I. 2:, ‘; {$\s. . 1."?0.‘ ..-_... ‘2‘. .,r:'. ','.‘.5‘\‘r\ ~ '1, ( \ "m Eggs 'tL'; ' xgn't}. *2in I 3 s 3“?” t' 3.3", ‘ '63. .1313. 33. '4' ‘ a 3 r?! 'I' 233.1%.3‘1 ' I I ‘n “I . {-39% ”333%31-33 3:3... I , X 11 Ia. 1;" "'2: 2'94 L; ’ :l wk.“ ‘- t’b'x #rf‘lf‘m I . A . ujs’; “3‘3; ' “5.65. 6::- :i- 2‘ '3 .1: I '4 fiméttlh.6§ . "- If". )(r‘ I; "E %’$'fi 'l'ov-v 3.173? ‘ :‘l Wkltk.r.:{.:q.n:nq'32.)..31.J"“,1n|'fi~:‘ 0A1:7A':1 .13. 4 I 3 {.Eov {f 35".. 3 '7'"; ' '9'“: §".‘.'"'-v*r,- “a” “27"}- II "\‘v 49"! 1‘ ”‘L- fidl'ri'fl" J ”1' H \'| 3": V‘J '\ . I 3 I ¢ 0 03M.|‘Ilr3»~:\.l .‘j“3 ';:'.I“;; .3" 0“. "Vi " ' . no - 3|I nu"... . 3 J | : ”W ';H t‘Ill " 'I‘A'lkil'w' [VJ I50‘1“"“IIIrQqQIg4: " “3:" "'3. MN 71515 “IN" IIq" I'f‘,‘ 3'33.‘ '3“ 323.43%... """,'.""I. 33:33 . ."“3l‘{33‘.‘;:;l . 11'..é(,w3vr~3_:.; """:‘ (z‘fl'uf'v;.'~:u '0" 3;; -_. m?- ., a .; LI ' "I ‘7 73 3" ‘lrL'lll‘Hl‘ 'l'- "" ""' 'L" " .‘ 1- .. .' \ 3‘ .. 33:35?- . :f'i. I . "II"' 7133-... .3. .J' 3 .3": 9(3),: '33 .3 ‘. 3 'A» . : "' 'I "'h' 3.).3.'.'. :. 31 .II ".\13.-‘333‘3-""'33, . ".3'3- ‘I 33' .. 33:33 :.---jII:fI._. S ‘3' 3 3 £33.33, I'LL: [M A'...33 ‘ ‘W I 'I3 3. 3 I‘.3 :03 .'.,I ' IIJ“.“ ‘+'l‘.~fl. 3I 3 3 "3:3 l. I03 . .4, 3. \ .. 4’ w b 3. , .:._ .3 - '.“ I- 1"'.'I1-'I'vi‘ " :3..." I‘; .1". 32‘ " . . Murmur 21:},‘3 " II' |{: .1” "‘ "' 'Efi‘". ““3 ' I 3"” 5| ' " "I I".l'~;n' "1'1"" "I("' V" WE'H J d. H“ 5"" k 9‘3: I' "~ 'I ‘ "W‘ ’ f. .'|."3l3.v'3l’l In.” 3.3.93.1." '1‘: "t'J‘I ‘35}... “r 3.. 3. .3. 3. J.- ‘zu. I .h 3'.33..3 3. ,3: I". ')",|' ' "SCRAEE l?"’( “'1'? “1‘1"“ 10“." "'."' "VJ. '- ; v .I . . I .. “It. I ’3 .13\' I3I':'I\3' 3. “high .'""'[:.::::"'L"3_ 3'3...3'.33 33: I , 3‘ . - ~ I :I ..33_ v.3 (3.3.. - I .3’13" I 3‘ .13: 30") ' ‘I, \ ‘c‘i'v r," .3.’ 0.61333 M} 'f-\' . .7333 II)? .‘fjg'3121'h I‘% “if." 'lu.':\": l‘. L. 1'4: 'I'.‘, 3'. "H' ' 3"." M; II} “Ir"'4'1-::.i"'">"" ' ‘3‘3‘3 M'w I3 5"‘3'3'3‘! """" 3.: I '-'";I"'r'.|\!3"¢"'} -I ' W '1' 3‘ ' ’ -3-v._~, ' qi I’-fi ' 3: . I1- I ,- up“... II' . c ' . ' " '0‘ 'l'" N'p'” '1" '3."';i: ' '1':I3 31'. 1"" "I'd I3:"'I' M. €33I3'I'N' QHI'I 3;"ng ' "' I I . .'1.3’.""- ’I .. ("U H3 'III3 I3'I—I1IP'\"'I" . .f"_t’\ II ”151 .I3's" . «31.12" .4' """ .' -" H). I ""‘3"1I’( '3." IV'v 1:3 :.I'. I'I :4 I |“"I '$""I:'c"";.','I"‘I|I".a I ""23; ',E‘433(:33I‘3333.:£_I3:'3 ' A ‘ I “I3 '1‘ I 'l3 3r 3 3 I .JI‘HI L'W ”3'3 I" 3 3'". 1'. l""' ":'I”I"\'"‘ 9'3 ' I 3 .3 3 3 3 3333‘ .3 I 3 33 I ' . 3E3"... I1 1 A. 3“” I” “Hot 'h‘ I "I.'.1-' i IIM-n" °J- ‘. 'NR 3'13-3 "-‘3 " «';’I*.3'.3.'.3.3 """"'r"'l 3 '.33 I I " 3'33 ' o '0 I. .I""3I'3.'I .3 'lu'w'j‘ ”43-“? ' If" "' '3 "'"III 3' N . My "'3'3.':3I' 3‘" 33 '3 " II I'?‘ '3‘ I" '3'3"."'--3'I.'i l' “'1'" '""' 3'! '3‘. luv'I'g 'I3h .'II"V.$'JIII“ :- I I """"II"' "'3 3 3 4 * 3'fi\‘.'9"33" "‘3' ' " " ""1IIIIi'I .I..l:""' ' ' ' ".H II l""|l|l"' 'lm "" """ " 3"” |' W "‘I' ”1'13"" THESIS Michigan Stars 5132 Univerii {y This is to certify that the thesis entitled AN AGROECONOMIC LAND RESOURCE ASSESSMENT FOR RICE PRODUCTION IN THE DOMINICAN REPUBLIC presented by Gary Stephen Kemph has been accepted towards fulfillment of the requirements for Ph. D. degree in Achicul tural Economics Date November 10, 1980 0—7639 IIIIIIIII III III III/IIIIIIIIIIIIII mm 9 {0044 9w: 25¢ per day per item RETURNIIG LIBRARY MTERIALS: Place in book retu urn tone-i0 charge from circulation records AN AGROECONOMIC LAND RESOURCE ASSESSMENT FOR RICE PRODUCTION IN THE DOMINICAN REPUBLIC By Gary Stephen Kemph A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Economics 1980 é/fla C/éo" ABSTRACT AN AGROECONOMIC LAND RESOURCE ASSESSMENT FOR RICE PRODUCTION IN THE DOMINICAN REPUBLIC By Gary Stephen Kemph The Dominican Republic has been importing increasing quantities of rice since 1972. In 1978 the government began a program to increase domestic rice production in order to reduce foreign exchange expenditures . and to increase employment. As a sub-component of that program, this study was undertaken to assess the agronomic and physical ("agrophysical') and economic ("agroeconomic") feasiability of rice production expansion in each land unit ("6055") in the Central Region of the country. The regional study was to serve as a prototype for a national rice land use assessment and for studies of other agricultural land uses. The analysis was of two types. First, an agrOphysical analysis of the land base was carried out in order to identify areas with potential for increased rice production. Rice plant tolerance limits for poten- tially limiting soil and water characteristics were estimated and cross- tabulated with the corresponding GDSS characteristics. Second, an economic analysis was made of the 60355 selected in the agrophysical analysis. Benefits and costs of the production of rice and of its two principal competitors for land in the Central Region (sugarcane and Gary Stephen Kemph cultivated pasture used for milk production) were analyzed from the producer cash expense ("monetary") and the national opportunity ("unsub- sidized") points of view. A typical current set of rice production techniques and an alternative set requiring increased labor and/or de- creased foreign exchange used were analyzed. The benefit-cost results were used in partial budgeting of a hypothetical 25,000 ta expansion of rice production area using 11 alternative strategies. The strategies consisted of various assumptions on expansion decision rules (maximiza- tion of rice production, maximization of rice monetary and unsubsidized returns to land and management, maximization of rice production labor use, and minimization of rice production foreign exchange use) and policy variables (rice production techniques, number of rice production cycles per year, and expansion in current or potentially available 60555). Conclusions drawn from the study can be divided into policy and methodological conclusions. There are two major policy conclusions. First, irrigated 60535 06A and 07A and rainfed 208 have the best pros- pects for rice area expansion. A 25,000 ta expansion of rice production in those GDSSs under strategy E (maximization of rice monetary returns with free choice of production technique and 0033) would increase annual labor use by 2.4 million hr and increase brown rice production by 4.4 million qq. The country could save $5.5 million in foreign exchange and $2.6 million total on the 4.4 million qq or rice through domestic pro- duction rather than importation. Second, adoption in current rice pro- duction areas of the alternative techniques analyzed in this study would both increase labor use and decrease foreign exchange use by about 50 percent. Gary Stephen Kemph There are three major methodological conclusions. First, the study methods seem to be appropriate to the current Dominican planning environ- ment. However, the secondary and judgmental data sources used in the study sould be supplemented by land use surveys at the farm level. Second, several of the critical assumptions made in the study should be given more detailed study prior to use in project planning. Third, other intra- and inter-sectoral production and consumption information in addition to this land use assessment should be incorporated in the rice expansion policymaking process in order to increase the probability that a feasiable and desirable expansion policy and resulting projects can be planned and implemented. DECIDATED to Julie, Peter, Erin, and SGT Johnny Snelson, RIP ii ACKNOWLEDGMENTS My sincere thanks go to the members of the SIEDRA (Sistema de Inventario y Evaluacion de los Recursos AgrOpecuarios) staff in the Dominican Republic for their forbearance in the conceptualization, plan- ning, and implementation of this study. I especially acknowledge the considerate support given me by my major professor, Dr. Glenn L. Johnson, in the final weeks of the dissertation writing. Thanks go to my other research committee members, Drs. H. Riley and M. Abkin, and Dr. D. Chappelle of the Resource DevelOpment Department for their help in evolv- ing the definition of a researchable t0pic. To my other guidance committee members. Drs. w. Vincent, and P. Strassman of the Economics Department go my thanks for their support and interest in the development of my graduate program. Several members of the Comprehensive Inventory and Evaluation System (CRIES) Project staff deserve a note of appreciation for their role as ”devil's advocate" during the early drafting of the study proposal. Carl Montano's comments were particularly incisive. The study was carried out as part of the overall CRIES/SIEDRA Pro- ject in the Dominican Republic, thus thanks for financial and admini- strative support for the study go both to the CRIES Project (a c00perative endeavor among USDA-ESCS, USAID, and MSU) and to the Subsecretaria de Recursos Naturales of the Secretaria de Estado de Agricultura of the Dominican Republic. Finally, I wish to express my loving gratitude to my wife, Julie, and children, Peter and Erin, for their sacrifices and support during my graduate studies. Julie typed and edited all of the early drafts. Lois Pierson typed the final draft of the dissertation. iv TABLE OF CONTENTS LIST OF TABLES ........................ LIST OF FIGURES ....................... Chapter I. INTRODUCTION ..................... Problem Statement ................. Study Background ................ . . . Study Objectives .................. Reader's Guide to the Dissertation ......... II. REGIONAL SETTING OF THE STUDY ........... Land Resource Base . .' ............... Land Classification ............... Physiography .................. Climate ..................... Soils ...................... Water ...................... Current Land Use ................. The Rice Industry ................. Rice Consumption ................ Rice Production ................. Domestic Production and Area ......... Input Use .................. Rice Imports .................. Rice Marketing ................. Rice Agronomic Research ....... ' ...... The Sugar Industry ................. The Cultivated Pasture/Milk Industry ........ III. REVIEW OF RELEVANT ECONOMIC THEORY AND LITERATURE . . . . Theoretical Framework .............. Agronomic Theory ............... Economic Theory ................ Efficiency ................ . . Multiple Production Goals and Values ..... Resource Valuation .............. Stages of Production ............. Input Substitution .............. Rice Adaptability Research ............. International Studies . . . .......... Domestic Studies ............... Page ix xii Chapter Page Development of CRIES/SIEDRA Data Base ........ 47 Chapter III's Relation to the Dissertation ..... 50 IV. RESEARCH METHODS . . . . . . .............. 51 Concept ....................... 51 Data Acquisition and Refinement . . ......... 53 Agrophysical Data ........... . . . 54 Rice Agrophysical Tolerance Limits . . . . . . 54 Land Resource Base ...... . . . . . . . . 55 Hater . . . ............ . . . . 55 Soils . . . ............ . . 59 Photoperiod and Temperature . . . . . . . . 59 Use of Agrophysical Data ........... 59 Economic Data .................. 61 Production Techniques and Yields ....... 61 GDSSs with Current Rice Production ..... 61 GDSSs with Potential But No Current Production ............. 68 Matrix of Potential Rice Yields . . . . . 68 Production Techniques .......... 70 Alternative Rice Production Techniques . . . 70 Monetary Production Cost and Income Estimates .......... . . . . . . . . . 72 Monetary Production Costs ......... 73 Monetary Income .............. 80 Monetary Returns to Land and Management . . 83 Data Analysis ................. . . . 83 Agrophysical Analysis . . . . . . . 84 Selection of GDSSs with Rice Production Potention ................... 84 Estimation of Potential Area for Rice Production ............... . . . 84 Potential for Multiple Rice Production Cycles .................... 86 Current Land Use ...... . . ..... 86 Potential Area for Rice Expansion . . . . . . . 86 Economic Analysis . . ...... . . . 86 Benefit- Cost Analysis and Input Accounting . . 87 Unsubsidized Production Cost and Income Estimates ............ 87 Unsubsidized Production Costs . . . . . . 88 Irrigation Infrastructure Investment . . 92 Unsubsidized Income ........... 92 Unsubsidized Returns to Land and Management .............. . 95 Comparisons of Production Coefficient Estimates .............. . . . 95 Partial Budgeting Analysis of Rice Area Expansion .......... . . ....... 95 vi Chapter Page V. ANALYTICAL RESULTS ................... 100 Agrophysical Results ................ 100 GDSSs with Rice Production Potential ....... 101 Limiting Factors ................. 103 Potential Physical Area for Rice Production . . . 103 Potential for Multiple Rice Production Cycles . . 105 Current Land Use ................. 105 Potential Area for Rice Expansion ........ 106 Economic Results .............. . . . 106 Benefit- Cost Analysis and Impact Accounting . . . 107 Farm Gate Yields ........ . . . . . . . 107 Monetary Costs of Production . . . . . . . . . 109 Unsubsidized Costs of Production ...... . 111 Rice Production Cost Subsidy ......... 111 Monetary Returns at the Farm Gate . . . . . . . 112 Unsubsidized Returns at the Farm Gate ..... 114 Labor Use in Production ..... . ..... 115 Foreign Exchange Use in Production ...... 115 Partial Budgeting Analysis . . . ....... 117 Rice Production Maximization Strategies . . . . 121 Maximization of Rice Monetary Returns at the Farm Gate ................. 122 Maximization of Rice Unsubsidized Returns at the Farm Gate ............... 123 Maximization of Rice Labor Use ..... . . . 124 Minimization of Rice Foreign Exchange Use . . . 125 Ranking of Expansion Strategies by Production Impacts .............. 125 Results of Sensitivity Analysis ........ 126 Chapter V's Relation to Chapter VI ......... 127 VI. DISCUSSION OF ANALYTICAL RESULTS . . ...... . . . . 128 In What Land Areas Can Rainfed and Irrigated Rice Production be Expanded? . . ......... 128 Which Expansion Areas Would Be the Most Profitable from the Producer's and the Nation' 5 Standpoints? . . 131 Where Can Labor be Increased and Foreign Exchange Use be Decreased in Rice Area Expansion? ...... 134 Are Alternative Rice Production Techniques Available Which Could Profitably Increase Labor and Decrease Foreign Exchange Use? ......... 135 How Much Are Rice Production Costs Currently Being Subsidized and What Would be the Impact of Subsidy Removal? ................. 136 Can a Land Assessment Procedure be Developed Which is ApprOpriate to the Dominican Planning Environment? .................... 137 Modifications Needed in the Regional Study for National Application ............ . . . . 140 The Need for Additional Information ......... 142 vii Chapter VII. SUMMARY AND CONCLUSIONS ................. Summary ...................... Conclusions ..................... BIBLIOGRAPHY viii Table 2-1. 4-8. 4-9. 4-10. 4-11. LIST OF TABLES Groupings of Dominant Soil Subgroups (GDSS) in the Central Region, 1979 .................. Nine Agrophysical Input Requirements for Seven Rice Varieties, Central Region, 1979 ...... Water Availability from Three Sources and Adaptability for Rice Production in 15 RPUs, Central Region, 1979 . . Soils Characteristics of the 60555 with Adequate Water for Rice Production, Central Region, 1979 ..... Sample SIEDRA Production Budget for Rice, Central Region, 1979 Typical Production Techniques ....... Labor Use per Cycle in Rice Production Activities, Central Region, 1979 Current Normal ......... Current Normal Annual Labor Use in Sugarcane and Cultivated Pasture/Milk Production, Central Region, 1979 ...................... . Inputs of Water Soil Slope and Depth Required to Attain Specified Levels of Yields for Five Rice Varieties, Central Region, 1979 Current Normal ..... Rice Yields Attainable with Various Combinations of Water and Woil Slope and Depth, Central Region, 1979 Current Normal ................ Comparison of Input Differences Between Representa- tive Rice Production Budgets with Typical and Alternative Production Techniques, 1979 Current Normal ...................... Annualized Monetary Costs for State Owned and Pri- vately Owned Machinery for Rice Production, Central Region, 1979 ......... . . ........... Current Normal Monetary and Unsubsidized Costs for Typical Private and State-Owned Rice Production Machinery, Central Region, 1979 ............ ix 58 6O 62 64 66 69 7O 71 75 76 Table Page 4-12. Monetary Costs of Sugarcane Production for Three Sugarmill Areas, Central Region, 1973 Costs Indexed to 1979 ......................... 76 4-13. Monetary Costs of Production of Cultivated Pasture Used for Pasteurized Milk in Rainfed GDSS 078, Central Region, 1973 Costs Indexed to 1979 ........... 79 4-14. Monetary and Unsubsidized Product Prices, Farm Gate and CIF Import, Central Region, 1979 Current Normal . . . 81 4-15. Derivation of Farm Gate Monetary and Unsubsidized Prices for Sugarcane, Central Region, 1979 Current Normal ......................... 81 4-16. Derivation of Unsubsidized Marketing Costs from Mone- tary Marketing Costs for Sugarcane, Central Region, 1973 Costs Indexed to 1979 . . . ............ 82 4-17. Calculation of Farm Gate Monetary and Unsubsidized Prices and Income for Livestock, Central Region, 1979 Current Normal ..................... 83 4-18. Derivation of Annual Monetary and Unsubsidized Costs of Production per Tarea for Sugarcane, in Four Rainfed and Four Irrigated 60855, Central Region, 1973 Indexed to 1979 ..................... 90 4-19. Derivation of Annual Monetary and Unsubsidized Costs of Production per Tarea for Cultivated Pasture/Milk in Four Rainfed and Seven Irrigated 00335, Central Region, 1973 Indexed to 1979 .............. 91 4-20. Derivation of Unsubsidized Farm Gate Prices from CIF Import Prices for White Rice and Reconstituted Milk, Central Region, 1979 Current Normal ........ 92 4-21. Derivation of Unsubsidized Marketing Costs for Rice, Central Region, 1979 Costs ............... 94 4-22. Assumptions on Five Decision Variables and Three Policy Variables for 11 Strategies for Expansion of Rice Production Area, Central Region .......... 96 4-23. Sample Partial Budgeting Calculations for Rice Expan- sion, Central Region, 1979 Current Normal ..... . . . 97 5-1. Agrophysical Adaptability of Rice to Nine Water and Soil Characteristics in 29 GDSSs, and Major Limiting Factors, Central Region ................. 102 Table 5-2. 5-3. 5-6. 5-7. 5-8. 5-10. 5-11. Four Agrophysical Characteristics of the 12 GDSSs with Rice Potential in the Central Region, 1979 . . . . Current Normal Farm Gate Yields for Rice Typical and Alternative Production Techniques, Sugarcane and Cultivated Pasture/Milk by Rainfed and Irrigated GDSSs, Central Region, 1979 .............. Current Normal Monetary and Unsubsidized Costs of Production for Rice Typical and Alternative Produc- tion Techniques, Sugarcane and Cultivated Pasture/ Milk in 12 GDSSS, Central Region, 1979 ........ Current Normal Rice Production Cost Subsidy for Typical and Alternative Production Techniques, Production Cycles Area, and Total Subsidy for 12 GDSSs, Central Region, 1979 .............. Current Normal Unsubsidized Returns to Land and Management Rice Typical and Alternative Production Techniques, Sugarcane and Cultivated Pasture/Milk in 12 GDSSs, Central Region, 1979 ........... Current Normal Labor and Foreign Exchange Use for Rice Typical and Alternative Production Techni- ques, Sugarcane and Cultivated Pasture/Milk in 12 60555, Central Region, 1979 .............. I Five Regional Impacts of 11 Rice Expansion Strate- gies and Two Modified Strategies, Central Region, 1979 Current Normal . ................ Current Normal Area and Area of Rice Expansion for 11 Strategies and Two Modified Strategies for 12 GDSSs, Central Region, 1979 ............. Choice of Typical and Alternative Production Techniques for 11 Strategies and Two Modified Strategies for 12 GDSSs, Central Region, 1979 Current Normal .................... Ranking of Rice Expansion Strategies on the Basis of Four Impacts, Central Region, 1979 Current Normal ........................ xi BERG 104 108 110 112 113 116 118 119 120 126 Figure 1—1. 1-2. 1-3. 2-1. 3-1. LIST OF FIGURES Domestic Rice Production, Consumption, and Imports for the Period 1962-1980 and Projections for 1985 and 1990 (SEA, 1977; INESPRE, 1979; AID, 1980) . . . . Location of the SEA Central Region in the Dominican Republic . . ................... Land Resource Assessment as a Single Input into Rice Policymaking .................. Resource Planning Units (RPU) and Major Rivers in the Central Region ................. Seed Production Response Surface with Two Variable Inputs and the Plant's Tolerance Limits (isoquant a = 0) for the Inputs with No Factor Interaction . . . . Geographic Distribution of Plant A Under Various Assumptions on Tolerance Ranges and Environmental Characteristics ................... Derivation of 1979 Input Price Index from BAGRICOLA 1973-78 Data (1973 - 100), Dominican Republic . . . . xii 11 17 31 32 78 CHAPTER I INTRODUCTION Problem Statement Rice is the most important staple in the diets of both rural and urban Dominicans. It provides an estimated 25 and 22 percent, respective- ly, of the daily caloric and protein intake (SEA, 1976). Domestic demand for rice has been increasing rapidly during the past two decades. This increasing demand is due to an average annual population increase of three percent and to a high positive income elasticity at low Dominican income levels coupled with increasing real incomes (Drilon, 1977). Domestic rice supplies have not kept pace with the demand increases, an imbalance resulting in the annual importation of rice since 1972 (Figure 1-1). This problem was exacerbated in 1973-74 when large increases in export prices for sugar caused many producers to shift out of rice and into more profitable sugarcane production (Juma, 1979). A 1979 hurricane contributed to a dramatic decrease in production and is expected to re- sult in the importation of large quantities of rice in 1980. As a result of the post-1974 weakening of world prices for sugar, a crop which traditionally has accounted for about 50 percent of Dominican foreign exchange, and the continuing rapid rise in petroleum import costs, the use of increasingly limited foreign exchange to import rice came under heavy questioning in the mid-19705. A vociferous nationalistic I 4 1' l /' /' /’ .I I I I I ’1 I: I / 2: 3 I lg’ I 8 / o / e" ,’ ' F. ' 1’ e, Consumpt1on I I +3 1‘. ’ ’ o I \ I I D. e I ’ E o T‘ I \ .I' " . ,..a .1 8 ° / .5 D. E :3 m C C U é‘ .2 4.1 U 3 13 O S. Q 73 3 C C < Imports 1” (I I I I I 1962 65 7o 75 80 2% 90 Figure 1-1. Domestic rice production, consumption, and imports for the period 1962-1980 and projections for 1985 and 1990 (SEA, 1977; INESPRE, 1979; AID, 1980). majority of agricultural spokesmen claimed that the country could become self-sufficient in rice production through implementation of stronger government rice production programs both to increase productivity on cur-. rent riceland and to expand current production areas (SEA, 1976a, 1976b). Reduction of high levels of rural unemployment and under-employment was consistently cited as a justification for increasing domestic rice pro- duction (SEA, 1979b). A small group of spokesmen suggested that continued rice importation was the most efficient alternative for meeting domestic consumption needs, as increased domestic production would require in- creased foreign exchange use for importation of production inputs, thereby offsetting the savings from discontinuing importation of the rice itself. In 1978 the controversy was made at least temporarily moot when the Dominican government (GODR) began a multiagency program to promote self- sufficiency in rice production (SEA, 1979b). No documentation of support- ing economic or agronomic analysis was published, and though no definition was stated for "self-sufficiency," it is presumed to have meant that the country would attempt to meet its rice consumption needs with domestic production as long as current relative input and output prices remain fairly stable. The specific purpose of the program was to raise the 1978 national average rice yield by 25 percent in four years, from 6.6qq/ta to 8.2 qq/ta (1 qq = 100 lbs.; 1 ta — 1/15.9 ha). This was to be accom- plished by increasing irrigated rice area by 18 percent and by eliminat- ing rainfed upland rice production. For the longer term, it has been estimated that at least 25,000 ta must be brought into rice production annually at the proposed 8.2 qq/ta yield level in order to meet projected domestic consumption needs without imports through 1990 (A10, 1980). To date, the governmental program has benefited little from scienti- fic analysis of either the positive or normative aSpects of the multitude of dynamic technical, institutional, and human factors that bear on the problem of sustained rice production increase. Quantitative and qualita- tive multidisciplinary analyses of such dynamic factors as present and future land resource base capabilities for agricultural production, pro- duction technologies, land tenure, input and output marketing, domestic and export demand, input and output substitution, income distribution, price policies, and sectoral and intersectoral comparative advantage are required in order to provide an economically sound and socially desirable answer to the questions of where, when, and how to increase domestic rice production. In view, however, of the limited qualified human resources available for research in the Dominican Republic and of costs of develop- ing each of these kinds of information, research efforts should be focused on developing the priority information known to have a critical bearing on the problem. Thus, a necessary early step in assessing the options for increasing domestic rice production is an analysis of the supply side of the rice production capacity of the Dominican land (soil and water) resource base as it relates to the critical factors of labor and foreign exchange use. As total production is a function of area and yield per unit area, studies are needed both of land base capabilities for extension of cur- rent areas and of alternative techniques to increase per ta productivity on current rice areas. Information on the agronomic and physical ("agrophysical") adaptability of rice plants to each land unit and on the profitability of current and alternative production techniques is needed in order to select priority areas for investment of financial and technical resources. Economic analysis of the benefits and costs of rice production increases both to the producers, in monetary (cash) terms, and to the national as a whole, in unsubsidized (opportunity cost) terms, would aid in providing a foundation on which a more efficient and realistic regional and national rice expansion program could evolve. Very little of these kinds of information on the land resource base pro- duction potential was available in the Dominican Republic when the present study was undertaken. Study Background Evaluations of projects dealing with the development of economic analysis capabilities in developing countries frequently point out that such projects often attempt to transfer inappropriate and overly SOphis- ticated analytical techniques to the thin veneer of local technicians. These technicians often are unable themselves either to apply these techniques appropriately to the typically rudimentary domestic data bases or to establish their credibility among the bureaucrats and poli- ticians who are asked to support the analytical work (Amin, 1979; Rossmiller, pt 21,, 1977). An alternative approach to developing the capability for economic analysis is to start with less sophisticated static and partial analyses and then to proceed, as local technical capability and reliable data is developed over time, to more sophisti- cated and complex static and dynamic general equilibrium analyses (Edwards, 1966; Kornai, 1975). This was the approach taken by the Comprehensive Resource Inventory and Evaluation System (CRIES) project when it initiated work in the Dominican Republic in 1977. At that time Dominican capability to under- take agaonomic and economic ("agroeconomic") analysis of agricultural land use alternatives was virtually non-existent. The CRIES project was initiated as a cooperative effort among the United States Department of Agriculture's Economics, Statistics, and Cooperative Service, the United States Agency for International Development, and the Michigan State Uni- -versity (CRIES, 1976). The project attempts to strengthen capacity in developing countries for development and analysis of data on land resource base capabilities for agricultural production. This data base will pro- vide information on food and fiber supply options for national and inter- national level decisionmaking. In the Dominican Republic the CRIES counterpart staff is referred to as the Sistema de Inventario y Evalua- cion de los Recursos Agropecuarios (SIEDRA). The author of this paper, while serving as the CRIES resident advisor to the SIEDRA staff during the period 1977-80, recommended to the Secretariat of Agriculture (SEA) that SIEDRA carry out the agroeconomic analysis of the various land areas of the country in order to identify areas appropriate for rice expansion. As this study was SIEDRA's first analytical effort, it was conducted on an experimental basis for the SEA administrative region closest to the SIEDRA offices in Santo Domingo. At that time the Central Region was the only one with sufficient soils and water data on which to base a study of this nature. It was also located near enough to the SIEDRA offices to provide cost efficient, coordinated field and office methodology testing and training for the SIEDRA staff. After review of the regional study results and incorpora- tion of any needed modifications in data collection and analysis methodologies, the rice study was to be expanded to the remaining six SEA regions. Approval of the regional study was given by the Secretary. Preliminary work began in the late spring of 1979, and was based on the existing SIEDRA benchmark land use data base for the SEA Central Region (Figure 1-2). It is this regional rice study that is the subject of this paper. Study Objectives The immediate purpose of the study was to make available to the Secretary of Agriculture information on locations where rice production could be expanded economically. This information was to be provided through an agroeconomic land resource assessment for selected rice pro- duction alternatives, with a focus on labor and foreign exchange use. The specific objectives of this study were to: 1. Develop a long resource assessment methodology appropriate to current Dominican technical and administrative capabilities; 2. Determine the agrophysical requirements for selected rice varieties; 3. Identify land areas with the required agrophysical character- istics; 4. Develop, for each suitable land area, rice enterprise budgets for typical current production techniques and for an alternative set of production techniques which would provide increased labor use and/or decreased use of foreign exchange; 5. Develop typical current normal budgets for the major land use alternatives (sugarcane and cultivated pastures) to rice production on the selected land areas; oanaamm :mowcwsoa ecu cw cowmmm Pmcpcmu on caduceuomcH auuucoanasoo ouououm cuesm AemamHmv «cosmoo0a< consumes nuououm wand Huoecnooa manuaau>< ceauuauomcH .:owvaasocu coaeoo 12 and periodic updating of the information is crucial to improving the effectiveness and efficiency of the rice policy. Reader's Guide to the Dissertation The dissertation is divided into seven chapters plus appended materials. Chapter I, the Introduction, contains a statement of the problem and the background and objectives of the study. Chapter II provides a general description of the land base of the SEA's administrative Central Region which was analyzed in the study. There is a brief explanation of the CRIES/SIEDRA land classification system and a description of each land unit ("GDSS"), with general infor- mation on soils, climate and rainfall, and surface and underground water availability. Land use in the Central Region is described in aggregate terms. There is a section covering the various aspects of rice produc- tion and consumption--area, output, irrigation, labor use, imports, and the institutional factors of land tenancy, price controls, and agronomic research. Concluding Chapter II is a similar discussion of the two primary competitors for land suitable for rice production: sugarcane and cultivated pastures. Economic theory and literature relevant to the study is reviewed in Chapter II, which begins with a discussion of the theoretical frame- work for analyzing the agroeconomic adaptability of rice plants to their environment and the economics of rice production. There follows a review of significant international and Dominican empirical research and, finally, a section explaining the CRIES/SIEDRA methods used to establish the benchmark land use data with which the study was initi- ated. 13 Chapter IV deals with the research methods utilized in the study and begins with an explanation of the overall concept of the research. Data acquisition and refinement into coefficients on which economic analysis was carried out are then discussed. First, the procedure for collection and refinement of the agrophysical data pertaining to rice agronomic requirements for critical soil and water inputs and the availability of those inputs in the Central Region are discussed. Described also are the corresponding rice yields in each land area ("GDSS"). Current GDSS use for the production of rice and its two major land use competitors, sugar- cane and cultivated pasture used for milk production, is then covered. Second, the collection and processing of the economic data pertaining to production costs, processing and transportation costs, labor and foreign exchange use, and output prices at the farm gate and import levels are detailed. There is a narrative concerning the application of the national opportunity cost concept to each input and output in order to convert monetary (cash) costs into unsubsidized (national opportunity) costs. Third, there is a discussion of how the agrophysical and economic data were analyzed in order to determine in which GDSSs it would be feasible to produce rice agrophysically and in which of those areas output, -monetary and unsubsidized profitability, and labor use could be maxi- mized and foreign exchange use minimized under current normal economic conditions. The chapter closes with a discussion of the means by which selected data uncertainties are tested in the study. In Chapter V the results of the agrophysical and economic analyses are presented. The 60885 to which rice is adapted on the basis of agro- nomic and physical requirements are given, along with the estimates of current rice production area in each GDSS, potential new GDSS area 14 available, number of rice production cycles possible, and expected yields for the current typical and alternative techniques for rice production. Production costs, farm gate income, and resulting returns to land and management (RLM) for the two sets of rice production techniques and for sugarcane and cultivated pastures are given for each GDSS from both the producer's (monetary) and the nation's (unsubsidized) points of view. Labor and foreign exchange use and rice subsidy levels for each produc- tion option are presented. Finally, the results of the partial budgeting of the costs, RLM, and labor and foreign exchange use under each of 11 expansion strategies for rice production area are given. Chapter VI contains a discussion of the implications of the analy- tical results for addressing the questions relating to regional rice policy which were posed in the first chapter. Consideration is given to the limitations inherent in the data acquisition and analytical methods used in the study. Included is a discussion of methodological modifications needed for a national level land resource assessment for rice production. The chapter closes with a reaffirmation of the need for additional types of information to augment this land resource assessment. Chapter VII contains a summary of the results of the study. Also, significant conclusions drawn from the study are presented. CHAPTER II REGIONAL SETTING OF THE STUDY The study deals with the SEA administrative Central Region, which has an area of 6,983 km2. The 1970 population was about 1.3 million, almost a quarter of the national total. Included within the boundaries of this region are the country's capital city, Santo Domingo, and two of the principal ports of the country. Land Resource Base Land Classification The land base of the Central Region has been classified according to the CRIES/SIEDRA land classification system (CRIES, 1979b; SIEDRA, 1979a). This system incorporates two interrelated conceptual units: the Resource Planning Unit (RPU) and the Grouping of Dominant Soil Sub- groups (GDSS). An RPU is described as follows: An RPU is a mappable unit of land relatively homogen- eous with respect to climate, vegetation potential, and soil distribution. It is useful for the planning of data collec- tion for national and regional level land use assessment and generally has readily discernible natural boundaries (soils and vegetation types). RPUs commonly consist of contrasting soil bodies, defined elsewhere as 60585, that are associated geographically in recognizable and definable patterns. The RPU is thus a cartographic unit within which GDSSs are identi- fied and is derived by overlaying a climatic zone map on a soils map and 15 16 delineating each unique soils/climate combination area. A GDSS is described as follows: A GDSS is a single, dominant and distinct soil subgroup or a grouping of agronomically similar subgroups relatively homogeneous with respect to climate and vegetation potential. It is characterized by soil parameter values from which pre- dictions can be made about agricultural land use, management practices, and potential levels of production as a basis for national and regional level land resource assessment and agri- cultural production planning. GDSSs can be represented by single-valued parameter estimates of agricultural factors such as plant adaptability and agronomic input-output coeffi- cients. The 60385 within individual RPUs are visually dis- tinguishable in the field by agriculturalists on the basis of important agronomic differences such as slope and surface drainage. 60555 are not mappable nationwide due to limited availability of soils mapping at the individual subgroup level and below, but are identified as percentage components of individual RPUs. The GDSS is thus the analytical unit which SIEDRA uses for national and regional level resource assessments in the Dominican Republic. Under this classification system the SEA Central Region is composed of 15 RPUs and 29 60555 (Figure 2-1 and Table 2-1). Physiography Approximately 65 percent of the Central Region is made up of level to rolling plains, the balance being part of the mountainous Cordillera Central. Elevations range from sealevel on the southern Caribbean coastline to almost 2600 m in the northwestern mountains. Most of the agriculturally significant watersheds originate in the western mountains, and all of the region's rivers drain southward into the Caribbean Sea. Climate The Central region is located three to four degrees south of the tropic of Cancer. Monthly extreme low and high temperatures typically 17 co wmmm Pmcucmu any cw mcm>wm Lewwz new Azmmv wows: mcwccmfia moc30mmm .H-N mazmwc 18 Table 2-1. Groupings of Dominant Soil Subgroups (GDSS) in the Central Region, 1979 . Distinguishing Estimated RPU GDSS Major Subgroup Characteristic Area (1,000 ta) 01 01A Typic pellustert level 240 018 Lithic ustropept hilly 129 02 02A Lithic dystropept mountainous 3,020 028 Typic dystropept valleys 1,009 06 06A Typic pellustert levels 212 068 Lithic ustropept hilly 141 07 07A Plinthic tropaquept poorly drained 855 078 Aquic dystropept well drained 700 09 09A Lithic eutropept hilly 676 098 Typic eutropept valleys 255 11 11A Lithic ustorthent hilly 46 118 Lithic ustrOpept valleys 31 16 16A Aeric fluvaquent poorly drained 70 168 Typic ustifluvent well drained 129 19 19A Aquic eutropept valleys 160 198 Lithic eutrOpept hilly 131 20 20A Typic tropudult undulating 450 208 Fluventic dystropept level 247 21 21A Aquic dystropept moderately drained 33 218 Fluventic dystrOpept well drained 15 22 22A Aquic dystropept undulating 242 228 Lithic eutropept hilly 204 23 23A Lithic ustorthent shallow 474 25 25A Typic camborthid level 227 258 Typic camborthid (fans) hilly 122 40 40A Typic dystropept moderately deep 207 408 Lithic dystropept shallow 254 41 41A Lithic ustropept hilly 117 418 Typic camborthid undulating 39 Total 11,103 19 vary from around 15-35 degrees C at the low elevations to 0-25 degrees C in the mountains. . Average annual rainfall varies from about 2000 11m in the northeast to 600 mm in the semi—arid southwest, with monthly peaks in May and October. Annual potential pan evaporation varies from approximately 1500 mm in the southwest to 800 mm in the northeast, with a monthly peak in July (SIEDRA, 1979b). Soils Soils characteristics vary widely, with textures ranging from sands to clays, pH from very strongly acid to moderately alkaline, depth from a few mm to over 5 m, and slope from level to over 100 percent. Most plains soils were formed from coral bedrock, while the mountain soils typically were formed from igneous and metamorphic rocks and limestone and shale deposits. There are no volcanic soils in the region nor in the rest of the country. Natural fertility levels vary from very low to moderately high. Water In addition to direct rainfall there is surface and underground water available for agricultural production in many parts of the Central Region. There are currently about 43,000 ta under (1 ta - 1/15.9 ha) irrigation in the region. The three major sources of surface water are the Ozama, Nizao, and Haina Rivers (Figure 2—1). The Nizao River is dammed and provides year-round irrigation water except during periods of extreme drought. Underground water currently is used very little for irrigation in the Central Region. The limited hydrogeologic 20 information available on the possibilities of developing further ground- water sources for irrigation in the region indicates high probability of success in several small areas (INDRHI, 1979). Current Land Use Recent land use in the Central Region is apportioned approximately as follows: 37 percent for cropland, 22 percent for rangeland and cultivated pastureland, and 13 percent for forestland (CRIES, 1977a). The cropland is devoted largely to sugarcane (60 percent) and to rice (15 percent). Rangeland consists mostly of non-cultivatable former savannahs now dominated by secondary series of brush, grass, and other herbaceous plants, and of dry mountainsides dominated by similar types of vegetation. Cultivated pastureland produces primarily three species of grass forages: guinea (Panicum maximum Jacq.), pangola (Digitaria decumbens Stent) and African stargrass (Cynodon plectostachyus K. Schum.). Forestland has not been exploited legally for commercial purposes since a 1966 logging ban was imposed, although semi-officially sanc— tioned "illegal" tree cutting for firewood and charcoal production is widespread. A 1974 FAO study indicates that the major forest species in the Central Region are mahogany (Swietenia mahogoni (L.) Jacq.) and Western White Pine (Pinus occidentalis Sw.). The Rice Industry Rice Consumption As was noted in Chapter 1, Dominican rice consumption has been increasing steadily for over two decades. Annual per capita consumption 21 of rice has increased from less than 0.7 qq (1 qq = 100 lb.) in the mid—19605 to about 1.1 qq in the late 19705 (A10, 1980). By far the largest consumption market in the country is the capital city of Santo Domingo which contains almost 20 and 90 percent, reSpectively, of the national and regional populations. A 1969 Central Bank survey in Santo Domingo indicated that the average family allocated 10 percent of total food expenditures to rice. The corresponding figure for low income families was 17 percent. Assuming that the Santo Domingo population is representative of Central Region rice consumers, over 90 percent of regional rice production is consumed in this metropolitan area. Rice Production Domestic Production and Area Rice has been produced in the Dominican Republic since before the time of Christopher Columbus' European discovery of the island at the end of the fifteenth century. Commercial rice production, however, began only about 40 years ago (SEA, 1968). In 1977 total commercial white (polished) rice production in the Central Region was approximate- ly 8,200 MT, or 15 percent of the national total. This accounted for about 15 percent of total regional agricultural production value (SEA, 1977a). Although time series data on which to establish trends are highly variable among sources, it is estimated by the World Bank (1978) that national rice production has been increasing annually by about five percent over the past two decades. Physical area utilized for regional rice production was approximate- ly 51,000 ta in 1977. Over 90 percent of the commercial rice is irri- gated, while less than 10 percent of the subsistence rice is produced 22 under irrigation. Nearly 70 percent of the irrigated riceland is double cropped, and about 10 percent produces three crops. Approximately 22 percent of the second and third crops was produced as a volunteer ("ratoon") crop in 1976 (SEA, 1977b). Only about three percent of the riceland is intercropped with one or more additional plant species. Small farmers with fewer than 80 ta of riceland produce almost 90 percent of the country's rice. A 1972 law called for expropriation of irrigated rice farms of more than 500 ta and resulted in the establish— ment of 465,000 ta of new collectivized agrarian reform settlements ("asentamientos") with an average family parcel size of 50 ta. The collectivized and single family types of asentamientos are controlled by the Dominican Agrarian Institute (IAD) and currently produce almost half of the country's rice (IAD, 1979). Agrarian reform asentamientos of both types produce approximately 80 percent of the rice crop in the Central Region (SIEDRA, 1979d). While government paternalism and "pork barrel" politics have dominated IAD administration since its inception a half century ago, recent emphasis has been on the fostering of the attitude among reform settlers that they are businessmen operating their own "cooperatives" with reduced government intervention. Input Use Irrigated rice paddies are typically quite small and are built with contour ridges, a Chinese innovation unique to the western hemisphere (AID, 1980). These small paddies do not permit the use of large machin- ery. Land preparation is accomplished by small tractor or draft animal- assisted plowing. Seeds are broadcaSt by hand on rainfed upland sites. Rice seedlings, of which almost 60 percent were modern varieties in 1976, are hand transplanted from on-farm nurseries to flooded paddies 23 (SEA, 2979a). Flooding is regulated in order to reduce the need for hand weeding, and insecticides and fungicides usually are applied to assure plant vigor. In 1976 close to 47 percent of rainfed and 94 per- cent of irrigated rice was fertilized (SEA, 1977a). Harvesting and threshing typically is done by hand. Thus, rice is a relatively labor intensive crop in the Dominican Republic, requiring approximately 5 man/days/ta annually for production and harvesting. The small farms of fewer than 80 ta use slightly more than average labor, with about 40 percent of it provided by the family. On large, multiple product farms of more than 560 ta only about four percent of the labor is provided by the producer's family (SEA, 1977b). Credit is provided for rice production primarily through the government's Agricultural Bank (BAGRICOLA). Interest rates are fixed at 9 and 11 percent, respectively, for loans of less than and more than RD$2,000. This rate differential ostensibly was established to benefit small producers. The legal maximum nominal rate of interest chargeable is 12 percent. In recent years BAGRICOLA has been increasing its lend- ing to rice producers, with around 70 percent of its food crop loans going into rice credit in 1978. Repayment for agrarian reform asenta— mientos have been about 75 and 100 percent, respectively, for individual family and collective farm loans (A10, 1980). Rice Imports Commercial rice has been imported annually since 1971, reaching a high of over 70,000 MT in 1974. The 1980 figure is expected to be at least 45,000 MT and could go as high as 90,000 MT. Import prices peaked in 1974 at US$576 per MT, and the 1980 prices are expected to be about 24 US$450 per MT. Foreign exchange use for rice imports was US$40 million in 1974 and is expected to reach US$19-38 million in 1980 (INESPRE, 19801 These figures represent, respectively, about 25 and 15 percent of total national commercial rice retail value. The National Price Stabilization Institute (INESPRE) controls all rice importation and coordinates it with its public food marketing and price control policies. All of the rice imports come from the United States. There have been no PL-480 rice shipments to the Dominican Republic to date (Agricultural Attache, 1980). Rice Marketipg Many aspects of rice marketing are controlled by INESPRE. Since 1974 INESPRE has functioned, as a result of private trade speculation during a period of rice shortages that year, as the monopsony buyer of milled rice from the millers. It also controls rice imports and operates a number of grain silos and warehouses. INESPRE has bought almost 90 percent of market rice production in recent years. There are over 100 rice mills of various sizes in the country, with an estimated annual milling capacity of 275,000 MT. Five of the larger mills are located in the Central Region. In general the mills produce good to high quality rice with a relatively low percentage of broken grain. Although the mills operate most of the year, their major Central Region processing load comes during the primary harvest period between August and January. Purchasing of rough rice at the farm, except in the case of large growers and IAD asientamientos, is handled by private intermediaries who deliver the rice to mills. Small farmers frequently complain about being cheated by the buyers on moisture 25 content and grading, and several compesino groups have responded by building and managing their own rice mills. Most of these efforts have ended in failure due to lack of adequate management skills and to under- estimation of mill operating costs. Rice Agronomic Research Rice agronomic research is carried out primarily at the Juma Experi- ment Station located about 10 km north of the Central Region's northern boundary. Recent research has concentrated on the selection of irrigated varieties of rice adaptable to labor intensive agronomic practices on small farms. Testing of rainfed varieties began in late 1979 (Juma, 1979). To date there has been relatively little use made of research results by the national extension service (Drilon, 1977). The Sugar Industry Sugar plays a dominant role in the Dominican agricultural sector and in the national economy. In recent years taxes on sugar exports have accounted for nearly 20 percent of the central government's total current revenues. Sugarcane occupies about 12 percent of national cultivated area and generates about 50 percent (US$100 million) of total exports. In the Central Region sugarcane occupies about 60 percent of cultivated land and accounts for nearly 70 percent of agricultural pro- duction value. The industry provides direct employment for an esti- mated 80,000 workers in production and processing and provides additional indirect employment through input marketing and related services (World Bank, 1979). About 65 percent of all sugarcane is produced on land owned by the State Sugar Council (CEA); the balance is produced on 26 private lands owned by the Dominican Vicini family and the US-based multinational Gulf and Western Corporation. Annual raw sugar production has averaged about 1.2 million MT in recent years, up only slightly from the 1960 average of 1.0 million MT. Average sugarcane yield declined by nearly 5 percent annually from 1963 to 1975. It has reportedly averaged about 3.8 MT per ta harvested since that time. Due to the use of low cost Haitian labor for hand harvesting of cane, Dominican sugarcane production is more labor inten- sive than in most other Latin American sugar producing countries. Hand harvesting also results in lower trash content and higher sucrose con- tent than that of machine harvested cane (Bookers, 1975). In recent years a growing portion of cultivated area has not been harvested because of limited processing capacity. Under normal condi- tions and full utilization of present milling capacity, CEA and private sugar factories can produce about 1.3 million MT of raw sugar annually. A 1979 World Bank loan is being used to finance an increase of CEA milling capacity from 0.8 million MT to 1.2 million MT. Approximately 14 percent of the annual sugar production goes into domestic consumption. In addition, dried cane pulp and molasses are consumed domestically, to a limited extent, by the CEA and by livestock producers. Domestic refined sugar prices are controlled by INESPRE. Sugar exports increased by about one percent annually between 1960 and 1978. Exports have averaged just under 1.0 million MT in recent years, marketed primarily under the International Sugar Agreement. The US bought as much as 93 percent of Dominican sugar exports in the late 19605, but its market share declined to about 75 percent in the late 19705. The Dominican Republic recently has diversified its export mar- kets to include Italy, Iran, Portugal, Sweden, and Venezuela. 27 The Cultivated Pasture/Milk Industry In 1976 the SEA estimated that approximately 59 percent of the milk produced in the Dominican Republic was consumed as fresh milk, 18 per- cent was pasteurized, 10 percent was used for cheese and butter, and the remainder was fed to livestock or was discarded as spoiled. Pro- ducers sold 65 percent of their marketed milk to wholesalers or retailers, 19 percent to milk processors, 10 percent directly to consumers and retailers, and the remaining 6 percent was sold to the Public Nutrition Program or processed on farm. There is currently a strong trend toward increased sales to pasteurizing plants (Associacion, 1979). The typical dairy herd produces around 275,000 lt of fresh milk annually. In recent years the number of milk cows with high genetic production potential has been increasing very slowly. Most milk cows do not produce at or near their potential due to poor herd and pasture management, unfavorable input-output ratios, and unfavorable climate. As a result, production of milk has been increasing at a slow rate for over a decade. Total milk production was 246 million lt in 1960 and about 383 million It in 1976, an average annual increase of around 3 percent. Based on this trend, milk production is estimated at 415 mil- lion lt for 1980 and 500 million lt for 1990. There were eleven major milk processing plants in the country in 1977, of which four were not in operation. The plants in operation were being utilized at 35 to 90 percent of capacity. Milk processors claim there is adequate demand to allow fuller utilization of capacity if greater supplies were made available. Current information on imports of powdered milk and processed milk products is very limited. The SEA (1977a) estimated that during the 23 early 19705 approximately 20 to 25 percent of total milk consumed was imported. The value of dairy imports in 1977 was US$1.5 million and con- sisted largely of 3.2 million kg of powdered milk. Trends indicate that at least 20 percent of total milk consumption will continue to be imported during the 19805. INESPRE controls retail milk prices and has kept them low in recent years, ostensibly to protect low income families. Many producers have abandoned dairy farming due to unprofitability, and others claim that the unfavorable price-cost ratios have not allowed them to invest in productivity-increasing inputs. As a result of lobbying efforts by milk producers, INESPRE in late 1979 announced a 30 percent increase in the retail milk price. It also took two additional steps aimed at benefiting low income consumers and at reducing the production-depressing effects of continued imports of powdered milk which were being dumped by exporting countries at less than their costs of production. The first step was to make available at low cost to low income consumers a low fat blend of fresh and reconstituted milk, and the second was to place the importation of powdered milk under the control of INESPRE, which was directed to coordinate the imports with its coverall programs of marketing and price control. CHAPTER III REVIEW OF RELEVANT ECONOMIC THEORY AND LITERATURE Theoretical Framework There are two bodies of theory of primary relevance to this study: agronomic theory having to do with the plant's adaptability to specific environments and economic theory dealing with the difference between monetary (producer) and unsubsidized (national) benefit-cost analysis and its implications for resource allocation. These theoretical con- siderations are incorporated in the selection of methods in Chapter IV and in the discussion of study results in Chapter VI. Agronomic Theory Individual plant varieties have genetically determined physiologi- cal input requirements for normal growth and development. Within specific ranges of tolerance for input levels above and below the nor- mal requirements, a plant will grow but will exhibit abnormal growth habits. Outside of these tolerance ranges, plants cannot grow and reproduce (Evans, 1963). The tolerance ranges, together with the cor- responding ecological characteristics of the environment, determine the natural distribution of a plant (Holdridge, 1968; Odum, 1971). The major environmental characteristics affecting the natural geographic distribution of plants are photoperiod, temperature, percipitation, 29 3O evapotranspiration, and soil. Man's manipulation, either consciously or unconsciously, of these physical variables can alter natural plant dis- tribution significantly (Weaver, gt gl., 1939). A concept which has been used widely for analyzing plant geographic adaptability is that of "limiting factors" (Odum, 1971). Limiting factors are those inputs necessary to a plant's growth but which are available in quantities falling outside of the plant's tolerance ranges and thus do not permit the plant's growth in a given area. The concept of limiting factors is based on two ecological "laws“: Shelford's (1913) Law of Tolerance and Leibig's (1840) earlier Law of the Minimum. Leibig's theory, as expanded by later researchers, states that plant growth is controlled by the physiological input available to it in mini- mum quantity. Shelford's theory recognized both that too much as well as too little of an input factor and that interaction can affect a plant's tolerance ranges for specific inputs. In production function terms, Leibig's theory assumed perfect complementarity of inputs while Shelford's theory allowed for changing rates of substitutability among inputs and a "Stage III" of the production function (see discussion of stages of production later in this chapter). Limiting factors are those inputs necessary to a plant's growth but which are available in quanti- ties falling outside of the plant's tolerance ranges and thus do not permit the plant's growth in a given area. The concept of limiting factors is illustrated graphically in the hypothetical response surface in Figure 3—1. In the figure the iso- quants represent plant seed production with two variable inputs where: d c b a = 0. Under the assumption of no factor interaction, the isoquants are rectangular. For plant A, the ranges of tolerance for 31 Figure 3-1. Seed Production Response Surface with Two Variable Inputs and the Plant's Tolerance Limits (isoquant a = 0) for the Inputs with No Factor Interaction x2 7 5 .- 5 ,-———T 4 3 d 2 C b 1 a 1 2 3 4. 5 6 X1 / X3...Xn inputs X1 and X2 are 3-5 and 1-6 units, respectively. Thus, if plant A were grown in an environment which contained 9 units of X1 and 2 units of X2, the plant would produce no seed, and X2 would be the limiting factor. Along the vertical and horizontal axes of the isoquants the two inputs are perfectly complementary. With Shelford's factor inter- action the isoquants would be rounded at the corners as the inputs become imperfect substitutes and as diminishing marginal physical output reduces the tolerable input levels from the plant's tolerance extremes. The application of these principles to the spatial distribution of plant A is displayed in Figure 3-2. The base map in Figure 3-2a is a map of the soils containing X1 at levels of 0-6 units per unit area. The darkened areas are those areas which have the required 3-5 units of X1 for seed production on plant A. Figure 3-2b indicates soils with 32 Figure 3-2. Geographic Distribution of Plant A Under Various Assumptions on Tolerance Ranges and Environmental Characteristics 61 a adaptability to x1 ** 7' a adaptability to x; V c adaptability to X, and 112 d adaptability with modified tolerance to X; ///‘ e adaptability with modified x, available f comer-icon of current distribution with . ' - potential distribution 33 0-6 units of X2. Figure 3-2c is an overlay of the X2 map on the X1 map and indicates the area to which plant A is adapted on the basis of both factors X1 and X2, assuming no factor interaction. With factor interaction the resulting adaptable area in Figure 3-2c would be reduced slightly depending on the marginal rates of substitution between X1 and X2. This potential distribution of plant A could be modified by man through breeding to select or create genetic varieties with different tolerance limits for specific environmental factors. For example, if a variety of plant A could be selected which could tolerate 1—4 units of X1 and the same amounts of X2, with no factor interaction its poten- tial distribution would appear as in Figure 3-2d. Similarly, the potential distribution of plant A could be modified through environmen- tal modification to adjust the levels of X1 and X2 available to plant A. As an example, if the same geographic area as represented in Figure 3-2c had 1 unit of X1 applied to it uniformly, the potential distribu- tion of plant A would appear as in Figure 3-2e. Again, factor inter- actions would modify the results. These principles for determining potential plant geographic distribution have been used to identify areas to which plant species and varieties are adapted but in which they are not currently being produced (Burton, 1968; Kemph, 1973, 1975; Klages, 1942). The reasoning is as follows: if plant A's potential distribution based on inputs X1 and X2 were as in Figure 3-2e, but its current distribution as in the area marked "CD" in Figure 3-2f, then the area that could be considered for expanding plant A's current distribution would be represented by the darkened areas in Figure 3-2f. Factors besides X1 and X2 (i.e., X3.. 34 Xn) which may be limiting to the growth of plant A must be taken into consideration in analyzing current and potential geographic distribution. A current distribution as in Figure 6f may indicate that factors other than X1 or X2 are in fact limiting to plant A's distribution. This would be the case, for example, if an area's soil and water inputs were within a plant's tolerance limits, but no plants were growing there because of biological factors such as insect infestations or weed com- petition or because of economic factors such as unprofitable input-output price ratios or an unfavorable land tenancy situation. Economic Theory Five concepts from the theory of production economics are particu- larly relevant to this study: (a) efficiency, (b) resource valuation, (c) multiple production goals, (d) stages of production, and (e) input substitution. Each is discussed below. Efficiency For the purposes of this study, efficiency is defined as the maxi- mization of the difference between utility and disutility. Utility in turn is defined as the satisfaction of human wants and needs. If utility were measurable in units of common denomination which were comparable among economic agents, efficiency would be attained by simply maximizing the difference between the utility people derive and the utility they give up from alternative actions. However, there is no widely accepted-interpersonally comparable common denominator for utility, and actions which provide utility for some people while taking utility from others cannot be analyzed quantitatively from the effi- ciency standpoint without first defining the utility measure to be used in the analysis. 35 In a market economy with a given pattern of resource ownership and where economic agents have perfect knowledge and foresight and are free to buy and sell goods and services, prices represent the marginal utility that buyers and sellers receive from trading these commodities. These market prices also represent the marginal utility that a nation derives from the commodity trading. Under these conditions, efficiency in an activity would be attained by maximizing the net income derived from an activity. In an agricultural production activity efficiency would be attained by maximizing the difference between total income and total costs--profit maximization. Efficient resource allocation on a single-product farm requires that production inputs be allocated such that ratios of the marginal value of the last unit of output (marginal value product, MVP) to the marginal cost of each input used in its production (marginal factor cost, MFC) be equal. For multiple product farms these ratios must be equal among products. Profit maximization takes place where the ratios are all equal to one. If a ratio does not equal one, resource alloca~ tion can be adjusted to increase output for a given quantity of inputs. A ratio of greater than one would indicate that production should be expanded to increase profits, while a ratio of less than one would in- dicate that production should be contracted. For example, if the MVP/MFC ratios for rice and sugarcane were 1.2 and 0.6, respectively, profits could be increased by reallocating sugarcane production re- sources to rice production. In the long run when all inputs are adjusted to their optimum levels, marginal and average value products and factor costs are all equal for each input, and profit is maximized at zero. 36 Under conditions of imperfect knowledge and foresight a producer cannot be certain of profit maximization. He must choose a production decision rule which accounts for his inability to know with certainty the outcome of his production activities. In theory, many decision rules are possible, including maximization of expected profits, maximi- zation of the probability of some specific minimum level of income, or the outcome of a coin toss. The rule chosen by each producer depends on his personal and family values and goals and on his aversion to risk. Empirical research indicates, however, that a decision rule of maximization of expected net income typically is used by agricultural producers in production decisionmaking. Consumption values and goals on family farms are discussed in the next section. Multiple Production Goals and Values Under conditions of uncertainty a producer may be unsure of the decision rule to use in his search for production efficiency. Maximi- zation of expected profits may be his overall goal, but only to the extent that it is compatible with his other goals such as sufficient leisure time to spend with his family and with values such as the pride and security of owning a new pair of hogs. Until he has a good idea as to the positive and normative implications of the application of each decision rule for each of his goals and values, the producer may be indecisive and inefficient in his actions. For instance, a rice producer would be reluctant to buy a new herbicide without knowing its probable impacts on production costs and output. On small farms where the family is both a producing and consuming unit, production decisions are influenced by family consumption goals and values. For example, 37 profit maximization in production is constrained by often uncertain family consumption needs such as medical care, church contributions, and vacations. A producer on a family farm typically has to reduce his pro- duction investment level in order to accommodate family consumption goals. As positive and normative information on these factors becomes available, uncertainty is reduced and the producer can make his decisions with more confidence in the probable outcomes. When the production decisionmaking unit is a region or a country made up of many individual producers the selection and application of an efficiency decision rule is an even more complicated process than for an individual producer. Many of the factors which are exogenous to the individual producer are endogenous at the higher decisionmaking levels. Taxes and subsidies, foreign exchange rates, interest rates, and domes- tic commodity prices can all be altered by the national government while, in the absence of monopoly/oligopoly, an individual producer cannot affect these variables. The government must try to satisfy many people, each of whom has his own set of values and goals. Choice of a decision rule which will satisfy one group of people may cause the rest of the people to protest, go on strike, or even to revolt violently against the government. For example, an agricultural minister might be hesitant to endorse a mechanized harvest project without first obtaining information on its probable output, employment, and foreign exchange implications. Yet a government is forced to make many sensitive decisions on a daily basis. This normally is a matter of whether the decisions are impli- cit or explicit, not whether they are to be made. A further uncertainty faced by a regional/national government is that of not being able to measure utility on an interpersonally valid 38 basis. If that were possible the government could simply inform its citizens of the aggregate utility and disutility associated with each decision and, as long as the people accepted the decision rule, the decisions would be accepted. However, in the absence of an interper- sonally valid utility measure, the best a government can hope for is to estimate and publicize the positive and normative tradeoffs related to each decision alternative and then to apply a decision rule accepted by its people. A government cannot sum the individually derived utilities and disutilities associated with, for example, an expansion of rice production areas and then simply apply a decision rule to determine whether the expansion meets the necessary conditions for efficiency (the sufficient condition being the specific expansion plan chosen pro- vided the highest utility possible for the associated disutility or that the lowest possible disutility would be required to obtain a speci- fied utility level). Yet, a government can estimate the expected impacts over time ("simulation") of a decision on relevant performance variables (e.g., rice production, farmer profit, national profit, employ- ment levels, foreign exchange use) and gradually reduce the uncertainties in the positive and normative information base to the point necessary to reach a decision through application of a generally accepted decision rule. Efficient development of the information requires that it be col- lected to the point where the expected marginal cost of each item of infonnation is equal to its expected marginal value in decisionmaking. Expected monetary costs (e.g., human resources, vehicles, data process- ing) may be relatively easy to estimate, but non-monetary costs (opportunity costs) and expected values are more difficult to evaluate. 39 Information developed for a specific decision may be valuable in making other current or future decisions not originally anticipated. The value of each bit of information also will vary among decisionmakers with different degrees of management capabilities. Thus, for example, soils information developed for rice production purposes may turn out to be of little value to a mediocre decisionmaker in the rice industry and yet be very valuable to a more competent rice producer or sugarcane pro- ducer or rancher. Multidisciplinary professional judgement is required to make reasonable estimates of the expected costs and benefits of infor- mation development for analysis of multiple production goals and values. Resource Valuation Once a decision rule has been selected for determining production efficiency, a decisionmaker is faced with the task of determining the corresponding values of the inputs used and the outputs expected to be produced in the production activity. This valuation is required whether the decisionmaker is an individual farmer or a national agricultural minister, and it involves both monetary and non-monetary values. For an individual producer in a market economy the input and output values for traded goods are the prices determined by the supply and demand conditions in each market. For non-traded goods and intermediate pro- ducts the values are determined by estimating the income which those goods could produce in their most profitable alternative use ("oppor- tunity cost"). In the absence of imperfect competition, no producer can affect input or output market prices through his own activities in the markets. He faces perfectly elastic supply curves for inputs and demand curves for his products. 40 With government intervention in market economies, prices may be altered by taxation or subsidization of input or output prices. When taxes are levied on goods and services, income is transferred from those who pay the taxes to the recipients of the tax revenues. Subsidies represent income transfers from taxpayers to those who buy subsidized products. In resource valuation an individual producer does not concern himself with permanent taxes or subsidies per se. Rather, he takes the government-altered prices as given and proceeds to allocate his resources according to his efficiency decision criterion. A government, on the other hand, concerns itself with taxes and subsidies and their impacts on production and consumption relations. Relative prices changes for inputs and outputs have both substitution and output/income effects which alter resource use and product con- sumption patterns. In order for a government to analyze the impacts of its tax and subsidy policies on agricultural production and, hence, on tax revenue generation, it needs information on probable resource allo- cation patterns in the absence of the tariffs. Then a comparison between the estimated "with" and "without" production coefficients would form the basis for a rational decision as to the desirability of the contemplated price alterations. Stgges of Production Given the law of diminishing returns, efficient (in the profit maximizing sense) production will take place at the point where the cost of an additional unit of input is just covered by the income derived from the sale of the resulting increase in output. In economic terms the marginal factor cost (MFC) is equal to the marginal value product 41 (MVP) at the profit maximizing output level on the value product function. The law of diminishing returns indicates further that a typical value product function can be divided into three sections or "stages" defined by the relationship between MVP and AVP (average value product). Stage I encompasses the area between minimum output and the point where MVP = AVP; Stage III lies between the points at which MVP = O and AVP = 0; while Stage 11 covers the area between Stages I and III. Under the static assumptions of profit maximization with perfect knowledge and foresight, and assuming that neither inputs nor outputs are free, pro- duction in Stages I and III is irrational because the producer can increase his profit by either increasing or decreasing, respectively, his input utilization in those stages. Rational production takes place only in Stage II, and profit is maximized in Stage II at the point where MFC = MVP. A number of special cases of the law of diminishing returns which involve the fixing of input levels of substitutable or comple- mentary inputs are possible, many of which result in extreme cases of the production relations discussed here. In reality uncertainty is the norm in agricultural production and rational production in Stages I and III is possible. For example, pro- duction in Stage III for water use would take place if a rice farmer initially flooded hi5 paddies with adequate water for Optimum plant- water relations with normal rainfall, then was inundated by a heavy rainfall of 50 year frequency (i.e., hurricane). Given the multitude of uncertainties affecting yields and prices, it is not realistic to assume that all production takes place in Stage 11. Rather, for analysis of actual production alternatives it is necessary to establish 42 empirically the ranges of utilization of physical inputs and outputs and their associated prices under expected physical (e.g., rainfall) and economic (e.g., income distribution) conditions for the time period of interest. Input Substitution The least cost combination (LCC) of inputs on a continuous function is found where the marginal rates of physical substitution (MRS) among inputs is equal to their price ratios. For each level of input expendi- ture there is a corresponding LCC, and the locus of points formed by LCCs over a range of expenditures is referred to as the line of least cost combinations (LLCC). The profit maximizing point (PMP) on the LLCC is where the ratios of MVP to input cost for each input is equal to one. A change in input prices will cause a shift in the LLCC and, therefore, in the PMP. The input combination would shift toward the relatively cheaper input. Thus, for example, government subsidization of the foreign exchange rate would make imported inputs cheaper relative to dbmestic inputs such as labor and would cause static profit maxi- mizing producers to substitute imports for domestic inputs. With uncertainty, it cannot be asserted that producers operate at the PMP or on the LLCC. Even assuming that the producer's ex ante estimates indicated production at the PMP, changes in physical, biolo— gical, and economic conditions during the production cycle invariably lead to inefficient production off of the PMP (due to the fixing of input use at unexpected levels) when viewed ex post. Thus, assuming a profit maximizing motive under conditions of uncertainty, production typically can be characterized as taking place on a subfunction of the production function. To the extent that production on subfunctions is 43 widespread, the implication is that programs which reduce uncertainty in a producer's search for the PMP will increase production efficiency. Well designed and implemented programs in the areas of research, exten— sion, and education can be expected to help producers to produce more consistently at or near the PMP. Once the actual production function parameters for each production unit under study are estimated and assumptions (projections) are made as to their probable behavior during the relevant time frame, it is possible to estimate the impacts on input use of increases in future production levels. Increased production could be effected through an increase in the productivity of current areas, through an expansion of current areas, or through a combination of the two. Assuming that cur- rent producers are motivated by a desire to maximize profits and that current relative prices of inputs and outputs are constant, geographic areas should not be expected to change their output except in the long term as a result of technological changes or resource deterioration. Increased production in the shorter term would have to come from expan- sion of current areas and, in the absence of suitable unused land (zero opportunity cost), would require the cessation of existing production activities in the proposed expansion areas. The national opportunity (real economic) cost of the new land would be the net income foregone in shifting to the new land use. The opportunity cost principle is equally applicable to other inputs which must be bid away from current activities in order to be incorporated into a new production activity. 44 Rice Adaptability Research International Studies In the early 19705, as part of a national policy of self-sufficiency in rice production, Korea carried out a study of its land base in order to identify areas suitable for producing a new modern variety of rice Shin, 1978). Agronomic requirements for the plant were determined from experiment station field trials which indicated that highly fertile and moderately well drained soils were needed. These requirements were crosstabulated with the technical soils descriptions developed for a reconnaissance level national soils map, and priority areas for rice expansion were designated. The resulting agrophysical adaptability information then was disseminated through extension agents who were given specialized shortcourses in identifying and developing suitable land areas. The author gives no indication that either water availabil- ity or economic factors were studied ex ante, though ex post evaluations indicated that farmers' incomes increased by over 45 percent after adopting the new variety. Suryatna, pt gl., (1979) reported a study of the Indonesian land base in 1974, the purpose of which was to identify areas suitable for future expansion of irrigated and dryland rice production. Agronomic requirements were detennined from experiment station field trials. These requirements were matched with a national soils map, and a land capability classification was established which indicated soils capa- bilities and limitations for rice production in each suitable map unit. The authors give no indication of explicit consideration of economic factors or of utilization of the study results by policymakers. 45 Somasiri, pt 21,, (1979) reviewed Sri Lanka's national land classi- fication and rice adaptability studies carried out under the government's long standing policy of elimination of rice imports. A regional study of land areas suitable for rice production was discussed in detail. The investigation initially used a combination of detailed field-verified soils and climatic data to classify the regional land base, then shifted to more aggregated data when costs of the detailed data proved to be prohibitive. In the modified system, visually identifiable land charac- teristics and aerial photographs were relied on more heavily than was originally planned for classifying the land base. The authors report that the modified riceland classification system was easily understood and used by the local extension service. No rice production economic information was reported to have been incorporated in the study. Winch (1976) studied rice area expansion for rice self-sufficiency in northern Ghana. He placed heavy emphasis on economic factors but dealt superficially with agr0physical factors. The land base was divided into 'upland" and "bottomland", with an additional subdivision based on water availability. Static economic benefit—cost and partial budgeting analyses were conducted on data derived from a rice fann survey. Producer and national ("financial and economic") input and output prices were estimated. The impacts of machinery and fertilizer subsidies on production, employment, and profitability in each of the land strata were analyzed. 0n the basis of the study results, the author recommended that the government eliminate subsidies because of their adverse effects on employment and on regional production effici- ency. 46 Domestic Studies Little domestic work has been done on determining combined agro- physical suitability and economic feasibility of individual plant species in the Dominican Republic. No work of this type has been pub- lished specifically for rice. An FAO (1974a) study of part of the northern Cibao Valley used a land classification system based on climate, water quality, soils characteristics, and benefit-cost analysis to establish priority pro- duction zones for major area crops. Lack of documentation of methods and data sources severely limited the credibility and utilization of the report. National tobacco production was studied recently in the Dominican Republic (INTABACO, 1978). Emphasis was placed on varietal differences in agronomic requirements. Economic factors and geographic distribution of the currently used soils and water resources were discussed only briefly. No consideration was given to identifying new areas with tobacco production potential. Kenah (1980), using CRIES/SIEDRA land base descriptions, studied the agrophysical adaptability of lime trees to the southcentral part of the country. Lime tree agrophysical requirements were matched with the land unit descriptions in order to identify areas with lime production potential. The author compared his results to 1970 farm census data in order to identify areas with lime production potential but with little or no 1970 production reported. He also noted areas which reported high total production but which did not have high productivity accord- ing to his analysis, and he recommended that these areas gradually be shifted over to more productive species. No economic analysis was con- ducted. 47 In 1972 Murphrey qualitatively analyzed the national riceland soils and climatic characteristics. He used those characteristics to identify similar regions in the country and reported several possible new areas for rice production. Economic factors were not considered and no men- tion was made of proposed use of the study results. Development of CRIES/SIEDRA Data Base As mentioned in Chapter I, the CRIES project began in 1976 with the purposes of: (a) improving USDA capability to analyze current and poten- tial food and fiber production in developing countries, and (b) inter- nalizing in developing countries the capability for land use data base development and analysis (CRIES, 1976). Implementation of the project was designed to take place in two phases. Phase I was the one-year US—based development, using secondary information sources, of a first generation consistent national data set, economic analysis model, and geoprocessing program (CRIES, 1977b; 1977c; 1977d). Phase II was carried out in the Dominican Republic for two years (since extended to three years at SEA request). This Resident advisor management phase of the project consisted of the transfer, adaptation, and improvement of the first generation CRIES data set and methods by a local counterpart staff ("SIEDRA"). Phase I preliminary analytical units were the Resource Planning Units (RPUs), originally defined as follows (CRIES, 1977a): An RPU is specifically defined as a unit of land with components sufficiently homogeneous with respect to agro- physical factors of soil, climate, and water resources to be depicted by single estimates of agricultural factors such as crop adaptability and input-output coefficients. 48 Early Phase 11 attempts at field verification of production coef- ficients met with opposition from agriculturalists who criticized the project for referring to the RPU as a homogeneous unit to which single valued input-output coefficients could be assigned, when in the field one could typically note, for example, that both hills and valleys with large differences in agricultural production potential were lumped within a single RPU boundary. The CRIES resident manager, after studying the problem and discuss- ing it with a CRIES soils consultant, decided that the grouping of the agronomically similar dominant phases of soil taxonomic subgroups within each RPU would capture the major differences noted in field observation (USDA, 1976). Given the expectation that SIEDRA would have to rely on non-soils trained field technicians for production coefficient estimates, the major soil subgroups within each RPU were grouped on the basis of visibly identifiable characteristics (SIEDRA, 1978). The new analytical units are referred to as Groupings of Dominant Soil Subgroups (GDSS) and are identified as unmapped percentages of each RPU map unit. Once the analytical units were modified, a program was developed to field verify RPU mapping and GDSS descriptions and to obtain current normal estimates of land use, production techniques, yields, costs, and other coefficients for agricultural land use analysis. Field verifi- cation of the RPU map and GDSS descriptions was begun in June,1978, in the SEA Central Region and resulted in the modification of the original map by the inclusion of two new RPUs. It was decided, based on both US experience and a lack of funds for a farm level SIEDRA land use survey, to interview regional level SEA agricultural specialists in order to obtain judgmental estimates of 49 needed coefficients. In each of the SEA regional offices there are two types of specialists: subregional geographic experts who are responsi- ble for all agricultural production within their areas and individual product experts who are responsible for all geographic areas in which their particular product grows. These technicians were believed by the SIEDRA staff to be the best available source of data in the absence of a farm level land use survey. Questionnaires and explanatory documents were developed for col- lecting crop yield estimates by GDSS and were field tested in July, 1978. Based on this pretest, the original questionnaires and supporting documents were modified and expanded to include limited information on pasture, range, and forest land use. In December, 1978, after brief training on interview techniques, SIEDRA personnel conducted interviews of Central Region SEA personnel and collected preliminary estimates of typical ("current normal") product yields, area, intercropping combi- nations, and production budgets for the major crops (excluding, for political reasons, sugarcane and cultivated pasture) for irrigated and non-irrigated GDSSs (Niehaus, 1978; SIEDRA, 1979c). The Central Region was selected for preliminary interviews because of its proximity to Santo Domingo. This permitted relatively inexpensive followup inter- views to fill data gaps resulting from initial lack of SIEDRA interview and data base development experience. Central Region interview results were checked for completeness and consistency, and followup interviews to complete the estimates were conducted in early January, 1979. Interview results were compared with original CRIES estimates and with the most comparable published SEA estimates in order to determine the reliability of each source of data SO (SIEDRA, 1979d). The SIEDRA regional interview results were judged to be of acceptable accuracy for CRIES/SIEDRA purposes, and interviews in the other six SEA regions were planned for the 1979-80 period. Chapter III's Relation to the Dissertation In this chapter, agronomic and economic theories relevant to the proposed rice production study were reviewed. A sample of related international and Dominican empirical research was then discussed. This information will be used in the following chapter to establish a method- ology for the agroeconomic analysis and in Chapter VI to assist in the interpretation of the analytical results. CHAPTER IV RESEARCH METHODS Concept The overall procedure of this study is, first, to conduct an agrophysical analysis for identification of Central Region land units (GDSSs) with potential for rainfed or irrigated rice production. Second, an economic analysis is carried out in order to identify those GDSSs in which expansion of rice production would be profitable from both the producers' and nation's standpoints under specific production parameter assumptions using alternative expansion decision rules. In the agr0physical analysis, agr0physical tolerance limits of rice for potentially limiting environmental factors are determined and crosstabulated with GDSS characteristics in order to identify those GDSSs with potential for rice production. For each selected GDSS, estimates are made of physical area with potential for rice production, current land use, and area with potential for rice expansion. Possi— bilities for multiple rice production cycles are analyzed. Definitions of two key terms are necessary to an understanding of the economic analysis: monetary costs and benefits, and unsubsidized costs and benefits. Monetary costs and benefits are those production and marketing values based on market prices reflecting producer oppor- tunity costs and shadow prices for inputs and products, though they may 51 52 include governmental price alterations (income transfers) such as taxes or subsidies. They establish the cash expense/income ratios used by maximizing agricultural producers to allocate inputs. Unsubsidized costs and benefits are those production and marketing values calculated by removing implicit (due to governmental price alter- ations) and explicit (direct government payments to producers) subsidies and taxes in valuation of inputs and products from the nation's stand- point. National government price alterations are netted out. Price alterations made by foreign governments on exports (such as milk) to the Dominican Republic may be involved but are not consiered quantitatively. National values provide the basis for estimating real economic costs and benefits to which monetary costs and benefits are compared in order to evaluate the impacts of government price alterations. In the economic analysis, production techniques and yields, pro- ducer and unsubsidized costs of production and income, and producer and unsubsidized returns to land and management (RLM) for rice and its two primary land use competitors (sugarcane and cultivated pasture as an intermediate input in milk production) were estimated for each selected GDSS at the farm gate level. A yield matrix relating rice yields to GDSS characteristics was synthesized to provide reasonable rice yield estimates on GDSSs with production potential but no current production. Monetary input and output prices were converted to unsubsidized prices by removing implicit subsidies. An alternative set of rice budgets requiring increased labor and/or decreased foreign exchange use was established. This benefit-cost information was then employed in analyzing the regional and GDSS-level impacts on rice production, labor and foreign 53 exchange use in production, and monetary and unsubsidized RLM for 11 alternative strategies for a hypothetical expansion of current rice areas, using five alternative expansion decision rules. Sensitivity analysis of the assumptions on multiple production cycles for rice was carried out on two selected strategies. Data Acquisition and Refinement Judgmental estimates by agricultural field specialists played a major role in data acquisition. If farm level empirical data had been available for all production coefficients of interest in the study, quantitative statistical estimates of population parameters and confi- dence intervals could have been made. Use of consensus judgmental estimates (specifically, estimates of population modes) does not permit statistical analysis of data reliability. In effect, these estimates were "experience based", single "educated guesses“ about population modes; hence, mechanical statistical estimates could not be made of their accuracy. However, the judgmental estimates as well as other coefficients used in the study were crosschecked with secondary data and field verified by the SIEDRA staff whenever possible. This procedure was intended to improve data acquisition and processing reliability by giving the staff a down-to-earth feeling for the strengths and weaknesses of the various sources of data by establishing qualitative confidence intervals on the data as time and other resources permitted. It was also intended to increase study credibility among both the government administrators sup- porting the study and the field technicians at the project implementation 54 level by demonstrating that SIEDRA was making a critical effort to obtain realistic data rather than simply using the first "wild guess" that came along. Agrophysical Data In this section the collection and iterative refinement of data relating to rice agrophysical input tolerance limits and corresponding GDSS characteristics are discussed. These agr0physical data provide the basis for the agrophysical analysis leading to the final selection of GDSSs in which rice might be profitably produced. Rice Agrpphysical Tolerance Limits Agrophysical tolerance limits for rice varieties "adaptable“ to the GDSSs of the Central Region were determined both from field interviews of production specialists at the Juma national rice experiment station and from published documents (Juma, 1979; IRRI, 1978; 1979; SEA, 1975). Rice was assumed to be adaptable to a GDSS if its expected yield in that GDSS was greater than the lowest current normal yield reported as commercially marketed by the region's rice producers (SIEDRA, 1979d). Even though rice undoubtedly could be produced with a yield greater than zero in all GDSSs, this assumption allowed research efforts to be con- centrated on those GDSSs with expected yields sufficiently high to indicate a reasonable possibility of profitable production. Thus, no arbitrary assumption of production in Stage II was assumed in establish- ing minimum acceptable yields (Heady, 1952). See Chapter III for a discussion of stages of production. This screening procedure resulted in the preliminary identifica- tion of seven rice varieties as being currently or potentially usable 55 in the region (Table 4-1). Major agrophysical factors identified initi- ally as being potentially limiting to rice production were: water (cycle and salinity), soils (slope, pH, depth, water holding capacity, and salinity), photoperiod, and air temperature extremes. Variety level data for these factors were obtainable only for crop cycle water use. Water input ranges were estimated for five of the seven varieties for which quantitative data were available. Species level requirements were estimated for the remaining eight characteristics. Tolerance limits for water use during the crop cycle for commercialb/ marketed output levels were derived from experiment station estimates of water quantity required to produce maximum (i.e., beginning of Stage III) yields (Juma, 1979). This was accomplished by multiplying these esti- mates for each variety by a constant factor of 0.7 which was obtained by dividing the estimated production cycle "dependable" precipitation (532 mm) in the GDSS of lowest commercial yield by the corresponding experiment station Stage 111 water input level for the same rice variety (758 mm). Dependable precipitation is defined by Hargreaves (1977) as that which occurs with 75 percent probability. Land Resource Base Collection and refinement of data on the land resource base (GDSS) characteristics which correspond to the potentially limiting agrophysical factors identified in the previous section are discussed below. Water: Two water input requirements were used to screen RPUs (and their associated 60585) for appropriate water quantity and quality for rice production: water availability during the crop cycle and water salinity 56 opnmpwm>m yo: mpmn H mmEmcpxm a oniofi Uo mcapmcoaemu cw< mfluflfi xmu\cc vowcmaouocm w-o mEo\H-o;sE xpwcwpmm __om mmcmoo auwomamo # yo: mcszmp mcwupo; cope: Fmom m.o Aeeev e eeeee P_om 2-3 In In :8 mHIO & maopm Pwom Ar wio mso\ -ocEE xuecwpmm cope: E 2: 5: § 5 3:3 zumcmF opoao N8 gm :2 22 E . am 25 Axeevezcg 223 8:8 tee: omen; mopmcH meowcmb omH opomcwz mmcm ocoh mmm23w “mesaq wave: pcmswcwacmm mucmEmcvzcmm psacH mmmfi .cowmmm Focucmu .mmeumwcm> mowm cm>mm cow mpcmsocwacmz usacm _mowmx;aocm< mcwz .Hiq oFQMF 57 (Table 4-2). Water sources considered were rainfall, surface, and under- ground water (INDRHI, 1976, 1978; SIEDRA, 1979b). Rainfall availability was estimated on the basis of dependable monthly rainfall (Hargreaves, 1977). The minimum production cycle water require- ments (138 mm/mo for 4 mo) for the most drouth tolerant rice variety, Juma-57, was used as the criterion for minimum acceptable rainfall. Use of this criterion meant that GDSSs eventually selected as acceptable for ”rice" production might in fact be usable only for the most drouth re- sistant variety. This factor is considered in Chapter VI. Surface water availability was deemed acceptable if surface flows from rivers or dams were at least 3 m3/sec in GDSSs with areas large enough for the construction of secondary irrigation canals of at least 4 km in length. Lower surface flows and smaller irrigable areas were assumed, on the basis of local experience, to be economically unviable for public investment (INDRHI, 1976). Underground water acceptability was determined on the basis of water table depth, chloride content and yield. Specific criteria were: depth to water table less than 80 m; chloride content less than 100 ppm; and yield of at least 600 gpm. Water with characteristics beyond these limits was assumed. on the basis of Dominican Water Resources Institute (INDHRI, 1979) studies, to be uneconomical for use on public irrigation projects. Water loss through evapotranSpiration was not considered quantita- tively because of constraints on data availability. However, this factor was given qualitative consideration in the estimation of potential physical production area in GDSSs with potential for rice production, as explained later in this chapter (Reyna and Paulet, 1979). 58 Hw 1 1 pooimz a: .8: 1 1 3 e801zez - ueo-»ez oeeez - 1 mm 1 1 1 1 1 1 mm 1 1 1 1 1 1 mm 3013: 1 3013: 03 E 1 1 3 >02 12:2 1 p818: a: .5: >oZ1m=< 3:723 cm 1 1 1 1 1 1 2 8.31.32 1 301me 253 1 1 3 881:2. oe com 818 ”Beta: Ego 1 1 S 9&1me 1 1 1 @3132 1 mo omoéfl. 03 co: m: 3013: 0383 .253 @248: 1 B emcee: 1 301?: omfiz 1 1 no 3013: 1 301.3: 2.8: .268 96.1%: 1 NO 39.82 1 301.3: 082:; .253 1 1 S :5 .203 325 .83 253 $5 e... fl... ._ _Fete>o gee: meomtee new; meoeeee sea Loam: ucaocmcmnc: cope: momwcam Fme:_em memfl .eoemee Peeeeeo .maee mH e: carnageOLa mow: com xpWFwnmuamc< new mmoczom awash scam auwpwanme>< comm: .N1e mpnep 59 Soils: Soils data were taken directly from SIEDRA (1979a). Only the GDSSs selected as having acceptable water availability were further screened for tolerable slope, pH, waterholding capacity, salinity and depth to bedrock or water table (Table 4-3). Specific tolerance ranges were: sl0pes 0-3 percent for irrigation and 0-15 percent for rainfed produc- tion; pH 4.5-7.5; texture (as a proxy for waterholding capacity) non- coarse; salinity 0-7 mmoh-l/cm3; and depth to bedrock or water table at least 0.5 m. Slopes over three percent had previously required unecon- omical investments for leveling or terracing for irrigation. Slopes over 15 percent for rainfed production were assumed to pose an unaccept- able erosion hazard under economically feasible conservation practices (Paulet, 1979). Photoperiod and Temperature: Photoperiod was eliminated as a potentially limiting factor for the seven rice varieties studied, as all seven had proved in field trials to be adapted to the 18 degrees south latitude of the Central Region. This factor is considered again in Chapter VI, however, in a discussion of the possible introduction of new rice varieties. Temperature extremes were handled in the same manner. Use of Agrpphysical Data The refined agrophysical data obtained through the procedure des- cribed above is used in the agrophysical analysis to make final selection of GDSSs with characteristics within the tolerance limits for rice production and to estimate the physical areas potentially usable for rice area expansion. The agrophysical analysis is discussed later in this chapter. 6O o ovum ococpm move .ooe N1m.o om1o~ ammo.ooe m o ovum ococom mow» .ooe m.O1H.o om1mH 3oFFogm m o .Poopo .ooE move .ooe m.O1H.o mfi.n mPPF: «mo 0 ovum .m >co> mow» m1m m1m :woco PFoz m o owoo .m >cm> on?» m1m m1o cFoco cooo m O atom oeoeem mete .eoe m.o-H.o mHA. mp_ee <~o o .Foo_o .Pm mow: m.01~.o mH1m mPF_: m o _ecoeee ee_c m-fl m1o ee_c go one; reason acoocoo om uxou cw oooPo—oxm mo .couoz o» unsound; monotonom no owns: a 5 one; oncoguxo vo~vamnamco oo.Hmo¢ n oo.me:m umoo xcouoooe pope» to acoocoo on m co" u mum“ .osa n xooc_ woven msmfio xooop opow» ocoucooom an ooumowoo mumoum oEoNO n o .ocvo: n : .xocouou n um muooooo omega each ooumonoo mumoo ooquoc oPop>~ om.m~ om.oo mm.oo mo.¢o m~.oo mo.oe ou.mm H~.~m Aou\uom msmflv umoo oo~Powmoomoo Pouch ~m.m on.“ om.n co.“ om.“ om.o N~.o wo.m mumoo poocouov oo~wowmnomco om.mo oo.mm oo.mm Ho.~m om.Ho oo.oo oo.om m~.oo pouch mm.o mm.o mo.o mo.o o o o o “Aop\momv umoo coon: cowuoowccw oo~+owmnomc= Hm.oo ~o.om so.om wo.~m om.~o mo.oo oo.om mH.oo Pouch 3.2 meow 8.8 8.2 3.8 8.3 2.3 no.2 @339: 32:; oofooep co 38 33333:: mo.~m ~o.mm No.mm w~.~m w~.mm m~.o~ No.om om.m~ Aou\»omv umou page? ovumosoo om.m~ -.- HN.- mm.oH -.ofl o~.- mm.m~ o~.oH mfiou\»m=v umoo uooom ooucoQEH ”mumou oo~wowmoomco m~.om mo.om mo.om m~.mo oo.mm om.nm mm.oo o~.mo oaou\mom oNoHv pmoo acouocoe pouch mm.mm m~.m~ m~.m~ mo.w~ -.om mo.- om.m~ fio.m~ Aou\»ma muomv umou pouch ~m.~ mm.~ m~.~ Nw.~ mm.m mm.m mm.~ ~w.~ :o_uomoocooo om.om om.o~ oo.- ow.m~ om.w~ o~.m~ os.o~ mm.- Aou\»o¢ whoa: u> Pouch o.m m.m H.m ~.m ~.m H.~ ~.m m.~ Aou\uev opo.> E E. E . E E E. E E 3:59. So: o> :38 mo.m o~.m mo.m oo.m m~.m m~.~ um.~ m~.~ umo>co: Ho.o Ho.m -.m Ho.o m~.m m~.m Ho.o Ho.o cop o>_upou umoo opoowco> 11111111111111111111111oooe_umo pooPorecooU1111111111111111111111 "mumou xuouoco: o No N: o No 9 o o Huomoom _pwz omom moo «me to”: «so mo~ oouommccm oomcwom memo a» oaxeoea memo .ea_ma¢ _ateeeo .mmmoo cabeotttm teat oea esteem“ Lace ea .ocoocooom Loy notch coo :owuooooco mo mumou ooNPoPmoomoo one xuooooo: poooc< co cowoo>wcoo .mH1o o—nop 91 uxmu c? umcwapnxo ma .quoz 0» uncommog acougmmam :0 women was; wucmzuxm um~wu+mnamcz ow.~noa n oo.~»m=o mung poaccm pcmugmq me s umou xgmumcoe Papa» $0 acmugma emm cog u mum” .omH xoucw ou_ga anode Possum pcmugma m mung ummgmucmm _u=cco acougma m mung ummgauc_ mxovcw upmw» xpws x5 umumanua mpmou~ ammuan mmon «Pg» sou» umumanua mumou umumpmg upmm>m A~H\»om mhmflv no.5m ~o.¢m no.5o No.em mm.~o -.mo “H.mm mm.vm mH.me mo.¢m nm.~m umou umNVuwmnamca Pouch om.m Hm.m om.m Hm.m mo.m m~.m w~.m sm.~ m¢.m mm.~ mm.“ mflmu\»omv «moo ummgmucw u ~wuwmn=mc= 2.3.. 3.3.. kgm :ém 8.3 3.8 8.3. 8.3 3.8 8.3 3.3 p33 ---------------------m~.e -n-------------------- o o o o Amu\»omv umou . Lmumz cowummwmgw umNVqunamca 3.3 E 3.3 modw 3.3 m . e mmdm E 36m 8.: whee p33 m~.o~ mo.mH o~.o~ mo.mH mm.mH mm.mH ~m.m~ mm.~H m~.qa mo.- HH.NH ofimu\»omv manna? vmugoaep yo umou um~wuwmgzmc= Eda 3.8 2.2 Edm 86m 3.8 3.8 mmdm wmam mmdm 3.3 €335-38 2.9: “93358 mm.oH am.mH mm.oH am.mH e¢.mH mm.mH .oo.mH mm.¢H om.~H o~.¢H o~.¢H mfiou\mm=v umou Haas? nonsensm "mumou ou~_cwmn:mc= oo.~¢ om.oe mo.m¢ om.m¢ me.m¢ -.oe NH.cv Hm.me m~.mm m~.me mm.Ho ¢?a\»om mumfivamou xgmuocoe Page» owém 86m 8.? 86m 3.3 35m 91mm 26m 3.8 San m~.~m A322 m3: :8 P33 lllllll ll'llllllllllllllllllllllllllllllllloooo llllllllllllllllllllllllllI ow.mm oo.nm amymm oo.om um.wm wv.mN we.mw unnumm mmwnmm khmumm “mnuwm um co ummgmucr .copuawumggma m Amp\*oa msmflv u> PmHOF em.H~ mn.m~ em.H~ mn.mH H~.m~ N~.mH N~.mH ~m.~H Rm.HH Ho.~H oo.oH “magmpcw .Loaop .vmmu ----------u--------------------------------eN.oH-n-----------------------u- mucmcmucmms mgaummq .ucmsawacu «moo mpnmPLm> "mumou xuouocoz mNH ofia am” mHH ofifi mfifi mHH moH mm be” em xmccm upmw> xpwz .mflm ¢~mfi on xou=~ .mumfi .cowmam _~Lu=mu .mmou vaamwgg~ =u>um can umwcvua Lao“ c, 3,.z\ogaumam umum>vppau so» mung» gun copuusuoga we mumou vmnpuvmnamcz ucm xgmumcoz Fm=c=< mo copua>wgmo .m~-¢ m—nm» 92 Irrigation Infrastructure Investment: No irrigation infrastructure costs were estimated for the rice area expansion. This was based on the assumptions that there were at least 25,000 ta of irrigated sugarcane and cultivated pasture which could be converted to rice production in each selected GDSS and that sugarcane and cultivated pasture irrigation infra— structure was a close substitute for rice infrastructure. These assump- tions are discussed in Chapter VI. Unsubsidized Income: Unsubsidized income was calculated by multiply- ing yields by the unsubsidized output prices. The unsubsidized output price for each product at the farm gate was derived from the CIF import price, after that price was adjusted for the foreign exchange differen- tial by netting out marketing charges and converting from processed to unprocessed product (Tables 4-14, 4-15, 4-17, and 4-20). Table 4-20. Derivation of Unsubsidized Farm Gate Prices from CIF Import Prices for White Rice and Reconstituted Milk, Central Region, 1979 Current Normal Parameters Rice Milk (per qq) (per lt) CIF import price (US$) 21.90 0.10 Foreign exchange conversion rate, unsubsidized x 1.2 x 1.2 Unsubsidized CIF import price (RD$) 26.28 0.12 Marketing charge - 6.10 - 0.12 20.18 0 Conversion factor, processed to unprocessed x 0.65 x 1 Unsubsidized farm gate price (RD$) 13.12 .0 1US$1.00 = RD$1.20 93 Current normal import prices for rice were estimated from INESPRE (1979) data. Refined sugar CIF import prices were taken from World Bank (1979) estimates of long tenn expected world prices. In the absence of an import market for pasteurized milk, the unsubsidized cost of milk was estimated on the basis of the CIF import price of reconstituted powdered milk (Associacion, 1979). The assumption was made that recon- stituted milk was a close substitute for pasteurized milk and is dis- cussed in Chapter VI. Unsubsidized marketing charges for rice, sugarcane, and cultivated pasture/milk were derived from monetary cost estimates (Tables 4-16, 4-17, and 4-21). Regional average monetary costs per qq of white rice were calculated from INESPRE (1979) data for five major mills in the Central Region (Table 4-21). Milled white rice with 14 percent moisture content was calculated at 65 percent of rough rice wieght with 20 percent humidity (Murphrey, 1972; SEA, 1977a). Regional average rice transpor- tation costs per qq to the local mill and from there to the Santo Domingo retail markets were calculated from INESPRE (1979) data. Rice wholesal- ing and retailing charges were estimated at 16 percent of current normal retail price as based on a 1977 SEA marketing report. Monetary marketing charges for sugarcane and livestock were dis- cussed in the data collection section of this chapter (Tables 4-16 and 4-17). A monetary charge for milk marketing was based on SEA (1977a) marketing margin estimates for 1975. These costs were calculated at 36 percent of the 1979 retail pasteurized milk price. Unsubsidized charges for marketing the three products were calcu- lated by multiplying estimated monetary costs for imported inputs by the unsubsidized exchange rate of R0$1.20 = US$1.00, and adding an interest 94 Table 4-21. Derivation of Unsubsidized Marketing Costs from Monetary Costs for Rice, Central Region, 1979 Costs Monetary costs: ($/qq) Processing - Machinery 1.41 Labor 0.35 Total processing 1.76 Transportation Farm to mill 0.45 Mill to Santo Domingo 0.30 Total transportation 0.75 Total processing and transportation 2.51 Sales margin1 3.24 Total monetary marketing cost 5.75 Unsubsidized costs: Imported input charge (30%) 1.72 Foreign exchange adjustment (1.2) 2.07 Unsubsidized marketing cost 6.10 116% of 1979 retail price of $20.25/qq charge of 16 and 18 percent, respectively, for sugarcane and the other two products. These adjusted costs were added to domestic input costs to obtain total unsubsidized marketing costs. Rice imported input costs were calculated to be 30 percent of total marketing costs. Based on this estimate, unadjusted imported input costs for sugarcane and milk/cull animals were calculated at 30 percent of total marketing costs. Unsubsidized labor costs were estimated at the local unskilled wage labor rate, as explained earlier. 95 Unsubsidized Returns to Land and Management: Unsubsidized RLM was calculated by subtracting unsubsidized costs of production from unsubsi- dized income. Comparisons of Production Coefficient Estimates: Comparisons were made among the three land uses of monetary and unsubsidized production costs, monetary and unsubsidized (RLM), and pro- duction labor and foreign exchange use on a per tarea (ta) basis. Rice production cost subsidies were calculated, and privately-owned and state- owned machinery costs were compared. The purpose of the comparisons was to characterize production activities in each GDSS as a prelude to the partial budgeting analysis discussed below. Partial Budgeting Analysis of Rice Area Expansion The economic data developed in this study was used to analyze the regional and GDSS level impacts on production, labor use, foreign ex- change use, and producer monetary and unsubsidized RLM of 11 alternative strategies for a hypothetical 25,000 ta expansion of current rice area (Table 4-22). The strategies were identified from among the various possible combinations of decision rules (maximization of rice production, monetary and unsubsidized RLM at the farm gate, and production labor, and minimization of foreign exchange use) and policy variables (single and multiple production cycles for rice, current normal and alternative production techniques for rice, and increased production on current GDSSs, expansion in current 60355, and expansion in potential 60535). The first three strategies (A-c) involved maximization of rice pro- duction under various assumptions on the policy variables. The next four strategies called for maximization of rice monetary RLM (D and E) and rice 96 Table 4-22. Assumptions on Five Decision Variables and Three Policy Variables for 11 Strategies for Exapnsion of Rice Production Area, Central Region Strategy Variable CN A B C 0 E F G H I J K Decision Rule Production max a x x x Rtn L&M, Monetary a x x Unsubsi. x x Labor max a x x For Exchange min a x x Policy variable Pdn Technique Typical a x x x x x Alternative x Tb T T T Production Cycles A A A A Current Normal a x N x N x N x N x Potent multiple x Y Y Y Y GDSS Curr Normal area a x x x x x x Expand curr norm Potential unsubsidized RLM (F and G). The final four strategies required maximiza- tion of labor use in rice production (H and I), and maximization of forefiyi exchange use in rice production (J and K). Partial budgeting of the net impacts of each strategy (production, RLM, labor, foreign exchange use) were carried out at the GDSS and region- al levels (Table 4-23). Added costs and reduced benefits associated with each rice expansion alternative were subtracted from the added benefits and reduced costs which were expected to occur. In each case, rice was assumed to substitute in each GDSS for the least profitable of the sugarcane and cultivated pasture/milk land use alternatives. The strate— gies then were ranked on the basis of net regional impact on each of the studied variables. 97 coo Hmm mPDmpwm>m Ho: annum mu\ca~ ca coo He no 8o;H mmmm ommm mme mm gouge? Pacowmwg pmz owed- cm 0 o mpvmmcmn xpwe\au Lo mcmugmmam cmozumm mHmH ommm mme mm muwmmcmn our; umcn< ... m mmoq coca LLH mzsm wuwm mm_o»u Hmuwm made mauve mm_osu maaxm am: aces \cwmm mmgw (~29 Eogm m cmng :o_uzpom mgmumsmgma Fmsgoz pcmcgau mumfl .cowmmm ngucmu .cowmcmaxm mowm Low mcowpmpsupmu mcwummuzm meucma mpqsmm .m~-¢ mFQQH 98 Table 4-23 (Continued) Strategy E assumed maximization of rice monetary RLM with free choice among other production parameters. The computation procedure for deter- mining the net regional impact of the rice expansion was as follows: 1. select the GDSS and rice production technique with the highest annual monetary RLM per ta (i.e., irrigated GDSS 07A using alternative pro- duction techniques with 3 cycles/yer x $36/cycle = $108/yr). . if there are less than 25,000 ta of expansion area available in the GDSS selected in (1), select the GDSS and rice production technique with the second highest RLM (i.e., irrigated GDSS 06A using alterna- tive techniques with 3 cycles/yr x $28/cycle = $84/yr). . calculate added rice production benefits: a. for the GDSS and technique selected in (1), multiply number of possible rice production cycles by the expansion area available up to 25,000 ta total, then multiply the result by the parameter values for each of the impacts (rice yield, monetary and unsub- sidized RLM in the example) to obtain the added rice benefits in the GDSS (e.g., 3 cycles x 20,000 ta x $36/ta monetary x RLM = $2,160,000 monetary RLM). b. repeat (3a) for the GDSS and technique selected in (2) up to 25,000 ta combined total (e.g., 3 cycles x 5,000 ta x $28/ta monetary RLM = $420,000 monetary RLM). c. for each impact, sum across 60585 to obtain the regional added rice benefits (e.g., $420,000 + $2,160,000 = $2,580,000 monetary RLM). . calculate reduced sugarcane or CP/milk benefits: a. for the GDSS selected in (1), determine whether sugarcane or CP/milk has the lowest monetary RLM, then multiply the expan- sion area by the corresponding impact parameter to obtain the reduced benefits (e.g., 20,000 ta x $2/ta monetary RLM (cp/ milk) = $40,000 monetary RLM). b. repeat (4a) for the GDSS and technique selected in (2) (e.g., 5,000 ta x $2/ta monetary RLM (cp/milk) = $10,000 monetary RLM). c. for each impact, sum across 60535 to obtain the regional re- duced sugarcane and CP/milk benefits (e.g., $40,000 + $10,000 = $50,000 monetary RLM). . calculate regional net impact of the rice expansion: subtract the reduced sugarcane and CP/milk benefits in (4c) from the added rice benefits in (3c) to obtain the net regional impact (e.g., $2,580,000 — $50,000 = $2,530,000 monetary RLM). 99 A sensitivity analysis of the assumptions on potential for multiple cycle rice production was conducted on two of the "best" overall strate- gies for rice expansion. The best overall strategies were selected on the basis of most consistent high ranking among the five regional impacts monitored. GDSSs with possibilities for three rice production cycles were restricted to two cycles in the sensitivity analysis. The purpose of the analysis was to determine the effect on study restuls of a possi- ble field determination that a maximum of only two cycles were possible in the Central Region. CHAPTER V ANALYTICAL RESULTS In this chapter the results of the agrophysical and economic analyses are presented. Agrophysical results include the selection of 00555 with rice production potential, major limiting factors to rice production, and, within each selected GDSS, potential physical area for rice produc- tion, rice multiple production cycle potential, current land use, and potential area for rice expansion. Results of the economic analyses include both per tarea benefit-cost and impact accounting estimates for each GDSS selected in the agrophysi- cal analysis and partial budgeting estimates of the regional and GDSS level impacts of alternative strategies for rice production expansion. The partial budgets monitored, for 11 strategies for a hypothetical 25,000 ta expansion of rice production, regional impacts on rice produc- tion, net (combined rice, sugarcane, and cultivated pasture/milk) mone- tary returns to land and management (RLM), net unsubsidized RLM, net labor use in production, and net foreign exchange use in production. Completing the chapter are results of the sensitivity analysis of two selected strategies as they are applied in testing the assumptions related to potential for rice multiple production cycles. Agrophysical Results The agrophysical results begin with the selection of 00555 with rice 100 101 production potential and a discussion of the major limiting factors pre- venting rice production in many of the 60555 in the region. Then follow the results of the analysis of each of the selected 00855 for determina- tion of physical area with rice potential, potential for multiple rice production cycles, current land use, and potential area for rice produc- tion expansion. 60555 with Rice Production Potential Crosstabulation of the rice agronomic requirements with the water and soil characteristics of each GDSS, considered with additional rele- vant information, resulted in the identification of four 00535 as having rainfed rice production potential and six 00555 and two inclusions of two 00555 as having irrigated potential (Table 5-1). The four rainfed 00535 were 07A, 07B, 20A and 20B, and the six 60335 with irrigated rice poten- tial were 06A, 07A, 16A, 16B, 21A and 21B. In addition, a swampy area of rice production reported in 07A was field checked and determined to be a single area of atypical soils ("inclusion"), which is referred in the rest of this paper as GDSS "07As". A second inclusion occurs in GDSS 113. Soil descriptions in the SIEDRA (1979a) land resource base docu- ment indicate that GDSS 118 has slopes over 3 percent and, therefore, according to the 3 percent maximum slope criterion used in this study, has no irrigation potential. However, SIEDRA (1979c) interview results indicated a current normal area of 10,000 ta of irrigated rice in 118. A field check of this apparent paradox revealed that the production in this GDSS takes place in a unique inclusion of naturally subirrigated soils. This area henceforth will be referred to as GDSS "118i". Initially 00555 01A and 028 were identified as having rice produc- tion potential. Field checks indicated that the areas with rice potential 102 Table 5-1. Agrophysical Adaptability of Rice to Nine Water and Soil Characteristics in 29 60555, and Major Limiting Factors, Central Region 41A Nater Soils M939r - lelt- GDSS Rain- Sur- Under- Slope Tex- Sali- FQEEI. fall face ground Rai n- Irri - Depth ture PH ni ty fed gated R I ---Adaptability2;-- ........... Adaptabilityz ............ 01A n A n A A A A A A w d B n A n A n A A A A w 5 02A A A n n n A A Q A s s B A A n A n A A Q A d d 06A n A n A A A A A A w A B n A n A n A A A A w 5 07A A A A A A A A Q A A A B A A n A n A A Q A A 5 09A A n n n n A A A A s s B A n n A n A A n A p 5 11A n A n A n A A A A w s B n A A A n A A A A w 5 16A n A n A A A A A A w A B n A n A A A A A A w A 19A n n n A n A A A A w w B n n n A n A A A A w w 20A A A n A n A A Q A A s B A A n A A A A A A A L 21A n A n A A A A Q A w A B n A n A A A A A A w A 22A n n n A A A A A A w w B n n n A n A A A A w w 23A n n n A n A A A A w w 25A n, A n A A A A n A w p B n A n A n A A n A w 5 40A n A n A n A A Q A w s B n A n A n A A Q A w s n n n n n A A A A w w n n n A n A A A A w w R = rainfed A = adaptable I = irrigated Q = questionable d = dispersion and size n = not adaptable s = slope w = water p=PH L = legal A = adaptable 103 within these 60555 are very small and dispersed and thus are not appro- priate for consideration for a major rice expansion program. For this reason they were dropped from further quantitative analysis. GDSS 203 also was identified in the preliminary analysis as having irrigated rice potential. However, in the gathering of additional infor- mation it was learned that the available surface water in 208 is limited by a 40 year old law to industrial use only. This GDSS was dropped from further consideration for irrigated rice production under the assumption that the law is not likely to be changed in the foreseeable future. Limiting Factors The most common limiting factor for rainfed production was water, which eliminated from consideration 21 of the 29 (69 percent) of the GDSSs in the Central Region (last column of Table 5-1). Slope eliminated two rainfed GDSSs and size (dispersion) and pH eliminated one each. The major limiting factors for irrigated production were slope, which caused 41 percent of the regions 60555 to be withdrawn from consi- deration, and water, which precluded 28 percent from further analysis. Size eliminated two GDSS, and pH and legal factors eliminated one each. Potential Physical Area for Rice Production Total physical land area in the Central Region with rice production potential was estimated at 314,000 ta rainfed and 106,000 to irrigated (Table 5-2). Potential rainfed land area by GDSS ranged from 34,000 ta in GDSS 208 to 119,000 ta in 07A. Potential irrigated land area ranged from a minimum of 9,000 ta in three 60585 to a maximum of 21,000 ta in 21A. 104 NNN NO ONO HOHOH .mm: .mm .mpw O O H H O NHN ON H H H HN O No: ONOO H OO O n¢ mm Om O Nm mm mHN OO O «O Hm NO O ON Om NNHOO mcmugmmzm mowm OmpO>NuPOO mcmugmmam muwa mmOO OONNONNOONOO xgmumcoz Ocm FOUNONH mowm Low cowuosvogm No mpmou OONNONNOONOO Ocm Ngmpmcoz HOELoz pcmggsu ONOH .833 2228 ..NNNOO NH E v:._.z\m..5ummm umpm> .5 :5 can wcmugmmam .mmacwcnumh cowuuzuoga m>3mcgmp~< .v-m mpnmb 111 Unsubsidized Costs of Production The unsubsidized production costs are monetary costs adjusted for implicit subsidies. The unsubsidized cost of estimates for rice using typical production techniques ranged from $26 to $65 per cycle per ta, a figure which averaged about 20 to 40 percent higher than the correspond- ing monetary costs (Table 5-4). In all cases except in rainfed GDSS 07A the unsubsidized production costs were higher for the typical rice pro- duction techniques than for the alternative techniques. This was due primarily to the use of fewer imported inputs in the alternative set of production techniques. Sugarcane unsubsidized costs of production ranged from $45 to $74 per ta per yr. Cultivated pasture/milk unsubsidized costs of production were estimated at $45 to $68 per ta per yr. In all GDSSs the cultivated pasture/milk unsubsidized costs were slightly ($0 to $15) lower than those for sugarcane. Rice Production Cost Subsidy The difference between the unsubsidized costs and the monetary costs of rice production are subsidies. They are largely income transfers from taxpayers and consumers to rice producers. Subsidies for all rice production ranged from $7 to $21 per cycle per tarea (Table 5-5). Rainfed subsidies averaged about $8 per ta less than irrigated subsidies for typical rice production techniques and about $6 less for the alternative techniques. There was little difference in subsidy levels between production techniques for rainfed GDSSs, but the alternative technique subsidies were slightly lower than the subsidies for typical techniques in irrigated GDSSs. Estimates of subsidies for machinery are discussed later in the chapter. 112 Table 5-5. Current Normal Rice Production Cost Subsidy for Typical and Alternative Production Techniques, Production Cycles, Area, and Total Subsidy for 12 GDSSs, Central Region, 1979 cuss sum dy de? Khys SuTbostiad] Typical Alternative yc es rea (typ teéfi) ----$/ta/c_yc'le ------ (110.) (1,000 ta) ($1,000) Rainfed: 07A 11 12 1 4 44 07B 0 0 0 20A 7 5 0 0 0 203 0 0 Irrigated: 06A 21 16 3 5 315 0 0 07As 18 17 1 14 252 113i 29 14 2 10 580 16A 18 14 2 8 288 163 15 16 0 O 0 21A 16 15 1 1 16 213 18 16 0 0 0 Regional Total 1,495 Monetarngeturns at the Farm Gate Monetary returns represent the difference between monetary costs and monetary income. Rice monetary returns to land and management (RLM) at the farm gate were estimated at $4 to $38 per cycle per ta using typical production techniques (Table 5-6). Rainfed returns ranged from $10 to $35/ta and irrigated returns from $4 to $38/ta. The highest rainfed returns were in GDSSs 20A and 203 and the highest rainfed returns were in 07A and 07As. In half of the rainfed 00555 and in all but two of the 113 mHOOHHO>O Ho: OHOO H NO- H- m- O O OH O OHN mm. m O O O N O OH ON : m m>_. 3 v u .3: 0 $05 U u. .2. 0 mo; mmow OONHOHNOONOO Ngmumcoz ONOH .OOHOOO HOLHOOO .mOOOO NH OH HHHZNOLOHmOO OOHO>HHHOO Ocm OOOOLOOOO .mozcwccomN cowuuauogm m>HpmchpH< OOO HOOHONN OOHO Hcmsmmmcmz OOO OOOH o» mcgaumm OONHOHNOONOO Hmsgoz HOOLLOO .O-O OHOON 114 irrigated 30585 the alternative production techniques provided returns at least as high (difference of $0 to $15) as those for the typical techniques. Sugarcane monetary returns ranged from $27 to $37 per ta per yr. Cultivated pasture/milk monetary RLM varied between (-) $6 and $4 per ta per yr. The negative returns in some of the GDSSs are consistent with results reported in a recent dairy association report (Associasion, 1979). In all GDSSs both single cycle rice and sugarcane provided higher mone- tary returns than did cultivated pasture/milk. Unsubsidized Returns at the Farm Gate Unsubsidized returns are the difference between unsubsidized costs and unsubsidized income. Rice unsubsidized RLM at the farm gate were all lower than monetary RLM and were estimated at (-) $13 to $33 per cycle per ta using typical production techniques (Table 5-6). Rainfed unsubsidized returns varied from $3 to $33/ta and irrigated returns from (-) $13 to $33/ta. The alternative techniques provided returns at least as high as those for the typical techniques in half of the rainfed 60555 and in all but one of the rainfed GDSSs. Sugarcane unsubsidized returns varied from $25 to $46 per ta per yr. Unsubsidized returns to cultivated pasture/milk were all negative and varied from (-) $62 to (-) $40 per ta per yr. The negative returns to milk production result from the assumption that reconstituted imported powdered milk is an acceptable substitute for fresh milk among Dominicans. That assumption results in an opportunity price of $0 at the farm level, which in turn causes all production costs to become negative returns. Sugarcane provided the highest and cultivated pasture/milk the lowest 115 annual unsubsidized returns among the three land uses in all GDSSs, with the exception that rice provided the highest return in rainfed GDSS 20A. Labor Use in Production Using current normal rice production techniques, labor use estimates varied from 18 to 41 hr per cycle per ta (Table 5-7). Labor use in rain- fed production averaged slightly over half of that for irrigated pro— duction. The highest labor use in rainfed GDSSs was in 073 and in irrigated GDSSs was in 06A. In all GDSSs except irrigated 06A the alter- native techniques were slightly more labor intensive than the typical techniques. Sugarcane labor use varied from 26 to 31 hr per ta per yr. Labor use in rainfed production averaged about 10 percent higher than that for irrigated production. Labor use in cultivated pasture/milk production ranged from 5 to 7 hr per ta per yr, and was slightly higher for irri- gated production than for rainfed production. Cultivated pasture/milk used only about 20 percent as much annual labor per ta as did sugarcane. Foreign Exchange Use in Production Estimates of foreign exchange use in rice production with current normal production techniques ranged from $5 to $16 per ta (Table 5-7). Rainfed estimates varied from $6 to $10/ta, and irrigated estimates varied from $5 to $16/ta. Foreign exchange use in rice productiontnfilizing the alternative production techniques ranged from $3 to $8 per cycle per ta. With the exception of rainfed GDSS 07A, the alternative production techniques used approximately half as much foreign exchange as did the typical tech- niques. No foreign exchange is produced through rice exportation. OHOOHHO>O Ho: OOOO 116 O O O OH N O OO HO OHN O O O O O O OO HO OpHOO OOOOLOOOO moOO OmpO>HHHOO OOOOLOOOO muwm OOOO OONHOHOOOOOO Ogmumcoz ONOH .OOOOOO HOLHOOO .OOOO NH O_ OHHZNOLOHOOO OOHO>OHHOO OOO OOOULOOOO .OwOOHOOOOO OOHHOOOOLO O>OHOOLOHH< Ocm HOOOONH moOO go» ON: OOOOOuxO OOmeom OOO LOOOH Hmsgoz acmggau .N-O OHOON 117 Sugarcane foreign exchange use was negative in all GDSSs and ranged from (-) $52 to (-) $102 per ta per yr. Negative values resulted from the assumption that all sugarcane would be exported as sugar and from the fact that sugarcane income exceeds costs in all GDSSs. Cultivated pasture/milk foreign exchange use varied from $3 to $9 per ta per yr. The cultivated pasture/milk land use produces a small amount of foreign exchange through meat exportation. No milk is exported. Thus, sugarcane is by far the lowest user (i.e., highest producer) of foreign exchange among the three land uses. Partial Budgeting Analysis This section compares the current normal (base) annual production coefficients with the results of the partial budgeting of 11 alternative strategies for a hypothetical 25,000 ta expansion of the current rice pro- duction area in the Central Region (see Chapter IV for explanation of partial budgeting methods). Results are presented in Table 5-8 accord- ing to alternative expansion decision rule: maximization of regional rice production (Strategies A, B, C); maximization of regional monetary returns to land and management (RLM) for rice (Strategies 0, E); maxi- mization of unsubsidized RLM for rice (Strategies F, G); maximization of regional labor use in rice production (Strategies H, I); and minimization of regional foreign exchange use in rice production (Strategies J, K). Each strategy had, in addition to a specific decision rule, restrictions on rice production techniques, numbers of production cycles, and GDSS availability as explained in Chapter IV. Regional impacts (changes from current normal) are estimated for each strategy and compared to the current normal coefficients. 118 :OOHOUHOHOos OHNOHOOO OHH>HHHOOOO H O.O O.O O.O N.O N.O N.O N.O N.O N.O N.O O.O HOOOHHOO OO mm: .Ouxm .LOO O.O O.O N.H O.N H.N O.H O.N O.N H.N O.N N.N HO; OOHHHHOO LOOOH O.H N.H O.N O.N N.N O.N N.N H.N O.H O.N O.O OONHOHNOONOO N.O O.O H.H O.H O.N O.H N.N O.N N.H O.H N.H NOOOOOOs HOOOHHOO OO zOO N.O N.O O.N N.O H.O O.N O.O O.O O.N O.N O.N AOO OOO.OOHO OOHHOOOOLO muHO x O HH H O HO O O OOOO O NOOHOOHO zu HOOOEH HOOOOZ OOOLOOO ONOH .OOHOOO HOOOOOO .NOHOOHOOON OOHOOOOO ozN OOO NOHOOHOLHO OOHOOOOxO comm HH OO OpuOOEH HOOOHOOO O>OO .O-O mHOON 119 In each section, overall regional impacts are presented, followed by impacts at the GDSS level (Table 5-9). Use of the alternative produc- tion techniques is discussed (Table 5-10). The section closes with the results of the yield sensitivity analysis of rice multiple production cycle assumptions. Table 5-9. Current Normal Area and Area of Rice Expansion for 11 Strategies and Two Modified Strategies for 12 GDSSs, Central Region, 1979 Base Area Added Area aoss ("Cougnrafi")t S‘tra tegy ABDFHCEEIGIIIJK ------------------------- 1,000 ta-------------------------- Rainfed: 07A 4 (4) 25 07B 20A 25 203 5 Irrigated: 06A 5 16 16 5 5 16 16 07A 20 20 20 20 9 4 07As 14 113i 10 16A 8 5 5 16B 21A 1 4 4 213 Total 42 25 25 25 25 25 25 25 25 25 25 1 Sensitivity analysis modification 120 Table 5—10. Choice of Typical and Alternative Production Techniques for 11 Strategies and Two Modified Strategies for 12 GDSSs, Central Region 1979 GDSS Production Strategy Technique A BDFH C E E1 G I 11 J K """"""""" Technique---------------____-- Rainfed: Alt (x) Alt Alt x 208 Typ x x Alt Irrigated: 06A Typ X X X x Alt x x A1t X X X x x x Alt llBi Typ Alt 16A Typ x Alt x x Alt Alt x Alt 1 Sensitivity analysis modification 121 All partial budgeting estimates of rice expansion impacts are net impacts arrived at by adding the rice coefficient changes to those of sugarcane or cultivated pasture/milk. Except for rice production change itself, these net impacts should not be confused with impacts on rice alone. Rice Production Maximization Stratggies 0f the 11 rice expansion strategies, strategies A, B and C incor- porated maximization of regional rice production as the expansion decision rule. Strategies A and B assumed current normal parameters except for a specific restriction: A was restricted to alternative rice production techniques and B was restricted to maximum number of potential multiple production cycles. Strategy C allowed free choice of any combination of production techniques, number of production cycles and GDSSs. All inputs for strategy 3 were equal to those of D, F, and H. This is accounted for by the fact that the same GDSSs (irrigated 06A, 16A, 21A) had the highest parameter values for each of the four decision rules used in strategies 3, D, F, and H. Thus, no other GDSSs entered the solutions, and the aggregate impacts totals were equal. Strategy C rice production of 441,000 qq was 128 percent more than the current normal production of 343,000 44. and strategies A and B resulted in close to 100 percent increases, respectively. Production labor use increased roughly in proportion to output for all three stra- tegies. Increases in foreign exchange use for strategy 3 was slightly less than the current normal $843,000, while A and C resulted in 20-30 percent increases from current normal. 122 Monetary RLM at the farm gate level increased by 185 percent for strategy C. Strategy A and B returns increased by a much lower percent- age. Current normal monetary RLM were about $1.3 million. Current normal unsubsidized RLM at the farm gate were $435,000. Strategy C increased that value by about $3.1 million, while the A in- crease was about $2.4 million. The strategy 3 increase was $1.8 million. These large increases in unsubsidized returns over the monetary returns were due to rice's replacing cultivated pasture/milk, a process which produces unsubsidized losses in all GDSSs under the assumptions of the study. Strategies A and 3 resulted in expansion of rice production in cur- rently irrigated GDSSs 06A, 16A, and 21A. Production in rainfed GDSS 07A would have resulted in the same production increase as in 21A; however, net returns were higher in 21A so it rather than 07A entered the solu- tion. Strategy C expanded in currently irrigated 06A and potentially irrigated 07A. The difference between C and strategies A and B was due to the strategy C option of expansion into the potentially irrigated 07A, which had a higher yield than irrigated 16A and 21A. Strategies A and B were restricted, respectively, to alternative and typical rice production techniques. Strategy C allowed free choice between techniques, and the result was selection of typical techniques in irrigated 06A and alternative techniques in irrigated 07A. Maximization of Rice Monetary Returns at the Farm Gate Strategies 0 and E used maximization of rice monetary RLM at the farm gate as the expansion decision rule. Strategy 0 was restricted to typical production parameters and E allowed free chOice among parameter alternatives. 123 All impacts for strategy 0 were the same as those for strategies 3, F and H. Increases in monetary RLM for strategy E were about $2.7 million greater than current normal returns of $1.3 million. The strategy E returns increase was $1.3 million greater than the 0 increase. In- creases in unsubsidized returns for strategies 0 and E were $1.8 and $3.3 million, respectively, over the CN returns of $435,000. The increase in white rice production for strategy E was about 127 percent over that for current normal and strategy 0 production. Strategy 0 impacts were similar to those of E in terms of production labor use. Net foreign exchange use increase for strategy 0 was slightly less than the current normal value of about $843,000. Strategy E showed a 25 percent increase. The GDSS level impacts of strategies 0 and E were the same as those for B and C, respectively. Selection of rice production technique dif- fered, however, in that the alternative techniques were selected in free choice strategy E. Maximization of Rice Unsubsidized Returns at the Farm Gate Strategies F and G employed maximization of rice unsubsidized RLM at the farm gate as the rice expansion decision rule. Strategy F was re- stricted to current normal production parameters. Strategy G allowed free choice among parameter alternatives. Strategy F results for all coefficients were the same as those for strategies 3, D and H, as explained previously. Current normal monetary returns were $1.3 million. These were increased by $1.2 million and $2.5 million with strategies F and G. Current normal unsubsidized returns were increased by about $1.8 million and $3.3 million with strategies F and G. Strategy G impacts were generally similar to those of E. 124 The implication of this similarity in comparisons between strate— gies D, E and F, G is that it made little difference at the regional level whether monetary or unsubsidized returns were selected as the ex- pansion decision rule. However, at the GDSS level there was a difference in impacts, with strategies 0 and F expanding production in irrigated GDSSs 06A and 21A, strategy E expanding in irrigated 06A and 07A, and strategy G expanding in rainfed 203 and irrigated 07A. These differences were relative ones in monetary and unsubsidized RLM at the GDSS level which were offsetting when aggregated to regional totals. Strategy F was restricted to use of the typical rice production techniques. Strategy G selected typical techniques in rainfed 203 and alternative techniques in irrigated 07A. Maximization of Rice Labor Use Strategies H and I assumed maximization of production labor use in rice production as the rice expansion decision rule. Strategy H was restricted to current normal production parameters and strategy I allowed free choice among parameter alternatives. All impacts for strategy H were the same as those for strategies 3, D, and F. Labor use increased by nearly 100 percent for strategies H and I in comparison to the current normal labor use of 2.3 million hr. Rice production increased 99 and 127 percent over current normal production for both strategies. Monetary returns increased by $1.2 and $1.9 million and unsubsidized returns increased by $1.8 and $2.4 million, respectively, for strategies H and 1. Foreign exchange use increased by approximately 90 percent for both strategies. Both strategies expanded rice production in GDSS 06A, but strategy I also expanded in irrigated 07A rather than in 16A and 21A. This was 125 due to the potential labor use in 07A being higher than the current normal labor use in the other GDSSs. Typical production techniques were forced in strategy H. Strategy I selected typical techniques in irri- gated O6A, and alternative techniques in I. Minimization of Rice Foreign Exchangg Use Strategies J and K assumed a rice expansion decision rule of minimi- zation of foreign exchange use in rice production. Strategy J was re- stricted to current normal production parameters, and strategy K allowed free choice among production parameter alternatives. An added restric- tion which required a minimum of a single rice production cycle in each GDSS was placed on these two strategies. Both strategies resulted in only slight differences from the current normal foreign exchange use of $843,000. Rice production increases for strategies J and K were about 19 percent over current normal production. Increases in monetary RLM were half or less than half of current normal returns for both strategies. Unsubsidized returns increased about 100 percent over current nonnal for strategies J and K. Labor use in- creased less than a half million hours over the current normal level of 2.3 million hr. All strategy J rice expansion took place in rainfed GDSS 07A, and all of the strategy K expansion was in rainfed 20A. Typical production techniques were forced in strategy J, while K selected alternative pro- duction teqhniques. Ranking of Expansion Strategies by Production Impacts Rankings of the top three alternative expansion strategies according to their impacts on the studied production parameters differed depending 126 on the expansion criterion selected (Table 5-11). The ranking using maximization of regional rice production as the expansion decision criterion was C, E, I. Rankings using monetary returns as the decision criterion were E, G, and C, and those using unsubsidized RLM were G, E, and C. When making rice labor use the criterion the ranking was I, C, E, and using the foreign exchange criterion resulted in a ranking of E, G, A. Table 5-11. Ranking of Rice Expansion Strategies on the Basis of Four Impacts, Central Region, 1979 Current Normal Rank Expansion Impact 1 2 3 -------------------- strategy------------------- Rice production C E I RLM, monetary E G unsubsidized G E Labor use I C E Foreign exchange use E G A Only strategy E ranked among the top three for all expansion deci- sion criteria, and it was ranked first for two of the five criteria. Strategy C ranked in the top three for four of the five criteria, and in three instances G was among the t0p three. Results of the Sensitivity Analysis One of the sets of assumptions believed to be most critical in influencing the outcome of the study was that which led to the estimates Of the number of potential multiple cycles of rice production in each 127 GDSS. Most of the solution GDSSs for each alternative rice expansion strategy were ones with three potential cycles. In order to test the sensitivity of the study results to the multiple cycle assumptions, two of the high ranking expansion strategies, E (maximization of monetary returns) and I (maximization of labor use), were rerun with the follow- ing change in the multiple production cycle assumptions: all GDSSs were restricted to a maximum of two production cycles. The modified strategy E results indicated, as expected, that rice production and net regional monetary returns, unsubsidized returns, and labor use all were reduced over the original strategy results by about 33 percent (Table 5-8). Rice expansion took place in two GDSSs with both strategy E and modified E. Both strategies expanded production in irrigated GDSS 07A. However, expansion also took place in rainfed 203 with modified strategy E compared with the irrigated 06A with original strategy E. The difference was due to rainfed 203's having a higher total annual production than irrigated 06A when 06A was restricted to two rice crop cycles per year. Comparisons between strategy I and modified strategy I production parameters are similar to those between E and modified E. At the GDSS level, both the I and modified I strategies expanded rice production in irrigated 06A and 07A, but modified I also expanded production in irri- gated 16A (with reduced area of production in 07A). Chapter V's Relation to Chapter VI The analytical results presented in this chapter will be discussed in Chapter VI. Their implications for rice expansion policy will be covered in light of the data and methodological limitations of the study. CHAPTER VI DISCUSSION OF ANALYTICAL RESULTS Chapter VI contains a discussion of the analytical results presented in the previous chapter. The discussion is organized in the same order as were the major objective-related questions asked in the first chapter. Each question is restated. Then the implications of the study results for answering each question are discussed. Assumptions used in the study and significant limitations of data quality and analytical procedures, as they might have affected study results, are covered. Comments are made as to the appropriateness of the study methodology for the Dominican land use planning environment. There is a discussion of methodological modifications required to expand this regional study to the national level. Finally, the need for additional information on intra- and inter-sectoral factors bearing on Dominican rice production and consumption is emphasized. In What Land Areas Can Rainfed and Irrigated Rice Production Be Expanded? The results of the agroeconomic analyses indicate that rice poten- tially can be produced at commercially marketable yield levels in four rainfed GDSSs (07A, 073, 20A, 203), six irrigated GDSSs (06A, 07A, 16A,16B, 21A, 213) and two subirrigated inclusions (07As, 113) in the Central Region. The most promising of the potentially available GDSSs are 128 129 irrigated 06A and 07A. It is to these two areas that priority attention should be directed for future project planning. Among the rainfed GDSSs, the overall agroeconomic results indicate that 203 and 07A should be given priority attention for more detailed field analysis leading to pro- ject identification and planning. Rice currently is produced in one of the four rainfed and three of the six irrigated GDSSs and in both inclusions. There is substantial physical area potentially available for rice expansion in each of the rainfed GDSSs (34,000 to 115,000 ta), a findings consistent with records indicating much greater earlier rainfed production in the Central Region in the 19605 than is current (Murphrey, 1972). Physical area potentially available for expansion of irrigated rice production in each GDSS is much smaller than that for rainfed production, ranging from 0 to 20,000 ta. Results of the sensitivity analysis of the multiple cycle assumptions indicate the desirability for more detailed study of those assumptions prior to using them in the planning of a rice expansion project. The results confirm the importance of the GDSSs mentioned above as potential rice expansion areas but indicate that irrigated 16A also would be a high priority area for rice expansion if it happened that triple produc- tion was not feasible in the other GDSSs. The term "potentially" should be emphasized here because of the fact that institutional factors such as long ownership, as well as other factors not analyzed in the study might prevent rice expansion or other land use changes in a number of the GDSSs. For example, the analysis assumed current efficiency levels in canal maintenance and water distri- bution to the field level. Given the many uncertainties in water 130 distribution consistency in many areas, due to lack of disciplined control of canal flood gates and lack of an economic incentive for efficient water use (i.e., pricing of water by volume rather than by area irrigated), this assumption must be given careful attention in future project plan— ning. Soils characteristics must be analyzed in the field at a level of detail much greater than that used in this study in order to identify the specific areas within 60555 with sufficient homogeneity to allow uniform water distribution and drainage patterns. In addition to the 60555 and inclusions identified as having rice production potential, there are other small, dispersed GDSSs and inclu- sions which have rice potential in the Central Region. These areas may be very useful and appropriate for production of locally consumed rice, but they are believed to be too small to justify consideration for a nmjor governmental rice expansion project. Atypical years of heavy rainfall, too, may give the illusion that additional areas are suitable for rice production. However, those areas would not be suitable in years of typical rainfall and should not be considered for further rice expansion. Additional areas could become adaptable to production in the future as new rice varieties are developed at the Juma rice experiment station. Increased tolerance ranges for such limiting factors as consumptive water use and soil pH would allow rice expansion in areas not identified in this study as having rice production potential. Varieties with shorter production cycles could open up possibilities for multiple cycle produc- tion in areas that currently have only a single production cycle. Such research advances at the experiment station, however, will have to be complemented and coordinated with technical extension services and input and output marketing assistance in appropriate land areas if a rice 131 expansion project is to have any chance of success. Social infrastructuna investment in such projects as schools and medical clinics will also have to be given careful attention if a major expansion project is to be undertaken. Which Expansion Areas Would Be the Most Profitable from the Producers's and the Nation's Standpoints? In this study, returns to land and management were used to estimate "profitability". More correctly, these are returns to land, management, and all other factors not explicitly specified in the benefit-cost calcu- lations. Results of the benefit-cost analysis indicate that rice can be grown profitably in all of the GDSSs identified in the agrophysical analysis. The most consistently profitable GDSSs under the expansion strategy alter- natives studied were irrigated 06A and 07A and rainfed 20A. Partial budgeting results indicate that profitable rice expansion in those GDSSs could best be achieved through strategies E, C and G. Each of those alternative strategies nearly doubled the regional net monetary returns to land and management and resulted in proportional increases in net unsubsidized returns. Strategies E, which assumed maximization of rice monetary returns, and C, which assumed maximization of rice production, not only resulted in the highest monetary returns and output, respective- ly,but also ranked very high in terms of the other production expansion impacts. Strategy G also had high returns but had relatively lower rice production and labor use. Regional net returns increased substantially due to rice substitu- tion for cultivated pasture/milk under all of the alternative rice 132 expansion strategies considered in the study. Cultivated pasture/milk is notoriously unprofitable in the Dominican Republic (Associacion, 1979). Expansion of rice into sugarcane areas would reduce net monetary returns as sugarcane is more profitable than rice in all GDSSs except rainfed 20A and 203 under current normal conditions. Rice expansion into sugarcane areas would not have taken place, from the standpoint of net unsubsidized returns, except in rainfed GDSS 20A. One of the implications of these net returns impacts is that, from the economic standpoint, the country should rethink its policy of rice self-sufficiency and production area expansion, at least in the Central Region. As long as rice replaces cultivated pasture, the region will be better served in terms of net returns on its resources. However, replace- ment of both cultivated pasture/milk and rice by sugarcane at current normal relative prices would increase regional net returns even more. Distribution of the returns and regional comparative advantage, among other factors, must be considered, but strictly from the standpoint of regional profitability agricultural planners should consider the loss from sugarcane profits which the region is giving up in order to produce or to expand production of rice. . These comments must be considered in light of several critical assumptions made in the study. Perhaps the most critical assumptions were those which established the relative prices of rice, sugarcane, and milk. Used in the study were current normal prices which did not indi- cate the volatile instability of sugar and rice prices in international trade. Sugarcane prices, for example, have varied for approximately $150 to over $1000 per mt during the past decade. The current normal price is $331 per mt. At the lower end of the price scale sugarcane 133 production would be unprofitable in all GDSSs. Rice producers can respond fairly rapidly to price fluctuations due to the short three to four month rice production cycle. Sugarcane and milk producers have less flexibility with products which have multi-year production cycles. These price variability factors should be studied thoroughly prior to land use policy changes or project development. Also important were the estimates of potential for multiple cropping of rice which were tested in the sensitivity analysis. The assumptions that new rice areas can be brought into production with no additional investment costs for irrigation development and no added water costs for dry season pumping of underground water may be unreasonable in specific GDSSs. The large capital investments in sugarmill facilities and the optimum mill operating levels must be considered before proposing any sig- nificant change in production in specific 60555. Project planners should give attention to those assumptions in more detailed project planning for rice expansion. Another set of assumptions critical to the economic analysis was the assumption of perfect. elasticity (If input supply and output demand assumption. Further analysis of those elasticities might in fact indi- cate that input prices would rise or output prices would decline as rice production was increased in the region. More elasticity input supply and output demand functions would in turn mean that equilibrium among the three land uses would occur with something less than complete replace- ment of cultivated pasture/milk by rice or complete replacement of rice and cultivated pasture/milk by sugarcane. Considering, however, the relatively small size of the Central Region and the relatively small percentage of national rice production analyzed in this regional study, 134 the assumption of perfectly elastic functions is probably reasonable except on a localized monopoly/monopsony basis. It should be emphasized before closing this section that maximiza- tion of returns to land and management was only one of the expansion decision rules considered in this study. Use of expansion decision rules other than returns to land and management resulted in different ranking of the "best" strategies. For example, Strategy 1, which assumed maxi- mization of rice labor use, resulted in the highest labor use but was substantially lower than several other strategies in terms of the other impacts monitored. These different rankings of expansion strategies, depending on the decision rule chosen, points to the importance of decisionmakers' goal identification as the first step in the iterative rice policymaking process. Depending on the specific goal and correspond- ing decision rule chosen, the study results identify several different GDSSs to which priority attention should be given for rice expansion project planning. Only by clear definition of policy goals early in the planning process can resources be allocated consciously and most efficiently toward the attainment of those goals and the avoidance of undesirable side effects. Where Can Labor be Increased and Foreign Exchange Use be Decreased in Rice Area Expansion? The analytical results indicate that current normal labor use of 2.3 million hours per year could be increased by over 100 percent by expansion of rice production in irrigated GDSSs 06A and 07A through appli- cation of alternative expansion strategies I, C, and E. All of those 135 strategies resulted in the selection, in one or both of the relevant GDSSs, of the alternative techniques for rice production rather than the typical techniques. Regional net foreign exchange use per qq produced can best be mini- mized through expansion in irrigated GDSSs 06A and 07A. Three of the alternative strategies for expansion of rice production led to an in- crease of at least 100 percent in rice production, with only a 25-38 percent increase in the use of foreign exchange. All of the three stra- tegies selected the alternative rice production techniques over the typical techniques in one or both of the relevant GDSSs. These results clearly indicate that there are land areas and technical production alternatives which could increase labor use and decrease foreign exchange use (per qq of output) in the Central Region. More will be said about the technical alternatives in the next section. Are Alternative Rice Production Technigues Available Which Could Profitably Increase Labor and Decrease Foreign Exchange Use? The results discussed in the previous section indicate that there are alternative production input combinations which could profitably both increase labor use and decrease net foreign exchange use per qq of rice produced in the Central Region. Adoption by current rice producers of the alternative techniques analyzed in this study would increase labor use by 46 percent and reduce foreign exchange use by 58 percent. Rice production planners should take this into account in future project planning and promote the dissemination to producers of infor- mation on labor-using and foreign exchange-saving techniques. At the 136 same time, experiment stations should place more emphasis on research to identify and field test alternative rice production techniques which produce these desired impacts and which are profitable both to the local producer and to the national economy as a whole. The alternative tech- niques analyzed in the study were but an illustrative sample of innumer- able options potentially available to rice producers. It is up to the research, extension, and agricultural information specialists to see that that potential is turned into reality. How Much Are Rice Production Costs Currently Being Subsidized and What Would be the Impact of Subsidngemoval? Current normal rice subsidies average about one-third of total pro- duction costs and amount to about $1.5 million annually in the Central Region. These subsidies are income transfers from taxpayers and con- sumers to rice producers and are not real economic costs to the national economy. Whether they are viewed as right or wrong is a judgment requir- ing consideration of both positive and normative information. The results clearly indicate that the Dominican government views rice sub- sidies as "right." Objectively, the income transfers decrease taxpayer and consumer utility and increase that of rice producers. There are also differential impacts on the producers and consumers of potential rice substitutes which are not subsidized. A number of possible impacts of removal of rice production cost subsidies can be anticipated. It is worth noting that the government could not remove subsidies only from rice inputs, as many of the inputs are used in the production of other agricultural products. There would 137 be no way to prevent the use of subsidized inputs by rice farmers, who would simply obtain them from their non-rice producing, subsidized neighbors. Assuming, however, that the government removed the subsidies (i.e., charged full economic costs) on all inputs used in rice production regardless of their eventual use, the empirical results indicate that rice profitability would decrease in all GDSSs and would be negative in several. The ratios of the marginal value products (MVP) to marginal factor costs (MFC) for the formerly subsidized inputs would decrease, leading to a general producer shift away from the use of those inputs and toward the use of inputs with higher MVP/MFC ratios. Producers with economically fixed assets (Pa MVP Ps) would not change input use. Marginal producers would shift out of rice production if their unsubsi- dized unit costs were higher than the government price support levels or if other land uses (in this case sugarcane) appeared to be more profit- able. Production enterprises which used the highest amounts of previously subsidized credit, foreign exchange, and water (e.g., irrigated GDSSs 06A and 113i) would be most affected. Can a Land Assessment Procedure Be Developed Which Is Appropriate to the Dominican Planning Environment? It is too early to tell whether the land assessment methods developed in this study will be used effectively by the Dominicans after U.S. tech- nical assistance is terminated. A conscious and continuous effort was made by the author to limit the sophistication of the data collection and processing and agroeconomic analysis techniques to a level commensurate 138 with the expected availability of Dominican human and financial resources. It was the belief of the author that if the study methods were to become permanently accepted as useful planning tools, the study's level of technical SOphistication had to be matched both to the expected techni- cal understanding of the somewhat transient SIEDRA technical staff, and to that of the highly transient crops of agricultural sector administra- tors who have to be repeatedly convinced to allocate their scarce resources to support the SIEDRA work. The multidisciplinary nature of the study and its heavy dependence on secondary and judgmental data necessitated a time consuming continuing effort by the SIEDRA staff to explain project purposes, methods, and data needs to itinerant administrators of the diverse data agencies who provided published or judgmental data for the study. Although these procedures were very beneficial in effecting inter-institutional cooper- ation that was virtually unheard of previously, there is a real question as to the permanence of the linkages and, therefore, to the viability of these data collection procedures in the future because of the personnel turnovers among administrators and technicians. As a result of the uncertainties in these delicate inter- institutional linkages critical to the future viability of the study procedures, SIEDRA is being encouraged to integrate farm level land use surveys into its overall data acquisition procedures as resources become available. Early SIEDRA efforts at integration of its data collection efforts with those of a SEA agricultural analysis project (ANSE) have proven fruitless to date because of lack of adequate support for ANSE. The dialogue proposing integration will continue in the interest of eventual strengthening of both SIEDRA and ANSE. Farm surveys should not 139 replace the secondary and judgmental data but should serve to reduce the need for the complicated cooperative relationships required among agri- cultural institutions with very diverse operational orientations. Agricultural sector surveys, properly designed and executed to fulfill SIEDRA land use data needs, would likely result in much better data quality, particularly for cultivated pasture and sugarcane production and would assure direct compatibility of the complete sector data set. Because of the complexity and difficulties of the multidisciplinary nature of the SIEDRA work and because of institutional jurisdiction disputes vis-a-vis economic analysis, there has been an increasing feel- ing among the SIEDRA staff that SIEDRA should concentrate its efforts on agro-physical analysis and relegate the economic analysis to other institutions. That type of "division of labor" could prove beneficial to the extent that inter-institutional coordination could be effected to provide for the economic analysis on a timely basis. However, attempt; at such coordination efforts to date have come to nought. The danger in carrying out agro-physical analyses to the exclusion of economic analyses is that resulting recommendations could prove to be economically dis- astrous. For example, a large expansion of rice production in GDSS 07A, as based on the agrophysical results of this study, could prove an econ- omic boondoggle if no consideration is given beforehand to the economic questions of land opportunity costs for sugarcane production and of the impacts on sugarmill operating efficiency of a reduction of sugarcane production in the mill area. 140 Modifications Needed in the Regional Study for National Application There are several types of modifications that would be required in the study methodology in order to apply it at the national level. Five of the more critical considerations-are discussed below. First is the question of improvement in data acquisition procedures, discussed in the previous section. Land use surveys at the farm level, ideally as a component of a more comprehensive sector survey, should become a part of the SIEDRA data collection process as soon as resources can be made available. The agricultural ministry (SEA) already has some limited experience with farm surveys for obtaining agricultural sector data, and the SEA survey personnel could likely administer the land use surveys needed by the SIEDRA staff. Financial support for a separate SIEDRA land use survey has not been available to date. A second modification that should be considered is that of the assumption of perfectly elastic input supply and output demand functions. This assumption will have to be reevaluated in the light of the magni- tude of the contemplated changes in input use and resulting output. The assumption of constant input and output prices is valid only as long as the changes in quantities of inputs and outputs are not sufficient to affect market prices. In the event that production changes are of a magnitude sufficient to affect input and output prices accordingly. Due to the almost complete lack of credible data on either supply or demand elasticities, a separate study is needed to develop information on price responses for key inputs such as labor and each of the major products. 141 Another consideration for a national study is the relatively greater importance of transportation costs in moving inputs and outputs to multi- ple consumption markets. In the regional study an average transportation cost for all GDSSs to the Santo Domingo retail market was calculated but was a relatively insignificant componentiof total costs. At the national level there will be multiple consumption markets for the vari- ous products and many more GDSSs from which production~will be marketed. Transportation costs and their impacts on regional comparatiVE‘advantage for the different products will become increasingly more important as fuel costs rise. Gasoline which was $0.99 a gallon early in 1979 rose to $2.39 by mid-1980 and there is no end in sight for future price increases. A fourth factor which must be considered before undertaking a national level land base assessment is that of computerization of the SIEDRA data base. Almost all of the data refinement and analysis com- putations carried out in the study were done on hand calculators. For the regional study, which was limited to 29 GDSSs and three land uses, the use of hand calculators was both feasible and desirable from the training standpoint. The calculators forced the SIEDRA staff to pay continuous attention to every number processed and required more staff members to become involved in the computations than would have been necessary with the use of a computer. On the other hand, errors in data entry and in the copying of numbers between calculators were a continual nightmare that consumed enormous amounts of previous staff time. A national study, which would involve at least 100 GDSSs and up 3’ to 25 alternative land uses, could not be carried out efficiently with- out computerization of the data base. 142 A final primary consideration in going to a national study and to additional regional studies is that of increased sophistication of economic analysis techniques. Increased analytical 50phistication would require more highly trained analysts than those on the current SIEDRA staff. Because of the organizational structure of the SEA, in which there are separate sub-ministries for economic analysis and for land use analysis, SIEDRA has been quite successful in obtaining highly qualified personnel from the non-economic disciplines but has had very limited success in obtaining good economic analysts. This institutional obstacle to improved economic analysis by the SIEDRA staff will have to be overcome if SIEDRA is to progress over time into such analytical techniques as linear and non-linear programming, various econometric techniques, as components of more general systems simulation models of the agricultural sector. The Need for Additional Information The final comment in this discussion of the study results is a repetition of a statement made in the first chapter: Policymakers need to obtain information in addition to this land base assessment before making policy decisions related to land use changes. There are many critical factors vearing on land use alternatives which were not studied in this land assessment. For instance, a major rice area expansion into former cultivates pasturelands would, according to the study results, significantly increase labor use in the area. Labor availability was not analyzed in the study but is obviously a critical consideration in a land use change requiring labor intensification. Also, a significant 143 migration of laborers into an area implies that the government will have to invest in certain social overhead capital, such as schools, medical facilities, and roads if the laborers are to become permanently settled in the area. Investment in social overhead capital was not analyzed in the study but is a factor which policymakers must take into consideration if a major land use change is to be successfully made. History is strewn with examples of costly, unsuccessful attempts at land use changes (e.g., colonization in several Latin American countries) in areas where the principal selection criterion was that they "had good soils." CHAPTER VII SUMMARY AND CONCLUSIONS Summary The Dominican Republic has been importing increasing quantities of rice since 1972. In 1978 the government began a program to increase domestic rice production in order to reduce foreign exchange expendi- tures and to increase employment. As a sub-component of that program, this study was undertaken to assess the agronomic and physical ("agro- physical") and economic ("agroeconomic") feasibility of rice production expansion in each land unit ("GDSS") in the Central Region of the country (see Figure 2-1). The regional study was to serve as a proto- type for a national rice land use assessment and for studies of other agricultural land uses. The Central Region was selected for study because of data avail- ability and because of the region's proximity to the Santo Domingo office of the technical staff ("SIEDRA") conducting the study. This proximity permitted relatively inexpensive field training of the SIEDRA staff and verification of questionable data. The analysis was to two types. First, an agrophysical analysis of the land base was carried out in order to identify areas with potential for increased rice production. Rice plant tolerance limits for poten- tially limiting soil and water characteristics were estimated and crosstabulated with the corresponding GDSS characteristics. For each 144 145 of the 60855 selected through that process estimates were made of the area potentially available for rice expansion, current normal (expected) rice yields and production, and possibilities for multiple rice produc- tion cycles. Second, an economic analysis was made of the GDSSs selected in the agrophysical analysis. Benefits and costs of the production of rice and of its two principal competitors for land in the Central Region (sugar- cane and cultivated pasture used for milk production) were analyzed from the producer cash expense ("monetary") and the national opportunity ("unsubsidized") points of view. A typical current set of rice production techniques and an alternative set requiring increased labor and/or decreased foreign exchange use were analyzed. The benefit-cost results were used in partial budgeting of a hypothetical 25,000 ta expansion of rice production area using 11 alternative strategies. The strategies consisted of various assumptions on expansion decision rules (maximiza- tion of rice production, maximization of rice monetary and unsubsidized returns to land and management, maximization of rice production labor use, and minimization of rice production foreign exchange use) and policy variables (rice production techniques, number of rice production cycles per year, and expansion in current or potentially available GDSSs). Results of the agrophysical analysis indicated that rice production was adaptable to rainfed GDSSs 07A, 073, 20A and 203, irrigated GDSSs 06A, 07A, 16A, 16B, 21A and 213, and to subirrigated inclusions 07As and 113i (see Tables 4-2 and 4-3). Total physical area potentially available for rice expansion was estimated at 310,000 ta (1 ta - 1/15.9 ha) for rainfed and 75,000 ta for irrigated production (see Table 5-2). Multiple rice production cycles were determined to be feasible in eight of the twelve 60555 and inclusions. 146 Economic analysis results indicated that irrigated GDSSs 06A and 07A and rainfed GDSS 203 should be given priority attention for rice expansion projects in the Central Region. Expansion in those areas would be profitable both to producers (monetary returns) and to the nation (unsubsidized returns) using a combination of typical and alternative rice production techniques. Labor use would increase substantially and foreign exchange use per unit of rice produced would decrease if rice were substituted for cultivated pasture used in milk production. Rice substitution for sugarCane would not be economically rational under the assumptions of the study. Conclusions Conclusions drawn from the study can be divided into policy and methodological conclusions. There are two major policy conclusions. First, there are land areas in which rice could be produced profitably from both the producer's and the nation's standpoints and at the same time provide increased employment opportunities and decreased use of foreign exchange per unit of rice produced. Irrigated 60555 06A and 07A and rainfed 203 are the best prospects. A 25,000 ta expansion of rice production in those GDSSs under strategy E (maximization of rice monetary returns with free choice of production technique and GDSS) would increase annual labor use by 2.4 million hr and increase brown rice production by 4.4 million qq. The unsubsidized cost of domestic production would be $3.2 million of which $0.3 million would be used for foreign exchange. An equivalent quantity of imported rice would cost $5.8 million. Thus, domestic production of the 4.4 million qq of rice 147 which would otherwise be imported could save the country $2.6 million total and $5.5 million in foreign exchange. The $2.6 million in savings could be interpreted as the economic value of this rice study. Second, there are profitable alternative rice production techniques which could increase labor use and decrease foreign exchange use. Adoption in current rice production areas of the alternative techniques analyzed in this study would both increase labor use and decrease foreign exchange use by about 50 percent. Such techniques must be introduced to producers through the Dominican research and extension institutions if they are to bring about the desired impacts in the field. There are three major methodological conclusions. First, the study methods seem to be appropriate to the current Dominican planning environ- ment. However, the secondary and judgmental data sources used in the study should be supplemented by land use surveys at the farm level, preferably as components of more comprehensive agricultural sector surveys. This integration of surveys would help to assure data compat- ibility and completeness for overall sector analysis. Second, several of the critical assumptions made in this paper should be given more detailed study prior to use in project planning. The order of priority should be: (a) number of rice multiple production cycles possible in each GDSS; (b) relative product prices; (c) irriga- tion infrastructure costs; and (d) unsubsidized prices. Third, other intra- and inter-sectoral production and consumption information in addition to this land use assessment should be incorporated in the rice expansion policymaking process in order to increase the probability that a feasible and desirable expansion policy and resulting projects can be planned and implemented. More detailed land base analysis in the field 148 is required to locate, within the selected GDSS, specific areas suitable for project development. BIBLIOGRAPHY BIBLIOGRAPHY AID. "Agricultural Sector Assessment." Santo Domingo: USAID Mission, 1980. Agricultural Attache. Personal conversation. U.S. Embassy, Santo Domingo, 1980. Amin, G. A. "Cost-Benefit Analysis and Income Distribution." DeveTOp- ment Digest, April, 1979. Associacion Dominicana de Produccion de Leche. "Comportamiento y Situacion Actual del Subsector Productor de Leche en la Republica Dominicana." Santo Domingo: Secretaria de Estado de Agricultura, 1979. BAGRICOLA. "Indices de precios de algunos insumos agricolas." Santo Domingo: Banco Agricola, 1979. Bookers Agricultural and Technical Services, Ltd. "Study of Rehabili- tation and Expansion of the Sugar Industry." Santo Domingo: Consejo Estatal del Azucar, 1975. Brown, M. L. "Budgets for Analyzing Farm Income and Agricultural Pro- jects (Draft)." Economic Development Institute, World Bank, 1977. Burton, G. "Breeding Better Forages to Help Feed Man and Preserve and Enhance the Environment." Bioscience, December, 1973. Castle, E., Becker, M. and Smith, F. Farm Business Management. New York: Macmillan, 1972. CEAGANA. Telephone conversation. Santo Domingo: Consejo Estatal del Azucar-Ganaderia, 1980. CRIES. "Proposal for a land resource analysis in the Dominican Republic." East Lansing, Michigan: USDA-ESCS-APPS, 1976. . "Land resource base report (draft)." CRIES 77-1, East Lansing, Michigan: USDA-ESCS-APPS, 1977a. . "Comprehensive Resource Inventory and Evaluation System." CRIES 77-5, East Lansing, Michigan: USDA-ESCS-APPS, 1977b. . "A user's guide to the SIEDRA national mapping system." CRIES 78-1, East Lansing, Michigan: USDA-ESCS-APPS, 1977c. 149 150 "Visual interpretation of LANDSAT imagery for land cover and land use of selected test sites in the Dominican Republic." CRIES 77-2, East Lansing, Michigan: USDA-ESCS-APPS, 1977d. . "User's guide to the CRIES analytical model for the Dominican Republic." East Lansing, Michigan: USDA-ESCS-APPS, 1979a. Drilon, J. 0., Jr. "Dominican Rice Strategies.“ Santo Domingo: USAID Mission, 1977. Edwards, C. "Budgeting and programming in Economic Research." in Methods for Land Economics Research. University of Nebraska press, 1966. Evans, C. Foundations of Plant Geography. New York: Harper, 1963. FAO. La Zonificacion de los Cultivos en el Valle Cibao, Santiago: Instituto Superior de Agricultura, 1974a. "Estimacion de los Costos de algunos cultivos." Santiago: Institu- to Superior de Agricultura, 1974b. A Framework for Land Evaluation. Rome, 1976. "Plan Internacional para la Coordinacion del Fomemoto Lechero (PICFL) y Plan Internacional para la Coordinacion del Fomento de la Carne (PICFC). Santo Domingo: Direccion General de Ganaderia, 1978. Hargreaves, G. H. Water Requirements Manual for Irrigated Crops and Rainfed Agriculture. Utah State University, March, 1977. Heady, E. Economics of Agricultural Production and Resource Use. Prentice-Hall, 1952. Holdridge, L. R. Life Zone Ecology. Turrialba, Costa Rica, 1968. Hunt, 0. Farm Power and Machinery Management. Iowa State University, 1977. IAD. Programa-1979. Santo Domingo: Instituto Agrario Dominicano, 1979. 103. Estudio sobre costos de produccion de leche para plantas pasteurizadoras. Santo Domingo: Inter-American Devel0pment Bank, 1973. INDRHI. Plan NaCional de Aprovechamiento de Recursos Hidraulicos Superficiales: Tomo III-Datos Basicos. Santo Domingo: Instituto Nacional de Recursos Hidraulicos, 1976. "Distritos de Riego." Santo Domingo: Instituto Nacional de Recursos Hidraulicos, 1979. 151 INESPRE. "Commercializacion del arroz.” Santo Domingo: Instituto de Estabilizacion de Precios, 1979. Telephone conversation. Santo Domingo: Instituto de Estabilizacion de Precios, 1980. INTABACO. "Zonificacion y tipificacion del tabaco negro en la Republica Dominica." Santiago: Instituto del Tabaco, 1978. IRRI. Rice: Soil, Water, Land. Los Banos: International Rice Research Institute, 1978. Rainfed Lowland Rice: Selected papers from the 1978 IRRI Confer- ence. Los Banos: International Rice Research Institute, 1979. JUMA. Personal interview with Dr. Hsieh, Chief of Tawainese Technical Mission. Juma Experiment Station, 1979. Kemph, G. S. "Growth of Selected Plants Under Trans-Pecos Conditions." Master's Thesis. Texas A & M University, 1975. "Potential distribution of selected cool season range forage plants in the Trans-Pecos." (mimeographed paper) Texas A & M University, 1973. Kenah, M. "Cartographic analysis of lime production in the Dominican Republic." Master's Thesis. University of Florida, Gainesville, 1980. Klages. Plant Environmental Tolerances. New York: Macmillan and Sons, 1942. Knox, E. Personal correspondence. Soil and Land Use Technology, Incor- porated. Colombia, Maryland, 1979. Kornai, J. "Models and policy: The dialogue between model builder and planner," in EconomyeWide Models and Development Planning. Oxford University Press, 1975. Leibig, J. Law of the Minimum. Germany, 1840. Morales, J. Personal conversation. Santo Domingo: Consejo Estatal del Azucar, 1980. Murphrey. "Rice situation." Santo Domingo: USAID Mission, 1972. Niehaus, R. 0. "Normalized Prices for Resource Planning: Theory and Measurement." USDA-ESCS-NRED, February 1978. Odum, E. P. Fundamentals of Ecology, 3rd edition, Philadelphia: Saunders Company, 1971. Paulet, M. Personal conversation. Santo Domingo: Instituto Inter- americano de Ciencas Agricolas, 1979. 152 Reyna, E. and Paulet, M. "Requirimientos de agua para la agricultura segun el clima de la Republica Dominicana." Santo Domingo: Secretar- ia de Estado de Agricultura, 1979. Rossmiller, G. E. Agricultural Sector Planning--A General System Simu- lation Approach. Michigan State University, 1977. SEA. "El cultivo del arroz." Santo Domingo: Secretaria del Estado de Agricultura, 1975. ____. "Documentos Basicos--Arroz, Maiz, Sorgo." Santo Domingo: Secretaria de Estado de Agricultura, Marzo, 1975. Estrategia de Desarrollo del Sector Agropecuario, 1976-86. Santo Domingo: Secretaria de Estado de Agricultura, 1976a. Situacion del cultivo del arroz en Republica Dominicana. Santo Domingo: Secretaria de Estado de Agricultura, 1976b. Diagnostico del Sistema de Mercadeo Agricola en Republica Dominicana. Santo Domingo: Secretaria de Estado de Agricultura, 1977a. Aspectos del Empleo Rural en Republica Dominicana. Santo Domingo: Secretaria de Estado de Agricultura, December, 1977b. "Rice Statistics." Santo Domingo: Secretaria de Estado de Agricul- tura, 1977c. "Aspectos de la produccion de algunos cultivos." Santo Domingo: Secretaria de Estado de Agricultura, 1979a. "Programa de Fomento Arrocero." Santo Domingo: Secretaria de Estado de Agricultura, 1979b. Shelford, V. E. The Law of Tolerance. New York: Macmillan, 1913. Shin, Y. H. "Use of Soil-Survey Data in Korea: Land Selection for TongilJ' in Soil Resource Data for Agricultural Development, edited by L. D. Swindale. Honolulu: Hawaii Agricultural Experiment Station. SIEDRA. "Agrupaciones de subgrupos dominantes de suelo (borrador)." Santo Domingo: Secretaria de Estado de Agricultura, 1978. "Uso potencial de la tierra: Evaluacion del recurso suelo-- Regional Central." SIEDRA No. 2. Santo Domingo: Secretaria de Estado de Agricultura, 1979a. “Use potencial de la tierra: Evaluacion del recurso agua-- Regional Central." SIEDRA No. 3. Santo Domingo: Secretaria de Estado de Agricultura, 1979b. "Metodologia para la recoleccion de datos agropecuarios mediante entrevistas regionales." Santo Domingo: Secretaria de Estado de Agricultura, 1979c. 153 . "Confiabilidad de los datos agroeconomicos--Regional Central." SIEDRA No. 4. Santo Domingo: Secretaria de Estado de Agricultura, 1979d. Somasiri, S. "Evaluation of Rice Lands in Mid-Country Kandy District, Sri Lanka." in Soil Recourse Inventories and Development Planning. Agronomy Mimeo No. 79-23, Department of Agronomy, Cornell University, 1979. Suryatna, E. S. Rainfed Lowland Rice in Indonesia, in Rainfed Lowland Rice: Selected Papers from the 1978 International Rice Research Con- ference. Los Banos: International Rice Research Institue, 1979. Tilmann, S. "Proposal for the rational use of the CEA land resource base." Santo Domingo: USAID Mission, 1979. USDA. Soil Taxonomy. USDA-SCS, AH-436, 1975. Weaver, J. E. and Clements, F. C. Plant Ecology. New York: Wiley and Sons, 1939. Winch, F. E., III. Costs and Returns of Alternative Rice Production Systems in Northern Ghanai: Implications for Output, Employment and I Distribution. Ph.D. Dissertation, Michigan State University, 1976. World Bank. Country Profile--Dominican Republic. Washington, D.C.: World Bank, 1978. . The Sugar Industry in the Dominican Republic. Washington, D.C.: World Bank, 1979. GENERAL REFERENCES Bradford, L. and Johnson, G. L. Farm Management Analysis. New York: Macmillan, 1952. Johnson, G. L. and Zerby, L. K. What Economists Do About Values: Case Studies of Their Answers to Questions They DonTt Dare Ask. East Lansing, Michigan: Michigan State University, 1973. HICHIGRN STRTE UNIV. LIBRRRIE IIIII IIIIII III IIII |3IIII|IIII II|IIIIIIIIIIIIZIIIIIIIIIIII