , pp. 427-467). Boisvert and McCarl (1990) provided an extensive survey of the literature on the applications of risk modeling techniques in agriculture. Most of these techniques are direct or indirect app lications of expected utility theory, as developed by von Neumann and l“(>l'genstern (1947). The two most-utilized approaches in applying the expected utility t1"eox'y are the mean-variance (E-V) analysis and the stochastic dominance rules (da Cruz and da Fonseca Porto, 1988, p. 381). However, these a1"5’3Il‘oaches tend to be complex, require extensive data, and their results are difficult to interpret by non—economists. If the prime users of the I."><1e1's results are policy makers or farmers, then the main criterion for choosing a particular model is its simplicity and the ease with which the results can be explained to decision or policy makers (15‘3iimsmvert and McCarl, 1990, p. 44). There are less complex methods that have been used to examine agricultural risk. A common approach involves manipulating the values 0f the most uncertain parameters to evaluate the consequences of optilhistic and pessimistic scenarios (see, for example, Savoie and ‘59be 1980- , , and Adams, Hamilton, and McCarl, 1986). Such an approach constitute the main basis for treating risk and uncertainty in this 8t. y . This is justified by two basic concerns related to the decision IIIIIIIIl-m-._ 55 model proposed in this study: First, the proposed model should he s imple enough to be used by Syrian planners with limited economics or mathematical backgrounds; and, second, the results from the model should be easy to interpret and to explain to policy makers. As mentioned earlier, rainfall is the most uncertain parameter affecting the profitability of fertilizer use in Syria. The calculation of the VCR associated with optimum fertilizer rates is based on “average" yield response functions. However, farmers growing rainfed crops are usually more concerned with the profitability of their investments in fertilizers in the event of a dry year. Therefore, in assessing the profitability of the estimated optimum rates on rainfed crops, the 1.5 minimum-VCR criterion will be applied to VCR calculations based on estimates of yield increase due to fertilizer use in the event of a dry year. The question of what constitutes a 'dry" year is subjective. Recent surveys of farmers in northern Syria suggest that two to three years out of ten are considered as 'bad" years by the responding farmers (“litaken 1990). Based on an analysis of rainfall data from several IhetIeorological stations in northern Syria, it is possible to define four rainfall scenarios, 'good", “normal", 'dry", and “very dry“, for the three zones of interest (see Table 3.1). The inclusion of the ”very dry“ scenario seems also important a. J“tl'nough its probability of occurrence is only one in eight years. This 1.3 because farmers would probably decide not to use fertilizer rates t hat may result in financial losses in the event of a very dry year. In C) ther words, an additional criterion for assessing the profitability of ‘ 56 Tab 1e 3.1: DEFINITIONS OF RAINFALL SCENARIOS Syria, Agricultural Stability Zones 1b, 2, and 3. Ra. infall Scenario Scenario Definition Zone lb Zone 2 Zone 3 “Good" Mean (mm/year): 500 350 300 Probability (1): 29 32 35 'Nornal' Mean (mm/year): 400 300 250 Probability (1): 47 42 41 'Dry' Mean (mm/year): 350 250 200 Probability (1): 12 13 12 "Very dry' Mean (mm/year): 300 200 150 Probability (1): 12 l3 12 L/ Refer to chapter 2 for the definitions of the Agricultural Stability Zones. Sources: Derived from Agtitnitntgi Statistical, Ab§§£§££§. various years. a given N-P combination is the requirement of a minimum VCR of 1.0 in the event of a very dry year. It is possible to estimate the effect of the calculated optimum rates on yield and profitability under each of the above defined 8(:et‘larios. This can be done by including the level of rainfall as an e3°5>3L43natory variable in the production function‘. Assuming that r'afittlfall is the only environmental factor affecting yield (i.e., the "Z" variables in Equation 3.1), four different production functions as 8t>Qiated with the four rainfall scenarios can be estimated for each ta infed crop: \ ti“ 1 Refer to the discussion on the specification of production = tions later in this chapter. 57 Good year: Y5 - G(N,P,X) Normal year: Y. - N(N,P,X) Dry year: YD - D(N,P,X) Very Dry year: Y; - E(N,P,X) where, Y5 is yield in a good year; Y. is yield in a normal year; YD is yield in a dry year; Y; is yield in a very dry year. Optimum fertilizer rates (N' and P') will be computed based on the ”normal year“ production function (YN). These optimum rates are then included in the other three production functions to solve for the yield levels, Y's Y} Y}, that would be obtained under the corresponding scenarios. Thus, the above minimum VCR criteria can be written as fo 1 lows: PyG * (Y‘s ‘ Yoc) VCR,3 - 2 1.5 (3.7) (wn * u“) + (up * P') Pyll * (‘1‘: ' You) vca, - z 1.5 (3.8) (w, * N') + (up * P‘) P3,D * ()r‘D - Y°D) VCRD - 2 1.5 (3.9) (wn at u') + (wp * P') Pr: * (Y's ‘ Yon) VCR; - 2 1.0 (3.10) (w, * N') + (H9 * P') Vb are P”, P”, P” and P3,: refer to the output price under the various t Q 1 11 scenarios . In addition to the above minimum-VCR criteria, the analysis will Q1 QQ consider the effect of the optimum fertilizer rates on yields in (I b ‘5’ and very dry years. This is based on discussions with wheat and Q: ley farmers in the drier zones of northern Syria. When these farmers g 58 were asked for the reasons for not increasing their current application rates, a frequent response was that higher rates may "burn" the crop (i. - e., reduce yield) in the event of a dry year. These concerns can be included in the analysis by introducing a further criterion: the optimum N and P rates should not exceed the rates that would maximize yield ( i - e., Stage III of the classical production function) in dry and very dry years. This can be written as N. S N2” 5 ND.“ and P" s P,” 5 PD” where maximum N and P (Num and P‘w‘) are computed by setting the first partial derivatives of the production functions, YD and YB, equal to zero_ In looking at the issue of fertilizer use on rainfed crops, barley is a special case that deserves some further discussion. Fertilizer use on barley in Syria is still very limited, with recent surveys indicating that: less than 152 of farmers apply any fertilizer on barley (ICARDA/FRHP, 1988, p. 148). In comparison, most farmers apply fertilizer on wheat and, as indicated by recent survey findings, they have been doing that for an average of 10 to 15 years (Whitaker, 1990). Therefore, farmers probably consider barley fertilization as a r e:Latively new and untested technology whose profitability is not yet 1) 1:.°\’en. Moreover, barley is mostly grown in the drier zones where :- aihfall is extremely variable. This further exacerbates the uh QQ:rtainty characterizing farmers' decisions about fertilization. l“& 15 ' Q 1 ht‘ers would be primarily concerned with the fate of their investment in evaluating the profitability of fertilizer use on barley, §I|e event of a dry year. ¥ 59 Such concerns can be incorporated into the analysis by assuming a “worst-case scenario” approach, or what is often referred to in the literature as the 'maximin' assumption (see, for example, McInerney, 1967). According to this approach, rather than assuming average rainfall, risk-averse barley farmers would assume that a dry year would occur. Therefore, they would choose optimum fertilizer rates that would max 1mize their net returns based on yield increases expected during a dry year. In other words, the calculation of N' and P' for barley will be based on the "dry year“ production function, YD, rather than Y... 3.1.4 W The procedure outlined in the previous section for the calculation of Optimum N and P rates is based on unconstrained optimization. In othe 1' words it is assumed that if farmers are willing to pay the price, they could buy as much fertilizer as they wish and, thus, they would choose to apply the rates that would maximize their profits according to equa tions 3.3 and 3.4. Similarly, it is assumed that if farmers decide to apply the rates that would maximize their profits, the government is capable of supplying all the quantities of fertilizer demanded by f a“Tillers, either from domestic production or from imports. However, the above assumptions are unrealistic given the serious problems facing domestic fertilizer production, and given the limited availability of foreign exchange for the imports of fertilizers. Under Sth constraints the government can rarely supply 811 the fertilizer lgi‘.~‘ua‘t‘tities that farmers would demand at the existing official prices. N tead, farmers are issued fertilizer rations based on the crops grown Qua the area planted to each crop. Therefore, the problem faced by ¥ 60 farmers is how to allocate their fertilizer rations in such a way as to maximize their returns from the limited quantities of fertilizer available to them. The simple model of profit maximization outlined earlier can be modified to solve the farmer's constrained optimization problem. Assming that only two crops are grown, crop l and crop 2, and that the farmer has perfect knowledge about the production functions and about input and output prices, the constrained optimization problem can be formulated as follows: Maximize NR - P” * A1 * (Y1 - Y1°) + Pyz * A2 * (Y2 - Yzo) " Vn(N1 * A1 + N2 * A2) - wpw1 * A, + P; * A2) SUbjeCt to: N1 * A1 + N2 * A2 - NT P1 * A1 + P2 * A2 - Pr F(N1, P1, X1, 2) " Y1 C(Nz, P2, x2' 2) - Y2 where NR are net returns from fertilizer use for the whole farm; Y1 and ‘12 are the per ha yields of crop l and crop 2; Y1° and Y2° are the yields of the unfertilized treatments; F(N1, P1) and C(Nz, P2) are the production functions for Y1 and Y2; N1 and P1 are the per hectare fertilizer rates applied on crop 1; N2 and P2 are the per hectare fertilizer rates applied on crop 2; N, and P, are the total quantities of fertilizer available; 9,, and V, are fertilizer prices; P” and Pyz are output prices; A1 and A2 are the fertilized areas planted to crop l and crop 2; X1 and X; are variable inputs other than fertilizers; 2 refers to environmental factors. Such a problem is often set up in a form known as a Lagrangean tuneup" (1.). as follows:1 I. (N1. N2, P1, P2, 1,, 12) - P” * A1 * (Y, - Y1°) + P,2 * A2 * (Y2 - Y2°) - Vn(N1 * A, + N2 * A2) - N901 * A1 + P2 * A2) ' ‘1(N1*A1+N2*Az ' NT) ' 12 (P1*A1+ P2*A2 ' PT) (3.11) \ Q 1 The following section is adapted from Appendix II in Doll and zen (1984). ¥ 61 where 11 and 1.2 are the Lagrangean multipliers, defined as 1, - dL/dN-r and 12 - dL/dP, In other words, 1, and 12 represent the amount by which net returns would increase if N, or P, are increased by one unit. To maximize L, the first partial derivatives with respect to N1, P1 , N2, P2, 1,, and 12 are set equal to zero (first order conditions): aL/aN, - 1,0,, * aY,/aN, - w, - 1,) - o (3.12) aL/ap, - A,(P,, * aY,/ap, - w, - 1,) - o (3.13) aL/an, - 1,0,, * aY,/aN, - w, - 1,) - o (3.14) aL/ap, - 1,0,, * aY,/ap, - w, - 1,) - o (3.15) aL/a1,- - N,*A,+N,*A,+N,- o (3.16) aL/a1,- - P,*A,+ p,*A,+ P,-o (3.17) To calculate the optimum fertilizer rates on crop l and crop 2 (N1- , Pl', Nz', and Pz') the above six equations with six unknowns (N1, P1 . 11,, P2, 11, and 12) are solved simultaneously. In the above system of six equations, the first four (eq. 3.12 to 3 - 15) can be written as VMPn, pl 3 llll C at ‘13 a: °r mu, - mu, - w, + 1, (3.18) - VHsz - up + 12 (3.19) The above two equations reflect what is often referred to as the W for profit maximization, i.e., the marginal revenues t o the application of fertilizer should be equal across all crops. A 18° . according to this rule, these marginal revenues should be equal to the Illarginal cost of the applied fertilizer. However, under fertilizer ta bioning, the fertilizer purchase price does not represent its true that g11ml cost given that the farmer cannot buy any additional quantities ‘ 62 beyond the ration purchased from the government. The true value to the farmer, or shadow prices, of the limited fertilizers are given by (Wu + 11) and (Wp+ 12). These values would constitute the maximum prices that the farmer would be willing to pay if he could purchase additional quantities from the parallel market. Combining eq. 3.18 and 3.19 we get: VHF“, VMPnz VMPp, vupp, (3.20) w,+1, wn+1, wp+1, wp+1, Equation 3.20 reflects a more general statement of the equimarginal rule specifying that “the ratio of the value of the marginal product to the price of an input be equal for all inputs in all uses" (Doll and Orazem, p. 191). Under constrained optimization, this rule would require the use of the shadow prices of fertilizers, as in eq - 3.20. In contrast, the purchase price would be the relevant price if there are no limits on the quantities of fertilizer that the farmer could buy (the Lagrangean multipliers would be equal to zero). 3-2 0 8 08 V8 3.2.1 W The same conceptual approach for fertilizer allocation at the farm level can be used to model the fertilizer allocation problem at the natic“rial level. This can be done if all cropped areas in Syria are treated as a single farm with one decision maker responsible for e eiding on the optimum fertilizer rates that would maximize net returns to the limited aggregate fertilizer supplies. In other words, the two- 63 crop model of constrained optimization is extended to cover all crops grown in Syria. Such an approach to the fertilizer allocation problem at the national level was proposed in an FAO fertilizer manual (FAO, 1966, pp. 239-40). Based on this approach, the national fertilizer allocation model is formulated as follows: Maximize NR - E, [P,, * A, * (Y, - Y°,)] - 2, (W, * F“) (3.21) Subject to: 2, (17,, * A,) - F“- (3.22) N,(F1t, F21, ..., Fti’ X,, Z) - Y, (3.23) where NR is aggregate net return from fertilizer use in Syria Y, is the per ha yield of crop i; Y°, is the per ha yield of the unfertilized treatment; N,(F,,, F2,, ..., F“, X,, Z) is the production function for Y,; F1, is the per hectare application rate of fertilizer f on crop i; X, refers to inputs on crop i other than fertilizers; 2 refers to environmental factors; F“ are aggregate available supplies of fertilizer f; W, is the price of fertilizer f; P” is the price of crop i; A, is total fertilized area planted to crop i. The Lagrangean function would be set up in a similar way as in the f‘a‘rnl-level model, and optimum fertilizer rates on all crops would be calculated by simultaneously solving a system of (i*f + f) equations. This national fertilizer allocation model is based on several 8 imp 1 181210 assumptions , including: 1'- There exist adequate reliable data to estimate production functions for all the crops grown in Syria. 2‘. For each crop, the estimated production function is representative of the crop's growing conditions in all regions. 3 Farmers will actually apply the fertilizer rates on each crop as recommended by the official fertilizer allocation plan. 64 A - All fertilizer supplies will be available for distribution at the time when farmers need them. 5 - All farmers have access to the official fertilizer distribution network and they have the necessary financial resources to cover all their fertilizer purchases. 6 - The optimum fertilizer rates are scale-neutral, i.e., they are equally applicable to small and large farms. These assumptions clearly suggest that the results of the model may not be a very accurate representation of actual conditions in Syria. However, the above working assumptions had to be made given the limited data availability in Syria. Although the model may not be as accurate as one would desire under ideal circumstances, the results would represent rough approximations of actual and/or simulated conditions. Tine 8e approximations would still constitute information that could ass iat policy-makers in formulating more efficient fertilizer allocation 8 tr'ategies . 3.2.2 W In the earlier discussion on the procedure to estimate ee<>l'ld:>gj,¢:ally optimum fertilizer rates, the term ”optimum“ was often used- This term is somewhat vague since it does not specify from whose point of view these rates are optimal, the farmer's or the economy as a w11°1~e17 This distinction becomes important if there are relatively large :1 1ftetrences in the prices that farmers pay or receive, as compared to th e true costs to the economy as a whole. 65 For instance, if the highly subsidized fertilizer producer prices are to be used in the analysis, then the optima calculated would be relevant only if the objective is to maximize farmers' income. If, on the other hand, the objective is to maximize national income (or any other measure of aggregate welfare), then these farmers' optima may be economically non-optimal. This would be the case if the input-to- output, or W, price ratio faced by farmers is significantly different from the true relative price ratio. In other words, farmers' optima may lead to an economically inefficient allocation of the limited fertilizer resources. In this case, it would be more appropriate to use international market prices, which generally represent the true opportunity cost of resources used and of outputs produced. The main objective of this study is to develop a national decision model for allocating the limited fertilizer resources in the most economically efficient way. This is based on the stated policy Obj ective of maximizing net economic returns from the limited fertilizer 8‘1pplies. This focus on “economic efficiency" necessitates that the litialysis be based on the true value of resources used and of output produced, in order to achieve national welfare objectives. Thus, whenever fertilizer or crop prices are significantly different from their true economic values, these prices should be adj usted to make them more closely represent the opportunity costs to t he economy as a whole. These adjusted prices are referred to as W (the term “shadow price" is also often used; see, for e xQI‘Dle, Gittinger, 1982, p. 243), whereas unadjusted prices (i.e., 66 actual prices faced by farmers) are referred to as financial prices‘. T‘kaerefore, the estimation of “optimum“ fertilizer rates, whether based on constrained or unconstrained optimization, will be based on economic prices throughout this study. In addition to the economic efficiency objective, an important objective of this study is to analyze the impact of alternative fertilizer allocations on farmers' net returns from fertilizer use. T1113 is of particular importance since the ultimate decision on which policy alternative to adopt is essentially a political decision. Such a decision would be influenced by many factors, including the potential impact of the policy in question on farmers. Therefore, the estimation of Optimum fertilizer rates will be based on economic prices, while the impact of these rates will be evaluated using both financial and economic prices. In other words, for the economically optimum fertilizer rates to be considered acceptable, the VCR values calculated based on both economic and financial prices would have to satisfy the minimum VCR requirements mentioned earlier. 3.2.3 W In addition to their concern about the impact of alternative fel-tillizer allocation strategies on farmers' income, policy makers are a 18° interested in the impact of these strategies on key macroeconomic Policy objectives. These objectives include: (I) reducing foreign \ 1 Refer to chapter 4 for a detailed discussion of the procedures in estimating financial and economic prices, and their underlying tions. ‘38 67 exchange expenditures, (2) reducing the government's budget deficit, and ( 3) increasing food self-sufficiency. l) e d t e The main reason for the current constraints on fertilizer supplies is the government's decision to limit fertilizer imports in an attempt to reduce its foreign exchange expenditures. Fertilizer policy affects the government's foreign exchange expenditures in three direct ways: (a) fertilizer imports, (b) crop imports, and (c) crop exports. For instance, if fertilizer imports are lowered to reduce expenditures in hard currencies, the resulting lower fertilizer use may reduce the output of crops that substitute for imports, and/or reduce the output of exportable crops. Therefore, the main focus should be on increasing n_e_t; foreign exchange earnings, defined as 353(2Wt2. foreign exchange earnings - Earnings from increased crop exports + Savings from reduced crop imports - Expenditures on fertilizer imports As will be discussed later (see chapter 4), all crops covered in this study are treated as traded goods, i.e., they are either exported or they substitute for imports. Since the government has a complete monopoly on foreign trade in fertilizers and most crops, any increase or decrease in aggregate crop output, due to changes in fertilizer policies, will be reflected in an equal increase or decrease in net crop exports. This would apply only to crops, such as cotton and sugar bee 1:3 ’ which are completely controlled by the official marketing system. I h the case of cereals, however, significant proportions of total output 68 are sold in the parallel market‘, while potato output is entirely sold in the recently legalized private market. The net increase in foreign exchange earnings as a result of fertilizer use can be expressed as follows: NFE - 2, [19,, * A, * n, (Y, - Y°,)] - 2:, (IP, * F“) (3.24) where NFE is net increase in foreign exchange earnings relative to no fertilizer use; I?” is the international price of crop i; A, is the total fertilized area planted to crop i; H, is the proportion of output of crop i sold to the government; Y, is the per ha yield of crop i; Y°1 is the per ha yield of the unfertilized treatment of crop 1; IP, is the international price of fertilizer f; F“ are aggregate available supplies of fertilizer f. Given that crop yields would vary depending on rainfall levels, foreign exchange earnings from net crop exports are also expected to vary. It is, then, possible to compute four different equations for net foreign exchange earnings based on the four rainfall scenarios discussed earlier. These equations can be incorporated into the allocation model as additional constraints with lower limits specified to reflect policy obj ectives. For instance, minimum foreign exchange earnings during a nor-.31 year can be set to be equal to the current level of net foreign e=":<-‘-1‘£ange earnings associated with fertilizer use. Such explicit constraints, however, might impose too much rigidity on the model solution. Thus, the general approach followed in this 8 tudy will be to calculate net foreign exchange earnings associated with e th fertilizer allocation examined, without a priori restrictions on t he lower limits. The results will be presented and their implications \ of 1 Refer to chapter 4 for more details on the estimated proportions Qereal out ut sold throu h official channels. P 8 69 will be discussed, but the decision on whether these earnings are acceptable or not will be left to policy-makers. 3) v ' e In addition to concerns about the government's foreign exchange expenditures, the impact of alternative fertilizer strategies on the overall government budget is also an important policy issue. This is of particular importance given the heavy subsidies on fertilizers in Syria. Therefore, in comparing the feasibility of alternative fertilizer allocations, it is important to estimate the impact of these strategies on the government budget. To estimate this budgetary impact, we need first to calculate the net taxes or subsidies associated with the fertilizers and crops covered by this study. If the official price of a crop is much lower than its true economic value, the government would be making additional revenues from 98c}: additional kilogram produced (i.e., the crop is implicitly taxed) as a result of fertilizer use. In contrast, the government would be 1t1<-‘—t..irring additional expenditures for each additional kilogram of fertilizer applied. If, on the other hand, the official crop price is higher than its true economic value (implicit subsidy), government expenditures would increase as a result of increased yields due to feIrtlllizer use, besides the higher expenditures on fertilizer subsidies. The net increase in government expenditures as a result of f artilizer use can be expressed as follows: N GE ~ 2, (s, * Fa) - 1:. [13. * A. * M. (Y. - Y°1>l (3-25) 70 where N63 is net increase in government expenditures relative to no fertilizer use; S, is subsidy per unit of fertilizer f; F“ are aggregate available supplies of fertilizer f; T” is implicit tax or subsidy per unit of crop i; A, is the total fertilized area planted to crop i; H, is the proportion of total output sold to the government; Y, is the per ha yield of crop i; Y0, is the per ha yield of the unfertilized treatment. As in the case of foreign exchange earnings, explicit constraints on maximum government expenditures might impose too much rigidity on the model solution. Thus, net government expenditures associated with each fertilizer allocation will be calculated and their implications discussed. But the decision on whether these expenditures are acceptable or not will be left to policy-makers. 3) W The main concern about food self-sufficiency in Syria relates Primarily to wheat production and, to a much lesser extent, the Production of sugar beets. Wheat imports represent the single most important food import item, accounting for an average of 251 to 301 of total wheat consumption (Agricultural Statistical Abstracts, 1987). This percentage can be as high as 751 in a very dry year like 1984, when a Ire(:ord 1.5 million tons of wheat was imported. Therefore, increasing wheat self-sufficiency is a prime policy objective often influencing the design and implementation of other policies, including fertilizer policy, This is clearly implied by the high priority given to wheat in the Current fertilizer allocation strategy adopted by the government. Such concerns can be incorporated into the fertilizer allocation “to del by setting up constraints specifying lower limits on aggregate 71 output for each crop. Also, these constraints can be formulated to reflect specific self-sufficiency requirements for each rainfall scenario. Although such constraints can be easily incorporated into the model, the minimum production limits would be difficult to determine given the vagueness of policy statements related to self-sufficiency. These statements are often expressed in very general terms, referring to the need to increase self-sufficiency in food, feed, and industrial crops, in addition to the objective of increasing agricultural exports. Explicit self-sufficiency constraints may also reduce the model's flexibility in finding an optimum solution that would maximize economic net returns to fertilizer use. Thus, the general approach followed in this study will be to calculate the impact of alternative fertilizer allocations on aggregate crop output and to discuss their implications on food self-sufficiency. 3.3 oca ode 3.3-1 WWII]: The general formulation of the fertilizer allocation model, given by equations 3.21, 3.22, and 3.23, may be too complex to solve based on standard calculus techniques. This would be particularly true if a large number of crops, or crop varieties, are to be included in the model. As mentioned earlier, such a model would require solving a Constrained optimization problem consisting of (1*f + f) constraints, "here i is the number of crop activities and f the number of fertilizer nutrients covered by the model. For instance, a model with 15 crop act1:|.vities and two fertilizer nutrients would include at least 32 c0tlstraints, which can be computationally difficult to achieve 72 particularly if additional constraints are to be incorporated into the model. Fortunately, there exist other methods to solve constrained optimization problems. One of the most commonly used methods is linear programing (LP). In addition to finding an optimal solution to large constrained optimization problems, the LP solution can also generate other useful information such as the shadow prices of limited resources and information on non-optimal activities. A detailed review of literature on the theory and' applications of linear programming is provided by Schrijver (1986). Hazell and Norton (1986) focus on the applications of programming techniques to agricultural problems. Bronson (1982), Hills (1984), and Taha (1987) provide practical introductions to LP including many agricultural examples. Linear programming is an optimization technique to solve for “allocation problems in which limited resources are allocated to a “timber of economic activities” (Taha, p. 50). The LP problem consists of optimizing (maximizing or minimizing) a specific quantity, called the “Objective”, which depends on a finite number of input variables, 8‘~l-7t)ject to a set of constraints (Bronson, p. l). The following section 18 adapted from Taha (p. 50), and provides a brief introduction to the f"firmilation of a general LP model. For a maximization problem, the LP model, in its general Illat‘hematical form, is expressed as follows: 73 Maximize Z - c1.X1 + c2.X2 + + c,,.}I{u subject to: 811.}{1 + 312.1(2 + ... + 8m.xn 5 b1 821.}(1 + 822.XZ + ... + 82“.Xn ‘ b2 8.1.}(1 + 8.2.x2 + ... + 8m.xn S b. XI, X2, ..., Xn Z O The above model includes n activities X1, X2, ..., Xn and m resources with maximum amounts available given by b1, b2, ..., b... The first line of the model is the objective function, which can be expressed as Haximize Z - 2, (c,.x,) mis function represents the combined contributions of each activity to t$tal profit 2, where cJ represents the profit or net return per unit of aetivity j. Under the objective function there are m constraints that can be e3(I>ressed as follows: 2., (au.XJ)sb1 i-l, .....m This means that each unit of activity j uses an amount a“ of reSource i, and the summation of all an}: represents the total use of resource i by all n activities, which cannot exceed b*. The resource 11‘I:l.ts or bi's are often called Right Hand Side (RHS), referring to the it positions with respect to the inequality signs. LP constraints can include 's', 'z', or '-" signs. However, strict inequality signs (>’ or <) cannot be included in the formulation of an LP problem. 7h The last line, X,, X2, ..., XD 2 O, is what is often referred to as “non-negativity constraints“, and it specifies that none of the activities in the solution can be negative, which is self-explanatory. 3.3.2 M An LP formulation of the fertilizer allocation problem was proposed by Nordblom and Al-Ashram (1989), who developed a conceptual model for the centrally planned allocation of limited fertilizer in contrasting production zones. The model's 8 upplies to crops Potential is illustrated through its application to a hypothetical three-crop country, using coefficients based on actual data from fertilizer experiments in Syria. In fact, Nordblom and Al-Ashram's Study represents the first stage of a larger research project which is the basis of this dissertation. The LP model presented in this section 18 based on Nordblom and Al-Ashram's model, with some modifications that wi 11 be discussed later. The fertilizer allocation model, given by equations 3.21, 3.22, and 3.23, can be re-formulated as an LP model, as follows: Maximize NR - 2, [9,, * A, * (Y, - y°,)] - 2, (u, * F“) (3.26) s‘-lt)ject to: 2, (P1, * A,) s Pa (3.27) (3.28) N‘(F11, 1:2,, ..., F11, X,, Z) '- Y, F11, F21, ..., Pg, 2 0 Y, and Y°, z 0 where NR is aggregate net return from fertilizer use in Syria Y, is the per ha yield of crop i; Y°, is the per ha yield of the unfertilized treatment; N,(F,,, Pu, , F“, X,, Z) is the production function for Y,; 1“,, is the per hectare application rate of fertilizer f on crop i; X, refers to inputs on crop i other than fertilizers; 2 refers to environmental factors; 75 1“,, are aggregate available supplies of fertilizer f; U, is the price of fertilizer f; P,, is the price of crop i; A, is total fertilized area planted to crop i. In this simple or 'basic' version of the LP model, the objective is to maximize net returns from the use of fertilizers on all the major crops in Syria. This is subject to the constraints imposed by the quantities of fertilizer available, and the physical input/output relationship between the amounts of fertilizer applied and yields obtained. The model would solve for the optimum fertilizer rates on each crop, 5"“, that will maximize net returns from fertilizer use given the constraints on aggregate supplies. Crop and fertilizer Prices, crop areas, and the upper limits on fertilizer supplies are all exOgenous to the model, i.e., they are fixed by the analyst. The input/output coefficients are based on the estimated parameters of the p r oduction func tions . 3.3.3 W A key determinant of the complexity of the above model is the Specification of the production function constraints. These functions we re expressed using a general formulation such as in eq. 3.1: Y, - N,(F,,, Pu, ..., F“, X,, Z) The production functions need to be formulated in more specific teI‘ms. This refers to the functional forms to be used and the e"‘planatory variables to be included in the production functions. F91 76 3.3.3.1 W The true relationship between applied fertilizer nutrients and yield is never known. Thus, a key step in the specification of 'production functions is the choice of an appropriate functional form. This has typically been difficult in applied research conducted by soil scientists, agronomists, and agricultural economists. This task is further complicated by the growing number of available functional forms to choose from. Griffin, Montgomery, and Rister (1987) compiled twenty different categories of functional forms that have been used in the production economics literature. Therefore, the problem of choosing the "best' function cannot be solved from a simple set of rules. The quadratic production function, and the polynomial function in general, has been successfully applied to a large number of fertilizer 8tudies listed in the literature. The general approach is based on studying experimental data by statistical methods and an empirical Polynomial equation of “best fit" is estimated, with no assumption or hyPothesis as to the underlying causes (Mason, 1956, p. 77). This approach emerged from the extensive efforts of agricultural economists during the 1950's. Heady and Dillon (1961) examined the estimation methods, and the mathematical and economic characteristics of tIllree most commonly used production functions, power models (Cobb- I><>\xglas function), exponential models (Mitscherlich and Spillman f\lfl'nctions), and polynomial models (quadratic and square-root functions). Se\reral studies during that period attempted to compare the Q‘Ppropriateness of these functions by comparing how well they fit fertilizer experiment data (see for example Johnson, 1953; Heady, 1954; Hutton, 1955; Hutton and Elderkin, 19514; and Heady, Pesek, and Brown, 77 1955). The conclusions reached in these studies were obviously specific to the unique fertilizer experiments. Nevertheless, for the majority of these studies the polynomial quadratic and square root models generally gave the best fit. In this study, the quadratic polynomial functional form will be used in the estimation of production functions. This functional form has been the standard one used in fertilizer trials by agronomists from both the Soils Directorate (SD) and the International Center for Agricultural Research in the Dry Areas (ICARDA). Given that this dissertation is based on a collaborative project between the two institutions, a functional form that can be easily estimated and understood was desirable. The quadratic function postulates a smooth, concave, and differentiable function, possessing a point maximum, with substitution aalong all nutrients (Baum, Heady and Blackmore, 1956; Baum, et al., 1957). The concavity assumption conforms with the empirically observed diminishing marginal response to fertilizer applications (Mason, 1956, p - 81). Empirical observations also confirm the assumption of 8\J-IJstitution among nutrients (see, for example, Dumenil and Nelson, 19<18). In Syria, these assumptions seem also to be confirmed by e“pirical observations, as noted by many SD agronomists working on fertilizer trials. The assumption of point maximum generally applies to nitrogen 81rice excess N application could result in excessive vegetative growth Q‘lusing lodging and lower yields. However, in the case of phosphate and potassium applications the most commonly observed response is that of it\tzreasing yield until a plateau is reached (Lanzer, Paris, and 78 Williams, 1987, p. 2). Yield depression usually occurs at quantities far beyond the minimum needed to attain the yield plateau. Such responses are more accurately represented using other functional forms, such as the Mitscherlich function or the linear response and plateau (LRP) function proposed by Gate and Nelson in 1971 (see also, Anderson and Nelson, 1975; Perrin, 1976; and Lanzer and Paris, 1981). These models are usually more complex than the quadratic function and they require more data, including accurate information on soil nutrient content. Therefore, the application of such models in Syria would be difficult given the data problems that relate to the inaccuracy of the soil testing procedures. 3.3.3.2 Explanatory Variab1e§ Multiple regression is the standard approach in estimating the parameters of a quadratic production function. Yield per ha is the dependent variable, while the levels of applied nutrients per ha are the independent variables. The production function has the following general form: Y-ao+a,N+a2P+a3N2+a,P2+a5NP (3.29) where Y is estimated yield per ha; N and P are the applied fertilizer rates per ha; N2 and P2 are quadratic terms; NP is an interaction term between N and P (i.e, N*P). The parameters a0, a,, a2, a3, a.” and a, are the estimated regression coefficients. The first term, a0, is the estimated yield of the unfertilized or control treatment (i.e., when applied N-P-O); a, and a; are typically positive reflecting the increase in yield in response to fertilizer applications, and a3 and a, are typically negative 79 reflecting diminishing response to increasing applications of N and P. The term a, is usually positive reflecting the positive interaction beWeen N and P, though it is not uncommon for a, to be negative. The list of independent variables is not necessarily limited to N and P. Ideally, potassium (K) should have also been included in the production function. However, given that Syrian soils are usually rich in K, the use of potassium fertilizers is very limited, with the exception of root crops and fruit trees. Therefore, this study will be limited to nitrogen and phosphate fertilizers. If the data are available, it is desirable to add several eXplanatory variables. These variables reflect the specific conditions th'ader which each experiment was undertaken such as soil type, residual 8011 nutrients, rainfall level, temperature, and cultural practices ( including the level of other inputs). It is also desirable to include interaction terms that reflect the empirically observed interaction between N or P and some of the explanatory variables. Most trials ut‘adertaken by the SD include data on residual soil nutrients before Planting and growing season rainfall levels. However, SD officials have ft‘equently questioned the accuracy of their soil analysis results which Widenines the usefulness of these data. Therefore, the only variable for which reliable data is available is seasonal rainfall. Rainfall level is expected to have little impact on irrigated ct‘ops (e.g., cotton, sugar beets, potatoes, corn, and irrigated wheat). HOVBVGI’, in the case of rainfed wheat and barley, rainfall is by far the I.Ost important determinant of yield and the largest contributor to the Variance in yield (SD/ICARDA, 1989 and 1990). Rainfall should be 1tioluded in the analysis in order to estimate zone-specific production 80 functions for wheat and barley. This can be done using two alternative approaches. One approach is to estimate a production function for each zone in the form given by eq. 3.29, based on trials from that zone. However, since the trial sites are unevenly distributed among the three zones, there are not enough data from each zone to estimate appropriate zone-specific functions. An alternative approach is to pool all the data from the fertilizer trials and to estimate a single equation that would include rainfall as an independent variable. Then, by using the value for average rainfall in each zone, it is possible to compute three different zone-specific equations. Such an approach assumes that the only difference between zones is seasonal rainfall, while differences in other variables are ignored. Hence, for rainfed crops, the estimated production functions will have the following general form: ‘Y - be + b, N + b, P + b, EN + b, EP +b5N2+b6£N2+b7P2+bOEP2 + b, NP + 13,, ENP + 13,, 2 + b12 22 (3.30) where E is total seasonal rainfall, and EN, EP, ENZ, EPZ, and ENP are all interaction terms between E and N or P. This equation is a generalized form and some terms may not be included in all estimated regression equations. In fact, as noted by Puss, McFadden, and Mundlak (1978, p. 224), too many variables may exacerbate multicollinearity problems. Therefore, it is desirable to reduce the number of terms in the equation to a minimum, even if this means that the equation may lose some of its explanatory power. The choice of which terms to retain in the equation will largely depend on the crop in question and the data used. 81 The following two simpler forms will be used as a starting point for the analysis and will be subject to modification as the needs arise: Y-bo+blN+bzP+b3EN+b‘EP + b, N2 + b, P2 + b, NP + b, E + b, 22 (3.31) and, *< I b, + b, EN + b2 EP + b, N2 + b, P2 b, NP + b, 2 + b, 22 (3.32) 4. Production functions based on the above two formulations will be estimated for each of the rainfed crops or varieties. The decision as to which formulation to use will depend on the statistical performance of the two formulations. This will include the value of the coefficient of determination (adjusted R3), the standard error of the regression, the standard error of the coefficients, and whether the estimated coefficients have the expected signs. Once the production function for a rainfed crop is estimated, based on either eq. 3.31 or eq. 3.32, zone-specific production functions can then be estimated by replacing 'E' by its corresponding average value for the zone in question. These zone-specific functions will be then incorporated into the linear programming model to allow for the calculation of optimum fertilizer rates for each zone. For instance, if a production function of the form expressed in eq. 3.31 was estimated for rainfed local (LYV) wheat varieties, then, in order to compute production functions specific to Zones 1, 2, and 3, the mean value of rainfall (E) for each zone1 is entered into the equation. This would give a production function of the form given by eq. 3.29: 1 Refer to rainfall data in Table 3.1. 82 Y-a,+a,N+azP+a,N2+a,P2+a5NP Where, a0 - (b0 + b, B + b9 £2) a, - (b, + b3 E) a; - (b; + b, E) 83-135 s, - b, 85-b7 b0, ..., b, are the estimated coefficients in eq. 3.31. 3.3.3.3 W A basic assumption of linear programming is that the objective function and the constraints are linear. Since the production function constraints, given by eq. 3.28, are based on quadratic functions, linear approximations of these functions are needed before the model can be solved using standard LP methods. Hazell and Norton (1986) describe two general procedures for approximating nonlinear functions in linear programming, or what is often referred to as separable linear programing methodslz The first procedure consists in dividing a nonlinear, concave and separable function Y - f(X) into linear segments which are defined over intervals (X,2qu) on the X axis and corresponding intervals (Y,JYrq) on the Y axis. The slope of the linear segment, 5,, in the ith interval is defined as Y1 ' Yrq s, - —-——-—-— X, " xi-l Let V, denote variables that measure the value of AX over the corresponding ith interval, such that O s V, s X,éxp1. Since the s, are 1 This section is adapted from Hazell and Norton (1986), pp. 73- 75. See also Kilmer (1978). , 83 predetermined, then a linear approximation to Y is obtained based on the following equation system: MaxY'-2,V,s, subject to X - 2, V, and 0 s V, s X,-Xp4, all i. A similar approach is used by Nordblom and Al-Ashram (1989) to obtain linear approximations for the quadratic production functions included in their fertilizer allocation LP model. Such an approach, however, may create computational problems if the LP model includes a large number of crop activities, as discussed by Hazell and Norton (p. 74): the degree of accuracy of the approximation depends on the number of segments introduced, but the associated costs is an extra column (V,) and upper bound row for each segment. While extra columns add little to computational costs with modern linear programming computer codes, extra rows are expensive. Consequently, the cost of introducing many nonlinear approximations would soon become prohibitive. Hazell and Norton suggest using another more efficient procedure for linearizing a nonlinear function. By defining new variables W,, or weights, for each interval 1, a linear approximation to Y - f(X) is then given by: Max 2' - 2, Y, H, (3.33) subject to E, W, s l (3.34) x - 2, w, x, (3.35) and W, 2 0 84 This procedure is computationally superior to the first one since it requires adding only two constraints, and it is not affected by the number of segments. 3.3-4 WHO—ck]. 3.3.4.1 u e n o v e Crop responses to varying levels of fertilizer applications can be estimated based on results from experimental fertilizer trials. In Syria, the Soils Directorate (SD) has been conducting fertilizer trials since the 1950's with an initial focus on cotton (for a summary of these early studies, see Kanbar and El-Hajj, 1974, pp. 6-9; see also Loizides, 1968). The most important research effort took place in the late 1960's and early 1970's. This involved a large number of on-farm and on- station fertilizer experiments primarily on wheat and cotton (FAO, 1970; Xanbar and Bl-Hajj, 1973, 1974, 1975a, and 1975b; and El-Hajj, 1985a and 1986). Between 1975 and 1980, most of the SD staff were involved full- time in the Syrian soil survey and classification project and thus very few fertilizer experiments were conducted. Fertilizer research efforts resumed in the early 1980's, with the main focus of comparing alternative forms of nitrogen fertilizers (urea and ammonium nitrate) on yield. This research interest was influenced by the plans to construct a urea plant which would result in urea replacing ammonium nitrate as the primary source of nitrogen. Fertilizer trials on wheat and cotton also resumed but on a much smaller scale than those conducted during the previous decade. The SD decided to shift its fertilizer trials to other economically important 85 crops such as sugar beets, potatoes, and chickpeas, and to newly introduced crops such as corn, soybeans, and sunflower. Also, a limited number of trials were started on vegetables (tomatoes and cucumbers) and on fruit trees (olives, apples, and peaches) (MAAR, 1980, 1981, 1982, 1983, and 1987). The results from most these experiments are not yet available for general use, with the exception of initial results of sugar beets (Kanbar and El-Hajj, 1986). In addition, two major fertilizer research projects were undertaken by ICARDA in collaboration with the SD in the 1980's. The first project was on fertilizer use on barley in northern Syria conducted from 1984/85 to 1987/88 (SD/ICARDA, 1990). .A similar study was also conducted in northern Syria on durum wheat (HYV variety Sham 1) from 1986/87 to 1988/89 (SD/ICARDA, 1989). The above data sources do not cover all the crops grown in Syria. But there exist enough data to estimate production functions for the following crops: wheat, cotton, barley, sugar beets, potatoes, and corn. These are the main crops grown in Syria and they currently account for about 862 of nitrogen and 781 of phosphorus consumption by field crops, or about 661 of total fertilizer consumption in Syria. In the case of wheat, there are adequate data from irrigated and rainfed experiments to allow for the estimation of distinct production functions for irrigated and rainfed wheat. Similarly, the data allow for the estimation of different functions for high-yielding (HYV) and local (LYV) wheat varieties. Since virtually all irrigated wheat varieties are HYV, only three distinct functions are needed for wheat: irrigated, rainfed HYV, and rainfed LYV. Similarly, sugar beets and potatoes can be planted either in fall or in spring, with distinct 86 growing patterns, yields, and fertilizer requirements. Since fertilizer trials exist for both seasons, separate production functions can be estimated for fall- and spring-planted sugar beets and potatoes. Given the available data, production functions for ten different crops will be estimated. These are: irrigated wheat; rainfed HYV wheat; rainfed LYV wheat; rainfed barley; irrigated cotton; irrigated corn; irrigated fall sugar beets; irrigated summer sugar beets; irrigated fall potatoes; and irrigated spring and summer potatoes. Moreover, in the case of the rainfed crops (wheat and barley), the SD provides fertilizer recommendations that are specific to each of the agroclimatic zones. These include zone-specific rates for rainfed LYV wheat and for barley in Zones lb, 2, and 3. Since very little HYV wheat is grown in Zone 3, specific rates for HYV wheat are provided for only Zones lb and 2. Zone la is excluded from the analysis since it is used mostly for fruit and vegetable production, with insignificant amounts of wheat and barley (see Appendix A, Table A.2). Zones 4 and 5 are also excluded since they are not covered by the fertilizer allocation system. Most of the above experiments focused on yield response to N and P205. Potassium (K) was included in the 1970's cotton trials but the findings strongly suggest that cotton does not respond to X application (Kanbar and El-Hajj, 1974 and 1975b). The results are reasonable since most Syrian soils are naturally rich in potassium. This makes it unnecessary to apply any X fertilizer on most crops, with the exception of root crops (sugar beets and potatoes) which are fairly responsive to K (Kanbar and El-Hajj, 1986). Thus, given the current limited use of X fertilizers in Syrian agriculture, only N and P will be included in this study. 87 3.3-4.2 NW The mathematical formulation of the actual or "expanded" linear programming, model for fertilizer allocation used in this study is described below. It is based on the “basic“ model given by Equations 3.26, 3.27 and 3.28, with several modifications that are discussed in details in this section. These modifications include a change in notations to coincide with those used in the computer input file presented in Appendix C. The full model specification appears at the end of this chapter. b e v un o : In. this model the objective is to ‘maximize the economic net returns from fertilizer use on the main crops grown in Syria (wheat, barley, cotton, corn, sugar beets, and potatoes). These crops are grouped into 15 crop activities based on whether they are irrigated or rainfed, on Agricultural Stability Zones, and whether they are planted in the fall or spring. The objective function (eq. 3.36) is specified in terms of the economic value of the aggregate increase in crop output due to fertilizer use, minus the economic value of total nitrogen (N) and phosphorus (P505) fertilizers applied on these crops. The increase in crop output is based on the assumption of “normal“ rainfall, except for barley, where a 'dry" rainfall scenario is assumed. This is in accordance with the 'maximin' assumption about the behavior of risk- averse barley farmers, which was discussed earlier in this chapter. 88 W: The model's constraints include eleven groups of equations: 1) u s a n These equations (eq. 3.37 to eq. 3.42) specify the input/output relationships between the fertilizer rates applied and the resulting yield increase relative to no fertilizer use. Since the production relationships are based on quadratic functions, linear approximations of these functions are included in the LP model based on the procedure suggested by Hazell and Norton (1986), as discussed earlier (see Equations 3.33, 3.34 and 3.35). To obtain linear approximations for the 15 production functions included in the model, the crop activities were grouped into three categories: (1) crop activities with relatively low expected optimum fertilizer rates, i.e., rainfed wheat and barley; (2) crop activities with relatively high expected optimum fertilizer rates, i.e., irrigated wheat and cotton; (3) crops whose production functions were estimated in terms of N only because of data limitation problems, i.e., corn, sugar beets, and potatoes. In the first category, each production function was divided into 100 linear segments corresponding to 100 different N-P20, combinations (10 rates of N by 10 rates of'IQOS) ranging from 0 kg/ha for both N and P205 to 90 kg/ha N and 65 kg/ha P205. The production functions in the second category were divided into 143 linear segments corresponding to 143 N-P205 combinations (11 rates of N by 13 rates of P205) ranging from 89 O kg/ha for both N and P205 to 230 kg/ha N and 130 kg/ha P205. In the third category, the yield response functions to nitrogen were divided into 34 linear segments corresponding to 34 N rates (0 to 220 kg/ha) by one rate of P205 (0 kg/ha). The number of segments in the above categories was selected in such a way as to cover the range of N-P205 combinations from zero to the rates that would maximize yield (i.e., Phase II of the production function). The accuracy of separable programming depends to a large extent on the number of segments per production function. Thus, the above segmentation approach attempted to include the largest number of segments allowable by the memory available on standard personal computers, which is the technology currently available at the Soils Directorate. It should be noted that the linearization of production functions in the third category does not allow for the estimation of optimum P205 rates on corn, sugar beets and potatoes. Therefore, arbitrary assumptions are needed as to how optimum P205 rates would change when the amounts of fertilizer available are varied. The assumption adopted in this study is that optimum P20, rates would change at the same rate as the change in optimum N rate. In other words, it is assumed that the ratio of optimum P203 to optimum N (PNRATIO,) is constant. Assuming that the current P205 recommendations on these crops represent the unconstrained economic optimum rates (ECONOPT,.,), this constant ratio is defined as: ECONOPTL-P» PNRATIO, - 200N0P2,,... 90 Optimum P505 rates are then estimated by multiplying the optimum N rate (OPTF,,1~) by PNRATIO,, rather than OPTFran‘which will always be equal to zero according to the segmentation of the third crop category mentioned above. 2) Uppgg Lipigg pg ngimum Fegtilizez Bates: Two sets of constraints are imposed on the maximum values of the estimated optimum fertilizer rates. 'The first constraint (eq. 3.43) specifies that the economically optimum fertilizer rates should not exceed the optimum rates calculated based on financial prices. This constraint is needed since farmers will not apply rates beyond those maximizing their net returns. The second constraint (eq. 3.44) specifies that the optimum rates should not result in a yield decline in the event of a very dry year. This is in accordance with farmers' concerns about the possibility that fertilizers may ”burn” the crop in very dry years. 3) t due 0 e v This group consists of three definitional equations (equations 3.45, 3.46 and 3.47), related to the above three categories of crop activities, which enables the calculation of the aggregate increase in output for each crap activity. This is done by multiplying the total area fertilized by the yield increase due to the application of the estimated optimum fertilizer rates. 91 4) e at 0 Cu u : This group of equations (eq. 3.48 to eq. 3.51) adds the total output of crop activities by crop. For instance, the outputs of all wheat activities are aggregated together to give the total wheat output (TOTWHEAT.). Similar aggregations are done for barley, sugar beets and potato activities. 5) We: As discussed earlier, one criterion for assessing the economic feasibility of the estimated optimum fertilizer rates is that the Value- Cost-Ratio, calculated based on these rates, should be equal to at least 1.5. This applies to VCR's calculated based on ”good", ”normal“, and 'dry' rainfall scenarios, while in very dry years the minimum VCR limit is reduced to 1.0. These minimum VCR conditions apply to VCR calculations based on either economic prices (eq. 3.52) or financial prices (eq. 3.53). In these two equations, MINVCR. is a vector of minimum VCR values corresponding to each of the rain scenarios. 6) WWW: Total fertilizer use for each crop activity is calculated by multiplying the estimated optimum fertilizer rates by the total fertilized area planted to each crop. Aggregate fertilizer use by all crop activities is then calculated by adding up total fertilizer use by each crop. Since fertilizer allocation decisions for winter and summer crops are frequently made independently of each other, aggregate fertilizer use for each season needs to be computed. These calculations 92 of aggregate fertilizer use by season are given by eq. 3.54 to eq. 3.57, while eq. 3.58 adds up total fertilizer use over the two seasons. The calculation of aggregate N use is straightforward, as shown in eq. 3.54 and eq. 3.56. On the other hand, the calculation of aggregate I505 use is slightly more complicated. This is related to the procedure for estimating optimum P50, rates on corn, sugar beets and potatoes, as described earlier. Since the optimum value of OPTTkpqr will always be zero for these crops, the estimated P505 rates on these crops need to be added to the equation to avoid underestimating aggregate P205 use, as shown in eq. 3.55 and eq. 3.57. 7) e ' e vs ab ns 8 nt : The upper limits on total fertilizer supplies available are given by equations 3.59, 3.60 and 3.61. These limits are expressed as a percentage (FPER,) of 'ideal" total fertilizer requirements (FLIMf), i.e., total requirements if there 'were no constraints on. supplies. These constraints allow for examining different fertilizer availability scenarios for the winter and summer seasons. For instance, if the policy issue of interest is fertilizer availability for the ‘winter season only, the percentage of total fertilizer requirements for the summer season (SPER,) would be set at 100. That for the winter season (WFPER,) would be varied according to the assumed levels of fertilizer availability. 8) legplggipps pf Eggnomig Neg fietuzns: The optimum fertilizer rates are estimated based on the objective of maximizing net economic returns from fertilizer use, assuming a dry 93 year for barley and a normal year for all other crops. As discussed earlier, the dry year assumption for barley was made to account for the relatively higher risk aversion among barley farmers. However, the value of the objective (i.e., maximum 2) does not represent net economic returns in a normal year, because of the dry year assumption for barley. Therefore, to calculate the impact of the estimated optimum rates on net economic returns in a normal year, a normal year rainfall should be assumed for all crops, including barley. Similarly, net economic returns to the application of the estimated optimum rates are calculated under the good, dry, and very dry rainfall scenarios. These calculations are given by equations 3.62, 3.63 and 3.64, which also provide a breakdown of net economic returns by season (winter vs summer). 9) We: The calculation of financial net returns (or net increase in farm income) for the four rain scenarios is identical to that of economic net returns except for the use of financial instead of economic prices. These calculations are given by equations 3.65, 3.66 and 3.67. 10) legplgtipn pf Egg Epzeign Exghapgg flappingsz The net foreign exchange earnings (DOLLARS.) associated with fertilizer use are calculated by adding the value of additional crop exports and lower crop imports due to fertilizer use, minus the import value of fertilizers. Since not all crop output is sold to the government, foreign exchange earnings per unit of crop output produced (FECROPL.) are weighted by the proportion of total output sold to the 94 government. For instance, if the import value of wheat is 224 $US/ton, and assuming only 401 of total wheat output is sold to the government, then the foreign exchange earnings per ton of wheat produced would be equal to 89.6 $US/ton (i.e., $224 * 401). The calculation of net foreign exchange earnings under the four rain scenarios is given by eq. 3.68. 11) v e d As discussed earlier net government expenditures related to fertilizer use need to be calculated to assess the impact of the optimum rates on the government budget. These net expenditures (GOVEXP.) are calculated by subtracting the increase in government revenues, due to indirect taxes on crops, from total expenditures on fertilizer subsidies, as shown in eq. 3.69. Since taxes on crops are proportional to yields, government revenues from crop taxation would vary depending on the level of rainfall. Therefore, for each rain scenario there would be a specific level of net government expenditures. 12) - a v on a t : These are the standard constraints specifying which decision variables cannot be negative. 3.3.4.3 o t 0 he am: A. 1.115112254292111: 1- where crop activities (WIRR, WHYV1, WHYVZ, WLYVl, WLYV2, WLYV3, BARLEYl, BARLEYZ, BARLEY3, COTTON, MAIZE, FALLBEET, SUMBEET, FALLPOT, SUMPOT) WIRR - Irrigated wheat WHYVl - rainfed HYV wheat Zone lb WHYVZ - rainfed HYV wheat Zone 2 WLYVl - rainfed LYV wheat Zone lb WLYV2 - rainfed LYV wheat Zone 2 WLYV3 - rainfed LYV wheat Zone 3 BARLEYl - rainfed barley Zone lb BARLEYZ - rainfed barley Zone 2 BARLEY3 - rainfed barley Zone 3 COTTON - irrigated cotton MAIZE - irrigated yellow maize (corn) FALLBEET - irrigated fall sugar beets SUMBEET - irrigated summer sugar beets FALLPOT - irrigated fall potatoes SUMPOT - irrigated spring and summer potatoes winter crops (WIRR, WHYV1, WHYVZ, WLYVI, WLYV2, WLYV3, BARLEYI, BARLEYZ, BARLEY3, FALLBEET, FALLPOT) rainfed crops (WHYV1, WHYVZ, WLYVl, WLYV2, WLYV3, BARLEYl, BARLEYZ, BARLEY3) irrigated crops (WIRR, COTTON, MAIZE, FALLBEET, SUMBEET, FALLPOT, SUMPOT} irrigated wheat and cotton (WIRR, COTTON) wheat (WIRR, WHYV1, WHYVZ, WLYVl, WLYV2, WLYV3} rainfed wheat (WHYV1, WHYVZ, WLYVl, WLYV2, WLYV3} barley {BARLEY1, BARLEYZ, BARLEY3} summer crops (COTTON, HAIZE, SUMBEET, SUMPOT} 96 r - maize, sugar beets, and potatoes - (MAIZE, FALLBEET, SUMBEET, FALLPOT, SUMPOT} sg - production function segments for rainfed crops - (G001, ..., G100) sc - production function segments for irrigated wheat and cotton — (C001, ..., C143) sr - production function segments for maize, sugar beets, and potatoes - (R01, ..., R34} f - fertilizer nutrients - (N, P) e - rain scenarios - (GOOD, NORMAL, DRY, V-DRY} ingg papa (Expgepous Variables): CPRICEL. - economic field price of crop i under rain scenario e (sum FPRICE,‘- economic field price of fertilizer f (SL/kg) CPRICEFL.- financial field price of crop 1 under rain scenario e (SL/kg) FPRICEFf- financial field price of fertilizer f (SL/kg) FECROPL,- foreign exchange earning per unit of crop i under rain scenario e ($US/kg) FEFERTr- foreign exchange expenditures per unit of fertilizer f TAXL. ($US/kg) - indirect tax on crop i under rain scenario e (SL/kg) SUBSIDYf- subsidy on fertilizer f (SL/kg) AREA, FLIM, WFLIH, SFLIM, FPER, - total fertilized area planted to crop i (million ha) total requirements of fertilizer f (thousand tons) total winter requirements of fertilizer f (thousand tons) total summer requirements of fertilizer f (thousand tons) percentage of total requirements of fertilizer f assumed to be actually available (I) 97 WPER, - percentage of total winter requirements of fertilizer f assumed to be actually available (I) SPER, - percentage of total summer requirements of fertilizer f assumed to be actually available (I) FINOPTL, - optimum rate of fertilizer f on crop 1, based on financial prices (kg/ha) ECONOPT,',- optimum rate of fertilizer f on crop 1, based on economic prices (kg/ha) VDRYMAXLJ- the rate of fertilizer f on rainfed crop g that maximizes yield in a very dry year (kg/ha) PNRATIO; ratio of optimum P205 rate to optimum N rate, on crop r, based on economic prices FERTG,J‘- rate of fertilizer f associated with production function segment sg (kg/ha) FERTCL,5- rate of fertilizer f associated with production function segment sc (kg/ha) FERTR,J5- rate of fertilizer f associated with production function segment sr (kg/ha) YC yield increase of rainfed grain crop g, under rain scenario e, associated with production function segment sg (kg/ha) YCcm“ - yield increase of crop c (irrigated wheat or cotton) associated with production function segment sc (kg/ha) Yeru - yield increase of maize, sugar beet, or potato associated with production function segment sr (kg/ha) MINVCR. - minimum Value-Cost-Ratio under rain scenario e (no units) e d ou V ab GWGHTL“ - optimum weight associated with production function segment sg for crop g (no units) CWGHTC'“ - optimum weight associated with production function segment sc for crop c (no units) RWCHTL" - optimum weight associated with production function segment sr for crop r (no units) 0PTF,A - optimum rate of fertilizer f on crop i (kg/ha) ”1,0 - roman,- TOTBAR. - TOTCOT - TOTMAIZE - TOTSUG - TOTPOT - TFU‘ - “Tm! - STFU, - RETURNS, - warms,- SRETURNS - FINCOME, - WFINCOME. - SFINCOME - nouns, - covm, - 98 total production increase in crop 1, under scenario e, due to the application of the calculated optimum N and P205 rates (thousand tons) aggregate increase in wheat output under rain scenario e (thousand tons) aggregate increase in barley output under rain scenario e (thousand tons) aggregate increase in cotton output (thousand tons) aggregate increase in maize output (thousand tons) aggregate increase in sugar beet output (thousand tons) aggregate increase in potato output (thousand tons) total utilization of fertilizer f (thousand tons) total utilization of fertilizer f for the winter season (thousand tons) total utilization of fertilizer f for the summer season (thousand tons) net aggregate economic returns from total fertilizer use under rain scenario e (million SL) net aggregate economic returns from total fertilizer use in winter, under rain scenario e (million SL) net aggregate economic returns from total fertilizer use in summer (million SL) net aggregate financial returns from total fertilizer use under rain scenario e (million SL) net aggregate financial returns from total fertilizer use in winter, under rain scenario e (million SL) net aggregate financial returns from total fertilizer use in summer (million SL) net foreign exchange earnings associated with fertilizer use under rain scenario e (million $US) net government expenditures associated with fertilizer use under rain scenario e (million SL) D.l D.2 D.3 99 Z - The objective to be maximized; aggregate net economic returns to fertilizer use assuming a normal year for all crops, except for barley, where a dry year is assumed (million SL) V on: Mimize Z - 3,, (CPRICE".~m-0 * TY“.~m.-) + 2,, (CPRICEhumu * WWW.) + 2,, (CPRICEir,”NORHAL” * nunm") - 2, (FPRICE, * “mm on ant: u u t o Constraints: 2“ (GWGHTLu) - l 2.. (wean...) - 1 2.. (worm...) - 1 ., (GWGHTLu * FERTGL“) - 0PTFL, Z“ (CWGHTmu * FERTCL“) - OPTFt'c 2,, (momma * FERTRL") - OPTFL, W911 Optimum Fertilizer Rates: 0PTF,', s FINOPTL, OP‘I‘FL, s VDRYMAXL, 8 to V ”I! (CWCHTLu * YGLL“ * AREA‘) - TY... zac: (CWGHTmu * ch,e,sc * We) " TY“. 2s: (chmral: * YRr,e,sr * AREA!) - TYL- (3. (3. (3. (3. (3 (3 (3 (3 (3 (3 (3 (3 36) 37) 38) 39) .40) .41) .42) .43) .44) .45) .46) .47) D.4 D.5 D.6 D.7 D.8 100 MW Crop Output: nIt: (TYIh,e) + TYWIRR",0 - TOMEAT. 3b (”134) ' TOTBAR. TY-Pwm~.'m~ “' TY"smmm","nmx-1AL~ " TOTSUG TY‘PALLPOT“,”WL' + TY"SLHPOT”,'N®ML" ' TOTPOT Wm: (AREA, * 2, (PPRICE, * 0PTP,,,)) * MINVCR, - CPRICEL, * TY,” s o (AREA, * 2, (PPRICEP, * 0PTP,,,)) * MINVCR, - CPRICEFL, * TY,” s o Calgplgtipn pf Aggregagg Egzgilize; Use: 2,, (OPTF~.~.,, at AREA.) - w'rPu..,,.. xv (OPTF’°P",w * AREA”) 4’ (PWTIO’TALLBEIT‘ * OPTF"I",”PAIJ.BEET” * AREATALLBEET") + (PNRArIo..,,,,,,o,. * 0PTP...,.,,,,,O,. * AREA-Tum") - WTFUup 2, (0PTF~.~_, * AREA,) - 53m..- 2, (0PTP.,..,, * AREA,) + (PNRATIO-mm— * 0PTP...,.S,,,,,,,. * AREA.S,,,,,,,,..) "’ (PNRATIO~sm1Por~ * OPTF”I",”SIHPOT“ * AREA”SIHPOT”) + (PNRATIdmw * OP'rP-..,.,,,,,,. at AREA.,,,,,,..) - STFU-opu "Tm! + STFU: - TFU, e e va ab onstraints: TFU, s 0.01 * FPER, * FLIM, WTFU, s 0.01 * WPER, * WFLIM, STI-‘U, s 0.01 * SPER, at SFLIM, COO!!! 81.111: 2, (CPRICBL, * TY,',) - 2, (PPRICE, * TFU,) - RETURNS, (3. (3. (3. (3. (3. (3. (3. (3. (3. (3. (3. (3. (3. (3. (3. 48) 49) 50) 51) 52) 53) 54) 55) 56) 57) 58) 59) 60) 61) 62) D.9 D.10 D.ll 0.12 101 2, (CPRICEL, * TY”) (CPRICE~,,m~., * nm~,.) (CPRICEWALLBEET'J * Tymwmnd (CPRICE”PALLPOT',e * Tye/11.nov,.) 21 (FPRICE: * WTFU‘) - WRETURNS. .+++ 31: (CPRICEir,”m1AL” * “unsound ' (CPRICEMRR~.~m1AL" * TY"HIRR“,”NmHAL") ' (CPRICE"PALan~,'-m~ * TY"PALLBEEY~,"RORHAL») ‘ (CPRICE"PALLPOY','RORHAL- * TY”PALLPOT”,”NORHAL”) - t, (FPRICE: * STFU‘) - SRETURNS W: 2, (CPRICEFL, * TY,,,) - 2, (PPRICEP, * TFU,) - FINCOME, 2, (CPRICEFL, * TY”) + (CPRICEPm.,, * Wanna.) 4' (CPRIC£F”FALLB£ET”,0 * TY”FALLBEET”,0) + (CPRICEFupmm-o'. * TY"PALLPOT”,0) - 2, (PPRICEP, * WTFU,) - WFINCOHE. 31: (CPRICEFir,”mL” * “it,"uomuv) ' (CPRICEF~wIRR".~m1AL~ * TY"WIRR”,”NORMAL“) ' (CPRICEF"PALLBEEY","IORMAL" * TY”FALLBEET”,"NORHAL") ’ (CPRICEF”PALLPOT“,”NORHAL” * TY”FAL1.POT”,”NORHAL") - 2, (PPRICEP, * STFU,) - SFINCOME e can 8 2, (FECROPL, * TY,,,) - 2, (PEPERT, * 2211,) - DOLLARS, 0V 9 en U 2, (SUBSIDY, * 'rPU,)- 2, (TAXL, * 'rY,,,) - GOVEXP, 0' V T1 at: GWGHTL“ z 0, CWGHTCO“ z 0, RWGH'I‘L“. 2 0, TY,” z 0, GNP!" 2 0. (3. (3. (3. (3 (3 (3. (3. 63) 64) 65) .66) .67) 68) 69) and CHAPTER 4 ESTIMATION OF FERTILIZER AND CROP PRICES The conceptual differences between financial and economic prices were discussed in chapter 3. Financial prices are used to estimate costs and benefits associated with fertilizer use as seen from the farmers' viewpoint. Economic prices, on the other hand, refer to the true economic costs and benefits of fertilizer use from the point of view of the economy as a whole. The main purpose of this chapter is to present the procedures involved in estimating financial and economic prices and to discuss the procedures' underlying assumptions. These procedures are used to estimate financial and economic prices for nitrogen and phosphorus fertilizers and the crops included in this study. Foreign exchange earnings or expenditures associated with each crop and fertilizer included in this study are also estimated in this chapter. This is needed to assess the impact of alternative fertilizer allocations on the government's foreign exchange budget. Similarly, taxes or subsidies on crops and fertilizers are also estimated to assess the impact of alternative allocations on the government's general budget. 102 103 4.1 na e Financial prices of crops and fertilizers are the actual prices that farmers receive from the sale of their crops or pay for their fertilizer purchases. However, for the same commodity there exist several prices: official producer price, market price, farm price, field price, and so forth. In this study, the main interest is to estimate the farmer's net benefits from the application of fertilizers. Therefore, the most appropriate price to use would be the field price, which is defined as follows (CIMMYT, 1988, p. 25): The field price of the crop is defined as the value to the farmer of an additional unit of production in the field, prior to harvest. It is calculated by taking the price that farmers receive (or can receive) for the crop when they sell it, and subtracting all costs associated with harvest and sale that are proportional to yield, that is, costs that can be expressed per kilogram of crop. Ideally, the crop wholesale market price should be used as the basis for calculating financial field prices. However, in Syria the prices of most field crops are officially set by the government. This system of official pricing often involves compulsory delivery of all or a large proportion of the crop to the government. Official crop prices are generally lower than their international market equivalents. However, recent trends in official prices suggest that the government intends to increase these prices gradually to align them with prevailing international prices. In spite of these trends, parallel markets in most controlled agricultural comodities continue to operate. These markets tend to play a much greater role for cereals than for industrial crops such as 104 cotton and sugar beets. Given that the public sector has complete monopoly on. cotton. ginning and sugar refining .activities, parallel markets in cotton and sugar beets are almost non-existent. On the other hand, although the government has a legal monopoly over cereals marketing, an average of only 351 of total production is procured through official marketing channels (FAO, 1989, p. 50). The remaining is either sold in the parallel market, retained for seeds, or consumed on the farm. Discussions with farmers suggest that parallel market prices of cereals are, on the average, about 201 higher than official prices. During dry years these prices may be 40 to 601 higher than official prices, while in good years the margin declines to less than 101. Therefore, the basis for estimating crop field prices should be a weighted average between official and market prices, accounting for the relative shares of official deliveries and market sales. A similar approach also will be used in estimating fertilizer field prices. Given that fertilizers are sold to farmers at subsidized prices, parallel markets in fertilizers appear only if quantities demanded by farmers exceed the actual amounts distributed by the government. Situations of excess fertilizer demand have been most visible during very wet years such as 1987/88. However, fertilizer production problems and limited foreign exchange available for fertilizer imports have often contributed in the appearance of shortages even in normal years. During dry years, such as in 1988/89 and 1989/90, there was no visible sign of any fertilizer shortages. This may also be related to the recent substantial reduction in fertilizer subsidies, further lowering the quantity of fertilizer demanded. 105 Thus, except for dry years, farmers are expected to continue to rely on the parallel market for part of their fertilizer needs. Discussions with farmers indicate that they purchase an average of 201 of their fertilizer needs from the parallel market at average market prices 201 higher than official prices. It should be noted that whenever farmers purchase fertilizer from the Agricultural Cooperative Bank (ACB) they usually incur some additional transaction costs. These include the costs of the several trips that farmers usually have to make to the nearest ACB branch before they receive their fertilizer allotments, in addition to any illegal payoffs to ACB employees. Therefore, based on the above discussion, field prices of craps and fertilizers are estimated according to the following general approach: Crop field price - average producer price - transport costs - harvest costs - handling costs Fertilizer field price - average producer price + transaction costs + transport costs + handling costs + application costs 4.2 MW As mentioned earlier, financial field prices will be used in this study to estimate the impact of alternative fertilizer rates on farmers' net income. However, the calculation of these rates will be based on economic field prices. This is to ensure that the true economic value, or opportunity cost, of crops produced and of fertilizers used is taken into consideration. Therefore, whenever financial prices are suspected to be significantly different from their true economic values, then they need to be adjusted to make them more closely represent the opportunity cost to the economy as a whole. 106 Gittinger (1982) provided a detailed step-by-step procedure to use in adjusting financial prices to economic values (see pp. 250-271). This procedure involves three main steps: (1) adjustment for direct transfer payments; (2) adjustment for price distortions in traded items; and (3) adjustment for price distortions in nontraded items. A first step in estimating economic prices is to decide whether the crops and fertilizers in this study are to be treated as “traded" or “nontraded" goods. Traded goods include imports, exports, import substitutes, and diverted exports. Nontraded goods are those for which the domestic cost of production is higher than the export price but lower than the import price (ibid., p. 253). They also include goods that are not traded due to government policy. Nontraded goods are often bulky goods such as straw, or highly perishable goods such as fresh vegetables or fluid milk. The commodities covered by this study include exports, such as cotton and potatoes; imports, such as urea and TSP fertilizers; and import substitutes, such as wheat, sugar beets, and corn. Substantial barley exports occur only in good years, while Syria is usually self— sufficient in normal years. During dry years such as 1989 and 1990, there were clear indications of domestic barley shortages. In spite of these shortages the government maintained a ban on barley imports given the severe limitations on foreign exchange. Despite this ban on imports, it still seems appropriate to treat barley as a traded good. The estimation of economic field prices of traded goods is based on the calculation of Import or Export Parity Prices (IPP or EPP) (ibid., p. 269). These prices refer to commodity prices at the point of entry or exit of the country, or border prices, adjusted for any 107 domestic costs incurred in transferring these commodities from or to the main point where they are to be used. Therefore, for an imported commodity such as fertilizer, the import or CIF price (Costs, Insurance, and Freight) is adjusted by adding domestic costs incurred by the government. These costs include unloading, storage, transport, distribution, administration, and so forth, i.e., from the harbor to the main warehouses of the Agricultural Cooperative Bank (ACB). This would add up to the economic value of fertilizers at the warehouse. To compute the economic field price, transport, handling, and application costs incurred by farmers need also to be added. The same rationale is also applied to exports, such as cotton. The economic value of exported cotton is equal to the export or FOB price (Free on Board) minus (1) domestic transfer costs from the warehouses and cotton gins of the Cotton Marketing Organization (CMO) to the harbor, and (2) ginning costs. This would give the economic value of cotton at the gin. When farmers' harvesting and handling costs are subtracted, this would give the economic field price. In estimating the true economic values of costs or benefits faced by the economy, it is important to adjust these values by considering direct or indirect taxes or subsidies and all other distortions included in actual costs of fertilizers and crops. The Syrian government provides farmers with a wide range of subsidies, including subsidies on crop transport, credit and other inputs. Thus the main adjustments to be included in this study are those for (1) transport subsidies, (2) credit subsidies, (3) subsidies on other inputs, and (4) exchange rate adjustment. 108 1) WM: Transport subsidies are most commonly used on wheat, barley, and cotton. In the case of wheat and barley, government trucks usually collect the harvested grain at the farm gate at no cost to the farmer. A similar arrangement is made with cotton growers. Cotton farmers have to pay the cost of transport for the first 100 kilometers, with the government covering any additional transport costs. 2) ed ub d Credit subsidies are essentially in the form of interest-free loans for fertilizer purchases, which are to be repaid at harvest time. In other words, the average eight-month delay in loan repayments constitutes a real cost to the government and, by implication, to the economy as a whole. In Syria, all the formal financial institutions are entirely controlled by the government. There exist several official interest rates applying to different sectors of the economy. The annual interest rate on short-term agricultural loans is set at 41, while the interest rate on savings accounts is fixed at 7.52. The highest rate that can be obtained in the formal sector is the rate on government bonds, set at 91 annually. This is in contrast to annual interest rates ranging from 251 to 40! charged by local money lenders in the informal financial sector. Since the Opportunity cost of fertilizer loans are incurred by the government, the use ‘of the market rate would be inappropriate. Since the government relies heavily on issuing bonds to finance its budget deficit, the appropriate interest rate to use would be the highest rate that the government would have to pay, i.e., 91 per year. 109 3) b e n ut : Direct input subsidies cover several farm inputs such as pesticides, diesel fuel, seeds, farm equipment, bags, and so forth, besides fertilizer subsidies. Ideally, all these subsidies ought to be accounted for in estimating the true value of cr0ps. However, the main focus of this study is to assess the value of the increase in yield due to fertilizer use. In other words, the inputs of interest are only those that increase incrementally with any increase in yield. These are essentially limited to harvesting labor, transport, and bags. The government provides farmers with bags at a subsidized price equivalent to 160 SL per ton of grain, but with the condition that an equal number of filled bags be delivered after harvest. Therefore, farmers would need to buy bags from the market at prices twice as high as official prices for all their parallel market sales. 4) tme A final issue related to the estimation of economic prices is the question of which exchange rate to use in converting border prices (in US dollars) into Syrian Liras. As in many developing countries, the foreign exchange policy adopted in Syria is based on a fixed official exchange rate. This policy also imposes strict restrictions on private transactions and transfers of foreign currencies abroad. Several exchange rates are currently officially in use in Syria. Cowitt (1991, p. 770) lists the following official exchange rates in operation as of December 30, 1988 (all rates in Syrian Liras per U.S. Dollar): 110 A. Basic rate (theoretically defined at 336.375 milligrams of fine gold); inoperative . . . . . . . 2.19 B. Effective (Official) rate; applicable to official loans, grants and budgetary receipts, most exports, some travel earnings, public sector imports and invisibles payments (except travel) and capital transactions, with buying and selling rates of SL11.20/11.25 . 11.225 C. Promotion rate; applicable to private remittances, most travel and. tourism transactions, transfers of Syrian workers abroad, some export proceeds and medical expenses abroad . . . . . . . . . . . . . . 21.00 D. Special export rates; based on 1. Tax of 81 on shipments of fruit, vegetables and vegetable oil . . . . . . . . . . . . . . . . 10.35 2. Tax of 71 on shipments of all other agricultural products . . . . . . . . . . . . . . . . . . 10.46 E. Airline rates’; applicable to l. Purchases of airline tickets by residents and nonresidents . . . . . . . . . . . . . . . . 16.25 2. Airline company transfers abroad . . . . . . 18.00 Parallel market trading in the Syrian Lira has been in existence since the introduction of exchange controls in 1961. The parallel market rate is essentially determined in the Beirut foreign exchange market, where the value of the Syrian Lira is freely determined by supply and demand forces. The Beirut market rate (and by implication the parallel market rate inside Syria) witnessed a period of extreme fluctuations since the mid-1980's. The market rate increased from 13.85 SL/$US in December 1985 to 27.25 in December 1986, 45.00 in December 1987, and reached a maximum of 60.00 SL/SUS in May 1988 (ibid., p. 771). The Syrian Lira then went into a period of steady improvement, stabilizing at a range of 42 to 46 SL/SUS, with an average of 44 SL/SUS at mid-1989. 1 Abolished on October lst, 1990. 111 As a result of this renewed stability in the market exchange rate, the government recently introduced a “neighboring countries” (NC) rate. The NC rate is officially set at 40 SL/SUS, i.e., about 101 below the actual rate prevailing in the Beirut market. Although the official rate of 11.225 SL/SUS is still widely used in public transactions, the shift to the NC rate is becoming more common, especially in valuing public sector imports. Moreover, all signs suggest that this shift will prevail in the coming years, provided that the market rate maintains its current stability. Therefore, although the NC rate is slightly below the free market rate, its use by many public agencies justifies its use as a basis for estimating economic prices of fertilizers and crops in this study. To sum up the above discussion on the estimation of economic field prices, Import and Export Parity Prices (IPP and EPP) will be estimated based on the following approach: IPP (SL) - CIF Price ($US x 40 SL/SUS) + Government Costs (SL) - Farmer's Costs (SL) EPP (SL) - FOB Price ($US x 40 SL/$US) - Government Costs (SL) - Farmer's Costs (SL) Where, IPP - Field-level Import Parity Price (SL/kg) EPP - Field-level Export Parity Price (SL/kg) CIF Price - Costs Insurance and Freight ($US/kg) FOB Price - Free on Board ($US/kg) Government Costs - Administrative, transport, and other local costs (SL/kg) Farmer's Costs — Transport, handling, bags, and harvesting costs 112 4.3 go: ., a- -v sine! '- -. - a; f :n:; z . .,- a 2 -:: As discussed earlier, an important criterion in assessing the economic feasibility' of alternative fertilizer allocations is their impact on two key policy concerns: (1) foreign exchange earnings, and (2) the government's budget deficit. 4.3.1 prgign Exgbapgg Earnings egg Expendigpzeg To estimate the impact of alternative fertilizer allocations on the government's foreign exchange budget, we need first to estimate how much each additional unit of fertilizer used and crop produced would increase or decrease foreign exchange earnings. The underlying assumption is that, for any increase in fertilizer requirements, the government would need to import an equal amount to satisfy that increase. Thus, for each additional kilogram of fertilizer used the government's foreign exchange expenditures would increase by the import price of fertilizer plus any additional expenses paid in hard currencies (e.g., commissions to foreign banks or importing agents). Similarly, it is assumed that every additional kilogram of export crop produced would be exported and thus constitute additional foreign exchange earnings. These earnings are equal to the crop's export price minus any additional costs incurred in hard currencies. Conversely, any increase in the output of imported crops implies an equal decline in imports and savings in foreign exchange. These savings are equal to the import price plus any additional import-related hard currency expenses. As noted earlier, the above assumptions would apply only to crops that are completely controlled by the government (e.g., cotton and sugar beets). However, farmers usually sell only part of their cereals (e.g., 113 wheat, barley, and corn) to the government. Therefore, it is assumed that for any increase in cereal output, only part of that increase (i.e., whatever is sold to the government) would substitute for cereal imports. Thus, the contribution of each additional kilogram of cereal output to foreign exchange earnings would be equal to the import price multiplied by the fraction of total output sold to the government. Therefore, the general approach used in estimating the impact of fertilizers and crops on the government's foreign exchange budget is as follows: 1) For fertilizers: Foreign Exchange Expenditures ($US/kg) - CIF Price ($US/kg) + Other Import Costs ($US/kg) 2) For crops substituting for imports: Savings on Foreign Exchange ($US/kg) - (Fraction of total output sold to government) * (CIF Price ($US/kg) + Other Import Costs ($US/kg)) 3) For export crops: Foreign Exchange Earnings ($US/kg) - (Fraction of total output sold to government) * (FOB Price ($US/kg) - Other Export Costs ($US/kg)) 4.3-2 mm To assess the impact of alternative fertilizer allocations on the government's general budget, we need to estimate the implicit taxes or subsidies on the fertilizers and crops covered by this study. Implicit taxes or subsidies are defined as the difference between the commodity's true costs or revenues faced by the government and the official producer price. Therefore, the general approach used in estimating net taxes or 114 subsidies on a given commodity is to subtract the official price and the indirect subsidies (mainly transport subsidies and subsidies on bags) from the true economic value incurred by the government in purchasing or selling the commodity in question, as follows: 1) For fertilizers: Net Subsidies (SL/kg) - CIF Price ($US/kg x 40 SL/SUS) + Government Costs (SL/kg) + Indirect Subsidies (SL/kg) - Official Sales Price (SL/kg) 2) For crops substituting for imports: Net Taxes1 (SL/kg) - CIF Price ($US/kg x 40 SL/$US) + Government Costs (SL/kg) - Indirect Subsidies (SL/kg) - Official Purchase Price (SL/kg) 3) For export crops: Net Taxes (SL/kg) - FOB Price ($US/kg x 40 SL/SUS) - Government Costs (SL/kg) - Indirect Subsidies (SL/kg) - Official Purchase Price (SL/kg) 4.4 Isrtilizsr_Zrisss 4.4.1 Egzgilizgz Einancial Ericeg A list of 1989/90 official prices for all the nitrogen and phosphorus fertilizers sold in Syria is provided in Table 4.1. Of the five types of fertilizers sold, urea and. TSP are the predominant fertilizers used, with the remaining types constituting only a minor proportion of total use. Therefore, all the calculations included in this study will be based on urea and TSP. Based on the above official prices, fertilizer financial field prices are calculated by adding 1 A negative tax would indicate a subsidized commodity. 115 Table 4.1: OFFICIAL FERTILIZER SALES PRICES Syria, 1989/1990 Nutrient Content (1) Price Fertilizer Name N 1505 (SL/ton) Ammonium Nitrate (local): 30 -- 3400 Ammonium Nitrate (imported): 33.5 -- 3800 Urea: 46 -- 4900 Di-Ammonium Phosphate (DAP): 18 46 7100 Triple Superphosphate (TSP): -- 46 5200 Source: Decisions of the Higher Agriculture Council's meeting of 26 August 1989. fertilizer application costs and the costs of loading, unloading, and transport, to the weighted average of official and market prices (see Table 4.2). 4.4.2 E§I§11122I_E£222El£_211££§ Calculations of economic field prices for N and P205 fertilizers are summarized in Table 4.3. 4.4.3 WWW Foreign exchange expenditures per ton of imported fertilizer are equal to the border price plus all other import-related expenses incurred in hard currencies. This is given by the ”Port Prices" of 156 $US/ton and 228 $US/ton for urea and TSP, respectively (see Table 4.3). These prices are equivalent to 339 $US per ton of N and 496 $US per ton Of P205 . 116 Table 4.2: FINANCIAL PRICES OF N AND P205 FERTILIZERS Syria, October 1989. UREA TSP (SL/Ton) Official Price: 4900 5200 Margin between Market and Official Price (1): +20 +20 Market Price: 5880 6240 Market Purchases as a 1 of Total Purchases: 20 20 PRODUCER PRICE (Weighted Avg.): 5096 5408 Transport, Loading, and Unloading Costs: +240 +240 Transaction Costs: +28 +28 FARM PRICE: 5364 5676 Application Costs: +200 +200 FIELD PRICE: 5564 5876 FIELD PRICE OF PURE NUTRIENT 1: 12096 12800 l/ Urea contains 461 N; TSP contains 461 P20, Sources: Based on informal interviews with farmers in northern Syria. 117 Table 4.3: ECONOMIC PRICES OF N AND P505 FERTILIZERS Syria, October 1989 UREA TSP BORDER PRICE (CSF $US/ton): 130 190 Insurance (1.251): +2 +2 Commissions for Foreign Banks (7.51 for 3 months): +2 +4 Commissions of Importing Agency and Local Expenses (16.8751): +22 +32 PORT PRICE ($US/ton): 156 228 PORT PRICE1 (SL/ton): 6240 9120 Transport Costs (Port to Warehouse): +120 +120 Bags, Storage costs, and Insurance on Storage: +24 +24 WAREHOUSE PRICE: 6384 9264 Interest (91 for 4 months): +192 +278 Administrative Costs (41): +255 +371 PRODUCER PRICE: 6831 9913 Transport, loading, and unloading: +240 +240 Interest on fertilizer loan (91 for 9 months): +461 +669 FARM PRICE: 7532 10822 Application costs: +200 +200 FIELD PRICE (SL/ton): 7732 11022 FIELD PRICE OF PURE NUTRIENT2 (SL/ton): 16809 23961 l/ Converted at the “Neighboring Countries“ exchange rate of 40 SL/$US. 2/ Urea contains 461 N; TSP contains 461 P20, Sources: Soils Directorate internal documents. 118 4.4.4 Ne; Subeigiee en Eczeilizeze To calculate net subsidies on fertilizers, we need first to determine how much of the true economic value of fertilizer is incurred by the government. The true cost to the government of one ton of fertilizer is equal to the economic Producer Price (see Table 4.3) plus the interest on fertilizer loans, which are incurred by the government. In contrast, government revenues from the sale of one ton of fertilizer are equal to the official producer“ price. ‘Thus, net subsidies on fertilizers are then computed by subtracting the official sales price from the true economic value incurred by the government, as follows (refer to Table 4.2 and Table 4.3): 1) Net subsidies on urea - 6831 + 461 - 4900 - 2392 SL/ton. or Net subsidies on pure N - 2392/0.46 - 5200 SL/ton. 2) Net subsidies on TSP - 9913 + 669 - 5200 - 5382 SL/ton. or Net subsidies on P205 - 5382/O.46 - 11,700 SL/ton. 4.5 W 4.5.1 WW Official prices of wheat grain for the 1989/1990 season were set at 8.5 SL/kg for durum (or hard) wheat and 7.5 SL/kg for soft (or bread) wheat. Shortly prior to harvest, and given projections of a poor harvest due to low rains, the government decided to add 1 SL/kg to the above prices. This was done to provide additional incentives to farmers 119 to increase their wheat deliveries to the government1. However, these bonuses will not be included in the estimation of field prices. This is because fertilizer application decisions were made by the farmer based on the original official prices. In the fertilizer requirement schedule no distinction is made between durum and soft wheat. Instead, fertilizer recommendations are based on whether the planted wheat is irrigated or rainfed; on the Agricultural Stability Zone for rainfed. wheat; and on ‘whether the planted wheat varieties are local (LYV) or high-yielding (HYV). The 1989/1990 fertilizer allocation plan included the following six categories for wheat grain: Irrigated HYV Rainfed HYV Zone Rainfed HYV Zone Rainfed LYV Zone Rainfed LYV Zone Rainfed LYV Zone mmwar-o UNHNH However, no data are available on the relative shares of durum and soft ‘wheat varieties according, to the above categories. Based. on discussions with officials from the Seed Multiplication Establishmentz, it is estimated that about two-thirds of the HYV varieties grown in Syria are soft wheat and one-third are durum. In contrast, the vast majority of local varieties are durum wheat. Therefore, the official price of LYV wheat grain is assumed to be equal to that of durum wheat 1 Based on the decision of the Higher Agriculture Council's meeting of 22 May, 1990. 2 The Seed Multiplication Establishment is a semi-autonomous agency of the Ministry of Agriculture and Agrarian Reform responsible for the production and distribution of certified seeds, including HYV wheat seeds. 120 (8.5 SL/kg), while the official price of HYV wheat grain is 7.83 SL/kg (based on a weighted average of 2/3 soft and 1/3 durum). To calculate wheat grain 'producer prices, a. weighted average between official and. parallel market prices needs to be computed. Discussions with farmers indicated that market prices are usually about 201 higher than official prices in a normal year. This margin would increase to about 301 and 401 during dry and very dry years, respectively. During good years, the margin between market and official prices would decline to about 101 (see Table 4.4). The level of rainfall is also expected to affect the proportion of total output delivered to the government. Discussions with government officials suggested that, for HYV wheat, this proportion is around 401 in good and normal years, and 301 in drier years. Given that consumers in the rural areas prefer durum to soft wheat, a relatively smaller proportion of durum wheat is sold to the government. This proportion is around 301 in good and normal years and 201 in drier’ years (see Table 4.5). Weighted averages are computed for wheat grain producer prices under different rainfall scenarios, based on the above estimates of the margins between official and market prices and the shares of total output sold to the government (see Table 4.4 and Table 4.5). To estimate field prices, harvesting, handling, and transport costs need to be subtracted from producer prices. However, farmers do not pay any transport costs on official procurements since the crop is delivered at the farm gate to government collectors. Also, the number of the subsidized government-supplied bags is proportional to the volume of official deliveries. Therefore, as can be noted from Table 4.4 and Table 4.4: FINANCIAL PRICES 0F HYV WHEAT GRAIN Syria, October 1989. RAIN SCENARIOS1 GOOD NORMAL DRY V. DRY (SL/Ton) Official Pricezz 7830 7830 7830 7830 Margin between Market and Official Price (1): +10 +20 +30 +40 Market Price: 8613 9396 10179 10962 Official Sales as a 1 of Total Output: 40 40 30 30 PRODUCER PRICE’: 8300 8770 9474 10022 Transport Costs‘: ~96 ~96 ~112 ~112 FARM PRICE: 8204 8674 9362 9910 Harvesting costs as 1 of Gross Revenues: 10 10 12 12 Harvesting Costs: ~820 ~867 ~1l23 ~1189 Loading and Unloading Costs: ~60 ~60 ~60 ~60 Government-Supplied Bags‘: ~64 ~64 ~48 ~48 Market Bags‘: ~192 ~192 ~224 ~224 FIELD PRICE: 7068 7491 7907 8389 1/ See Table 3.1 for the definitions of rain scenarios. 2/ Weighted average between hard and soft wheat prices. 3/ Weighted average between official and market prices. 4/ These costs are already' adjusted. to account for the relative shares of official vs market sales. Sources: Based on informal interviews with farmers in northern Syria. Table 4.5: FINANCIAL PRICES OF LYV WHEAT GRAIN Syria, October 1989. RAIN SCENARIOS1 GOOD NORMAL DRY V. DRY (SL/Ton) Official Price: 8500 8500 8500 8500 Margin between Market and Official Price (1): +10 +20 +30 +40 Market Price: 9350 10200 11050 11900 Official Sales as a 1 of Total Output: 30 30 20 20 PRODUCER PRICEZ: 9095 9690 10540 11220 Transport Costsa: ~112 ~112 ~128 ~128 FARM PRICE: 8983 9578 10412 11092 Harvesting costs as 1 of Gross Revenues: 10 10 12 12 Harvesting Costs: ~898 ~958 ~1249 ~133l Loading and Unloading Costs: ~60 ~60 ~60 ~60 Government-Supplied Bags1: ~48 ~48 ~32 ~32 Market Bagsaz ~224 ~224 ~256 ~256 FIELD PRICE: 7753 8288 8815 9413 1/ See Table 3.1 for the definitions of rain scenarios. 2/ Weighted average between official and market prices. 3/ These costs are already' adjusted. to .account for the relative shares of official vs market sales. Sources: Based on informal interviews with farmers in northern Syria. 123 Table 4.5, a weighted average was used in computing the cost of transport and bags. Also, it should be noted that harvesting costs are expressed as a percentage of gross revenues. This is because most grain harvesting in Syria is usually done by independent harvesting contractors. These contractors retain an average of 101 of the harvested grain in return for their services. During drier years these contractors charge a higher rate, averaging 121. 4.5.2 Eeonomie Epices of,Wheet Grain Given that Syria imports substantial amounts of wheat every year, wheat import (CIF) prices are used as the basis for calculating economic prices for wheat grain. The CIF price of soft wheat is estimated at 200 $US/ton and that of durum wheat at 212 $US/ton. As mentioned earlier, about two-thirds of all HYV varieties are soft wheat varieties and one~ third durum varieties, while all local varieties are durum wheat. Thus, by taking the weighted average the CIF price for HYV wheat grain would be equal to 204 $US/ton, while the price of LYV wheat grain would be equal to 212 $US/ton. The calculations of economic field prices of wheat grain are presented in Table 4.6 and Table 4.7. 4.5.3 Eingngi§l_£11££§_2fgflhfié£_§£lé! The above estimations of financial and economic prices of wheat referred to wheat grain only. However, fertilizer use is expected to increase both grain and straw yields. Therefore, the value of straw ought to be included in the economic analysis of fertilizer use. Straw is usually stored on the farm and fed to livestock during the winter season. There are few economic incentives for farmers to sell straw in Table 4.6: ECONOMIC PRICES OF HYV WHEAT GRAIN Syria, October 1989. RAIN SCENARIOS1 GOOD NORMAL DRY V. DRY BORDER PRICE CIF ($US/ton)2: 204 204 204 204 Local Expenses (101): +20 +20 +20 +20 PORT PRICE $US/ton: 224 224 224 224 SL/ton: 8960 8960 8960 8960 Transport Costs (Port to Warehouse): +120 +120 +120 +120 PRODUCER PRICE: 9080 9080 9080 9080 Transport Costs (Farm to Warehouse): ~160 ~160 ~160 ~160 FARM PRICE: 8920 8920 8920 8920 Harvesting costs as 1 of Gross Revenues: 10 10 12 12 Harvesting Costs: ~892 ~892 ~1070 ~1070 Loading and Unloading Costs: ~60 ~60 ~60 ~60 Bags: ~320 ~320 ~320 ~320 FIELD PRICE: 7648 7648 7470 7470 1/ See Table 3.1 for the definitions of rain scenarios. 2/ Weighted average between the import prices of soft and durum wheat Sources: Based on discussions with officials from the Soils Directorate. Table 4.7: ECONOMIC PRICES OF LYV WHEAT GRAIN Syria, October 1989. 125 RAIN SCENARIOS1 GOOD NORMAL DRY V. DRY BORDER PRICE CIF ($US/ton): 212 212 212 212 Local Expenses (101): +21 +21 +21 +21 PORT PRICE $US/ton: 233 233 233 233 SL/ton: 9320 9320 9320 9320 Transport Costs (Port to Warehouse): +120 +120 +120 +120 PRODUCER PRICE: 9440 9440 9440 9440 Transport Costs (Farm to Warehouse): ~160 ~160 ~160 ~160 FARM PRICE: 9280 9280 9280 9280 Harvesting costs as 1 of Gross Revenues: 10 10 12 12 Harvesting Costs: ~928 ~928 ~lll4 ~lll4 Loading and Unloading Costs: ~60 ~60 ~60 ~60 Bags: ~320 ~320 ~320 ~320 FIELD PRICE: 7972 7972 7786 7786 1/ See Table 3.1 for the definitions of rain scenarios. Sources: Based on discussions with officials from the Soils Directorate. 126 the local market unless market prices are high enough to justify the very high costs of transport and bags. As shown in Table 4.8, the average market price of straw in normal years is 3200 SL/ton. This is slightly higher than the total costs that farmers would have to incur for harvesting, bagging, and transporting straw to the market. In good years, the supply of grain and straw is usually abundant enough to drive straw 'prices down to around 1750 SL/ton, which is less than total farmer's costs. Table 4.8: FINANCIAL PRICES OF WHEAT STRAW Syria, October 1989. RAIN SCENARIOS1 GOOD NORMAL DRY V. DRY ———_ (SL/Ton) MARKET PRICE: 1750 3200 5600 8000 Transport Costs: ~480 ~480 ~480 ~480 FARM PRICE: 1270 2720 5120 7520 Harvesting costs as 1 of Gross Revenues: 50 50 50 60 Harvesting Costs: ~635 ~1360 ~2560 ~4512 Loading and Unloading Costs: ~60 ~60 ~60 ~60 Bags: ~960 ~960 ~960 ~960 FIELD PRICE: 0 340 1540 1988 1/ See Table 3.1 for the definitions of rain scenarios. Sources: Based on informal interviews with farmers in northern Syria. In other words, for most farmers it would be uneconomical to market their straw under such conditions. This is confirmed by the phenomenon of straw burning in farmers' fields often observed during 127 good years such as in 1988. In drier years, on the other hand, feed shortages would drive market straw prices up to 8000 SL/ton. These prices would constitute a greater economic incentive for farmers to harvest and sell their straw in the market. 4.5.4 W The above discussion illustrates the importance of including straw prices in valuing the increase in wheat yield as a result of fertilizer use. This is particularly important in dry and very dry years. Given that there is no indication that straw prices are affected by price distortions, economic and financial prices of straw are assumed to be equal. Research results at ICARDA suggest that for each ton of wheat grain harvested, an average of 470 kg of straw is also produced (Nordblom and Thomson, 1987, p. 9). This ratio is taken into consideration in the estimation of financial and economic field prices of total wheat output, as shown in Table 4.9. 4.5.5 WW For every ton of additional HYV wheat output, 300 to 400 kg are sold to the government, depending on seasonal rainfall (see Table 4.4). As mentioned earlier, wheat imports are assumed to decline by an amount equal to the quantities of wheat marketed through official channels. Given a port price of 224 $US/ton (see Table 4.6), an increase of one ton in HYV wheat output would result in 89.6 $US (i.e., 224 $US * 401) of savings in import costs in normal and good years, and 67.2 $US (i.e., 224 $US * 301) in dry and very dry years. Table 4.9: PRICES OF TOTAL WHEAT OUTPUT Syria, October 1989. RAIN SCENARIOS1 GOOD NORMAL DRY V. DRY (SL/Ton) (1) Financial Price of HYV Wheat Grainzz 7068 7491 7907 8389 (2) Financial Price of LYV Wheat Grain3: 7753 8288 8815 9413 (3) Economic Price of HYV Wheat Grain‘: 7648 7648 7470 7470 (4) Economic Price of LYV Wheat Grain’: 7972 7972 7786 7786 (5) Straw Price‘: 0 340 1540 1988 (6) Quantity of Straw Harvested per Ton of grain (tons): 0.47 0.47 0.47 0.47 (7) Straw Value per Ton of Grain (5)x(6): 0 160 724 934 FINANCIAL PRICE OF TOTAL HYV WHEAT OUTPUT (l)+(7): 7068 7651 8631 9323 FINANCIAL PRICE OF TOTAL LYV WHEAT OUTPUT (2)+(7): 7753 8448 9539 10347 ECONOMIC PRICE OF TOTAL HYV WHEAT OUTPUT (3)+(7): 7648 7808 8194 8404 ECONOMIC PRICE OF TOTAL LYV WHEAT OUTPUT (4)+(7): 7972 8132 8510 8720 1/ See Table 3.1 for the definitions of rain scenarios. 2/ From Table 4.4; 3/ From Table 4.5; 4/ From Table 4.6; 5/ From Table 4.7; 6/ From Table 4.8. 129 The same approach is used to estimate foreign exchange savings from increased LYV wheat production. Given a port price of 233 $US/ton (see Table 4.7), an increase of one ton in total output would result in 69.9 $US (233 $US * 301) in foreign exchange savings in normal and good years. In dry and very dry years only 201 of total LYV wheat output is delivered to the government (see Table 4.5). Thus, net foreign exchange savings per ton of additional output would be equal to 46.6 $US (233 $US * 201). 4.5.6 Ee§_1exes on Wheat Grain To calculate net taxes (or subsidies) on wheat grain, we need first to determine how much of the true economic value of wheat is incurred by the government. In other words, the wheat purchased from farmers has an Opportunity cost to the government equal to the import price plus all additional government costs. Given that the government incurs the transport costs from the farm gate, the economic Farm Price would be the opportunity cost of the wheat purchased by the government. Also, the subsidies on bags need to be included in estimating net taxes on wheat, as follows (see Table 4.6 and Table 4.7): Net taxes on wheat - Economic Farm Price ~ Subsidies on bags ~ Official Price Net taxes on HYV wheat - 8920 ~ (320 ~ 160) ~ 7830 - 930 SL/ton and Net taxes on LYV wheat - 9280 ~ (320 ~ 160) ~ 8500 - 620 SL/ton The above estimates of implicit taxes on wheat apply only to official government purchases. However, as mentioned earlier, only 401 of total HYV wheat output is sold to the government in good and normal years and 301 in drier years. The figures for LYV wheat are even lower, 130 with an estimated 301 of total output sold to the government in good and normal years and 201 in drier years. Therefore, in estimating the average tax paid per ton of additional wheat produced, an approach based on a weighted average needs to be used, as follows: Average Net taxes on HYV wheat: A) Good and normal years - 930 x 401 - 372 SL/ton B) Dry and very dry years - 930 x 301 - 279 SL/ton Average Net taxes on LYV wheat: A) Good and normal years - 620 x 301 - 186 SL/ton B) Dry and very dry years - 620 x 201 - 124 SL/ton 4.6 W 4.6.1 Eipapeiei Ezicee pf Eagley Grain The official price of barley for the 1989/1990 season was set at 5.5 SL/kg. As mentioned earlier, an additional 1.5 SL/kg was announced prior to harvest as an incentive to increase farmers' crop deliveries to the government. However this increase will not be included in the estimation of field prices given that it was announced after fertilizer application time. To calculate barley grain producer prices, a weighted average between official and market prices is computed, as shown in Table 4.10. This is based on the assumption that market prices would be 201 higher than official prices in a normal year. This percentage would increase to 401 and 601 during dry and very dry years, respectively. During good years, on the other hand, market prices tend to be very close to official prices. Therefore, in comparison to wheat prices, barley grain prices are much more susceptible to fluctuations in rainfall than wheat prices. Table 4.10: FINANCIAL PRICES OF BARLEY GRAIN Syria, October 1989. RAIN SCENARIOS1 GOOD NORMAL DRY V. DRY (SL/Ton) Official Price: 5500 5500 5500 5500 Margin between Market and Official Price (1): 0 +20 +40 +60 Market Price: 5500 6600 7700 8800 Official Sales as a 1 of Total Output: 50 40 30 25 PRODUCER PRICEZ: 5500 6160 7040 7975 Transport Costs3: ~80 ~96 ~112 ~l20 FARM PRICE: 5420 6064 6928 7855 Harvesting costs as 1 of Gross Revenues: 8 10 12 15 Harvesting Costs: ~434 ~606 ~83l ~ll78 Loading and Unloading Costs: ~60 ~60 ~60 ~60 Government-Supplied Bags3: ~80 ~64 ~48 ~40 Market Bags’: ~160 ~192 ~224 ~240 FIELD PRICE: 4686 5142 5765 6337 1/ See Table 3.1 for the definitions of rain scenarios. 2/ Weighted average between official and market prices. 3/ These costs are already' adjusted. to account for the relative shares of official vs market sales. Sources: Based on informal interviews with farmers in northern Syria. 132 This is 'because barley' is grown. in. the drier' areas ‘where rainfall fluctuations are more accentuated. Also, barley prices are very much affected by the erratic rainfall levels in the steppe. This is because of the close interrelationship between feed prices and the availability of natural pastures. Furthermore, barley prices are more sensitive to rainfall than wheat prices because little barley is imported to offset domestic production shortfalls. Similarly, the level of rainfall is also expected to affect the proportion of total barley output delivered to the government. As shown in Table 4.10, official deliveries are estimated to decline from around 501 of total output during good years down to 251 in very dry years. 4.6.2 Eeenpmie Prieee pf Barley Craig As discussed earlier, barley poses some complications in deciding whether to treat it as traded or nontraded good. Until the late 1970's, Syria regularly exported substantial amounts of barley. These exports have been declining steadily due to increased domestic demand for feed driven by high meat prices. During the 1980's, significant barley exports continued only in good years, while imports were on the increase especially during dry years such as 1984. As a result the government has attempted to ban barley imports to save on foreign exchange. However, after two consecutive dry years and a substantial increase in ‘barley market prices, the government allowed some barley imports in 1990. Therefore, despite the current ban on barley imports, it is expected that such a ban would be partially lifted whenever there are signs of significant barley shortages. Ba FIJ § 1/ Table 4.11: ECONOMIC PRICES OF HARLEY GRAIN Syria, October 1989. RAIN SCENARIOS1 GOOD NORMAL DRY V. DRY BORDER PRICES CIF ($US/ton): ~~ 160 160 160 FOB ($US/ton): 135 ~~ ~~ ~~ Local Expenses (101): ~14 +16 +16 +16 PORT PRICE $US/ton: 121 176 176 176 SL/ton: 4840 7040 7040 7040 Transport Costs (Port to Warehouse): ~120 +120 +120 +120 PRODUCER PRICE: 4720 7160 7160 7160 Transport Costs (Farm to Warehouse): ~160 ~160 ~160 ~160 FARM PRICE: 4560 7000 7000 7000 Harvesting costs as 1 of Gross Revenues: 8 10 12 15 Harvesting Costs: ~365 ~700 ~840 ~1050 Loading and Unloading Costs: ~60 ~60 ~60 ~60 Bags: ~320 ~320 ~320 ~320 FIELD PRICE: 3815 5920 5780 5570 1/ See Table 3.1 for the definitions of rain scenarios. Sources: Based on discussions officials from the Soils Directorate. 134 Therefore, in dry and very dry years the estimation of the economic price of barley grain would be based on the import (CIF) price, while in good years the export (FOB) price would be used. In normal years, Syria is usually self-sufficient in barley. However, with the continued increase in the demand for meat, Syria is expected to import barley even during normal years. Thus, any increase in barley output due to fertilizer use would substitute for such barley imports. This would justify the use of the CIF price as the basis for estimating economic prices in normal years. The estimation of the economic prices of barley grain is presented in Table 4.11. 4.6.3 Eigeneiel Erieee pf Barley firrew As in the case of wheat, fertilizer use on barley is also expected to increase both grain and straw yields. In fact, barley straw has a greater economic value than wheat straw given its higher nutritional content (see Nordblom and Thomson, 1987, p. 17). This is often reflected in an average market price for barley straw 251 higher than that of wheat straw. Except for the differences in market prices, the estimation of the field price of barley straw is identical with the procedure used for wheat straw, as shown in Table 4.12. 4.6.4 WWW For each ton of barley grain harvested, an estimated 530 kg of straw are produced (Nordblom and Thomson, 1987, p. 7). Based on this ratio, the value of straw is added to the estimated financial and economic field prices of barley grain (from Table 4.10 and Table 4.11) to obtain the field prices of total barley output (Table 4.13). a: me 10] of cum ($99 Paula would gIVES 135 Table 4.12: FINANCIAL PRICES OF BARLEY STRAW Syria, October 1989. RAIN SCENARIOS1 GOOD NORMAL DRY V. DRY ———_ (SL/Ton) MARKET PRICE: 2200 4000 7000 10000 Transport Costs: ~480 ~480 ~480 ~480 FARM PRICE: 1720 3520 6520 9520 Harvesting costs as 1 of Gross Revenues: 50 50 50 60 Harvesting Costs: 860 ~1760 ~3260 ~5712 Loading and Unloading Costs: ~60 ~60 ~60 ~60 Bags: ~960 ~960 ~960 ~960 FIELD PRICE: 0 740 2240 2788 1/ See Table 3.1 for the definitions of rain scenarios. Sources: Based on informal interviews with farmers in northern Syria. 4.6.5 Ne; Sevinge in Eereigg Exehange pp Barley Qreig The contribution of additional barley production to the foreign exchange budget is based on the Port Price of barley grain. As mentioned earlier, barley is usually exported in good years. Thus, foreign exchange earnings from barley exports would equal the FOB price of barley grain minus all other export-related expenses paid in hard currencies. This would give a Port Price of 121 $US/ton in good years (see Table 4.11). In normal and dry years, additional barley output would substitute for barley imports. Thus, savings in foreign exchange would be equal to the CIF price plus import-related expenses, which gives a Port Price of 176 $US/ton. Table 4.13: PRICES OF TOTAL BARLEY OUTPUT Syria, October 1989. RAIN SCENARIOS1 GOOD NORMAL DRY V. DRY (SL/Ton) (1) Financial Price of Barley Grainzz 4686 5142 5765 6337 (2) Economic Price of Barley Grain3: 3815 5920 5780 5570 (3) Straw Price‘: 0 740 2240 2788 (4) Quantity of Straw Harvested per Ton of grain (tons): 0.53 0.53 0.53 0.53 (5) Straw Value per Ton of Grain (3)x(4): 0 392 1187 1478 FINANCIAL PRICE OF TOTAL BARLEY OUTPUT (l)+(5): 4686 5534 6952 7815 ECONOMIC PRICE OF TOTAL BARLEY OUTPUT (2)+(5): 3815 6312 6967 7048 1/ See Table 3.1 for the definitions of rain scenarios. 2/ From Table 4.10; 3/ From Table 4.11; 4/ From Table 4.12 137 The above port prices need to be further adjusted to reflect the share of total barley production sold to the government (see Table 4.10). This would allow the estimation of the amount of foreign exchange contributed by each additional ton of barley output: Foreign exchange earnings from barley: A) Good years - 121 $US/ton x 501 - 60.5 $US/ton B) Normal years - 176 $US/ton x 401 - 70.4 $US/ton C) Dry years - 176 $US/ton x 301 - 52.8 $US/ton D) Very dry years - 176 $US/ton x 251 44.0 $US/ton 4.6.6 Net Taxes_on_Bar1ey Greip As in the case of wheat grain, the net taxes (or subsidies) on barley grain are calculated based on the economic Farm Price. This price is equal to 7000 SL/ton in normal and dry years (see Table 4.11). As discussed earlier, the estimation of the economic price of barley in good years is based on its export price. This would give an economic Farm Price of 4560 SL/ton, as shown in Table 4.11. Thus, net taxes on barley are computed as follows: Net taxes on barley - Economic Farm Price ~ Subsidies on bags ~ Official Price In good years - 4560 ~ (320 ~160) ~ 5500 - ~1110 SL/ton In normal and dry years - 7000 ~ (320 ~ 160) - 5500 - 1340 SL/ton The above figures are further adjusted to reflect the share of total barley production sold to the government (see Table 4.10). This would allow the estimation of the average tax paid per ton of additional barley produced, as follows: Average net taxes on barley: A) Good years ~1100 x 501 - ~550 SL/ton B) Normal years 1340 x 401 - 536 SL/ton C) Dry years - 1340 x 301 - 402 SL/ton D) Very dry years 1340 x 251 - 335 SL/ton 138 4.7 Cerreg Brice; 4.7.1 Eigeneial Eriee pf Raw Cotton Cotton is the most important export crop in Syria. Several detailed studies on cost of production and marketing have been conducted by the Cotton Bureau and the Cotton Marketing Organization (CMO). As mentioned earlier, the CMO has total monopoly on domestic marketing, ginning, and export of cotton, with insignificant quantities of raw cotton sold in the parallel market. Therefore, the official price will be the basis for estimating the financial field price of raw cotton. Official prices for raw cotton for the 1990 season were set as follows: 19 SL/kg for deliveries before the end of October (base price); 17 SL/kg for deliveries before the end of November; and 15 SL/kg for later deliveries. Based on figures from the last five seasons (1984 to 1988), CMO officials estimated that the price received by farmers is, on average, equal to 90.71 of the base price. Thus, for a base price of 19,000 SL/ton, the average producer price for the 1990 season would be equal to 17,233 SL/ton. Transport and harvesting costs are subtracted from the average producer price to give a financial field price of 13,525 SL/ton (see Table 4.14). 4.7.2 W The CMO makes its export price projections based on expected spot prices of cotton lint at the end of the season (December). In January 1990, for instance, the future price of lint cotton for March 1990 deliveries was 67.91 US cents/1b (FOB New York). The future price for December 1990 deliveries was 63.87 cents/lb, i.e., a decline of 4c04 cents/1b. Therefore, it was expected that the January 1990 spot price 139 Table 4.14: FINANCIAL AND ECONOMIC PRICES OF RAW COTTON Syria, January 1990. FINANCIAL ECONOMIC BORDER PRICE1 (FOB $US/ton): 572.1 Local Expenses (51): ~28.6 PORT PRICE: ($US/ton): 543.5 (SL/ton): 21740 Ginning Costs (51 of producer price): ~862 Interest (91 for 6 months): ~775 PRODUCER PRICE (SL/ton): 17233 20103 Transport Cost Differentialzz ~~ ~100 Farmer's Transport Costszz ~45 ~45 FARM PRICE (SL/ton): 17188 19958 Harvest Labor Costs: ~2500 ~2500 Transport Costs of Harvest Labor: ~750 ~750 Bags: ~180 ~267 Bagging: ~133 ~133 Loading and Pressing: ~100 ~100 FIELD PRICE (SL/ton): 13525 16208 1/ Includes transport costs from the gins to the port. 2/ Farmers pay for the first 100 Km. and the Cotton Marketing Organization pays the extra transport costs. Sources: Based on discussions with officials from the Cotton Bureau and the Cotton Marketing Organization. 140 of 76.75 cents/1b (CIF Northern Europe) also would decline by 4.04 cents/lb to give an expected spot price of 72.71 cents/lb in December 1990. Shipping costs from the cotton gins in Syria to Northern Europe are estimated at 3.5 US cents/1b, which gives an FOB (Syria) price of 69.21 cents/1b, or 1524.4 $US/ton. However, prices need to be expressed in terms of raw cotton rather than lint cotton. According to the CMO, one ton of raw cotton produces, after ginning, an average of 343.3 kg of lint cotton, 610 kg of cotton seeds, and 46.7 kg of waste. All the cotton seed produced in Syria is sold locally at an average price of 3.25 SL/kg, or 80 $US/ton (based on the exchange rate of 40 SL/SUS). Therefore, the FOB price of raw cotton is estimated as follows: FOB Price - (1524.4 x 0.3433) + (80 x 0.61) - 572.1 $US/ton To estimate the economic field ‘price of raw' cotton, the costs of ginning, transport, harvesting, and other local costs are subtracted from the border price to give a field price of 16,208 SL/ton (see Table 4.14). 4.7.3 WW Since cotton marketing is completely controlled by the government, any increase in cotton production will be reflected by an equal increase in exports. Thus foreign exchange earnings from additional cotton production would be equal to the Port Price of raw cotton estimated at 543.5 $US/ton (see Table 4.14). 141 4.7.4 Ne; Ieree pp Raw Cotton As for the estimation of net taxes on raw cotton, this is done by subtracting the official price and the subsidies on transport and bags from the economic Producer Price, as follows (see Table 4.14): Net taxes on raw cotton - Economic Producer Price ~ Transport Subsidy ~ Subsidies on bags ~ Official Price - 20103 ~ 100 ~ (267-180) ~ 17233 - 2683 SL/ton 4.8 9231111221 4.8.1 W The official price of corn grain for the 1989/1990 season is 7000 SL/ton, while the market price was expected to be 201 higher. Given that about 501 of all corn output is sold to the government, the weighted average producer price would be 7700 SL/ton (see Table 4.15). Transport, handling, and harvesting costs are subtracted from the average producer price to give a financial field price of 6910 SL/ton. 4.8.2 W As for the estimation of economic prices, the expected 1990 import (CIF) price of corn was 170 $US/ton. After adding local expenses and transport costs, the economic producer price would amount to 7600 SL/ton (see Table 4.15). The field price is computed by subtracting farmers' costs, which would give an economic field price of 6340 SL/ton. 4.8-3 WW For an increase of one ton in corn output, the government receives 500 kg that would substitute for corn imports. Given a port price of 142 Table 4.15: FINANCIAL AND ECONOMIC PRICES OF CORN Syria, March 1990. FINANCIAL ECONOMIC BORDER PRICE (CIF $US/ton): 170 Local Expenses (101): +17 PORT PRICE: ($US/ton): 187 (SL/ton): 7480 Transport Costs (Port to Warehouse): +120 Official Price: 7000 Margin Between Market and Official Price (1): +20 Market Price: 8400 Official Sales as a 1 of Total Output: 50 PRODUCER PRICE (SL/ton)1: 7700 7600 Transport, Loading, and Unloading Costs: ~200 ~200 FARM PRICE (SL/ton): 7500 7400 Harvesting Costs (101): ~750 ~740 Bags: ~2402 ~320 FIELD PRICE (SL/ton): 6510 6340 1/ Weighted average between market and official prices. 2/ Weighted average between market and government-supplied bags. Sources: Based on discussions with officials from the Soils Directorate. 143 187 $US/ton, the foreign exchange earnings for each additional ton of corn output would be equal to 93.5 $US (i.e., 187 $US/ton * 501). 4.8.4 Ne; Ieree op Qprp The estimation of net taxes on corn is done by subtracting the official price and the subsidies on bags from the economic Producer Price, as follows: Net taxes on corn - Economic Producer Price ~ Subsidies on bags ~ Official Price - 7600 ~ (320 ~ 240 ) ~ 7000 - 520 SL/ton Given that an estimated 501 of total corn production is sold to the government, the average net tax on corn would be equal to 260 SL/ton (i.e., 520 SL/ton * 501). 4.9 W 4.9.1 Elpageiel Eriee pf Buger Beere The official price for sugar beet roots was set at 1250 SL/ton for the 1989/1990 season. The General Establishment for Sugar Refining (GESR) has total monopoly on sugar refining and it is the sole legal buyer of sugar beet roots. Although, in the past, farmers have been observed to sell their beets as feed for livestock, the recent increases in official prices have drastically reduced such practices. Therefore, the official price is the actual price that most farmers receive. By subtracting harvest and transport costs from the official price, the financial field price would then be equal to 1025 144 Table 4.16: FINANCIAL AND ECONOMIC PRICES OF SUGAR BEET Syria, March 1990. FINANCIAL ECONOMIC BORDER PRICE OF REFINED SUGAR (CIF $US/ton): 400 Local Expenses: +25 PORT PRICE OF REFINED SUGAR: ($US/ton): 425 (SL/ton): 17000 Transport Costs: +120 Extraction and Refining Costs: +1737 REFINERY PRICE OF REFINED SUCAR (SL/ton): 18857 PRODUCER PRICE OF SUCAR BEET ROOTS1 (SL/ton): 1250 1544 Harvest and Transport Costs: ~225 ~225 FIELD PRICE (SL/ton): 1025 1319 1/ One ton of sugar beet roots yields an average of 81.9 kg of refined sugar Sources: Based on discussions with officials from the General Establishment for Sugar Refining. 145 SL/ton (Table 4.16). Given an estimated 225 SL/ton in harvesting and transport costs, the field price of sugar beet roots would be equal to 1025 SL/ton. 4.9.2 Beppemie Eriee pf Bpgar Beere The most recent price of refined sugar imported by Syria was 400 $US/ton (CIF). It should be noted that international market prices of sugar frequently underestimate the true cost of production in the producer countries. In these countries heavy producer subsidies and frequent dumping of surplus production in the international market are common practices. Although the international price of sugar may not represent its true economic value, it still represents the opportunity cost of importing countries such as Syria. Thus, for the purpose of this study, the international market price will be the basis for estimating the economic price of sugar in Syria. Adding an estimated 25 $US/ton in local expenses, this gives a border price of 425 $US/ton or 17,000 SL/ton. .Adding transport, refining, and extraction costs would give an economic producer price of 18,857 SL/ton for refined sugar (Table 4.16). One ton of beet roots gives, on average, 81.9 kg of refined sugar (see General Establishment for Sugar Refining, 1989). Therefore, the economic producer price for beet roots is estimated at 1544 SL/ton. Subtracting transport and harvest costs would give an economic field price for beet roots equal to 1319 SL/ton. 146 4.9.3 Eereign Bxehepge Saving; from Spgar Beere Since sugar production and refining is completely controlled by the government, any increase in the production of refined sugar would lead to an. equal decline in sugar imports. Thus, the additional production of one ton of refined sugar would result in 425 $US savings in sugar imports. Since one ton of sugar beet roots produces an average of 81.9 kg of refined sugar, the foreign exchange savings for each additional ton of sugar beet roots would amount to 34.81 $US (i.e., 425 $US/ton * 0.0819). 4.9.4 Ne; Iaxee on Sugar Beete Taxes on sugar beet roots are estimated by subtracting the official producer price from the economic producer price (or refinery price), as follows (see Table 4.16): Net taxes on sugar beets - Economic Producer Price ~ Official Price - 1544 ~ 1250 - 294 SL/ton. 4.10 Wises 4.10.1 W In the 1989/1990 season, potato prices were not fixed by the government. Therefore, the proper price to use in the estimation of financial field prices would be the expected wholesale potato price shortly after harvest time, expected to be around 6000 SL/ton. If transport, handling, and harvest costs are subtracted, this would give a financial field price of 4860 SL/ton (Table 4.17). 147 Table 4.17: FINANCIAL AND ECONOMIC PRICES OF POTATOES Syria, March 1990. FINANCIAL ECONOMIC BORDER PRICE (FOB $US/ton): 200 Local Expenses (101): ~20 PORT GATE PRICE ($US/ton): 180 PORT GATE PRICE (SL/ton): 7200 Transport Costs (SL/ton): ~100 EXPECTED WHOLESALE PRICE (SL/ton): 6000 7100 Transport Costs (SL/ton): ~100 ~100 FARM GATE PRICE (SL/ton): 5900 7000 Harvesting, and Handling Costs (101): ~590 ~700 Bags (SL/ton): ~450 ~450 FIELD PRICE (SL/ton): 4860 5850 Sources: Based (n1 discussions with officials the Directorate. 148 4.10.2 Bconomie Eriee pf Eotatoee Potato exports have increased during the past few years primarily due to the rapid devaluation of the Syrian Lira. The expected potato export price (FOB) for 1990 is 200 $US/ton. Subtracting 101 in local expenses would give a port price of 180 $US/ton, or 7200 SL/ton. If transport, handling, and harvest costs are subtracted, this would give an economic field price of 5850 SL/kg (Table 4.17). 4.10.3 generipurien pf Eerarpee £2 rhe Governmentfs Eoreign han ene ud et In 1989, the potato marketing and export functions were completely transferred from the public to the private sector in an attempt to promote potato exports. Thus, any increase in potato production or exports will have no direct effects on the government's budget. This applies to the foreign exchange budget as well as the general budget. 4.11 WW Table 4.18 provides a summary of financial and economic prices of the fertilizers and crops covered in this study. Net taxes (or subsidies) and foreign exchange earnings (or expenditures) are also presented in the same table. A comparison between the financial and economic prices indicates some significant distortions in domestic prices. This is particularly true for fertilizers, which are highly subsidized. In contrast, domestic crop prices are generally below their international market equivalent. As shown in Table 4.18, all crops are implicitly taxed, except for barley in good years and potatoes. Table 4.18: SUMMARY OF FINANCIAL AND ECONOMIC PRICES, 149 NET TAXES, AND FOREIGN EXCHANGE EARNINGS OF FERTILIZERS AND MAIN CROPS Syria, 1989/1990. FINANCIAL ECONOMIC NET FOREIGN PRICES PRICES TAXES1 EXCHANGE (SL/kg) (SL/kg) (SL/kg) EARNINGSZ ($US/ton) e ers: ~ N: 12.10 16.81 ~5.20 ~339.0 ~ 1505: 12.80 23.96 ~11.70 ~496.0 Crops: ~ HYV Wheat: Good: 7.07 7.65 0.37 89.6 Normal: 7.65 7.81 0.37 89.6 Dry: 8.63 8.19 0.28 67.2 V. Dry: 9.23 8.40 0.28 67.2 ~ LYV Wheat: Good: 7.75 7.97 0.19 69.9 Normal: 8.45 8.13 0.19 69.9 Dry: 9.54 8.51 0.12 46.6 V. Dry: 10.35 8.72 0.12 46.6 ~ Barley: Good: 4.69 3.82 ~0.55 60.5 Normal: 5.53 6.31 0.54 70.4 Dry: 6.95 6.97 0.40 52.8 V. Dry: 7.82 7.05 0.34 44.0 ~ Raw Cotton: 13.53 16.21 2.68 543.5 ~ Corn: 6.51 6.34 0.26 93.5 ~ Sugar Beet Roots: 1.03 1.32 0.29 34.8 ~ Potatoes: 4.86 5.85 0.00 0.0 1/ Taxes minus subsidies. 2/ A negative sign indicates foreign exchange expenditures. AN ECONOMIC ANALYSIS OF FERTILIZER ALLOCATION AND IMPORT POLICIES IN SYRIA Volume II By Maurice Emile Saade A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Economics 1991 CHAPTER 5 FERTILIZER RECOMMENDATIONS UNDER NO CONSTRAINTS ON FERTILIZER SUPPLIES This chapter presents the main findings related to fertilizer recommendations for the main crops in Syria. The principal assumption underlying all the results in this chapter is that farmers have unlimited access to fertilizers, given the prevailing prices. In other words, based on the optimum fertilizer rates estimated in this chapter, ”ideal“ recommendations are proposed. These optimum rates are defined as those maximizing net returns to fertilizer use. They are computed by equating marginal costs with marginal revenues calculated based on economic prices. These recommendations will constitute the maximum rates, which would need to be adjusted downward depending on the amounts of fertilizer supplies available. The issue of fertilizer recommendations under limited supplies will be discussed in the next chapter. The chapter begins with a summary of the estimated production functions. Based on these functions economically optimum fertilizer rates are estimated. These rates will constitute the basis for proposing any adjustments in the current fertilizer recommendations for the main crops in Syria. Next, the economic feasibility of the proposed rates is discussed. The last section of this chapter presents comparisons between the proposed fertilizer rates and the current 150 151 recommendations. This comparison is done in terms of impact on national and farm incomes, aggregate crop output, and the government foreign exchange and general budgets. 5.1 a ma d c o u t n Production functions for the main crops were estimated based on pooled analysis of available data from fertilizer trials undertaken in Syria since the 1960's. These crops are: (l) irrigated wheat; (2) rainfed high-yielding wheat varieties (HYV); (3) rainfed low~ yielding or local wheat varieties (LYV); (4) rainfed barley; (5) irrigated cotton; (6) irrigated fall-planted sugar beets; (7) irrigated summer-planted sugar beets; (8) irrigated fall-planted potatoes; (9) irrigated spring and summer-planted potatoes; and (10) irrigated corn. The estimated coefficients of the production functions and some of their statistical characteristics are presented in Table 5.1. A detailed discussion of the step-by-step estimations and the data sources were presented in an earlier report (Saade, E1~Hajj, and Meda, forthcoming). The report also includes a discussion of the assumptions underlying each estimated function, alternative formulations, and evaluations of their statistical performance. 5.1.1 W The acceptable range for the values of the coefficient of determination (adjusted R2) is very difficult to determine based on previous studies. If the analysis is based on data from a single fertilizer experiment (i.e., with little variability in soil and 152 .noo%\aa an one Amy oao>o~ HHoucnon Hocomoom nouoa - .Ao£\mcouv moououoe coo mnoon nowsm now unooxo .or\wx On one apnoea Han uo£\wx on one oouon nonnanunow HH< - .mconuoovo one an oHonEhm oru mo oCOHUHCnuoo oca now n noueono on nowom - .oanoonadoo no: I <2 - .ucooanswnu no: I mz ”no." I s “an I «a ”an I «as ”o~o>oH oocoonuncwnm - .muconoHuuooo onu mo mnonno onoocouo one one ouoxoonn an oonswnm - «as own «man «mnn a so.~ne 36.2nm oe.oeoH eo.~oen nonnm .eum one.o mee.o onm.o soo.o 42 .fie< <2 mznmooo.ovm~ooo.o <2 <2 222 m2n6~o.ov memo.o <2 manAHo.ov snoo.o. mznmno.oc nano.o 22 Renewo.ov moeo.o. mznenn.ov ooen.o. ««2~neo.ov moon.o. «Rnoeo.ov nnoH.o- «a «sAoqo.ov NHHH.o- «Amma.ov mmu~.o- «ssAono.ov ammo.o- «ssAoao.ov Nomo.o- «z mznmoo.ovn~noo.o- mznmuo.ov -~o.o mzneoo.ov oeoo.o <2 22 «eenmoo.ov semo.o «snnno.oc mueo.o «Reneoo.oc nwuo.o <2 22 «samom.~v meoo.o <2 «Aoom.cv «can.» sssAamm.ev oeuo.na m mznemn.mc moo~.o <2 mznmme.~v nnmm.~ seeneen.av ennm.en 2 «eennoo.oe enmo.o- <2 «eennoo.ov «ano.o. <2 «2 «Rennao.nv mmqa.e~ «Rename.ov «Neo.n seeneoo.nc oon~.mn <2 2 «esnm.nonvmn.mn~m- «44am.en~v nn.nmm- «+«n~.em~v ~.noa~- +«2A2.~uaa~ CO‘QH 1 U'‘ 4 2 F‘F‘F‘P‘ N 2. 2. P‘P‘hih‘ P‘F‘P‘P1 z 50 1 7~5.0 l~6.3 9~5.8 6~4.8 2-3.5 7-5.0 5-4.5 2-3.6 .0-3.2 .5-4.4 .3-3.9 .0-3.1 6-7.9 2 7-5.2 1 3~6.8 2 3~6.8 2 7-8.2 2 6~7.8 2 F‘P‘NJP‘ P‘P‘P‘P‘ h‘h‘h‘h‘ z 35 1 6~3.4 0~4.2 5-3.9 2-3.2 l~2.4 6~3.4 5-3.l 2-2.4 0~2.1 4~3.0 3~2.7 0~2.1 6~5.4 l 7~3.5 0 .2-4.6 l .2-4.6 l .7~5.5 1 .5-5.3 1. 0000 00°C 0000 MN I 050‘ 0‘0 abeb O\¢~\J\J O\¢‘\J\I : 701 on oooo 1/ See Table 3.1 for the definitions of rainfall scenarios. 170 5.4 v on As mentioned earlier, the proposed fertilizer recommendations were based on the estimated economic optimum rates, i.e., the rates computed based on economic prices. The underlying assumption for this approach is the objective of achieving an economically efficient allocation of resources, from the viewpoint of the economy as a whole. However, the final decision on fertilizer allocation is a political decision that may be subject to influences from many potential interest groups, particularly farmers. Therefore, if the above objective is modified to maximize farmers' income, then fertilizer recommendations would have to be based on the financial optima. Thus, the objective of this section is to compare the economic impact of these financial optima with the proposed rates based on economic optimum rates. This comparison would provide estimates of the potential costs to the economy, should the government decide that the objective of fertilizer policy is to maximize farmers' income. As can be noted from Table 5.2, the estimated financial optimum fertilizer rates are, on the average, 5 kg/ha higher than the economic Optima. The only exception is 1505 rates on 'barley, where the differences are more significant, amounting to 20 kg/ha. The financial optima represent, by definition, the rates that would maximize farmers' net returns from fertilizer use. In other words, these rates are based on equating the farmer's marginal costs and marginal revenues. However, this does not necessarily imply that these financial optima ‘would satisfy the minimum rate of return and the risk concerns of farmers. To address these concerns, the financial VCR values associated ‘with the economic and financial optima are compared in Table 5.7. This 171 Table 5.7: FINANCIAL VALUE-COST-RATIOS OF FERTILIZER USE BASED ON ECONOMIC AND FINANCIAL OPTIMUM RATES Syria, 1989/1990. Financial VCR based on Financial VCR based on Economic Optimum Rates Financial Optimum Rates Crop: Good Normal Dry V. Dry Good Normal Dry V. Dry Irrigated Wheat: 5.6 5.4 Rainfed Wheat: HYV Zone 1: 4.9 4.2 4 2 3.8 4.6 3 9 3.8 3.4 HYV Zone 2: 3.9 3.8 3 7 3.3 3.7 3 5 3.4 3.1 LYV Zone 1: 7.6 5.6 4.9 3.7 6.9 4.9 4.1 2.9 LYV Zone 2: 5.2 4.4 3.6 2.4 4.7 3.9 3.0 1.7 LYV Zone 3: 4.6 3.8 2.9 1.6 4.1 3.2 2.2 0.9 Rainfed Barley: Zone 1: 4.9 4 3 4.6 4 2 4.3 3.8 4.1 3.7 Zone 2: 3.5 3 5 3.6 3.2 3.0 3.1 3.2 2.8 Zone 3: 3.1 3 1 3.2 2.8 2.7 2.7 2.8 2.5 Cotton: 5.2 5.0 Corn: 4.3 4.2 Sugar Beets: Fall: 4.2 4.1 Summer: 4.2 4.1 Potatoes: Fall: 5.4 5.3 Summer: 5.2 5.1 1/ See Table 3.1 for the definitions of rainfall scenarios. 172 comparison shows that the VCR values for the financial optimum rates would be slightly lower than those associated with the economic optima. Nevertheless, the financial optimum rates would have VCR values above the 1.5 limit for all crops, except for LYV wheat in Zone 3 with a VCR of 0.9. Therefore, the financial optimum fertilizer rates would represent profitable and relatively safe investments for farmers. The only exception is the rates on LYV wheat in Zone 3, which may be too risky given the possible financial losses in very dry years. 5.4.1 lmpeer pi Economie and Financial thimpm Ratee en Aggregage W Another reason policy-makers may want to base fertilizer recommendations on financial instead of economic optima is the current policy objective of reducing food imports and increasing agricultural exports. In Syria, wheat imports represent the most important food import item, with 1.5 million tons imported in 1989 (Arep_figgjeplrpre, 1990, p. 70). Cotton, on the other hand, represents the major source of agricultural export earnings, with 85 thousand tons of lint cotton exported in 1988/89 (Jaber, 1989, p. 192). Thus, policy makers may favor the higher financial optimum fertilizer rates in the hope of reducing wheat imports and increasing cotton exports. To estimate the impact of the economic and financial optimum rates on aggregate crOp production in Syria, yield estimates are made based on the production functions presented in Table 5.1. These yield estimates are computed by solving the production equation for the calculated economic and financial optimwm N and £505 values. It should be noted that, throughout this analysis, the impact of fertilizer application 173 will be measured in terms of the increase in yield over the unfertilized treatment. This is done by basing the analysis on the difference between the estimated yield and the intercept term in the production functions. The main reason for this approach is to minimize the potential bias in the results caused by what is often referred to as the yield gap. This refers to the commonly observed large differences between yield estimates based on experimental data and actual yields obtained by farmers (see, for example, ICARDA/FRMP, 1988, pp. 143-150). To estimate the impact of fertilizer application on aggregate crop production, the increase in yield due to fertilizer use is multiplied by the total fertilized area planted to each crop. Based on discussions with agronomists from the Soils Directorate, it is estimated that all areas planted to irrigated crops and to HYV wheat are currently fertilized. In contrast, it is estimated the percentage of total LYV wheat area fertilized is approximately 901 in Zone 1, 701 in Zone 2, and 501 in Zone 3. The corresponding figures for barley are 501 in Zone 1, 401 in Zone 2, and 301 in Zone 3. The estimated percentages for barley probably overestimate the current fertilizer use by barley farmers. However, it is assumed that if barley farmers had a greater access to fertilizer, then more farmers would start applying fertilizer on barley. Thus, the above percentages for barley should be viewed as realistic targets rather than figures representing, current fertilization practices. As shown in Table 5.8, the financial optimum fertilizer rates on wheat would result in an additional 28 thousand tons in total wheat output in a normal year, as compared to the lower economic optimum rates. This increase in wheat production would represent approximately 174 Table 5.8: IMPACT OF ECONOMIC AND FINANCIAL OPTIMUM FERTILIZER RATES ON AGGREGATE CROP PRODUCTION Syria, 1989/1990 Aggregate Production Increase1 in a Normal Year ('000 tons) Economic Financial Net Crop: Optima Optima Change Irrigated Wheat: 482 488 6 Rainfed HYV Wheat Zone 1: 237 245 8 Rainfed HYV Wheat Zone 2: 104 108 4 Rainfed LYV Wheat Zone 1: 67 69 2 Rainfed LYV Wheat Zone 2: 93 100 7 Rainfed LYV Wheat Zone 3: l4 l6 2 TOTAL WHEAT 997 1025 28 Rainfed Barley Zone 1: 25 27 2 Rainfed Barley Zone 2: 183 205 22 Rainfed Barley Zone 3: 92 107 15 TOTAL RAINFED BARLEY 300 339 39 Cotton: 284 287 3 Corn: 118 119 1 Fall Sugar beets: 257 258 1 Summer Sugar beets: 228 229 1 Fall Potatoes: 52 52 0 Spring and Summer Potatoes: 39 39 0 1/ Increase over the unfertilized treatment. 175 1.11 increase over the current average wheat output of 2.5 million tons. Similarly, in comparison to the economic optimum fertilizer rates, the financial optima would lead to an increase of 39 thousand tons in aggregate barley production in a normal year. This additional barley output would represent an extra 3.51 in total barley production over the current average level of 1.1 million tons. The results in Table 5.8 also indicate that the difference in the impact of the economic and financial optima on the aggregate output of cotton, corn, sugar beets, and potatoes, is very limited. Cotton production would increase by only 3000 tons (0.61 increase), sugar beets by 2000 tons (0.21 increase), corn by 1000 tons (0.91 increase), and no change in potato production. Therefore, the use of the higher fertilizer rates implied by the financial optima, instead of the economic optima, would increase aggregate crop output, particularly barley and wheat. However, this increase in production would be relatively small, leading to a slight reduction in food imports and an insignificant increase in exports. However, this increase in aggregate output would be accompanied by an increase in fertilizer use and, thus, in fertilizer imports. 5.4.2 WWW 11821211214112.2122 If fertilizer recommendations were based on financial optima instead of economic optima, aggregate N requirements for the crops in this study would increase by 71 and P205 requirements would increase by 181 (see Table 5.9). These additional fertilizer requirements would necessitate an extra 600 million SL worth of fertilizer imports, representing a 121 increase in total fertilizer costs. The value of the 176 Table 5 . 9: AGGREGATE ECONOMIC IMPACT OF ECONOMIC AND FINANCIAL OPTIMUM FERTILIZER RATES Syria, 1989/1990 Economic Financial Net Optima Optima Change Fertilizer Requirements: N (tons): 163,691 175,102 11,411 P205 (tons): 96,977 114,007 17,030 ECONOMIQ ERICES (Million SL) Fertilizer Costs: 5,075 5,675 600 Gross Returns: ~ Good Year: 16,622 17,235 613 ~ Normal Year: 16,255 16,784 529 ~ Dry Year: 15,708 16,112 404 ~ Very Dry Year: 14,771 15,017 246 Net Returns: ~ Good Year: 11,547 11,560 13 ~ Normal Year: 11,180 11,109 ~71 ~ Dry Year: 10,633 10,437 ~196 - Very Dry Year: 9,696 9,342 ~354 W (”111100 SL) Fertilizer Costs: 3,222 3,578 356 Gross Returns: ~ Good Year: 15,399 16,021 622 ~ Normal Year: 14,983 15,474 491 ~ Dry Year: 15,214 15,617 403 ~ Very Dry Year: 14,700 14,951 251 Net Returns: - Good Year: 12,177 12,443 266 ~ Normal Year: 11,761 11,896 135 - Dry Year: 11,992 12,039 47 - Very Dry Year: 11,478 11,373 ~105 1/ See Table 3.1 for the definitions of rainfall scenarios. 177 additional crop output would be equal to 529 million SL in a normal year. Therefore, the use of financial instead of economic optimum fertilizer rates would result in a net loss in national income, or “dead-weight loss", amounting to 71 million SL in normal years and up to 354 million SL in very dry years. 5.4.3 lmemmflnmmmmmmm mm The results in Table 5.9 also show the impact of the economic and financial optimum fertilizer rates on farm income. The additional fertilizer use implied by the financial optima would cost farmers an additional 356 million SL. These additional expenditures would result in an increase in the value of crop production amounting to 491 million SL in normal years. Therefore, in a normal year, farm income would increase by 135 million SL, with a range of 47 million SL in dry years to 266 million SL in good years. In the event of a very dry year, however, the net returns from using the financial Optimum fertilizer rates would be 105 million SL below those obtained by applying the lower economic optimum rates. Thus, although the financial optima would increase farm income, the variability in farm income would also increase. This is shown in Table 5.9, where financial net returns from fertilizer use based on the financial optimum rates would vary from 11,373 million SL in very dry years to 12,443 million SL in good years, i.e., a range of 1070 million 31.. If we divide this range by net returns in a normal year (11,896 million SL), this gives a ratio of 0.09. This ratio can be used as a proxy for measuring variability in net returns to fertilizer use. 178 In comparison, financial net returns based on the economic optima would vary from 11,478 million SL in very dry years to 12,177 million SL in good years, i.e., a range of 699 million SL. Dividing this range by net returns in normal years (11,761 million SL) would give a ratio of 0.06. Thus, the variability in net returns to fertilizer use would increase by about 501 (from 62 to 91) if the financial optima were used as a basis for fertilizer recommendations instead of the economic optima. 5.14.4 m a c nom nd nancial O timum xc a a in The higher fertilizer use associated with the financial optimum rates implies 12 million $US more in fertilizer import expenditures than the economic optimum rates (see Table 5.10). These additional expenditures would increase the value of net crop exports by 3 to 8 million $US. Therefore, if financial optima are used instead of the proposed economic optima, net foreign exchange earnings would decline by an amount ranging from 4 million $US in good years to 9 million $US in very dry years. 5.h.5 Impgcc cf Eccnomic cmg Efimgncigl Optimum Eczciuzc: Ectec on ov e ud e Finally, the increase in fertilizer use associated with the appIication of the financial optimum rates would result in an increase in government expenditures on subsidies. As shown in Table 5.11, these expenditures would be 258 million 81. higher than those associated with the use of economic optima. The corresponding increase in revenues from indirect taxes on crops would amount to 38 million SI. in a normal year. 179 Table 5.10: IMPACT OF ECONOMIC AND FINANCIAL OPTIMUM FERTILIZER RATES ON FOREIGN EXCHANGE EARNINGS Syria, 1989/1990 Economic Financial Net Optima Optima Change (Million $US) Import Value of Fertilizer: 104 116 12 Value of Increased Crop Exports and Reduced Importslz - Good Year: 301 309 8 - Normal Year: 289 296 7 - Dry Year: 25h 258 4 - Very Dry Year: 244 247 3 Net Increase1 in Foreign Exchange Earnings: - Good Year: 197 193 -h - Normal Year: 186 180 -6 - Dry Year: 150 142 -8 - Very Dry Year: 140 131 -9 1/ Relative to no fertilizer use. 2/ See Table 3.1 for the definitions of rainfall scenarios. 180 Table 5.11: IMPACT OF ECONOMIC AND FINANCIAL OPTIMUM FERTILIZER RATEs ON THE GOVERNMENT BUDGET Syria, 1939/1990 Economic Financial Net Optima Optima Change (Million SL) Expenditures on Fertilizer Subsidies: 1,986 2,244 258 Increase1 in Revenues from Indirect Taxes on Crops: - Good Year: 1,113 1,111 -2 - Normal Year: 1,431 1,469 38 - Dry Year: 1,265 1,290 25 - Very Dry Year: 1.212 1,231 19 Net Increase1 in Government Expenditures: - Good Year: 873 1,133 260 - Normal Year: 555 775 220 - Dry Year: 721 954 233 - Very Dry Year: 774 1,013 239 1/ Increase due to fertilizer use. 2/ See Table 3.1 for the definitions of rainfall scenarios. 180 Table 5.11: IMPACT OF ECONOMIC AND FINANCIAL OPTIMUM FERTILIZER RATES ON THE GOVERNMENT BUDGET Syria, 1989/1990 Economic Financial Net Optima Optima Change (Million SL) Expenditures on Fertilizer Subsidies: 1,986 2,244 258 Increase1 in Revenues from Indirect Taxes on Crops: - Good Year: 1,113 1,111 -2 - Normal Year: 1,431 1,469 38 - Dry Year: 1,265 1,290 25 - Very Dry Year: 1,212 1,231 19 Net Increase1 in Government Expenditures: - Good Year: 873 1,133 260 - Normal Year: 555 775 220 - Dry Year: 721 954 233 - Very Dry Year: 774 1,013 239 1/ Increase due to fertilizer use. 2/ See Table 3.1 for the definitions of rainfall scenarios. 181 Therefore, if fertilizer recommendations were to be based on financial instead of economic optima, government expenditures would increase by an amount ranging from 220 million SL to 260 million SL. 5.4.6 Financial vs Ecomcmic Qpcima; Summary In summary, if policy makers were to decide to base fertilizer recommendations on financial instead of economic optimum rates, the main effect of such a decision would be a net income transfer from the government budget (the foreign exchange budget in particular) to farmers. In a normal year, net government expenditures would increase by 220 million SL (see Table 5.11) resulting in 135 million SL increase in farm income (i.e, increase in financial net returns in a normal year; see Table 5.9). The 85 million SL difference between these two figures would be partially accounted for by the 71 million 81. in dead-weight loss (i.e., decline in economic net returns in a normal year; see Table 5.9). The remaining 14 million SL would represent income transfers from the government to other sectors of the economy, particularly fertilizer and crop traders in the parallel market. Thus, if income redistribution in favor of farmers is considered an important policy objective, then other income distribution strategies based on direct income transfers to farmers might be more economically efficient" Moreover, the 12 million $US in. additional fertilizer imports, because of the higher financial optimum fertilizer rates, would offset any gains from the slightly lower barley and wheat imports and the minor increase in cotton exports. The net effect would be a decline in foreign exchange earnings by 6 million $US in a normal year (see Table 5.10). Therefore, basing fertilizer recommendations on the 182 economic instead of the financial optima would be the most economically rational approach. This is especially true under the current situation of chronic fertilizer shortages, limited foreign exchange resources, and government budget deficits. 5.5 u v 03 ecommendat o In this section, the impact of the proposed fertilizer recommendations, based on economic optima, is compared to that of the current rates recommended by the Soils Directorate. This is done in order to estimate the impact of the shift from the current to the proposed recommendations on yields and aggregate crop production, national and farm incomes, and the government's fOreign exchange and general budgets. 5.5.1 Impac; cf szzcnc and Ercposed Eczcilizc: B§§§§ cn Aggzcggcc W The current and proposed fertilizer rates on the crops covered by this study were presented in Table 5.2. Fertilizer rates that would maximize yield, or biological maximum rates, are also included in Table 5.2. A closer look at the current recommended rates on wheat indicates that all I505 rates and most N rates exceed the corresponding biological maximum rates. In other words, given the quadratic formulation of the production functions, the application of the current rates on wheat may result in a yield decline (Phase III of the production function). This may be an accurate representation of actual yield responses to excessive levels of nitrogen. However, in the case of phosphorus, 183 rates that are higher than the maximum level would rarely cause a decline in yield. The most commonly observed yield response to excess £505 is that of a plateau maximum, with yield declines only because of extremely high F50, applications (Lanzer, Paris, and Williams, 1987, p. 2). Such a response would have been more appropriately modeled using other functional forms, such as the Mitscherlich function or the linear response and plateau (LRP) function proposed by Gate and Nelson (1971). Therefore, in assessing the impact of the current fertilizer rates on wheat output, it is more realistic to ignore the possibility of yield depression and to assume that a yield plateau would be obtained whenever these rates exceed the biological maximum rates. Based on this assumption, the increase in aggregate wheat output as a result of applying the current fertilizer rates would amount to 1045 thousand tons in a normal year (Table 5.12). In comparison, the proposed fertilizer rates on wheat would result in 997 thousand tons of additional wheat output. Thus, compared to the current rates, the proposed fertilizer rates on wheat would lead to a decline in aggregate wheat output in a normal year by about 48 thousand tons. This is equivalent to an average 31 increase in current wheat import levels. The results in Table 5.12 also indicate that increasing the current fertilizer rates on barley would result in a net increase of 57 thousand tons. This is equivalent to an increase of about 52 in total barley output in a normal year. Approximately two-thirds of this additional barley output would originate in Zone 3, where the proposed rates are almost twice as large as the current ones. In the case of cotton, the proposed fertilizer rates imply an increase in the current N rate and a reduction in the rate of P505. The 184 Table 5.12: IMPACT OF CURRENT AND PROPOSED FERTILIZER RATES ON AGGREGATE CROP PRODUCTION Syria, 1989/1990 Aggregate Production Increase1 in a Normal Year ('000 tons) Proposed Current Net Crop: Rates Rates Change Irrigated Wheat: 482 494 ~12 Rainfed HYV Wheat Zone 1: 237 249 -12 Rainfed HYV Wheat Zone 2: 104 113 -9 Rainfed LYV Wheat Zone 1: 67 70 -3 Rainfed LYV Wheat Zone 2: 93 103 ~10 Rainfed LYV Wheat Zone 3: 14 16 -2 TOTAL WHEAT 997 1045 ~48 Rainfed Barley Zone 1: 25 20 5 Rainfed Barley Zone 2: 183 167 16 Rainfed Barley Zone 3: 92 56 36 TOTAL RAINFED BARLEY 300 243 57 Cotton: 284 278 6 Corn: 118 116 2 Fall Sugar beets: 257 253 4 Summer Sugar beets: 228 226 2 Fall Potatoes: 52 51 1 Spring and Summer Potatoes: 39 36 3 1/ Increase over the unfertilized treatment. 185 net impact of these adjustments in cotton rates would be an increase of 6,000 tons in total cottonx output, which represents about 1.21 of aggregate cotton production in Syria. Similarly, increasing the current N rate for corn would increase production by 2000 tons, or about 1.8% increase in total output. As for sugar beets, an increase in the current N rates would result in 6,000 tons of additional output, representing about 0.62 increase in total production. Finally, increasing the current N rates for potatoes would increase output by 4,000 tons, an increase of about 11 in total production. 5.5.2 Impact ct gmlrcmc and Echoscc Eaceg cm Ngcicmcl Incomc The above results clearly indicate that the shift from the current to the proposed fertilizer rates would have a somewhat limited impact on aggregate crop production. However, the most significant impact would be in the substantial reduction in aggregate fertilizer requirements. As shown in Table 5.13, most of this reduction would come from the drastic decline in fertilizer use on wheat. Total N use on wheat would decline by 30 thousand tons, while F505 use would decline by 42 thousand tons. A further decline of 5.2 thousand tons in P20, use on cotton would bring the total reduction in P505 use to 47 thousand tons. These drastic declines are offset somewhat by the increase in N and Paos'use on barley and N use on cotton, besides the minor increases in N use on corn, sugar beets, and potatoes. Therefore, the shift from the current to the proposed fertilizer rates would result in a net reduction in total fertilizer use by approximately 17 thousand tons of N and 45 thousand tons of'laos. This would represent a 101 decline in total N requirements for the crops 186 Table 5.13: AGGREGATE FERTILIZER REQUIREMENTS BASED ON CURRENT AND PROPOSED RATES Syria, 1989/1990 Aggregate Fertilizer Requirements (tons) Proposed Rates Current Rates Net Change Crop: N P205 N P205 N P205 Irrigated Wheat: 36156 17408 40173 26782 -4017 ~9374 Rainfed Wheat: HYV Zone 1: 23148 11574 28935 23148 -5787 -11574 HYV Zone 2: 10832 6319 14443 10832 -3611 -4513 LYV Zone 1: 4451 3895 8903 6677 -4452 -2782 LYV Zone 2: 7957 6366 19098 19098 -lll4l -12732 LYV Zone 3: 1480 1110 2220 2220 -740 -1110 TOTAL WHEAT: 84024 46672 113772 88757 -29748 -42085 Rainfed Barley: Zone 1: 1604 963 1069 856 535 107 Zone 2: 12937 10349 10349 10349 2588 0 Zone 3: 6962 6092 3481 3481 3481 2611 TOTAL BARLEY: 21503 17404 14899 14686 6604 2718 Cotton: 39150 20880 34800 26100 4350 ~5220 Corn: 9050 5569 8354 5569 696 0 Sugar Beets: Fall: 3305 1844 3074 1844 231 0 Summer: 2780 1756 2633 1756 147 0 Potatoes: Fall: 2210 1560 1950 1560 260 0 Summer: 1669 1292 1292 1292 377 0 GRAND TOTAL 163691 96977 180775 141564 17084 44587 187 Table 5.14: AGGREGATE ECONOMIC IMPACT OF PROPOSED AND CURRENT FERTILIZER RATES Syria, 1939/1990 Proposed Current Net Rates Rates Change ECONOMIC PRICES (Million SL) Fertilizer Costs: 5,075 6,431 -1356 Gross Returns: - Good Year: 16,622 16,917 -295 ~ Normal Year: 16,255 16,139 116 - Dry Year: 15,708 15,420 288 - Very Dry Year: 14,771 14,327 444 Net Returns: - Good Year: 11,547 10,487 1060 - Normal Year: 11,180 9,709 1471 - Dry Year: 10,633 8,990 1643 - Very Dry Year: 9,696 7,896 1800 W (Million SL) Fertilizer Costs: 3,222 3,999 -777 Gross Returns: - Good Year: 15,399 15,610 -211 - Normal Year: 14,983 14,931 52 — Dry Year: 15,214 14,953 261 - Very Dry Year: 14,700 14,226 474 Net Returns: - Good Year: 12,177 11,611 566 - Normal Year: 11,761 10,932 829 - Dry Year: 11,992 10,953 1039 - Very Dry Year: 11,478 10,226 1252 1/ See Table 3.1 for the definitions of rainfall scenarios. 188 included in this study and a 46% reduction in P505 requirements. These drastic reductions would decrease aggregate fertilizer costs by 1356 million SL, which would constitute a potential 27% savings in total fertilizer costs (Table 5.14). These savings in fertilizer costs are the net result of the following: (1) reduced N and.I505 use on wheat by 1509 million SL; reduced P205 use on cotton by 125 million SL; (2) increased N and P50, use on barley by 176 million SL; and (3) increased N use on cotton, corn, sugar beet, and potatoes by 102 million SL. As discussed earlier, the shift from the current to the proposed recommendations would result in minor reductions in wheat output but would slightly increase the production of other crops, particularly barley. The net effect of these changes in production would be an increase in the economic value of aggregate output by 116 million SL in normal years (Table 5.14). However, in good years, the shift to the proposed rates would lead to a 295 million SL decline in the economic value of aggregate output. Thus, by adding the savings in fertilizer costs to the changes in the value of output, the net impact of the shift to the proposed rates on the GDP would be equal to 1.5 billion SL in a normal year. This is equivalent to an average increase of 2.62 in the agricultural GDP, estimated at 56.9 billion SL in 1988 (Al-Akhrass, 1990, p. 6). 5.5.3 IaEa2t_2f;Q2rrEat_aDd_2r2E2EEQ_REEE§_ED_EEEE_IDEEEE The results in Table 5.14 also show the impact of the shift from the current to the proposed fertilizer rates on aggregate farm income (i.e., costs and returns based on financial prices). As noted earlier, 189 the main impact of this shift would be the significant reduction in aggregate fertilizer use. This would translate into a 777 million SL decline in farmers' fertilizer costs. In a good year, this shift to the proposed rates would reduce the value of aggregate farm output by 211 million SL. However, in normal and dry years, the value of aggregate farm output would increase by an amount ranging from 52 million SL in normal years to 474 million $1. in very dry years. Thus, in a normal year, the net impact of the shift from the current to the proposed fertilizer rates would be an increase of 829 million SL in aggregate farm income. In addition to increasing farm income the shift to the proposed rates also would reduce the variability in farm income. As shown in Table 5.14, financial net returns from fertilizer use based on the current rates would vary from 10,226 million SL in very dry years to 11,611 million SL in good years, i.e., a range of 1385 million SL. If this range is divided by net returns in a normal year (10,932 million SL), this gives a ratio of 0.13. As discussed earlier, the corresponding ratio for the financial net returns based on the proposed rates is equal to 0.06. Thus, the variability in financial net returns to fertilizer use would decline by more than half (from 131 to 62) as a result of shifting fertilizer recommendations from the current to the proposed rates. 5.5.4 Cu rent and o ose ates 0 As mentioned earlier, the main impact of the shift from the current to the proposed rates would be the substantial decline in 190 Table 5.15: IMPACT OF PROPOSED AND CURRENT FERTILIZER RATES ON FOREIGN EXCHANGE EARNINGS Syria, 1989/1990 Proposed Current Net Rates Rates Change (Million $US) Import Value of Fertilizer: 104 131 27 Value of Increased Crop Exports and Reduced Importslz — Good Year: 301 300 -l - Normal Year: 289 286 -3 - Dry Year: 254 249 -5 — Very Dry Year: 244 239 -5 Net Increase1 in Foreign Exchange Earnings: - Good Year: 197 169 -28 - Normal Year: 185 154 -31 - Dry Year: 150 118 -32 - Very Dry Year: 140 107 -33 1/ Relative to no fertilizer use. 2/' See Table 3.1 for the definitions of rainfall scenarios. 191 aggregate fertilizer use. This decline would lead to 27 million $US in reduced fertilizer imports (see Table 5.15). Also, the shift to the proposed rates would result in a slight increase in the value of net crop exports (increased exports and reduced imports) ranging from one to five million $US. Therefore, the net impact of the shift from the current to the proposed rates would be an increase in foreign exchange earnings ranging from 28 to 33 million $US. This would represent an increase of about 161 in total foreign exchange reserves, which amounted to 191 million $US at the end of 1988 (Imccxpmcicmgl__£1mmmc;ml Spacigticc, April 1991, p. 510). 5.5.5 Impact of Current and O 05 d tes on the Govcznmcnc was: The drastic reduction in fertilizer use as a result of the shift to the proposed rates would lead to an equally drastic reduction in fertilizer subsidies. As shown in Table 5.16, these subsidies would decline from 2,596 million SL under the current rates to 1,986 million SL under the proposed rates, a 24! decline in fertilizer subsidies. Also, government revenues from the indirect taxes on crops would slightly increase because of the shift to the proposed rates, except in good years. Therefore, the net impact of the shift to the proposed rates would be a decline in government expenditures ranging from 636 million SL in very dry years to 642 million SL in normal years. In good years, this decline would be smaller (559 million SL) since indirect taxes on crops 192 Table 5.16: IMPACT OF PROPOSED AND CURRENT FERTILIZER. RATES ON THE GOVERNMENT BUDGET Syria, 1939/1990. Proposed Current Net 1 Rates Rates Change Change (Million SL) Expenditures on Fertilizer Subsidies: 1,986 2,596 -610 -23.5 Increase1 in Revenues from Indirect Taxes on Crops: - Good Year: 1,113 1,164 -51 -4.4 - Normal Year: 1,431 1,399 32 2.3 - Dry Year: 1,265 1,237 28 2.3 - Very Dry Year: 1,212 1,186 26 2.2 Net Increase1 in Government Expenditures: - Good Year: 873 1,432 -559 ~39.0 - Normal Year: 555 1,197 -642 -53.6 - Dry Year: 721 1,360 -639 -47.0 - Very Dry Year: 774 1,410 -636 -45.1 1/ Increase due to fertilizer use. 2/ See Table 3.1 for the definitions of rainfall scenarios. —-——.u—-—.—— 193 are usually lower than in normal and drier yearsl. These reductions in expenditures are equivalent to a decline of 391 to 542 in net government spendings associated with fertilizer use on the crops covered by this study. Also, these declines in expenditures would represent a reduction of approximately 1.71 in total government expenditures (excluding capital and military spendings), estimated at 35.7 billion SL in 1989 (Bouzo, 1990, p. 48). 5.5.6 gmggcmc v§ Proposec Eczcilize: Becommendationc; Smmmary In summary, the above analysis clearly indicates that the shift from the current to the proposed fertilizer recommendations would greatly enhance the economic efficiency of fertilizer use in Syria. Such a shift would increase the agricultural GDP by an average of 1.5 billion SL per year, or a 2.6! increase. This shift also would increase farmers' income by an average of 829 million SL, in addition to significantly reducing the variability in farm income. The shift to the proposed fertilizer rates would substantially reduce aggregate fertilizer requirements, especially £50, requirements, which would decline by 461. This would imply substantial declines in fertilizer imports and, thus, savings in foreign exchange amounting to 31 million $US, or a 161 increase in total foreign exchange reserves. Also, net government expenditures associated with fertilizer use would decline by an amount ranging from 559 to 642 million SL, which is equivalent to a 391 to 54! decline. 1 In good years the export price is used as the basis for computing indirect taxes on barley. Since the official barley price is slightly higher than the export price, barley' would 'be implicitly subsidized in good years (refer to the discussion in chapter 4). 194 The only potential disadvantage of the shift from the current to the proposed fertilizer rates is the decline in total wheat production. However, this decline would be somewhat limited, amounting to an average 22 decrease in total wheat output. Also, this reduction in wheat output would be offset by expected increases in aggregate output of other crops, particularly barley. 5.6 chpcc; summary This chapter presented the main findings related to ”ideal” fertilizer rates, i.e., the rates to be recommended if there were no constraints on fertilizer supplies. These rates were computed based on the estimated production functions and based on the true economic value of fertilizers and crops. The results have clearly shown that these proposed fertilizer recommendations would constitute profitable and reasonably safe investments for farmers and for the economy as a whole. The analysis also examined. the possibility of increasing the proposed fertilizer rates so as to maximize farmers' income from fertilizer use (i.e., optimum rates computed based on financial prices). Such an increase would result in limited increases in aggregate crop production, but the main effect would be a net income transfer from the government budget to farmers amounting to 136 million SL. This would cost the government an average of 220 million SL and would result in a net loss (dead-weight loss) of 71 million SL to the economy as a whole. A comparison between the proposed and current fertilizer rates has indicated the need for some major readjustments in the current recommendations. Current fertilizer rates on wheat need to be substantially reduced, while barley rates need to be increased 195 significantly. The differences between the proposed and current fertilizer recommendations for the other crops are close. The results indicate that the N rate on cotton should be increased, while the rate of’IfiO, should be reduced. As for corn, sugar beets, and potatoes, N rates should be increased, whereas the current P205 rates are not modified since the production functions were estimated in terms of response to N only. The shift from the current to the proposed fertilizer recommendations would greatly enhance the economic efficiency of fertilizer use in Syria. Such a shift would increase the agricultural GDP by an average of 1.5 billion SL and farmers' income by 829 million SL. This is mainly due to substantial reductions in aggregate fertilizer requirements, which would reduce fertilizer imports by 27 million $US and fertilizer subsidies by 610 million SL. The only disadvantage of such a shift would be the somewhat limited decline (about 21) in total wheat production. However, this reduction in wheat output would be offset by expected increases in aggregate output of other crops, particularly barley. It should be noted that all the above estimates of the impact of the proposed fertilizer recommendations are hypothetical. In other words, these estimates would be relevant only under the ideal situation of unlimited fertilizer supplies. Despite this hypothetical nature of the analysis in this chapter, the results and conclusions could have some important practical implications. Although fertilizer shortages have become a normal occurrence over the past few years, policy makers and government planners continue to treat fertilizer shortages as temporary aberrations. Thus, most 196 planning aspects of fertilizer policy are still based on the hypothetical situation, or 'base scenario“, of unlimited fertilizer supplies. Depending on the magnitude of fertilizer shortages at the beginning of the season, government planners would then modify this base scenario by reducing some or all the recommended rates by a given percentage. However, if the recommended fertilizer rates in the base scenario are inaccurate, then fertilizer allocations computed by taking a given percentage of the recommended rates are also likely to be inaccurate and, thus, economically inefficient. Also, the results and conclusions based on the above mentioned base scenario would provide policy makers and government planners with more accurate estimates of the potential impact of fertilizer use under ideal circumstances. These estimates would be particularly useful in providing more realistic targets in medium- and long-temm planning of future fertilizer use in Syria. CHAPTER 6 FERTILIZER ALLOCATION STRATEGIES UNDER LIMITED FERTILIZER SUPPLIES The analysis in the previous chapter was based on the assumption that there exist sufficient fertilizer supplies to satisfy' all the optimum fertilizer requirements calculated based on the rates that would maximize net economic returns. However, this is not the case since the actual quantities of fertilizer distributed to farmers have often been below the planned fertilizer requirements. This is due to technical and managerial problems facing domestic fertilizer production, and import restrictions because of increasing foreign exchange constraints facing Syria since the mid-1980's. Thus, the estimated differences in economic impact between the current and proposed fertilizer recommendations are hypothetical. The calculated increases in agricultural GDP, farm income, and foreign exchange earnings should. be viewed as maximum potential gains. They would be attainable only if the government could ensure that all the estimated fertilizer requirements would be available to farmers. The main objective of this chapter is to analyze the impact of the constraints on fertilizer supplies and to compare alternative fertilizer allocation strategies under varying levels of available supplies. Given the constrained optimization nature of the problem, most of the analyses 197 198 in this chapter will be based on the fertilizer allocation linear programming (LP) model discussed in chapter 3. This chapter starts with a brief overview of the fertilizer supply situation in Syria during the past few years. The next section addresses the issue of the accuracy of the LP model. This is done by comparing the results of the model under unlimited supplies with the unconstrained optimization solutions presented in chapter 5. This is followed by an analysis of the impact of limited fertilizer supplies on aggregate crop production, agricultural GDP, farm income, foreign exchange earnings, and government expenditures. The next section addresses the issue of fertilizer pricing and its potential role in strategies aimed at reducing the constraints on fertilizer supplies. This is followed by a comparison of alternative fertilizer allocation strategies for the winter season, under existing levels of limited fertilizer supplies. The chapter concludes with a summary of the main findings. 6.1 W112; As mentioned above, the amount of fertilizer actually distributed to farmers has often been below planned requirements. These gaps between planned and actual fertilizer use have become more accentuated in recent years. As discussed in chapter 2, between 1984/85 and 1988/1989 the percentage of planned fertilizer requirements actually used declined steadily from 911 to 551 for N, and from 891 to 532 for P205 (refer to Table 2.9). It should be noted that this decline occurred in spite of a gradual increase in actual fertilizer use. Therefore, the increasing difference between planned and actual use is 199 partly due to the rapid increase in planned requirements, especially in 1987/88 and 1988/89. As discussed earlier, the reason for the large increases in planned requirements since 1987/88 is the change in the rules used by the Soils Directorate (SD) in estimating fertilizer requirements (refer to the discussion in chapter 2). Planned requirements based on this new rule assume that all areas planted to field crops are fertilized. However, this would overestimate total requirements given that significant areas of field crops, particularly barley, are not currently fertilized. In this study, more realistic assumptions were made about the percentage of barley and wheat areas actually fertilized, based on which optimum fertilizer requirements were estimated (see chapter 5). Thus, for the crops included in this study, optimum requirements in 1989/90 were estimated at 164 thousand tons of N and 96 thousand tons of P505. When the requirements of fruit trees and other crops not included in this study are added, optimum requirements would amount to 245 thousand tons of N and 148 thousand tons of P205 (see Appendix B). Actual fertilizer use in 1988/89 was 161 thousand tons of N and 109 thousand tons of P205, or approximately 601 of optimum requirements (refer to Table 2.9). In 1989/90, total fertilizer use declined to 154 thousand tons of N and 95 thousand tons of'FQOS. Such a decline, however, could not be predicted at the beginning of the 1989/90 season, especially given the upward trend in fertilizer use during the preceding years. Therefore, in estimating fertilizer availability at the beginning of the 1989/90 season, it was assumed that the level of actual fertilizer use would be 200 at least equal to that of 1988/89, i.e., around 601 of optimum requirements. This assumption will be the basis for most analyses presented in this chapter. 6.2 ELEMWWM 6.2.1 MW As discussed earlier, the fertilizer allocation model was developed based on separable programming (SP) techniques. SP approximates the curvilinear shape of the production functions by subdividing each function into several linear segments. Since SP is an approximation technique the results may differ from the true analytic solution that would be obtained if the problem was solved using calculus techniques. Theoretically, the fertilizer allocation problem could be solved using constrained optimization calculus techniques. In practice, however, such techniques could be very difficult to use given the complexity and large number of variables included in the fertilizer allocation problem. SP provides a more feasible approach to solving this problem. Furthermore, this approach can closely approximate the calculus results if a reasonable number of segments are used to approximate the production functions. One way to test the accuracy of the allocation LP model is to solve the model with the assumption of unlimited fertilizer supplies, and then compare the solution with that obtained by unconstrained oPtimization. This comparison is done using both economic and financial Prices. As shown in Table 6.1 (compare the results in columns 1 and 3 for the economic optima, and in columns 2 and 4 for the financial 201 Table 6.1: OPTIMUM FERTILIZER RATES BASED ON CALCULUS AND LP SOLUTIONS. Unconstrained Unconstrained Constrained2 Calculus Solutions LP Solutions LP Solution Economic Financial Economic Financial Financial Prices Prices Prices Prices Prices Crop: N P205 N P205 N P205 N P205 N P205 Irrigated Wheat: 136 65 142 72 135 65 140 70 134 65 Rainfed Wheat: HYV Zone 1: 82 38 87 45 80 40 90 45 80 40 HYV Zone 2: 62 35 67 41 60 35 70 40 60 35 LYV Zone 1: 41 36 43 42 4O 35 4O 40 4O 35 LYV Zone 2: 27 21 29 26 25 20 30 25 30 25 LYV Zone 3: 21 14 23 20 20 15 20 20 20 15 Rainfed Barley: Zone 1: 74 46 79 66 70 45 8O 65 70 45 Zone 2: 51 4O 55 60 50 40 6O 65 50 40 Zone 3: 39 37 44 57 40 35 4O 65 40 35 Cotton: 226 118 231 129 225 120 230 130 220 112 Corn: 130 803 135 803 115 71 125 77 115 71 Sugar Beets: Fall: 216 1203 218 1203 200 112 205 114 190 106 Summer: 192 1203 193 1203 175 111 180 114 170 107 Potatoes: Fall: 171 1203 173 1203 160 113 165 116 155 109 Summer: 157 1203 159 1203 145 112 150 116 140 108 1/ A11 fertilizer rates are in kg/ha. 2/ Assuming the same aggregate fertilizer requirements obtained by the unconstrained LP solution based on economic prices. 3/ Current recommended rates. 202 optima), the LP solutions for optimum fertilizer rates on wheat, barley, and cotton are very close to the analytic solutions, with differences averaging less than 5 kg/ha. However, the LP solutions for optimum N rates on corn, sugar beets, and potatoes are consistently lower than the analytic solutions, with differences of up to 18 kg/ha. Therefore, it is possible to conclude that the LP model is accurate in estimating optimum fertilizer rates on wheat, barley, and cotton, but would slightly underestimate optimum N rates on corn, sugar beets, and potatoes. As mentioned earlier, the production functions for corn, sugar beets, and. potatoes were estimated in terms of' N only, given the peculiarities of the data sets. Therefore, the current I505 rates on these crops were assumed to be the economically optimum rates, given the assumption of unlimited supplies. Under limited supplies, the LP model was set up to estimate Optimum P503 rates on these crOps by assuming a constant N:PZO, ratio. Thus, the model would estimate optimum N rates first, and would then compute optimum P20, rates based on the fixed- ratio assumption.‘ 6.2.2 LE ScImcIcn vs EQIEQIS' Qpcimc Another issue that needs to be addressed is whether the optimum fertilizer rates computed based on economic prices would also be optimal from the farmers' viewpoint. In other words, given that fertilizer allocations are based on economic prices while farmers' decisions are based. on financial prices, farmers may decide to reallocate their 1 Refer to chapter 3 for a more detailed discussion. 203 rations to maximize their net returns. Given these possible reallocations 'by' farmers, actual fertilizer' use may' not ‘necessarily represent an economically optimum allocation. One way to check whether farmers' reallocations would be significantly different from the LP optimum solution is to solve the model based on financial prices. This is shown in Table 6.1, whereby the upper limits on fertilizer supplies were set to be exactly equal to the aggregate requirements obtained with the unconstrained LP solution based on economic prices. The results indicate that the difference between the two solutions are minimal (compare column 3 and column 5 in Table 6.1). These results suggest that farmers would apply the same rates on wheat and barley as those obtained with economic prices. The only exceptions are the rates on LYV wheat in zone 2, with 5kg/ha of N and P20, more than the unconstrained LP solution based on economic prices. These extra amounts would be obtained by reducing fertilizer rates on cotton, corn, sugar beets, and potatoes by an average of 5 kg/ha. Therefore, these farmers' reallocations are somewhat minor and, thus, they are not expected to cause any significant deviations from the economically optimal solutions. 6.3 Wiles The fertilizer allocation model developed in this study is designed to answer two basic questions related to the problem of limited availability of fertilizers. The first question is how to allocate the limited fertilizer supplies among the various crops in a way that would give the highest economic returns to fertilizer use. In other words, given that total fertilizer supplies are lower than the actual 204 requirements, the optimum rates estimated earlier are no longer feasible. These rates need to be adjusted downward in such a way as to maximize net economic returns from the use of the available quantities of fertilizer. The second question is to estimate the impact of the current restrictions on fertilizer imports on the economy as a whole, on farmers' income, and on the government's foreign exchange and general budgets. 6.3.1 Qpcimmm EQIEIILZEI Rates Undez Limiced SuppIIec To answer the above two questions, and given that fertilizer supplies are expected to vary from year to year, optimum fertilizer rates are computed under several scenarios of fertilizer availability. The economic implications of these scenarios are then assessed 'by comparing them with those obtained under the “base” scenario of unlimited fertilizer supplies. For the crops included in the model, the base scenario (1002 of total fertilizer requirements available) would require 164 thousand tons of N and 97 thousand tons of’I50, (refer to chapter 5). The next step is to make gradually decreasing assumptions about the percentage of the above quantities actually available (901, 801, 702, 601, and 501). Based on the allocation model, it is then . possible to compute the optimum fertilizer rates that would satisfy the upper limits on supplies imposed by the above five constrained scenarios. Optimum fertilizer rates for the base scenario and the five constrained scenarios are presented in Table 6.2. For instance, if we examine optimum P20, rate on irrigated wheat, this rate gradually declines from 65 kg/ha under unlimited supplies to 55 kg/ha if 701 of 205 mm odd Hm mad ma mHH me omH an or om ass 0 o o o o o o o oH om om cm ow cm am mm on oca “can 2 "on an mHH «a and «a ems am omH me on om oca o o o o oH on oa ca om ca mm on ma ma en es an oca moan 2 now so no No on as OCH 0H 0H ma 0H mH mu am On mm Roam "on mNH mma ned ooH Om owH ow On On 0H On On mm Om OCH 2 Hoa omH on ooa no cmH no 05H mm Om oca com ma om mm on On oo ca ON ma oN 0m 0m 0m oc mm Co Co ONH III Roma 2 now cod oea moH nnH «0H mod oca omH mo oHH oHH oHN On On um no es oh as ow RH o~ an as mm ow an OR oa ona noaa 2 non «ad was "uoaasm mHH ooa "dash ”mooueuom Haa mma ”useEdm NHH oo~ ”Hash nausea woman an mag HF30 oma mmm "souuoo no es ”m soon 0: cm ”N OGON me oh "H coon ”Amanda oomcuem ma om "n ocou >xq cm mm "u «CON >>A mm oe "H oGON >>u mm 09 ”a econ >>= OS on “a moon >>z "UQOFS COMCHQ.“ as was ”acorn aaaamauau noun 2 “demo nooH oHQOHHO>< mucanuasoox mosqauuuom Hooch me u oaaaxamaa .aaasm VHHAHm< MMNHAHHmmm mo mAm>mA ozH>m<> «MQZD Ae£\wxv mma24 N ocoN >>Q H 0:0N >>u AN ocoN >>: "a crow >>= ”uses: consaom "uses: vouewuuua ”amazoaoauuooo Seauoaom oa\amaa .aaaAm “mmHHHUHHmHH<2202 MMNnAanMh no .o manna 224 6.4.2 Qciipicicn cf Ecpciiizc; Epicc Scenariog Based on the above fertilizer demand functions and price elasticities, it is possible to estimate the impact of increasing fertilizer prices on crop output, farmers' income, national income, and the government's foreign exchange and general budgets. To do that, five alternative fertilizer price scenarios are defined as follows (see Table 6.9): 1. Marin: This scenario represents the existing situation with official prices of N and P50, set at 10.65 SL/kg and 11.30 SL/kg, respectively. These prices represent 672 and 492 of the true economic value of N and I50, fertilizers, which were estimated to equal 15.85 SL/kg and 23.00 SL/kg, respectively (refer to chapter 4). In other words, subsidies account for about 332 of the true cost of N and 512 of the true cost of P20,. Also, this scenario assumes that only 602 of total fertilizer requirements are available and that the limited supplies ‘would ‘be allocated based on economic prices, as discussed earlier in this chapter. 2. Steam: This scenario assumes no change in current official fertilizer prices, combined with unlimited fertilizer supplies and unlimited farmers' access to these supplies at current prices. Thus, farmers are assumed to apply the fertilizer rates that would maximize their income. In other words, this scenario represents the ideal scenario from the viewpoint of farmers. Table 6.9: 225 DEFINITIONS OF FERTILIZER PRICE SCENARIOS Fertilizer Price Scenarios Base Official Prices N (SL/kg): 10.65 P205 (SL/kg): 11.30 2 of 1989/1990 Official Prices: N: 100 P2052 100 2 Subsidyl: N: 33 P205: 51 Constraints on Fertilizer Supplies: No Optimization Based on Financial Prices: NO3 1 10.65 11.30 100 100 33 51 Yes No3 2 10.65 11.30 100 100 33 51 No Yes 3 11.40 16.50 107 146 28 28 No Yes 12.35 17.97 116 159 22 22 No Yes 15.85 23.00 149 204 No Yes 1/ Refer to chapter 4 2/ 602 of total fertilizer requirements assumed available. 3/ Based on economic Prices. 226 3. Sccngric 1: Under this scenario fertilizer supplies are also assumed to be unlimited, with farmers applying the fertilizer rates that would maximize their income. The official price of N is assumed to be 72 higher than the current price, while the official price of'I50, would be 462 above the current price. The reason for assuming a larger increase in the price of P20,, relative to the price of N, is the concern of reducing the level of subsidy on P20, down to approximately the same level as the subsidy on N. Thus, the above assumed fertilizer prices imply a decline in government subsidies down to about 282 of the true cost of both N and P20, fertilizers. The assumed official fertilizer prices in this scenario were selected in such a way as to reduce subsidies to a point where they would be approximately equal to the increase in government revenues from indirect taxes on crops in a normal year. As mentioned earlier, these indirect taxes on crops refer to the difference between international and official prices, plus adjustments for other taxes (or subsidies) that affect the crop's cost of production (refer to chapter 4). Therefore, scenario 3 was defined in such a way that, in a normal year, net government expenditures associated with fertilizer use would be approximately equal to zero‘. In other words, under this scenario farmers would be able to purchase all their fertilizer needs, given the assumed official prices, without any increase in net government expenditures. 1 This result will be shown later (see Table 6.14). 227 4. Sccnaric_4: The assumptions of this scenario are very similar to those in Scenario 3. The only difference is that the official price of N is assumed to be 162 higher than the current price, while that of P50, would ‘be 592 above the current price. 'These prices would reduce subsidies to about 222 of the true costs of N and P50, fertilizers. Given that government revenues from indirect taxes are lower in good. years compared. to normal and. drier’ years1. the above assumed prices would ensure that expenditures on fertilizer subsidies would be approximately equal to the increase in tax revenues in a good year. In other words, this scenario would ensure that farmers will be able to purchase all their fertilizer needs, given the assumed official prices, without increasing net government expenditures even in good years. As will be shown later, in normal and drier years net government expenditures based on this scenario would actually decline (refer to Table 6.14). 5. Sccnagic_5: Under this scenario official fertilizer prices would be equal to their true economic value, which would lead to the complete elimination of fertilizer subsidies. This would require increasing the price of N by 492 and the price of P20, by 1042, compared to current official prices. Official crop prices are assumed to remain unchanged and, thus, indirect taxes on crops would substantially reduce government expenditures. In. other words, this scenario represents the ideal scenario from the viewpoint of the government budget. 1 Refer to chapter 4. 228 6.4.3 Qpcimmm EéE§§ mpde: Al£9128§1V§ Eccciiize: Ericc Sccpacicc The above five fertilizer price scenarios are compared in terms of their impacts on fertilizer use, crop production, farmers' income, national income, and the government's foreign exchange and general budgets. These five scenarios are also compared to a “base” scenario, which is defined. as the scenario that.‘would. maximize economic net returns from fertilizer use. This is the same base or ideal scenario defined in section 6.3, which assumes unlimited fertilizer supplies with optimum fertilizer rates computed based on economic prices. Optimum fertilizer rates calculated based on the six examined scenarios are presented in Table 6.10. With the exception of the constrained scenario (Scenario 1), optimum fertilizer rates do not show great variations between the different scenarios. The rates obtained under Scenarios 2, 3, and 4 are slightly higher than the base scenario, while those obtained under Scenario 5 are slightly lower. In other words, even with a 162 increase in the price of N and 592 increase in the price of P20, (Scenario 4), fertilizer use would be still slightly higher than the base scenario. Only when the price of 150, is doubled and the price of N is increased by about 502 (Scenario 5) that fertilizer use would be slightly lower than the base scenario. Therefore, as predicted by the low price elasticities, increasing fertilizer prices would have a somewhat limited impact on fertilizer use by farmers. The only exception to this observation seems to be in the case of P50, rates on barley. As suggested by the results in Table 6.10 and by the estimated price elasticities, these rates would decline by 402 to 452 if the price of P20, is doubled. 229 .uoHuoooom OOHuo HONHHHuuom mo nooHuHchoo one you m.o OHnoa com \H ONH mHH NHH «HH om HHH cm mm on >H o« m« H« OCON >>H mm 04 HH OGON >>H on cm H« OCON >>2 o: 0» HH OCON >>= Masons vouoHom no mnH "uses: oouemHuuH noam 2 “ mono oHuecoom omen ooo ~\aaaa .uaasm mmonm MmNHAHammm HmH 02H>2<> 22oz: Ao£\wxv mmH<¢ mmNHHHammm ZDZHHQO H0H6 OHAQH 230 6-4-4 IMWW Since higher fertilizer prices have a limited impact on fertilizer use, the impact on aggregate crop output is also expected to be limited. This is clearly shown in Teble 6.11, where the results indicate that higher fertilizer prices would virtually have no impact on the production of cotton, corn, sugar beets, and potatoes. Under Scenarios 2, 3, and 4, wheat and barley production would increase slightly, compared to the base scenario, while barley output under Scenario 5 would be slightly lower than the base scenario. However, the most important thing to note from Table 6.11 is that the current situation of limited fertilizer supplies (Scenario 1) would result in significantly lower levels of aggregate output for all crops, in comparison with the other scenarios. 6.4.5 Impccc cf flighc; Eezciiizc: Eziccc cm Epciongi Inccmc It is clear, then, that higher fertilizer prices (i.e., Scenarios 3, 4 and 5) would have a limited impact on fertilizer use and crop production. Thus, the impact on national income is also expected to be limited. In fact, compared to the base scenario, net economic returns from fertilizer use would decline by an insignificant amount due to higher fertilizer prices. In normal years, net economic returns under Scenarios 3, 4, and 5 (11164, 11172, and 11156 million SL, respectively) are only 8 to 24 million SL lower than the base scenario (11180 million SL). In good years, net returns under these scenarios (Scenario 3: 11605 million SL; Scenario 4: 11605 million SL; and Scenario 5: 11588 million SL) would be 41 to 60 million SL higher than the base scenario (11547 million SL) (see Table 6.12). 231 Table 6.11: AVERAGE CROP PRODUCTION INCREASE UNDER VARYING LEVELS OF OFFICIAL FERTILIZER PRICES Syria, 1939/1990 Fertilizer Price Scenarios1 Base 1 2 3 4 5 Crop: Irrigated Wheat: 482 432 490 488 487 481 Rainfed Wheat: HYV Zone 1: 237 173 244 241 241 235 HYV Zone 2' 104 73 109 108 107 104 LYV Zone 1: 67 58 69 69 68 67 LYV Zone 2: 93 74 100 99 98 96 LYV Zone 3' 14 10 15 15 15 14 TOTAL WHEAT 997 819 1027 1020 1016 998 Rainfed Barley: Zone 1: 25 17 27 26 26 24 Zone 2: 183 0 205 196 191 173 Zone 3: 92 0 106 100 97 84 TOTAL BARLEY 300 17 338 322 313 282 Cotton: 284 259 287 285 284 280 Corn: 118 87 119 119 118 118 Sugar Beets: Fall: 257 225 258 258 258 256 Summer: 228 197 229 229 229 228 TOTAL SUGAR BEETS: 486 422 487 487 486 484 Potatoes: Fall: 52 48 52 52 52 52 Summer: 39 36 39 39 39 39 TOTAL POTATOES: 91 83 91 91 91 91 1/ See Table 6.9 for the definitions Of fertilizer price scenarios. 2/ All figures refer to production increases (in thousand tons) relative to no fertilizer use. 232 Table 6.12: AGGREGATE ECONOMIC IMPACT OF FERTILIZER USE UNDER VARYING LEVELS OF FERTILIZER OFFICIAL PRICES Syria, 1939/1990 Fertilizer Price Scenarios1 Base 1 2 3 4 5 Fertilizer Use ('000 tons): - N: 164 98 174 173 171 163 - P50,: 97 58 116 106 103 92 EQQNQNIQ_£BIQ£§ (Million SL) Fertilizer Costs: 5,075 3,038 5,702 5,444 5,330 4,933 Gross Returns: - Good Year: 16,622 12,822 17,220 17,049 16,935 16,521 - Normal Year: 16,255 12,339 16,792 16,607 16,502 16,089 - Dry Year: 15,708 12,187 16,138 15,966 15,885 15,524 - Very Dry Year: 14,771 11,862 15,062 14,912 14,864 14,580 Net Returns: - Good Year: 11,547 9,783 11,518 11,605 11,605 11,588 - Normal Year: 11,180 9,301 11,090 11,164 11,172 11,156 - Dry Year: 10,633 9,148 10,437 10,522 10,555 10,592 - Very Dry Year: 9,696 8,823 9,360 9,468 9,534 9,647 1/ 2/ See Table 6.9 for the definitions of fertilizer price scenarios. See Table 3.1 for the definitions of rainfall scenarios. 233 The results in Table 6.12 also clearly show that a strategy based on higher fertilizer prices coupled with farmers' unlimited access to fertilizers (i.e., Scenarios 3, 4 and 5), would be much more economically efficient than the current situation of limited supplies and heavily subsidized prices (Scenario 1). In a normal year, such a strategy would increase national income by at least 1855 million SL (difference in net economic returns between Scenario 5 and Scenario 1), relative to the current policy . 6.4.6 12W Since higher fertilizer prices would have little impact on national income, their main effect would be to increase farmers' costs and to reduce government expenditures on fertilizer subsidies. Thus, the main impact of higher fertilizer prices is to redistribute income from farmers to the government ‘budget (i.e., other sectors of the economy). As shown in Table 6.13, farmers' net returns would be highest under the current levels of fertilizer prices with unlimited supplies (Scenario 2). An increase of 72 in the price of N and 462 in the price of P50, (Scenario 3) would reduce farmers' net returns in a normal year by 442 million SL (11763 vs 11321 million SL), compared to the base scenario. This potential loss in farm income would gradually increase with higher fertilizer prices, amounting to 764 million SL under Scenario 4 (11763 vs 10999 million SL) and 1858 million SL under Scenario 5 (11763 vs 9905 million SL). However, in spite of these substantial reductions in farm income relative to the base scenario, farmers' net returns would be still higher than under the current levels of fertilizer prices with Table 6.13 234 IMPACT OF VARYING LEVELS OF FERTILIZER OFFICIAL PRICES ON FARMERS ACCRECATE NET RETURNS TO FERTILIZER USE Syria, 1989/1990 Net Returns Fertilizer Price Scenarios1 Base 1 2 3 4 5 per Cropzz (Million SL) Irrigated Wheat: 3,027 2,796 3,040 2,941 2,879 2,661 Rainfed Wheat: HYV Zone 1: 1,382 1,069 1,392 1,327 1,286 1,145 HYV Zone 2: 584 436 594 557 535 463 LYV Zone 1: 459 410 465 442 432 394 LYV Zone 2: 610 509 630 589 570 503 LYV Zone 3: 89 64 91 84 81 69 TOTAL RAINFED WHEAT: 3,123 2,488 3,172 2,999 2,904 2,574 Rainfed Barley: Zone 1: 106 80 109 102 99 87 Zone 2: 725 0 763 691 654 544 Zone 3: 344 O 368 322 302 236 TOTAL BARLEY 1,175 80 1,241 1,116 1,055 867 Cotton: 3,101 2,972 3,106 2,992 2,924 2,687 Corn: 596 476 594 566 549 491 Sugar Beets: Fall: 202 187 201 192 186 167 Summer: 179 165 179 170 165 148 Potatoes: Fall: 206 197 205 198 193 178 Summer: 154 146 154 147 144 132 Aggregate Net Returns: - Good Year: 12,179 9,580 12,416 11,835 11,490 10,351 - Normal Year: 11,763 9,456 11,893 11,321 10,999 9,905 - Dry Year: 11,995 9,781 12,056 11,481 11,176 10,104 Very Dry Year: 11,480 9,716 11,414 10,852 10,577 9,567 1/ See Table 6.9 for the definitions of fertilizer price scenarios. 2/ Net returns in a normal year, based on financial prices. 3/ See Table 3.1 for the definitions of rainfall scenarios. 235 limited supplies (Scenario 1). In fact, even with the complete elimination of subsidies (Scenario 5), farmers' net returns to fertilizer use in a normal year would amount to 9905 million SL, or 449 million SL higher than their current levels. It should be noted, however, that the elimination of subsidies would reduce the income of farmers growing irrigated crops in comparison to Scenario 1. 6.4.7 Impacc cf fligher Eezciiizc; Egiccs cm chc QQVQIBEQDE Smcgec As mentioned earlier, the main impact of higher fertilizer prices would be to redistribute income from farmers to the government budget. Thus, the decline in farmers' income due to higher fertilizer prices (Scenarios 3, 4, and 5) would be coupled by a parallel decline in government expenditures. As shown in Table 6.14, fertilizer subsidies would gradually decline from 1986 million SL under the base scenario to 1455 million SL under Scenario 3, 1115 million SL under Scenario 4, and zero under Scenario 5. In comparison, the current levels of fertilizer subsidies (Scenario 1) would amount to 1188 million SL. It should be noted that Scenarios 3 and 4 were defined in such a way as to achieve specific alternative impacts on net government expenditures. As discussed earlier, fertilizer prices under Scenario 3 were specified in such a way as to equate fertilizer subsidies with the increase in government revenues from indirect taxes on crops in a normal year. Prices under Scenario 4, on the other hand, would result in equating fertilizer subsidies with government revenues from indirect taxation in a good year. This is shown in Table 6.14, where the increase in revenues from indirect crop taxation under Scenario 3 would amount to 1454 million SL in a normal year, compared to 1455 million SL 236 IMPACT OF VARYING LEVELS OF OFFICIAL FERTILIZER PRICES ON THE GOVERNMENT BUDGET Syria, 1989/1990 Table 6.14: Fertilizer Price Scenarios1 Base 1 2 3 4 5 (Million SL) Expenditures on Fertilizer Subsidies: 1,986 1,188 2,119 1,455 1,115 0 Increase2 in Revenues from Indirect Taxes on Crops: - Good Year: 1,113 1,126 1,112 1,113 1,115 1,115 - Normal Year: 1,431 1,126 1,469 1,454 1,446 1,411 - Dry Year: 1,265 1,041 1,291 1,279 1,273 1,248 - Very Dry Year: 1,212 1,028 1,233 1,222 1,218 1,197 Net Increase2 in Government Expenditures: - Good Year: 873 62 1,007 341 O -l,115 - Normal Year: 554 62 650 l -331 -1,411 - Dry Year: 721 147 828 176 -158 -l,248 - Very Dry Year: 773 160 886 232 ~103 -1,l97 1/ See Table 6.9 for the definitions of fertilizer price scenarios. 2/ Increase due to fertilizer use. 3/ See Table 3.1 for the definitions of rainfall scenarios. 237 in fertilizer subsidies. In contrast, fertilizer subsidies under Scenario 4 would be equal to 1115 million SL, which is approximately equal to the increase in tax revenues in a good year. Therefore, Scenarios 4 and 5 would ensure that net government expenditures associated with fertilizer use would gcclimc if the government decides to adopt a strategy of unlimited fertilizer supplies. This is particularly important under Scenario 5, where net government expenditures would decline by 1115 to 1411 million SL. Under Scenario 3, on the other hand, ensuring unlimited fertilizer supplies would still constitute a burden on the government budget, except in normal years. 6.4.8 Impgcc ct Higher Eertiiize; Ericcg cm Eorcign Exchange EarnmEE Although higher fertilizer prices will reduce the government budget deficit, foreign exchange expenditures on fertilizer imports would have to increase in order to ensure farmers' unlimited access to fertilizers. Under the current restrictions on imports (Scenario 1), the import value of fertilizers would amount to 62 million $US (Table 6.15). In comparison, strategies based on higher prices and unconstrained supplies (i.e., Scenarios 3, 4 and 5) ‘would require increasing fertilizer imports by: 49 million $US under Scenario 3 (i.e., 111 - 62 million $US); 47 million $US under Scenario 4 (i.e., 109 - 62 million $US); and 39 million $US under Scenario 5 (i.e., 101 - 62 million $US). These additional fertilizer imports would lead to higher levels of net crop exports (increased exports and reduced imports) whose value 238 Table 3.15: IMPACT OF VARYING LEVELS OF OFFICIAL FERTILIZER PRICES 0N FOREIGN EXCHANGE EARNINGS Syria, 1939/1990 Fertilizer Price Scenarios1 Base 1 2 3 4 5 (Million $US) Import Value of Fertilizer: 104 62 116 111 109 101 Value of Increased Crop Exports and Reduced Importszz - Good Year: 301 242 309 306 304 298 ~ Normal Year: 289 235 296 293 292 286 - Dry Year: 254 213 258 256 255 251 — Very Dry Year: 244 210 247 246 245 241 Net Increase2 in Foreign Exchange Earnings: - Good Year: 197 180 192 195 195 197 - Normal Year: 186 173 180 182 183 185 - Dry Year: 150 151 142 145 146 150 - Very Dry Year: 140 148 131 134 136 140 1/ See Table 6.9 for the definitions of fertilizer price scenarios. 2/ Relative to no fertilizer use. 3/ See Table 3.1 for the definitions of rainfall scenarios. 239 would offset the increased foreign exchange expenditures on fertilizer. For instance, under Scenario 3, the value of net crop exports would be 293 million $US in a normal year, compared to 235 million $US under limited fertilizer supplies (Scenario 1), a difference of 58 million $US. Such a difference would more than compensate for the 49 million $US increase in fertilizer imports, resulting in a 9 million $US gain in foreign exchange earnings. Similar gains in foreign exchange would also be obtained under Scenarios 4 and 5 (see Table 6.15). Therefore, under the higher fertilizer price scenarios (Scenarios 3, 4 and 5) net foreign exchange earnings would be slightly higher than their current levels in spite of substantial increases in fertilizer imports needed to ensure farmers' unlimited access to fertilizers. However, this would be the case in normal and good years only. In dry and very dry years, net foreign exchange earnings under the current polioy of subsidized but limited fertilizer supplies (Scenario 1) would be slightly higher than those obtained under Scenarios 3, 4 and 5. The differences, however, are very small and would be largely offset by increased foreign exchange earnings in normal and good years. 6.4.9 WW ROW The above results have clearly shown that a strategy based on ensuring unlimited fertilizer supplies and farmers' unlimited access to these supplies, combined with higher fertilizer prices, would be highly recommended. Compared to the existing situation of limited but heavily subsidized supplies, such a strategy would significantly increase 240 aggregate crop output and farmers' income, would. reduce government expenditures, and would slightly increase net foreign exchange earnings. The net impact of this strategy would amount to an increase of at least 1.8 billion SL in national income (Table 6.12), which is equivalent to a 3.22 increase in the agricultural GDP. The question of how much fertilizer prices ought to increase is essentially a political question. From a pure economic efficiency viewpoint, increasing prices within the range covered by the above five scenarios would have a very limited impact on national income. The main impact of higher prices would be to redistribute income from farmers to the government budget. If the political objective is to reduce government spending to control inflation, then the complete elimination of fertilizer subsidies (Scenario 5) would be the answer. Given the existing indirect taxes on crops, such a scenario would imply that farmers may end up subsidizing the rest of the economy. Thus, farmers groups are expected to strongly oppose any such move, particulary farmers growing irrigated crops whose income would decline if fertilizer subsidies are eliminated. If, on the other hand, the objective is to maximize farmers' income, then fertilizer prices ought to remain unchanged (Scenario 2). This would require increasing fertilizer subsidies by about 900 million SL (see Table 6.14). Given the current constraints on the government budget, these extra expenditures could only come at the expense of public spendings in other sectors. Therefore, a realistic solution to the fertilizer pricing problem would be closer to Scenarios 3 and 4. The difference in impact between these two scenarios would amount to about 330 million SL, which would be added to farm income under Scenario 3, or to government revenues under 241 scenario 4. Thus, an 'optimum' fertilizer pricing strategy’ would require an increase of 72 to 162 in the current official price of N and 462 to 592 in the price of P20,. Such a strategy would be desirable only if the government allocates the foreign exchange needed to import enough fertilizer to fill the gap between domestic production and total fertilizer requirements. 6.5 om-a or . . -_ a v- ' . 7. . 01 . — -- ~ . .- Wags 6.5.1 W The earlier comparison between the scenarios with varying levels of fertilizer supplies was based on total (winter and summer seasons) fertilizer requirements and ayailability (see section 6.3). However, constraints on fertilizer supplies have tended to be much more serious for fertilizer use on fall-planted crops (wheat, barley, and fall sugar beets and potatoes) than on spring-planted crops (cotton, corn, and summer sugar beets and potatoes). The peak period of fertilizer demand for winter crops, especially phosphate, occurs very early in the season (October-December). Therefore, any delays in. importing fertilizers and/or any early disruptions in production would cause serious reductions in fertilizer availability for the winter season. In fact, for the past few years, the impact of fertilizer shortages was mostly felt during the winter season, with spring-planted crops usually receiving all their fertilizer requirements. As mentioned earlier, fertilizer allocation strategies adopted by the Soils Directorate (SD) are based on a system of policy-based priorities (refer to chapter 2). According to this system irrigated 242 crops, including irrigated wheat, should receive their optimum fertilizer requirements. As for rainfed crops, the first priority goes to HYV wheat followed by LYV wheat, with barley fertilization having the lowest priority. As discussed earlier, given the serious fertilizer availability constraints for the winter season, the main. practical implication of the above ranking system is that fertilizer is rarely allocated to barley, especially in Zone 3. TO illustrate the kind of problems facing SD officials in planning fertilizer use on fall-planted crops, the fertilizer supply situation at the beginning of the 1989/90 season is described in some detail1. On September lst, 1989, the SD estimates of fertilizer supplies for the coming winter season were 126.9 thousand tons of N and 97.6 thousand tons of 150,. These estimates were based on existing stocks, realistic estimates of domestic fertilizer production between September lst and December 31st, and fertilizer import contracts signed. with foreign suppliers. In comparison, the SD estimated planned fertilizer requirements for the winter season at 175 thousand tons of N and 160 thousand tons of I50, (assuming all cropped areas will be fertilized). In other words, 732 of planned N requirements and 612 of planned P20, requirements were expected to be available for distribution to farmers for fall-planted crops. Based on the above estimates of available fertilizer supplies, the SD formulated its fertilizer allocation plan for the 1989/90 winter season as follows: 1 Based on discussions with officials from the Soils Directorate (SD) and internal documents from the SD and the Agricultural Cooperative Bank. 243 1. Irrigated wheat, sugar beets, and potatoes should receive all their current fertilizer recommended rates. 2. Rainfed HYV and LYV wheat in Zones 1, 2 and 3, should receive 802 of their current recommended rates. 3. Barley in Zone 2 should receive only 502 of its current recommended rates. 4. No fertilizer is allocated to barley in Zones 1, 3, and 4.1 5. Fertilizer allocations to fruit trees and vegetables are postponed until further notice, except fall-planted vegetables in the coastal areas, which should receive all their requirements. Based on the current fertilizer recommendations, the above allocation plan for the 1989/90 winter season would have been feasible if the SD estimates of available supplies were realistic. These supply estimates were based on the assumption that all fertilizer imports would be delivered in time for distribution to farmers. However, due to delays in payments to foreign suppliers, only a very small proportion of fertilizer imports was delivered early enough to be applied on fall- planted crOps. Thus, actual quantities of fertilizer applied. on fall-planted crops were around 78.4 thousand tons of N and 42 thousand tons of P50,. Subtracting the requirements of vegetables in the coastal areas, actual fertilizer use by the winter crops included in this study would amount 1 It is unclear why no fertilizer was allocated to barley in Zone 1. One possible explanation is that barley in Zone 1 is primarily grown for straw, with little amounts of grain sold through the official channels. As suggested by SD officials, fertilizer use for straw production. is considered. as a low' priority’ in. comparison. to grain production. 244 to 77.8 thousand tons of N and 41.5 thousand tons of £50,. The actual requirements of these crops, estimated in chapter 5, were 111 thousand tons of N and 67.5 thousand tons of 150,. Therefore, actual fertilizer use represented about 702 Of actual N requirements and 602 of P20, requirements. The delays in the delivery of fertilizer imports were not totally unpredicted by SD officials when the allocation plan was formulated. That is why the plan specified that the fertilizer requirements of irrigated crops would be satisfied first, followed by rainfed wheat, with fertilizer distribution to barley farmers conditional on the early delivery of fertilizer imports. Thus, based on the above allocation priorities, the actual amounts of available fertilizer supplies could only satisfy the requirements of irrigated crops and only part of the requirements of rainfed HYV wheat. Although there exist no data on actual fertilizer use by crop, the above figures on actual fertilizer supplies clearly suggest that very little fertilizer was applied on LYV wheat and barley. 6.5.2 Eczciiize; AIIocctiop SEIQEQELCS fc; che E1022: Seascn The main implication of the above fertilizer allocation strategy for the 1989/90 winter season is that very little fertilizer would be allocated to LYV wheat and barley. Thus, the question is posed as to the economic rationality of the above fertilizer allocation strategy. More specifically, one should ask whether, given the current fertilizer availability constraints, it would not be more economical to reduce fertilizer allocations to irrigated crops and to increase those for LYV wheat and barley, including barley in Zone 3? 245 To answer the above question, the fertilizer allocation model is used to compare the economic impact of alternative allocation strategies, given the quantities of fertilizer actually used during the 1989/90 winter season. For this purpose, three alternative fertilizer allocation strategies for the winter season are examined: - §£I££££1_A is the above mentioned strategy adopted by the SD. - Scraccgy S is based on the same system of priorities as in Strategy A, but all the fertilizer rates are based on the new recommendations proposed in chapter 5 (refer to Table 5.2), rather than the current rates recommended by the SD. - Scxcccgy_§ is based on the principle of equating marginal revenues across all crops, with no a priori conditions. This strategy is essentially based on the solution obtained by the LP model for fertilizer allocation. All the above strategies are based on actual fertilizer use on fall-planted crOps in 1989/90, which amounted to about 77.8 thousand tons of N and 41.5 thousand tons Of P20,. These quantities are approximately equal to 702 of optimum N requirements and 602 of P20, requirements for the fall-planted crops included in this study. Fertilizer rates included in the above three strategies are presented in Table 6.16. Economically optimum fertilizer rates under unlimited supplies, or base scenario, are also presented for the sake of comparison. As mentioned earlier, the current SD strategy (Strategy A) stipulates that irrigated crops should receive all their fertilizer 246 Table 6.16: FERTILIZER RATES UNDER ALTERNATIVE ALLOCATION STRATEGIES FOR THE WINTER SEASON Syria, 1989/1990 Fertilizer Rates (kg/ha) Base Scenario Strategy A1 Strategy B Strategy C Crop 2 N P205 N P205 N P205 N P205 Irrigated Wheat: 135 65 150 100 135 65 111 55 Rainfed Wheat: HYV Zone 1: 80 40 75 24 64 32 60 3O HYV Zone 2: 6O 35 60 20 48 28 40 25 LYV Zone 1: 40 35 O O 20 18 30 25 LYV Zone 2: 25 20 0 O 13 10 2O 10 LYV Zone 3' 2O 15 0 O 10 8 10 10 Rainfed Barley: Zone 1: 75 45 0 0 O 0 50 10 Zone 2: 50 40 O 0 0 0 30 10 Zone 3: 40 35 0 O 0 0 0 0 Sugar Beets: 215 120 200 120 215 120 160 90 Potatoes: 170 120 150 120 170 120 135 95 1/ Refer to text for the definitions of fertilizer allocation strategies. 247 requirements, rainfed wheat should receive 802 of the current recommended rates, and barley in Zone 2 should receive 502 of its recommended rates. However, after the allocations for the irrigated crops were distributed, the quantities of fertilizer left would have been barely enough to cover the plan's stipulations for HYV wheat. In fact, as shown in Table 6.16, there would be enough fertilizer left to provide 752 of the recommended N rates on HYV wheat in Zones 1 and 2, and only about 332 of the recommended P20, rates. Strategy B assumes the same set of priorities stipulated by the current SD fertilizer allocation strategy. The only difference is that Strategy 8 is based on the proposed fertilizer rates, which. were computed by equating marginal costs with marginal revenues, based on the true economic value of crops and fertilizers (refer to chapter 5). Compared to the current recommendations, these proposed rates are significantly lower for wheat and higher for barley, sugar beets, and potatoes. Given the substantial reduction in the fertilizer rates on irrigated wheat in particular, enough fertilizer would be left to apply 802 of the proposed rates on rainfed HYV wheat. However, the remaining fertilizer quantities would allow the application of only 502 of the proposed rates on LYV wheat, instead of the stipulated 802. As in the case of Strategy A, there would be no fertilizer left for barley fertilization. Unlike the above two strategies, Strategy C imposes no a priori conditions. Therefore, it is more flexible since it allows for reducing the fertilizer rates on irrigated crops. As shown in Table 6.16, this would, in turn, allow more fertilizer to be allocated to LYV wheat and to barley in Zones 1 and 2. However, as with the above two strategies, 248 no fertilizer would be allocated to barley in Zone 3. This provides further support to the current policy of excluding barley in Zone 3 from the fertilizer allocation plan. It should be noted that fertilizer rates on irrigated crops are only 152 to 252 lower than their corresponding rates under the base scenario. Therefore, by reducing the rates on irrigated crops by a relatively small percentage, enough fertilizer would be left to allow for moderate levels of fertilization on barley. 6.5.3 Impccc cf bicepnative Eeztilize; Aiiocacicn Stpmccgicg cm W The impact of the three fertilizer allocation strategies on aggregate crop production is presented in Table 6.17. The results indicate that the current high rates of fertilization on irrigated wheat would contribute only 3000 tons more in wheat output than the much lower rates under the proposed base scenario. On the other hand, the lower rates on rainfed HYV wheat and the lack of fertilization of LYV wheat would result in 210 thousand tons in foregone wheat output. This foregone *wheat. output. due to limited. fertilizer’ supplies would be less than 100 thousand tons had Strategy B been adopted instead of the current strategy. By applying the proposed lower rates on irrigated wheat under Strategy B, enough fertilizer would be left to fertilize LYV wheat. This would lead to the production of an additional 113 thousand tons of wheat compared to the current SD strategy. Similarly, under Strategy C wheat output would be 91 thousand tons lIigher than the output produced under the current strategy, but it would ‘be 22 thousand tons lower than in Strategy 8. Also, since the output of 249 Table 6.17: IMPACT OF ALTERNATIVE FERTILIZER ALLOCATION STRATEGIES 0N AGGREGATE PRODUCTION OF FALL-PLANTED CROPS Syria, 1939/1990 Average Increase1 in Aggregate Output ('000 tons) Base Scenario Strategy A Strategy B Strategy C Crop: Irrigated Wheat: 482 485 432 449 Rainfed Wheat: HYV Zone 1: 237 213 214 206 HYV Zone 2: 104 92 93 84 LYV Zone 1: 67 O 45 58 LYV Zone 2: 93 O 61 74 LYV Zone 3' 14 O 9 10 TOTAL RAINFED WHEAT: 515 305 421 431 TOTAL WHEAT 997 790 903 881 Rainfed Barley: Zone 1: 25 0 O 17 Zone 2: 183 0 O 110 Zone 3: 92 0 0 0 TOTAL BARLEY 300 O O 127 Sugar Beets: 257 253 257 232 Potatoes: 52 51 52 49 1/ Increase relative to no fertilizer use. 2/ Refer to text for the definitions of fertilizer allocation strategies. 250 irrigated wheat under Strategy B is 33 thousand tons higher than Strategy C, there would ‘be less uncertainty about aggregate wheat production under Strategy B. The impact of the three allocation strategies on barley production is more straightforward than their impact on wheat output. Since no fertilizer would be allocated to barley under Strategies A and B, there would be 300 thousand tons in foregone output in comparison to the base scenario. In contrast, under Strategy C fertilizer allocation to barley in Zones 1 and 2 would result in the production of 127 thousand tons more than the unfertilized barley under Strategies A and B. Table 6.17 also shows the impact of the above mentioned allocation strategies on the production of fall-planted sugar beets and potatoes. The results Show that under Strategies A and B aggregate output of these two crops would be essentially the same as in the base scenario. However, under Strategy C, sugar beet production would be 25 thousand tons lower than the base scenario, while potato output would decline by an insignificant amount. In summary, the above discussion has shown that the constraints on fertilizer supplies for fall-planted crops would result in 207 thousand tons in foregone wheat output and 300 thousand tons in foregone barley output, based on the current SD fertilizer allocation strategy (Strategy A). 'Under Strategy B, aggregate output of ‘barley, sugar beets, and potatoes would be the same as the current strategy, but wheat production would be 113 thousand tons higher. This is a clear indication that Strategy B would be superior to the current SD strategy. Barley production under Strategy C would be 127 thousand tons higher than the other two strategies. However, this increase in barley output 251 would be at the expense of wheat and sugar beet production, which would be slightly lower than under Strategy B. 6.5.4 Impacc of AIIcccciom Stcgtcgiec om Naciomai Imcomc The impact of the three fertilizer allocation strategies on national income is summarized in Table 6.18. The results show that, for approximately the same investment in fertilizer, the resulting increase in the economic value of crop production would be highest under Strategy C. In other words, the value of the additional 127 thousand tons in barley output would more than compensate for the slight decline in wheat and sugar beet production relative to Strategy B. In ‘normal years, net economic returns under Strategy C (6047 million SL) would exceed those under Strategy B (5473 million SL) by 574 million SL. These net returns would be 1532 million SL higher than those obtained under the current SD strategy (i.e., 6047 - 4515 million SL). Therefore, the results clearly indicate that Strategy C would be the most economically rational fertilizer allocation strategy. The results in Table 6.18 also show that the combined economic impact of the constraints on fertilizer supplies and the inefficiencies of the current SD allocation strategy would amount to 2375 million SL in a normal year (difference between the base scenario and Strategy A). This would be equivalent to an increase of approximately 4w22 in the agricultural GDP. This economic impact can be broken down into three separate components:1 1 All figures refer to net economic returns in normal years (refer to Table 6.18). 252 Table 6.18: AGGREGATE ECONOMIC IMPACT OF .ALTERNATIVE FERTILIZER ALLOCATION STRATEGIES FOR THE WINTER SEASON Syria, 1939/1990 Base Scenario Strategy A Strategy B Strategy C Fertilizer Use: N (tons): 111,042 77,731 75,958 77,801 150, (tons): 67,480 40,742 40,905 40,049 ECONOMIC ERIQES (Million SL) Fertilizer Costs: 3,483 2,283 2,257 2,267 Gross Returns: - Good Year: 10,704 7,174 8,272 8,705 - Normal Year: 10,374 6,798 7,730 8,315 - Dry Year: 9,827 6,775 7,595 8,051 - Very Dry Year: 8,890 6,601 7,283 7,533 Net Returns: - Good Year: 7,258 4,891 6,015 6,438 - Normal Year: 6,890 4,515 5,473 6,047 - Dry Year: 6,343 4,492 5,338 5,783 - Very Dry Year: 5,407 4,318 5,026 5,266 1/ See Table 3.1 for the definitions of rainfall scenarios. 2/ Refer to text for the definitions of fertilizer allocation strategies. 253 1. Impact of the constraints on fertilizer supplies for the winter season amounting to 843 million SL (difference between the base scenario and Strategy C). 2. Impact of a priori conditions imposed by the priority-based SD allocation strategy, amounting to 574 million SL (difference between Strategies B and C). 3. Impact of basing fertilizer allocation decisions on the current SD recommended rates, which would amount to 958 million SL (difference between Strategies A and B). 6.5.5 Impccc cf Allocaciom Sczacegiec om Eapm Imcomc Similar conclusions can be reached as to which allocation strategy would contribute most to farm income. As shown in Table 6.19, under Strategy C, farmers' net returns in a normal year would be approximately 1.5 billion SL higher than net returns under the current SD strategy, and 500 million SL higher than Strategy B. The combined impact of limited fertilizer supplies and allocation inefficiencies on farm income would amount to approximately 2.6 billion SL in a normal year (difference in farmers net returns between the base scenario and Strategy A). It should be noted that although Strategy G would substantially increase aggregate farm income in comparison to the current SD strategy, farmers growing irrigated crops would be worse off under Strategy C. 'This is especially true in the case of farmers growing irrigated wheat. ‘The results clearly suggest that reducing the rates on irrigated wheat would not affect yields and, thus, would not affect farmers' gross :returns. However, these farmers would be losing part of their 254 Table 6.19 IMPACT OF ALTERNATIVE FERTILIZER ALLOCATION STRATEGIES FOR THE WINTER SEASON ON FARMERS NET RETURNS TO FERTILIZER USE Syria, 1989/1990 Base Scenario Strategy A Strategy B Strategy C Net Returns per Crop1: (Million SL) Irrigated Wheat: 3,027 2,878 3,026 2,888 Rainfed Wheat: HYV Zone 1: 1,382 1,280 1,291 1,253 HYV Zone 2: 584 526 540 498 LYV Zone 1: 459 O 325 410 LYV Zone 2: 610 0 422 509 LYV Zone 3: 89 O 61 64 TOTAL RAINFED WHEAT: 3,123 1,806 2,638 2,734 Rainfed Barley: Zone 1: 106 0 O 80 Zone 2: 725 O 0 480 Zone 3: 344 O O 0 TOTAL BARLEY 1,175 0 0 560 Sugar Beets: 202 200 202 192 Potatoes: 206 202 206 199 Aggregate Net Returns: - Good Year: 8,148 5,092 6,181 6,767 - Normal Year: 7,732 5,086 6,072 6,573 - Dry Year: 7,964 5,520 6,452 6,904 - Very Dry Year: 7,449 5,606 6,423 6,716 1/ Net returns in a normal year, based on financial prices. 2/ Refer to text for the definitions of fertilizer allocation strategies. 3/ See Table 3.1 for the definitions of rainfall scenarios. 255 fertilizer allotments, which they could have applied to other crops and fruit trees or sold on the parallel market. 6.5.6 Impgcc cf Aiiocacicm Scrategieg cm Eoccign Smchamgc Eaznimgg The three alternative allocation strategies would require essentially the same amount of foreign exchange expenditures on fertilizer imports. However, as shown in Table 6.20, the differences in the value of net crop exports (increased exports and reduced imports) are substantial. Compared to the current SD strategy (Strategy A), the value of net crop exports under Strategy C are 13 million $US higher in normal years. Therefore, the shift from the current SD strategy to an allocation strategy based on equating marginal revenues across all crops (Strategy C) would increase net foreign exchange earnings in a normal year by 14 million $US (i.e., 47 - 33 million $US). This increase would range from 5 million $US (i.e., 15 - 10 million $US) in very dry years to 16 million $US (i.e., 55 - 39 million $US) in good years. 6.5.7 IWWMMMW Finally, the impact of the three alternative fertilizer allocation strategies on government expenditures is presented in Table 6.21. Expenditures on fertilizer subsidies under the three strategies are essentially the same. However, given the differences in the composition of aggregate output between these strategies, revenues from indirect taxes on crop output would also differ. As mentioned earlier, the difference in aggregate output between Strategies A and B would be an additional 113 thousand tons of wheat produced under Strategy 8. Thus, given that wheat is implicitly taxed, government revenues in a normal 256 Table 6.20: IMPACT OF ALTERNATIVE FERTILIZER ALLOCATION STRATEGIES FOR THE WINTER SEASON ON FOREIGN EXCHANGE EARNINGS Syria, 1939/1990 Base Scenario Strategy A Strategy B Strategy C (Million $US) Import Value of Fertilizer: 71 47 46 46 Value of Increased Crop Exports and Reduced Imports1: - Good Year: 127 85 95 101 - Normal Year: 116 80 88 93 - Dry Year: 80 59 64 66 - Very Dry Year: 71 57 6O 61 Net Increase1 in Foreign Exchange Earnings: - Good Year: 56 39 49 55 - Normal Year: 45 33 42 47 - Dry Year: 9 13 18 20 - Very Dry Year: 0 10 14 15 1/ Relative to no fertilizer use. 2/ See Table 3.1 for the definitions of rainfall scenarios. 3/ Refer to text for the definitions of fertilizer allocation strategies. 257 Table 6.21: IMPACT OF ALTERNATIVE FERTILIZER ALLOCATION STRATEGIES FOR THE WINTER SEASON ON THE GOVERNMENT BUDGET Syria, 1989/1990. Base Scenario Strategy A Strategy B Strategy C (Million SL) Expenditures on Fertilizer Subsidies: 1,367 881 874 873 Increase1 in Revenues from Indirect Taxes on Crops: - Good Year: 255 392 416 313 - Normal Year: 574 367 389 437 - Dry Year: 407 283 296 321 - Very Dry Year: 355 272 284 293 Net Increase1 in Government Expenditures: - Good Year: 1,112 489 458 560 - Normal Year: 793 514 485 436 - Dry Year: 960 598 578 552 - Very Dry Year: 1,012 609 590 580 1/ Increase due to fertilizer use. 2/ See Table 3.1 for the definitions of rainfall scenarios. 3/ Refer to text for the definitions of fertilizer allocation strategies. 258 year would be 22 million SL higher under Strategy 8 (i.e., 389 - 367 million SL). As mentioned earlier, under Strategy C there would be more barley and less wheat produced relative to Strategy B (see Table 6.17). Given that, during normal and drier years, indirect taxes on wheat are lower than those on barley, government revenues under Strategy C would be 48 million 81. higher than under Strategy 8 (i.e., 437 - 389 million SL). During good years, barley is implicitly subsidized (refer to the discussion in chapter 4) and, thus, government revenues under Strategy C would be lower than the other two strategies (see Table 6.21). Therefore, except in good years, Strategy C would result in the smallest increase in government expenditures associated with fertilizer use. In normal years these expenditures would amount to 436 million SL compared to 514 million SL under the current SD strategy, a net decline of 78 million SL. 6.5.8 Alternative Egztilize; Allocation Strategies for the Hinge; Season; Summary and 2911;! Implications The above analysis has clearly shown that an allocation strategy based on equating marginal revenues across all crops (Strategy C) would be the most appropriate strategy for fall-planted crops, given the existing constraints on fertilizer supplies. Such a strategy would require reducing the optimum fertilizer rates on irrigated crops by 151 tx> 251. This would leave enough fertilizer to be allocated to LYV wheat axui to barley in Zones 1 and 2. However, under the current constraints or! fertilizer supplies, fertilization of barley in Zone 3 would not be economical. The reduction of fertilizer rates on irrigated crops would 259 have a somewhat minor negative impact on aggregate production. However, this decline would be compensated by the substantial increase in rainfed wheat and barley output. In a normal year, this proposed strategy would increase wheat and barley output by 91 and 127 thousand tons, respectively, compared to the current SD strategy. If the current SD strategy is replaced by the proposed strategy, national income in a normal year would increase by 1532 million SL, which is equivalent to a 2.7% increase in the agricultural GDP. Also, in normal years, farm income would increase by 1487 million SL, foreign exchange earnings would increase by 14 million $US, and government expenditures would decline by 78 million SL. 6.6 W The main objective of this chapter was to analyze the impact of the constraints on fertilizer supplies and to compare alternative fertilizer allocation strategies under the current fertilizer supply constraints. Based on the linear programming model presented in chapter 3, the results in this chapter provided an illustration of how fertilizers would. be allocated among competing crops under 'varying levels of the constraints on fertilizer supplies. These results were also used to estimate the impact of the constraints on fertilizer supplies on aggregate crop production, national income, farm income, and the government's foreign exchange and general budgets. Current fertilizer supplies constitute approximately 601 of total fertilizer requirements of fall and spring-planted crops. The analysis has shown that the main effect of these supply constraints would be a substantial reduction in the fertilizer rates applied on rainfed crops, 260 particularly on barley. In fact, the results clearly suggest that barley fertilization in Zone 3 would be uneconomical under the current constraints on fertilizer supplies. Thus, approximately 282 thousand tons of barley and 178 thousand tons of wheat would be foregone because of reduced fertilizer application. Although fertilizer import costs would decline by 41 million $US, the value of increased crop imports would amount to 54 million $US in a normal year, a 13 million $US decline in foreign exchange earnings. This would imply a reduction of 1.9 billion SL in national income, while farm income would decline by 2.3 billion SL. These foregone incomes could be potentially obtained if the government would decide to import enough fertilizer to fill the gap between domestic production and total requirements. However, this would require an additional 600 to 800 million 51. in fertilizer subsidies. Given the current severe constraints on the government budget, such a substantial increase in government expenditures would be very difficult to implement. A feasible policy option would be to increase official fertilizer prices, which are currently highly subsidized, particularly the price of P205. The analysis has shown that, with unlimited fertilizer supplies, an increase of 71 to 161 in the price of N and 451 to 601 in the price of P20, would have a limited effect on fertilizer use, aggregate crap output, foreign exchange earnings, and national income. The main impact would be to redistribute income from farmers to the government budget. However, the unlimited access to fertilizer would allow farmers to increase their output, which would lead to higher revenues that would offset any increase in fertilizer costs. 261 Compared to the current situation of constrained fertilizer use, farmers' income would increase by at least 1.5 billion 81. in a normal year, in spite of the higher fertilizer prices. On the other hand, total expenditures on fertilizer subsidies would not be significantly affected, while government revenues from indirect taxes on crops would substantially increase due to higher crop output. Thus, the net effect would be a reduction in net government expenditures associated with fertilizer use. Given the frequent disruptions in domestic production and delays in imports, fertilizer shortages in the past few years have had serious effects on fall-planted crops in particular. Thus, the analysis focused on comparing alternative fertilizer allocation strategies for the winter season, with the 1989/90 season as an example. The current strategy adopted by the Soils Directorate (SD) gives priority to the fertilization of irrigated crops, followed by rainfed HYV wheat, LYV wheat, and barley. Based on this strategy, and given available fertilizer supplies for the 1989/90 winter season, no fertilizer would be left to be allocated to barley or LYV wheat. In contrast, a strategy based on equating marginal revenues across all crops, with no a priori conditions, would allow moderate fertilization levels for LYV wheat and for barley in Zones 1 and 2. Such a strategy would require reducing fertilizer rates on irrigated crops by 151 to 251, which would result in relatively minor declines in aggregate output. However, such declines would be offset by the substantial increases in wheat and barley production amounting to 91 thousand tons and 127 thousand tons, respectively. 262 Thus, a shift from the current SD strategy to the above proposed strategy would significantly improve the economic efficiency of fertilizer application on fall-planted crops. Given the existing levels of fertilizer supplies, such a shift would increase farm income by approximately 1.5 billion SL in a normal year, an increase of about 2.7% in agricultural GDP. Furthermore, this shift would increase foreign exchange earnings by 14 million $US and reduce net government expenditures associated with fertilizer use by 78 million SL. Therefore, this proposed fertilizer allocation strategy would be the most economically rational strategy from the viewpoint of the economy as a whole and for farmers, in. addition. to ‘being feasible under the government's foreign exchange and budgetary constraints. CHAPTER 7 SUHHARY, POLICY IMPLICATIONS AND FUTURE RESEARCH 7.1 Wain The main purpose of this study was to develop a national model for the centrally planned allocation of limited fertilizer supplies in Syria. This economic decision model was developed as a tool to be used by Syrian decision makers to plan more economically efficient annual fertilizer allocation schemes. Also, the model was used to estimate the economic impact of the proposed fertilizer allocation strategy in comparison with the current strategy adopted by the government. Furthermore, the model served as the main analytical framework for estimating the economic implications of the current constraints on fertilizer supplies, and for evaluating possible means for removing these constraints. The main underlying hypothesis of this proposed model is that a fertilizer allocation strategy based on equating marginal revenues across all crops is more economically efficient than the current government strategy. The current strategy is strongly influenced by policy concerns such as national food self-sufficiency and the balance of trade deficit. Thus, priority in fertilizer allocation is given to major export crops (e.g. cotton) and to crops that substitute for the main food imports (e.g. wheat). Also, fertilizer allocation to 263 264 irrigated crops has priority over rainfed crops, particularly rainfed barley in the drier zones, which is rarely allocated any fertilizer. Given the constrained optimization nature of the fertilizer allocation problem, the economic model was formulated in terms of a linear programming (LP) model. The objective function consists of maximizing the economic value of the increase in aggregate crop output due to fertilizer application, minus the value of total fertilizer quantities used. The main constraint in the model is the upper limit imposed on aggregate fertilizer supplies available for distribution to farmers during a given growing season. Such a formulation of the model allows for estimating economically optimum fertilizer rates for all crop activities covered by the model, under alternative levels of available supplies. The model's input/output matrix is based on the results of fertilizer experiments undertaken in Syria between 1965 and 1989. These experiments provided sufficient reliable data to estimate quadratic production functions for the most important crOps grown in Syria: wheat, cotton, barley, sugar beets, potatoes, and corn. These crops account for approximately two-thirds of total fertilizer consumption. Given the limited use of potassium fertilizers in Syria, only nitrogen (N) and phosphate (P205) fertilizers were included in the model. The first step in the research approach consisted of estimating production functions for the crops in the model. Based on these functions, economically optimum fertilizer recommendations for each crop were computed. These recommendations were calculated by equating marginal costs with marginal revenues, assuming no constraints on fertilizer supplies and based on the true economic value of fertilizers 265 and crops. In the next step, linear approximations of the estimated quadratic production functions were obtained by dividing the functions into linear segments, which were then incorporated into the LP model using separable programming methods. The optimum solution of the LP model provided estimates of the fertilizer rates that would maximize national economic net returns to fertilizer use under the existing constraints on aggregate supplies. These rates were then compared to the rates stipulated in the current allocation strategy adopted by the government. The policy objective of maximizing net economic returns from fertilizer use was used as the key criterion in evaluating alternative allocation strategies. These strategies were also compared in terms of their contribution to achieving other important policy objectives, including: (1) increasing aggregate crop output, (2) increasing farmers' income, (3) reducing foreign exchange expenditures, and (4) reducing the government budget deficit. The same criteria were used in assessing the impact of the current constraints on fertilizer supplies. This was done by comparing the results of the model under the current supply levels with those obtained under the ideal scenario of unconstrained supplies. These comparisons constituted the basis for identifying some of the obstacles to achieving the objectives of fertilizer policy and for assessing possible means to removing these obstacles. 266 7.2 umma n n 7.2.1 WWW Fertilizer recommendations were defined as the economically optimum rates under no constraints on aggregate fertilizer supplies. These rates were computed by equating marginal costs with marginal revenues based on the estimated production functions and the true economic value of fertilizers and crops. The results showed that these proposed recommendations would be profitable for farmers and economically feasible from the viewpoint of the economy as a whole. Furthermore, a sensitivity analysis of the results, using a minimum Value-Cost-Ratio (VCR) criterion of 1.5, clearly suggested that the proposed recommendations would constitute reasonably safe investments for farmers and for the economy as a whole. A comparison between the proposed and current recommendations clearly indicated the need for some major readjustments in the current recommendations, particularly on wheat and barley. Current fertilizer rates on wheat, especially P205 rates, need to be substantially reduced. The current N and P205 rates on low-yielding wheat varieties (LYV) are particularly excessive. These rates are, on the average, twice as large as the rates proposed in this study. In contrast, fertilizer rates on barley need to be increased significantly, especially P205 rates. This is particularly true in the drier areas (Zone 3)‘, where the proposed N and P205 rates are almost twice as large as the current ones. The 1 Refer to chapter 2 for the definitions of agro-climatic zones. 267 differences between the proposed and current fertilizer recommendations for the other crops are close.1 7.2.2 AW The constraints on fertilizer supplies during the past few years were generally more serious during the winter season (fall-planted crops) than for spring-p1anted crops, which usually receive all their fertilizer requirements. Thus, the analysis focused on comparing alternative fertilizer allocation strategies for the winter season only. Using the 1989/90 winter season as an example, three alternative fertilizer allocation strategies were examined: - Erma: This is the allocation plan adopted by the government for the 1989/90 *winter season. It: stipulates that irrigated. crops should receive all their fertilizer recommendations, rainfed, wheat should receive 801 of the recommended rates, and barley in the wetter areas should receive 502 of the recommended rates. No fertilizer was allocated to barley in the drier areas (Zone 3). However, given the delays in the delivery of fertilizer imports, the actual quantities of fertilizer available were barely sufficient to satisfy the plan's stipulations for irrigated crops and only part of the stipulations for rainfed high-yielding wheat varieties (HYV). No fertilizer was left to be applied on local wheat varieties (LYV) or barley. 1 Refer to Table 5.2 for a listing of the proposed and current fertilizer recommendations. 268 - W: This strategy is based on the same set of priorities stipulated by Strategy A, but all the fertilizer rates are based on the proposed recommendations estimated in this study, rather than the current recommendations. Given that the proposed recommendations are significantly lower for' wheat, particularly irrigated. wheat, enough fertilizer would be left to apply 80! of the proposed rates on rainfed HYV wheat, as stipulated by the plan. However, the remaining fertilizer quantities would allow the application of only 501 of the proposed rates on LYV wheat, instead of the stipulated 802. As in the case of Strategy A, there would be no fertilizer left for barley fertilization. - W: This strategy is based on the optimum solution obtained by the LP fertilizer allocation model. It is based on the principle of equating marginal revenues across all crops, with no a priori conditions. Therefore, this strategy is more flexible since it allows for reducing the fertilizer rates on irrigated crops. Given the available supplies, the results indicated that an economically optimal allocation would require reducing the unconstrained optimum rates on irrigated craps by 152 to 251. This would allow moderate fertilization levels for LYV 'wheat and for barley in the wetter areas (Zones 1 and 2). However, as with the above two strategies, no fertilizer would be allocated to ‘barley in the drier areas (Zone 3). A comparison of the economic impact of the above three strategies showed that a strategy based on equating marginal revenues across all crops (Strategy C) would give the highest economic net returns to the use of the limited fertilizer supplies. In a normal year, these 269 economic net returns would amount to 6.1 billion Syrian Liras (SL)1, compared to 4.5 billion SL under the current government strategy (Strategy A). This represents a 352 increase in the economic efficiency of fertilizer use and an increase of 2.71 in the agricultural GDP. Also, compared to Strategy A, farm income under Strategy C would increase by an average of 1.5 billion SL and foreign exchange earnings by 14 million $US, while net government expenditures would decline by 68 million SL. 7.2.3 W The main effect of the current constraints on fertilizer supplies is the substantially reduced fertilizer use on rainfed cereals, particularly barley. Thus, an average of 282 thousand tons of barley and 178 thousand tons of wheat are foregone because of the current constraints on fertilizer use. As noted earlier, the results clearly suggest that barley fertilization in the drier areas (Zone 3) would be uneconomical under the current constraints on fertilizer supplies. The value of the additional crop imports needed to compensate for this foregone crop output would amount to 54 million $US in a normal year. This represents 13 million $US more than the 41 million $US in additional fertilizer imports needed to satisfy total fertilizer requirements. This implies a decline of 71 in foreign exchange a reduction of 1.9 billion SL in national income, and a 2.3 reserves, billion 81. decline in farm income. These economic losses could be regained if the government adopts a policy of allocating enough foreign 1 1 $US - 40 SL (refer to chapter 4). 270 exchange to import all the fertilizer needed to fill the gap between domestic production and optimum requirements. An important precondition to implementing this policy is the to increase its current government's willingness and ability expenditures on fertilizer subsidies by 600 to 800 million SL. Given government budget, such a the current severe constraints on the substantial increase in expenditures would be difficult to implement. A potential solution to this problem is to reduce fertilizer subsidies by raising official fertilizer prices. These prices are currently highly subsidized, representing only 671 and 491 of the true economic value of However, higher fertilizer prices N and P20, fertilizers, respectively. This, may force farmers to reduce their fertilizer application rates. in turn, could lead to lower crop output and may result in reduced aggregate national income. Also, higher prices would increase farmers' production costs and may, thus, result in reduced farmers' income. Consequently, an important policy question is whether a strategy based to satisfy total requirements, in on increasing fertilizer imports conjunction with higher fertilizer prices, would be more economically efficient than the current policy of subsidized but limited fertilizer supplies. with unconstrained fertilizer The analysis has shown that, supplies, an increase of about 71 to 151 in the price of N and 451 to 601 in the price of P20, would have a limited effect on fertilizer use, aggregate crop output, foreign exchange earnings, and national income. The higher fertilizer prices would increase farmers' fertilizer costs. However, the unlimited access to fertilizer would allow farmers to increase their output and lead to higher revenues that would offset any 271 increase in fertilizer costs. Compared to the current situation of constrained fertilizer use, farmers' income would increase by at least 1.5 billion SL in a normal year, in spite of higher fertilizer prices. Host of this gain in farm income would benefit farmers growing rainfed crops, particularly barley. The combination of lower fertilizer subsidies and increased fertilizer use by farmers would result in a situation where total expenditures on fertilizer subsidies would not be significantly different from their current levels. In contrast, government revenues from indirect taxes on crops would substantially increase due to higher crap output. Thus, as shown by the results, the net effect would be a reduction in net government expenditures associated with fertilizer use, amounting to at least 60 million SL in a normal year. 7.3 W IllWWmamm The main stated objective of fertilizer policies in Syria is to increase agricultural production, which in turn would increase the income of farmers, increase food self-sufficiency, and reduce the balance of trade deficit by increasing agricultural exports and reducing imports. Currently, there are three main constraints preventing the full implementation of such policies: (1) technical and managerial problems facing domestic fertilizer production; (2) limited foreign exchange resources available for importing fertilizers; and (3) heavy fertilizer subsidies that constrain the government budget. During the past few years, the government has resorted to limiting fertilizer imports in an effort to reduce its foreign exchange deficit 272 and the trade deficit in general. Such a strategy can be effective only if domestic fertilizer production levels are increased to substitute for any decline in imports. This could be achieved by solving the problems facing fertilizer production and/or by expanding production capacity. However, the recent downward trends in fertilizer production clearly suggest that production problems are likely to continue to hamper the Also, any planned Syrian fertilizer industry for the next few years. capacity would require several years to expansion in production implement and would still be subject to the same foreign exchange constraints facing fertilizer imports. the option of increasing fertilizer production levels Therefore, In the meantime, any is feasible only in the medium or long run. reduction in fertilizer imports could only lead to fertilizer shortages. In fact, the combined quantities of fertilizer domestically produced and imported are at present barely sufficient to satisfy 60% to 702 of aggregate fertilizer quantities demanded by farmers, given the existing heavily subsidized prices. The results of this study showed that these fertilizer shortages have resulted in reduced aggregate crop production, including export crops and crops that substitute for major food imports. in net crop exports offset any savings in foreign These declines as indicated by exchange due to lower fertilizer imports. Therefore, the results, the net impact of the current policy of limited fertilizer imports is to further exacerbate the balance of trade deficit and the exchange deficit. Furthermore, this policy government's foreign significantly reduces the aggregate income of farmers and national income in general. 273 Therefore, the results of this study clearly suggest that the objectives of fertilizer policies would be more effectively achieved if current fertilizer import and allocation policies are modified. Based on these results, several policy options involving various possible adjustments or modifications to current policies are examined. These policy options are: (1) current fertilizer allocation strategy based on the proposed fertilizer recommendations; (2) equimarginal allocation of limited supplies; (3) unrestricted fertilizer imports at current official prices; and (4) unrestricted fertilizer imports with higher official prices. 7.3.2 MW Egztilizgz Eggommendggigns The current approach used by Syrian government planners in allocating the limited fertilizer supplies among various crops is to reduce the recommended rates by a given percentage. This percentage varies from crop to crop depending on the amounts of fertilizer available and depending on the importance of the crop based on the set of policy-determined priorities discussed earlier. Therefore, the accuracy of the current fertilizer allocation strategy depends to a large extent on the accuracy of the fertilizer recommendations. The results. of this study have indicated the need for some major readjustments in the current recommendations, particularly those on wheat and barley. Should the government decide to maintain the current restrictions on fertilizer imports and to continue allocating the limited supplies based on the existing priority system (i.e., Strategy B), then the mere readjustments in the fertilizer recommendations would 274 substantially enhance the efficiency of fertilizer use in Syria. As shown by the results, this would increase national income by an average of 958 million SL per year. 7.3.3 W A further improvement in the efficiency of fertilizer use can be obtained by replacing the current priority-based strategy for allocating the limited fertilizer supplies with a strategy based on equating marginal revenues across all crops. Such a strategy can be implemented using the linear programming allocation model developed in this study. The results have shown that if this model is used as a basis for allocating the limited fertilizer supplies, national income would increase by an additional 574 million SI. (i.e., in addition to the 958 million SL increase in national income due to the proposed readjustments in the fertilizer recommendations). It should be noted that the abandonment of the policy-determined priority system for fertilizer allocation does not contradict the policy priorities themselves. These priorities are based on the policy objectives of increasing national food self-sufficiency (mainly in wheat) and reducing the balance of trade deficit. The results have indicated that these policy objectives would be better achieved under the proposed allocation strategy. Compared to the current government strategy, the proposed strategy would increase wheat and barley output and foreign exchange earnings. Therefore, the results of this study have clearly demonstrated that an allocation strategy based on equating marginal revenues across all crops would ensure a more efficient use of the limited fertilizer resources as well as increasing wheat self- 275 sufficiency and reducing the balance of trade and foreign exchange deficits. 7.3.4 W A further step that the government could take to improve the efficiency of fertilizer use is to allocate sufficient foreign exchange to allow for importing the fertilizer quantities needed to fill the gap between domestic production and total requirements. It should be noted that such a policy would not necessarily eliminate the need for fertilizer rationing. This is because fertilizer recommendations and aggregate requirements were computed with the objective of maximizing net returns to the economy as a whole rather than farmers' net returns from fertilizer use. Farmers' optimum rates are generally higher than the proposed economic rates and, thus, the quantities demanded by farmers are likely to exceed the calculated total requirements. Also, rationing would be necessary since unlimited access to the heavily subsidized fertilizers may encourage smuggling to neighboring countries where fertilizers are more expensive. As mentioned earlier, that such a policy would require approximately 41 million $US in additional fertilizer imports. As a result, the value of increased net crop exports (increased exports and reduced imports) would amount to 54 million $US in a normal year. Therefore, the net impact of the current policy of restricted fertilizer imports would be to reduce the government's net foreign exchange earnings. This is in contradiction with the initial objective of reducing the government's foreign exchange deficit by restricting fertilizer imports . 276 Therefore, as long as domestic production cannot satisfy all of Syria's fertilizer requirements, the results of this study clearly suggest that the government ought to increase fertilizer imports to This policy of unrestricted fertilizer imports The cover the difference. would increase national income by an average of 843 million SL. combined economic impact of lifting import restrictions, the equimarginal allocation of fertilizers, and the proposed readjustments in the fertilizer recommendations would amount to an average of 2375 million $1. per year. This is equivalent to an increase of 4.22 in the agricultural GDP. Although such a policy option would substantially enhance the efficiency of fertilizer use in Syria and the productivity of the agricultural sector in general, there exist two potential obstacles to implementing this policy. The first obstacle is the centrally planned allocation of the scarce foreign exchange resources. This implies that fertilizer imports are competing with other sectors of the economy that on raw materials , are equally constrained by import restrictions equipment, spare parts, and so forth. This is particularly important given the magnitude of the additional fertilizer imports needed, which would consume up to 201 of total foreign exchange reserves. The results of this study have demonstrated that the economic returns to the additional fertilizer imports can be substantial. However, investing the scarce foreign exchange resources in other sectors of the economy might result in even higher returns than fertilizer imports. Also, the decision on how to allocate the scarce foreign exchange among the different sectors of the economy is essentially a political decision, with economic returns to investment 277 often playing only a secondary role in that decision. Thus, the scope of this study is limited to estimating the potential economic impact of the proposed policy of increased fertilizer imports. However, it is up to policy makers to decide whether this impact is sufficiently large to justify the implementation of such a policy. The second obstacle to implementing a policy of increased fertilizer imports is the heavy burden of fertilizer subsidies on the government budget. Such a policy would require the government to spend up to 800 million SL in additional fertilizer subsidies. facing the Syrian government, However , given the serious budgetary constraints increasing expenditures on fertilizer subsidies would exacerbate the government budget deficit. Given that the government budget deficit is usually financed of the money supply, increased through inflationary expans ion might further exacerbate the expenditures on fertilizer subsidies already high inflation rate. In Syria, the most direct effect of inflation is to reduce the purchasing power of public sector employees, who constitute the majority of the urban middle and lower-middle Thus, higher inflation would be highly unpopular and, hence, classes. any increase in expenditures on politically undesirable. Therefore, fertilizer subsidies would have to be either at the expense of reducing public expenditures on other sectors of the economy, or through tax As with the allocation of foreign exchange resources, the increases. decision of how to allocate public spendings is influenced by relative economic returns to public investments in each sector as well as political factors . ‘h—‘nh-JI 278 7.3.5 2 - -- e -- moo t w , :h- 0 It is clear from the above discussion that the burden of fertilizer subsidies on the government budget constitutes an important obstacle to implementing a policy of increased fertilizer imports. This budgetary constraint can be eliminated if official fertilizer prices are increased. As shown by the results, an increase of about 71 to 151 in the current price of N and 451 to 601 in the price of P205 would ensure that the current expenditures on subsidies would remain approximately the same. Furthermore, these substantial reductions in subsidies would bring domestic fertilizer prices closer to their international market equivalents. With the proposed higher prices, subsidies would still represent 20 to 301 of the true value of fertilizers. However, this subsidy level would significantly reduce the incentives for smuggling fertilizers out of Syria. Thus, a policy of increased fertilizer imports in conjunction with higher official prices would not require fertilizer rationing. In other words, farmers would be able to purchase all the fertilizer they need to maximize their net returns. Given the differences between farmers' prices and the true value of fertilizers and. crops, this unlimited. access to fertilizers may result in an economically inefficient use of fertilizers. However, the results have shown that, with the higher fertilizer prices, farmers' optimum rates would be very close to the calculated economic optima. Also, it should be noted that the proposed increase in the price of P505 is more drastic than for N. This is because the current price of P505 represents only 492 of its true economic value, compared to 671 in the case of N. Recent research findings strongly suggest that farmers may 279 be applying excessive P20, rates, particularly on wheat (Soils Directorate/ICARDA, 1988, pp. 10-13). Therefore, the more drastic increase in the price of P205 may dissuade farmers from applying excessive P20, rates. The results of this study suggest that the proposed higher prices would have a limited effect on the fertilizer rates that would maximize farmers' net returns. Farmers' unlimited access to fertilizers would allow them to apply these rates and to increase their revenues from the higher yields. These higher revenues would offset the increase in fertilizer costs. Therefore, in spite of higher fertilizer prices, farmers' net returns would be higher than under the current situation of subsidized but limited fertilizer use. It should be noted that the above results were based on the fertilizer demand equations computed based on the estimated production functions. These demand equations were formulated in terms of the effect of fertilizer prices on the quantities of fertilizer applied. However, fertilizer prices are not the only factors affecting the demand for fertilizers. Other factors, such as crop prices and the farmer's cash flow situation, could be equally important. Furthermore, the normative demand equations computed in this study were derived from production functions based on data from fertilizer research experiments, which are seldom representative of farmers conditions and constraints. The above discussion suggests that the results of this study should be viewed with caution, especially with regard to farmers' response to higher fertilizer prices. Thus, if the proposed price increases are to be implemented, this should be done gradually over several years. If there is evidence that farmers are reducing their 280 fertilizer rates in response to higher prices, compared to the current rates under limited supplies, then a re-evaluation of the price-increase policy would become necessary. This cautionary note on the possible negative impact of higher fertilizer prices is particularly relevant with respect to crops, such as barley, where fertilizer use is still limited. The Ministry of Agriculture and Agrarian Reform has recently initiated a program to promote barley fertilization through demonstration plots located in the main barley-growing regions. Also, major efforts by the Agricultural Extension Directorate to encourage farmers to adopt barley fertilization are planned for the next few years. The success of these efforts depends to a large extent on the high rate of return that farmers would expect from barley fertilization. The results of this study have confirmed previous research findings showing that barley fertilization is profitable and reasonably safe (see Soils Directorate/ICARDA, 1990). However, higher fertilizer prices will reduce the farmers' expected rate of return and may, thus, act as a disincentive to fertilizer use. The rate of return to barley fertilization can be increased through higher official barley prices that would offset the increase in fertilizer prices. However, such an option may not be feasible given its potential heavy burden on the government budget. Also, higher official barley prices would contribute to a further increase in feed costs, which could have a serious negative impact on livestock production in Syria. Furthermore, higher feed prices may encourage livestock producers to rely more on natural pastures as substitute sources of feed. This could further exacerbate the already serious problem of overgrazing in the steppe. Therefore, given the importance 281 of barley in Syrian agriculture and its complex linkages with the rest of the economy, the option of increasing official barley prices should be examined in a much broader context than the limited objective of encouraging fertilizer use on barley. This is beyond the scope of this study and can. only' be addressed through a comprehensive empirical analysis of the barley sub-sector. Thus, there might be a strong argument in favor of maintaining the lower fertilizer prices as long as a large number of farmers do not apply fertilizer on their crops. However, cheap fertilizers may lead to excessive use on crops, such as irrigated wheat and cotton, where application rates may be already too high. This would not only result in uneconomical uses of the limited fertilizer resources, but may also lead to water pollution problems such as the nitrate buildup in the watertable. One possible solution to the above dilemma, which needs to be addressed in future evaluation of fertilizer price policies, would be to implement a two-price system for fertilizer sales. According to this system, a portion of total fertilizer supplies would be allocated to farmers at current low prices, while the remaining supplies would be sold at the higher prices. This system would be similar to the present system of fertilizer rationing and allocation. The only difference is that farmers would have the option of purchasing all their fertilizer needs at the higher prices, if their rations are not sufficiently large to maximize their net returns. A two-price system for fertilizers would ensure that all farmers would have unlimited access to fertilizers. Such a system could also be useful in easing the transition from low to high fertilizer prices. 282 Thus, while prices are gradually being increased, farmers would still have access to cheap fertilizer rations that would dampen the impact of higher prices. However, as long as subsidized fertilizers are rationed, the fertilizer parallel market would continue to operate. The importance of the parallel market would depend to a large extent on the margins between the two sets of fertilizer prices. If these margins are gradually reduced, then the parallel market could ultimately disappear. A two-price system may also allow the government to promote fertilizer use on specific crops, such as barley, or to favor poorer farmers in the drier zones, as suggested by El-Sherbini and Sinha (1978, p. 94)‘. It should be noted that if farmers make production decisions on the basis of marginal costs, and if their ration does not fulfill all their fertilizer needs, they will make their fertilizer application decisions based on the higher prices. Thus, a two-price system would not necessarily induce barley farmers to increase their fertilizer rates. 00 the other hand, the two-price system would reduce W fertilizer costs and would, thus, increase the average rate of return to barley fertilization. This would encourage farmers who are currently applying no fertilizer on barley to experiment with and ultimately adopt barley fertilization. 1 See also Akinola (1987) for a similar suggestion concerning input subsidies in Africa. 283 7.3.6 Qtnez Enligy Issues 7.3.6.1 o i ers :1 c e s a - flinging: The policy objective of increasing national food self-sufficiency, particularly in wheat production, and the role of fertilizers in attaining this objective are important policy issues that deserve further discussion. The results of this study indicated that all the current fertilizer recommendations on wheat are excessive. Host of these rates, i503 rates in particular, are so excessive that they exceed the rates that would maximize yields. Based on discussions with agronomists from the Soils Directorate, two important factors that contributed to this situation of possible excessive fertilizer use on wheat were identified. The first factor is that most of the wheat recommendations were computed based on fertilizer trials conducted in the early 1970's. These trials were performed on research and farm plots not previously fertilized. Thus, the calculated rates for the most part reflect initial low soil fertility levels. By the late 1980's, after two decades of fertilizer application by farmers, there is clear evidence of substantial buildup of soil nutrients, particularly phosphate, resulting in a reduced response of wheat to fertilizer application (ICARDA/FRMP, 1988, pp. 10-13). The second factor that contributed to the excessive fertilizer recommendations on wheat is the political pressure on policy makers. Given that wheat self-sufficiency is an important political issue in Syria, policy makers and planners at the Ministry of Agriculture and Agrarian Reform are under constant political pressure to increase wheat self-sufficiency. Thus, fertilizer requirement and allocation decisions 284 during the past decade were often influenced by such political pressure, leading to the gradual increase in the recommended fertilizer rates on wheat. The results of this study strongly suggest that the role of fertilizers in increasing wheat self-sufficiency in Syria has become increasingly limited. Therefore, the excessive use of the limited fertilizer supplies to further increase wheat yields could only lead to an inefficient use of resources. Policy makers, planners, and agronomists should explore other means to increasing aggregate wheat output. These include increasing the official price of wheat and intensifying the use of other inputs such as irrigation, high-yielding varieties, pest and weed control, cultural practices, and so forth. 7.3.6.2 We: Another important issue, which is directly related to concerns about the potential excessive use of fertilizers on wheat, is the current policy debate on whether to reallocate some fertilizer from wheat to barley grown in the drier areas. This is based on results from recent fertilizer experiments on barley in northern Syria conducted jointly by the Soils Directorate and ICARDA. The results of these experiments suggest that barley fertilization in the drier areas (Zone 3) might be much more profitable and less risky than was previously thought (Soils Directorate/ICARDA, 1990). Based on these results, economic optimum rates on barley in Zone 2 were estimated at 54 and 49 leg/ha for N and P205, respectively, whereas optimum N and P205 rates on barley in Zone 3 were estimated at 56 and 44 kg/ha (ibid. , p. A16, Table 20). 285 In contrast, the results of this study showed that, under the current constraints on total fertilizer supplies for fall-planted crops, optimum N and P50, rates on barley in Zone 2 would be 30 and 10 kg/ha, respectively, whereas barley fertilization in Zone 3 would be uneconomical. Therefore, the results of this study suggest that the current government policy of not allocating any fertilizer to barley in Zone 3 is economically sound given the existing constraints on aggregate fertilizer supplies. However, these results also suggest that fertilizer rates on wheat should be reduced to allow for moderate fertilization levels on barley in Zones 1 and 2.1 It should be noted that, in this study, the production functions for barley were estimated based on the assumption that barley farmers, being risk averse, would make their fertilizer application decisions by assuming a worst-case scenario (maximin assumption). That is, they would assume that the coming season is going to be a dry one and would, thus, apply the rates that would maximize their net returns in the event of a dry year. These rates are necessarily lower than those calculated based on the expectation of a normal year, had we assumed that barley farmers were risk neutral. This assumption of worst-case scenario may be one of the reasons why the optimum rates on barley estimated in this study are lower than those estimated in the SD/ICARDA study mentioned earlier. Another possible reason for this discrepancy is that the optimum rates in the SD/ICARDA study were computed based on financial fertilizer and crop 1 Refer to Table 6.15 (Strategy C) for the details of the proposed fertilizer reallocations from wheat to barley, given the current constraints on total fertilizer supplies available for the winter season. 286 prices, i.e., the actual prices faced by farmers. However, the results of this study showed that the use of financial prices may lead to an economically inefficient use of fertilizers. This is particularly true in the case of barley, where the relative product-to-fertilizer price ratio is substantially higher when calculated based on financial prices compared to a ratio based on economic prices. Since the fertilizer recommendations on barley proposed in this study' were based. on somewhat conservative assumptions, these rates should, thus, be viewed as minimum levels. The results clearly suggested that there is ample space to gradually increase these rates in the future. Furthermore, an essential step in increasing fertilizer use on barley would be to expand fertilizer application to most barley areas. If the required fertilizer supplies are made available, the results have shown that the use of fertilizers by most barley farmers would result in an average production increase of 800 thousand tons over the current level of 1.1 million tons in aggregate barley output. These potential increases in barley production are in line with stated government policy objectives. However, for these objectives to be attained, the government should follow a policy of ensuring the availability of sufficiently large fertilizer supplies to economically justify barley fertilization. Furthermore, the current efforts by the agricultural extension services to expand the adoption of barley fertilization should be intensified. Also, since barley is grown mostly in the drier and more remote areas of Syria, increasing farmers' access to fertilizer retail outlets would be crucial to the success of any policy aimed at increasing fertilizer use on barley. 287 The current government strategy to increase barley production is based on expanding its area of cultivation. This has meant expanding barley cultivation into the ecologically fragile lands of the steppe (Zone 5). Also, farmers are currently encouraged to replace the traditional barley-fallow rotation with continuous barley cultivation. This, however, may cause rapid depletion of soil nutrients and the buildup of diseases and insects, as suggested. by discussions ‘with agronomists from the Soils Directorate and ICARDA. This would ultimately lead to a decline in yields and, possibly, in total barley production. Therefore, in the long run, barley fertilizationr may constitute a more economically and ecologically sound policy option for a more sustainable growth in barley production in Syria. 7-4 W This study was entirely based on secondary data. No attempt was made to generate any primary data given the resource and time limitations. The existence of relatively large sets of fertilizer trial data made it possible to entirely rely on these secondary data in estimating the production functions which constitute the core of the fertilizer allocation model. Thus, the main idea was to make the best use of all the available fertilizer trial data. By relying on pooled statistical analysis techniques, it was possible to use data from different years and from different locations to come up with what we believe to be the most accurate estimates possible, given the existing data. Hence, the accuracy and relevance of the results presented in this study are largely dependent on the accuracy and relevance of the existing fertilizer trial data sets. 288 For this reason, future research efforts should concentrate on generating yield response data to fertilizer application that are more accurate and more representative of actual farm conditions. These research efforts on fertilizer yield responses should include the following: 1. Future fertilizer experiments should be designed to estimate fertilizer recommendations for more specific recommendation domains for each crop or variety. Host current recommendations are made for the country as a whole. However, yield responses to fertilizer application may vary tremendously from region to region and within the same region due to variation in soil type, rainfall levels and distribution, cultural practices, and so forth. Although some region-specific data already exist, especially in the case of wheat and cotton, most of these data are outdated and uneven in terms of quality and research design. Thus, future research should include a more systematic and concentrated effort at designing experiments aimed specifically at the development of region-specific fertilizer recommendations. Such experiments should be undertaken on farmers' fields and be preferably managed by farmers themselves, with minimum interference from researchers, to ensure that yield response data are representative of actual farmers' conditions. After decades of fertilizer trials in research stations, with the likely buildup of soil nutrients, yield responses observed in experimental plots are probably becoming less and less representative of farmers' conditions. Furthermore, the design of these experiments should include measurements of the residual effects of 289 fertilizer applications in order to incorporate fertilizer carry-over effects into the economic analysis of fertilizer use. 2. More attention should be given to fertilizer research on crops and fruit trees that have not been adequately studied yet, if at all. This is especially relevant with respect to fruit trees, such as olives, grapes, and citrus, which are becoming increasingly important in terms of total area, contribution to farm income, and their potential consumption of large quantities of fertilizer. The same lack of fertilizer trial data can also be observed with respect to barley in Zone 4, vegetables, food and feed legumes, forage crops, and natural pastures. Given that the fertilizer allocation model developed in this study should, ideally, include all fertilizer use activities, the exclusion of the above crop and fruit tree categories represents a serious weakness that ought to be addressed in the future. 3. Attempts should be made to try alternative formulations for estimating the production functions. The quadratic polynomial formulation gave generally acceptable results. The main problem encountered as a result of the quadratic formulation was in the case of wheat. When the estimated production functions for wheat were solved for the current excessive recommended rates, the results obtained suggested that yields would decline if these rates were actually applied. Although most agronomists from the Soils Directorate and ICARDA seem to agree that the current rates are too high, they have questioned the validity of the yield-decline implication. They suggested that a 290 more commonly observed response to very high fertilizer rates, especially phosphorus rates, is that of a yield plateau. This problem was addressed in this study by assuming that the current rates on wheat would give the same yield as that obtained by the lower yield-maximizing rates (refer to chapter 5). A more appropriate approach would have been to re-estimate the wheat production functions using other functional forms such as the Hitscherlich function or the linear response and plateau (LRP) function (refer to the discussion of functional forms in chapter 3). Furthermore, alternative functional forms will be needed if the residual effect of applied fertilizers is to be incorporated into the analysis. This would allow more appropriate modeling of yield response to fertilizer application and would lead to more accurate economic analyses of fertilizer use. However, this would require the implementation of more accurate soil testing procedures by the Soils Directorate to generate the required data. 4. In addition to the basic reliance on on-farm fertilizer trials to determine optimum fertilizer recommendations, more use should be made of farm surveys to complement the information obtained from fertilizer trials. Farm surveys would be most useful in the case of crops where fertilizers have been used for a relatively long time. This is the case of most irrigated crops and rainfed HYV wheat in the wetter zones. Such surveys could provide a cost-effective method that will give a more accurate picture of what rates farmers actually apply and the yields obtained across agro-climatic regions. It is even possible to estimate yield response functions based on these surveys, provided that enough 291 variation exists within the survey sample to cover the entire range of the production function. If these surveys are repeated over several years, yield responses under varying rainfall levels can be estimated providing a strong empirical basis for risk analysis. In addition to the above issues related to fertilizer trials, future research is needed in the area of price and cost estimations to allow for more accurate economic analyses of results from fertilizer experiments. Such research should focus on monitoring seasonal and regional variations in crop prices as well as prices of related agricultural byproducts (e.g., straw). This would allow the formulation of more accurate region-specific fertilizer recommendations. These price-monitoring efforts should cover the domestic parallel markets as well as prices in the international market, including spot and future prices. This would provide a more accurate basis for making price projections needed to be incorporated in the fertilizer allocation model. Other marketing-related issues that need to be addressed in the future include detailed studies to estimate storage and transport costs of crops and fertilizers. These would provide more accurate estimates of financial and economic field prices. Future research should also examine the policy option of legalizing the private domestic trade in fertilizers. Such an option would be in line with recent political statements stressing the need for more coordination and complementarity between the activities of the public and private sectors. Private fertilizer marketing, especially retailing, could enhance farmers' accessibility to fertilizers and contribute to reducing many of the inefficiencies in the official 292 fertilizer distribution system. This would be particularly relevant if fertilizer rationing is eliminated and subsidies are substantially reduced. As for the fertilizer allocation model, several possible improvements can be suggested for future modifications in the model, provided that the relevant data requirements become available in the future. These improvements include the following: 1. Despite the current low levels of potassium fertilizer use in Syria, the inclusion of potassium fertilizer in the allocation model will become crucial in the future. This is important given the expected rapid increase in the application of potassium fertilizers, particularly on fruit trees. At present, most soils in Syria are considered sufficiently rich in potassium not to require the application of any potassium fertilizers on most crops, with the exception of root crops (e.g., sugar beets and potatoes). However, regular monitoring of soil potassium levels is needed in order to detect any signs of potassium depletion in the soil and, if needed, to recommend its application to prevent future potassium mining that may lead to serious yield declines. 2. The allocation model developed in this study does not differentiate between the costs of domestic and imported fertilizers. The use of import parity prices for all the fertilizer used in Syria was justified based on the fact that any increase or decline in fertilizer use would be reflected by an equal increase or decline in fertilizer imports. That is, the opportunity cost of fertilizer was assumed to be its import cost. Such an assumption can be maintained as long as Syria 293 remains an importer of fertilizer. However this situation is expected to change in the medium run given the plans to substantially increase current production capacity. Therefore, in few years, Syria might resume exporting significant amounts of its fertilizer output, which would require the use of export parity prices. 3. The model used in this study is static. The time dimension is not explicitly addressed either in the formulation of the response functions or in the design of the model itself. The current formulation of the model requires annual updating of the input/output coefficients, prices, and the upper limits on available fertilizer supplies. Although data on prices and fertilizer supplies are readily available, updating of production functions would be more problematic given the limited new fertilizer trial data expected to become available every year. This static nature of the model does not allow for the appropriate treatment of several important variables that are time-dependent. These include crop rotations, fertilizer carry-over effects, seasonal distribution of rainfall, fertilizer and crop inventories, and so forth. Future modifications of the model should also address possible farmers' reactions to any changes in the system. The use of a dynamic programming model would allow for the incorporation of some of these variables, which would greatly enhance the accuracy and usefulness of the fertilizer allocation model. 4. A related issue is the treatment of risk and uncertainty in the model. The present model incorporates risk considerations in an :indirect way by enabling to find solutions to the fertilizer allocation 294 problem under different rainfall scenarios. Also, assumptions were made about the risk-averse behavior of farmers with respect to barley fertilization. These very simplistic assumptions were made given the limited empirical information on farmers' risk management strategies with respect to fertilizer rates used, timing, of fertilizer applications, and the number of applications. thure research should focus more on understanding farmers' behavior under uncertainty and identifying their risk strategies. This would allow the formulation of more realistic fertilizer recommendations and would provide the basic information needed for a more systematic treatment of risk in the fertilizer allocation model. 5. The present fertilizer allocation model does not explicitly incorporate farmer's constraints and their potential impact on the farmer's decision on how to allocate the limited quantities of fertilizer among various crops and fruit trees. In order to include such considerations in future modifications of the allocation model, detailed estimates of whole-farm budgets would be needed. These estimates would allow the construction of several model farm budgets, or farm modules, representing the main farming systems in Syria. As a first step, the model would solve for the fertilizer allocation problem at the farm level, for each farm type. This would be followed by the aggregation of all farm modules into a national model that would provide a solution to the allocation problem, given the constraints on fertilizer supplies at the national level. 295 6. Finally, an important weakness in the present allocation model is the lack of accurate information on the percentage of total cropped areas that are actually fertilized. As discussed in chapter 5, assumptions were made for each crop (in each zone) about the percentage of total area actually fertilized. These assumptions were based on very rough estimates provided by agronomists from the Soils Directorate. More accurate estimates could be obtained from a survey of a representative sample of farmers in the main agricultural regions in Syria. Such a survey could be repeated every few years to get a clearer idea about potential trends in fertilizer use. Based on these trends, specific growth rates for each crop (in each agro-climatic region) could be estimated and incorporated into the procedure for estimating total fertilizer requirements. This would replace the arbitrary 20% annual growth rate in fertilizer consumption for all crops, which is currently used in estimating national fertilizer requirements. APPENDICES APPENDIX A LAND USE AND CROP NIX APPENDIX A 296 .94..» 36:2, .333 ”common .e-mm aausoaaoaue< oo\oo.~ up» no vc.-a .voee-am \a moo.oos.n unu..co.c one.~n. noo.ooo «ou.-s o~..ouo.~ ~n~...~.u o~..ooo “neon «no.0na «a..ane aco.~ aon.o ouo.n~ on¢.oo~ oun.occ .nn.o- .ooua asses «do..no.n ana.oe~.c u«...~. ~o..oao one.aca oo..nnc.a nua.nn~ ooo.~oa .mouo .ooxoosa oon.~no.c co..cea.n n~o.eon on~.omn moo..sc osn.amu.a uoo.no~.~ oon.e.o deuce coo.o~a -n.ooo a...“ ado.o «so.n~ nno.o.~ nec.n~. .nn.o~« segue eases ooa..oo.n nao.eas.n oeo.non oha.oon .«a..ac ccn.ooo.a onn..na auo.o~a cacao ooxoooa ono.oac.c on~.aco.n ..u.aoa «so.~o. ouc..on asa.ona.s as...o~.a 0.5.... deuce .ca...o oo..oon eon.u u~o.o aao.~u no..o~m on...~. noo.no~ .ooue asses .n~.~na.n on».ono.n o....ou ooo.ca. unn.~o. .o«.ono.m oaa.oo~ .o..n.o cacao oo\aooa ano.~a~.. ooo..on.n oos.~n~ oo..nan .oo.... .ma.nna.a nn...nm.— coo.aoa deuce oom.nno o-.onn can.“ so..~ ooo.o~ o~a..~m o~o.~o. nus.n. ...ue oases one.-o.n ana.n...~ cm..on« oo..non oom.o~. noo.o~o.m as..ana nu.._ao a2o aoxoooa o...ono.. “on.aon.n .m..~.~ an..asn no~.—a. .nn.ooo.~ moo.nc~.a “...".a «each ona.cmo ncn..«n o~n.~ mac.» oao.oa ~c..-m ao...on n~¢.no ...up cause ooo.nnc.n on~.oaa.~ non.o.~ o.n.non "an.~on uaa.n.o «oa.cna unn.sn. cacao oo\nooa ..o.ooa.. oso.~nn.n can.aa~ ....oo. soo.n.. n05.5ao.a .Hu.no«.m ....aoa deuce «an.oon no..non coo no... on..." 5.".ao« coo..an oaa.ue .oouu amass o~n.¢mn.n non..~o.~ ooo.o- can.oo. on~.a~. one.oao an..ooa ouo.coo cacao noxcoou 430 .50qu 000.00 000.00 -- 000 000.0 .00.00 000.0 000..0 0.00.0 0e000uo 0.0.00 0...00 -- -- -- 00..0 0.0.00 000.. 000.» 0:00.00 000.0. 000... -- -- 000.0 .00.00 .00... 000.0 ..as0oa 000 a... 0.000 000.00 000.00 -- -- .00.. 000.0 00..0 -- :00.» 0.0000 000.00 000.00 -- -- 00... 000.. 000.0 00 000.> 0e00aaau .0..00 00...0 -- -- -- 000.0 .00.00 000.0 euo0 0e00.3000 000.000 000.000 -- -- 00..0 000.00 000.000 000.00 ...:0.0 .00 0°00 00000 000.. -- -- -- -- -- -- 000.. 0:000 0000000 nun 000 -- -- -- -- -- -- 000 00.0 .ma 000.00 000.. -- -- -- 000 000.0 000.0 0:0.0 0.000 0mm 000..0 000..0 -- -- 000 000.0. 000.00 000 00oea0006 00..000 000.000 -- -- 000.0 000.00 000.000 00. 0000c.a .00.000.. 0.0.000.0 00...0. 000..00 0.0.000 000.000.0 000.00. .00.0.0 000.0.0 00000 00..00 00... -- -- -- -- 00... 000 .500000 000... 00 -- -- -- -- 00 000.00 .000: 3000.. .00.0.0.0 0.0.000.0 000.00. 000.000 0.0.000 000.0.0 000.0. .00.. 0.0000 000.0.0 .00.000 000..0 000.00 000.0.0 000..0. 0.0.000 00..00 00.33 00000 000.000 000.... -- -- -- 000.000 000..00 00..0.0 00.0: s»: 0000. 0.000 0 .co0 . .co0 0 0:00 0 0:00 0 .eo0 0.0.0.000 aouu 00<¢e 00cc» uaaaaaaum aeusuasodue< ha .veusueu .030 883 0.0.0.0 .0250 "03.8.: Bug 5 ”06 3.1.. APPENDIX A 298 00.00000000 000..00.0 000.000.. 000.00. 00..000 000.0.0 000.00..0 000.000 000.000 ..0000 0.00:. 0.000 000.00 .00.0 -- -- 000.0 0.0.0 .00.0 0.0.0 .0000 .0.02 .00.000 0.0.00 -- -- -- 0.0.00 0.0.00 000..00 ..00.0...> 0.000 00..00 0.0.00 -- -- -- 000.. .00.00 000... ..00.0.0.> .o.0s 000.0 000 -- -- -- -- 000 000.0 0000.0 .00 000.00 000.00 -- -- -- .00..0 000.00 000.0 0.0.32.0: 0 000.30...m 000.00 000.0 -- -- -- 000 000.0 000.00 00.300 000.00 000.0 -- -- -- -- 000.0 000.00 00.000 000.00 000.00 -- -- 000.. 000.00 000.00 .00.0. .0006 000 0.000 000.00 000.00 -- -- 000.. 000..0 0.0.0 000.0 .a...0 000.00 -- -- -- -- -- -- 000.00 .000... 000.0 000.0 -- -- -- 000.0 000.0 000.0 0.3000050 0.00-.0 000.. -- -- -- -- -- -- 000.. 0.3000050 000 .0..00 -- -- -- -- -- -- .0..00 0....00 000.000 000.00 -- -- -- -- 000.00 000.000 .0000 0.000.0000 0.000 000.00 000.00 -- -- -- -- 000.00 000.. 000.000 000.00 -- -- -- -- -- -- 000.00 ...: 0.030 000..00 -- -- -- -- -- -- 000..00 eouaou 0.000 0.000 0 .000 . .000 0 .000 0 .000 0 .000 0.0.00000 0006 00<0u .0030 00003.3 0.03.00.30.02 ha 600000.0- 4 5.0-Ive» u.< Odin APPENDIX A 2S99 .luouou :i0uauu< val ousaaso0uu< uo mus-«a0: “couaom ..03.00.>. no: .0 one. 55 00.0» 00:00 voua0ou a» cocaiaa acou- uo clouucoun 05h . ..0coaou0sv.u 0.0000000u nc0uql0uno :0 oeuv- ouo cuacuao oca-.0000 we .5 000.0 vcsouc £908! on \0 0.0.000.0 NNN.coo.c ono.Nnc “00.000 Noo.-n n~v.ouo.u NnN.co~.~ Ouc.ooo Aud0 A