OPTIMAL LAND AND WATER USE AND PRODUCTION RESPONSE UNDER ALTERNATIVE TECHNOLOGIES 1 IN BANGLADESH -A PROGRAMMING APPROACH - r Dissertation for the Degree Of Ph. D. ° MICHIGAN STATE UNIVERSITY MOHAMMAD FAISAL 1977 ‘ 3- guy LIBRAR Y “ m: .13- I In.“ 3.72:1 State I* U": r-‘,o-A;I.' . i. k 7-! ‘3 " £$$VLL3Avf 1' This is to certify that the thesis entitled Optimal Land and Water Use and Production Response Under Alternative Technologies in Bangladesh -- A Programming Approach presented by Mohammad Faisal has been accepted towards fulfillment (7 , of the requirements for 1/7 P _degree in ' "43 ~ AZW / ' ' ,. gar-mm» as,» ‘ /,7 21/ 1' I A /" 7 C (,’ ’ ' Vf/ .I Lu .. ['7 Major professor Date M474” W77 0-7639 -..4-.--_,. _, i (Lava-eso‘B‘o/s ? ABSTRACT OPTIMAL LAND AND WATER USE AND PRODUCTION RESPONSE UNDER ALTERNATIVE TECHNOLOGIES IN BANGLADESH — A PROGRAMMING APPROACH - By Mohammad Faisal The agriculture sector of Bangladesh consists of complex interactions related to regional production possibilities, and comparative advantage in crop alterna- tives. Consequently, an improved allocation of productive resources in agriculture requires continuous analysis of new regional possibilities and interregional relationships. The aim of this research was to investigate the efficient allocation of productive resources and to determine the optimal land use and production pattern in each region of Bangladesh under two technological levels, existing agri- cultural technology and technology projected for 1985. Specifically, the objectives were: 1) to evaluate and compare the effects of alterna- tive technological adjustments on net national farm income, employment, and resource use; Mohammad Faisal 2) to determine improved land use and production patterns for the four regions of Bangladesh with particular attention to food grains and commercial crops (e.g., jute and sugarcane); 3) to examine the effects of alternative govern- ment price and input distribution and policies on land use, farm income, and employment; and lastly, 4) to estimate the supply response for rice under two technological levels. The analysis was carried out on two technological levels, within four regions, which were defined as (i) Transitional agricultUre (Technology level-I) -- a level which uses relatively small amounts of improved inputs and modern technology. approximately the present technology level. (ii) Improved agriculture (Technology- II) -- this level of technology was defined by input- output coefficients which were changed significantly from their present values and include substantial increased use of HYV Aman and increased amounts of chemical fertilizers. The two technological levels represent two plan periods used in the analysis - viz the Technology-I as Plan I (1976) and Technology-II as Plan II (l985). The research finding that the farmers' profita- bility increases significantly, as a result of the Mohammad Faisal introduction of HYV Aman, reinforces the possibility of success of a government policy to encourage expansion of areas under this crop. As the existing policies with respect to supply in inputs (e.g., fertilizer, water, etc.) are mainly geared to the Boro season, a similar sort of provision for HYV Aman will necessitate the reorienta- tion of these policies. All the programming results under alternative government policies and technological constraints retained rice production at a very high level. Therefore, high pro: fitability in rice production indicates a substantial scope for reduction of input subsidies involved in irriga- tion water, fertilizers, and pesticides with little signi- ficant adverse effects on farmers' incentives. The existing support price appears high relative to the prices of other competitive crops. Currently rice support prices appear fixed at a level where supply re- sponse of rice is inelastic. 0n the other hand, these high levels have severe detrimental impacts on production incentives for jute, sugarcane, and wheat. The finding of this research indicates that the existing support price for rice could probably be reduced by one third without appreciable adverse effects on rice production. The breakthrough in seed technology for HYV Aus rice has increased considerably the competitive strength Mohammad Faisal of Aus rice relative to jute. With high yield potential of HYV Aus rice, the jute acreage will likely be further adversely affected, unless the relative price of jute is adjusted proportionately to the price of rice to maintain its competitive position. OPTIMAL LAND AND WATER USE AND PRODUCTION RESPONSE UNDER ALTERNATIVE TECHNOLOGIES IN BANGLADESH - A PROGRAMMING APPROACH - By Mohammad Faisal A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Economics 1977 GIG IQCICILI Dedicated to my grandfather (Mr. Mohammad Ansar Khan) and mother (Mst. Zabun Nessa Khanum) ii ACKNOWLEDGMENTS The author wishes to express his sincere appre- ciation and gratitude to several persons who have made significant contributions to this dissertation. The re- search and thesis preparation benefited greatly from the insights, suggestions and constructive criticisms of my thesis advisor and chairman of the Guidance Committee, Dr. Robert D. Stevens, and the other committee members, Dr. Larry J. Connor, Dr. L. V. Manderscheid, Dr. J. Ferris, and Dr. Milton Taylor. The author is especially indebted to Dr. Robert D. Stevens. His continuous guidance, en— couragement and confidence in me played a significant part throughout my entire graduate program at the Michigan State University. The author is also grateful to the Ford Foundation for general financial support for conducting graduate work and officials of Bangladesh Government such as, Mr. Salauddin Ahmed, Mr. Anisuzzman and Mr. A, K. M. Obiadullah Khan for helping at various stages of this study. The author is extremely grateful to Dr. Robert Picciotto, Director, World Bank, and Dr. Risto Harma, Senior Economist, World Bank, for their continuous assist- ance and insights throughout the study. Special thanks is expressed to Mrs. Torres for typing and assembling the final manuscript. Without the background provided by my grandfather and parents, and the love, encouragement, and support of my wife, Bela, and son, Faiaz, this endeavor would have remained undone. Their sacrifices are deeply appreciated. Finally, I wish to express my gratitude to "ALLAH", The Almighty, for the gift of life and all His help in the world. iv DEDICATION TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES . LIST OF FIGURES CHAPTER I II INTRODUCTION Problem Situation . . Objectives of the Study Scope of the Study . . Background of Research . . Technological Change and Theory of Farm Level Adjustment . . . The Analytical Framework of Enterprise Combination Between Farm- Firm and Regions Analytical Tools for Regional Analysis . Limitations of the Programming Model Models to Deal with Aggregation Bias in Regional Model . SALIENT FEATURES OF AGRICULTURE AND ITS POTENTIAL FOR FUTURE DEVELOPMENT IN BANGLADESH. . A. Topography and Soils Climate and Rainfall Farm Structure and Tenure . . Land Use and Cropping Intensity Cropping Pattern and Relative Importance of Crops Potential for Agricultural Development. . Traditional Technology . . . New Technology and Seed- Fertilizer Revolution . . . Regional Land Suitable to New Seed- Fertilizer Technology as Determined by Soil Reconnaissance Survey . . . . . . . . . . Page ii iii ix xii d .u—a N—iSON 20 25 28 29 31 32 35 37 38 4O 41 46 46 48 CHAPTER Page 111 FRAMEWORK AND STRUCTURE OF MACRO MODEL . . 55 The Objective Function . . . . . . . . 57 The Activity Set . . . . . . . . . 59 The Production Constraints . . . . . . 66 IV GENERATION OF INPUT-OUTPUT COEFFICIENT . . 8l The Representative Farm . . . . . . . 83 Land Input . . . . . . . 84 Fertilizer Input and Yield . . . . . 84 Labor Input Requirement . . . . . . 9l Bullock Power Requirement . . . . . 94 Water Input Coefficients . . . . . . 97 Credit Requirement . . . . . . . . . 98 Factor and Production Prices and Cost of Production . . . . . . . . . l00 V OPTIMUM CROPPING ORGANIZATION UNDER ALTERNATIVE TECHNOLOGIES: IMPLICATIONS FOR FARM INCOME. LAND USE AND RESOURCE UTILIZATION . . . . . . . . . . . 105 Optimum Organization With Existing Resource Constraints Under Plan I (Technology-I) . . . . . . . l06 Net Returns to Fixed Factors in Farming . . . . . l06 Optimal Regional Land Use and Cropping Pattern . . . . . . llO Utilization and Marginal Value Product of Resources . . . . l16 Optimal Organization With Techno- logical Change in Plan II (Technology- II) . . . . . . . . . lZl Net Returns to Fixed Farm Resources . . . . . 124 Utilization and. Marginal Value Productivity of Resources . . . . . 131 VI SUPPLY RESPONSE AND IMPACTS OF ALTERNATIVE POLICIES ON LAND USE. PRODUCTION PATTERN AND NET NATIONAL INCOME . . . . . . . . . . . 136 A. Supply Response of Rice . . . . . . . l38 Analytical Procedure . . . . . . . . . l39 vi CHAPTER Optimum Organization Under Various Price Combinations of Rice Under Technology I and II (Plan I and II). . Results of Technology- I (Plan- I). Results of Technology- II (Plan- 11): B. Alternative Policies With Respect to Fertilizer and Irrigation Water Constraints . Policy Alternatives Considered Analytical Procedure Land Use Pattern and Net National Farm Income Under Alternative Policies Under Plan I and Plan II C. Implications of Alternative Yield Assumptions on Production Pattern and Net National Farm Income. . . . . . . . . VII SUMMARY, POLICY IMPLICATIONS AND CONCLUSIONS . . . . . Summaries of Background and Methodology . Limitations of the Analytical Procedure . . . The Optimal Land USe and. Production Pattern and Its Impact on Net National Farm Income . A. Results of Plan I B. Results of Plan II Impact of Alternative Rice Price Policies on Supply Response of Rice in Plan I and Plan II A. Plan I - Rice Suppiy Response B. Plan II - Rice Supply Response Results and Impacts of Alterna- tive Fertilizer and Irrigation Water Policies on Land Use Pattern vii Page I40 I40 I45 I48 I49 ISI I51 I62 I72 I72 I77 I80 180 I82 I84 I84 I85 186 CHAPTER A. Results of Plan I B. Results of Plan II. Results and Implications of Alternative Yield Assumption On Production Pattern in Plan II . . . Policy Implications and Conclusions . Further Research APPENDIX A Estimated Per Acre Growing and Harvesting Cost for Different Crops Used in the Programming Model (Technology I - Plan I). . . . B Estimated Per Acre Growing and Harvesting Cost for Different Crops Used in the Programming Model (Technology I - Plan II). . . REFERENCES viii Page 186 . 187 . I88 189 191 I93 . I94 . 195 Table 1.1 LIST OF TABLES Per Capita Availability of food grains in Bangladesh . . The Wholesale Prices of Different Agri- cultural Commodities . Comparative Statement of 5 years Average Acreage, Production and Per Acre Yield of Different Agricultural Commodities Present Adaptation of Cropping Pattern to Land and Soil in Bangladesh. . Land Utilization in Bangladesh Land Use (Average of 1969- 1973) of Bangladesh . Area Under Main Crops in Bangladesh Land Capability and Improved Varieties Summary Estimates of Acreages Suitable for Specified Crops by Region Schematic Representation of LP Tableau of the Model. Estimated Regionwise Fertilizer Availa- bility During Phases I and II Hypothetical Long- Range Water Develop- ment by Region . . Projected Irrigation Availability During Two Phases of Development . . . Estimated Subsistence Consumption Requirement of Farm Population During l975, 1980 and 1985 . . . . . ix Page 36 39 42 45 50 54 61 70 74 76 79 Table 4.l Estimated Regression Coefficients for Fertilizer . Estimated Fertilizers Input-Requirement for Different Crops in Alternative Stage of Development . Estimated Yield Projection for Alter- native Phases of Development . Estimated Monthly Labor Requirements for Various Crops. Estimated Bullock Power Requirements for Different Crop Activities in Bangladesh Estimated Irrigation Requirement Per Acre in the Low, Medium and High Rainfall Regions in Bangladesh . Characteristics of Optimal Organization of National Agricultural Plan for Bangladesh - Technology-I (Plan 1) Total Regional Land Use Pattern of Food Grains and Commercial Crops in Bangladesh Existing and Optimum Land Use Pattern Under Tech-I in Four Regions of Bangladesh . . . . . . . Resource Availability Levels and Their Utilization Under Optimum Plan Technology I . The Characteristics of Optimal Organi- zation of National Agricultural Plan Under Technology II (Plan II) Existing and Optimum Land Use Pattern Under Technology II in Bangladesh Existing and Optimal Regional Land Use Pattern of HYV's Under Tech II Page 90 92 93 95 96 99 108 111 113 117 123 125 128 Table 5.8 Proportion of Suitable Land Identified by Soil Reconnaissance Survey Uti- lized in the Optimum Plan for Tech I and II . Resource Availability Levels and Their Utilization Under Optimum Plan - Tech II . . . . Results of Variable Rice Price Program- ming Under Tech-I in Bangladesh Results of Variable Rice Price Program-r ming Under Tech—II in Bangladesh. Comparative Statement of Optimum Land Use Pattern for Different Agricultural Crops Under Alternative Resource Combinations Assumptions in Technology I. . . . . . . . . . Optimum Credit Requirement and Marginal Value Product of Resources Under Alternative Policy Alternatives in Plan I in Bangladesh Comparative Statement of Optimum Land Use Pattern for Different Agricul- tural Crops Under Alternative Resource Combinations Assumptions in Technology II. Optimum Credit Requirement and Marginal Value Product of Resources Under Alternative Policy Alternatives in Plan II (Tech II). Alternative Yield Assumptions for Sensitivity Analysis Under Technology II . . . . Production of Different Crops Under Alternative Assumptions in Plan 11 xi Page 130 132 142 I47 153 158 161 163 165 170 LIST OF FIGURES Figure/Graph Figure I II III IV Graph I II Production Functions of Fertilizer With and Without Technological Change . . Optimum Value Product and Costs With Changed Fertilizer Prices . . Best Combination of Resources Optimum Resource Utilization and Product Combination in-between Regions Supply Response for Rice Under Different Technologies in Bangladesh . . . . . The Impact of Resource Expansion on Net National Farm Income xii Page 19 21 23 26 144 157 CHAPTER I INTRODUCTION Precise objectives of development planning differ from country to country and over time. In a general sense, however, all seek to promote human welfare through econo- mic growth. Planning attempts to connect ends and means. In economic terms the efforts at development are designed to optimize use of existing resources among alternative means to achieve maximum production over a specified time period. In LDC's the knowledge of strategic parameters (e.g. relations between production inputs, soil, and output, etc.) associated with transformation of the existing pro- duction structure need to be delineated clearly so as to achieve optimum resource utilization and enhancement of farm and national income. Reliable economic analysis of agricultural sector in LDC's is confounded by the complexities due to the existence of: (1) old technological structure; (ii) inade— quate supply of higher yielding inputs (e.g. seeds, ferti- lizer, water, etc.); (iii) inflexibility of the production structure; (iv) high risk and uncertainty surrounding 2 farmers' decision environment; and (v) lack of emphasis on scientific investigation and economic research. Problem Situation Over the last two decades total food consumption has doubled in Bangladesh. This has been due to an unpre- cedented population upsurge at the rate of 3.1 percent per annum. Food production, however, has failed to keep pace with population growth. Rice production increased from 9.52 million tons in 1960-61 to 10.96 in 1970-71, an annual growth rate of 1.2 percent ( 44) and declined to 10.4 million ton during 1970-75 (5 years average). As a result of the widening gap between supply and demand, the food prices soared up to an average of Rs. 3264 per ton in 1970-75 as against Rs. 880 per ton in 1965, an increase of 271 percent due partly to inflation. To meet increasing demand, the government of Bang- ladesh has laid tremendous emphasis on enhancing food pro- duction. Government policies have consisted of: (a) Input subsidies -- To accelerate the use of a substantial quantum of required production inputs by farmers, the government used high subsidy rates -- ranged from 40 percent and 60 percent for fertilizers and water respectively, anleO percent for pesticides. High produc- tion incentives were sought particularly for high yielding varieties. (b) Support price -- Provision for dual procure- ment prices for food grains was made so that a higher price could be paid for the high yielding varieties than for local varieties. (c) Supervised credit -- Increased flows of credit have been provided through co-ops. An appropriate supervision and guidance was attempted to assure its pro- per utilization. (d) Research and extension —- Crop Research Institutes and agricultural extension services have been strengthened. In addition, large amounts of high yielding seeds were made available to farmers. Though this intensive production effort has con- tinued over the last decade and produced certain achievements, it has posed certain serious problems to decision makers: (i) In spite of large input subsidies and high procurement prices, food production has lagged conti- nuously behind population growth. As a result, the per capita availability of food grains has declined over the years (Table 1.1). (ii) Secondly, the relative increases in food grain prices in comparison to other competing crop prices has resulted in continuous transfer of lands to rice crops, thereby leaving a smaller percentage of total land ¢.NP m.m_ _~.mn omm._F mum.m Npm.m murmnmP P.N_ o.m_ mm.mn _.¢.FF wmm.~ m_o.m Nurpum_ _.m_ N.¢_ um.mn omm.o_ oom._ omo.m FNIONmF m.cp a.m~ ¢.NN opp.mp oeo._ omm.op ourmoop «.mp F.m_ ~.o5 oom.o_ omm._ omm.m marmom_ m.mp ¢.¢P o.mo omo.op omo.— om~.m mmruomp w.~_ m.qp ~.oo omm.m oom._ o-.m umrmomp ¢.q_ m.m_ m.¢o ope.op onm. oFm.m mmrmomp m.<~ N.o_ m.Fm omp.op com. om~.m moroomp u o m e m N _ uganF “Lone? paonuwz gap; xov\:omcma\muc:o Acowppws cwv :owuqszmcoo Lo» mucoasH cowpuauoca mcmm> xpv__nm_wo>o cowpopaaoa zuwpwampwm>m pmz Pouch upummEoc upwamu com umz Amcop cowPFPE cpv mcwmgw uoou mo zuw—wnm—ww>< muwamu Lam .—.P wpnmh for the other food and cash crops. Rice prices show an increase of 196 percent from 1964 to 1970, whereas price increases of other crops only ranged from 5 to 20 percent (Table 1.2). In the case of potatoes, the price has declined by 60 percent during the period under comparison. Due to low prices of other crops relative to rice at the farm level, the acreages of sugarcane and tobacco have declined from 0.42 million acres in 1965-70 and .131 million acres in 1950-55 (5 years average) to 0.36 million acres and .110 million acreage in 1970-75 res- pectively; whereas the areas under rice crop have increased by 3 million acreage (a 15 percent increase) during the same period. (iii) Thirdly, the great increase of food grain prices has caused farmers to become less careful about topographic and hydrological characteristics of land with respect to its suitability for food grain produc- tion vis-a-vis other alternative crops. This has two dampening and vulnerable effects on production: V (a) As lands are being transferred to food grains, only marginal lands which are inferior in quality are being left for other crop production. Consequently, the Marginal Physical Product (MPP) of these crops have shown a continuous declining trend. For example, Table 1.3 indicates that the Table 1.2. The Wholesale Prices of Agricultural Commodities Different Rupees/md Rice2 Potato Jute3 Tobacco Sugar (Course) 1964 26.38 31.51 22.52 107.92 74.77 1965 32.05 29.52 31.47 119.38 85.32 1966 41.10 23.13 27.39 99.17 69.87 1967 42.20 25.32 31.0 103.42 67.14 1968 47.66 23.09 26.0 158.77 84.42 1969 67.64 24.15 26.0 174.59 90.00 1970 78.10 18.50 27.0 110.50 90.00 Percentage inc.(+) or dec.(-) over 1964 196 -41.3 19.8 2.4 20.4 1975*3 120.0 45.0 100 N.A. 154.5 (72-73) *The price series from 1971 to 1974 are omitted as the data of these years do not represent a meaningful indi- cator for economic analysis. The reasons for not incor- porating the data for these years are: (1) War during 1971-72; (ii) unstable political situation; (iii) exist- ence of rampant inflation, etc. According to experts of World Bank and USAID, etc., economy appears to have stabilized during later part of 1974 and beginning of 1975. For this reason, price data for 1975 is included. Source: 1Directorate of Agricultural Marketing. Ministry of Agriculture. 2Wholesale prices of rice in E. Pakistan - Pak., p. 4, Bangladesh Bureau of Ag. Statistics. Quarterly Eco. Indicator for E. 3 of Agriculture, Bangladesh. Bangladesh Agriculture Statistics, Ministry Auzoom ooaav =aom.xoa co .o.>oo .oxao: ac. .oc co maoo.=_x .oo.oo.ooom .m< .gmuuapmcom mo .a.>oo .mu.um_uoum .a< mo anucac .opnu—va>u we: a .<.z m w .caumvxaa .u mo .u.>au we «mom_a poo—umwuoum ace coum.xaa mo .u.>oo .acu .capa a =0-uuum .p.gm< noon. "augaom ~.~_ o~.o_ sm.m~ ~.~ msw.m ~mm.. msrcua— ~.~_ v~.cp ca.n~ m.~ cum.o cm~.~ osrmmmp w.m om.~ mm.c~ m.n sme.m pwm._ Ao>o menu» av mmromm_ uu—onxoguo Acowp—vsv vcaoe\mguu mcou copp—_e Acopp—pa. u—o.» mm—oa A=0+-—.sv can u—o—> co_uu=coga mocu< «sun so; co.uu:uccm mogu< nu u _ a ~u h = a ass ~.om a... v.5— o—n.o mm.o msrcNm— pmu m.om opp. a. mum.~ ~o.° curmma_ _.m “.mm _m_. .<.z .<.z .<.= Aus. «Loo» my mmromm— mn—\u—awa on. :o_p—.a Acc.—__I. Amcoavupupa Amcau co.—__Iv Ago—p__av moon sonI :o—uusuoca acu< «can so; covuusuoLa mowwn Lou» o u u < m c h mu 2 < u a < a = m uo—upeoesou posse—au.co< acocouupa uo v.0.» ugu< so; we. :o_uu:aoca monocu< maucu>< meow» m mo «casuuaum o>muncaasou .m.— m_ao» per acre yield of jute has decreased from 3.6 bales/acre in 1950-55 to 2.9 bales/acre in 1965-70 and to 2.7 bales/acre in 1970-75. Similarly, the tobacco yield declined from 811 lbs/acre in 1950-55 to 770 lbs/acre in 1970-75. The per acre yield of sugarcane has declined from 18 tons in 1965-69 to 17.4 in 1970-75. (b) Due to the increased use of land, a signifi- cant portion of land (nearly 10% of rice land) which is at present being used for food grain production is located in low areas subject to recurring floods every year. It is estimated that since 1964-65, on an average, about 0.8 to 1.0 million tons (10) of rice as against the actual average imported requirement of 1.5 million tons, have been destroyed by floods annually. (iv) Fourthly, in spite of increasing emphasis on rice production and large transfer payments to farmers in the form of input subsidies, it was found that real per capita rural income declined from Rs. 275 in the early 1950's to Rs. 268 in the 1960's. Over the same period, the urban income rose from Rs. 619 to Rs. 677 per annum(3). The problems of Bangladesh agriculture as seen by the Annual Plan for 1974-75 were: (1) How to increase production of cereals (e.g., rice and wheat) to the maximum extent from the minimum acreage of land; (ii) How to enhance production of noncereals and important cash crops like jute, tobacco, pulses, sugarcane; (iii) How to release land from rice to other cash and food crops production so as to enable the country to earn foreign exchange needed for economic development and bridge the nutritional gap of the population; (iv) How to improve the allocative efficiency and optimum use of economic inputs, e.g., ferti- lizer, water, pesticides, and land; and above all, (v) How to increase farm profit in order to pro- vide sufficient production incentives to farmers. Objectives of the Study The general purpose of this study is to develop an empirical base for analyzing alternative agricultural poli- cies for Bangladesh. The investigation centers on four specific objectives. They are: 10 (i) To evaluate the effects of alternative levels of agricultural technology on net national farm income, employment, and resource use; (ii) To determine optimal land use and production patterns for the four regions of Bangladesh with particular attention to food grains and commer- cial crops (e.g., jute and sugarcane); (iii) To examine the effects of alternative government price and input distribution policies on land use, farm income, and employment; and lastly, (iv) To estimate the supply response for rice under two levels of agricultural technology. It is important to clarify some terminology and concepts used here, which might confuse the reader in understanding this analysis. (a) Net National Farm Income (NNFI): is defined as the gross revenue received from the production of eleven commodities in the nation minus the variable cost incurred from the production of these commodities. (b) Employment: is related to the total farm labor force (both family labor and hired labor) that is absorbed in the production of eleven crop activi- ties studied. (c) Resource use: includes both fixed and variable in- puts which are limited in supply and act as 11 constraints on production activities. The con- straining resources considered in the study were: agricultural land, family labor, hired labor, bullock power, irrigation water, fertilizer and credit. (d) The optimum allocation of land: this condition has been defined as one, that with given physical and technical resource constraints , indicates which crop activities would be undertaken and how much land would be allocated to each crop activity so that net national farm returns are maximized in an annual cycle. (e) Procurement price: refers to the government fixed prices for rice, wheat and sugarcane whereas farm gate prices are used for potato and jute. Scope of the Study The entire area of Bangladesh is included in this investigation. The analysis concentrates on two technolo- gical levels of farming within the four regions of Bangla- desh identified by the soil reconnaissance survey project. The regions are: (i) central; (ii) eastern; (iii) north- west; and (iv) southwest. The two levels of agricultural technology used in this study are defined as follows: 12 (i) Transitional Agriculture -- Technology Level-I The level which uses small amounts of improved inputs and modern technology. The existing transi- tional system of operation is regarded as technolo- gical level-l. This level of technology is included in Plan I. (ii) Improved Agriculture -- Technological Level-II The second level of technology is defined by input-output coefficients which have changed signifi- cantly from their existing values and include use of larger quantities of high yielding (HYV) Aman seeds and increased use of chemical fertilizers as projected for 1985 (Plan II). Theoretical Background of the Research The agricultural sector consists of a series of complex interactions relating to regional production possi- bilities, comparative advantage, competitive crop alterna- tives and a marketing system that is both uncertain and complex. The importance of this sector stems largely from its contribution to: (i) the domestic food supply; (ii) the supply of domestic savings needed for investment; and (iii) the supply of foreign exchange required for im- portation of capital goods. The failure of Bangladesh agriculture to produce adequate requirements of food grains for the entire population, and to generate l3 increasing rates of savings and foreign exchange earning capacity may result from: (i) underallocation of resources to the agriculture sector; (ii) inhibition of free play of market forces due to regulation and administrative action or other constraints; and (iii) misallocation of resources associated with inadequate technical information required for optimal use of resources. In agriculture production typically takes place on multi-product farms where several activities compete for common resources. This suggests that it is not possible to regard supply response to prices, costs and other variables of one crop as being independent of the response to other crops grown on the same farm. Besides, regional characteristics influence the predominant size of the pro- duction units, the methods of production that can be used and crops or livestock alternatives that can be considered. The regional characteristics are controlled by such factors as the existing climate, topography and hydrological con- dition available to the farm or region. Specialization resulting from regional comparative advantage is a function of relative yield as affected by soils and climate, regional resource supplies and the relative profitability of dif- ferent crop and livestock enterprises for the individual farmer. Due to the presence of these technical relation- ships, it is found that a crop suitable for a particular 14 region or farm may be less productive in other areas. Allocation of production inputs by government agencies which disregard these regional technical characteristics are likely to cause inefficient use of resources, which, in turn, results in less production and increased costs. In the past, very little technical information has been available to the agricultural policy decision makers of Bangladesh. Consequently, the growing of a particular crop in a region as well as distribution of production inputs among alternative crops through governmental chan- nels was primarily a function of administrative and poli- tical choices. As a result, it is hypothesized that agri- culture has not been able to produce its greatest poten- tial results in spite of vigorous government efforts in continuously augmenting the supply of production inputs. The availability of improved scientific information about technical relationships between crop physiology, topogra- phic and hydrological characteristics of particular regions or soil strata can result in improved allocation of pro- duction inputs, optimum cropping patterns and increased production. The "Soil Reconnaissance Project" of Bangla- desh/FAO/UNDP, which has completed a survey of 80 percent of agricultural land during the last ten years was a step in that direction. The survey provides an improved esti- mate of technical production relationship in different regions in Bangladesh. The extent to which Bangladesh 15 agriculture can satisfy internal food needs and is able to compete in the international market depends to a large extent on the utilization of her productive resources where these can produce highest returns and the extent to which crop and livestock production is distributed among regions according to their comparative advantage. Technological Change and the Theory of Farm Level Adjustment in Bangladesh This theoretical framework shows that farm level adjustment processes are a result of changes in the produc- tion and cost functions under alternative technologies con- sidered in this study. Over the years, the term "technolo- gical change" has been given a wide range of meanings and interpretations. In the field of agriculture in LDC's, development strategy often looks to the "seed-fertilizer revolution" as the primarysource of technical change which includes three key elements: (i) agricultural research leading to the development of high yielding varieties (HYV) which have the capacity to respond to high levels of soil fertility; (ii) greatly increased application of chemical fertilizers; and (iii) activities which promote both the widespread use of new varieties and the associated changes in the farm practices which are needed if these varieties are to realize their high yield potential (29). Recently, the FAO/UNDP soil reconnaissance survey has indicated that 16 about 22 million acres (11) of land can be brought under improved HYV rice varieties of Irri-aman and Irri-Aus/boro in Bangladesh. This new potential means that technical change resulting from the introduction of HYV leads to a production function shift which will enable the agricultu- ral sector to increase output associated with greater use of inputs, particularly fertilizers. Consequently, in this study the introduction of HYV seeds accompanied with in- creased use of chemical fertilizers is regarded as a prime- determinant of technical change in agriculture in Bangladesh. Separate production functions are, therefore, spe- cified with and without technological change based on HYV seeds and chemical fertilizer. In a production function the fixed factors are always specified as representing not only a specific quantity but also a specific quality as well. Hence, whenever either of these two aspects change for one or more of the fixed factors, the production func- tion will shift upward/downward depending on the nature of the change. However, a change in the quantity of variable factor is reflected in movements along the production function (4,6). In Bangladesh, under transitional techno- logical conditions, the output will be a function of land, labor, bullock-power, water, fertilizer, seed and pesti- cides, etc. Under an improved technological situation re- sulting from introduction of HYV seeds, the production 17 function shift upwards enabling greater output to be pro- duced with increased use of other variable inputs. How- ever, in order to avoid data constraints for the purpose of this study, it is assumed that the use of only one variable factor, i.e., fertilizer, would change under the two technological adjustments. The production function with HYV and local seed varieties is specified as follows: Y f(X]/X2,X3,X4,X5,X6,X7) where Y = output X1 = variable quantities of fertilizer, X2 = land, X3 = labor. X4 = bullock power, X5 = seed, traditional or HYV X6 = water, X7 = pesticides. In addition to the specification of the production func- tion, the following assumptions are necessary for the pur- pose of theoretical analysis. First, the levels of inputs per acre other than fertilizer are known for each crop activity; second, prices of inputs and output per unit remain constant; third, the farm operators are profit maxi- mizers and will, therefore, produce at that output level where the marginal factor cost (MFC) of fertilizer equals the marginal value product (MVP) of fertilizer. 18 With the production function under HYV seeds and assumptions as specified above, it is possible to present a theoretical exposition of the farm adjustment process under the technological alternatives considered in this study. Figure 1 illustrates an upward shift in the produc- tion function from TVP0 to TVP1 due to introduction of HYV seeds. The total value product curve is obtained by mul- tiplying total physical product (TPP, i.e., output of crops) by the prices of the product. The two marginal value product (MVP) curves representing two technological situations were obtained from TVP curves (MVPF = 3%3 - P0). The marginal factor cost (MFC) of fertilizer is assumed to be constant. First, notice the impact of a shift in the production function has on the MVP of F in the production of Y. With the old technology, a price of F at Fx, the ideal level of factor use was FO units of fertilizer input with a corresponding output of Y0. With technological change due to the introduction of HYV seeds, at F unit of O fertilizer, the MVP1 > Fx. So, the rational entrepreneur would increase use of F from F0 to F]. This would result in total output Y1 instead of Y3, that would otherwise re- sult if the entrepreneur did not increase the use of F. The TVP curve originates from N and N1 because 0N and 0N1 amounts of value products are assumed to result from natural fertilizer for local and HYV seeds respectively. yI yé/ TAflfi 1'3 I—‘ls 2 V-\ HYV 8 a, TVPo"F ISLOCAL 3 Q 3 0: Ni N \ Fx / M FC 0 F0 F2\\ f:IS MVPO MVPl Fertilizer Doses in Pounds per Acre Figure I. Production Function of Fertilizer With and Without Technological Change 20 Figure II shows that in the case of a farm- adjustment process, a change in fertilizer price will shift the TVC curve upwards/downwards as well as the quantum of fertilizer to be used increase/decrease depending on the nature of the change. Figure II shows the impact of a change in fertili- zer price on the farm adjustment process. As the fertili- zer price increases from Fxo to Fx], the quantum of ferti- lizer used will decline from F0 to F]. The optimum output will change from V0 to Y]. The Analytical Framework of Enterprise Combination Between Farm-Firm and Regions This analytical framework includes the conceptuali- zation of procedures to determine the most profitable re- source allocation when two and more products are being produced on a farm. In the latter part of this section, this theoretical idea is extended beyond micro-production economics to demonstrate the theoretical rules of enter- prise combination between regions. This is particularly relevant when the national planning objective is to ensure regional specialization in order to derive maximum output from a limited amount of resources. The search for optimum enterprise combinations is necessary to determine the best combination of products to obtain maximum use for a given outlay of resources or the 21 y yo 8 Y. if: Q. a) 0 O. O. = 0: IS: MFC I I Xofl’,/’r 0 F, F0 \ F MVP' Fertilizer Used in Pounds per Acre Figure 11. Optimum Value Product and Costs With Changed Fertilizer Prices 22 best use of resources for a given combination of products within or in between farms or regions. Assume that a farm or region in Bangladesh produces two crops, rice (Y1) and jute (Y2), with their production functions specified as follows: Y] = (x1 ... xd/xd+], ... xg/lxg+] ... xn) Y2 = (x1 ... xd/xd+], ... xg/lxr+] ... xt) where: x1 ... xd = inputs which variable for the farm as a whole as well as each enterprise such as fertili- zer, etc. Xd+1 x9 = resources fixed to the firm but allocable between Y1 and Y2 such as land, plough and family labor. xg+1 xn = inputs fixed in the production of rice xr+1 xt = inputs fixed in the production of jute. These hypothetically specified production functions for rice and jute are shown in graphical form in Figures 3(a) and 3(b). Assume that the farm has a given amount of resources allotted to variable inputs x1 ... Xd which would enable the farm to produce: (a) only rice; (b) only jute or (c) some combination of both. If the entire budget were forced into jute production the farm would be producing at "a" (Figure 2(b)) and no rice. By reducing expenditure on jute to "b" (Figure 2(b)), the farm would be producing F 23 RICE JUTE V, y2 P Lu E / LI L o' b'd d' d c b 0 index x' ....... de xd” ....... x0“ x9.“ ..... xn index XI"..de xd‘l. ..... xgi xfflumux' (o) (b) VI so-cost curves Best Combination of Resources Product Complimentarity RI CE Product CompetItIon Isa-revenue Product ‘5. Campmnmfimmw JUTE Y2 (c) Figure III 24 quantity of jute and L quantity of rice. Similarly, if the entire resources are spent (d1 in Figure 3(a)) on rice, the farm would be producing aILH quantity of rice and no jute. By continuing the above process it is possible to construct a production possibility or iso-cost curve repre- senting the possible combination of Y1 and Y2 which could be produced for given resource level (shown in Figure 3(c)). On the production possibility curve (PPC), below the point C and to the right of N, the production function would be in stage III, and thereby represent an irrational boundary of production choice. By combining iso-cost and iso- revenue line determines the profitable enterprise combina— tion. For the best enterprise combination the variable resources should be used such that: MVPxi(Y]) = MVPxi(Y2) PXl Pxi and for best level of production with the optimum enter- prise combination, the variable resources should be used such that: MVPxi(Y1) = MMPxi(Y Pxi Pxi 2)_1 The above mentioned idea of micro-production eco- nomics can be used to obtain enterprise combination in between regions. Assuming the production functions per- taining to the production of the two commodities in each region is known, it is possible to derive regional 25 production possibility curves for the two regions. From these production possibility curves best combination of products Y1 and Y2 are determined where the price ratio is tangent to the PPC (shown in figure 4a ). Determina- tion of the combination of production which would yield the maximum total output from the two regions combined is demonstrated in figure 4b. In this diagram the produc- tion curve for region I has been inverted and plotted tangent to the production possibility curve or region II. The point of tangency indicates the combination of pro- duction which would be the maximum of both Y1 and Y2.1 Analytical Tools for Regional Analysis The whole spectrum of Bangladesh agriculture, its present problems and full potential under alternative regional resource constraints, technological change re- sulting from "HYV seed-fertilizer revolution" and its capacity of meeting internal demand for food, etc., needs to be evaluated for formulation of appropriate policy guidelines for future development. Answers to these wide range of problems require an analytical tool that gene- rates both sufficient empirical results at both regional as well as national levels. Detail by region is needed 1For details see G. L. Johnson, Enterprise Combination -- Economics and Management in Agriculture, edited by W. H. Vincent, pp. 82-114. 26 yI (4a) Region-I y't: yl Isa-revenue Reg ion II yz' ya" ya ya"! y' (40) Region-II Best combination of Products y. and y2 yllll _ _ _ _ ______ — yIIu Region-I Y2. - Y2 Figure IV. Optimum Resource Utiliza- tion and Product Combina- tion in-between Regions 27 so that the flexibilities of or restraints on production and resources can be measured and impact can be expressed at the local level. National detail is required so that market impacts on prices can be measured and interdepen- dence among regions in terms of their comparative advan- tage can be established. Thus, if either resource and production potentials under various technologies, resource or production policies are to be evaluated, an analytical tool is needed that allows measures and generates results at the level of both individual regions and the nation as a whole. This type of detail can be provided by a linear programming model incorporating relevant production possi- bilities and various regional constraints under alternative technological adjustments. L.P. is a method of combining resources, whereby it is assumed that the production process can be broken down into a series of straight line relationships. A unique solution is obtained to a set of simultaneous linear equations which represent the productive relation- ships of the various resources. It is a method of finding the most profitable combination of alternative production processes within a given set of restrictions. Linear optimization models in the agriculture sector are expected to attain a double aim: (i) to help agricultural planning by indicating optimal distribution of the available resources between and within geographic regions (branches) 28 and (ii) to make calculations in order to estimate for agriculture the national economic resources it can fruit- fully utilize. Regional resource availabilities can be altered by policies affecting land use, water availabili- ties or transfer and the availability of other endoge- nously allocated inputs such as fertilizer, pesticides, etc. These policy alternatives will be evaluated in this framework by (i) incorporating new restraints necessita- ted by a new policy; (ii) changing the values in "c" vec- tor to reflect changes in the relative activity costs resulting from a different policy or assumptions; (iii) changing the coefficient in the "A" matrix to reflect a changed level of interaction between an activity and any of the relevant markets. Limitations of the ProgrammingyModel Programming techniques represent a useful and versatile method of evaluating agricultural policies. But L.P. has its limitations that restrict the scope of its use and the interpretation of results. Its limitations are inherent in its assumptions which are: (i) additivity and linearity; (ii) divisibility; (iii) finiteness; (iv) single-value expectations; and (v) nonnegative output. 29 Methods to Deal with Aggregation Bias in Regional Model Aggregation error is said to exist when the weighted sum of production of representative farms does not equal to the total production of the region. The sufficient condition for zero aggregation bias are: (i) B], ... 82, ... 8n = 8. Condition (1) states that all firms must have identical matrices of input- output coefficients. (ii) 29 = YgZ, where Yg is a scalor greater than zero for all j. The vector of net returns for every farm has to be proportional to the vector of the aggregate; (iii) C9 = AgC where Ag is a scalor greater than zero and less than one for all g. Vectors of resources of every farm must be proportional to corresponding vector (i.e., proportional variation in constraint vector). The term proportional variation means that if farms differ in one resource, say labor, by a certain ratio, then they must differ in all resources by that ratio. Millerz, on the other hand, in his qualitative homogeneous output vector (QHN) approach indicated that in order to have a bias free aggregation the following conditions need zMiller, T. A. "Sufficient Conditions for Exact Aggregation in Linear Programming Models", Agricultural Economic Research. ERS/USDA, 18: 52-57, April 1, 1966. 30 to be met: (a) all farms must have an identical coefficient matrix and (b) all farms must have the same activities included in the solution. Conventional procedures have invariably used "farm size" as the selection criterion for bias free aggregation. The weakness of this procedure is that it takes into account only one factor of production, namely, land, and ignores all other resources of the farm-firm. In this study both land and labor have been considered as strategic factors for aggregation. The master survey of agriculture (7th edition, 67) indicated that nearly 71 percent of farmers hold, on an average, 3.1 acres of land and only 3 percent hold more than 12 acres. Recent estimates show that nearly 85 percent of the farm fami- lies hold about 3.1 acres of land. With respect to labor, the "Farm Mechanization Survey Report" indicated that 98 percent of the farm families had on an average, about 2.1 adults working population. Judging from these two resource points of view,since 85 percent to 98 percent of farm families possess uniform land and labor resource endowments, it is assumed that their a.-'S would not be 1.1 significantly different from each other. CHAPTER II SALIENT FEATURES OF AGRICULTURE AND ITS POTENTIAL FOR FUTURE DEVELOPMENT IN BANGLADESH The combination of soils, climate and hydrology greatly influence the crop production environment in Bangladesh. The first part of this chapter focuses on the important existing elements of regional diversity in res- pect to topography, soils, climate and hydrology and its impact on land use patterns and cropping intensity, etc. The second part of this chapter deals with the potential for increased crop production with the innovation of new "seed-fertilizer" technology and the availability of de- tailed technical information regarding soil strata, topo- graphy and hydrology in respect of each region through Soil Reconnaissance Survey. This discussion will help in providing a basis as to what extent modern technology under given information would enable Bangladesh agriculture to surmount the natural as well as physical constraints to crop production that it is encountering today and usher a new hope and aspiration in the future. Bangladesh is traditionally and predominantly an agricultural country stretching over 55,000 square miles with a population density of 525 per square mile. It is 31 32 situated on the confluence of three big rivers of the world -- the Ganges, Brahamputra and Megna. A. Topography and Soils The characteristics of topography and soils in Bangladesh have influenced the evolution of the cropping pattern and are important in determining irrigation and other agricultural development possibilities. The alluvial plains of Bangladesh rise with very low gradients of .03'-.06' per mile, with few parts of the country being more than 50 feet above sea level. The only significant elevation, the Sythet hills in the northeast and the chit- tagong hills in the southeast. The terrain is crossed by 15,000 square miles of waterways including major rivers. The overall picture disguises considerable variations in local relief. Complex local differences in elevation are important in planning irrigation and drainage especially as they are associated with important differences in soil per- meability and associated crop suitability. Bangladesh soils have, in general, a high potential for increased crop production. This is particularly due to high natural fertility, as evidenced by the extensive agro- nomic trials which show that virtually all soils respond well to fertilizers. The Soil Reconnaissance Survey (SRS) has identified the main soil tracts in each region including their physiography and important characteristics which are: 33 (1) CENTRAL REGION (a) Northern Piedmont Plains -- are mainly loamy sediments subject to shallow and intermittent flooding and these are adjoined by clay plains or basins subject to deeper flooding. (b) The Old Brahamputra Flood Plain -- the sedi- ments are mainly silty on the ridges and clay in the basins but there are important sandy areas in the west of mymensing and northeast of Dacca. The highest ridges and basin in the west and in Dacca are shallowly flooded in the monsoon area. (c) The Madhupur Tract -- is a heavy slicken sided clay, but gives rise to both loamy and clay soils. They are few valleys which have a general pattern of relatively well drained soils or have the higher edges and of poorly drained or seasonably shallowly flooded soils on slightly lower terrace interiors. (d) The Jamuna Flood Plain -- comprises a typical meander flood plain pattern of broad ridges and river channels. This unit also includes the temporary alluvial formations (chars) within and adjoining the main river channels. (II) EASTERN REGION (a) Northern Piedmont Plain -- consisting of mainly loamy sediments subject to shallow or intermittent 34 flooding and clay basins subject to deeper flooding. (b) Sythet Basin -- consists mainly of heavy depo- sits of clay soils. (c) Young, Middle and Old Megna Estuarine -- mainly alluvial land. The sediments are predominantly highly silty and finely stratified with clays in some basins. Flooding varies from shallow to deep. (d) Chittagong Coastal Plain -- consists of loamy alluvial from adjoining the hills and level clay plains adjoining the three rivers. (III) NORTHWEST REGION (a) The Old Himalayan Plain -- is a complex pattern of broad sandy or loamy ridges intermixed with numerous shallow channels with mainly loamy soils. (b) Tista Alluvial Flood Plain -- conSists of silty sediments and heavy clays. (c) The Barind Tract -- is mainly level with slowly permeable soils overlying little-weathered madhu- pur clay. (d) The Ganges River Flood Plain -- where clay soils predominate in basins and on the greater part of most ridges, but silts and occasionally sands occupy higher ridge crests. 35 (IV) SOUTHWEST REGION (a) Ganges River Flood Plain -- as a whole, com- prises a typical meander floodelain landscape of ridges, basins and old channels. The Ganges channel itself is constantly changing, eroding and depositing large areas of new alluvium which is calcareous. Clay soils predomi- nate in basins and on the greater part of most ridges, but silts and occasionally sands occupy higher ridge crests. (b) Gopalganj-Khulna Peat Basins -- occupy peren- nially west basins and are covered by clay. (c) The Ganges Tidal Flood Plain -- the sediments are mainly noncalcareous clays, but they become more silty in the east and usually have a buried peat layer in the West. The above mentioned physiological characteristics of land, soils and hydrology and its influence on existing cropping patterns is shown in Table 2.1. A.l. Climate and Rainfall Bangladesh has a tropical climate with mean monthly temperatures of 60°F and hence permit the growth of a wide range of tropical and sub-tropical crops such as rice, sugarcane, jute, potatoes and tea, etc., throughout the year. Rainfall ranges from 50 inches in the west to over 200 inches in the northeast and the southeast. The monthly distribution of rainfall follows the monsoon pattern of 36 Table 2.1. Present Adaptation of Cropping Pattern to Land and Soil Land Soil State of flood Present Cropping Pattern Type Land Elevation permeability During monsoon season Spring SummerTI’fill winter 1 High High Free Aus or Millet Fallow or Kharif Rabi or fallow vegetables Fallow 2 High (Flood Low Free Aus or Jute Fallow Rabi or plain) Fallow 3 High (Barind Low Water poundable Fallow or Aus T. Aman Fallow tract) by bunding 4. Medium High Water poundable Aus or Jute T. Aman Rabi or by bunding Fallow 5. Medium Low Very shallowly Aus or Jute T. Aman Rabi 0r flooded Fallow 6. Medium Low Flooded up to Aus or Jute T. Aman Rabi or 3 feet. Fallow 7. Medium Low With flood hazard Aus Tall T. Aman Fallow 8. Low High or low Deep. but not Mixed Aus 8 B. Aman Fallow hazardous B. Aman. (continue) Jute Fallow Fallow 9. Low High or low Deep and B. Aman at Aman Fallow hazardous (continue) Fallow Fallow Boro 10. Low, char High or low Flood hazard Fallor or Fallow Fallow land Jute or Aus. Fallow or or millet or B. Aman(cont) B. Aman 11 Low. char High or low Flood hazard Fallow or Fallow Fallow land Jute. B. Aman Fallow B. Aman (cont) 12 Low, flood High or low Flood hazard Aus or Jute Fallow Rabi or plain ridges or Millet Fallow outside of embankment 13- Low. Flood High or low Flood hazard Mixed Aus B. Aman Fallow plain depres- and B. Aman (cont) sions outside Jute Fallow Fallow of embankment Source: Report compiled from data extracted from (1) East Pakistan Soil Survey Reports. Bogra District and (2) Report on the Potential for Rainfed HYV Rice Cultivation in Bangladesh, by H. Brenner. 1974. 37 South Asia, with heavy rains starting in June and ending in October. From November through March, there is vir- tually no rain. April through May are transitional. The abundant monsoon rainfall, with the resulting widespread water logging and flooding, give a comparative advantage to wetland crops, rice and jute over dry land crops. However, wherever drainage conditions are good, there is a great diversity of cropping opportunities during the monsoon season. A.2. Farm Structure and Tenure The average farm size, as found in the 1960 census, is 3.50 acres; out of which about 3.1 acres are cultivable land (13). Distribution of the number of farms under various size groups indicates that 83 percent of the farms are of the size of 3 acres and below and that 0-48 percnnt are of the size of 25 acres and above (14). In addition to small size, farms in Bangladesh are highly fragmented. About 90 percent of the farms are fragmented of which 29 percent of them having more than two (10) fragments. There are three types of tenure groups, namely: (a) owner-cultivator; (b) owner-cum-tenant cultivator; and (c) tenant-cultivator. Farm classification on the basis of tenure indicates that 61 percent of the farms, account- ing for 52 percent of the total cultivable area are owned by owner-cultivators. The owner-cum-tenant and 38 tenant-cultivator poses about 37 percent and 2 percent of farms comprising 47 percent and 1 percent of cultivable area respectively. A.3. Land Use and Cropping Intensity Land is the major constraint to expanding agri- cultural production in Bangladesh. Topography, soils, climate and depth of flooding determines to a large ex- tent the pattern of land use as well as cropping inten- sity. Data on land utilization shown in Table 2.2 indi- cates that increases in cultivated area have come about by a decrease in area classified as cultivable waste from an average 1.95 million acres in 1955-60 to only .67 million acres in 1973-74. Increased cropping of 'culti- vated area' has taken place through a decrease in 'current fallows' from 1.32 million acres in 1955-60 to .73 million acres in 1969-70. However, area under current fallow has increased to 1.55 million acres in 1973-74 just after the war of independence as farmers were compelled to keep their land fallow due to lack of production inputs. More significant than the increase in cultivated area has been the growth of 'cropped area'. It has in- creased from 25.91 million acres in 1955-60 to 32.84 mil- lion acres in 1969-70 (Table 2.2). This was achieved by a dramatic increase in multiple cropping which raised cropping intensity from 120 percent to 146 percent, but 39 Amumoz m_no>wu_:u m:_a coca noaa>wu_:uruav coca opnm>vupzu _nuo~ cop x coca commocu “oz \m oo— x «cm soaaocu .ouop coca mammoeu um: \m .m~m_ .zmouo_m:am mo acoscgm>om .me:u_:u_cm< do xcum_c_z .mu_um_uaum .mc=u_=u_cm< do :omeam “oucaom woo mam Gem yam New \mxo.o=oo=. do: use; .m mom. Nom_ new. Nee. acq_ new. no~_ omoocooaoa “waged -cmuc_ ceaqocu .~ me.om ...mm mm._m om.Nm v¢._m mm.- .o.m~ coca nuanced page» .0 mm.- sv.N~ me.- ae.- m¢.- _m.- mo..~ 4vu_au .m mm.m um.o Nm.o cm.o e~.o Nm.m mm.m :o_uo>.u_:u to» opaopeo>o uoz .N .m.m pm.m om.m om.m cm.m oe.m oc.m anode; .— efirm~m_ ~n1_~m~ -ro~¢P okrmom_ moro~m_ marcoa_ oormmm_ mgo_3u_ucoa ac_ucm acmvco moccm>u moccm>a somxrm couarm Amoeba mo mcow_—_ev =o_uo~+__u: ucod ammuopmcom .~.N mpnmh 40 the same has been declined to 136 percent in 1973-74. Similarly, land use intensity has increased from 86 per- cent in 1955-60 to 94 percent in 1970. Nearly half of the cultivated area is double cropped. The nature of the double crop system is defined by relief or depth of flooding shown in Chart I. On medium and highland, cropping intensity is usually above 140 percent (shown in Table 2.3), where aus or jute precede T. Aman, which precede dry season fallow or rabi crops. 0n lowland (flooding 3-6 feet), the cropping intensity average about 120-130 percent. A.4. Croppingypattern and relative Importance of Crops Chart 2 indicates graphically the existing crop- ping pattern and how the cultivated area in Bangladesh is used throughout the year. Rice dominates the chart and the agriculture of Bangladesh. About 90 percent of the cultivated area is devoted to rice (Table 2.4). It is grown in three seasons -- aus, aman and Boro. Aman rice is predominantly rainfedand accounts for 56-58 percent of rice acreage and production. Aus is the second largest rice crop, predominantly rainfed and occupied about 32 percent of rice acreage during 1974-75. Boro, however, is entirely dependent on irrigation and is grown in winter season. There has been a gradual increase of boro acreage *over the years from 5 percent of the total rice acreage in 41 Chart I fl fl . I—1 w _ q . . _ . . . fl . _ . . _ _ . z u _ n _ _ _ . . " $0.33 ....o a .61: co ” u _ 2.01.: no . 3.. 9.0“. . .mau _ ..max _ chi“... . £0.31 " are ~ a #3; a: . _ . . _ m _ . . _ . . _ . _ . — . . _ . _ . . . . _ _ It'o‘ — . . . _ 1.. I . . ..eJUU" 2.1.2.4 u — _ r. uoii .Q . ~ . " 5.35...an _ :20...“ 35.}. ”we...“ _ 21...}; .....duh . u . rm... . III a an}: . a”. 53.1 u 640....» «a $0.4: “owe/443.31....“ cpl...“ _ a r :3. “.....3. n)... - _ . M _ _ - a .. N . . . _ L - _ p . o — . . . 2.3. . . _ _ _ . kmduuannu . . 2.01.}: S ...o, 4.3.) S . £013 «.0 . do a»)... no “ 7.1.3. ”9693.5 2 .m12 v.3» _ «MW: our... _ um». I u 34 _ .8... . 0.... 3.. 852 _ mu... “ mono .m i _ c o)"; $35. I! . n . u . ~ . . _ . _ . e . . _ — . . U.- “HUN-I . _ _ _ . . _ , . _ . _ w . _ . _ . . . _ M . — — . - . . n u _ H . _ ~ . H _ m _ u _ _ . _ I. H . . _ . a _ m m . _ . . 2 our...” . h r .61 h... * F u _ o r..( .u _ n u m n ” .01 n _ _ _ o u .0 tinmn _ _ _ . _ _ _ _ . . . . _ . . ~ . P b V b a . A A . . . . ct. 5; . _ . _ _ _ m . _ of) _ 9.4.. _ 053 _ . . . . ..Cu 2 5.23., _ so... . .. _ z _ _ _ . u _ M _ m _ _ w «.an «(0.3.3.05 33.x .33.": 33 >9...» ..eatur 55¢ 341...} Ewhmerm 02_&OEU no ZOdehm344. 0:4: .Eoumod‘a 42 Table 2.3. Land Use (average 1969-1973) of Bangladesh Land elevation Crop Area cropping Intensity acr.'000 1 1 High and medium Jute 2,140 7 Aus (single)* 5,730 19 T. Aman* 9,280 30 Boro* 500 l Rabi* 1.500 5 Sugarcane, fruit 0 other 2,530 8 Sub-totail/ 21.680 70 143 Low and B. Aman* 4,900 16 very low Aus§/* 2,000 6 Boro* 1,750 Rabi* 640 2 2/ Sub-total- 9.290 30 127 Total cropped 30,970 100 140 Total cultivated 22.460 - - Source: Bureau of Agricultural Statistics, MA. (Report No. 455a-BD *Divison approximate l/Approximately 15.2 M ac. g/Approximately 7.3 M ac. alSown together with B. Aman as a mixed crop. AREA IN MILLIONS OF ACRES 43 Chart II BANGLADESH AVERAGE USE OF CULTIVATED AREA (1989 - 73) M IS SEASON AMANIMIIAOOI Si ASHN n,.-..- em .. I?! l- Plv 1". "no .r A-. r . . ”Md-“ ¢-..\_ .HCIJ»-~ . .....r. 'mevmwm- .... um tel!IOWIOIOOOOOOIOIeoiueeeelI Ice IIIIOOIIIHO-~e~.I......u...I. vO‘IOIIOIII race. one... a ‘0’... .W~o-eeoeueeeoI-oeeo -We-ooeaeeae-ee.eeoo”-oeln 0"”... I no. I .I .... e. 0....-IIIQOIOCOIIOIOOUOOOHIOOI aeoem.uc0. COW-loeoe. I.'00....I‘ll-IDIOQICIODIOWSI......OOOOD 00.... .e O I ......IOIWW INCIOOCI .'.l IIIIII: . :III. mu: ANI man ANNNAI I Hill" 2;? . .e.0.0.e.I.O I.I.I.I...I.O.... ......I.I.II ‘ no.0. . 0.0-0.0 0.0.I. I O O o -:e a. e O O O c e o :‘u'.’.'." '0'. . ‘.:.u.o.o.e.e.’.e. I o.::- a ...':.u.u':.:.:.e.e.o.e.e” I O. I. I u. I. ' .: . IA fi'f.l,|”w JAAJ»! - an.» .1411; i 1'.",, ‘ ‘ t,__,"‘ . . 633.76.]th .III II I 's I. (3.37 ..;.' '. .:.:-:o:e:o:o: o :-:|I,\“| (‘|.()|”.:o:o::: e:e:e:I:o:e . . ~ - rag-“w... "OH" I” AMI SF ASON FALLOW N-a'h‘flgf’l.’ [(9) {WA ’o' WI} r: R FALLOW IALI OW . ~\ \\ ‘. I , , Aw“.4.44.-J:.LL-.L_LL~J ,g.."..q.m_ . _. .... ...—...... .35 .- Leer ‘mi mm .CIW I ' ' j ’ ' .Lm ‘ U . “we‘- '1 >‘ Z '"flv W ‘- O .. 'owoeooe oeflooleioceoc evereoue: FALLOW . . ' .. H.132; _. ‘ . l I ' . 1 .‘ fl; . .- ’ .1 if ' y L- “a. .2 a __y- t’ :r‘ "7‘ ‘ iO—m'. 9 -—,' ' FALLOW C d .- . f W—Tyfi'l‘ ’7'! I..o'eooeaonee.ooe.oeoeeeo'eeeee.e . ., -..!..!.I—L—..J. ..JJ-J..W ‘ Clix-MM, 4H - ’ I. ‘- ‘ a—nl _‘ I . , ‘ ' 7 Mle II An" , _ ‘ ’1'. Mt ("My ' .- 1 .. - . 9- . ,‘ o_r “can - .~ ,. , . _ . r; ‘ i 'v ‘ ‘. 'I k,‘, . . .--- r..,l.4.(... ...-2_ ' FALLOW 5 at ' .' .~1 ~ . 0 MI "MAUI, ‘_ ‘ --. I mama; mrit I)" ‘ 0 ..v . .... .. -r, Tr—vv .- 7"- .177 -—y 3 —l -M‘-4‘d—Ifl"" "‘r «4.! q | 4‘; A I" p c I. - 9 HQ- :—~ I I . . 3 . '. . I ' .- _ \.‘- ‘ \ e ‘a. “\ ' ""..’ S P‘ |\“ .-“‘:.‘a ‘5 I "~ ‘l‘ . ‘\ -"“.“- :‘\\‘ I .II I ‘n. r.'\ "3.3.1""‘.~r-"C'I-\‘S""r ” I'-, t"‘ "."3‘5131‘!" i." IIIII'I ‘ . t n ‘ . o ‘ e - e e ‘ . . ‘ a ' a, 2 --. " -’ I't. ‘3 s ”r. I." 1‘ n1 h‘a‘ ' I " I.".I.’i-\ I'\ “'1‘ ..-,” . ..-.1 In.” II‘1~"‘|\"' ‘. -'I~\‘ I IN}. 'X‘ ..I.':‘. ‘ ..‘T- .' ' .I‘.’ :W.‘ .3 I. 3 ‘S. ’t " .. ‘1' I; . I 1:.CC'II:..'.C..;I: . mo..- 0.. .M e o Note: The cumming masons have been sompIIlII-rl in this chart. Sworn-I and harvesting actually take plot-I with greater (WI-that than stwwwt have. 44 l960-65 to 9 percent in 1969-70 and over l2 percent in 1974-75, Table 2.4 shows that the total rice area has increased from 22.2 million acreage in 1960-65 to 24.2 in 1974-75, an increase of 9 percent. Most of the increase in rice area was gained through expansion of Boro and Aus rice. Jute is the second largest crop and occupies about 2.2 to 2.4 million acres of total cultivated land. In terms of regional distribution of crops, whereas rice is grown in all parts of the country, Jute is mainly concentrated in the districts of Mymensing, Dacca, Faridpur, Rangpur and Camilla. About 72 percent of the total area under Jute is located within these five dis- tricts. The sugarcane area is mainly concentrated in the Northwest region and a part of Kushtia, Kishoreganj and Mymensing districts. Potato is mainly grown in the North- west region followed closely by central and eastern re- gions. The Southwest region hardly accounts for 5 percent of the total potato area. B. Potential for Agricultural Development This section deals with the existing states of art and technology in the realm of agricultural crops, parti- cularly with reference to rice, and assesses its produc- tion potential with the advent of new 'seed fertilizer' technology and the results of soil reconnaissance surveys. 45 .ouauuaueuod no mun-«aux .ouuu-«uouu «unauuauuuud no season "ummmmw .uonum-u»ou« \M nan" "on" n “A won~ uoqn nod. "on“ nos. o4.- uq.«~ ~4.- Be.u~ oa.~« oa.~u ~e.~« aa.- u“.nn an. n au.on s~.mu na.~n an.~n a«.~n aq.un p“ H ”n.” o4.o mm o ev.o an.o Ia.a. «v.0 n”.o “a.“ an.“ nu.o “..o "H.o u..o d~.o I”.n "H.o “v.0 I”.o a”.o -.o ou.o o“.u “0.0 “u.o do.” no.0 ~o.o no.0 no.o 46.0 Imus.” an.” an.” no.” o~.~ as." . an." an." "v.0 an.o mn.n mm.o no.u ao.o no.o 00.0 “H.n «2.0 ou.o h".o a”.o o~.o .~.o ofi.o nu.” o~.o ”.o .N.o —~.o ‘~.o "5.0 o~.o mn.u m~.o u .o «m.o 04.0 oq.o "4.0 ~4.o mn.n nn.u mn.o on.o 04.0 ~«.o ...o ~a.o Auv.o v A": o v “mo.” v “mo.c v Ao~.o v Anh.c v .oa.o v .n~.o . nu.o m ”0.0 an.” .0.” mo." . mo.o mo.o ec.o «v.0 _ -«.n “H.“ a~.o n~.o o~.o .~.o o“.o no.0 * no.0 ”I.o nu o an.o qa.o an.o n».o h." . n“.¢ h.o on.n "..o ~¢.o ~a.n oo.o u.” _ «N.» ”.0 an.“ o».o an.o a~.e ".0 «w.o _ on.o Un.a "n.o ~u. os.o on.o o~.o as: _ .3..." 3.2:. 3:: 322 8.23 :3: 5...: w hm.“ . no." no.“ H.” ~4.u aw." dc.~ an.“ . "4.n« _. n«.qu ~«.4~ fin.nfl .n".a. .qn.cu 4.4a no.an . em.. as,“ an.» a.» on.” oq.. oo.. «u.» p.4ho” _ a..nho_ nh.uaaw ~5.H~an oa.uom« no.moon p_.u F ~F_EV smmumpmcmm :p uptake“ autocr— maogo Non“ nu.- 3 c1 N .[O’cfl a n.1uoI-h.ao~oo.g £515 c>¢>¢3 rottcsczcn Onh~flfll~ A u N O V 23 c><$ oq.o oo.o 4~.o c».o ..q.-. on." no.4" a... nonmmou :wmz Lone: mmg< Nun~ Hun" an." ac." 4n.ou m 10 I . N Inr‘t) ‘5 D.-. ‘3 Otdfl , Jnt-Oplnl‘ls MH~OIWN!JH-Ov\ OOOO'J'COQUC OQfl'llflfi '00-... ¢>¢J¢o c$c>-— fic)v;r- “N' A 0‘ O O O u A 'l . O U “Qt. 00¢: \, '0 f) n D . <3 N c>¢3<> “(.M ($¢$¢3 ‘1 .- an—.n«v Anu.nuv a".~ vu.~ no.4" um.qu un.» «o.o emumeo~ avIIVo. \"o.coo~ ououo>< sauna-"nu uuuanouu aqua-h unohuau ucuvauuau nou< monnuuu uvz nuouu uu< "upon .3 .3 a ..u ..n 40..- . 4. nauuou .u “an .N coupon-«op one an".Lu . " cannon nounu .. ouqun .n ououuqmuu .. .uuuum .o woo-awn and u.t ”my vuounnfl can can» any nun-"Io noxuu .n uaupu .n «anon anon no” nun! ”v .n< “a. cu.“ .“ rnuuu .¢.N mpnmh 46 8.1. Traditional Technology Bangladesh rice production technology has long been dominated by varieties with limited capacity to utilize fertilizer efficiently. Traditional varieties, like in most other areas of Asia are relatively tall, weak stemmed plants and not receptive to chemical fertilizers. For this reason, farmers use a very negligible quantum of fer- tilizer. The end result has been that cultivation tech- nology used on traditional varieties remained largely unchanged for decades. It is observed that though the area under rice has increased by over 3 million acres from l960-65 to 1975, production has increased by only 1.1 percent during the same period. Thus, the increased rice output in Bangladesh remained dependent on major modifi- cation in the crop environment through increased cropping intensities or a shift from broadcasting to transplanting techniques. 3.2. New Technology_-- "Seed-Fertilizer Revolution" The major breakthrough in rice production techno- logy was achieved with the introduction of IR-8 varieties during Boro season. On the average a yield of 1.5 tons of rice was obtained in the Boro HYV crop with proper fertilization and pest control. The area under the crop has increased from an annual average of 0.83 million 47 acres in l955-60 to 2.87 million acres in l974-75. In spite of its tremendous success Boro constitutes only l2 percent of the total rice area and its further expansion has been extremely difficult due to the following reasons: it is (a) largely dependent on irrigation water which is very difficult to meet; (b) very susceptible to disease and pest attacks, and (c) requires a longer growing season than the traditional Boro variety, thus interfering with Aus crop. The overriding constraints to the expansion of Boro crops led to the innovation of HYV varieties for Aman season which constitutes nearly 60 percent of the total rice acreage. The experimental findings suggest that these new varieties are (a) better adjusted to the rainy season than traditional Aman; (b)matures in l20-l35 days as compared to 150 days or more required for tradi- tional varieties; (c) highly resistant to bacterial leaf blight, tungro virus, and partly resistant to leak streak and stem borers; (d) has a production potential of 3 tons or more, and above all, (e) requires much less water than ll2-8. Similarly, the innovation of varieties like 122-72-4-l-2, etc., for Aus season offered similar pros- pects. This recent dramatic breakthrough achieved in the production technology of rice ushered in an enormous pros- pect and potential for a large increase in rice production,‘ thereby, for solution of food shortage problem. 48 8.3. Re ional Land Suitable to New "Seed Fertilizer"TechnologygasDetermined by Soil Reconnaissance Survey With a view to devising a better national agricul- tural development strategy to increase agricultural produc- tion, the government in 1964-65 undertook a soil survey project assisted by FAO/UNDP. The major purpose of land capability evaluation is to facilitate the identification of suitable areas for agricultural intensification. Since land and water resources determine the potential for future agricultural development (i.e., where production of exist- ing crops can be increased or replaced; where new high yielding varieties can be introduced with or without irri- gation, etc.), it is anticipated that a comprehensive soil reconnaissance survey would provide a sound basis for allocation or land among its alternative agricultural and nonagricultural uses. By the end of l969-70, the soil survey had com— pleted a detailed reconnaissance survey for 70 percent of Bangladesh. In addition, sufficient information has been obtained for the remainder of the country through aerial photo interpretation and exploratory surveys to extrapo- late the results of the survey to the entire area of Bangladesh. The survey provides information not only on soil conditions but also on present land use and on physical factors limiting agricultural development, such as depth of natural flooding, surface relief, erosion 49 hazard. This information has been interpreted to deter- mine the suitability of individual soils for production of particular crops, with and without irrigation. The survey recognized five classes of land, ranging in terms of quality from very good to very poor. The survey divided the country into four regions, namely, (1) Northwest (NH); (2) Central (C); (3) East (E); and (4) Southwest (SN), primarily on hydrological and topo- graphic (soil strata) considerations. The study indicates the considerable scope for increased rice production pos- sible from the introduction of IRRI rice varieties in all regions of Bangladesh. In addition, soil, land capability and hydrological information reveals that there is ample suitable land available that could be irrigated where necessary for cultivation of strategic crops such us jute, sugarcane, potato. The most important finding is that some ll mil- lion acres out of 22.5 million acres of agricultural land , are suited for agricultural intensification with an input package consisting of improved crop varieties (especially IRRI rice), fertilizers, and pesticides without major investments in drainage or flood protection works (see Table 2.5). About 6.2 million acres would be suitable for the existing IR-20 Aman variety following either IRRI- Aus, or jute. On a further l.7 million acres in the dry western portion of the country, where currently only a 50 Table 2.5 LAND CAPIMILITY AND IMPROVED VARIETIES Alternatives offered by land .0gpabiliticn and Improved Varieties Uithout Irrigation Additionaf and Drainage/Flood with Irrigation Control Works onlL PDOTAcmsY (m Acrefl Rice ' Bom/l‘ransflmted Aus: a) Traditional Varieties 1.5 1 .h - b) mu Varieties M _:_ 15. rot-.1 3:1 1.2% 1 .6 Broadcast Aus: 3‘) Traditional Varieties 8.5 l .o - b) mu Varieties (IR-176 (Chandina) , m-hhz) _;__ L5 - Total 8.5 8.5 from linked Aman: raditional varieties 9.3 - - b) IRRI Varieties: . 1) 13-20 0.2 ' 602* ~ ‘.7 11) 111-20, Ill-1:112 Ill-176 (Chandina) and non-photosensitive varieties in research pipeline - 1.7.. .. iii) IR-th‘, its improve- mtn .:.. 2:1“ -——." alt-total '9.5 11 .o - Broadcast Aman 5.5 h.o - m1 1&9. 15.2 . . Rice - Total 25.6 - - grape Other Than Rice jute 2.5 L5" 0.5 "belt 003 - 5018*. Dresden: (Mustard and related crops) 0.5 12.0! - “maxing!” 00‘ - o e) Rabi Season ma. - 9.8 b) Kharit Soaeon 9;.1' 3.1 - Totlll 0.1 - .- Sugarcane 0J1 3.9 3.9 G Yields will be impmvc- d if mpplementary irriration is available. *I Assuming that wheat is conflnod to areas N of lattitudo 25°" ' _/ Source: Soil Survey by Soil Survey of Bangladesh. MOP/[A0, 51 single local rice crop is grown, a HYV rice either IR-20 or IR442 would be able to replace the traditional variety. In addition, 3.l million acres of deeply flooded broadcast Aman could be substituted by IR442. Moreover, of the 8.5 million acres of B Aus, some 7.5 million acres could be replaced by HYV varieties. Besides rice, substantial acres of about 4 and 5 million acres were found to be suitable for sugarcane and wheat, respectively. The development potential for further intensifi- cation of crop production in different regions as revealed in the study is enumerated briefly below: (i) The Northwest Region (NR) The NR has higher potential for rapid agricultural development than any other region. This is mainly because of the great extent of highland and shallowly flooded land suitable for conversion to IRRI Aman varieties without irrigation. Much of this land, except on the Briand tract, could also produce a broadcast IRRI-Aus crop without irrigation to replace the local varieties. Nith irriga- tion, Boro or T. Aus could be grown on all except the permeable ridge soil covering much of the Dinaipur and on the highest flood plain land elsewhere. This permeable land is also suitable for different Rabi crops and sugarcane. 52 (ii) The Central Region (CR) In CR IRRI varieties could be substituted for local Aus and T. Aman varieties throughout the region. Irriga- tion would make possible a vast extension in Boro cultiva- tion, replacing deep water Aman on basin land subject to flood hazard and B. Aus in higher land where drought is the main hazard. Irrigation would be particularly benefi- cial on the Madupur tract, especially for sugarcane, dry- land cereals and vegetables. (iii) The Southwest Region Irrigation in this area, as within the existing G.K. project area, could be used to enhance both rice and dryland crop production: sugarcane, tobacco, and vege- tables, 8 Aus and T. Aman together with jute and rabi crops on lower ridges. Irrigation also would make it possible to follow a T. Aman crop (which may not need irrigation) with IRRI Bora. (iv) The Eastern Region The highest potential exists on the old Meghnan flood plain, especially in the higher east where IRRI-Aus and T. Aman could be substituted for existing varieties without the need to provide supplementary irri- gation. Irrigation in this area would permit Boro to be substituted for B. Aus; alternatively it could be used to provide high yielding dryland rabi crops. 53 The areas suitable for different crop production activities in each region as identified by land survey is A shown in Table 2.6. m T—TT L 54 Table 2.6. Summary Estimates of Acreages Suitable for Specified Crops By Regioni Region Total area CVOP (000 acres) of region IRRI IRRI Boro b/c/ (000 acres) Transplanted Transplanted Nheat- — Sugarcane Aman Aus NH 8047 2l50 4548 2752 l469 (2628) CN 5254 1326 2978 2082 109) (1720) EST 10917 1996 4069 2456 438 (662) SN 9304 2450 3992 1814 870 (400) TOTAL 335229/ 7922 l5587 9104 3868 (5410) llAcreage of crops are gross, exclusive of settlement and water. The estimates relate solely to the suitability of soils and land for crop production with irrigation (except for Kharif groundnuts, and for most of the IRRI Aman acreage). They do not take into account availability of water for irrigation. Drainage is not assumed. 9’The estimates relate to the crops individually, not as compo- nents of crop rotations, e.g., on most soils, dry land rabi crops cannot be grown if the land is also used for IRRI Aman. E[There are indications that high wheat yields may not be obtainable south of about latitude 24°N, but agronomic trials are needed to check this. If confirmed, the acreage suitable would be reduced to the figure given in parentheses. gJGross area of Bangladesh is 35,280,640 acres, including water in the major rivers and estuaries. Source: Soil Survey by Soil Survey of Bangladesh, and UNDP/FAO. CHAPTER III FRAMEWORK AND STRUCTURE OF MACRO MODEL Chapter I has provided a conceptual framework for these studies. In this Chapter a detailed programming model is described to operationalize that conceptual framework. Linear programming is used as the basic tool to investigate the likely impact of alternative technolo- gical adjustments at the regional level on the allocation of resources and on the estimation of production response of different crops. Profit maximization is selected as the objective function over cost minimization because it seems more realistic that farmers in Bangladesh are gen- erally concerned with increasing their profits. There is ample literature which purports to test the economic rationality of agricultural producers (lO,33,54,55). Such studies generally concluded that producers, even in the most backward areas, act as a profit maximizer within their technological and institutional constraints. Even Dillion and Anderson in an article generally critical of the methodology employed in this area concluded that using a probability technique does not change this con— clusion. The important reasons for choosing profit 55 56 maximization in the objective function are: (i) in the past, empirical studies explicitly employing the profit maximization hypotheses (e.g.,in linear programming studies of individual farm) have generally provided results consistent with observed and plausible behaviour (47,52) of farmers in Bangladesh; (ii) product prices and techno- logy, the most risky components of a production function, are mostly controlled and determined by the government. The modern inputs such as HYV seeds, sophisticated machine- ries for irrigation, fertilizer and pesticides, etc., are all procured and distributed to farmers through government supply and service agencies. The procurement price for important agricultural crops is determined by the ministry of agriculture in consultation with other ministries; (iii) previous studies (l6) found that farmers in LDC's do respond to price incentives though the magnitude of such response varies from one environment to another. However, the present study has not altogether omitted the risk- factor. But, to a certain extent, the risk factor has been implicitly considered in the model by incorporating the minimum consumption constraints which have to be pro- duced by the farmers. The whole nation consists of four regions which is used as the unit of optimization. A region is considered homogenous with regard to climate, tODOTOQYa farm size and 57 resource distribution. For such a homogenous region, it is possible to view a regional model with the following three elements: (l) the objective function; (2) the activity set; and (3) the constraint structure. The mathematical LP model can be expressed in the following form: maximize n = C'X subject to restrictions AxfB xio where A is an MXN matrix of technical coefficient C is an nXl vector of prices X is an nXl vector of activities 8 is an le vector of resources and other constraints C'X = n in the objective function. Objective Function The model maximizes the following objective function: um]: 11 : Max ( t) [Pit Sit + zit Cit ‘ Dit] i l ”(t) is the objective value. Pit is the harvest price for the final output in year t. S. is the actual level of ith sales it activity in year t (in pounds). 58 Zit is the price of crop output consumed at farm level Cit is the actual level of ith consumption activity in year t in pounds. Dit is the actual amount of cost incurred for purchasing different variable production inputs like seeds, fertilizer, hired labor, etc., in year t. This objective function assumes that each farmer (i) expects the same output prices; (ii) incurs the same amount per unit production and marketing costs; (iii) pos- sesses initial endowments of land, labor, and money capital in the same proportion, and (iv) responds the same way to price and income changes in making production and consump- tion decisions. The above linear programming model can be used to represent the sum of the decisions for all the farm-firms in a particular homogenous region. The objective function maximizes the short run net national farm income (NNFI) on fixed factors. The variable inputs to the agricultural production process are expressed in terms of seeds, fertilizer, pesticides and hired labor and accounted for by assigning unit cost per acre to each of these items. The interest cost is also included in the variable cost. The cost associated with the fixed factors of production such as land, family labor, bullock power, and fixed irrigation rental were excluded. 59 Output is expressed in terms of crop yields. A farm house— hold generally uses its output in two ways -- part of it is consumed at the farm level and the residual is sold in the market. Consequently, the output price is determined in the model by the manner in which it is utilized. The marketable surplus (sales) is accounted for at the minimum procurement price fixed by the government for each commo- dity whereas the quantity which is consumed at the farm is accounted for at the buying price. (For details see Chapter IV, Product and Factor Prices.) The Activity Set A typical farm household in Bangladesh is expected to be engaged in certain activities like (i) production activities; (ii) labor hiring activities; (iii) subsis- tence consumption activities; (iv) sales activities; (v) buying activities, and (vi) financial activities. The activity set used in the model is denoted by ~ IP]---Pg, H]---HL, C]---Cg. S]---Sg. V]---Vq, Mfl where P]---Pg are production activities measured in crop ‘ acreage sown to l---g crops in the region. H]---HL are activities associated with the purchase of outside labor to the extent family labor could not fulfill the labor input requirement for various crops. 60 C ---C are the consumption activities of l---g crops. The household consumption requirement is re- tained from total farm output. S]---S are sales activiiies of final output for cash. V]---V activities involving the purchase of v outputs for domestic consumption. associated with farmers borrowing activity. The following assumptions were deemed essential for the model: (i) input-output coefficients considered are consistent with the farmers' cultural practices; (ii) prices are held constant at current prices; (iii) the input subsidy policy of the government would continue in the same form and rate in l975 except for fertilizers. The matrices for technology I is presented in Table 3.l. The basic structures are the same for an- other technological stage. Consequently, they have not been duplicated here. Each of the model activity as mentioned above is discussed in detail below. (1) Crop Production Activities: Eleven (ll) crop production activities are considered in the model, which consists of seven different varieties of rice and four other food and cash crops: wheat, potato, jute, and . 4.3- :. 00-0.. u. 00.... ulfil: one...- a.e.c.-.o0c coo-... go. anal: till-III.-- it'l' [.1 3. :33. . 0. O.~ ¢.-.oe..oo. a. ...aooo.o - =33. 0.. .0 ... ...-8 3. CC... CV... O. - v3.0. 61 n11. (frilrl. .l-Ili 09-1-9sn. .3 1...:- 3...3 . .... 3...: ......- ...: 2:33 1.18.354 ll) vilvl... lnlulnll.)t'u.. will '1')... 3c :- .033 - ... ... :- 3.l: 2. .3...- to.‘ a :3: Iiooco..s. . .02! 3. 3...: .413: you...-o. :- .33: . '3‘. 7:1 I... s! c'..!. =- 33‘. ‘4-i.l. Q. .3333 cell. 2.31: a so... v8.1 2 00.9: :. {...- A. nae n. ~.§ .- . s _. 8.3: . 2...... -llTill ll. llllllll'l.1."l (I .I11 I). o 3.33 ......aoo. o .00.. ...... u 3.31 a. ...-l. u. ..I ...... I: '00. .. owl... .1 oo- 3- .... ...... .. .... ..o.....¢ou loo-caaqv‘ 1...! so .9 ear—04— ea .0 :..---~:...- 3.00.0.3 ,. a 0.... 62 sugarcane. These crops entered into the model in two time periods viz summer and winter seasons. The net price shown in the Cj row for crop growing and harvesting activities is negative. The latter functions, when considered sepa- rately, subtract from income. The income from different crop activities derives from its sale whose coefficient in C row is positive. Thus, the Cj row entry for each crop activity (P1--- to P44) is negative by the amount of allo- cable costs, other than land and labor, entailed in crop growing and harvesting. (2) Labor Hiring Activities: There are two main sources of agricultural labor -- family labor and hired labor. Farmers hire labor only when the household is not adequate to perform all the agricultural operations in any particular month of the year. In all, there are twelve labor hiring activities, i.e., one for each month. Conse- quently, labor hiring activities deal with labor from outside the families at the prevailing wage rate and are included in the model through transfer activities. The total amount of hired labor depends on the level of pro- duction activities and the amOunt of family labor available from the households. Since the hiring activi— ties add to the supply of family labor rows, their coef- ficients are negative. However, the labor hiring acti- vities in their row column (JHRLBRZl to JHRLBR32) have positive coefficients, indicating that each man-day of 63 labor hired reduces the 8 column value by one man-day. Net prices for labor hiring activities are negative because the labor hiring activities themselves substract from the value of the program. (3) Consumption Activities: Subsistence consump- tion activities describe the consumption of farm commodi- ties by the farm household. Farmers grow food crops mainly to satisfy their household consumption requirements. In this model the minimum household consumption requirements for each farm commodity are determined exogenously from the farm household budget studies conducted jointly by USAID and University of Dacca. In the consumption activities in the row column for rice (CONRICE 49) no specific distinc- tion has been made in between different qualities of rice. In other words, all different qualities/varieties of rice will hold uniform taste in the preference list of farm households. The Cj value of consumption of different crops are positive (CONRICE 65 —-- CONPOT 74) as the price of crops has not been taken into account in the objective value of each crop production. (4) Selling Activities: The marketable surplus of various farm commodities is determined as a residual of production and planned consumption decisions taken by farm households. As the consumption of different farm commo- dities is determined outside the model, the marketable surplus function of farm household can be obtained. 64 > ” > 51(t) ‘ [03(t) ‘ Cgc3m QHm LmNVFPugmu\m .commmc some cw »_aq:m cmNPFPpsme _m:pum mmmcm>m Amnuosmpv meme» m ummp mo mvmmn on» :o mums mew: mmmp use omm. mcwgsu mLmNPqucmm eo cowumuop mmpzcowmmg m;»\fl F meo._ mN©.F mom oom._ mme 4< ca~epeeeai \Iamrzcoe am umeas_pmu .N.m ainae 71 two phases of development are shown in Table 3.2. Water Constraints In Bangladesh, any future plan for increased agricultural production is crucially linked up with the feasibility of expanding irrigation facilities. At pre- sent, irrigation facilities are being provided through (1) low lift pumps; (2) shallow tube-well; (3) tube-wells; and (4) indigenous methods. These different modes of irrigation provides about 4.88 million acre feet of water and irrigate about 1.434 million acres of land. Irriga- tion potential is, however, constrained by the availabi- lity of surface water as well as the intrusion and occurrence of saline groundwater in the south. These prb- blems have been aggravated further due to diversion of considerable quantities of water from the Ganges during the winter season through the construction of Farraka barrage by India. The plan indicates that the major rivers carry about 50 million (46) cusecs of water. This water is more than adequate to irrigate summer crops in- cluding high yielding varieties like Irri-Aman. 0n the other hand, during the long dry months of November through April, the rainfall hardly accounts for more than 4 inches. 4Source: Compiled from Bangladesh Agril. in Statistics, Ministry of Agriculture, Bangladesh, Nov., 1973. pp. 57—61. 72 Consequently, it is in the winter season that the need for irrigation is deemed very essential. The flows of water in all the major as well as minor rivers add up to only 22,500 cusecs of water during the winter season, which could possibly utilize a maximum 5 in the number of 45,000 single stage low lift pumps winter season. The past trend of using low lift pumps indicates that the number of pumps has increased from 2,024 in 1962-63 to 35,427 in 1974-75. However, coverage per cusce pump has decreased from 38.5 acres in 1962-63 to 20.0 acres in 1972-73. This is partly due to the fact that pumps are being provided without careful survey of surface water availability and partly due to the lack of government institutional and organizational abilities in organizing farmer groups large enough to make maximum potential use of the pumps' capacity. These are basic inherent bottlenecks in any organization and require a considerable gestation period to manage. Understanding these situations, it is expected that a total of 45,000 low lift pumps of full capacity would be fielded by 1980. However, by 1985 no enhancement of fielding of low lift pumps is envisaged due to a paucity of available surface water. The maximum potential use of the estimated 73 existing surface water of 225,000 cusces would be utilized only by expanding the coverage of each pump by 1985. Groundwater represents another potential source of irrigation where surface water is not available. The com- prehensive ground water supply has identified about 9.5 million6 acres for prospective tube-well development. From recharge considerations alone, it is estimated that about 47,000 (2 cusces) tube-wells can be accommodated in the area. However, there are serious limiting factors on which devel- opment of groundwater depends: (i) safe yield; (ii) econo- mies of pumping; (iii) diminution effect of surface water on ground recharge capacity and saline water intrusion due to diversion of water through Farraka barrage, and (iv) in- stitutional andorganizational capacity of government. These factors tend to have a tremendous impact on the possibili- ties of expanding the tube-well program. A hypothetical long range water development program, as estimated by Harvard University, indicates that a total of about 8.2 million acres would be brought under alternative modes of irrigation (shown in Table 3.3) as compared to the planned target of 19,000 tube-wells and 15,000 frictional pumps covering 1.369 million7 acres. In addition to the above mentioned projection and programs, the following factors Ibid., p. 144 Ibid., p. 144-146. 74 Table 3.3. Hypothetical Long-Range Water Development by Region Region Low lift Tube-wells Polders Total pumps Northwest 0.13 2.21 0.13 2.47 Central 0.25 1.15 0.25 1.65 East 0.40 0.27 0.98 1.65 Southwest 1.02 0.39 1.001/ 2.41 TOTAL 1.80 4.02 2.36 8.18 lIIncludes 500,000 acres Coastal Embankment Project and 300,000 acres Genges - Kobadak Project. 75 have been considered in estimating tube-well projection for this study: (i) last 5 years average growth rate of tube- wells; and (ii) groundwater availability. Considering these factors, it is estimated that about 10,000 tube-wells and 10,000 frictional pumps are likely to be sunk by 1980, which would increase to 20,000 tube-wells, and 20,000 fric- tional pumps by 1985. The detailed estimates of irrigation possibilities under alternative set of assumptions and technological adjustments are given in Table 3.4. Credit Constraints The credit supply is extremely inadequate and often available at an interest rate as high as 30 percent to 40 percent in the noninstitutional market. The supply of credit from institutional sources accounted for only 14 percent while 86 percent is available from noninstitutional sources like the well-to-do rural people and friends, etc. The total credit constraint can be specified as Lij 5 Aij - Rij i = crops j = regions where Lij = potential credit available to farmers for use in the ith crop in jth region from both institutional and noninsti- tutional sources; A.. = area under ith cr0p and jth region; 13 76 Table 3.4-Proiected Irriggtion Availability’ouring Three Phases of Develgpment. 1975 Low Lift P s Tube Wells Shallow Tube Hamil Traditional Method Nos Acre Ft, 865. Acre Ft 95s Acre ft. 000 Esra ft Tbtil acre water 2/ 2/ water Water Water feet water Sumner - - 1 385 208 670 14 240 480 702 CENTRAL . Winter 9505 795 1385 208 670 14 765 1530 2547 Summer - - 275 41 295 6 600 1200 1247 EASTERN Winter 11473 960 275 41 295 6 775 1860 2867 SUMMDP - - 1050 158 1580 33 250 500 691 NORTHHEST Winter 3831 321 1050 158 1580 33 985 970 1482 Salli? - - 190 29 455 10 350 700 739 SOUTHHEST Winter 10191 853 190 29 455 10 467 934 1826 1 .2 Summer - - 2400 600- 2400 50 100 200 850 CENTRAL Winter 10700 1338 2400 600 2400 50 380 760 2748 Summer - - 400 100 400 8 700 1400 1508 EASTERN Winter 12710 1589 400 100 400 8 765 1865 3557 Sill“? - - 10200 2550 10200 213 - - 2763 NORTHUEST Winter 4400 550 10200 2550 10200 213 - 3 3313 SI”? - - 7000 1 750 7W0 146 1 00 200 2096 SOUTHHEST Winter 17190 2149 7000 1750 7000 146 100 200 4245 1/ The regionwisa powarpumps figures are three years average from 72-76 (sourca- Bangladesh Agril. in Statistics, Ministry of Agriculture) 2/ The nos of tubewolls and shallow tubawalls to be sunk in each region in 1985 were not available. The total national figure estimates are 20,000. The ragionwise fi ras wars approximated from the percentage of most favorable aeras exist in each region as outlined in the p an. The total prospective area in each region as mentioned in the p an .a.: 4000 sq. ads. in NH region . .66 percent x 20.000 total tubewalls to be sunk in 1965 - 10.200 nos. of tubewells 2600 sq. ads. in SW region - .36 percent x 20,000 total tubawells to be sunk in 1985 - 7,000 nos. of tubewells 500 q. mls. in Central region - . 07 percent x 20,000 total tubawalls to be sunk in 1985 - 2,400 nos. of tubawells 2] In total 150 days in the Winter season (December to April) it is observed that a power pumps and tubewells supply about 1004 and 1800 acre inchs of water per season respectively, and runs about 6.7 and 6 hr. each day. Considering the heavy underutilization of these resources and a recent emphasis by the government for increasing capacity utilization of pumps and tubewalls. it is inputad that per day utilization would be increased to 8 hr. and lo hr. during 1980 and 85 respectively, (i.e. supply would be 1200 and l500 acre inch of water through power pumps and 2400 and 3000 through tubawalls.) 5/ One of pump operate about 502 hr. per season and supply about equivalent amount of water. So, if a pump operate about 502 hr. would supply about 250 acre inch of water in one season par shallow tubewalls. g/ A total about l400 thousand acres of land irri tad through traditional methods. Survey in Commilla shows that about 74: of land irrigated through traditiona methods. Which comes about 6 to 7 thousand acres of land. Source: Winter Croo Survey in Coanlla Kotwali Thana 6. N. Soloiman, BARD, Aug., 1970, p. 10 6/ Since the Eastern region is not suitable fbr1tuba~wcll irrigation, it is assumed that the traditional method of irrigation would continue in Plan II at the same rate as in Plan I. 77 Rij = average per crop credit supply under ith crop and jth region. The average supply of credit per crop acre for both local and irri-varieties are ob- tained from the Survey Reports under- taken by the Cooperative directorate and BIDE (75,2). According to these surveys, a farm family on an average, is able to secure a per acre loan of Rs. 200 for local varieties and Rs. 400 for HYV from both institu- tional and noninstitutional sources. Thus, the total potential credit availability for the entire country is arrived at by adding the estimates of credit supply of different regions and over different crops. That is: '2‘ 'i‘ TNSC = L.. ; i=1 i=1 ‘3 where TNSC = total national credit supply. On the basis of these equations, the estimated credit con- straints for Plan I comes to about Rs.6830 million. For Plan II TNSC is assumed to be 30 percent higher than Plan I. The per acre credit availability for other cash crop is also assumed to be equivalent to the credit supply of local varieties. 78 Behavioural Constraints, Consumption Constraints The food consumption levels of the model were exo- genously determined. The household food consumption and nutrition survey conducted jointly by USAID and the Univer- sity of Dacca serves as the basis for estimating total rural consumption constraints. (Table 3.5.) Nonnegative Constraints None of the activities discussed above can be oper- ated at negative levels. Acreage Constraints Estimates of acreage for different varieties of rice and other crops included in the model were made sepa- rately for two levels of technology. (i) TECH I: For Plan I solution, the acreage of each crop activity has been constructed by taking a mean of last three years' actual harvested acreage statistics received from the Bureau of Agricultural Statistics. (ii) TECH 11 (Plan II); The only meaningful way the total production can be increased is through transfer of land from local to HYV varieties as well as by in- creasing the cropping intensity. The FAO/UNDP Soil Survey Scheme has identified suitable areas for 79 Table 3.5. Estimated Subsistence Consumption Requirement of Farm Population During 1975, 1980 and l985_ ' Crops/ Per Capitag/ 1975 1980 1985 Commodities Consumption (in (in (in per day million million million (in ounce) tons) tons) tons) Rice .9375 7.7570 8.9916 10.3221 Wheat .0625 0.0517 0.0599 0.0668 Potato .09375 0.7757 0.8992 1.0322 Sugar .0625 0.0517 0.0599 0.0668 l/Farm population for 1980 and 1985 were projected on the basis of 3 percent and 2.8 percent growth rate, respectively. Current farm population was estimated from total number of farms[(2,l76,400/3 acres) = 7,254,667 num- ber of farms x 7 members each farm holding = 50.78 (million)]. $0 for l980 and l985 it comes about 58.86 and 67.57 million, respectively. g/Household consumption survey conducted by USAID and Dacca University. Source: A. Jalimian - Consumption Requirement for a Balanced Diet in Bangladesh 1975-80. Ministry of Agriculture, Government of Bangladesh. 80 agricultural intensification (see Chapter II for details). Consequently, the areas delineated as suitable for different crops serve the basis for acreage constraints in Plan II. CHAPTER IV GENERATION 0F INPUT-OUTPUT COEFFICIENTS In Chapter III alinear programming model was set forth to determine the optimal land use pattern which would serve as a guide to profitable crop adjustments in four regions of Bangladesh. In this Chapter the input-output coefficients and prices of inputs and outputs used in the model are estimated. The first part of this Chapter dis- cusses the problems associated with the estimation of input-output coefficients and procedures used by different researchers to alleviate these problems. The sources of data used in this study are also mentioned in this section. The second part of this Chapter describes the estimation procedures of input-output coefficients for all crop acti- vities incorporated in this model. The main inputs to the agricultural production process are expressed in terms of land, human labor, draft animal power, fertilizers, water, seeds, and pesticides and accounted for by assigning unit costs to each of these items. Output is expressed in terms of crop yields and accounted for at the minimum-procurement price fixed by the government for each commodity. 81 82 The application of LP involves the accurate estimate of input-output coefficients and prices of inputs and outputs. However, the extensive and detailed enterprise budgeting information needed for constructing input-output coefficients were difficult to obtain. The problems often become complicated when the data is needed for macro model, because the input-output coefficients of a particular crop is usually affected by climate, soil differences, and re- source availabilities. In addition, the principal cropped areas among regions exhibit considerable dissimilarities in terms of absorption of inputs and production of outputs. The degree of information required to take into account all these differences is hardly possible by available means. Thus, budgets and technical coefficients underlying the L.P. model were either (I) assembled and synthesized from information gained at meetings with personnel from the applied and technical agricultural services or (2) obtained from research publications. For macro models the data are best suited if they are collected in terms of 'typical or representative' farm-firms. Heady (2l) has mentioned that where the solution is to provide for 'typical or represen- tative' farm-firm the best sources of data are likely to be census data, farm record surveys and technical studies of production relations, information from extension person- nel and conference with interested farmers. Considering the relative advantage of this approach over others, for 83 the purpose of this study the input-output coefficients are estimated in terms of average 'typical or representative farm' in Bangladesh. Brief details about the representative farm approach including the sources of data are described below. The Representative Farm The input-output coefficients generated for the model were in terms of average representative farms in the four regions of Bangladesh. The development of a "representative" farm involves the estimates of input-output relationships, prices of inputs and outputs, net returns obtained from different crop enterprises, and the level of available resources relevant for this type of farm. The "Synthetic Firm Technique" is used to make these estimates. This technique allows the development of average represen- tative (or hypothetical) farms by using estimates based on data the researcher judges to be most relevant under the specified conditions. The synthetic farms developed here are designed to be "representative" under the physical and economic environment it operates. This is "representative" in the sense that it displays identical internal and external characteristics with respect to resource endow- ments, production facilities, production constraints en- countered, input-output relationships, and input and pro- duct market situations. 84 The estimates were made by examining data from a number of sources and making judgements based on these data. The data were obtained from many sources: published and unpublished reports and studies from: (i) government minis- tries; (ii) semi-government agencies like ADC and HAPDA, and (iii) educational and research institutes, e.g., Rice Research Institutes (BRRI), Mymensing Agricultural Univer- sity, Academyifin~Rural Development (Comilla), and Institute for Development Studies (BIDS), etc. However, certain input coefficient requirements which are of strategic sig- nificance with respect to output determination, such as use of fertilizer, were estimated from the data obtained from fertilizer experimental stations. The required input coefficients for all crop acti- vities incorporated in the model are discussed below: (”L119 The resulting coefficients(Table 3,1,Chapter 111) are simply a description of the period (i.e., Summer and Winter) during which a particular crop occupies the land. (ii) Fertilizer input and yield The input coefficient of fertilizers for each crop activity is contingent upon a wide range of variables such as crop varieties grown, the availability of irrigation water, regional climatic differences, and the availability of auxiliary inputs such as credit and extension programs. The following procedures were followed to generate 85 fertilizer yield coefficients for use in the programming analysis at two alternative levels of technology -— TECH I (BENCHMARK), and TECH II (l985). (a) TECH I (BENCHMARK) Data on existing farmer's fertilizer applica- tion rate for different crop activities were not available but were imputed by relating monthly fertilizer sales to cropping patterns.8 The yield rate for all crops were estimated by taking averages of the last three years' yield received by the farmers during l972-75. (b) TECH II The fertilizer input-output coefficients and yields of different crops were estimated from the secondary data of crop-fertilizer trials received from the Soil Fer- tility Institute and Comilla Academy.* 8The per-acre use fertilizer for different crops is imputed from BADC's monthly fertilizer sales to cropping pattern during 1972-73. *The experiment was conducted through 200 fieldmen based on 200 thanas (out of a total of 417) throughout the country. Each fieldman is attached to 5 plots. The design is laid out in randomized block design. One advantageous feature of the randomized block is that experiemntal units are first sorted into homogenous groups called blocks. Such grouping of the experimental units tends to minimize experimental error within each block. Two variables were used where nitrogen was in the form of urea and K20(P) in the form of TSP. There were nine treatments and there were three observations on yield for each treatment. The selec- ted treatments were then allocated at random anong experi- mental units. In most of the experimens the surface region of interest was bounded by 0 S N S 200 and o S P S 200. 86 Various research studies indicate that by using statistical estimating procedures one can obtain reliable estimates of the relevant portions of the fertilizer-crop yield production surface to be able to predict input-output relationships at any relevant point on the surface. Ferti- lizer use depends on cultivators demand for it. Demand is an outcome of two decisions: whether to use fertilizer and how much to use. These decisions are governed by cultiva- tor returns. The cultivator return depends on: (l) ferti- lizer production functions; (2) prices of crops, and (3) cost of fertilizer. Understanding these three factors, the farmer will demand the optimum amount of fertilizer. In Bangladesh, the Rice Research Institute and FAO Soil Fertility Investigation Project conducted a{series of fer- tilizer experiments on different varieties of rice and other crops in four regions. The fertilizer experiments were done both in experimental and farmers' plots and on repre- sentative soils which had the widest possible relevance to established soil conditions. In the experiments, all resources or inputs except ferti- lizer, were held constant. The crop was grown under con- trolled irrigation. In all, there were (2x9x3) = 54 plots, each of which was the siée l/40 of an acre. For grain yield determination, lOm per plot was harvested from the center of each plot. Yields were calculated and expressed in terms of a per maund per acre basis. The average yield of the experiment was made available to the author from where the fertilizer coefficients and yields for different crops were estimated. 87 The average results of these trials were used to obtain the various optimal combination of plant nutrients. The yield of the fertilizer trials obtained in the experi- mental stations were deflated by 30 percent to make it realistic vis-a-vis farmers' plot. The yield was deflated because research results based on experiment data often overestimate yield forthcoming on a farmer’s plot where nontreatment variables are not controllable to the same degree as in the experiment plot. In Bangladesh no such experiment has been conducted thus far, so as to provide an indicative figure regarding the difference in yield between the experimental and representative plots (farmers' plot). Similar sorts of experiments have, however, been conducted in the Philippines where it has been observed that in the dry season the yield from representative plots would average l7 percent lower than the maximum attainable from experimental plots (25). Consequently, for the pur- pose of this study yield figures received from the experi- mental stations were deflated by 30 percent. The func- tional forms used to determine the optimum doses of ferti- lizers for different crops were as follows: _ v Y — x + Bl" + 32" .... (l) _ v O Y - x + BIN + BZP + B3N + Bap + 85" P .... (2) where Y = yield 88 N and P = two plant nutrients -- nitrogen and phosphate NP = interaction term. Taking the partial derivatives of the two nutrients N and P in the equation (2) with respect to Y gives av = 5N b]+2b3N+b5P.... (3) BY _ to obtain economic optima in the case of two inputs, pro- fits would be maximized where the value of the marginal product (MPPx - P = VmPn) of each input is equal to the Y price (PM) of that input. This occurs where: 3V _ 5— Py — PN .... (5) VmpN = PN .... (6) Dividing (5) by Py gives P N =——.... (7) PY Q) .< 8 2 Utilizing the partial derivatives of the previous equation (3) ad (4) gives: b1 + 2b3N + bSP P N (8) b2 + 2b P + b N 4 5 (9) II «:15? 89 Solving by use of simultaneous equations which is in matrix notation Ax = b, where A is an n2 by n2 matrix of coefficients, X = col., (X1, X2) is a column vector of n2 variables and b = col., (b1, b2) is a column vector of the given constants in the n2 equations. On pre-multiplication by A"1 we obtain A-]Ax = A'lb since A'IA = I and since Ix = x, x=-A'1 b. Using Cramer's rule gives the optima combination of plant nutrients for a given set of product and factor prices. The regression equations obtained by fitting the two different functional forms were given in Table 4.1. The Table shows that the equation (2) in respect to all crops provide higher Rv than equation (1). Consequently, equation (2) was used to obtain the optimal input-output coefficients for each crop activity. In case of certain crops like T. Aman, where equation (2) could not provide satisfactory results due to high multi-collinearity between N and P only equation (l) was used omitting variable P. The optimum profit points were analyzed with common price levels of Py = Rs - 120.00 per maund (see factor and pro- duct price of this Chapter), Pn = Rs - 0.51 and Pp = R5 - 1.38. The cost of nitrogen and phosphate 9() Table 4.1. Estimated Regression Coefficients for Fertilizer Crops Rv Sig ' 1. Boro (local) (8) 1 17.1044 + .1927N - .00096NV .51 .009 (2.2042 (.0668) (.0005) (b) v 17.0242 + .lZSON + .0661? - .001011U - .007?v + .OOOBNP .55 .10 (2.5443) (.1229) (.1284) (.0009) (.0009) (.00012) 2. IRRI Boro v 24.994 + .lSSSN - .0003Nv .60 .015 (3.6823) (.1249) (.0008) v 24.0610 + .1491N + .0085? - .0011):V - .0007?V + .OOlSNP .90 .006 (2.3777) (.1203) (.1185) (.0008) (.0002) (.0005) 3. T. Aman v 14.9544 + .2282N - .OOlBNV .69 .001 (1.3210) (.0711) (.0009) 7 14.8000 + .0159N + .1292? - .00032NV - .0004?V + .OOOOSNP .83 .003 4. Aman (HYV) v 15.1689 + .2131u - .000911v .52 .006 (2.6718) (.0831) (.0006) 1 14.9502 + .1647u + .0634? - .0014):v - .0008?v + .0014"? .62 .03 (2.7734) (.1012) (.1060) (.0010) (.0007) (.0013) 5. Jute v 7.9966 + .1456N - .00018N” .89 .0005 (1.0045) (.0544) (.003) v 7.8384 + .1634N + .0137? - .0013N’ - .0008?” + .ooozsu? .94 .0005 (.9996) (.0443) (.0663) (.0004) (.002) (.00015) 6. wheat v 7.097) + .1523N - .0005711v 79 .0005 (1.1272) (.0419) ( 00027) v 8.1757 + .09o7u + .0732? - .000711v + .0012?v + .0015"? .97 .0005 (.7388) (.0315) (.0558) (.00016) (.0008) (.0005) 91 considered for optimum profit points were the actual pro- curement and distribution costs incurred by the government. All the equations (in Table 4.1) were significant at 10 per- 2, and NP. The cent or less for coefficients of N, P, N equation for local Boro and HYV Aman yielded slightly higher doses of N and P than the current imputed doses computed from fertilizer sales to cropping pattern. The optimum doses, however, were below the recommended doses of 1:50:50 (i.e., 1 maund of nitrogen, 1/2 maund of phosphate and 1/2 maund potash) and 2:2:1 for local Boro and HYV Aman res- pectively. The estimated fertilizer doses for other crops are quite reasonable as well as consistent. The fertilizer input-output coefficients for each crop activity under alternative technological adjustments are shown in Table 4.2 and 4.3. In the L.P. model (Table 3.1, Chapter III), the P l to P coefficients for fertilizer (R to R40) reflect the 44 37 fertilizer input required to produce one acre of each crop in four regions of Bangladesh. The yield of each crop, expressed as production from one acre of crop harvesting activity (P1 to P44), which is entered in the balance equa- tion (R53 to R64) as the output transfer row. (iii) Labor Input Requirement One of the crucial problems in budgeting crop pro- duction is the estimation of labor input requirements. The significance of reliable estimates of labor use for 92 Table 4.2. Estimated Fertilizers Input - Requirement for Different Crops in Alternative Stage of Development Crops 1974-751’ 1980 1985 Fertilizer use Fertilizer use Fertilizer use Lb/acre Lb/acre Lb/acre Rice Aus (local) 10 12 15 Aus (HYV) 100 170 253 Aman B - - - Aman T 29 26 24 Aman (HYV) 120 150 190 Boro (local) 35 50 80 Boro (HYV) 150 170 253 Wheat 10 25 4O Jute 22 45 73 Potato 5 20 50 Sugarcane 73 80 110 1/The per acre use of fertilizers for different crops is imputed from ADC's monthly fertilizer sales to cropping patterns. It is actually the figure of 1972-73 as the monthly sales of later years are not available which is obviously higher. Table 4.3. 93 Estimated Yield Projection for Alternative Phases of Development (in acre/maund) 1970-751/ 1980 19853/ (5 yrs. ave.) Estimated Aus (local) 8.6 8.6 8.6 Aus (HYV) 30.3 34 40 Aman B 9.2 9.2 9.2 Aman T 12.5 13.5 15.0 Aman (HYV) 25.5 26.0 29.0 Boro (local) 15.0 18.0 22.0 Boro (HYV) 33.5 35 42.0 Nheat 9.2 10 11.0 Jute 13.78 13 13.0 Potato 105.0 111.30 117 Sugarcanei/ 473.0 473 500 l/Source: Bangladesh Agriculture in Statistics, Ministry of Agriculture, Gov't. of Bangladesh. Z/Estimated from the data obtained from Fertilizer Experi- mental Stations and Comilla Academy. 3/ — The 1970-75 yield data for sugarcane and jute appears to be highly inflated. Consequently, has not given much weight in estimating yield figures for 1970-75. 94 each crop activity can hardly be overemphasized due to the fact that it accounts for significant portion of the total production cost. In spite of such importance, dependable data on this particular aspect are scarce. Most of the farm management studies reveal that employment on HYV farms is about 1 1/2 times that of traditional rice farms (all on a per acre basis) (26). For the purpose of this study almost all of the crop production surveys9 so far underta- ken in different parts of Bangladesh were consulted and realistic estimates were made on the basis of these reports and my personal experience in the area. The labor input requirements were calculated per acre by crop for each month of the growing season. The monthwise labor require- ment for each crop activity is shown in Table 4.4. (iv) Bullock Power Requirement Bullock power is the traditional source of cultiva- tion in Bnagladesh. Mechanization has hardly been used; neither is its introduction on a significant scale contem- plated in the future. Most of the farm management studies conducted thus far have been consulted and bullock power requirements for each crop activity have been worked out. Table 4.5 shows the bullock power requirement for each crop activity. 9(i) Gumail Bill Survey report; (ii) Mymensing area Survey; (iii) Comilla data; (iv) Rural Development Project I Survey Data. 955 .m eN m_ _ p _ . _ _ N, NN _ oeeocooem Mo 8N NN N mm oueooa No, N. NN N. e. N. ooze no .N mN N N e. smog: o__ e. mN a. o. ca m A>>zv m=< om o. N_ p N NN N A_ooo_v a=< ea c. .N N m. NM 5 A>>zv e-eee< oN 8 ON N N, NN N h-eee< em 4H m. N N NN e N-eee< o_P m NN a. c. on m. A>>=V been we a a. e a NN e. .ooo_v oboe .uuoh uuo .>oz auo gum m=< pza caa xa: 25¢ co: .aou cow Amxoutco: :_v mqogu m=o_co> cc» mucoEoc.=co¢ can»; >_=u:cz ouaeNumm .v.v a—aah 96 Table 4.5. Estimated Bullock Power Requirements for Different Crop Activities in Bangladesh Crops , Bullock Power Requirement (in Bullock pair-days) Aus local 20 Aus HYV 26 B. Aman 20 T. Aman (local) 21 Aman HYV 24 Boro local 22 Boro va ‘ 26 Wheat 19 Sugarcane 25 Potato 23 Jute l7 97 (v) Hater Input Coefficients The climatic and weather conditions of different regions of Bangladesh vary considerably from each other. For this reason, the use of country-wide average water re- quirement coefficients for all regions would likely provide an erroneous policy conclusion regarding the availability as well as requirement of total quantum of water for various crops. Considering this fact, the irrigation water re- quirement for major crops estimated by Trafdar10 on the basis of the Penman approach was used for this analysis. The irrigation water requirements for crops were computed as: n if] Iri = 1)] Et] + Pdi + Lsi - Mb] - Re1 "here= . . . . .th Iri = Irrigation water requirement 1n the 1 period; Eti = Evapo-transpiration requirement in the ith period; Pdi = Deep percolation loss in the ith period; ”bi = Initial moisture at the beginning of the ith period; Re = Effective rainfall in the ith period. Monthly data on rainfall and evapo-transpiration have been 1°0p.oit., p. 116—120. 98 used. The reason for using this study is that it recognized the climatic and weather differences between different regions and estimated water requirements for each crop grown in high, medium and low rainfall regions sepa- rately (shown in Table 4.6). Indicative figures have been taken from the theoretical estimates and the results used as region-wide averages of the depth of irrigation water required to bring a crop to maturity. The estimates of irrigation water requirements were divided into two time periods: (i) Summer; and (ii) Winter, which correspond to the cropping pattern of the country. (vi) Credit Requirement A clear consensus of opinion as to the actual quantum of credit required per acre for each crop activity is not available. A number of studies mentioned a lump sum requirement for each farm family. The Cooperative Survey StudiesH indicated the required quantum of credit by per family ranging from Rs. 178 to Rs. 475 in the Eastern re- gion, Rs. 185 to Rs. 271 in the Central region and Rs. 278 to Rs. 300 and Rs. 135 to Rs. 221 in the Southwest and Northwest regions, respectively. The study was, however, representative of the requirement for credit only for indi- genous crop varieties, as the survey was undertaken at a time when HYV was barely introduced. Understanding this HAgril. Credit in East Pakistan: A Survey, 1966. Op.Cit., p. 61-70. 99 Table 4.6. Estimated Irrigation Requirement Per Acre (EgiflfisLii’offigliflésfifi ”‘9“ Ram” (Acre feet of water) Low rainfall Medium rainfall High rainfall region region region Aus (HYV 2.75 2.6 2.4 Aus (local) - - - Aman B - - - Aman T - - - Aman (HYV) 1.5 1.0 0.167 Boro (HYV) 2.75 2.6 2.4 Boro (local) - - - Nheat 1.6 1.0 0.9 Sugarcane - - - Potato 1.5 1.0 0.96 Jute - - - l/Source: Trafdar R. Islam - Low Lift Pump Irrigation in Bangladesh - Ph.D. Thesis in Michigan State University, 1973. p. 105-122. 100 problem the credit requirement was assumed to be 30 per- cent of production cost of each crop activity. Factor and Product Prices and Cost of Production Factor Prices The concept of cost that is relevant for economic analysis differs considerably from that which is relevant for private profitability analysis. The gap between social and private cost increased or decreased depending on extent to which imperfection and externality existed in the market- ing mechanism and inhibited the free play of market forces. In Bangladesh a significant degree of distortion exists in factor market prices due to large transfer payments made to farmers through heavy subsidization of all inputs. Throughout the last decade it has been the practice of the government to provide modern inputs to farmers either free or at heavily subsidized prices. The extent of subsidy allowed to farmers is estimated to range from 56 percent to 65 percent for fertilizers, and between 60 percent to 65 percent and 40 percent to 50 percent for pesticides and irrigation water, respectively. Recently, the government has emphasized the need for decreasing subsidies toward their gradual elimination. But the rate at which it would be reduced and the time it would take for its elimination were very difficult to pre- dict. Such a proposal has often in the past been deemed 101 economically feasible but regarded as politically vul— nerable. Skepticism often shrouded the minds of the deci- sion maker regarding its impact on demand for inputs, per- formance level of farmers and thereby on production. All these problems pose a serious difficulty in projecting different input prices during alternative phases of develop- ment. However, it has been observed that the farmers have recognized the importance of fertilizers for crop production and evidence indicates that they are willing to pay higher prices for it. With these perspectives, for the purpose of this study it is assumed that the present rate of subsidy for pesticides and water will continue whereas fertilizer subsidies would be eliminated during terminal phase of dev- elopment (i.e., 1985). This would mean that farmers would be charged at Rs. 8.00 per pound of pesticides and Rs. 80.00 per acre feet of irrigation water. Labor is the single most important input. Produc- tion costs fluctuate widely due to variation in wage rates. It has been observed that because of seasonal peak demand, the wage rate increased to Rs. 9.00 in 1975. The data of the Bureau of Statistics indicate that the average wage rate of agricultural labor in seasonal peak demand oscil- lated around Rs. 8.00 during the last two years (1974-1975). Consequently, the wage rate of Rs. 8.00 per day was consi- dered for the model. 102 Product Prices The relevant price that is usually accounted for in determining returns from crop activities is what farmers actually receive for their crops, that is, the farm gate price. But, the farm gate price is not collected in Bang- ladesh. Only wholesale prices were collected from district markets which are compiled and published in the form of national averages. Estimates for farmers selling prices for outputs in the model were, therefore, based on average wholesale prices. There are several other difficulties regarding which product prices are to be considered for estimating farmer returns. It has been observed that prices for rice went up from Rs. 37 in 1970 to as high as Rs. 300 in 1974. The prices for other commodities also went up in a similar fashion. The reasons for abnormal increases in prices were (i) scarcity and high prices of production inputs in the international market; (ii) breakdown and suspension of production in local fertilizer factories; (iii) high rates of inflation both at home and abroad; (iv) large smuggling of food grains across the border, etc. The logical conse- quence of these facts suggests that any price projection will hardly be expected to provide a realistic estimate for economic analysis of the future. However, to alleviate the wide fluctuations in the product market, the government has fixed the price of rice at Rs. 120 in 1975. Recent 103 conversations with officials of Bangladesh's Planning Commission and the World Bank indicate that the present market price for rice is more or less stabilized around the price fixed by the government. Like rice, the jute price has also undergone simi- lar sorts of fluctuation. Jute is the primary cash crop and produced mainly for export. .Consequently, past govern- ment policy was to stabilize the jute market by fixing a statutory minimum internal price to enable it to compete in the international market. The statutory minimum floor price for jute is at three levels; primary (farmer's level), secondary and terminal markets are fixed by the government at the beginning of each jute season. There are, however, controversies and complaints regarding jute price policy, as it is often alleged that the farmers do not receive the minimum prices prescribed by the government. For the pur- pose of this study, however, the current statutory minimum prices fixed for rice was considered for analysis. However, for jute the current market price was considered. Regarding wheat, its total internal production is very insignificant in comparison to rice. The larger per- centage of wheat is usually received either as commodity aid or procured from abroad and distributed at a government controlled price. The farm gate price of wheat appears to be closely related to the controlled price in Dacca. The present controlled price of wheat is Rs. 55.00 per maund. 104 Similarly, the sugarcane price is determined by the pro- curement price fixed by the sugar mills which is Rs. 4.00 per maund at present. The farm gate price (FGP) is con- sidered for potatoes. The FGP is derived by deducting the price spread between farm yard and wholesale markets from the average annual wholesale price. Based on factor price assumptions mentioned earlier, the growing and harvesting cost of each crop activity is estimated (Appendix A,B). The product prices are incorpora- ted directly into buying, selling and consumption activities of the model in order to estimate farmer returns. CHAPTER V OPTIMUM CROPPING ORGANIZATION UNDER ALTERNATIVE TECHNOLOGIES: IMPLICATIONS FOR FARM INCOME, LAND USE AND RESOURCE UTILIZATION Chapter I has provided a conceptual framework for the economies of resource use on the basis of regional com- parative advantage. In Chapter III a detailed programming model was elaborated to operationalize that conceptual framework. In light of that model two alternative plans were developed representing two technological levels. In Plan I, the net returns were optimized with existing national and regional resources and existing maximum crop acreage constraints. These constraints were determined on the basis of an average of last three years. In Plan II, at higher technological levels, the net returns were optimized with limits on maximum acreages suitable for different crops as identified by the FAO/UNDP soil reconnaissance survey and projected regional resource constraints for 1985 as esti- mated in Chapter III. The purpose of this Chapter is to present the basic results of the analysis of these two models. The first part of the Chapter employing Plan I focuses on three topics: (i) the possibilities of increasing net national farm income 105 106 through improved allocation of existing resources, (ii) the determining of an optimal land use pattern under existing sets of regional resource constraints; and (iii) the extent of resource use such as land, labor, etc., under this plan. The second part of the Chapter employing Plan II focuses on (i) the possibilities of increasing net national farm in- come through efficient allocation of projected resources; (ii) the determining of optimal land use patterns, and (iii) the utilization capacity of the Bangladesh agricul- ture with the estimated resources for 1985. Optimum Organization with Existing Resource Constraints [(Technology-I) (Plan 1)} In exploring the characteristics of the optimum organization in Plan I the model results were examined by focusing on three separate economic criterions, viz (i) changes in net national farm income to fixed farm re- sources; (ii) land use and the cropping pattern of food grains and commercial crops; and (iii) total resource use, i.e., land, labor, fertilizer, etc. A. Returns to Fixed Factors in Farming The model results indicate that there exists signi- ficant scope for increasing national agricultural farm in- comes by devoting more lands to HYV cultivation (Table 5.1). In this table the relative performance of the agricultural 107 sector of the nation is measured on the following economic criterions: return on fixed factors, net return per acre, net return per man-days and net return per unit of borrowed capital. The net returns to fixed factors obtained in the optimal organization is Rs. 43127.994 million as against an estimated Rs. 3878312 million from the existing plan, a 11.2 percent increase. The average per acre income of the entire crop mix combined, as estimated in the plan is Rs. 1526.49. The resulting crop plan is specialized in rice production, particularly in different HYV's including local improved varieties due to their comparatively higher net returns than other competing crops. 12The data for Net National Farm Income (NNFI) actually derived from the eleven crops during 1972-1975 is not available. Consequently, the NNFI for these years were indirectly estimated on the basis of the following equation with a view to making an income comparison between the actual (average of 1972-75) and the optimal organization in Plan I. n n E NNFI = _2 (Pin, - dit) i 1 i=1 where Pi = price of the ith crop assumed or esti- mated in the model, Qi = actual last 3 years average production of ith crop dit = the variable cost of ith crop as esti- mated in the model. 1(38 —m.o “mm c_v—Nu_aou con cczumm umz mm.NNAmm :_onuicoz con ccaumm umz m¢.o~m_ A.mm c_v mcu< gwq www.mm—me NmNMON mw_oom Naomvm ooo~om~ vmnmwm caesvmm ommmmm cm.mo o.mmp mwmn mcczumm umz A.ma cowppvsv mcczumm umz cw :. ummzcuaom Lousy” cou:_3 ummxgusoz seesaw :cmummu “Wynn“ co :. pmeucau towel“ Au» mauoz gonouuo consouqmm amam=< x—aa «and wcza mm: .Nee< saga: zeoacnmu xcoacmn cmu:_z cmEE:m cou:_3 possum cmucwz cmsaam cmu:_: gossam- eees ._ ummzzuaom umeogusoz ccmumou ooo _ocucmu mucmeumcou mmmm N A_eeo_seev beech meoN__NNeoa .e gone. xw_Eau .N mmm omm emm use ~mo~ mmmN m_op momm .Nem v.~mm moms Ame ommo A : amm— m.ao— unm— wmom ommm 1' musw canyon mcmucnmam “was: >>= ocom _ouo_ osom >>= ces< cme<.h coa<.m >>z ma< _auo_ mz< cue ooov momwcmcmucm mocu .e moons; o_——_EV u_vmcu zoLcom .m Amocue mcouv mm_an cyan anon: ouauom occucomam mu_¢ Aooov .1- mzo» mcv__mm .N Langmuma e acezcoa mmmmeiemmwm mAcuizoz a vmmogucsa .— mw*u*>_uu< to» cope mc=u~:u_cm< A_ eases _-xoo_oc;ump . :mmvo_m:mm Po:o_uoz mo :o_uoN_comco _os_uao mo mu_umvsmuumcmzu .—.m m—am» 109 In the optimal plan, a total production of about 13,215 thousand tons of rice is obtained as against an actual last five years average production of 10,700 thousand tons, an increase of 23.5 percent. Out of a total estimated rice production of 13,215 thousand tons, 7829 thousand tons were consumed and 5386 thousand tons were sold in the market. A total quantity of 161.8 thousand tons of wheat was produced, of which 52 thousand tons were consumed and 109.8 thousand tons were sold. Sugarcane production was, however, de- creased from an actual production of 6635 thousand tons in 1974-75 to 3628 thousand tons in the optimal plan, an approximate decrease of 45 percent. This has happened par- ticularly due to very low market price for sugarcane, com- pared to rice, both at farm as well as factory gate. The entire sugarcane production was sold in the market, obviously to sugarcane industries. The optimum plan also indicates that there is great potential for enhancing income by in- creasing potato production. Over the years cultivation of potato has become an increasingly profitable crop. Its pro- duction has increased from 348 thousand tons in 1961-65 to 866 thousand tons in 1974-75. In the optimum plan a total production of 2132.6 thousand tons of potato was estimated, of which 195.6 thousand tons were consumed at farm level and the rest 1937 thousand tons were sold in the market. The return per man-days to labor obtained in the optimal plan is Rs. 22.85 as against actual wage rate of 110 Rs. 8.0 in agriculture and Rs. 10 in nonagricultural sector. The model results indicate that (a) first, productivity of agricultural labor force could be sufficiently high if re- sources are allocated efficiently in between different crops according to their highest returns and (b) second, agricul- ture could be a profitable enterprise in comparison to other competitive sectors if it is properly organized. B. Optimal Regional Land Use and Cropping Patterns The previous section described the results of opti- mal organization of national agricultural plan and its im- pact on net national income. This section focuses on opti- mum land use pattern of each region under different crops. In the optimum plan food grains increased their share in the cropping pattern in all regions except the Northwest where the area under food grains declined from the existing situation (average of last three years as estimated in Chapter III) by 15.4 percent (Table 5.2). In the optimum plan the total area under foodgrains in Central (CN), Eastern (EST), and Southwest (SW) regions registered an in- crease of 37 percent, 7 percent, and-15 percent respectively over existing plan with an overall national increase of 13 percent. The cropping pattern obtained from the model shows that the food grains would account for about 95 percent of the total cultivated area as compared to 90 percent in the 111 eemNmNN OOQGON. Neommmn NNNNNNN N.N.N.v mN_NmNNN NNN_mNeN .N . «N _esoe - - Ne A-V _NeNNN_ ceomemN ...eo_aeev.eao» - - - - - Na .-v cocoa oooN_e aeoseasom - - - - - - e. .-v ceNNNN cocooa aaoseusoz - - - - - on “.3 ..memm sac—8N etoueeu - - - - - NNe .-v oceam. coo—5N .esueou mmmuw .awucoeecu Am N.m_ ..V N.oemNN NNNNNNN NN A+. someNeeN NNNNNNNN ...eo.aeev_ehoe mmN A43 oemamNN cooch_ N.N. A-V nee_mN NNNoee _m_.+v maammNN mmneeno amuseasom mNm .4. cmN_aN cccm.. °.Na A.“ e_NomN mnemN. ..m. .-. cocoeam mneNpce aaozeasoz NN ‘4. NoquN ooca.. N.oN ..v _m_oNN— oocmNe_ N ... N_eopoN accomme eeoaneu em A4. Nememo cocoa. N.cN ..V o.NNNo_ cocaNN NN A+V NNeNNoe cQONNNe .esueou m:_ogm uoom A< 1 nomuucuoc so (womaocuou so i omoosumc co M+ nomaocuc. »o + oncogu=_ so m+ amousuc. we «moucousoa asepuqo m=_um.xu mooucuucoa e:s_uac mepumwxu oooucougum e:s_uao oc.umwxu mo_uo.ca> uo>ccqe_ .aqu mo_uo_co> >>z pouch eos< emanate .auo» «co—aux Amucuu c. aoc>= econ .8 .-. Nee—“N oocmee Nee.mN cocoo— - agave o eoemon - eeeoN_ 5.: 45< a .4. e°°NmmN caeemme oceNe.N ccoe°.N cacaoNN cocoa—N ceoeoe. oeceoeN eeo..e. oocao.. ..eoo.. as< _~m_aqo a=.um.xu —as—uac mcwum_xu peevuao m:..m.xu .aa.uac m:_um.xu pee—ago ncwum.xu .aNOp Fuco.uom)11|1 umuxguaom umozgucoz ccoumou pacacou Amocuo c. nocc. snows—ace: mo m=o_mo¢ Lac. :4 .-zuop cane: change; on: use. aas_uao uco me.um_xu .m.m opac— 114 from 1.6 million acres in the existing situation to 2.09 million acres in the optimal plan, whereas the local T. Aman crop has declined by 15.7 percent. The released acreage from local T. Aman is transferred to Aman Irri pro- duction, the area under which has increased from 0.96 mil- lion acres to 1.61 million acres in the optimal plan, an approximate increase of 68 percent. The total area brought under HYV rice cultivation is 3.85 million acres in the optimal plan as against 3.2 million in actual situation. ut also appears from Table 5.2 that the adoption rate of HYV is not uniform in all regions, ranging from 20.7 percent in CN and EST regions to 98 percent increase in the NW re- gions. The latter region has high potential for bringing more areas under HYV because of the availability of adequate ground water. During the last three years, the total area of HYV in NW region has increased from 0.259 million acres in 1969-70 to 0.649 million acres in 1973-74. The total area of HYV both in actual and optimum plan were, however, much greater in the eastern region. The local improved rice variety (LIV) has increased in all regions. The total national LIV acres has increased from 1.2 million acres to 2.89 million acres in the optimum plan with a share of 579 and 795 percent increaseS'hiNW and SW regions respectively. The CN region appears to have a comparative advantage in producing wheat. In the optimum plan, the entire wheat pro- duction has taken place in the CN region. 115 Potato cultivation appears to be the most profitable of cash crops. In the optimum plan the total acreage ob- tained under potato was 0.55 million as compared to .231 million acres in the existing plan, whereas sugarcane area had decreased from 0.38 million acres to 0.22 million acres in the optimum plan. About 90 percent of sugarcane culti- vation is taking place in the NW region and the remaining 10 percent in the EST region. Jute, the principal cash crop and foreign exchange earner for Bangladesh, was found to be the most unprofitable among cash crops, given the technical coefficients and prices used in the model. The actual situation of jute corraborates the findings of optimal plan. The actual acreage under jute has decreased from 2.28 mil- lion acres in 1965-70 to 1.95 million acres in 1970-75. The per acre yield of jute has also declined from 3.6 bales/acre in 1950-55 to 2.7 bales/acre in 1970-75. In spite of signi- ficant importance to the national economy, the government policy for jute over the last two decades fails to provide sufficient incentives to farmers for jute production or to make it more viable and competitive vis-a-vis other crops. In the optimum plan, the total jute acreage has declined to 0.55 million as against 1.95 million acreage in the existing situation, a decrease of 71.6 percent. About two thirds of jute is produced in the NW region and the remaining one- third is located in the Eastern region. 116 C. Utilization and Marginal Value Products of Resources The extent to which available regional resource en- dowments (i.e., land, water) of each region are utilized for production of different crop activities and the estimated marginal value product (MVP) of each unit of resource used is given in Table 5.4. The total cultivable land increased from 26.35 million acres to 28.47 million acres in the optimal plan. The proportion of total available summer land utilized ranged from 79 percent in the NW region to 96 per- cent in the EST region, whereas winter land has been fully utilized in all regions except in the EST region. The short supply of winter lands in the CN, NW and SW regions are re- flected by their positive shadow price or marginal value pro- duct (MVP) in column 6 of the Table. The MVP indicates the possible gains in income through acquisition of scarce re- sources. The MVP's of winter land ranged from Rs. 52 in the CN region to Rs. 913 in the NW region. Thus one acre in- crease in winter land disposal in the CN and NW regions, with land remaining constant at 2.0 million and 1.40 million acres in those regions, profit would decrease by Rs. 52 and Rs. 913 respectively and vice-versa. The more limiting the resources the higher the MVP. The positive MVP indicates that the farmer would find it profitable to acquire that scarce resource if MVP > MFC of that resource. The optimal solution is also provided (lower and upper activity) or 117 Table 5.4. Resource Availability Levels and Their Utilization Under Optimum Plan Technology-l Unit Resource Resource Percentage of Shadow Price Income Penalty for (acres Availability Utilized Utilization for activity Activity No: in in solution Solution (in as.) {in Rs.) 1 2 3 4 5 7 8 1. and LEllie" land c."t,.‘ 5463000 4827424 88 o 80 7.99 Hintar ieno 2000000 2000000 100 o Sun-er land “n." 5537000 5297379 96 o 7.9 1116 winter 1end 2087000 2086862 99.9 0 7.9 563 Summer land N rthwest 6915000 5514325 79 0 111 31 uinter land ° 1405000 1405000 100 913 0 0 Summer land 5 th'.,t 6670000 6346000 95 o 552 198 uinter land °" 1000000 1000000 100 597 0 o 2. Perm n n labor Nan-days ‘31?02%7&""" 396104826 106105431 27 0 3.17 .58 February 396104826 170356165 43 0 .28 26.8 March 396104826 55179622 14 o .46 17.67 April 396104826 89163280 22 0 7.85 1.83 May 396104826 331363663 84 0 4.13 1.18 June 396104826 89018144 22 o .50 2.33 July 396104826 175379732 44 o 3.20 .26 August 396104826 383224029 97 0 6.77 69 Seote-ber 396104826 187817827 47 o 5.33 6.26 October 396104826 107252726 27 0 8.0 7.99 November 396104826 139823170 35 o .49 4.47 December 396104826 52627440 13 o 8.0 1.08 3. Ni r Nan-days oanuary 7922070 0 o 4 8. . . o 8.0 . ‘ o 8. December 7922070 0 o ‘- £232158l_11531111851 "IUflGS Central region 3030000 3030000 100 1985 0 0 testern region 4263000 4263000 100 670 o 0 wertnoeet region 2229950 2229950 100 1612 0 0 Southwest region 1856400 1856400 100 2147 0 0 5. rri ati n we r Acre/ r feet 702000 335236 48 0 80 94 81::5r C'""" 2547000 2547000 100 68 0 0 Summer 1247000 322784 2 0 3967 - winter """“ 2867000 38:7:g0 1?: 3099 363 - sue-er 6910000 4 6 - Minter '°"”""‘ 1482000 900185 61 g 59 Egg on r 739000 - -- ' ~ 11nt1r 5°“‘”"“ 1826000 703752 38 6 109 257 . ull w r 6 1:55375T2755Ton 343940856 145426485 42 0 1-72 “39 Eastern region 323221584 156664366 48 0 .32 28.24 Northwest region 310546152 140207706 45 0 6.85 1.95 Southwest region 241075464 166263772 65 0 23.74 12-68 7. grggig Rupees million 6830 6830 100 .48 o 0 118 range over which the shadow prices of Rs. 52 and Rs. 913 are relevant. One of the important uses of linear programming is that it can estimate the shadow prices of excluded activi- ties. Columns 7 and 8 of the Table show the shadow prices of the excluded activities. It indicates cost of forcing of an extra unit of activity into the solution. The shadow prices of the excluded activities also provide information regarding competitive position of these activities in the optimal solution. The income penalty of summer land in the Eastern region ranges from Rs. 7.9 on the lower side to Rs. 1116 on the upper side. The range over which the in- come penalty extends is obtained in the optimal solution [(5267449 acres on the lower side - to 5297379 on the upper side) -- Appendix Optimum Solution]. This means that on the upper side any quantity of land could be brought under cul- tivation at a loss of Rs. 1116, whereas reducing the culti- vation of summer land below the 5267449 acres used in the plan would involve an income penalty of only Rs. 7.9 for each acre of land withdrawn. The lower the income penalty, the higher is the competitive position of that activity to enter into the optimum solution and vice-versa. The permanent labor force employed in the optimum plan ranged from 13 percent in the month of December to 97 percent in August. August is usually the peak month for employment, as it is the harvesting period of 119 B. Aus - T. Aus; and sowing period for local transplanted and Irri Aman, the two largest crops in Bangladesh which account for about 30 percent and 56 percent of the total rice crop respectively. The income penalty likely to be incurred for using an additional unit labor beyond the opti- mum employment given in the solution is Rs. .69 for August labor, as against 1.08 for December labor. In other words, to force an increase of one man-day of labor employment beyond 52627440 man-days in the month of December would re- duce profit by Rs. 1.08 in the optimal plan. The magnitude of labor employed or remaining surplus as shown in the Table, however, does not represent the actual employment or unemployment figures in agricultural sector, as there are about 7 million acres of land devoted to six other crops which were not incorporated into this model. However, the present study would help in providing a reasonable approxi- mation of the rate and magnitude of unemployment in differ- ent months of the year in agricultural sector. Employment is also relatively high in the months of May, July, and September, and absorb about 84 percent, 44 percent and 47 percent of the available labor supply during those months. The labor supply is not a limiting factor in the optimal solution and is reflected by their zero MVP's in column 6 of the Table. Fertilizers and winter irrigation water are critical elements in the production of HYV varieties and are scarce 120 in supply. Most farmers recognize and understand the impor- tance of these inputs for increasing production. In the optimal plan, the entire available fertilizers were exhaus- ted in all four regions. Its MVP ranged from Rs. 670 in the EST region to Rs. 2347 in the SW region. This indicates that it is profitable to use more fertilizer under assumed ferti- lizer prices and output conditions. Scarcity of winter irrigation water in the CN and EST regions is reflected by their positive MVP's of Rs. 68 and Rs. 1099 respectively in the Table. However, the same is in excess supply in the NW and SW regions as indicated by their zero MVP's; whereas in actual situation irrigation water is scarce in those regions. The optimum plan shows that there is misutilization of irrigation water in those regions and either profit could be increased or cost of pro- duction could be decreased by efficient allocation of water resources. The last few years data confirm these findings which show that the area coverage of per cusec pumps has declined from 38.5 acres in 1962-63 to 20.0 acres in 1974-75. Borrowed capital appears to be not very profitable, as anticipated. Bullock power is in excess supply in all four regions and is reflected by the zero MVP's in Table 5.4. 121 Optimal Organization with Technological Change [(Technology II)*(Plan 11)]? 1985 The first part of this Chapter shows to what extent the optimal allocation of existing regional resources under present state of technology would increase the net national farm income and change existing land use patterns. This section provides the empirical findings of the differen- tial impact of technological change (as defined in the model -- see Chapter I for details) on (a) net national farm income, (b) regional land use and cropping pattern and (c) resource utilization. The latter part of this section will present a comparative analysis of potential utilization capacity of Bangladesh agriculture under the two technolo- gical levels in making use of total land, as identified suitable for different crops by FAO/UNDP. A. Net Returns to Fixed Farm Resources Under Technological Change (Plan II) The relative performance of the optimal organization of national agriculture is examined using the following economic measures: net national farm income, net return per acre, net return per man-days and net return per unit of capital. The model result indicates that there exists a sig- nificant scope for increasing net national income through adoption of the new HYV Aman varieties and the extension of existing HYV varieties. In the optimal organization under 122 Plan II, net national income is estimated at Rs. 65465.97 million as against Rs. 43127.99 million under Plan I, an increase of approximately 52 percent (Table 5.5). The optimum crop plans in each region were not diver- sified, but found to be specialized in the production of two crops -- rice and potato -- because of their higher relative net returns. Rice production increased from 13.215 million tons under Plan I to 19.33 million tons under Plan II, an increase of 46 percent. Out of the total estimated produc- tion of 19.33 million tons, 10.41 million tons are consumed at the farm level and 8.92 million tons are marketable sur- plus. The crop sales figures in the activities section in Table 5.5 were the net after consumption withdrawals. Potato production increased from 2.13 million tons under Plan I to 7.05 million tons in Plan II, including a marketable surplus of 6.8 million tons. Jute is found to be a most unprofitable crop and its production remained, more or less, stagnant around 1.5 million bales, both in Plan I and II as compared to actual production of 6.0 million bales in 1973-74. Sugar- cane production gained a minor thrust in Plan 11, and its production increased to 4.93 million tons as against 3.62 million tons in Plan I. The net per acre average return obtained in the opti- mal plan under Plan 11 is Rs. 2337.6 as against Rs. 1526.49 under Plan I, an increase of about 53 percent. The increase in income that occurred in the optimal Plan 11 is largely 123 no.5 mv.om o.nmmm som.mowmo —mmw mpw vow o—eo— cmnomvm omv~o~_ oo0m—mn omoomom ooo~mmm wcmm—m ooowvum oooomm oomvmms oonmmom ooommmum oonmm—m ..mm c_. ..aN e.. ..aN c.. woman; co..—_a mamas; co..__e .000. week umxwLu< wucam: mmxmocmw .ouo. .e..aau - 866.1211) \cczuoz “oz meommm wasp \ccuow-cae qeemNo_c~ anoxzusom .Nmnmo. oueuoa u a “oz wwwwmwwwm unexcucoz ~mm~¢~ occucomam :cedmou qummmn >>= ocom ocuo\:c=uo¢ uoz ammovmnqm .ocucou vemmo__ .ouop ccom cs a .8446 ..meNmm >>= cae< : u a uoz icvaa c..cm:oa.xuo_—:m .n cocommv cosc .h mNmmwoo coe< .a WMWHWMQ . cmmeemcm ~— coaaaumo mesomm >>= mz< v m N oMmWMMwmm m_o_c—o .euo. ma< owmwun WNMNNaaa... .384 e... 8208813091.... .8 _ ppveo «6.x meme_-m_ .Nmm moans; macaw—mm o :1Ww11 . cowumE=mcou .o -Nm-o_m : __—.a :o..cm q z oaommoen— coo “can: Lmu=_ mmues mm umOI‘ufi-Om LQEEzm ~QOOQGMON am .M umozgugoz “Mam”: mace—awn. — acozcea m l . com. mo.an «Nan ..om acoumau smug.) mxou-coz cone. ..Eo. N Nome cacao; __mm “Mafiam o : snow «coucnoam _—um cement: .m mmwmmmw 32.5.6: “an...” 2:: 3%. .N fig”. we... a...“ . e ecoumou mom..m. coue.: . I. .ocacou _mmm~vv .acacou cmesam o _ xcoacon asaN_p_Nco. .v .mom. nee. .. mxav-:o= cone. cwc_z amendmmm .— m c < mu:_ocum:ou wad.ocumcom no.8.fiflmwm :e_a ~oc:..:u.cm< .aco.uaz mo co_NoN.cooco .ea..ao we mu_um_souuacazu mz~ 1 1.811111 ... :o—q. ..ixmo_oczum_ Emma: .m.m m.oeh 124 due to: (i) increased adoption of HYV varieties on 6.63 million acres of land which includes 4.6 million acres transferred from local to HYV varieties; (ii) increased regional specialization reducing the number of crops from eleven to seven crops in each region (see next section); and (iii) efficient use of production resources among competitive crop enterprises, say jute vs Aus. rice. In the Optimal Plan II the net return per man-days and per unit of borrowed money increased to Rs. 30.49 and 7.68 as against Rs. 22.85 and Rs. 6.31 respectively under Plan I. The higher returns to labor and borrowed capital indicates the greater productivity of resources with techno- logical change. B. Land Use Pattern Under Technology-II (Plan II) In the optimal plan at the higher technological level the area under food grains increased by 7.2 percent, whereas the total area under commercial crops has declined by 2.6 percent compared to the existing plan (Table 5.6). The op- timal plan as shown in the Table determined the magnitude of interregional shifts in production pattern and land use im- plied for the future under specified conditions of technolo- gical improvements and resource constraints. The decrease in the area under commercial crops (cc's) is particularly due to shrinkage of area of jute and sugarcane. In fact, excluding 125 ~.o .4. mommoomm mus—mmow ~nmamoo ~o_m~w~ nesosms m—moonocconmom o.~ .-. mmmooew ooomomw ommwe. occm_m omomo—_ oooamw mmmmom ace—ow cswemm coo—m. m.mo M-w maommm ochmmF . ooccav oommmm cccmoo «mammp coo... omovm oooooo muzw Now 4 .Nmmmo_ coo—mw oommv— oocm— mwmmwm cocoa wumcwm coomo vmmmNm cooem ounce; m.o~ .-. NmMNcN cooomm - occoc— Nonco— cccmow ooomw ooom. coo.m cease mcoucoazm mmomw .o.ucmSEMm .m N.N .4. mummommw mummmsmw .eowpmo mnnoemm osvnuso mnoupco memnmse coommmo memmmom ooownmq - i ooo~_m - occmo . cacao. i accov - oooom News: mNo .4. ..mcwmm unmcoo ooeNoN— ooowv— ewnseo— Nncom_ Nooewo— occ-v ccoomw ooooo~ >>z coe< we Mi. «moommv ooovanm omcmwo oco—~v~ mmvsnwp cocosmp poomwo occnem— moooo—— coo—mm— coe< .p v m 4. mammwoo oooemue mac—~—— cocoo~— «sumac. oocvm—_ occm~o_ oocoNc_ omQNms coon—m coe< .m m.» .1. «ammopp ocooou— cccom coams occco— cccm__ ccoose ococ.o cemmme cocoov page. ocom m.. .4. mwemmmm .o—mwo— sesame mnmwmp NNoQ—a mommc— av~m~—— cacao. oponmm coco—m >>= atom «w .4. wouowm ooomom ~on-o occac- i cocoa ocmwa— ocomom i coco“. >>z ma< w. .-. m—o_e_o oooommo seam—“p coomc.~ oasemo— cocoa—N mwneep— ooavmv— cemeomp ocomo._ .auo— ms< ammocuwu Miw Easvuqo mcpumpxm s:a_uao m=.um—xu Ens—Na: a:.um.xu E:s_uao m:.um_xu 5:5.uqc m:_um.xu m:_~co coo. .< neocwmwmuwuwucwa .>z 4o ccmuuua mm: was. .m:o.mo¢ _ue_uao ace mc_um_xm .N.m 6.86» 129 region. The increased adoption of HYV is largely due to the availability of adequate ground water in these two re- gions. In fact, most of the planned increase of tube-wells in coming decades will likely take place in these regions. The EST region, however, accounts for about 32 percent of the total HYV area increase. The total area of HYV in this region increased to 3.15 million acres as compared to 2.56 million and 2.47 million acres respectively in the NW and 5" regions. About 53 percent of the total HYV area in the EST region is devoted to HYV Aman varieties. A comparative statement of land use patterns under different technological adjustments hirelation to land iden- tified suitable by FAO/UNDP soil survey for different crops is shown in Table 5.8. One noticeable feature that emerged from this Table is that the higher technological levels (Plan II) utilize proportionately less area under tradi- tional rice varieties and devote comparatively larger areas to HYV varieties in comparison with lower technological levels (Plan 1). Out of 8.5 million acres of land identi- fied suitable for Aus local in the soil survey reports, about 6.14 million acres are utilized in the optimum plan for Tech-II, as against 7.55 million acres used in Tech-I. Similarly, the optimum Plan I used about 99 percent and 71 percent of the identified suitable land of B. Aman and T. Aman local whereas the utilization of the same decreased to 73 percent and 47 percent respectively in Plan II. A total '130 xpmuucoamm um_4.ucmu. uoz n .z 5N oeom¢m mm oomvmm cocoomm «New «.2 pwmnmop .z ooocmm i cacao; N NmmNcN m.o o_mNNN oocccmm oecocaasm o . m.m mmaosc cocooem Noun: 0. ..memmm e.o~ mo~m_m_ ooommmn >>= cos< Ne emommme o.o~ coommmo ooooomm cos< .» m. . mummmoe m.mm wmppsem ooooomm coe< .m «N «ammop— mm. ovmwmmm cocoom_ .ouo_ ocom Na mmvmmmm ow ppm—mow ooooOmN >>= ocom .. moNomm m.m nee—mm cocoomu >>z m=< N. m—opvpm 5m coonmmu cocoomm pouo. m=< ..-.ooh. copa s:e_uao ca—a Eaewuao as» c. uoN___u: as» c. uoN.——u: ecu. m.nouwam ...ixmo—ocguo». can. apnoupam ..ixmo—ocgumh. 4o amoucmucom aae.uao 4o mamacoucum E:e_uao xo>c=m ..om maocu .mmcuo c_. ..-zuup one .izuah so» cope asepuao as» :. um~.._.= >m>c=m muccmm_u:=oumm Ppom ma to.._ucau. new. o_naupam 4o covucoaoca .m.m e_aop 131 area of 4.6 million acres released from local varieties were devoted to HYV in the optimum plan for Tech-II. The soil survey report indicates that nearly 7.9 inillion acres suitable for HYV Aman, of which optimum plan 01~ Plan II utilizes 5.5 million acres, about 70 percent cxampared to 1.6 million acres used in Plan I. The utiliza- ‘tion of Boro HYV land is 20 percent higher in Plan II than iri Plan I. The utilization of land under commercial crops par- tricularly under jute and sugarcane both in Plan I and Plan 11’ did not make any noticeable difference. The only signi- ihicant increase in areas under two technological stages ccnnes from potato cultivation which increased from 0.55 inillion acres in Plan I to 1.63 million acres in Plan II. C. Utilization and Marginal Value Productivity of Resources The availability of national as well as regional Wesources was estimated for higher technological level in c”lapter III. The proportion of these resources utilized in lflle optimum Plan II is shown in Table 5.9. The proportion (VF total available summer land used for cultivation ranged from 74 percent in SW region to 94 percent in the EST region. The excess supply of summer land is indicated by their zero MVP in column 6 of the table. Changing the acreage of Summer land used in the optimal plan either forcing the plan 132 Table 5.9. Resource Availability Levels and Their Utilization Under Optimum Plan - Tech-II Unit Resoorce Resource Percent Shadow Income Availability Utilized Utilities Price Penalty for (in Activity Rs.) not in solu- tion (in Rs) 1. Land (in acres) Summer land Central 5463000 4473931 82 0 937 Winter land 3329774 1917588 57 O 1400 Summer land Eastern 5537000 5187677 94 O 835 Winter land 3141000 2211066 70 0 1117 Summer land Northwest 6915000 6285456 91 O 720 Winter land 1738000 1738000 100 1117 0 Summer land 50 thwest 6670000 4960531 74 0 44 winter land “ 1499000 1499000 100 59 o 2. Permanent labor (man-days) January 552854000 132310065 24 O 10.7 February 552854000 203446661 37 0 59.7 March 552854000 98274725 18 0 24.9 April 552854000 134655696 24 O 54.5 May 552854000 316779722 57 O 48.3 June 552854000 83199405 15 O 21.9 July 552854000 152714842 28 O 11.0 August 552854000 444411016 80 O 1.1 September 552854000 199922279 36 O 2.9 October 552854000 88139063 16 O 10.5 November 552854000 206306030 37 0 2.1 December 552854000 86564886 16 0 4.3 3. Hired labor Man-days January 113547000 0 0 O 8.0 ° . 0 0 8.0 O O 8.0 . . 0 0 8.0 December 113547000 0 O 8.0 4. Chemical Maund fertilizers Central region 6135300 6135300 546 0 Eastern region 9785800 9785800 546 0 Northwest region 8435700 8435700 546 0 Southwest region 7534800 7534800 1466 O S. Irrigation water Acre/ ft Summer CN 850000 850000 100 2134 Winter 2748000 2748000 100 1251 Summer EST 1508000 912348 61 0 Winter 3557000 3557000 100 1209 Summer NW 2763000 2470986 89 0 Winter 3313000 3313000 100 777 Summer SW 2096000 1202490 57 O 44 Winter 4245000 3460776 82 0 413 6. Bullock power (pair—days) Central region 343940856 139211425 40 0 38.5 Eastern region 323221584 156592597 48 0 31.6 Northwest region 310546152 171548648 55 O 65.8 Southwest region 241075464 152023737 63 0 5.9 7. Credit (in Rupees) 9400000000 8520907119 100 0 1.04 133 to make use of more or permitting less to be used, will reduce income. The decline in income ranges from Rs. 44 in the SW region to Rs. 937 in the CN region. The winter land is scarce in supply in the NW and SW regions. The MVP's of winter land in those regions are Rs. 1117 and Rs. 59 respec- tively. The positive MVP of winter land in the NW region indicates that income to the extent of Rs. 1117 can be increased by adding one acre of these restricting resources. The total employment obtained under Plan 11 in the agricultural sector is 7.352 million man-years as against 6.50 million man-years under Plan 1. Column 5 of the Table exhibits the seasonal employment and unemployment pattern. The farm labor force is surplus indicated by their zero shadow price in all seasons. The employment rate is highest, about 80 percent of the total available farm labor force, during the month of August. The highest employment rate in August corresponds with lowest income penalty of Rs. 1.1 likely to be incurred if one additional unit of labor is em- ployed during that month. The lowest income penalty in August also indicates the highest competitive position of that resource to come into optimal solution. The employment rate is about 57 percent in February. The unemployment rate is highest during the month of June and October, about 85 percent and 86 percent respectively. The excess supply of labor force suggests that there is scope for transfer of 134 farm labor to off-farm employment without affecting agri- cultural output significantly. Chemical fertilizer is scarce in supply in the op- timal solution in all four regions, which is reflected by the positive MVP. However, the MVP of fertilizers in Plan 11 is lower than that of Plan I because of its relative increase of supply in that period. The MVP of fertilizer obtained in optimal solution is greater than marginal factor cost (MVP > MFC) of fertilizers in all regions. In the previous Chapter, while estimating the fertilizer response functions, it was assumed that with given product and fertilizer input prices farmers would maximize their profit where the value of marginal product (Mppn - Py = VmpF) of fertilizer input is equal to the price of that fertilizer input (PF). This occurs where VmpF = PF' Thus, in the optimal plan the higher MVP implies that it would be profitable for a farmer to use more fertilizer if an additional quantum of fertilizer was made available at prevailing prices until the marginal value product of fertilizer becomes equal to marginal factor cost of fertilizer (MVPF = MFCF). It appears that the MVP of fertilizer in the SW region is Rs. 1466, which is about three times higher than the MVP's of fertilizers in the other three regions. This is due to the fact that in L.P., as the supply of another complementary resource is either increased or exist in excess supply, the MVP of the restricting resource resource increases tremendously (32). It may be seen that 135 the irrigation water is scarce in supply in all three regions where the MVP of fertilizer is low whereas the same is in excess supply in the SW region as reflected by its zero scarcity value where the MVP of fertilizer is much higher. The optimum plan indicates that credit (borrowed capital) is in excess supply in Plan 11, whereas the same is in scarce supply in Plan I. This is particularly true in a situation that at higher technological levels farms usually have increased savings generated from higher production levels whereas at lower technological level farms that usually have low levels of production and which are usually exhausted in satisfying basic needs, and consequently, have a low savings rate.14 14Source: This statement is similar to the findings of the study made by R. K. Eyvindson, E. 0. Heady -- in the U.S.A. -- where they mentioned that the shadow price for capital is frequently higher on small farms than on large farms. For details, see Spatial Sector Program- ming Model by E. 0. Heady, Iowa State University Press, p. 258. CHAPTER VI SUPPLY RESPONSE AND IMPACTS OF ALTERNATIVE POLICIES 0N LAND USE, PRODUCTION PATTERN AND NET NATIONAL FARM INCOME (NNFI) The foregoing Chapter has provided the results of optimum organization of the national agricultural plan in terms of its impact on net national farm income, regional land use, and cropping patterns under two alternative tech- nological levels. The empirical results of the previous Chapter suggest that under existing prices rice is highly profitable compared to other crops. Consequently, 90 to 95 percent of the total cultivable land is devoted to rice production in these estimates. Recognizing that the coun- try's objectives in agriculture are not limited to producing more rice but also to maintaining a balance between the pro- duction of food grains and commercial crops so as to enable the country to meet both its minimum food demand as well as the foreign exchange earnings needed for economic growth. An effective government price policy could help ensure that balance. However, before resorting to policies for bringing about a readjustment in the existing cropping pattern, it is imperative to have knowledge of supply response. An under- standing of supply response provides insight for more 136 137 effective agricultural policies and corresponding farm pro- granineeds. In this Chapter the effects of changes in pi‘ices, regional resource availabilities and yield on land use: patterns, and income are examined. The optimum plan for a. given situation depends exclusively upon resource availa- trility, the existing production possibilities (the input- output coefficients), and the prices employed in the pro- gramming. A change in any of these three components of the progyramming problem affect the optimum mix of crop activities and the value of the program. The specific objectives of this Chapter are to: (a) estimate the supply response of rice under alternative sets of price assumptions for the two technological levels; (b) estimate the effects of alternative levels of fertilizer and irrigation water resource availa- bility on optimal land use pattern of Bangla- desh, and (c) estimate the effect of alternative sets of yield assumptions at Technology-II on the pro- duction pattern. These [Malicy evaluations were intended to serve two purposes: (i) to help estimate the performance of Bangladesh agriculinare under alternative sets of government policies and constraints, and 138 (ii) t9 help examine government policies and to provide sufficient information for the formulation of new policies consistent with national economic goals. A. Supply Respgnse of Rice Rice accounts for about 90 percent of the total cul- tivable land in Bangladesh. In spite of the dominance of rice in the agricultural economy, inadequate knowledge about the supply response of this crop has hampered the formulation of sound national agricultural price and marketing programs consistent with economic goals. This is especially signifi- cant in a country like Bangladesh, where efforts to promote agricultural growth by introducing new technology has affected the crop enterprise. Moreover, knowledge of supply response enters into a number of policy calculations in- cluding price support levels, buffer stock programs, input subsidies, and food grain imports. Because rice is linked with all these policy ques— tions and is crucially important in determining the produc- tion and land use pattern of all other crops, the national supply response of rice was estimated (in this section of the Chapter) assuming the two technological levels specified earlier. It is expected that these will provide the govern- ment knowledge for efficient guidance of adjustment in agri- culture in four different ways: (i) by assisting the govern- ment in formulating an effective agricultural price policy 139 and marketing program consistent with national goals; (ii) by providing policy makers with better insights to prospective changes in agriculture and corresponding farm programs needs; (iii) by enabling government to plan the best use of its resources; to realize greater incomes, and help farm input supply agencies to more accurately predict the demand for their products, and lastly, (iv) by aiding agriculture to better adopt to market demand conditions, changing resource supplies, and new technologies. 8. Analytical Procedure Variable price programming was used to estimate the supply response for rice. Since the production possibility curve employed in L.P. is represented by a series of linear segments, adjustments to price involve a series of discrete shifts in production plans. In other words, one plan will be stable over a range of prices. However, a discrete shift between plans will occur at a critical price level at which it is profitable to shift resources between activities. A change in price alters the value of the objective function and the slope of the iso-price line. A price change does not, of course, alter the feasible region. The basic results in the foregoing Chapter were based on a government fixed price of rice per maund. In order to determine the supply response of rice, however, a price range is selected between a high and a low level. 140 The high price is assumed to be 30 percent higher than the existing price of Rs. 1.61 per lb. and the low price was based on the average of four years price (1966-69) when economy was more stable and free from inflationary spiral. For Plan II, the assumed low price was further reduced to the level of the 1965-67 average price. In estimating the supply response, however, about eleven different prices for rice -- ranging from a low price of Rs. 0.46 to a very high price of Rs. 2.06 as against the fixed price of Rs. 1.61 per pound of rice, were used. The objective was to locate the optimum farm organization with different relative rice prices under two technological levels. Optimum Organizations Under Various Price Combinations of’Rice Under Tech-I (Plan I) andTech-IITPlan II) The impact of variable relative prices of rice assumed in the model were examined with respect to: (i) farmer response to production of rice, as the relative price of rice is increased or decreased; (ii) optimal production patterns of different crops, and (iii) net national farm income (NNFI). Results of Technology-I (Plan I) It was observed from the model's results that with the increase of the relative price of rice from Rs. 0.64 to Rs. 2.46 the production of rice increased from 11.42 million 141 tons to 13.24 million tons (Table 6.1). The model result indicates that when the relative price of rice was lowest, the production of other competing crops was highest. On examination of Table 6.1, it appears that when the relative price of rice is set at Rs. 0.64, the model estimates about 20.6 million tons of sugarcane, 0.464 million tons of wheat, and 5.403 million bales of jute. But as the price of rice is raised to Rs. 0.86, production of wheat decreases to 0.136 million tons and jute to 2.543 million bales. Sugar- cane and potato production remained constant. When the price of rice was further raised to Rs. 1.23, production of sugarcane decreased from 20.6 million tons to 4.39 million tons, and jute from 2.5 million to 1.669 million bales, whereas rice production increased to 13.18 million tons. The model results show that with successive rises in the price of rice, there is a corresponding increase in rice production while at the same time the relative competitive position of other crops, compared to rice, decreases. Potato is found to be an extremely profitable crop even at very high prices of rice. Jute and sugarcane appear to be competitive with rice only within the rice price range of Rs. 0.64 to Rs. 0.86. The model results show that an increase in the price of rice beyond Rs. 1.23 would hardly provide any production gain. As the price of rice was raised from Rs. 1.23 to Rs. 2.46, production increased from 13.18 142 A.m¢ gov—p.2V ascoc_ seam uuz - ~uzz mmpna :o—__*e . muan «so» co___—s . :o_uu:uoca mo.cmmom 8 on mo.~ mom.o nm_.~ mw.m ~o..c ¢~.m_ ma mpm.pmmom mo.~ mn._ amm._ mm_.~ mm.m Ne—.o o~.m_ ea m~.~mm—m mm._ m~._ aoo.. mm_.~ mm.v om_.o m_.m— ma mo.~mm—~ n~._ mm. mem.~. mm_.~ oo.o~ om—.o ~m.~_ Na mm.memmp om. em. moo.m nm—.~ om.c~ ¢m¢.c we... .a mx ex mx Nx Px pm xo: cvz muaa cacao; occugamam anon: muwz woven Eaepcps meansaz a...“ :3 «a ~uzz wmcaz «owe; au_m mova*>.uo<. :a_m sumac—mean cw ”-xmo_oc;um» Love: u:_saacmoca mu_ga mu_¢ o—oo_ga> mo mu_=mmm ._.o o_no» 143 nii'llion tons to 13.24 million tons, an increase of only 0 - 3 percent. The low production response was particularly due to: (1) limited availability of critical production inputs '1 ‘ike fertilizer, water, etc., and (ii) constraints on maximum transfer of land possible in between competitive crops, etc. Column 7 of the Table indicates the stability of the optimum ralan within the minimum and maximum price range. The supply response of rice obtained under different aassumed prices is shown in graph 1. The graph shows that .above the price of Rs. 1.23 the supply response of rice is inelastic. As the price is raised from Rs. 1.23 to Rs. 1.38, {aroduction increased by .15 percent; and when price is fflarther increased from Rs. 1.38 to Rs. 2.06 production in- Cireased by 0.3 percent. Beyond the price of Rs. 2.06 produc- ti on remained constant. This indicates that farmers response E3] asticity is higher at lower relative rice prices and is 1c>wer at higher relative rice prices. These results suggest trlat the existing support price of Rs. 1.51 per pound of rice 'fi xed by the government is (i) too high, (ii) does not pro- V‘i de any additional rice production, and (iii) is injurious tc: the production of other competitive crops like sugarcane arid jute. In fact, the model suggests that almost the same liroduction of rice could have been achieved if the price had been fixed at Rs. 1.23. Rs. If the price were lowered to 0.86, the production loss of rice would be relatively 144 23. 8____s_ omm..m4:.._m~_m.v.m.w_ .68 m A m n «m I -N S2 C3 8988 C) ‘Q .888 :7; t 11 98 88888338. é @md: AVNQV 09 and £325 marmopoccumh ucmcmmewo Lang: muwm com mmcoammm apaaam .H zamcw 145 small -- only 7.0 percent -- compared to a large increase in jute and sugarcane production of 66 percent and 467 per- cent respectively. The resulting optimum organization shows that as the relative price of rice is increased from Rs. 0.64 to Rs. Rs. 2.06, the NNFI increased from Rs. 15948.87 million to Rs. 56330.5 million (Table 6.1). This occurrence is parti- cularly due to the fact that rice contributes about three- fourths of the total GDP. Consequently, an increase or de- crease of relative rice prices has a large impact on the (level of NNFI, though the total rice production remained, more or less, constant. Results of Technology:II (Plan II) The model results indicate that the new rice tech- nology exerts a profound influence on both the optimal level of output at current prices and the elasticity of farmer price responses. The two effects are readily discernible from graph 1 where the price of rice is varied parametri- cally. The graph shows that with the introduction of high yielding rice varieties, the supply curve of rice shift to the right and the supply response at a given price is much more elastic at the higher technological level (Plan 11) 15 than at lower technological level (Plan I). The model 15For details on technological change, see Chapter 1, pages 15-20. 146 estimates that at rice price of Rs. 0.64, about 15.04 million tons of rice is produced under Plan 11 as compared to 11.42 million tons produced under Plan I. As the price is raised in Plan II from Rs. 0.46 to 0.64, rice production increased by 15.7 percent. The graph also shows that when the price of rice is raised further from Rs. 0.64 to Rs. 0.86, rice production increases by 11.7 percent (Tech-II) as compared to 7.9 percent increase under Plan I (Tech-I). Thus, the model results show that farmer response to relative price change is higher in Plan II than in Plan I. This happens because farmers' production patterns are more flexible in the long run when they are able to adjust their enterprise com- bination more accurately. In order to maximize profits the detailed model results of variable rice price programming are presented in Table 6.2. There is an inverse relationship between the rise in rice prices and the production of other competing crops as expected, except for potato. As the rice price is raised successively from Rs. 0.46 to Rs. 2.06 there is a corresponding increase in rice produciton from 13.04 million to 19.36 million tons, and a gradual decline in pro- duction of all other competing crops except potato. In this rice price range sugarcane production declined from 66.71 million tons to 4.513 million tons; jute from 12.94 million bales to 0.997 million bales and wheat from 0.943 million tons to 0.695 million tons. Wheat production, however, re- mained constant within the rice price range between Rs. 0.64 1447 mamaam co..__a av 1 umzz manna ccp___e - cyan «cog :o..__e - :o_uu:uoL¢ po.mmmew a mo.~ ham.c smo.~ n_m.e mmo.c on.o. mg ~m.moeeo oo.~ .o.. em._ “mo.“ n_m.e mao.o mn.o_ ma mm.—moum .o._ mm.. m._ “mo.~ Nem.em mmo.o m~.m— ma om.me~_v mm._ .o._ m_o.n umo.~ ~—.me mmo.o .m.~_ ca oo.m_mmm po._ mm. mv_.e umo.n m~.me mmo.c mm.o_ ma n_.mm-~ om. co. moo.~ Nmo.~ mo.~o moo.c mo.m_ Na mo._ec- co. co. co.~_ smc.~ _~.wm mem.c co.m_ _a xa: cw: uuaa cacao; occugooam gums: ouwa ou_cn anew:_e mcoaaaz -mmldl 3. ~22: macaw 00.2wqou.¢ ma_a.>.uu< :~_a .__ =a_av smooa_m=.m :— _~-xmo_o:gump cons: m:_asoguoca mu_g¢ ou_¢ a—na_co> we mu—amoz .~.o o_ao» 148 to Rs. 2.06 because of the subsistence constraint that a minimum of .595 million tons wheat has to be produced. Potato production remained stable throughout the price variation plans because of its high profitability. 1. Alternative Policies with Respect to Fertilizer and Irrigation Water Constraints The preceding section of the Chapter discussed the impact of different assumed relative prices of rice on the (i) supply response of rice and (ii) production of other com- peting crops. In this section the effect of alternative regional fertilizer and winter irrigation water availabili— ties on land use pattern and Net National Farm Income (NNFI) were examined under two technological levels. It has been observed that fertilizer and irrigation water availabilities of each region are often altered by government policies affecting land use and production patterns of different crops. The changes in government policy results mainly from (i) high price of production inputs in the international market, (ii) lack of adequate foreign exchange, (iii) national and international economic crises and above all, (iv) lack of adequate information for making policy decision at the appropriate time. 149 11. Policy Alternatives Considered The previous analysis of Plan I as discussed in the foregoing Chapter incorporated an average of last three years of actual resource availabilities in each region. In this section the impact of resource expansion on land use . patterns and income is considered. Specifically, the effect I? of additional amounts of fertilizer and irrigation water on g; land use pattern and net national income is examined. Three ; sets of irrigation water and fertilizer availabilities, re- presenting three alternative government farm policies are analyzed. The three policy alternatives (PA) are: (l) PA-I: 10 percent increase of fertilizer and 20 percent increase irrigation water availability in all four regions; (2) PA-II: 20 percent increase of fertilizer and 10 percent increase of water availability; (3) PA-III: 30 percent increase of fertilizer and 20 percent increase of irrigation water availability. In case of Plan II, the implications of both reduction and expansion of resource availabilities from that of projected resource availabilities considered in the basic model (see Chapter III) were examined. Five different resource combi- nations of fertilizers and irrigation water representing five alternative government policies were analyzed to deter- mine their effects on both (a) land use pattern and (b) net 150 national farm income (NNFI), They were: PA-IV: PA-V: PA-VI: PA-VII: PA-VIII: reduction of 50 percent fertilizer in all regions and 50 percent winter irrigation water availability in NW and SW regions. decrease of 30 percent fertilizer from the basic results and 40 percent decrease of winter irrigation water availability in NW and SW regions. decrease of 10 percent fertilizer availa- bility in all regions and 20 percent de- crease of irrigation water availability in NW and SW regions. increase of 20 percent fertilizer availa- bility in all regions and 15 percent in- crease in winter irrigation water in NW and SW regions. increase of 30 percent fertilizer availa- bility in all regions and 25 percent in- crease in winter irrigation water in NW and SW regions. In the policy alternatives the availabilities of winter irrigation water is varied only. Because, winter irrigation is primarily dependent on mechanical devices such as power pumps, tube-wells, etc., whose availability, in turn, depends on various exogenous variables like foreign assistance, imported spare parts, and fuel, etc. Moreover, 151 while considering the variation of winter irrigation water, only the NW and SW regions were considered for the analysis as the plan envisaged that due to availability of adequate ground water, most of the increase in winter irrigation in the forthcoming decade would take place in those two regions.16 Credit was not assumed as a constraint and opti- mum credit requirements for each policy alternative were estimated from the model. III. Analytical Procedure The variable resource programming is used to examine the impact of alternative regional resource availabilities on the land use pattern and net national farm income (NNFI). IV. The Land Use Pattern and Net National Farm lncgme Under Alternatingolicies (PA) In Tech-I (Plan I and Tech-II_(Plan II) (a) Plan I In this section the impact of three alternative re- source combinations of fertilizer and irrigation water on land use pattern of different crops and net national farm income (NNFI) is presented. The model results indicate that as the resource availability increases beyond PAI, the 16For details, see Soil Reconnaissance Survey of Bangladesh-FAO/UNDP. Soil Survey Reports, Ministry of Agriculture, 1970-75. 152 area under food grains decreases, though the percentage of decline was very insignificant (Table 6.3). However, the total area under commercial crops increased from 28253 thousand acresin the basic run of Plan I to 28936 thousand acres in PAI, 29205 thousand acres in PAII and 30290 thousand acres in PAIII (Table 6.3). One noticeable feature of resource expansion is that as the resource level is ex- panded, the area under local varieties either remained con- stant or decreased in each subsequent policy alternative situations, while the area under HYV increased gradually. With the expansion of resources, the area under local Aus and local T. Aman remained constant at 7557 thousand acres and 6565 thousand acres respectively in all policy runs. The local 8. Aman do not usually compete for sophisticated production inputs like fertilizer and irrigation water. Consequently, the expansion or contraction of resource availabilities do not affect its cultivation. The results show that its area increased to a maximum limit of 5490 thousand acres in PAI and then remained constant in later phases of development (Table 6.3). The total area under local Boro decreased consistently from 2898 thousand acres in the basic run to 2563 thousand acres in PAI, 2418 thou- sand acres in PAII and 2318 thousand acres in PAIII as the resource level expanded. The acreage released from local Boro variety was allocated to HYV cultivation. Consequently, 153 . Km.~m- .m.a__~ o.cmco .aa =0._._av NN mean A acoe~e.:cug a_uogu ass—ago o. A4v -.ma.~. n.m A+V mo.~amme c.o A.. Km.on~mv om.-_ne A": co.._.s =_. .222 ~.~ A+v oa~cn ..n ... mo~a~ ..~ ... anQQN mmNmN bu.>= atom o~ -V o_n~ o.o. M- m..~ o... M-w nemw momw .aog. atom m._ A+V “no. m.~ + “no. m.~ + ~m¢~ «.0. >>= coa< .h - memo - mama . mono ammo ..uo. cas< .p n.o M+V ooqm n.o M+w omen n.o A+v coem .scm ..uu- coe< .o «m. +v m.“ cm. + .mo na A4v on. wa >>= m=< - nmms - hams - smmu Nmmu .auo_ »:< N o m c n ~ _ mw_atm goo; .e «nocu A-Vomaoguou uo Aiv oncoguou go Acvamaocuou Lo A+Vomoueoc_ mo ___vua=gaup< “mucus ooov guano—ocam c, A_-co.a. .-xmo_ocgom» c. m=o_uaa:mm< Aswan: ecu Lo~__.ugouv «no.uo=,ceau oucaomoa a>_ua=cuup< smug: maogu pug:u_:o.ua< aeogouu*a Lou cguuuog on: need 333.“ 9 ea “gasouoam a).uugeascu .m.o o_aop 154 the total area under Boro HYV increased consistently from 2091 thousand acres in the basic run to 2303 thousand acres in PAI, 2420 thousand acres in PAII and 2605 thousand acres in PAIII with gradual increases of resource level. Simi- larly, the Aus HYV increased from 252 thousand acres in the basic run to 716 thousand acres in PAIII and T. Aman HYV increased from 1613 thousand acres to 1637 thousand acres in PAIII. Wheat appears to be an unprofitable crop and its competitive position with respect to other crops began to decline gradually as the resource levels increased. Though 477 thousand acres of wheat was estimated in the basic run, it did not enter into the optimum plan for PAIII (Table 6.3). Among commercial crops, potato was found to be an extremely profitable crop and its area increased from 231 thousand acres in the existing situation to a maximum acreage limit of 550 thousand acres in the basic run. The arearemainedconstant throughout the three policy runs. Jute, on the other hand, was found to be a relatively un- profitable crop, though its area increased from 1214 thou- sand acres in PAI to 3401 thousand acres in PAIII as against 555 thousand acres in the basic run of Plan I (Table 6.3). Most of the expansion in the jute area took place in PAIII only when most of the crops competing with jute already reached their maximum acreage limits. Similar to jute, sugarcane is not a very profitable crop, though its area 155 increased from 224 thousand acres in the basic run to 252 thousand acres in PAIII, which is 38 percent below the actual acreage of 400 thousand acres. The gain in Net National Farm Income (NNFI) for situation PAI to PAIII ranged from 6.0 percent to 10 percent (Table 6.3). A graphic representation of the magnitude of increase of NNFI as the resource levels are increased is shown in Graph II. The horizontal axis refers to a quantum of fertilizers between 0 and 600 thousand tons, while the vertical axis records at different points the winter irriga- tion water that was varied and the Net National Farm Income. Using the coordinates fertilizer and water, it is possible to locate the points in Graph II which correspond to re- sults of each policy alternative (PA) shown in Table 6.3. In Graph II, 0A represent the combination of fertilizer and water used in the basic run. Similarly, B, C, and D corres- pond to the PAI, PAII and PAIII with increased amounts of irrigation water and fertilizers. Since the present analy- sis is dealing with linear relations, it is possible to plot the points 0, A, B, C, D by straight lines to indicate the water and fertilizer necessary for maximum income in three alternative policies. Any water-fertilizer combina- tion lying above the line OABCD has the same optimum plan as the corresponding points on OABCD, except that surplus irrigation water will be allocated to disposal. The lines 156 AA', 88', CC' and 00' provide the basic result including three critical plans representing three policy alternatives which show income changes as the resource level increases. The fertilizer income graph is constructed by fixing irri- gation water at different levels. Graph 11 shows that as the resource level increases the magnitude of increase of NNFI diminishes (indicated by slope of the graph). The net return per maund of fertilizer and per acre feet of irriga- tion water increased up to a basic result then decreased consistently as the resource level is increased further in PAI to PAIII. The model also estimated the optimum credit requirements for each policy alternative situation. About Rs. 7117 million credit is required for PAI, Rs. 7258 million for PAII and Rs. 7344 million for PAIII (Table 6.3). Table 6.4 presents the MVP of farm resources under various levels of resource combinations assumed in three different policy alternatives. The MVP's of summer land in all regions were zero, suggesting that summer lands were not a binding constraint. As fertilizer resource levels increased gradually, their respective MVP's declined as expected. 0n the other hand, the MVP of other resources, which were not varied such as land and labor, etc., either remained constant or increased consistently as more ferti- lizer was obtained. The MVP of August labor rose from Rs. 6.37 in PAI to Rs. 34.36 in PAIII. The MVP of winter land in the SW region increased from Rs. 674 in PAI to 157 LT.’ g PA PA PA V I n m 3 60000 O ”A 500001- Basic _. 5& ResuliS\K-——6"5 ES 40000- A B 0:: us: 33; 30000- C 3 20000. O E 10000. 0 z o 0 F PA 3 PAII ,_ 800°" PAI 2 _ A §c 7000 Basic 2’. Resources :90- 392 .58 .5. .5 3 100 200 300 400 500 600 Fertilizers (000 tons) Graph 11. The Impact of Resource Expansion on Net National Farm Income (NNFI) 15£3 Table 6.4. Optimum Credit Requirement and Marginal Value Product of Resources Under Alternative Policy Alternatives in Plan I (Tech-I ) in Bangladesh (in Rupees) Policy alternatives Resource constrsints PA-I PA-II PA-III 1. Land (in acres/Rs.) CN Summer land 0 0 0 CN Winter land 0 0 0 EST Summer land 0 0 0 EST Winter land 1434 1434 1434 NW Summer land 0 0 0 NW Winter land 1756 1756 1756 SW Summer land 0 0 0 SW Winter land 674 1001 1756 2. Family labor (per man-days ) ' January labor 0 0 D . 0 0 0 August labor 6.37 8.0 34.36 . O 0 0 December labor 0 0 O 3. Hired labor January labor 0 0 0 . D O 0 August labor 0 0 26.36 December labor 0 0 0 4. Fertilizers (per maund/Rs) Central 1866 1737 0 Eastern 0 0 0 Northwest 0 0 0 Southwest 2516 1757 0 5. Irrigation water (per acre feet/Rs) CN Summer 0 O 0 CN Winter 278 368 1591 EST Summer 0 0 O EST Winter 0 0 0 NW Summer 0 0 0 NW Winter 866 866 866 SW Summer 0 0 0 SW Winter 0 230 764 6. Bullockgpower (per days/Rs) Central 0 0 0 Eastern 0 0 0 Northwest 0 0 0 Southwest 0 0 O 7. Credit 7116.84 7257.57 7345.22 (optimum requirement) (million rupees) 159 Rs. 1756 in PAIII. The winter irrigation water of the CN region was not varied in any policy situation. Conse- quently, its MVP rose gradually from Rs. 278 in PAI to Rs. 1591 in PAIII. However, it was found that the MVP of winter irrigation water in the NW and the SW regions either remained constant or increased even with the expansion of resource levels. This happened because fertilizer and water are complementary inputs and there is some specified production relationship existing between them. Unless both resources are increased proportionately, as per their spe- cified relationship, one would expect that the MVP of par- ticular resource would increase which has proportionately less of it and vice-versa. In the present policy alterna- tive situations the total fertilizer availabilities in- creased about two times than that of irrigation water in the NW and SW regions. Consequently, as the fertilizer level is varied the demand for water is also increased pro- portionally in those regions, which, in turn, increases its MVP. Results of Plan II The model results obtained for Plan II were similar to that of Plan I. As the resource levels (fertilizer and irrigation water) increased, the total area under food grains rose from 19548 thousand acres in PAIV to 25957 thousand acres in PAVII as against 25510 thousand acres in 160 the basic run. The result shows that the area under local rice varieties either remained constant or decreased gra- dually as the resource levels were relaxed (Table 6.5). The area under Aus local remained constant at 6141 thousand acres, local B-Aman at 4023 thousand acres and T. Aman local at 4355 thousand acres throughout the policy runs. However, the area of local Boro declined from 1109 thousand acres in PAIV to 929 thousand acres in PAVII, whereas Boro HYV area rose from 2608 thousand acres in PAIV as the re- source levels were increased gradually (Table 6.5). The total area under HYV Aus and Aman increased from 156 thou- sand acres and 1156 thousand acres in PAIV to 1287 thousand and 5648 thousand acres in PAVIII respectively. Wheat was not found to be a profitable crop, consequently, did not enter into optimal Plan II. As regards commercial crops, jute and sugarcane were found to be relatively unprofitable compared to rice and potato. Jute did not enter into the initial three policy alternative situations where the resource levels were re- duced from the basic model. Jute, however, entered into the optimal solution only when resource levels were substan- tially increased (in PAVII and PAVIII) and when most of the food crops already reached their maximum acreage limits (Table 6.5). Similarly, sugarcane did not find its place in the initial two policy situations and entered into opti- mal solution from PAVI onwards. 161 a. ... .. ... hep-ho. III-III... In- ... unease»: ... onco.u:. .c.oavo o~.-.—~ «..un .mmpm cm_~ e.v— o~m. .moou ween o~o even «an. -ov .m~— _v—o rllltl I I'...’ .o coo.couso. ...».c. . I. '91.. .I' 9.. ... .- ... par-pa. amN .. o .- ... .. s_.nnvo ou.o.~.. o—o—n anew on_~ vow— mne— ocom~ .oon o~o_ ovum mun. -ov ne—— ...0 .-. uncosuov ... 0no~.uc. .o oo-acuusc. ... «aqueous ... canoe»:— _.>-<. .o 09-.30950. om.m~o. ao.eooo o:..oom ...nmn~a a. ... ca....~n .n ... .~.o.o.- as... m. .-. a..n~ .~ ... ~e... imEm 1M7?» :fiB mfifi. 53. .no. . .... - .... a.n - . . ....N .. .-. coon. "N ... o...- ma~n ._ M- ..c. o. M.. coo. .o.. an . oa.. a. .. a... a... a. .- a... o. ... a... «on. - a... . a... -o. - ”we. . n~oc no. .. .-. on. .. .-. ... .... . .... - .... . tacosuvi ouoosuoi Mow incubuco ouoosuco ..-cs .9 oo-accuuo. p.«. .o nooacouuo. p.-¢. Quotas. so.__.l. acola.—::es ..tosu l:I..:c ...»c... ...o... co...... .... wean. ...g. .oco..~: ...o. no..- ...a. ... ..aa a... a...c. .o~ acousooan mumuw.awmmummm@w .- o.mm~ mmdmeu.mwe.-...a_ . ..aaa mama ..2 etc. _~o ‘09-; 9.00 m~nm up: .... can. -. cqa< n~ov a. cone .~. ..2 .a< .... ..oc. 4:. medueu-uwww .« Illa-llllaill- I lllt:--11;.§t ...-go... .099. u.uoo «ecu. .nosuq ace. .....a 9:. sou—...so.. :o..-c.aiau ouuaoao- o»..-c.-¢—¢ have: «a:.u _. .90-ocsuo— a. «so..alana .oasa.ao..u< «30.....9 so. stud... 0.: send I:I.u.e .o ace-o...“ o»..-.qaueu .m.c 0.3.. 162 The NNFI increased from Rs. 43069 million in PAIV to Rs. 71722.79 million in PAVIII. The optimal credit re- quirements enhanced from Rs. 5886 million in PAVI to Rs. 9490.87 in PAVIII, as the resource levels were increased gradually. Table 6.6 presents the MVP of resources under the five alternative resource restraints. As the resource level is reduced in the initial policy situations from PAIV to PAVI land became surplus in all four regions indicated by their zero shadow price. But when the resource levels were increased from PAVIII onwards it became a limiting resource in all regions as indicated by their positive shadow price. The effects of changing the levels of fertilizer and irriga- tion water availability on shadow prices of various re- sources can also be observed in Table 6.6. As the fertili- zer constraints were relaxed, its MVP decreased in all regions whereas those resources which were not changed such as land and water in the CN and EST regions, their respective MVP's increased due to increased demand. C. Implications of Alternative Yield Assumptions on Production Pattern and NNFI in Plan II In Chapter IV of this study the yield of different crops, particularly HYV varieties, for higher technological levels were estimated from the data received from the exper- imental stations. Numerous studies indicate that there are Table 6.6. 1(53 Optimum Credit Requirement and Marginal Value Product of Resources Under Alternative Policy Alternatives in Plan II (Tech-II) in Bangladesh Policy Alternatives PA- PA- PA- PA- PA- Constraints IV v VI VII VIII 1. Land (acre/Rs) CN Summer land 0' 0 0 0 982 CM Winter land 0 0 0 0 0 EST Summer 0 0 0 531 531 EST Winter 0 0 0 0 0 NW Summer 0 0 0 531 531 NW Winter 0 0 0 1354 3632 SW Summer 0 0 0 0 0 SW Winter 0 0 0 4081 4668 2. Family labor (per man-day/Rs) January labor 0 0 0 0 0 . 0 0 0 0 0 August labor 0 0 0 0 0 . 0 0 0 D 0 December labor 0 0 0 0 0 3. Hired 1abor(man-days/Rs) January labor 0 0 0 0 0 . 0 O 0 0 0 August labor 0 0 0 0 O . 0 0 0 0 0 December labor 0 0 O 0 0 4. Fertilizers (per maund/Rs) Central 1623 1489 1489 1407 0 Eastern 1511 1489 1489 0 0 Northwest 1489 1489 1406 0 0 Southwest 1489 1489 1439 190 o 5. Irrigation water (per acre feet/Rs) CN Summer 0 0 O 192 2473 CN Winter 0 159 159 258 1929 EST Summer 0 0 0 0 O EST Winter 0 27 27 1945 1945 NW Summer 0 0 0 0 0 NW Winter 150 150 243 1205 377 SW Summer 0 0 0 0 0 SW Winter 25 25 25 0 O 6. Bullock power (per day/Rs) Central 0 0 O 0 0 Eastern 0 0 0 0 0 Northwest 0 0 0 O 0 Southwest 0 0 0 O O 7. Credit (in million Rs.) 5885 6900.99 7925.58 9433.17 9490.87 (optimum requirement) 164 serious constraints responsible for the gap between poten- tial yield demonstrated under experimental conditions, and actual yield obtained in farmers' plots. The important 17 generally: (i) po- reasons for this yield difference are tential yields of the modern varieties are not fully ex- pressed under poor environmental conditions in which farmers operate; (ii) supply of certain inputs is less than needed to achieve the economically optimum output; (iii) farmers strive for economic optimum‘8, not maximum yield, etc. 'Considering the importance of these reasonings, in this section, alternative yield combinations for differ- ent crops were used to explore its impact on the production pattern and net national farm income. In order to determine these effects, five alternative levels of yield per acre, low I, low II, medium 1, medium II and high are assumed (see Table 6.7). 17Herdt and T. H. Wickham - "Exploring the gap between potential and actual yield in the Philippines". op. cit., p. 165-180. 18Production theory shows that, because of dimi- nishing returns, profits are always lower at maximum yield than at some lower levels of input use. In certain cases, farmers hesitate to use profit maximizing levels of inputs because the greater cost of inputs might leave them badly in debt if the crop failed. .coFuuca—axm «Eon 0g» u—o; «cacao .x—ga—.a.m ._ :64 gu>o v.0.» c. oncogue. am mucou.u:_ - 304 .m.-.a— ac.s:u vac—auao v.0.» owogu>a paauua ou u:~—a>.:co m. _ 3o4\M .maogu or» omega .9 =o.ou>.u—=u ocu so. moo.uuaga ago. um>ogna. Ace om: x—cgn: mgoEgo. ma cos< .m page. ecu ma< pouo— so. coasmma m. omuoguc. v.0.» cz\a 1(55 o.m_ m... o.m— m... m..n_ mesa pq— ow. m.m—_ o—— mo— cacao. mam AN. «we _¢v an. aeaacaaam ¢.~_ o.—— c.o~ ..m ~.o gang: mo o.oe m.om c.mn m.mm >>= atom om c.m— m.o_ ..m— o.m— pauo— ogom ¢.¢m o.om c.m~ m.o~ m.m~ >>z coa< .. ._ o.m_ . ..m— o.m— m.~. .auop gos< .. \M~.a .uuo. coe< .u .e ¢.om n.mm o.~n n.on >>z m=< \Ho.m .auo. ma< unamgucp amm amoosuc. umw omaosuc. uc— omuosuc. am «mouse:- o: «no.9 gm.: - e=.umx . 5:.uu: _. 30a — so; mco.uae:mma >-_-_ua=.oo_< .eeaae =.. ... =._a. __ moo—ocguo» Lace: m.m>—uc< Au.>.u.m=um so. mco.uas=mm< v.0.» o>.uuceoup< ...o m_a~h 166 (I) Low I -- represents existing yield levels (average of last five years used for Plan I). (11) Low II --represents five percent increases in yield from the existing yield level. (III) Medium I -- represents ten percent increase in yield from the existing one. (IV) Medium II —- represents 20 percent increase in yield, from the existing situation. (V) High -- represents 35 percent increase in yield from the existing situation. No change in yields of local Aus and B. Aman are made as farmers do not use any improved farm practices for culti- vation of these two crops. Table 6.8 presents the production patterns of dif- ferent crops with changed levels of yield. The Table shows that under low I assumption about 16229 thousand tons of rice is estimated in Plan II as against 19330 thousand tons in basic run. The resultant increase is mainly obtained from three sources: (1) relaxation of acreage constraints under each crop on the basis of FAO/UNDP Soil Survey; (2) projected increase in regional resource availabilities in 1985 and (3) substitution of HYV for local varieties as more fertilizer and irrigation water are available. Under high yield assumption (V), the rice production enhanced to 20530 thousand tons as compared to 16229 tons under 167 Table 6.8. Production of Different Crops Under Alternative Assumptions in Plan II (production - 000 tons) (jute - 000 bales) PA-I PA-II PA-III PA-IV PA-V Rice 16229 16830 17390 18690 20530 Sugarcane 3982 4690 5390 6790 8900 Potato 6332 6634 6968 7602 8508 Jute 1063 1118 1157 1280 1434 NNFI (mi11ion Rs) 54509.96 57117.12 59538.48 654420.00 73334.60 168 assumption I in Plan II. The overall tendency of optimum organization is that as the yield levels of different crops is raised, production increased gradually keeping the overall crop mix ratio the same throughout. In the optimum plan, there is an intense competition for land and fertili- zer between Irri-Aus rice and jute in the Aus season. Even in high yield assumption, Irri-Aus is likely to replace substantially the jute crop. Jute production has declined to 1434 thousand bales under assumption V as compared to 1500 thousand bales under basic results. The NNFI increased from Rs. 54509.96 million under assumption I to Rs. 73334.60 million under assumption V. CHAPTER VII SUMMARY, CONCLUSIONS AND POLICY IMPLICATIONS This Chapter is comprised of seven sections. In Section I a summary of the background of the study and the methodology employed is presented. In Section II, limita- tions of the analytical procedure are described. In Section III, a brief indication of expected land use patterns and resource use under transitional (existing) technology and with technological change are provided. Supply response of rice under two technological levels with respect to existing government pricing policy is summarized in Section IV. Expected changes in land use and patterns of different crops resulting from alternative regional resource availabilities and the changes in yields of different crops under alterna- tive assumptions are described in Sections V and VI respec- tively. The last section indicates directions for further research. I. Summaries of Background and Methodology The economy of Bangladesh is dominated by agricul- ture. The agricultural sector is the largest source of income and employment. Among crops, rice dominates 169 170 accounting for 90 percent of the total cultivable land. It contributes about 70 percent of the total G.D.P. Jute, sugarcane, and potatoes are the important cash crops. How- ever, jute is the principal export crop, earning about 80 percent of the total foreign exchange. Unfortunately for the pe0p1e of Bangladesh the growth of this sector has been very slow during the last decade. The important reasons for stagnation of the agricul- tural sector can be divided into three broad categories: (1) institutional; (2) informational and (3) market imper- fection. The institutional problems are of two kinds: (i) the misallocation of productive resources (e.g., fertili- zers, water and pesticides, etc.) among its alternative uses such as between crops and regions. Most of these productive inputs are distributed through government agencies. But the criterion governing the distribution of these inputs among crops and regions, are not based on an economic rationale. Consequently, the growing of a particular crop in a region, as well as the distribution of production inputs among alter- native crops have primarily been a function of administra- tive and political choices; (ii) the inefficient utilization of productive resources -- results mainly from imperfect incentive policies built into the system. For instance, the present emphasis of the government policy seems to be on the number of pumps placed in the field rather than on an increase in the area irrigated per pump. The present 171 system has generated a tendency to keep low the area irriga- ted.(53). Informational problem emanates from two sources which are (i) lack of knowledge on the part of the resource owner (government input distributing agencies) to prevent resources moving from low to highly profitable crops, and (ii) inadequate technical information regarding soils, cli- mate, topography and hydrology of each region that is im- portant for optimal use of resources. The market imperfec- tion results mainly from an inhibition of free play of market forces due to government laws and regulations both in product and factor markets. Recently, the FAO/UNDP Soil Survey project has com- pleted a detailed survey of about 70 percent of the total land in Bangladesh. The survey provided an improved basis for the allocation of land among its alternative agricul- tural and nonagricultural uses and among specified areas or regions where particular crops are suitable. In addition, sufficient information has been obtained for the remainder of the country through aerial photo interpretation and ex- ploratory surveys to extrapolate the results of the survey to the entire area of Bangladesh. The survey not only provides information on soil conditions, but also on present land use and on physical factors limiting agricultural development, such as depth of natural flooding, surface relief, erosion hazards, etc. The most important finding 172 was that about 11 million acres out of a total of 22.5 million acres of agricultural land are estimated suitable for HYV varieties. The entire area of Bangladesh was brought under the scope of this investigation. The analy- sis is of two technological levels of farming. These tech- nological levels are defined as: (i) transitional agri- culture (Tech-I) -- the existing system of farm operation, regarded as technological level-1. This level uses rela- tively small amounts of improved inputs and modern tech- nology; (ii) improved agriculture (Tech-II) -- this level of technology is defined by input-output coefficients which have changed significantly from their existing values and include substantial increases in the use of HYV Aman and increased amounts of chemical fertilizer. In this study the physical and economic characteristics were examined on a regional basis. These four regions are: (i) Central (CN); (ii) Eastern (EST); (iii) Northwest (NW); and (iv) Southwest (SW). Linear programming methodology was used to determine the optimal enterprise combinations and land use patterns of different crops. The entire nation was used as a unit of optimization. Each region was considered to be homoge- nous with regard to climate, farm size, and resource dis- tribution. The farmer was assumed to maximize monetary pro- fit from the allocation of land to various crops subject to 173 a set of regional and national constraints. The constraints were (i) regional land constraints; (ii) variable inputs constraints (e.g., water, fertilizer, labor); (iii) finan- cial constraints, and (iv) maximum acreage constraints for each crop. The activities in the model are: (i) regional production activities; (ii) labor hiring activities; (iii) subsistence consumption activities; (iv) selling activities; (v) buying activities; and (vi) financial activities. The analysis was carried out to represent two plan periods viz Plan I (1976) and Plan 11 (1985). In Plan I the net returns were optimized with existing national and regional resource availabilities and existing maximum crop acreage constraints. These constraints were determined on the basis of an average of the last three years. In Plan II the net returns were optimized using higher levels of agri- cultural technology, with limits on maximum acreages suit- able for different crops, as identified by the FAO/UNDP soil reconnaissance survey, and with regional resource constraints based on projections for 1985. On the basis of the results of the two analyses the impacts of various government policies and alternative technological con- straints on land use and production patterns were investi- gated focusing on three problem areas: supply response for rice, fertilizer and water availabilities, and yield levels. (i) Supply response for rice was estimated for the two 174 technological levels under alternative government price policies; (ii) The effects of alternative regional fertili- zer and irrigation water resource availabilities were analyzed, both for Plan I and II (iii) The effects on farm organization in Plan 11 of five alternative yield levels of different crops were explored. II. Limitations of the Analytical Procedure Recognition of the following limitations is required for the proper interpretation of the calculations of this study: (1) The "representative farm" was developed to represent average production situations with respect to resource ownership positions, input-output relationships, etc. The "representative farm" thus, cannot be taken to re- present all varieties of farms located in the country. In addition, the input-output coefficients for repre- sentative farms were developed by synthesizing secon- dary data available from diverse sources such as exper- imental findings, farm surveys, and macro-data col- lected by government agencies. (2) The fertilizer input requirement and yield coefficients for different crops, particularly HYV varieties in Plan II, were estimated from secondary data received from the Soil Fertility Testing Institute and the Comilla Academy. Detailed information regarding the (3) (4) 175 trials for each crop, design of the experiments, and the extent of the total trials conducted on farmers' plots were not accurately known. Due to the lack of this information, the yield and fertilizer input coefficients for different crops used in this study may differ from the actual situation, depending on the magnitude of error present in the data. The input- output relationships are considered to be single valued. In sensitivity tests only the effects of changes in crop yields, variation of rice prices and changes in fertilizer and irrigation water resources were analyzed. All other crop prices, input-output coefficients and resources were held constant throughout. The model is constructed under static economic assump- tions. In this static framework reference is not made to the management function, nor to interrelationships between the firm and household. The model assumes pro- fit maximization is the only motivation for produc- tion. The only management decisions beyond profit maximization considered in the model involve the choice among different enterprises acceptable to the manager. The potential increase in agricultural production en- visaged in the analysis is contingent upon the per— formance of the agricultural institutions created to ensure the supply of modern inputs, credit facilities and extension services to farmers. Not only do these (5) 176 institutions or inputs have their own independent im- pact on production, but also there are important com- plementarities among them. Irrigation, for example, enhances not only the profitability of improved seeds but also increases the return to fertilizer. The un- availability of these inputs, as per projected require- ments due to domestic or international politics or economic crisis would jeopardize production gains. The lack of risk and uncertainty considerations is another characteristic of the static economic assump- tion under which the model is constructed. One of the important characteristics of small scale farms in Bangladesh is that they are in transition from subsist- ence to commercial farming (shifting from local to HYV). But the cultivation of HYV is entirely dependent on the timely availability of sophisticated production inputs like fertilizer, water and pesticides, etc. Thus, a substantial element of risk is involved in the cultivation of HYV varieties as the supply of these inputs is often quite unpredictable due to political and economic reasons. This uncertainty makes the situation of the small holder quite vulnerable, as they have less flexibility to diversify their enterprise combination in order to minimize their expected income loss. 177 (6) The model recognizes only one category of land suitable for a particular crop in each region. In fact, there are important elements of regional diversity with res- pect to topography and soil characteristics. Even within each region there are small natural sub-regions with distinctive physiographic, hydrological and soil characteristics. The land use patterns in Bangladesh is preponderately determined by these different soil characteristics, and thereby provides a distinctive combination of development opportunities and physical constraints. III. The Optimum Land Use and Production Pattern and Its Impact on Net NationalTFarm Income A. Results for Plan I . (i) In Plan I, showing optimal allocation of resources, food grains account for 95 percent of the total cultivated area, as against 90 percent in the existing situation (i.e., an average of 1972-75). The total area under food grains in the Central (CN), Eastern (EST), and Southwest (SW) regions registered an increase of 37 percent, 7 percent, and 15 percent respectively over the existing situations. The resultant increase of area under food grains in the optimal plan is due to the higher relative profitability of food grains, as compared to commercial 178 crops. The net national farm income increased from Rs. 38,783 million in the existing plan to Rs. 43,128 mil- lion in the optimal plan. The increase in income obtained in the model (Plan I) is largely due to: (i) increased level of adoption of HYV varieties, (ii) increased regional specialization in crop production, and (iii) efficient use of production inputs among competitive crop enterprises (e.g., jute vs Aus riCe). (ii) The optimal land use pattern from Plan I results in considerable specialization of crops. Out of 11 crops included in the analysis only 7 or 8 crops in each of the regions appeared in the solution. Most of the shift in the cropping patterns was toward the HYV and local improved varieties (LIV) of rice. The total area of HYV and LIV increased from 3.2 million and 1.2 million acres in the existing situation to 3.85 million and 2.89 million acres respectively. The comparative advantage of certain regions in producing particular crops is also evidenced by the con- centration of wheat in the Central (CN) region and jute and sugarcane in the Northwest (NW) region. (iii) The adoption of HYV was not uniform in all regions; it ranged from 20.7 percent increase in acreage over the existing situation in the CN and EST regions to 98 percent increase in the NW region. The latter region has high potential for bringing more area under HYV because of the availability of adequate ground water for irrigation. 179 (iv) In the analysis the most profitable of the food and cash crops was the potato. In all regions, pota- toes tended to replace HYV Boro under irrigated conditions. This adjustment was, however, limited in the model by the crop rotation or acreage constraints. In the optimal plan the total acreage under potato went up to the maximum acreage limit of 550 thousand as compared to 231 thousand acres in the existing situation. (v) Under given yield and price assumptions, the Aus rice crop has a substantial edge over jute cultiVation. In optimal organization jute acreage was drastically reduced to 550 thousand, as against 1950 thousand acres in the existing situation. 8. Basic Results of Plan 11 (i) The results of Plan II, which attempt to in- dicate 1985 conditions have a similar trend as those in Plan I. In Plan II the relative competitive position of rice, compared to other crops, increases with the introduction of new HYV Aman varieties and the extension of existing HYV varieties. The net national farm income was estimated at Rs. 65,465 million, an increase of 50 percent over Plan I. Rice production increased to 19.33 million tons in Plan II as against 13.215 million tons in Plan I. 180 (ii) The optimum crop plans in all four regions were not diversified, but specialized in the production of two crops, rice and potato, because of their higher relative net returns. The increase in rice cultivation was largely due to: (a) increased adoption of HYV varieties on 6.63 million acres of land which includes 4.6 million acres transferred from local varieties; (b) increased specializa- tion reduced the number of crops from eleven to seven or eight crops in each region; and (c) more efficient use of production resources among competitive crop enterprises, like jute and Aus rice. (iii) Out of 7.9 million acres identified as suitable for new HYV Aman by FAO/UNDP soil survey, Plan II utilizes 5.5 million of these acres. (iv) The acreage under commercial crops, espe- cially jute and sugarcane, did not change very much between Plan I and Plan II. The only significant increase in areas under commercial crops between the two technological levels came from potato cultivation which increased dramatically from 0.55 million acres in Plan I to 1.63 million acres in Plan II. 181 IV. Impact of Alternative Rice Prige Policies on Supplleesponse of RTCeTTn Plan I’and Plan II A. Plan I - Rice Supply Response . (i) Supply response of rice was relatively high within the price range of Rs. .64 and Rs. 1.23. As the rice price rose from Rs. .64tm Rs. 1.23, rice production increased from 11.42 million tons to 13.18 million tons, an increase of 15.4 percent. However, supply response to increases in the relative rice price beyond Rs. 1.23 was very inelastic. Rice production increased by only 0.45 percent if the rela- tive price of rice rose from Rs. 1.23 to Rs. 2.06. (ii) These results suggest that the existing government support price of Rs. 1.51 per pound of rice may be unnecessarily high and provide little in the way of pro- duction incentives. The relatively high support price of rice is also detrimental to production of other competitive crops like jute and sugarcane which have declined substan- tially in recent years. In fact, it appears that almost the same production level of rice could have been achieved if the price had been lowered to Rs. 1.23 instead of Rs. 1.61. If the price were dropped further to Rs. 0.86 the production loss of rice would be relatively small, only 7 percent, as compared to large increases in the production of jute (66 percent) and sugarcane (467 percent). Even at the very low rice price of Rs. 0.64 per pound, production was estimated 182 to decline from only 13.18 million tons to 11.42 million tons. These results are a possible explanation of the in- flexibility of farmer rice production patterns due to its overwhelming importance as a subsistence crop and its rela- tively high net returns. (iii) Changes in relative rice price have a large effect on most of the competitive crops, particularly jute and sugarcane, but has hardly any impact on potato production. Potatoes were found to be highly profitable. They reached the maximum acreage limit, even at high rice prices. Potato production remained constant at 2.13 million tons throughout the rice price range of Rs. 0.64 to Rs. 2.06. 8. Plan II - Rice Supply Response (i) With the expansion of HYV in Plan II, the supply curve of rice shifts to the right and the supply res- ponse at a given price is much more elastic at this higher technological level. In Plan II when the rice price is set at Rs. 0.64, about 15.04 million tons of rice is produced under Plan II, as compared to 11.42 million tons produced in Plan I. As the price of rice was raised further from Rs. 0.64 to Rs. 0.86 rice production was enhanced by 11.7 percent in Plan II, as against 7.9 percent increase in Plan I. (ii) There is a similar inverse relationship between the increase in rice prices and the production of other competing crops in Plan 11 as expected, except for 183 potato. As the rice price is raised from Rs. 0.46 to Rs. 1.61 there is corresponding increase of rice production from 13.04 million tons to 19.33 million tons, whereas sugarcane production declines drastically from 66.71 million tons to 4.9 million tons and jute from 12.94 million bales to 1.5 million bales. Variation in rice price, however, hardly produces any adverse impact on potato production, its production remained stable at 7.1 million tons through- out the analysis. V. Besults and Impacts of Alternative Fertilizer and Irrigation Water Policies on Land Use Pattern Results of Plan I (i) Expansion of regional resource availabilities of fertilizer and irrigation water tends to have a favorable impact on HYV varieties, while it de- creases the area under local varieties in most of the subsequent policy alternatives (PA) studied. In Plan I, when fertilizer and irrigation water was increased by 30 percent, the total area under HYV increased from 8.9 million acres to 9.9 million acres, whereas the area under local varieties de- creased from 22.5 million acres to 21.9 million acres . (ii) (iii) (1) 184 Among commercial crops with greater resource availability the potato was found to be a highly profitable crop. Its area increased from 231 thousand acres in actual situation to a maximum acreage limit of 550 thousand acres. Jute and sugarcane, on the other hand, appear to be rela- tively less profitable because most of the expan- sion of the jute and sugarcane area took place only when all the food crops in competition with these crops have already reached their maximum acreage limit (Policy Alternative PA III). As the fertilizer resource level increased gradually, the respective marginal value products (MVP's) in all regionsdeclined as expected. The MVP's of resources, such as land and labor, increased con- sistently as more fertilizer was obtained. The MVP of August labor rose from Rs. 6.36 in PA-I to Rs. 34.36 in PA-III. Similarly, the MVP of winter land in the SW region increased from Rs. 674 in PA-I to Rs. 1756 in PA-III. Results of Plan II Implications of resource expansion on land use pat- terns were similar in both Plan I and Plan II. There was a tendency to substitute HYV rice for local varieties as more resources were vauired. (ii) (iii) VI. (1) (11) 185 The competitive position of jute and sugarcane rela- tive to rice and potatoes decreased substantially as the resource levels were reduced from the basic results. Both jute and rice entered into optimal solution only when resource levels were expanded considerably. The slope of net national farm income declined as more resources were obtained indicating, as expected, diminishing marginal returns of fertilizer and water. Results and Implications of Alternative Yield Assumption on Production Pattern in Plan II Even under the low yield assumption (average yield of 1972-1975), rice appears as the dominant crop. However, under low yield assumption production de- clined to 16,229 thousand tons, as compared to 19,300 thousand tons under basic results. The de- crease in production is mainly due to the low yield assumption. The overall crop acreage remained, more or less, constant in both situations. The overall tendency of optimum organization shows that as the yield levels of different crops changed upward, production increased simultaneously, keeping the overall crop mix ratio the same throughout. _ .l I m . a- (iii) 186 There is an intense competition for land and ferti- lizer between HYV Aus rice and jute in the Aus season. Even under the high yield assumption Irri- Aus is likely to replace substantially the jute crop. Jute production declined to 1434 thousand bales under assumption V, as compared to 1500 thousand bales under the basic results. POLICY IMPLICATIONS AND CONCLUSIONS The important policy implications of this study are as follows: (i) There is a tremendous scope for extending the areas under all HYV rice varieties, particularly HYV Aman. It appears that about 50 percent of the area under local T. Aman could be replaced by HYV Aman. The research finding that the farmers' profitability in- creases significantly, as a result of the introduc- tion of HYV Aman, reinforces the possibility of success of a government policy to encourage expan- sion of areas under this crop. As the existing poli- cies with respect to supply in input packages (e.g., fertilizer, water, etc.) are mainly geared to the Boro season, a similar sort of provision for HYV Aman will necessitate the reorientation of these policies. (ii) (iii) (W) 187 The high profitability in rice production indicates that there is substantial scope for reduction of input subsidies involved in irrigation water, fer- tilizers, and pesticides without any significant adverse effects on farmers' incentives. The existing support price of rice appears high relative to the prices of other competitive crops. Currently, rice support price appears to be fixed at a level where supply response of rice is inelas- tic. On the other hand, these high levels of rice price have severe detrimental impacts on production incentives for jute, sugarcane, and wheat. The finding of this research indicates that the existing support price for rice could probably be reduced by one third without appreciable adverse effects on rice production. The comparative gain to be re- ceived through large increases in production of jute and sugarcane would be much higher than the relative loss to be incurred from the decline in rice production. The breakthrough in seed technology for HYV Aus rice has increased considerably the competitive strength of Aus rice relative to jute. With high yield po- tential of HYV Aus rice, the jute acreage will de- cline greatly, unless the relative price of jute is adjusted to maintain its competitive position. Further 188 From the long run point of view the most viable proposition, of course, lies not on enhancing jute price, but on increasing the productivity of jute. In order to realize this objective, it is necessary to initiate a vigorous research program which would help to increase the productivity through varietal improvement in seed and related production practices. Research (i) (11) The basic prerequisite for building of any mathema- tical model for an agriculture sector in a particu- lar country is the availability of sufficient amounts of basic micro-data for various regions or sub-regions. The state of existing data sources in Bangladesh is too inadequate and consequently, serves as a tremendous constraining factor in model building and effective policy making. It is neces- sary, on the part of the government to accord top- most priority in the collection and compilation of basic farm management production data for different categories of farms and regions. Result of the study suggests, in further modeling, that other activities and restrictions should be added to the L.P. model used. These activities would include production activities of different (iii) 189 categories of farms and their adjustment to tech- nological change over a period of time. Different restrictions should be placed on off-farm employment opportunities to see which farms would go out of farming; such a restriction should permit full time off-farm employment. Research emphasis should be placed equally on econo- mic, physical, and institutional determinants of land use along with supply and demand considerations for farm products and farm inputs. This problem is multidisciplinary in nature and its solution requires concentrated cooperative efforts of dif- ferent branches of human knowledge connected with agricultural production such as economics, agrono- mics, soil science, hydrolOQY. etc. APPENDICES . .oUNLO Noxgae coca NoemNg new uoaamma wuan No. Nauuxu Nqocu ..N No. umoo uaomxm .n.-~sa— c. ca~.—.as~. .c on: .Oauuo as» .opuwmncc— .o o_uoc gm: .o NNNon as» :o vouaa.umo was Nmou co~.—.ucu.\M .pouol as» :. eouacoacouc. a—NuoLNv ago New». umou omega NO o—aau mpg» c. «enouoa ouc— cmxau NO: one ..uuo .umou Nona. u=NNN= nee umou :o.»au.sg. No guan New». NNoo o.no.ce> :NONcou Nqosu NaoNNN> .0 “Non u=NNNo>gas use ac.:ocu as. u=.uoa.umm c_\fi 19(1 N__ ON ON N ON OON NO N.ON OOOOLOOON ONO ON ON ON NNN OON O.— OO. OOOOOO O.— ON NO N NN O N. N. OOOO NN. ON ON O. OO OO O N Name: NON ON ON ON OO O. ON ON ..z. OLOO ONN ON . ON O. OO. O. ON N.N_ A_OOO_. OLOO NON ON - OO O. OO O. NO ON ..N:. OOON NN. ON ON O. OO O. N. N... A case OO_ ON ON - NO. ON - O Oaa< NNN ON OO ON OO O. ONO ON .>>=. NON OON ON ON - OO. OO 0.0 O ...OO.. NON N O N O N N _ Naocu ONOO ANN O.. ANN. .NNO ANOONO ON. ALOON. Ana. ALOON. NOOO o.ao.go> uNou umou NNou mac umou sac umou xuc o.na.go> pouch .uN.: uNocuNON NouNuNNNa. \mmuoum \wNLONNNNNNu. .o Nama— A_ c... 1 _ Nae—ocguo»v —ouoz m=.esugaoga ogu :. can: Naosu u:~c¢...: co. \Humou m=NNNu>Noz ecu me.zocm ocu< com augusNNNm < x~ozumm< .Nma. :. Nu.Na:N co~...acu. c: an v.36: ocogu Ncaoe N.g. ....os.uuomNue am. can Nucua .o ucaos can N.. Na can NO N. NON. Neoaecosou as. .o co..:a -.NNN.u can Neuaogauoca .oauuu .o N.Naa as. ea toast.NNu ago: NuNou ca~...»co.\fl 191 O.N.. ON ON N N. OON NO NN OOOONOOON 0.00N ON ON ON NNN ONN O.N ON. OOONON O.NN. ON NO N NN O NN N.NN ONOO O.NN. ON ON O. NN OO .N ON NOON: O.NNN ON ON ON ON O. ONN N.NN. ...z. OLON O.NON ON ON O. OO. OO NO OO ...OO.. OLON O.N.N ON ON OO ON O. NO. NO ...2. OOeO O.NN. ON ON O. ON O. N. N. . OOOO O.OO. ON ON - OO. OO - N OOOO O.N.N ON ON ON OO. OO ONN N.NN. ...z. NON O.NON ON ON - OO. OO N. N.N ..OOO.. NON N N O N- N O1 Naogu NNou AN: =.. .Ngv Ana. .coom. .Nz. .Noom. A... ..oom. uNou o.au..n> “nag «nag «Nou sac NNou mac umou xuo u.na..u> .Ouoh .uN.: «Nocoue. Nov.9.amoa Nuoom \Hmsou...ueu. .e mama. M.. O... - . .OO.OOOOO.. .ovcz 9:.35a. as. as. :— eoNz Naocu u:ocu...n so. NNou oc.uNo>ga: ecu 9:.3ocu «50¢ so. nouae.umu a x—ozm¢m< REFERENCES AHN, C. Y. and I. J. Singh (1972), "Distribution of Farm Incomes Under Alternative Policy Regimes: A Dynamic Analysis of Recent Developments in Southern Brazil (1960-70)," a paper presented at the seminar on Applied Welfare Economics at the Annual Meetings of the American Agricul- tural Economics Association, Gainesville, Florida, August 21, 1972. ASADUZZMAN, M. and Mohabat Hossain. "Some Aspects of Agricultural Credit in Two Irrigated Areas in Bangladesh," New series No. 18, BIDE, October, 1974. BOSE, S. R., "Trends in Real Income of Rural Poor in East Pakistan, 1949-65,“ Pakistan Development Review. BUSE, R. C. and D. W. Bromely (1975), Applied Econo- mics Resource Allocation in Rural America, Iowa State Press, p. 100-224. CHARLES, Kennedy, H. P. Thirlwall, March, 1972. Survey in Applied Economics: Technical Pro- gress, The Economics Journal. CHARNES, A. and W. W. Cooper (1959), "Chance Con- strained Programming,“ Management Science, Vol. 6, pp. 73-79. DAY, R. H. (1963), Recursive Programming and Pro- duction Response. Amsterdam: North Holland Publishing Company. DAY, R. H. and T. Heidhues (1966), "Towards a Micro- economic Model of Agricultural Production and Development," Farm and Market Workshop Paper No. 6702, Social Systems Research Institute, Universityof Wisconsin, 1966. 192 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 193 DAY, R. H. and I. J. Singh (1972), "A Dynamic Microeconometric Model of Agricultural Development," Paper No. 7135, Social Systems Research Institute, University of Wisconsin, March 1972. Department of Agriculture, 1969-70, Accelerated Rice Production Plan in East Pakistan, Government of East Pakistan. Department of Soil Survey, 1970, Soil Survey Reports of Bangladesh. DULOY, John and Roger 0. Norton (1971), "A Pro- gramming Model of the Agricultural Sector in Mexico: A Development Research Center, IBRD," A paper presented at a Conference on Agricultural Sector Analysis and Programming, Ames, Iowa: Iowa State University, May 1971. East Pakistan, Agricultural Census 0r anization. AgggculturaTCensus 1960, Government Press East Pakistan, Bureau of Statistics. Master Survey of Agriculture, 1967-68, Eden Buildings, Dacca. East Pakistan, The Registrar of Cooperative Socie- ties, Agricultural Credit in East Pakistan: A Survey Report, January, 1966. FALCON, W. P. and Carl H. Gotsch, Relative Price Response, Economic Efficiency and Technolo- ical Chan e: A Study of—Punjab AgriT.in evel. Po cy-II, The Pakistan Experience, Harvard University Press, Cambridge, Mass. 1971. PP. 160-185. GOTSCH, Carl (1971), "A Programming Approach to some Agricultural Policy Problems in West Pakistan," in Studies in Program Planning, H. P. Chenery, ed., Cambridge: Harvard University, 1971. GUNNAR, Mydral. Asian Drama, Part 5, Pantheon, New York, T968. HABIBULLAH, M., Pattern of Agricultural Unemploy- ment, Bureau of Economic Research, Dacca University, 1962. 20. 21. 22. 23. 24. 25. 26. 27. 28. 194 HAYAMI, Y. and V. W. Ruttan (1970), "Factor Prices and Technical Change in Agricultural Develop- ment: The United States and Japan 1880-1960," J. P. E. , Vol.78, No. 5, Sept.-Oct. 1970, pp. 1115- 1141. HEADY, E. O. and Chandler, Linear Programming Methods. PP. 225- 256. Iowa State University Press, 1969. HEIDHUES, Theodore(l966), "A Recursive Programming Model of Farm Growth in Northern Germany," Journal of Farm Economics, 48: pp. 668-684. HEIDHUES, Theodore (1969), "Recursive Programming in Agricultural Applications," in Carter and Brody (eds. ), Input- Output Techniqpes (in honour of Wassily Leontief), Vol.1: Contri- butions to Input-Output Analysis, North HollanHTPublishing COmpany. HENDERSON, J. M. (1969), "The Utilization of Agricul- tural Land: A Theoretical and Empirical Inquiry," The Review of Economics and Statistics, Vol. XLI, No.3 (August, 1959), pp. 2421259. HERDT, R. W. and T. W. Wickham, "Gap Between Poten- tial and Actual Yield," Food Res. Inst. Stud., Vol. XIV, No. 2, 1975. IFTIKAR, Ahmed, The Green Revolution in Bangladesh, A paper presented in a seminar on HYV, April 9-11, 1975, Comilla Academy,8angladesh. JOHNSON, Glenn L. , et a1. (1971), A Generalized Simulation Approach to Ag_icultural Sector Analys1s: withiReference to’Nigeria, Michigan State University, East Lansing, November, 1971. JOHNSTON, B. F. and J. W. Mellor (1961), "The Role of Agriculture in Economic Development," American Economic Review, September 1961. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 195 JOHNSTON, B. F., J. Cownie and B. Duff, (1970), “The Quantitative Impact of Seed Fertilizer Revolution in West Pakistan," Food Research Institute Studies in Agricultural Economics, Vol. 1, No. l, 1970. KAHLON, A. S. £3 31,, (1960), The Dynamics of Punjab Agriculture, Department of Economics and SociologY. Punjab Agricultural University, Ludhiana, India. KOOPMANS, T. C. (ed.) (1951), Activity Analysis of Production and Allocation, New York: John Wiley and Sons, Inc. ”DO-l ’ A ' I LEE, Y. C. Adjustment in the Utilization of Agricul- tural Land in Southeast Michigan, Unpublished Ph.D. Thesis, Dept. of Agricultural Economics, Michigan State University, 1975, p. 79. MASSELL, B. F. and R. W. M. Johnson, Economics of Small holder Farming in Rhodesia, A Cross Sectional Analysis of Two Areas, Stanford University, Food Res. Inst. Stud. Supplt. to Vol. 8, 1968. MELLOR, John W. (1965), "The Subsistence Farmer in Traditional Economies," Paper presented at the A.D.C. Seminar on Subsistence and Peasant Economies, East-West Center, Honolulu, Hawaii, February-March 1965. MILLER, T. A. (1972), "Evaluation of Alternative Flexibility Restraint Procedures for Recur~ sive Programming Models Used for Prediction," A.J.A.E., Vol. 54, N0. 1, Feb. 1972. Ministry of Agriculture, Bangladesh Agriculture in Statistics, Statistical Series No. 1, Nov., 1973, Gov't. of Bangladesh. Ministry of Agriculture and Works, Survey Report on Farm Power, Machinery and Equipment in Pakistan, Dec., 1967, Govt. of Pakistan, Rawalpindi. MUDAHAR, M. S. (1971). "A Dynamic Microeconomic Analysis of the Agricultural Sector: The Punjab," A paper presented at the Fifth In- ternational Conference on Input-Output Tech- niques, Geneva, Switzerland, January, 1971. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 196 MULLER, G. P. (1970), "Forecasting and Rolling Plans for Competitive Supply with Production Lags," BME 7052, Social Systems Research Institute, n versity of Wisconsin. NAKAJIMA, Chihiro (1957a), "Overemployment and Theory of the Family Farm," Osaka Daigaku Keisaigaku, March 1957. NAKAJIMA, Chihiro (1957b), "Equilibrium Theory of Family Farms," Osaka Daigaku Keisaigaku, March 1957. NAKAJIMA, Chihiro (1965), "The Subsistence Farmer in Commercial Economies," Paper presented at the A.D.C. Seminar on Subsistence and Peasant Economies, East-West Center, Honolulu, Hawaii, February-Marh 1965. ' _'.° "Ilia NAKAJIMA, Chihiro (1970), "Subsistence and Commer- cial Family Farms: Some Theoretical Models of Subjective Equilibrium," in Wharton, C.R., Jr. (ed.) Subsistence A riculture and Econo- mic Development, Aldine ubli§hing Co., Chicago, 1970. Planning Commission, Bangladesh, 1972-73. Annual Plan, Bangladesh Government. Planning Commission, Economic Development in 1973-74 and Annual Plan for 1974-75, Plan. Com. Gov't. of Bangladesh, July, 1974. Planning Commission. First Five Year Plan of Bangla- desh, Gov't. of Bangladesh, Nov., 1973. RAIS UDDIN Ahmed, Economic Analysis of Tube-well Irrigation in Bangladesh, Unpublished Ph.D. Thesis, 1972, Dept. of Ag. Econ. Michigan State University. ROBINSON, W. C., Disguised Unemployment Once Again: East Pakistan 1951-61, J. Farm Economics. SCHALLER, W. Neill (1968), "A National Model of Agricultural Production Response," Agricul- tural Economic Research, U.S.D.A., E.R.S., Vol.T20 (2). 50. 51. 52. 53. 54. 55. 56. 57. 197 SCHULTZ, T. W. (1964), Transforming Traditional Agriculture, New Havefik Yale Ohiversity Press. SINGH, I. J. (1971), "A Recursive Programming Model of Traditional Agriculture in Transition: A Case Study of the Punjab, India," Ph.D. Dissertation, University of Wisconsin, 1971. TARAFDAR, Rabiul Islam, "Alternative Low Lift Pump Irrigation Policies in Bangladesh: An Econo- mic and Financial Analysis," Unpublished Ph.D. Thesis, 1973, Dept. of Agril. Econo- mics, Michigan State University. TARAFDAR, Rabiul Islam, "Socio-Economic and Insti- tutional Factors and the Efficiency of Low Lift Pump Irrigation in Bangladesh. (Seminar paper), Michigan State University, Sept. 1976. YOTOPOULOS, P. A., "On The Efficiency of Resource Utilization in Agriculture," Food Res. Inst. Stud. 8: 125-135, 1968. HOPPER, W., "Allocation Efficiency in a Traditional Indian Agriculture," J. Farm Econ. 47: 611-624, Aug., 1965. WHARTON, Clifton R., J:. (1963), "The Economic Meaning of Subs stence," Ma a an Economic Review, Vol. VIII (2), OctOBer 1963. WHARTON, Clifton, R. Jr., (ed.) (1970), Subsistence Agriculture and Ecgnomic Development, Chicago: Aldine Publishing Company.