we. FEASIBILITY 0F USING smummn m EVALUATE ALTERNATIVE SYSTEMS 0F ' BEEF PRODUCTION IN NORTHEAST Emu. Thesis for the Degree. of M. 3. MB‘CWGAN STATE iiNNERfiW ‘ 5mm Nil LEHK£F€ 19?G ‘ ' -!- “wflulll-I.| ‘ I::‘ by I Eff-1i:: R514" J1;g3‘1ly «L Universitth { q“ ‘1 £13 '3 ALTERATIY" S THE FEASIBILITY OF USING SIMULATION TO EVALUATE ADTERNATIVE SYSTEMS OF BEEF PRODUCTION IN NORTHEAST BRAZIL By ’ If." {I John N;‘Lehker A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Economics 1970 TE T US$93 KL‘I‘FE This thesis min. the Niel Eesented 3“": “term-18, an economic :1 “501 is 1 QEhfies "ho i tang. t° the The Roi; filiaations 111?; Cash (2 15! D38 of 1 3413 avenue: :53er Of ‘ . 3 15+ I ABSTRACT THE FEASIBILITY OF USING SIMULATION TO EVALUATE ALTERNATIVE SYSTEMS OF.BEEF PRODUCTION IN NORTHEAST BRAZIL by John N. Lehker This thesis presents a study of the feasibility of using simulation to analyze the planning of beef production in Northeast Brazil. The nodal presented focuses on the detailed production relationships of the beef enterprise and other enterprises within the firm which have impor- tant economic interrelationships with beef production. The purpose of the model is to develop a conceptual framework which would be useful to agencies who are responsible for rendering technical or financial assis- tance to the entrepreneur in the process of making investment decisions concerning his beef enterprise. The model provides for detailed accounting of land use among various combinations of feed production and between feed production and compet- itive cash crOps, competitive meaning a competition between the two for ‘the use of land. The model further translates feed production into costs and revenues resulting from the production of cattle and provides a con- parison of the returns from the use of land for feed and cash crop pro- d notion. The model also studies the time lag between the time planning decisions ‘are implemented and the time the full benefits of such decisions are real- ized. By the attention given to this lag the model is able to ascertain the sesame effects : akematives but wives Herd manage scanners or ‘. in metal stab: "glides will t int becomes p John N. Lehker economic effects to the firm of not only the land use and herd management alternatives but also of the rate of change from present to future alter- natives. Herd management and sales policies are predetermined by the user and remain more or less fixed throughout the simulation cycle. In order to keep the model stable as the relationship between land and animals change, these policies will be changed once during the simulation cycle when the total herd becomes predominantly modern rather than traditional. The research as: Brazil and 5 mm .‘Yiduester the Brazil Sim} 5in appreciatf It has has Smite}. and Margit} . T Eli-awn. thesi ‘439 author owe 31'. Thomas J. 3339! Scienc “TITO:- worked d '33:. ACKNOWLEDGMENTS The research and writing of this thesis was conducted both in North- east Brazil and at Michigan State University. Its funding was supplied by the Midwestern Universities' Consortium for International Activity and the Brasil Simulation Project. To both organizations the author is grate- fully appreciative. It has been the author's privilege to have been associated with, supported. and guided by three outstanding professors at Michigan State ‘University. To Dr. Harold M. Riley, major professor, and Dr. Marvin L. Hayenga, thesis supervisor, both professors of Agricultural Economics. the author owes many thanks for their patience and guidance. It was with Dr; Thomas J. Manetsch, Associate Professor of Electrical Engineering and System.Science and the director of the Brazil Simulation Project, that the author worked most closely in this study. His help was vital to the whole work. Dr. Dale E. Hathaway and the Department of Agricultural Economics tieserve many thanks for their help and financial assistance to the author during his studies at Michigan State University. The author would like to express his loving thanks to his wife and children whose patience and endurance through this whole episode were decisive to its completion. ii Chapter I In Chapter II III TABLE OF CONTENTS Description INTRODUCTION.................. Background.................. The Problem. . . . . . . . . . . . . . . . . . Objectives . . . . . . . . . . . . . . . . . . General Review of Simulation, Its Advantages, and Previous Research. . . . . . . . . . . . . Research Technique . . . . . . . . . . . . . . THEBASICMODEL................. Introduction . . . . . . . . . . . . . . . . . The Traditional Firm . . . . . . . . . . . . . The Animal Nutrition Sector. . . . . . . . . . The Herd Management Sector . . . . . . . . . . The Overall Model. . . . . . . . . . . . . . . THE DETAILED MODEL . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . The Demographic Subroutine: The Determination of Herd.Size and Productivity. . . . . . . . . Cash CrOp Subroutine . . . . . . . . . . . . . "SUBROUTINEPLAST"............... "SUBROUTINE BOD". . . . . . . . . . . . . . . . Other Subroutines and Functions. . . . . . . . The Structural Equations e e e e e e e e e e 0 iii Page (n \0 r4 +4 15 21 21 25 26 33 39 39 #1 I+8 50 52 53 53 Chapter IV Append ix Chapter IV V Appendix I II III Bibliography Description MODEL TESTING. . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . Testing Criteria . . . . . . . . . . . . Tests of Alternatives. . . . . . . . . . The Sensitivity Tests. . . . . . . . . . Suggested Areas of Future Investigations SUMMARY AND CONCLUSIONS. . . . . . . . . . Introduction . . . . . . . . . . . . . . what Has Been Accomplished . . . . . . . What More Needs To Be Accomplished . . . Conclusions Concerning the Industry. . . Suggested Areas of Further Investigation Important Areas for Future Research. . . GmSARYOOOOOOOOOOOOOOOOO PRINTOUT OF THE COMPLETE MODEL e . . . . . FIFTEEN YEARS OF SIMULATION OF THE TRADITIONAL FIRMOOOOOOOOOCOOOOOOOO... iv Page 80 80 81 83 88 96 99 100 101 103 106 106 109 117 140 Table I Estil II cyniz1 II Igand IV Para Y Para VI Para HZ Table II III IV VI VII LIST OF TABLES Description Estimated Coefficients of Animal Production Characteristics of the Traditional Firm . . Land Alternative Costs. . . . . . . . . . . Parameters Incremented Minus 10 Per Cent. . Parameters Incremented Plus 10 Per Cent . . Parameters Incremented Plus 30 Per Cent . . Parameters Incremented Plus 50 Per Cent . . 1962. Page 27 32 9O 91 92 Figure VII Figure II III IV VII LIST OF FIGURES Description Northeast Brazil Drought Polygon. . . . Basic System. . . . . . . . . . . . . . Animal Nutrition Sector . . . . . . . . Herd Management Sector. . . . . . . . . The Overall Model . . . . . . . . . . . Traditional.Birth and Death Rate Curves Modern Birth and Death.Rate Curves. . . vi 31 35 “3 Since the c: 511de for sugar 91‘ the country, sea; food becam I?” '35 devot 1“ Interior to The intern 155‘. for many y‘ 153m one‘half ' $39311! all 0 3121a. CHAPTER 1 INTRODUCTION Background Since the coastal region of Northeast Brazil was found to be well suited for sugar production, it was one of the earliest settled regions of the country. As sugar production increased, a need for plentiful, cheap food became apparent to supply the labor force. Since the coastal region was devoted largely to sugar production, the settlers turned to the interior to produce the demanded food supply. The interior is subject, periodically, to severe droughts which may last for many years. The drought polygon of Northeast Brazil comprises about one-half the total area of the nine states of the region. It takes up nearly all of seven of the nine states and about half of the state of IBahia. (Figure 1) Only the state of Maranhao is excluded from the poly- gon.1 Due to the unevenly distributed and unreliable rainfall in the interior, the area was found to be unsuited to most types of agriculture, particularly cash crops. In years when rainfall is normal, nearly all of it comes in a three or four month period from mid-January to miqupril or iMay, and the rest of the year the interior receives little or no rainfall.2 18tefan H. Robock, Brazil's Developinngortheast (Washington, D.C.: Brookings Institution, 1963). p. 9+. 21b1de9 p. 72e II" ‘ll 1. ‘i'ul‘ll'ler. It', ‘NIIOII ' i | U.’ II. 't”l.’I"-“-C 1"... I: V in, .Q -I 9.- I! .‘ I." So .~, 4"“ w-nww— ~ 0 o A-‘ -4... . A. “MW W .o."- ‘Mrfiuu‘l'.’ A— 'w— v—w- m c V’umzow Rives FIGURE I .Nbrtheast Brazil: Drought'pown ,v , v .. . r 48‘ 46‘ 44" 55‘ I 4="'t:¢.nm:.\ . /.’ m “M“ \ mum. . , {”700 some ‘\ .34 ' 71/ ”(V “W ‘ 2 an? 8'08 Panama \ §$ 4030 Mm ! "I 3) _.' P I A U I 10 \,.-£,,,-‘ 'M‘L—xg Passes /. m - \I 3' “~ RECIPE ‘ ° ewvsfw-.assa29 . ’\ \\\:“ _’-' Jdi'm «an male 17"” IlACEIO “TN mm" \\\ WWI! \\\‘ e. ,\‘ “.9. «at; \ KN; 8mm" Atlantic dl“ “on? N“ , O c e o n 'o ‘0‘“ o u . ‘ \ .\ mu .9... r A _ o m 200200me "J MI NAS GERAIS ,-V 0 STATE camw. ~_ ' ' , 0 once cmts '4 ...... 5‘ /\ ['9 0.9 -—..— sure eomonnv p we mean: Ni measures ‘1 L I ‘/V 000.... as“ ea scenes: Sources Stefan H. Robock, Brazil's Develo (Washington, D Northeast .0“ Brookings Institution, 1&3), p. 71. Ibsen a severe drc MI hyped by a Even though 105ml to inprc louver, studie: for irrigation 1‘11 now in (zuj~ :35: ”0.98: can. my the pro. income in the $6 4331' Cent, 01 When a severe drought occurred in 1958, the estimated size of the cattle herd dropped by as much as 31 per cent in the state of Ceara.3 Even though cash crops are unsuited to the interior, it would seem logical to improve suitability through the use of widespread irrigation. However, studies sponsored by the government concluded that the potential for irrigation was very small, encompassing about 1.4 per cent of the land now in cultivation.)+ Because of the interior's unsuitability for cash crops, cattle quickly became an important agricultural product. Our- rently the production of cattle generates 2H per cent of the total farm income in the Northeast, while agriculture, forestry and fisheries generate “6 per cent of total income in this region.5 The Problem The Northeast is the poorest region in Brazil. The per capita income in 1960 was about 50 per cent of the national average. The mean income per capita for the Northeast was about stsluo.‘S The distribution of income is also highly skewed: for example, in the city of Fortaleza, about one-half the total population received only about 25 per cent of the total income.7 This low and skewed income is attended by low levels of caloric intake, 3Anuarios Estatisticos do Brasil (Rio de Janeiro, Brasil: Institute iBrasileiro de Geografia e Estatistica), Vols. l8, 19, 20. “Robock, op. cit., p. 6h. 5.1.259... PP- “9-55. 6M” p. 311'. 7Supginento de Generos Alimenticos Para a Cidade de Fortaleza (Fortaleza, Ceara, Brasil: fianco do Nordeste do Brasil§/l, Departamento de Estudos Economicos do Nordeste, December, 196a), pp. 12 ff. high ulnutritic: In spite of the total popula‘ 1950. Koreover, ligation and ma Sharla source of that the agl'iculi m Opportunitt If the ham Ina intuitiVQ Sc 31181151“ News menu! is tag. 113' encouraging 15 th! mV'lous Name". It “I “5 approact Th' CattL $19 SGCOnd 1a.:- high malnutrition and high infant mortality.8 In spite of a high rate of rural to urban migration, two-thirds of the total population of the Northeast still resided in rural areas in 1960. Moreover, the rural-urban income gap is widening, despite the migration and high urban unemployment.9 Since agriculture is the largest single source of income in the Northeast, the apparent deficiencies suggest that the agricultural sector is not providing sufficient income and employ- ment opportunities. If the basic problem is low income and insufficient employment, then one intuitive solution is to shift away from labor-saving land and capital intensive means of production toward more labor intensive means. This generally is taken to mean discouraging mechanized and pastoral agriculture and encouraging labor intensive cash crap production. The difficulty here is the previously noted unsuitability of much of the interior for cash crop production. Thus the intuitive solution does not seem to be a feasible way of approaching the problem. The cattle industry produces 24 per cent of total farm income, making it the second largest single industry in agriculture in the Northeast in terms of income generated.10 Since cattle production is well suited to the interior anomald w. Larson, ”A Diagnosis of Product and Factor Market Coord- ination in the Bean Industry of Northeast Brazil" (unpublished Ph.D. dissertation, Michigan State University, Department of Agricultural Economics, 1968), p. 4. 91bid., pp. h ff. IORobock, op. cit., p. #9. _____[ relative to other amt for deveb amino. if “1 systen which do mot say. 2 H eff. Conversely than labor may I The questic estel a proble: "fill improve thq an (3) Are the The first '~’ 0 I. . ..e Getulio nations were these funct‘ 931‘. + lMid coeffi :‘ {q a .at-le I. 11 In of Y - View: ‘5’ for QXa YhT X1; XZ§ with. 3 +' 8 other C:Qu relative to other agricultural industries, it would seem to be a logical target for development activity if one wished to improve low incomes. Similarly, if there are improved alternatives to the present beef production system which do not displace the labor the present system uses, then one cannot say, a_ iori, that labor, under such an alternative, would be worse off. Conversely if labor intensive alternatives can be found and implemented, then labor may be better off. The questions become: (1) Is low productivity in the beef production system a problem? (2) Are there alternatives to the present system which will improve the productivity of land, labor and capital (especially animals)? and (3) Are the alternatives such to increase the use of labor? The first question may be answered by referring to a study sponsored by the Getulio Vargas Foundation, in which a series of CobbéDouglas production functions were fitted to data by regions and types of farm specialization. In these functions the dependent variable was total farm income. The estimated coefficients for the Northeast and cattle production are shown in Table I.11 11The regression equations used in the study to estimate the values in TABLE I are of the following general type: _ a1 a2 a3 an Y x1 XZ XB OOOXn where, for example: Y-Total farm income Xl-Area in artificial and natural pasture (Ha.) 12-Feed originating from agricultural and....... The entire equation refers to farms or ranches specialized in beef production in either Ceara or Pernambuco. There is a separate equation for each state. he. in artificia Pastures FM Originating 3&3 industrf Vmines. nedici Labor 3911 value of 1 CIOps Zeal value of 1 and "0'3 at 31W cows Q \ u 1;; mi“‘13 exc aniF-als 333mm 13.. TABLE I COBBéDOUGLAS ESTIMATES FOR ANIMAL PRODUCTION 1962 Variable (xi) Coefficient (A1) Ceara Pernambuco Area in artificial and natural .0007 .0141 pastures Feed originating from agricultural .1889 .0212 and industrial by-products Vaccines, medicines, and disinfectants .0786 .0808 Labor e 1204 e 0446 Total value of land in permanent .1637 .1050 crops Total value of buildings, equipment, .2734 .41## and work animals Brood cows -.0058 .0508 Sows -.0281 .0208 All animals except brood and work -.0266 .0651 animals Sources Projections of Supply and Demand of the Agricultural Products of Brazil, Vol. 1 (Getulio Vargas Foundation, Brazilian Institute of Economics, Center of Agricultural Studies, 1966), P. 126e 1 I these data testers will be * tins in Tablq 123:3. The est in the size of In general 105 or cattle hereases p053 .15th means fi‘iitional nut Wheat Pastu‘. e 2trier to 1m amiable . These data show that increasing the acreage of artificial or natural pasture will be less effective in increasing farm income than other alter- natives in Table I. This is true for both regions except for animals in Ceara. The estimated coefficients for Ceara also show that further increases in the size of the cattle herd will reduce farm income. In general these estimates tend to show that further increases in land or cattle will not increase farm income significantly relative to increases possible from cash crOps. The small increases obtainable from pasture means that additional units of pasture would yield very little additional nutrition. Therefore, one may infer that the productivity of current pasture methods is very low. Therefore, the most significant barrier to improved farm income with respect to cattle is the nutrition available. If the productivity is low, are there any more productive alternatives? One answer can be found by a comparison of estimates of actual productivity with what has been accomplished under controlled conditions. Estimates show'that cattle are marketed at four to five years of age: but, under feedlot conditions and high quality feed, cattle from the same general breed have been marketed at eighteen months of age.12 Animal scientists at the University of Ceara estimate that market 'time can be reduced to three years given improvements in nutrition and 12Joint interview conducted with the Faculty and their counterparts from the University of Arizona Project, Escola de Agronomia da Univer- sidade Federal de Ceara, Departmento de Zootecnia, Fortaleza, Ceara, Brazil, September, 1969. rope: herd mane m Mediate prc given to any br theise‘ives. it: higher quality atagement. T] 3319:: leveled factor to imp: “M. ‘eae OVBfi -felevark t '3‘} 1 Ice-44.. decis .eve; in the. °Pec:fi 1. 137%1 ”\‘3 it"? ‘k‘a‘§\ ‘niu .rfejrk? Sla am; st. * ‘fi. ll ‘-e proper herd management conditions.13 Proper nutrition and management is an immediate problem which must be solved before consideration can be given to any breeding program to improve the efficiency of the animals themselves. More efficient breeds of animals generally require more and higher quality feed than native breeds as well as better care through management. This latter point is brought up in anticipation of a criticism often leveled at traditional cattle production: that is the limiting factor to improving productivity is breeding. The need for improved nutrition implies that labor intensive home grown feeds, given the unfeasibility of a high supplement ration, are the solution, thus answering the third question. Objectives The overall objective of this study was to develop a conceptual framework to study the long run and short run consequences of develop- mental decisions concerning beef production at the farm or sub-industry level in the Northeast section of Brazil.lu Specific objectives were: 1. To develop a model for evaluating alternative means of modernizing beef production in selected areas of North— east Brazil. 131bid . 1“Industry is defined as all farms or ranches breeding and raising slaughter cattle in the seven states of the Northeast. Seven states refer to the nine states of the region without Bahia and Maranhao. 2. T0 fo A Gene ‘ For the put tecimique for c certain types 0 Savior of a bus extended period n .olputer I 2. To formulate and test a computerized simulation procedure for estimating the effects of different systems of beef production. 3. To determine the usefulness of this procedure and specify how it might be further developed into an Operational analytical tool for development planning. A General Review of Simulation,_lts Advantages, and Previous Research For the purposes of this thesis simulation is defined as "a numerical technique for conducting experiments on a digital computer, which involves certain types of mathematical and logical models that describe the be- havior of a business or economic system (or some component thereof) over extended periods of real time."15 Computer simulation is one of a number of quantitative techniques for analyzing systems models. Simulation has a number of special features which should be of interest to the economist. The technique allows the economist to perform dynamic experiments on an economic system over which there is complete control. Using the technique with the relevant data, 'the experimenter can test virtually any compatible hypothesis or policy alternative for its impact on the socio-economic environment.l6 Further- more, this technique has utility to the economist in that it can greatly reduce the number of necessary assumptions concerning the desirability of 15Richard E. Dawson, ”Simulation in the Social Sciences,” in Simulation in Social Science: Readings, ed. by Harold Guetzkow (Englewood Cliffs, Ne'u‘J' erse'y":'""'Prontice-Ha11, 1967), pp. 1-15. 16Thomas R. Webb, "A Systems Model for Market Development Planning: Northeast Brazil" (unpublished Ph.D. dissertation, Michigan State University, 1969), pp. 26 ff. .Lr. —. 5“ :— goals to be achi mi the shape 0 351mg a similar. are mine a ‘ iii: With othe .53): he can SE firm about th. the 13 ava 19?! 515971510 i a 10 goals to be achieved, the normative desirability of one policy over another, and the shape or even the existence of the social welfare function.17 Using a simulation model (simulator) the researcher can attempt to describe and analyze a larger number of important interrelationships on the system than with other techniques. The policy maker is helped by a simulator in that he can select from among the various alternatives since more informa- tion about the political, social and economic consequences of each alter- native is available. By using a simulator rather than other models many more dimensions of the problem, economic and non-economic, can be viewed simultaneously and the normative and non-normative consequences of alter- native solutions noted.18 Simulation in its broadest definition has been a part of man's environment since the beginning of history. Cave drawings, for instance, are simulations of the animals they represent. Simulation has been used in this century in the training of air crews. With the development of high speed computers new applications for the concept were found.19 The speed of modern computers allows complex problems to be handled in a :reasonable amount of time. The business world found application in wedding 'the computer with the concept of simulation in solving such widely diverse jproblems as aircraft and electronic circuit design on the one hand to making improvements in their marketing and distribution systems on the other. 17Thomas H. Naylor, "Policy Simulation Experiments with Macro- econometric Models: The State of the Art," American Journal of Agricultural Economics 52 (May, 1970), p. 2 . 1aGlenn L. Johnson, "Discussion of Macro Simulation Models,” American Journal of Agricultural Economics 52 (May, 1970), p. 288. 19Dawson, op. cit., pp. 1 ff. There are a ecmnist which I sense. economic sheet of realit interacts with :‘ 3319mm genera “595 by economi ”finely by othe .‘aletionships ; “Wine and Whereas 3 Sim 1150, budget S coatinations ‘ f°T each of 1 of Each colbfi ithgn ' 11 There are a number of research techniques which are available to the economist which closely approximate simulation. However, in the broader sense, economic models themselves are simulations since they describe some subset of reality and show how that subset behaves internally and how it interacts with its environment.20 Computer simulation, as distinct from the more general term, simulation, has one advantage over other techniques used by economists. While some relationships can be studied quite ade- quately by other techniques, computer simulation can study many more such relationships simultaneously. For example, budget studies are capable of analyzing and comparing two or three different combinations of resources, whereas a simulator can study literally hundreds of different combinations. Also, budget studies are capable of determining the profitability of these combinations to the firm: simulators can determine the same information for each of its hundreds of combinations besides the social ramifications of each combination. Nithin economic development there has been a growing interest in the use of computer simulation to study the problems of development. Most of this interest has been focused on the study of macro-level problems. In .a macro-simulation development study the area of interest is the whole economy or some large subsector of it. ‘Two good examples of macro-develop- ment simulators are Webb's Systems Model for Market Development Planning: Northeast Brazilzl and a study conducted by the Consortium for the Study ZOIbid. 21Webb, loc. cit. m m _ of Nigerian Ru: in his mod relationship b< niel‘s struct' iistribution a services, anj The Xige] 39591 construc 5':qu in that :59 Outputs 0 Studied are cm to the j Participate j 12 of Nigerian Rural Development (CSNRD).22 In his model Webb structured a simulation of the overall marketing relationship between the city of Recife and the surrounding region. The model's structure provides a detailed description of the production, distribution and consumption sectors which emphasizes the flows of goods, services, and incomes with the market.23 The Nigerian Model, as implied by its title, is a subset of a larger model constructed around the Nigerian rural economy. The model is a macro study in that no discrete production unit can be isolated in the model and the outputs of the model are aggregated values. Likewise the alternatives studied are general policy alternatives and do not reflect alternatives open to the individual farmer except to the degree that he chooses to participate in the alternative systems. The use of simulation in studying general policy alternatives is the crux of the neglected area in simulation studies of developing economies. Essentially all simulation work has been aimed at helping high level policy makers arrive at optimum decisions for development strategy. Very little work has been done at the farm level. Yet there is a clear need for this type of simulation study. Overall policy decisions must be implemented by individuals and agencies who must translate such decisions 22Glenn L. Johnson, et al., A Simulation Model of the Nigerian Rural Economy: Phase I - The Northern Nigerian Beef Industry, Report to the .Agency'for International Development (East Lansing: Michigan State University, April 26, 1968). 23Webb, op. cit., Abstract. into concrete 3 intm, devell tins rational allocate funds natives and Si :onjitions. 5 consideration Other areas a also need to finer which 3'33], Thu: 1391' leVel . '5“ he mus There h Lamina tc Elites Of I he of Stu: “it? a gen. 3:. his ‘ ens: .‘5 .0391 13 into concrete actions. But in the process of doing so these agencies must, in turn, develop their own policies in order to carry out their instruc- tions rationally. Such agencies, for example, must develop policies to allocate funds and personnel among individual farmers, production alter- natives and subsectors and do so in conformity to the local socio-economic conditions. Simultaneously these agencies must, or should, take into consideration the externalities their decisions will impose upon such other areas as labor, input and commodity markets, and consumers. They also need to understand and anticipate the constraints upon the individual farmer which limit his ability to behave in the manner desired by the agency. Thus there is clearly a need for simulation studies to aid the lower level policy maker in understanding the nature of the system about which he must make his decisions. There have been a number of studies aimed at applying the simulation technique to the agricultural firm. All of these have been confined to studies of more develoPed countries (MDCs). One good example of this 'type of study is a simulator constructed by Hinman and Hutton. In this study a general simulation model for four firms was constructed based on the generally accepted theory of firm behavior. The objective of this model is "to provide a means of studying management problems using the simulation approach.”24 The objective of studies such as this has been (exclusively to help the individual farm manager meet his individual goals. 2“H. R. Hinman and R. F. Hutton, A General Simulation Model for Farm Firms, USDA, Agricultural Economics Research, Vol. 22, no. 3 (July, 1970), P. 1e hese studies am 13mins what 29. to varying studies do not lizitations whj s‘fial’s which sh: In additi a'5'.’iC'lltln'a,l m were fundame ’5 =21 isolated 9‘93 Wheres? The rang ; . 7 "new?” eco ’5 AKTXCultu: $133313 8‘ 53.9. 3m {wizavion fie 320.4 fits 14 These studies are therefore basically farm management simulators designed to analyze what are clearly important problems for developed agriculture and to varying degrees, for underdeveloped agriculture as well. These studies do not approach the basic problem of structural and technological limitations which are imposed upon underdeveloped agriculture or suprafirm goals which should be taken into account by policy makers. In addition to these limitations on the applicability of present agricultural micro simulators to less developed countries (LDCs) there is a more fundamental limitation. One cannot justify the study of the firm as an isolated unit of production with given states of technology and given ownership. The range of ownership is much broader in LDCs than in MDCs. More developed economies are well established institutionally and the structure of agriculture is well integrated with the rest of the economy. While radical change is possible in agriculture the process at best can be con- sidered slow and evolutionary in nature. This is not so in LDCs. The political system in most LDCs is perpetually in a state of flux. As long as change in the political systems is possible then changes in the patterns of agricultural organization and ownership are also possible and, indeed, inevitable should a revolutionary change in the political system take place. But even barring political change, the concept of development planning implies directed change and this in turn implies that alternative organizational patterns as a policy goal are not ruled out. The model deve10ped in this thesis is an attempt to provide govern- ment agencies responsible for implementing the national policies (and to alesser degree production math: it‘ferent techn. should be note-i lojel. Althoug 20931! in the reference to 5; here attempts 1 tier sectors. The aPDI‘O.‘ 1213211593 by H Init 1311 15 a lesser degree the farmer) with the means of studying alternative production methods given any state of technology or the means of studying different technology and production alternatives simultaneously. It should be noted that size and method of ownership are independent of the model. Although the "entrepreneur” and the "firm" are referred to fre- quently in the study, these terms are for convenience and make no real reference to specific organizations or ownership. The model developed here attempts to consider the externalities imposed by the firm on certain other sectors. The approach used in this study is in some respects very close to that used by Hinman and Hutton. It is not the purpose of the model devel— oped in this thesis to study the day to day management of the enterprise. Instead it looks at the planning phase of the management function. Its purpose is to study consequences, over a period of years, of planning decisions on the enterprise's profits, on labor, on government costs and returns, and the supply of primary products (1.6. meat) through changes in the firm's contributions to total product, taxes, and employment. The model in this study borrowed an important component from the Nigerian Model. The Nigerian Model has built into it a good framework :for determining the demographic situation of the cattle herd. This component was used with slight modifications in the model to be presented here 0 The Research Technique Initially background research was done on the cattle industry of the var 'M‘ _ mneast. Th1: available at Hi: inflividuals at T infustry in the crop production aéiitional infon Superintendancy of Northeast Bra the author to de mm, and f 01 available. At the beg to Kort beast Br 1'21 *‘ 31%., one d8 16 Northeast. This consisted of searching out all known reference materials available at Michigan State University25 and conducting interviews with individuals at the University who had firsthand knowledge of the cattle industry in the Northeast,26 or had firsthand knowledge of complementary crop production relationships in the Northeast.27 At the same time additional information, not currently.available, was requested from the Superintendancy for Development of the Northeast (SUDENE) and the Bank of Northeast Brazil (BNB/ETENE). The information thus obtained allowed the author to define the boundaries of the system to be studied more sharply, and formulate additional data requirements not, at that time, available. At the beginning of September 1969 a three week field trip was taken 'to Northeast Brazil. Two cities were visited, Fortaleza and Recife, and a brief, one day tripuwas made to a number of cattle production enter- prises to gain some firsthand impressions about their modes of Operation. IIn Fortaleza the author was able to conduct interviews with knowledgeable persons from the Bank of Northeast Brazil ,‘28 the School of Agronomy of the 25These materials are listed with an asterisk in the Bibliography for the readers' convenience. 26Lawrence A. Johnson, Associate Professor of Dairy Science, East Lansing: Michigan State University, a series of personal communications throughout 1969. 27Thomas J. Manetsch, Associate Professor of Electrical Engineering and System Science, East Lansing: Michigan State University, a series of personal communications from September, l968—June, 1970. 28Eduardo Bezerra, and others, a series of personal communications at Banco de Nordeste do Brasil, S/A Departamento de Estudos Economicos de Nordeste, September, 1969. University of C mommy and £200.31 In a fro: U.S.A.I.D. 3533939,?" and 03 these inter! nation relati. in Published f The model - iiichigan The 33301 ”f Careful, r \ 29Facu1t 17 University of Ceara,29 the University of Arizona Project to the School of Agronomy,30 and the Rural Extension Service for the State of Ceara (ANCAR- CEARA).31 In Recife interviews were conducted with agricultural officials from U.S.A.I.DJ’2 FAO,33 the Superintendencia do Desenvolvimento do Nordeste,3u and the Institute of Agronomic Experiments (IPA).35 The purpose of these interviews was to gain a more detailed knowledge of cattle pro- duction relationships in the Northeast, fill in gaps in the data available in published form and confirm important data values already obtained. The model was constructed using Fortran computer language compatible with Michigan State University's Control Data Corporation 3600 Computer. The major problem with the data was that little of it was the result of careful, rigorous experimentation or controlled survey research. The 29Facu1tye e e ’ lOCe Cite 30Raymond Anderson and Charles Manes, Livestock Specialists, Univer- sity of Arizona Project to the School of Agronomy, University of Ceara, personal interviews, September, 1969. 3J-Clinton Saboia Valente, Engenheiro-Agronomo, Fortaleza, Ceara, Brasil: Service De Extensao Rural de Ceara, personal interview, September, 1969. 32Dr. Henry Mike Kilpatrick, Agronomist, U.S. Agency for International Development/Ibec Research Institute, and Dr. Roy Carvalheira Handerley, Chefe de Zootecnia, IPEANE, personal communications, September, 1969. Elbert B. Bowen, and others, Food and Agricultural Officer, Division of Agriculture and Rural Development, U.S.A.I.D., a series of personal com- munications, September, 1969. 33Cleveland James Allen, FAO Animal Production Officer, Recife, Brazil, personal communication, September, 1969. 3"Technical Staff, Superintendancy for Development of the Northeast, Division of Agriculture and Supply, Recife, Brazil, a series of personal interviews, September, 1969. 35Kilpatrick, loc. cit. | u l —I A" MI 111?," I!“ bulk of the da‘. of iifferent k1 interviews: no“ technique. Th have been extra One advanta nation concerni: collection tech: mzible with th1 here contained a Elsie: as well : :3 guide the no The basic 18 bulk of the data was well informed opinion or the results of case studies of different kinds of farming operations. Except for that obtained through interviews, none of the information was in the form required for the modeling technique. Thus, most of the parameters and variables used in the model have been extrapolated from the original data. One advantage in using this kind of data is that the range of infor- mation concerning the overall system is much broader than for other data collection techniques where the data to be collected are in a form com- patible with the proposed model. The data collected in the manner described here contained a good deal of information concerning the structure of the system as well as its input-output parameters and therefore could be used to guide the model builder in constructing the model. The basic approach used in the model was to measure the resulting revenue, costs, employment, and wage bill given any combination of proposed alternatives for herd management and land utilization. In this model the basic unit of study is a single, large hypothetical ranch constructed by aggregating the data concerning land, labor, and animals from several real ranches studied by Dickerman from.the University of Arizona Project.36 This procedure may be questionable, especially from au1«econometric point of view, but since the Dickerman Report provided no tunable input-output data, it is not an important objection. Information 36A1an B. Dickerman, The Economic Structure and Analysis of a Ranching System in Northeast Brazil (Portaleza, Ceara, Brazil: University of Arizona in cooperation with U.S.A.I.D. and the Federal University of Ceara, 1968) e 19 obtained from Dickerman included an estimate of the number of animals which ,should be combined with the appropriate amount of land in a realistic com- bination, and a determination of the number of people necessary to tend the postulated amount of land and animals. Actual input-output data were obtained from various other sources and compared, as closely as possible, to the production relationship described in Dickerman's report. In this paper this hypothetical ranch will be referred to as ”the firm.” Conceptually there are three broad categories of production functions under which the firm may be operating. These three are classed according to the three major climatic zones of the Northeast. These zones are (l) the dry interior or sgrtag, (2) the transition zone or este, and (3) the humid coastal belt. For each of these three categories there are two additional possi- bilities. One is a high capital-labor ratio and the other is a low capital- labor ratio. In reality there are not two such alternatives but an infi- nite number: the theoretical boundaries being zero to infinity. If hypothetically there are two possibilities with respect to capital/labor, 'then there are thus far six possible production functions. For every one of these six production functions there are two more possibilities. These are: (l) the firm is also specialized in the pro- Accounting Labor Extension 3 L ;. Lo C Run it Legend 8 M-D H-D-man-days Land in _ c ._ v , . Improved rTrdditional Land Use I Native Pasture {“‘“ Land in I - Native , .__/~__+Inni_fi Pasture _ I Land in ' y Unirrigated I Forage .44 Production ' It Land in Q- ~ ‘ - Palms \____Landfl L__.—...a Production | t ’1 f I ...__I Sorghum . . Production '9H '6“: land in E | I Perennial ____,« L Land 4 Cotton (- ) Production _f A W | Land in A Irrigated l , Forage £4 I i in Production 4» Annual Cotton I L Production I i in j Hoderniz ' Artificial \ In ts L Pasture 32 .930» Men mwcaveso mo Mommas on» no coweonsm a ma ends» mange cannedammd peze n\s ca. s\c fi\: fi\s fi\c n\s emspuem o>avcz com n\s N 0H n\s mm. e\c sconce Hessc< com ma. N mm j\: dfi\n copeoo cone s\: nm.ofl m 0H mm mm mm manages s\e pm.mm s OH cos mm one «ceases eopsmanuH n\s N.N n\s e m: H H emspnem e>deez uo>ounsH s\s n~.ma e\s m mm mm mm .nssusm Huaosmdsne ax: pm.na n.~a me mm mm mm essmnom .na\.ws .upa coca usseugua amoz mosses any wadsmgsmumnnuuwsssuam HessasH msapsnoso Hsdmmmm nuaodw memomondsdom Honda apnea esmcH mo>aaacnopad Hmdyomm mam mamoo H>Ha1" 1‘ Cattle in "men R 2'5 Traditional Transi- Management ~ g, “1 :11 °‘ E memnt 4‘“ tion i f o :90: 8 g :. a [m S - f r labor - ‘: Require I 1 n1 , L ension V1 - nen s an s ' e uire- E for Sh r animals 5 { its a Run :‘ o n w an:- ‘/ Animal f - ' L ani Sales 8: ‘5 Input “1:11: ccunu- .P a, Byprodu Require _ animals lated D e .. .C_ nents .. W3 pita], _ «I | , ~ , (animal) ' “gab ’ V f . ._ ‘ ‘ Input ' .-_.~ ,9 Reg"! Legend: -. ments P-Policy Variable no, H Flow % (Land) $"c“h F1" Accounting ' ‘. Accounting ‘ " r H-D-Man-Day - - . u s a, (:f labor)m \U ; II M 3L Tim-’1‘ 0131 918”“- Meat Unit Prices Net Cash Flow ble Nutrien Byproduct Unit Costs Returns to Capital Cash Cro Credit Cost Fifi-033.11% Rate Interest Total Value of Capital Inflation Herd Size Wage Rate Present Value Government Revenues and Costs 37 2. Sales policies with respect to modern, traditional, and transi- tional herds concerning: a. The sex ratio the policy maker wishes to maintain; b. The rate at which the herd size is to be changed: 0. The rate at which animals will be sold in accordance with modern or traditional sales policies.37 3. Nutritional modernization concerning hectares and combinations of: a. Sorghum b. Forages c. Improved native pasture d. Irrigated forages 9. Artificial pasture h. The management of the nutritional sector with respect to: a. The rate at which land is to be reallocated among the various alternatives; b. The rate at which the various alternatives are to be depreciated out and replaced; 0. The amount of storage necessary for the forages being planted for silage. The major parameters taken to be fixed with respect to decisions are: l. The prices of products and inputs. 2. The rates of inflation. 3. The costs of credit to the government assuming there is a degree of subsidy from this activity. 37There is only one sales policy with respect to the three herds con- cerning nutritional levels to balance out the herd demands for feed. 5. 6. 7. 38 The interest rate: both returns to alternative investments not in the cattle sector and the cost of debt to the entrepreneur. The wage rate. The costs of government services not absorbed by the entrepreneur. The technical input, labor, and output coefficients. The major parameters which may be influenced by the decision maker but not controlled by him as an on-going management decision are: l. 2. 3. The l. 2. 3. h. 5. 6. 7. 8. 9. 10. ll. 12. The capital-labor ratio: By products of beef production whose returns are yielded to the firm, i.e., a decision between cheese and milk. The initial matrix of land, modern and traditional nutrients, and cattle production; The level of irrigation initially being used. major outputs from the model in any time period are: Cash profits per year Cash returns to capital Levels of employment of labor Total capital accumulated The size of the herd The present value at time (t-O)‘ of all returns Total tax revenue Total government costs Private debt balance Heat supplies By product supplies Quantity of cash crops supplied CHAPTER III THE DETAILED MODEL Introduction The modelling technique used in this thesis is essentially an eco- nomic-engineering approach. An alternative technique to analyze a similar problem was used by Dixon in his comparative budget studies of production methods used in cattle breeding herds in the Argentine Pampa.38 The main difference between Dixon's procedure and this one is that Dixon was able to obtain enterprise combinations and input-output data from actual ranch records for both modern and traditional practices. All data were obtained from ranches with resources in a combination similar to the budgeted one. In the present study this was not possible: instead, data were only avail- able from diverse sources. In this case the input-output data had to be synthesized from these data. The attempt, then has been to meld informa- tion from diverse sources into a meaningful structural framework approxi- mating that of the beef enterprises found on ranches and farms in the Northeast. The engineering approach and its attendant building block concept seems to lend themselves well to the required flexibility of a model purported to be applicable in any given cattle enterprise in the region. Also it was discovered early in the modelling phase that there were many 38.1. J. G. Dixon, "An Economic Analysis of Range Improvement in the Cattle Breeding Area of Buenos Aires Province," (unpublished Ph.D. disser- tation, Michigan State University, 1969). 39 #0 points of correspondence in the way seemingly dissimilar computations could be performed. This discovery made possible greater simplification of the model structure than would otherwise have been possible. The modelling procedure used was to select from whatever sources were available the present combinations of land use, herd size, birth rates, and death rates. These data came mainly from a budget study by Alan B. Dicker- man who was at the time of the study a member of the University of Arizona Project inBrazil.39 Alternatives to the present production process were selected on the basis of (1) what was actually observed by the University of Ceara personnel on the more progressive ranches and (2) what were thought to be potentially useful innovations by these and other personnel.“0 The alternatives considered were restricted to the two major areas of study, nutrition and herd management. Physical input-output relationships for each alternative were then selected from whatever sources were available, hence the building block concept. In some cases input-output relation- ships for one alternative had to be synthesized from more than one data source. Where possible the input-output relationships once accepted were confirmed in full or in part from additional sources. In certain cases the differences between the accepted relationships and those to which they ‘were compared were sufficiently dissimilar to require serious reconstruction of the relationships to resolve the differences as much as possible. ”Dickerman, loc. cit. uoKilpatrick, Faculty and Counterparts of Escola de Agronomia da IJniversidade Federal de Ceara, Anderson, and Hanes. #1 Input and output prices were determined from a third group of sources. These plus fixed costs were then applied directly to the input-output relationships to derive the resulting revenue-cost relationships. The Demographic Subroutine: The Determination of Herd Size and Productivity This subroutine is essentially the same as that incorporated in the Nigerian Macro Model II.’+1 Here the entire subroutine will be described and dissinilarities will be noted. 1 BR-TABLIE(VALB,SMALLB,DIFFB,KF,TDNA) 2 DR-TABLIE(VALD,SMALLD,DIFFD,KD,TDNA) 3 RB-PF*BR/(PF‘+PM) 3l ERP-RB-DR 32 ERPF-RBADR*Ch4 Equations 1 and 2 are identical to Macro Model II. Equations 3 and 31 are equivalent to that found in Macro Model II and Equation 32 has been added. The purpose of these equations is to determine the initial unlagged birth and death rates where: BR: The percentage of females calving per year (%»of PF). DR: The percentage of the total population dying per year (as of PF+PM). TABLIE(VALB,SMALLB,DIFFB,KF,TDNA): A table look up function in which the computer essentially reads the BR value directly from a graph (see Figures VI and VII). “lJohnson, et al., op. cit., pp. 17 ff. #2 On the graph TDNA is the independent argument and BR the dependent argument.“2 RB: A dummy variable to redefine the births as a percentage of the total herd. PF: Population of females (K animals).43 PM: Population of males (K animals). ERP: The unlagged extraction ratio of males. This is the per- centage of excess births over deaths. ERPF: Unlagged extraction ratio of the females. 044: A parameter which allows for a differential death rate for the females.“5 It is important to note the value C44 takes and in which equation it is added. If used in equation 32 then it will have the net effect of shifting the total death curve upward, thus causing distortion. If used in equation 31, 04% would be less than one and would have the effect of shifting the whole curve downward, again causing distortion. There- :fore, caution should be exercised when using this parameter so that its value does not become too large (or too small). If DF is greater than DM “akobert H. Llewellyn, Ford 9: An Industrial D amics Simulator (Raleigh, North Carolina: Typing Service, 1985), pp. E23 ff. 43K indicates thousands. “uNote that 3 and 31 would be combined in the Nigerian Model to form the following equation: ERP-PF*BR/(PF+PM)-DR. This equation assumed that the death rates of the males and females are identical which may not be true. Therefore the model presented here separates the two and makes allrnflances for separate death rates as was suggested by further work on the Nigerian Model, Manetsch, loc. cit. 1*5Note that this parameter would be about 1.7 for the Nigerian Model. Ibid. BR -DR .6 .5 .3 .2 .1 “3 FIGURE VI Traditional Birth and Death Rate Curves BR DR if VALl (1) - .06 VAL3 (1) - .55 VALl (2) - .19 VAL3 (2) - .22 m1 (3') - .27 m3 (3) - .13 VALl (n) - .33 VAL3 (a) - .11 VAL3 (5) - .lo VAL3 (6) - .09 VAL3 (7) - .08 Birth.Rat3 L _1 A 660 102C 1360 1700 zdho 2360 2720 TDNA Pounde7aninal-year Source: A Simulation Model of the Nigerian Agricultural Economy: Phase I - The Northern Nigerian Beef Industry, pp. 20-21. 4# FIGURE VII Modern BirthéDeath Rate Curves BR-DR .6 I Birth,Rate BR DR .5 1h- an2 (1) - .08 VAL“ (l) - .50 VALZ (2) - .29 mm (2) - .17 .4 '- VALZ (u) - .51» MIA (4) - .08 v2.1.4 (5) - .07 VAL“ (6) ' e06 .3 ‘- VAIA (7) - .05 .2 -* .l -' L Death Rate 4_ 0 A 4 l n d L j 6é0 1020 1360 1700 2050 2380 2720 TDNA Pounds/ani-al-syoar Source: A Simulation Model of the Nigerian Agricultural Economy: Phase I - The Northern Nigerian Beef Industry, pp. 20-21. 45 by some large factor then a weighted average should be used in order to average out the potential distortion. A more accurate but more time consuming method would be to use separate death curves for males and females. Equations four through twelve determine the actual animals born per year. a Al#BR*PF 5 AlP-A1P+(DT/.3)*(AlquP) 6 ......9.,......... 10 A2-A2+(DT/BRDEL)*(Al-A2) ll BF-.5*A2 12 BM-BF A1: A variable which redefines the birth rate from a percentage into a rate of animals. This is an unlagged variable (K animals/year). AlP: The preposed lagged birth rate in which it is explicitly assumed that the current birth rate is a function of past events (K animals/year). Equations 6 through 9 determine the birth rate delay (BRDEL) of equation 10. In this case if the long run proposed birth variable is less than the unlagged current births then the birth rate delay will be increased and vice versa. A2: The actual lagged births in the current time period (K animals/year). 46 Equations eleven and twelve reflect the obvious fact that one half the animals born will be females (BF) and one half males (BM). DT is the basic time unit of the model one DT-.l of a year. The next series of equations compute the deaths of animals per year. 121 DRF=C44*DRu6 13 A3-PF*DRF 14 DFdDF+(DT/D3)*(A32DF) 15 A4-PM*DR 16 DM-DM+(DT/D4)*(A4ADM) DRF: The death rate of the females. A3: A dummy variable which redefines the unlagged female death rate into an unlagged rate of animals (K animals/year). DF: The lagged current time period death rate. As in previous equations this explicitly assumes that current death rates are a function of past events. A4: Same as A3 but for males. DM: Same as DF but for males. The following equations compute the extraction ratio for males and females and augment the level of males and females.“7 17 ER-ER+(DT/D5)*(ERP-ER) 171 ERF-ERF+(DT/D5)*(ERPF-ERF) uéNote: Equation 121 does not exist in the Nigerian Model. 47Note: Equation 171 and statement 172 do not exist in the Nigerian Model. u? 172 IF(KKK.EQ.O.) GO TO 20 18 PFbPF+DT*(BF-DF-SF-RFT) 19 PM-PM+DT*(BM-DM-SM-RMT) ER: ERF: IF(...: The lagged extraction ratio showing the current percentage of males which may be extracted from a stable herd without changing the herd size. Same as ER but for females. Population of females. Equation 18 is a level equation in which SF (rate females are sold), DF, BF, and EFT (rate females are transferred into or out of the herd) are rates used to calculate the net change from the year at current rates and the incremental addition for the herd is made (K animals). Population of the males, the argument is identical with that above. This statement is a switch function designed to omit equations 18 and 19 at this time for the transitional herd which will be calculated elsewhere. Equations 20, 21, and 22 compute the amount and gross returns from the herd by-product if the herd is specialized in beef or the main product if 'the herd is specialized in milk. 20 QM-PF*PFCA*YMA*TABLIE(VALS,1360.,1360.,l,TDNA) 21 QCH-QM*TCFFC 22 YC-QCH*PRC QM: PFCA: v“A: o . .g ‘ELIE-i :0 e3 QCH: '71.me U ads. “2 (.3 TU. N‘ QM: PFCA: YMA: TABLIE( . . .: QCH: TCFFC: YC: PRC: 48 The quantity of milk produced by the total herd (K kg/yr.). The proportion of females lactating (%). The average yield of milk per animal (kg). Another table look up function where an upper limit of milk is the dependent argument and TDNA is the independent argument. A conversion variable which may be redefined as cheese or left as milk, or some other milk product as desired (K kg/yr.). A technical coefficient to convert rates of milk into rates of some other milk product. Total revenue from QCH (in K NCR$/yr.). The price of QCH in NCR$/kg. The Cash Crop Subroutine_(Subroutine Crop) This subroutine is used in calculating revenues, costs, quantities of output, and labor inputs for cash crops. As currently used in associ- ation with perennial and annual cotton, it may be expanded to any number of cash crops. Equations 1, 2, 3, and 42 compute the total quantity of the crop produced per year, the gross return, tax return, and sharecr0ppers' share. 1 TOTC-HECT*YIELD 2 A-TOTC*PRICE*EXP(RFFP*TDT) 3 B-A*TAX 42 C-(A-B)*SHAR TOTC: HECT: The total quantity of the cash crop produced (K kg/yr.). The amount of land in the crop (K ha). Yield: A: PRICE: 49 The average yield of the crop (Kg/ha-yr.) This is a constant value and therefore is the long run average. Gross return from the crop (K NCR$/yr.). Price of the cr0p (NCR$/kg.). EXP(RFFP*TDT): This is an exponential function defined to be equivalent B: TAX: C: SHAR: to cit where RFFP is the rate of inflation of farm products (in general) and TDT is time defined in tenths of years. In every iteration of the model, TDT is augmented by DT, i.e. TDT-TDT+DT. Initially, TDT is set to zero. This formulation allows the model to compound the inflation through time. Note also by changing RFFP relative to other inflationary parameters, the model user is able to change classes and/or specific relative prices continu- ously over time. Tax revenue (K NCR$/yr.) The tax rate. Net return after taxes and reduction of sharecroppers' share accruing to the owner (X NCR$/yr.) The inverse of the sharecroppers' share, i.e., if the share- croppers' share is a% then SHAR-l-.a (Equations 5, 6, 7, and 8 compute the returns to the sharecroppers: the total cost to the entrepreneur, and the amount of employment generated per year from this cash crap. Labor is calculated in man-days per year (M-D/yr.) where a man-day is defined as an eight hour day. One man employed for one year is defined as able to generate 300 man-days of labor per year. PFC: 5 “BBB-(A-B)-C 6 D-HECT*RPR SO 7 PTCD-D*(CTOI*EXP(RFLTI*TDT)) 8 EMPLiD*WP+TOTC*WH NB3D: PTCD: D: RPR: CTOI: RFLTI: EMPL: UP: The sharecroppers' share (K NCR$/yr.) The costs incurred by the owner from replanting the crop. Note that the model as formulated assumes no cash coats are incurred K Ha/yr. The rate is 1). Costs of The rate The rate includes from harvesting or marketing the cash crop.“8 being replanted. the crop is replanted (for annual cotton this value inputs for planting per hectare (NCR$/ha.). of input inflation. of employment per year from the cash crop. This sharecropper labor as well as hired labor. (K MAD/yr.) Labor requirements for planting (MeD/ha.). Labor requirements for harvesting (MqD/ha.). "Subroutine Plast" This subroutine may be considered in one of two ways. First it may be regarded as the costs and employment possibilities of reestablishing the land production alternatives as they complete their useful life cycle and require replanting. This is the actual meaning of this subroutine. “BThis assumption, while simplifying, is unrealistic and should be C hanged e . . 1.. '| "NM 71 An alternat the amount perioi. The 58) l A-PE‘) 2 3-14.21 3 Cami A: B: CTOIP: C: CPR ‘0‘: 4 [huge 53.0,: 6mg. 51 An alternative consideration is to regard this subroutine as determining the amount of reinvestment necessary in the nutrient sector in any time period. The seven equations comprising this subroutine are as follows: 1 A-HECT*RPR*HP 2 B-HECT*RPR*CTOIP*EXP(RFLTI*TDT) 3 C-FEC'I‘*CPR A: A variable to calculate the labor requirement for replanting (K M-D/yr) - HECT, RPR, NP NH, RFLTI, TDT are identical to those in SUB- B: CTOIP: C: CPR: 1+ D-C*HH ROUTINE CROP. The cash costs of replanting the alternative under con- sideration (x NCR$/yr.) . The cost of inputs for replanting (NCR3/ha). A variable denoting the amount of land in the alternative which will be harvested in any one year (K Ha/yr.).49 The number of cuttings to be accomplished in any one year. For grazed crops this value will be zero. The yield of the crop will be dependent in part upon this value. 5 E-C*CTOIH*EXP(RFLTI*TDT) 6 PTC-PTC+(E+B) 7 EMPL-EMPL+A+D These four equations compute the labor requirement for harvesting, “9This reflects the necessity of cutting forage crops. costs 0f ha in total 1 This 1 ”This SUbrou talculating “P the land 11 RPTC 2 HERE 3 REX! EPIC: 33w . CTQI : REV}: ‘-L E" 52 costs of harvesting, total costs of operating the alternative for one year and total labor requirement for one year. The variables are: D: Labor requirements for harvesting (K MAD/yr.). E: Costs of harvesting (K NCR$/Nr.). CTOIH: Cost of harvesting inputs (NCR$/ha). PTC: Total cost to the entrepreneur (K NCR$/yr.). EMPL: Total employment (K M-D/yr.). "SubroutineBod" This is the final subroutine to be discussed in detail in this thesis. This subroutine consists of three equations and performs the function of calculating the initial costs, labor, and extension requirements of setting up the land alternatives. 11 RPTC-RHECT*CTOI*EXP(RFLTI*TDT) 2 REMP-RHECT‘NI 3 REXTC-RHECT*EXTC RPTC: The rate of addition to total private cost from establishment of this alternative (K NCR$/yr.). RHECT: The rate land is put into production of the alternative in question (K Ha/yr.). CTOI: The costs of establishing the alternative (NCR$/ha). REMP: The rate of addition to total employment from the establish- ment of this alternative (K M-D/yr.). WI: The labor requirements for establishing the alternative (Men/ha). , .. 1h REXTC : EXTC: With r total gover 1'Wellir‘ement COSts to th 53 REXTC: The rate of addition to the extension load from the establish- ment of the alternative (K M-D/yr.). EXTC: The per unit extension requirements for the alternative (M-D/ha) . With respect to extension and extension costs, it is assumed that the total government extension costs may be translated directly into labor requirements. Thus, once the labor requirements are determined, the total costs to the extension service may be determined directly from this value. Other Subroutines and Functions Two other subroutines are used in the model for the sole purpose of calculating distributed delays. These are the DELAY and DELDT subroutines. Both are essentially the same but the DELDT routine allows a much smaller delay with a given DT value.50 A table look up function, FUNCTION TABLIE, is also used in conjunction with the Demographic Subroutine.51 The Structural Equations Statements 992 through 997 set up a mechanism to control the policy decisions. The basic policy decision with respect to nutrition is setting a land target. That is, regardless of the combinations of alternatives selected, the decision maker must select the number of hectares he thinks reasonable to include in the modern sector. This is not to imply that 50For discussion see Llewellyn. 511bid. concept was tence of a i The nodal m: facing nutr; of head in Includ 313 relax“ differentia “St also 1 “other is “3“sz Yates are {5% atlas and palicieSe the hem 1 meme cc '11.- State \ 5.235: such a land target must be selected independent of the knowledge of (1) the combination and relative size of the alternatives or, (2) the size of the herd. However, in order to keep the model simple and allow the innova- tional process to be stopped at the appropriate time, the land target concept was selected. The land target is important because of the exis- tence of a large amount of land and a relatively small number of cattle.. The model must be capable of ceasing the innovative process to avoid pro- ducing nutrition at a level which is far above that necessary for the number of head in the model. Included in the nutritional decisions is the selection of combinations and relative weights of the alternatives to be used. Since there are differential costs and benefits to these alternatives, these decisions must also be made with respect to the overall system and the land target.52 .Another group of parameters are used to select the rates at which the land is transferred from the traditional to the modern sector. These rates are vital in determining the degree of stability of the model. If these ‘rates are too low, the cattle population, depending upon the sales pol- icies and rate of management modernization, will cause the herd to decline .absolutely over time. If too high, subject to the management and sales (policies, the yearly returns will decline for a variable period allowing -the herd to build up. This increase in herd size to the detriment of the :revenue component will result in any case from the lack of a buying func- 'tion. There is, therefore, an optimum rate of increase in herd size 'which will optimize the long run returns to the entrepreneur. Statements 994 through 997 make provision for stopping the transfer 52068 is a parameter determining the land target in K ha. of cattle ‘- as level c then the n: the transf transfer-mi rate Of gro her] will b ferred : thi the IOdern the lower 1 the traJlsfe lint. “it 399‘“ be exc t“insofar rd in one? to Statem S‘lbroutine it is calla 55 of cattle to the modern sector. This mechanism depends upon the desired TDN level of the modern herd. This is a "target" or parameter value. when the nutrition available exceeds a limit above the target level, then the transfer mechanism is started at a parametric rate. As the herd is transferred to the modern sector, and coupled with the sales policy, the rate of growth of TDN over and above the TDN target in the traditional herd will be slowed. Depending upon the rate at which the land is trans- ferred, this growth may continue until the land target is reached or until the modern herd is of sufficient size to insure that the total TDN per animal (in the model GTDNA), modern and traditional, becomes less than the lower limit around the desired TDN per animal (DTDN). At this point the transfer of animals will stop until the CTDNA again exceeds the upper limit. with proper management of the model only the first of these limits need be exceeded. If it is desired to keep the herd constant, then land transfer rates, animal transfer rates, and sales policies may be adjusted in order to just exhaust the traditional herd as the land target is met. Statements and equations 1000 through 1014 are used to call the crop subroutine as often as necessary to cover all cash crops. In this model it is called twice, for perennial cotton and annual cotton.- 1012 PTC-PTHC+PCTC 1013 PTR-PCTR+PTRC 1014 NB3-NB3C+NB3H PTR: The total revenues after taxes and share payment, accruing -to the entrepreneur (K NCR$/yr.). EULL{,m: VB} Bquati from the su' sional cash period sine is subject Equati< Pariod from 1100 mm 1110 TDNA; 1120 TDNS 1130 TDNI m0 may 1170 TDNH In equ TDNINP: YINp, 111sz Equati Equati 56 PCTR,PTRC: Same as PTR but for the perennial and annual cotton sectors respectively. UB3: Sharecroppers revenues after taxes (K NCR$/yr.).53 Equations 1012 through 101# are necessarily separate and distinct from the subroutines so that the model can be generalized to an n-dimen- sional cash crop sector. The cash crops must be calculated in every time period since the amount of land available to these cropping alternatives is subject to change with time. Equations 1100 through 1170 compute the total TDN available in any time period from the modern nutritional sector. 1100 TDNINPHYINP*HINP 1110 TDNAP-YAP*HAP 1120 TDNS-YS*HS 1130 TDNIP-YIP*HIP llho TDNFHYF*HF 1170 TDNM-TDNINPWTDNAPWTDNS+TDNIPtTDNF In equation 1100: TDNINP: The total TDN available from improved native pasture (x 1bs./yr.) . YINP: The yield of TDN of improved native pasture (lbs./ha). HINP: The total land in improved native pasture (x Ha). Equation 1110 refers to artificial pasture. Equation 1120 refers to sorghum. Equation 1130 refers to irrigated forages. Equation 1140 refers to non-irrigated forages. Equation 1170 computes the total TDN from the modern nutritional sector 53The C and H subscripts denote H33 for perennial and annual cotton respectively. 1220 TC,“ . 1222 P133: 1223 EXP 1224 SR: TML ‘J” m, 57 in K lbs./yr. (TDNM). Equations 1202 through 1224 compute the impact of supplemental feeding and veterinary care on costs and employment. 1202 TMUeTHT*060+CS*(THM+THD) 1204 PTC-PTO+TMU*CM*EXP(RFI*TDT) 1220 TCU-CPHT*THT+CPH*(THM+THD) 1222 PTC-PTG+TCU*CC*EXP(RFI*TDT) 1223 EMP-EMP+TCU*WDC 1224 SRCC-TCU*08*(CC*EXP(RFI*TDT)) TMU: Total animal doses administered of a predefined combination of drugs (K doses/yr.). THT, THM, THD: Total population of the traditional, modern, and transitional 060: C5: CM: TCU: CPHTI CPH: RFI: HDC: herds respectively.5u A parameter determining the percentage of animals treated in the traditional sector. A parameter determining the number of animals treated in the modern sector (%). The cost of veterinary services per aggregated dose in the base year. The total amount of supplemental feed used per year (K lbs./yr.). A parameter determining the pounds per head per year of supplement fed in the traditional sector. Same as CPHT but for the modern sector. The rate of inflation of inputs. The labor required to handle and distribute the supplemental 5“THT-PFT+PMT . - Eh“ C8: CC: The fol Sector, 1270 Tom 1230 Twp. 13.60 ENG- 1290 TDNT. TDNNP, TDNp, 58 feed (M-D/lh.).55 SRCC: Credit requirements for feeding supplemental feed. Because the government suffers a net loss in handling the credit, a credit subsidy is assumed.56 This is credit for one year duration (K NCR$/yr.). C8: A parameter reflecting the percentage of the total supplemental feed bill that is lost to the government. It is assumed in this model that the whole cost is financed through a short term loan. CC: The cost of the supplemental feed per pound in the base year (T-O). The following equations compute the TDN available from the traditional sector. 1270 TDNNP-YNP*RCON*HNP 1280 TDNP-YP*HP 1160 TDNC-YTC*HC 1290 TDNT-TDNNP+TDNP+TDNC TDNNP: Total available TDN from native pasture (K 1bs./yr.) TDNPI The same, but for palma. Palma, a crop of some small impact 55Note that labor requirements for veterinary services are assumed ‘to be absorbed in the cost since it is assumed that this labor is accomplished by veterinary personnel. 56This loss is the difference between the rate of interest and service charges on the one side and the rate of actual costs per unit time in real terms to the government. asafeeds latter. A‘ nutritive 5 its popula: east from a zero hec* TDNC: RCON: 121‘0 CRT- 1250 BOON CRT I RCO“. CE 59 as a feed source, is approximately 92 per cent water and 8 per cent dry matter. At one time there was serious consideration of palms as a nutritive source, especially in dry areas; however it seems to be losing its popularity and the data did not indicate any in the area of the North- east from which the data were taken. Palma was, therefore, included at a zero hectare level. TDNC: The same but from the residue from tree cotton (perennial). RCON: A variable reflecting the degree of over grazing. This is a function of pasture land and herd size, defined below. 1240 GRT-HNP/( (THT+THD+THM) *(TDNT/(TDNT+TDNM) ) ) 1250 RCON-RCON+DT*C6*(CRT-CRE) CRT: Grazing rate in the traditional sector (Ha/animal). RCONz An exponential average of the difference in traditional grazing rate versus an equilibrium grazing rates If GRT-CREhO, then the limit as time approaches infinity of the range condition is one. C6: A parameter reflecting the impact in any time period of the disequilibrium on the total value. From the definition of CRT and the corresponding independence of 'the herd innovations from pasture innovations, it is easy to see that <1RT¥f(THT) only. Therefore it becomes necessary in equation 1240 to make CRT-f(THT,TI~fl),TI{M, percentage of total nutrition from traditional sources). In the manner shown in thO, CRT will remain a reflection of actual grazing conditions in the traditional feed sector. It has been explicitly assumed that the modern pasture practices incorporated in the 60 package of innovations will reflect proper grazing rates for the modern feed sector. It can also be expected, as pressure is taken off the tra- ditional sector, that the yield of the native pasture will improve. CTDNA-(TDNT+TDNM) /( THT+THD+THM) TDNAT-CTDNA+TDNAKT 'I'DNAM+CTDNA-TDNAK The above three equations determine the availability of TDN for the three herd components. As previously explained, the transitional herd is managed according to modern requirements with traditional results. ' Therefore there needs to be only two availability variables for the three herds. CTDNA is the total TDN per animal available from the two sources of farm produced feed. TDNAKT is the TDN per animal arising from supplemental feeding in the traditional herd. TDNAK is the same but for the modern herd. These last two parameters are calculated in the initial phase of the model using the following two equations: (1) TDNAKT-CPHT-TDNK (2) ‘I'DNAK-CPH*'I'DNK where TDNK is a parameter whose value depends on the percentage value by weight of the TDN in the supplemental feed used. GTDNA,TDNAT, and TDNAM are in pounds per animal-year. Equations 1292 through 1340 compute the amount of land to be taken out of traditional land use. These are rates in K Ha/yr. 1292 RLLHC-ClO-HHC 1293 RLIC-C12*HC 13110 R11. 2119,12 91“ l 61 l3#0 RLLNP-Clh*HNP RLLHC: Rate land leaves annual cotton. RLLC: Rate land leaves perennial cotton. RLLNP: Rate land leaves native pasture. ClO,12,lh: Policy variables discussed previously concerning the rate at which these alternatives are taken out of their current use. Equations 1350 through 1371 allocate the land taken out of native pasture to the four alternatives according to the way in which the decision maker has prescribed. 1350 RLDF-C15*RLLNP 1360 RLDINP-Cl6*RLLNP 1370 RLDAP-Cl7*RLLNP 1371 RLDNS-C7*RLLNP {The dependent argument in each equation is the rate native pasture is to “be modernized for use as non-irrigated forages, improved native pasture, artificial pasture and sorghum, respectively.57 Statements 1390 through 1hh0 route the value of the variable repre- senting the amount of land removed from traditional use through a series [of'distributed delays. Statement 1390 will be analyzed as an example. 1390 CALL DELAY(RLLHC,RLAIP,CROIP,DEL1,DT,K1) This statement calls the SUBROUTINE DELAY. In the case shown, the 57Note that the parameters are such that c15+c16+c17+c7 nuet equal 1. shut to t the output 38011 62 input to the subroutine is RLLHC, the rate land leaves annual cotton, and the output is RLAIP, the rate land is added to irrigated forages. CROIP: A multi-dimensioned variable initialized at zero and con- taining the intermediate values throughout the delay of the land in the delay. DELl: The length of the delay in years. K1: The order of the delay. The Subroutine DELDT is also used in much the same manner: as a sub- stitute for DELAY in statements 1420 and 1440 where the value of the delay falls below A, where A-2*DT*K.58 In this case the delay will become un- stable. The delays, as used here, represent the time difference between the introduction and completion of work on an innovation. A completed inno- ‘vation may mean either that a new innovational phase may begin or that the innovation is prepared to yield its intended services. Allowing a new innovational phase to begin does not imply a mere reallocation of resources used in the innovational process: but that the completion of the preceding innovation was a prerequisite to the second innovation regardless of resources available for the completion of the innovation. The use of a distributed delay allows one to predetermine the nature of the completion of the innovation process, given the nature of the input. Equations 1442 through 1640 compute the incremental additions and subtractions to land that are forthcoming from the distributed delays. 58Llewellyn, op. cit. a previous 1452 H3131 1460 HIE 1562 Hc-a .. remov 63 These equations are among the few level equations in the model. with respect to rate equations, the results of the model must be considered to be instantaneous reports of the condition of the system at the time and not the reports of level conditions resulting from activities over a previous time period. 1442 HHC-HHC-DT*RLAIP 1460 HIP-HIP+DT*RLAIP 1562 HC-HC-DT*RLLC 1600 EN P-HNP-DT'“RLI.N P 1610 HF-HF+DT+R8AF 1620 HINP-HINPWDT*RLAINP 1630 HAP-HAP+DT*RLAAP 1640 HS-HS+DT*(RLAS+RLLP+RLLC) Where I RLAIP: RLAF: RLAINP: RLAAP: RLAS: Equations The output of delay 1390. This is the rate of addition of land to irrigated forages (K Ha/yr.). The output of delay 1410. The rate land is added to forage (non-irrigated) production. The output from delay 1420. The rate land is added to native pasture. The output from delay 1630. The rate land is added to artificial pasture. From delay 1440. Rate land is added to sorghum. 1442, 1562, and 1600 represent traditional land use as it 1432reduced by innovative decisions. Note that annual cotton is assumed tc> be removed from production after the expiration of the delays. The great amount of work necessary to establish irrigation which implies a significantly greater time period when it would not be feasible to try and establish forage production and therefore cotton production will continue. {There will, however, be a time period for the establishment of forage when it will be necessary to stop production of cotton. This is not reflected in the model and some corresponding double counting results. It will probably be feasible in later models create a delay reflecting this estab- lishment time lag. For tree cotton and native pasture this reduction in .acreage takes place when the land is fed into the delay. The land must 'be taken out of current production once innovation has started. There is a degree of distortion here also, since grazing may continue on native pasture during a portion of the innovative period. The magnitude, however, is not now known but probably will not be large. Equation 1640 refers to the addition of land to sorghum. Only one source of this addition flows through a delay. The others from palma (and tree cotton are assumed to be fed directly into sorghum production. fPhis is possible from these cropped lands since fencing and brushing need :not be accomplished. However, this does ignore the time needed to estab- lish sorghum production. This, however, is a seasonal variable and is jprobably not relevant. Cotton will probably be taken immediately out of production after it is harvested in November. It would not be feasible 'to establish sorghum until at least January, the normal planting time. 'Therefore no conflict should result. Statements and equations 1652 through 1681 deal exclusively with zanimal transfers from traditional to modern production and the demography 4!” of the tr: The a tinue to p | and sales demagraphi birth rate: on the tra; “5 VII) b1 cormspcmd Ieat Yields on an inter Statem hm‘ NOte 19‘!le n01; iii‘f‘v‘l‘ent p 65 of the transitional herd. The animals which are transferred from the traditional to the modern sector must be fed into a distributed delay to account for the time nec- essary for these animals to come up to the full potential of the modern herd. During the time that they are in the delay, however, they will con- tinue to produce at some intermediate level. This mouse that births, deaths, and sales will result. It was, therefore, necessary to include a separate TPRFT063 b. EMF) EMF0 Ce “>W‘Bo d. CP>GPO 9e 0MS>OMS° f. PROFIT> PROFITO‘S“ 2. INFERIOR CLASS a. TPRFTerPRFTO 63This should read: "The variable TPRFT for the last reporting peri- od of the simulation run with the alternative included must be greater than the variable TPRFT for the last reporting period in the ”Zero Level" run. 6’"'Forthis variable the values must be compared in each reporting period.~ 83 b. PROFIT>O c. QJJIS>¢21HSo The inferior class, by absence of employment and government returns, reflects the possibility that a modernizing alternative which passes these criteria only will make government and/Or labor worse off in terms Of re- turns to the two sectors. This means that some of the goals of society, on the one hadn, represented by government and labor, and the entrepreneur on the other, may partially conflict. In order for a particular alternative to be profitable, labor may be forced to accept a smaller share of the return, and/Or government may have to subsidize the modernization process more heavily: but the process will yield government no additional revenues for the added expenditure. Tests of Alternatives The options to be tested were more or less arbitrarily selected. Guide- lines to selecting the Options were: (1) Previous experience with the model with respect to which options would probably have the highest payoff. (2) The parameter values, both yields and costs, with respect to Opportunity costs, cash costs, and expected returns.65 (3) A desire to test all the important options, and (4) A desire not to greatly upset the system of pro- duction as would have happened if, for example, all native pasture had been placed in the production of sorghum (implying a much higher labor require- ment). (5) An intuitive understanding of a realistic combination of options 650pportunity costs refer to both cash crops and nutrition foregone. given the major constraint of the model.66 In sum constructing the model was a long and difficult process: many problems were encountered and often these could be traced to imprOper combinations of alternatives. These improprieties had to be corrected before further structural problems could be identified. By this process, a reasonable combination of options began to emerge, and Option A resulted. After the testing of the options was completed a fault in the accounting mechanism for labor was found. The model was increasing the rate of employ- ment reported in the read out by a large non-constant factor. This was also reflected in the returns to labor but did not influence the returns to the farmer. Therefore the effects of the various Options on labor and the wage bill are marginally certain at best and cannot be assumed to be a valid comparison. Each subsequent Option was the result of analyzing the preceding option and determining the probable cause for its failure to meet the criteria. This is the fundamental reason for the clear dichotomy between the first three Options and the last four which will be discussed later. The options are defined below. 1. Option A The land target, i.e., the amount of land to be placed in modern production alternatives, was set at 1000 hectares. Annual cotton land is reallocated at 10 per cent per year of the re- maining land in annual cotton. Perennial cotton is reallocated 66This constraint is that there are too few animals to take advan- tage of any relatively great increase in nutrition. 2. 3. 5. 6. 85 at 5 per cent per year. Native pasture is reallocated at 2 per cent per year. Cattle are to be brought under modern herd manage- ment at the rate of 70 per cent per year of animals remaining under traditional management. Ten per cent of all land reallo- cated from native pasture will go to sorghum: the rest is evenly divided among the other alternatives for native pasture. Option B This Option has the same features as Option A except the land target has been reduced to 500 hectares. Option C The land target is set at 100 hectares. All innovations are to come from native pasture only. Native pasture is reallocated at 2 per cent per year with 10 per cent devoted to sorghum. Cattle are transferred to modern management at #0 per cent per year. Option D The same features as Option C are found here except the land target has been increased to 500 hectares. Option E The same features as Option C characterize Option E except the land target is 250 hectares. Option F The same features as are present in Option E characterize Option F except cattle are transferred at 60 per cent per year. 86 7. Option C The land target is set at 1000 hectares: cattle are transferred at #0 per cent per year: and native pasture is reallocated at .5 per cent per year with 10 per cent going to sorghum. The results of these tests show that all seven options fail the superior class of criteria. Options A through C fail the inferior class as well. Options D through G pass the inferior test. The major prdblem ‘with all tests was that yearly profits tend to drop Off for a period of years early in the innovative process. This may indicate that either credit terms are too high or the repayment period too short to make repay- ments profitable or that sales policy has been mismanaged or that cost jparameters are sufficiently unrealistic to be meaningful. Assuming sales jpolicy and parameter values to be correct, then terms or length of credit appear to be a significant barrier to modernization. Both types of credit «difficulties when approached from an income point of view appear roughly identical, since both are subtractions from yearly net income. when one considers the fact that the response of the cattle enterprise to moderniza- -tion is rather slight in the early years and only slowly becomes more jpreductive, then credit that is only partially subsidized can be a serious Zlimit on the rate of modernization. There are at least three partial solutions to this problem: (1) More fully subsidize credit by reducing :interest rates further: (2) Extend the repayment period: (3) Defer repay- :ment for a given number of years until the cattle enterprise becomes more productive. The dichotomy between the sets of Options A-C and D43 hinged on the 87 present value of future returns. The slow gain in productivity of the cattle enterprise means that high returns to modernization will be deferred for a greater number of years than would be the case if much faster increases in productivity were possible. Since the early returns in the simulation cycle are weighted more heavily than later returns in calculating present ‘value, one could expect that heavy expenditures coupled with small gains in :revenues would greatly outweigh even higher net profits in later years given a.high discount rate. The tests in fact show that while net profits are .higher in later years for all options as compared with the "zero level” :run, the present value of Options A-C are less than the zero level run. From this present value problem it would seem that there exists a Zlimit on the rate of innovation. In order to keep early returns from (dropping so low that the present value of future returns is less than the jpresent value of doing nothing, expenditures must be reduced. This then ‘translates into limiting the rate of modernization in the absence of sig- Irificant changes in inflation and/Or interest rates. An associated problem is the effect on revenue from reallocating «cash crop land to feed producing alternatives. The loss of revenue re- sulting from this land use does not seem to be made up by improved pro- 2:: .eape» one me see: one as uepeodusd endomeouea one an cease-enema mam mom on» no meansdmep one as envenomed one seam cache soaeedssam one no use one es commend meanness» one woman» one poeamen enemas: era .3. as: .. ET: 5: .. 8a.: are» sees 35 .mosHe> :a: as weeeenawe one uemaep HH<€ «and Nam.o man .0 $06 mama mead mu Nesta «moi. mam; $84. «as; send {him mduq<> no 9. a. o. o. a. are; BHhomm a. o. o. o. o. D..- Hmo.m mam man .a mate mass. malts man .a ma ~:mm.n Toma wnmw.m Nissan mums.m Tana mums.m mdoemm «mufi Nani «an .m NHN.m madam «mmfi mad Bzfio mum OH mDZHz nmezmzmmo2H mmmamxH mania cam.HH om:.NH 85.: omn.HH ommeHH are afifis so; can as are so no Foreman manage mndem<> Bamabo 91 made mma.m 93.0 mu main «mail mmoé. man .5 Nani. Emma 3mm.~ tamed... saw...” saw; :34 BHhomm Hzmo mmm 0H mam @205 gm .ENA a nmné 9&6 9:6 mmmé 3:6 «and m: > and“. «urea mafia «swim mumnfi «:3 .m gm mmwfi «3% «moi Namfi Nana «mmtm oar :5 0% 3H came: 8.1m.” came—H ES. mafia; has 0N0 are so me massages E555.” EaHm<> E0 92 936 .- muoew mo macaw «mix. «an 4. «Mai. Hamma .- a. .- scrim A also; new A sumo...” HHhomm Sufi ERA EMA mam mafia mmuam 9.36 nmmé mum .m m: H> mqmfla TEA «dim «lama minim Adoamm Sea mum on was Efizmmufi message 0. o. o- o. t o. .0 «Eden mad oawtfld 2:9 or mxo manzo m¢QH= hm: mm: no mommUZHza mmamzdmdm aHm abmabo 93 mood «also «on; 93.0 «and $06 so Nani. «~36 ~am.o N34. mail Ema also; seed smote Seems 0. .EN .m Hana Hanan g was mead mum .m mama mam .m mass mmmé man a: «dim TEA ~ama.a «new .m «dim «arm .n «no: gm usages—co 1|H> @348 Nana «swim «min as: «sea was 80.: 89.: came—H 83.3” 8025” ES. oxm mmm mw m2”? “2» um: um» magmas, mend mesa made mains 956 mu «min «and. use! «as; «mmaa ~m~.s mania «meg. mame o. .0 swim are” samem aam.a: $3.0.” Ewen : EQJMH aria $0..” m BHhomm Hzmo mum om mbnm QHHZHZHMUZH mmmsmz mqmda ~:Mm .m «Amoen «:mm .31 «:36 «Amara m:mwtm $3.“ m ameDo 95 prices, estimates of yields, and the grazing rate. With respect to prices, the parameters defined as rates of inflation are actually rates of inflation only if all parameters so defined have the same value. If given different values then they represent differential price trends. From the results it can be seen that the values of these parameters will greatly influence the outcome of the model and therefore special attention must be paid to determining their actual values in future research. When these parameters are used as experimental variables to test for the effects of changing market conditions, the experimenter will have to be very cautious about the results because of the impact these 'variables have on the system. The initial prices of beef and cotton also seem to have a significant effect on the outcome of the simulation. These factors will have to be given carefully investigated initial values. Since any error in these parameters will be enlarged as the simulation process proceeds, these factors are as important as the inflation parameters. Yield estimates of land alternatives seem also to be important but of slightly less importance than the initial prices. Their importance stems from their being direct inputs to the cattle herd. The equilibrium grazing rate parameter seems to have a rather destabi- lizing effect on the system, which is because of its significant influence on the herd nutrition in the long run. This parameter must be more fully investigated in future research not only with respect to its value but also 'with respect to its actual influence in a real situation on the nutrition available to the herd. 96 The other parameters shown in the tables have some destabilizing influences on the model but these influences are, in general, less than those previously mentioned. Nevertheless close attention should be paid to estimating these values also. Suggested Areas of Future Investigations On the basis of the results three areas are suggested for more inten- sive investigation. First, the equilibrium grazing rate is very important for overall system stability. This variable will affect the future results of the model by the degree to which the native pasture improves as the herd pres- sure is reduced. Also, there is a real question as to how much the nutri- tive value of native pasture can be expected to improve through this alone. {The range condition mechanism is on the whole suspect because of these effects. The second area concerns the yields to be expected from the modern lland alternatives. These yields will obviously have a great deal of effect Iipon the costs and herd sizes permissible within the system, and, therefore, -bhe total capital accumulation and cash profit. Since yields will vary :Erom area to area and will be a function of the expenditures on inputs, curreful investigation must be made with respect to any locality to ascer- ‘bain.the interrelationship between these two factors, yields and input costs. A third area is relative price trends. In the long run the directions idhe trends take, both relatively and absolutely, will determine the decision 97 to innovate or not to innovate. In the long run these variables are not only difficult to estimate, but there is also the possibility that the level of aggregation will have a distorting effect as the factors and commodities within any category group assume different relative prices. TPhis implies that substitutions will take place thus altering the pro- duction function as it is described in this model. A fourth area that needs more intensive investigation concerns the 'birth-death curves around which this model is built. These curves are as initially reported in the Nigerian Macro Model II.69 For lack of more appropriate data to this situation, the curves are assumed a fair .approximation of what one could expect in the Brazilian case. The simi- ‘larity of climatic conditions between Northern Nigeria and the gertag land the similarity of native breed types is significant. However, these curves should be more closely investigated to determine what the values should actually be. There are several quantitative methods which could 'be used. The main problem is to estimate values for the birth-death <3urves against TDN intake under actual conditions found in the Northeast. frhe inclusion of the transitional herd in this model also introduces a cluestion as to the wisdom of differentiating the demography of this herd. ESimilarly, there arises a question with respect to the wisdom of assuming iindependence between feed production and herd management. It is probably (correct to say that the effects of improved nutrition and improved manage- rnent yield higher gains when considered together than when considered —_' 69Johnson, et al., op. cit. 98 separately-—a multiplicative effect versus an additive effect because the two separate seetors are expected to complement each other. With careful model management this difficulty can be controlled within limits. How- ever, this assumed independence also produces a degree of distortion by the limitations placed on the land innovational process. In summary, while the above mentioned areas of further research are the most important with respect to the stability and usefulness of the model, it must not be overlooked that essentially all the parameters are subject to revision. There are six different conditions under which the model will have to be parameterized. Consequently, an entire library of parameters will have to be obtained to use the model generally. Also there are ranges of conditions among these six that will necessitate the discovery of good approximations for the parameters. CHAPTER V SUMMARY AND CONCLUSIONS Introduction The overall objective was to develop a conceptual framework to study the long run and short run consequences of deve10pmental decisions con- cerning beef production at the farm or sub-industry level in the North- east section of Brazil. Specific objectives were: 1. To develop a model for evaluating alternative means of modern- izing beef production in selected areas of Northeast Brazil. 2. To formulate and test a computerized simulation procedure for estimating the effect of different systems of beef production. 3. To determine the usefulness of this procedure and specify how it might be further develOped into an operational analytical tool for development planning. ‘The model is of use chiefly in making evaluations of alternative combinations of land use and herd management practices at the firm level. This evaluation need not be restricted to the m but may be used in virtually any region of the Northeast assuming the parameters for the region are available. It may be used also in evaluating an aggregation of several ranches using similar production techniques and located in the same geographical area. At higher levels, however, care must be taken to 99 lOO ascertain the impact the ranches have on the local assembly market and the local factor market since the model assumes a perfectly competitive production sector with respect to impact on market prices. The model is designed to report the rate of cash flow, employment, supplies, government revenue and expenditures, and levels of land by land utilization and herd size. The model itself does not optimize but merely tells the user what the effects of his decisions will be on the cattle production enterprise of any given ranch or set of ranches. The model cannot determine prices or supply responses over time. The price at the farm level for beef products and inputs must be given. ‘The model can predict the consequences of allocation decisions on the firm itself given changing relative prices over time. As the model is now con- stituted the rates of price changes must remain constant. However, with minor redesign the pricing mechanism can be made to perform in any desired manner so long as the performance preferred is independent of the model itself. In other words price must remain an exogenous variable if major :redesign is to be avoided. What Has Been Accomplished This study has produced a method potentially useful in calculating —the consequences of a given set of production relationships and the (corresponding decisions on the part of management concerning these rela- txionships. The consequences of this interaction between management and iihe production relationships may be observed for both short and long periods. 101 Specifically the study has created what is potentially a decision- making aid for an entrepreneur, credit lending agency, extension agency or development planner with respect to cattle production. It is a mecha- nism to calculate costs, revenues and employment levels in each of a given number of years given specific decisions and specific economic trends in the form of prices. A second potential use is in determining the impact on the individual enterprises of government intervention in the production and marketing sys- tems. Specifically it will measure the impact of successful government efforts at changing subsidies, manipulating prices in the commodity or factor markets. An example of its usefulness is the tentative conclusion discussed later concerning the inadequacy of the system for a rapid rate of profitable modernization. A most important area of research in the long run, given 'the present state of knowledge, will be to study the economic impact of successful research aimed at adapting foreign technology to the region, [especially biological technology. There is a third potential use of the model as a mechanism to study ‘the economic impact of natural or other uncontrollable phenomena on indi- ‘vidual enterprises. Examples of this are: (l) a sudden decrease in herd asize due to natural disaster; (2) undesirable price trends in one or more :factors without a corresponding change in commodity prices. What More Needs To Be Accomplished There is a high priority on the sales function in future work in the Inodel. As it is presently constructed, with price given, the sales function 102 attempts to study the technological decisions the entrepreneur must entertain in order to establish a balance between present and future returns when making decisions concerning whether or not to sell, and the amount and kind to sell. This basic approach to the marketing problem should be retained. However, the present sales function is too divorced from the market forces to be considered realistic. Another criticism of the current sales function is that it is very inflexible. To be useful it should allow greater free- dom concerning the assumptions of selling behavior than is now possible. {The sales function should be extensively redesigned to take into account this needed flexibility and to conform more significantly with reality. A second area of further work is in the market for feeder calves, cattle, and breeding stock. Although this market is neglected in this model, it is fairly well developed in the Northeast. The reason for neglect is -that the simple demographic subroutine is incapable of handling the large Ichanges in age composition which would result from a rational cattle buying (decision. The result of neglecting this factor market for cattle is that ‘both the supply function and the profitable rate of innovation are distorted. As more and more nutrition becomes available, a larger herd size is needed ‘to utilize the additional nutrition. In the absence of the ability to buy new animals, the only other alternative is to retain additional females :from the mature animals ready for sale. Since the potential productivity :of'the herd is lagged, fewer animals will be sold, thereby reducing revenues. But, the costs of innovations are increasing total cost. Alternatively, if neither reduction in revenues nor large increases in costs are desirable, 103 then the entrepreneur must select a properly reduced rate of innovation in order to keep supplies flowing at a constant rate while allowing the herd to grow. Herd size sales will increase more slowly, but there will be no wasted nutritional resources and no interim reduction in revenues. In- creased costs, however, are unavoidable, but may be partially mitigated by increased revenues. 0n the other hand, with capability of buying additional animals, gains can be realized more quickly and herd size can increase much more rapidly. The constraint of the rate and magnitude, then, is shifted to availability and terms of credit and the effect on the present value in the long run. One solution to the inadequacy of the Demographic Subroutine to handle this function might be to use a more detailed subroutine such as that sug- gested in the Nigerian Macro Model 11.70 The suggested subroutine is capable of handling animals by age groups and determining separate nutri- 'tion requirements, births and deaths. In sum, it keeps age composition separate and explicit allowing birth, death and nutritional calculations to be made on any age group. Conclusions Concerning the Industry’ Since the data cannot be considered very reliable because of the method of generation, these conclusions are by necessity general and superficial in nature. Before meaningful conclusions can be drawn, more rigorously generated data will have to be obtained. Parameters will then 10“ have to be more carefully estimated and the model subject to more extensive testing with respect to the actual input-output relationships. This means that rigorously constructed surveys will have to be made of ”typical" ranching operations in both the modern and traditional sectors both to generate the data and to compare with the results of the model. The results of the tests of the alternatives suggest that it may not ‘be profitable to transfer cotton land to cattle production. It further suggests that, in the Sgrtgg at least, cattle may not be able, in the short run, to compete with cash crops in general. In the long run this may not be true because of the greater ability of the cattle sector to survive periods of stress. This result is tentative; further study should be done on this using more reliable data. The ability to import feeder cattle from outside the region appears important in taking advantage of gains in the quantity of feed realized :from improved land use. This point has been discussed in the section, "What More Needs To Be Accomplished," in this chapter. In the long run most of the Options studied in the model resulted in.higher profits to the rancher: however, in the short run profits were jless than those in the traditional sector during the innovational process. frhis means that yearly cash returns less yearly cash expenses were lower .in the modern sector than in the traditional sector during the innovational jperiod. Assuming the correctness of the parameters during the innovation jperiod and a few succeeding years, either revenues were too low or costs 105 too high, Vises-vis the traditional sector. If returns were too low, then the absence of additional sources of cattle could explain this lower profit. However, if costs were too high, then the credit system may be at fault. In the process of innovation the greater the rate of innovation the less the short run profit rate. However, since all costs to the process were paid with the use of long run credit, it is not unrealistic to believe that the repayment period was too short. To lengthen the repayment period would require increased interest rates for the credit agency to break even due to inflation. Increased interest rates may either increase long run total cost or offer no change to the rancher. Therefore it may be necessary to increase the length of the repayment period with no change in interest rates in order to mitigate either partially or fully the reduced profits during 'the innovational process. Another alternative might be a direct cash sub- sidy to the rancher rather than the more indirect credit approach. This .analysis is meaningful whether or not inflation rates remain substantial. Given an actual wage rate of NCR$3.00 per day in 1968 prices, the inodern sector will increase the amount of labor utilized, but the modern cattle production sector appears to be a less lucrative alternative than «cash crops for labor. This may or may not be true since experience with the IBrazilian Cotton Simulation Model suggests the opposite.71 The difference zappears to hinge upon the estimation of relative prices for cattle and cotton sand their trends over time. However, in the long run there appears to be as substantial increase in employment, assuming no labor saving technology. _—i 711'. J. Manetsch, Com uter Simulation Anal sis of a Proggam for Mod- ernizi Cotton Production in Northeast Brazil East Lansing: Michigan State University, Division of Engineering Research, August, 1968). 106 Sugggsted Areas of Further Investigation Using This Basic Model In addition to the uses for which this model was intended, there are at least two other areas which may be investigated using a properly param- eterized model of this basic structure. The first area is to study the long run consequences of the drought upon the enterprise and labor. A comparison between one zero level run and another zero level run in which in a given year a drastic reduction in herd size and nutrition available can be made to determine these conse- quences. Two factors may be determined. First, what is the magnitude of -the economic loss to the region in terms of lost opportunity for profits over a period of twenty or thirty years? Second, what policies could be established to reduce the adverse effect of the drought on the entrepreneur, his laborers, and the region? The second high priority area of study is the consequences of changes in the terms of long term and short term credit. The model, by itself, will not determine the influence upon invest- ‘ment of alternative credit arrangements. The influence on the returns -to the investor, however, will be reported by the model and investment :rate changes can be made from that point to determine the maximum likely :rate of investment under the various credit alternatives which would be :feasible for long run profitability. Impgrtant Areas for Future Research There are at least four areas for future research which the model suggests as profitable. l. 2. 3. 1+. 107 Forage Grasses. The model results have suggested that the yields obtained from land in forage crops are very impor- tant for the overall model because of their effects on herd size and density. The overall inadequacy of the available forage grasses, especially perennials, is well known by the plant scientists in the Northeast (11). Continued research in productivity and viability in droughty conditions for grasses is urged by the author. Credit System. Research should be conducted in the adequacy and consequences of the current credit system and the terms of credit with respect to cattle production. Input markets for beef production. The innovational process as described in the model will require a great deal of in- puts. If the input markets are unable to handle such a demand, serious adverse effects on the process can be fore- seen. The net result of an inadequate market is a slowing of the innovational process in the aggregate by greatly increasing the price of necessary inputs. This suggestion does not come as a result of the model itself but because of what the model indirectly implies. Alternative methods of increasing the herd size. As TDN becomes available, larger herds can be supported on the same amount of land. The question then becomes: Is it more profitable, both in the short term and the long term, 108 to allow the herd to increase naturally or to increase herd size by purchasing feeder cattle? If natural in- crease is selected, then output (and thus revenues) will be reduced in the short term as females are withheld from sale for use as breeding animals. Natural increase could have unfortunate long term consequences if, for example, the discount rate remained high due to high risk; then the present value of an increased herd size in the distant future might not be sufficiently large to outweigh the re- duced revenues in the short run. Alternatively, if feeder cattle are bought to increase the size of the herd, then there is no short term reduction in revenue. There is, however, an added cash cost which will be reflected in the profit and loss statement and thus, eventually, in the present value of the increased herd. A third alternative would be to buy breeding stock. An added cash cost in the short run would not be repeated in the long run. The analysis is similar to that for a natural increase in herd size. To determine which of these alternatives, or what combination of choices is appropriate will be very impor- tant to the profitability of the firm and, consequently, to the future of beef production in the Northeast. Al A2 A1? ALPH ANETF ANETM BETA BF BM BR C1 CZ C3 Ch C5 (36 (37 APPENDIX I GLOSSARY Total live births/year (kilo animals/year) Live births (kilo animals/year) An exponential Average of A1 A correction factor to kee the sales of males in line with that of females (K animals? The net change of females in the transitional herd (K animals /yr. ) The net change of males in the transitional herd (K animals/yr.) A correction factor to keep the sales of females in line with the available nutrition Female births per year (K animals/yr.) Male births per year (K animals/yr.) Live birth rate-~proportion of all females calving per year Decision variable, a scale factor influencing the rate of sales of traditional females Decision variable, determining the rate that the sales of traditional males in equilibrium is to be changed Decision variable, a scale factor influencing the sales policy of modern females Decision variable, a scale factor influencing the sales policy of modern males Decision variable, the number of standard doses of medicine per animal in the modern herd A parameter that determines the extent in influence of grazing rate upon the range condition A decision variable, the percentage of land taken from native pasture which goes to sorghum 109 08 C9 010 C11 012 013 014 015 016 Cl? C20 C21 CZH 025 C26 CAI’ CBS 110 A parameter, the rate of subsidy of credit A decision variable, the percentage of forage production to be stored per year Decision variable, the rate at which land is taken out of annual cotton production (%/yr.) Decision variable, the percentage of remaining land in perennial cotton to be removed in the coming year Decision variable, the percentage of land remaining in palma to be removed in the next year Decision variable, determining the rate at which the storage deficit is to be made up Decision variable, the rate at which land is transferred from native pasture (%/yr.) Decision variable, the percentage of land taken from native pasture which goes to forage (%/yr.) Decision variable, the percentage of land taken from native pasture which goes to improved native pasture (%/yr.) Decision variable, the percentage of land taken from native pasture which goes to artificial pasture (%/yr.) A parameter determining the initial value of land per hectare (NCR$/ha.) Rate of repayment of debt Parameter: the minimum rate at which traditional females must be sold (%/yr.) Parameter: the minimum rate at which transitional females must be sold. Parameter: the minimum rate at which modern females must be sold. The desired amount of storage needed for silage (K# TDN/yr.) Input costs for building storage facilities (NCR$/#TDN) CC CF CFCM CFCT CHP CIAP CIC CIF CIHC CIINP CIIP CINP CIP CIS CMDAP CMDF CMDINI’ CMIP CMS CPH CPHT CS 111 Cost of supplemental feed (NCR$/¥TDN) Number of cuttings of forages per year Input costs for routine herd management (NCR$/animal) Input costs for routine herd management (modern herd) Cost of harvesting Palma (NCR$/¥TDN) Cost of inputs to artificial pasture (NCR$/ha) Cost of inputs to perennial cotton Cost of inputs to forages Cost of inputs to annual cotton Cost of inputs to improved native pasture Cost of inputs to irrigated forages Number of cuttings of improved native pasture per year (a dummy parameter) Costs of inputs to palma Cost of inputs to sorghum Cost of medicine/standard dose Cost of inputs to initially establish artificial pasture Cost of inputs to initially establish forages Cost of inputs to initially establish improved native pasture Cost of inputs to initially establish irrigated forages Cost of inputs to initially establish sorghum Pounds of supplement per head (bulk material) Decision variable, amount of supplement per animal (# sup- plement/animal) Cuttings of sorghum per year CTDN CUIP Cul CuZ cum C50 C60 C62 C6” C66 C68 C69 (370 (371 D1, D2 D3: D4. and D5, DELI, . . . , DEL12 DIST 112 See DTDN Number of cuttings of irrigated forages per year Decision variable, a scale factor influencing sales policy of transitional females Decision variable, a scale factor influencing sales policy of transitional males A parameter to differentiate the death rate of females from that of males An initialization parameter to balance the traditional herd with the native pasture Decision variable, the number of standard doses of medicine per animal in the traditional herd Decision variable which scales the value of ALPHl by a constant percentage Decision variable which scales the value of ALPHZ by a constant percentage Decision variable which scales the value of ALPH3 by a constant percentage Decision variable, the total number of modern land to be ob- tained by the innovational process Decision variable to determine what percentage of the tradi- tional females must be left to switch to a modern sales policy Decision variable, the rate at which females are brought under modern management Decision variable, the rate at which males are brought under modern management (Years) Time delays in determining birth rates, death rates, etc. (Years) Time delays in determining the gestation lag in the innovational process The rate of depreciation of capital DT DTDNA EDA? EDF EDINP EIP GRE GTDNA HC HHC HINP 113 The basic unit of time in the model 1/10 of one year A control parameter, the desired rate of nutrition/animal Extension labor units per pound of storage facilities Extension labor units per ha. of artificial pasture established (nan-days/Ha.) Extension labor units per ha. of forages established Extension labor units per ha. of improved native pasture established (nan-days/ha.) Extension labor units per animal transferred to the modern sector. Extension labor units per ha. of irrigated forages established Extension labor units per ha. of sorghum established Equilibrium grazing rate (animals/ha.) Total TDN available per animal per year lb./animal-yr.) Rate of employment (K nan-days/yr.) Rate of employment on land innovations (K n—d/yr.) Extraction ratio for males, the percentage of males which may be removed from the herd without decreasing the size of the herd. . Same as ER except for females Total rate of extension work required for the model (K n-d/yr.) Hectares of land in artificial pasture (Kha) Ha. of land in perennial cotton Ha. of land in forages Ha. of land in annual cotton Ha. of land in improved native pasture HIP HNP MONTH NYEAR PA PER PKG PRC QCH RC RF 114 Ha. of land in irrigated forages Ha. of land in native pasture Number of iterations per simulation cycle Number of cycles per simulation Average value of animals in the herd (NCR$/animal) Farm price of beef per head Farm price of cotton (NCR$/kg.) Percentage of the total females lactating Farm price of annual cotton Price of milk (cheese) NCR$/kg.) Number of females in the herd (K animals) Number of males in the herd (K animals) Level of current debt (K NCR$) Total accumulated costs (K NCR$) Total accumulated revenue (K NCR$) Total quantity of milk produced (K kg. /yr.) Rate artificial pasture is replanted (%) Rate annual cotton is replanted Rate forages are replanted Rate of inflation of cattle prices (%) Rate of inflation of farm product prices General rate of inflation Rate of inflation of input prices Rate of inflation of wages RFLND RHC RINT RINP RIP RCON RFIH RMAM SF‘ SM SC ISCS 31K} SR: SRI> STOR I?! TFCFTN) TTHTK 115 Rate of inflation of land prices Rate annual cotton is replanted - Rate of commercial interest Rate of replant of improved native pasture Rate of replant of irrigated forages Rate of replant of sorghum Range condition Rate animals are removed from the transitional herd (females) (K animals/yr.) Rate males are removed from the transitional herd (K animals/yr.) Rate of repayment of long term credit (%) Sales of females (K animals/yr.) Sales of males (K animals/yr.) Sharecroppers' share of the perennial cotton (%) Vaqueiros' share of the cattle sector (7;) Sharecroppers' share of the annual cotton (%) Sex Ratio (PM/PF) Sex ratio ”desired" Total storage capacity (KiTDN/yr.) Total medicine used (K doses/yr.) Tax revenue (K NCR$) Total herd (PF+PM) Technical coefficient changing milk into cheese TDN value from supplemental feed (%) TDNAK VALCAP VALC VALLND KGB NCO WCS WHC WHHC 116 TDN value per animal of supplemental feed (#/animal) Value of the capital accumulated through the innovational process (K NCR$/yr.) Value of the total stock of capital (K NCR$) Value of the land (market value) (K NCR$) Labor requirements to build storage facilities for silage (man-days/#TDN) Labor requirements for tending the cattle (m—d/animal) Total labor available in the herd management sector (K m-d/yr.) Labor requirements for planting forages Labor requirements for harvesting artificial pasture Labor requirements for harvesting perennial cotton Labor requirements for harvesting forages Labor requirements for harvesting annual cotton NOchU ---11|l?flllllli|-i!l.I.hCOe n GU ®.uuu |-.audu ‘ - - can mu cunnU can "U i-em.aus m CC CC” FZHUU 1 - ‘ -- v-1 - 00.ezacn- .m.o~m.xm.m.oaw.me~ e \m.onw. Cam.m.on~e.Mll: . n..mr .axm.m. ofimam.xm\m.o~m.Axm.m.o~mvo~ .XU\m. ofiu.m .xm. mocfimaofi.xm\m.ofim.axm.m. Camaofi.xu ‘ « \Am. odmxmaoflem.o~m .«edu.OI~.k*.xonmsmzsrxm.*&F&*.xo.*kzw*eXF.*UIJJ&*.XO.*u*XmuH\\\*wEHF*.X~.quaou 0C" a*¥*. m X-.*W*XHH.*~*XH—.*I*X-.*w*x~_.*u*xafi.*m*xu~.*Q*Xu~.*uvm quc*m*xda.*<*X0\\\*¥.b.~ mMJm<~U<> zoaawk~au*xmuc*I.0.u mme<—Q<>u Camoxmm L E mw III { M OF* WW T. fin i T. +b- I P M VA AU IC D WE. F. Mu P T. P. ill|!|LflILF nu nu We N I .h P «f huedoaaxummtw.uKIMhumhhah mhhhkuwDOElUPNKnmkxbmthhkk«Econ 00 AOCuCUU .AOCNF..0.dJ<>..OVMJ<>..0.NJ<>.AOC~J<> ZOHmZmzao illillliiulllllill!|l?||i|{il?OC§N-z OH H IFZCE -QN.NH¥. e u flay Ill!" KNCoHDmEII l- doccccuucz .!IstumU3 KCCOHFUUB H C O "EUUL: «CCCouEUB ’FuthmJ<>§ilili||!|ii l ouflanme<> :Imo.uahsenq>IIIIIIIIIII-I oo.n.o.eJ<> ho.MhmPNU<> mocuacch<> Imumamso44> FueuANCQJ<> mouahch<>.!I.Ilillli:Iz 122' omwmbW~NJ<> moonAFCMJ<> DDoHMDVMU<> ~.u.m.m4<> “Hemleem4§> I fluouamvm44> mmmmawamJ<>IIIIIIIIIIIII mm.ua~.m4<> emuueeem4<> deouanvNJ<> mo.u.~.NJ<> mmwmwe~a44> em.u.m.an<> oa.uamc_44> oo.u..._n<> mo. u yzoe euuuueuzouuoe summer a avenue a.ueuumoe m oux c¢1n0> ocwwuké> I :I--...IIII .:I-.||II.IIIIII -. I III ..II..-II III .1. . QCOHuSicxr ccmfiucz> oocmmmnu~> oCCRRHOZH> oCOVHUI> oCC6¢~Hu> occn n U> oCUn¢~HO<> ofiuCE .mHanE 0 annual: oCHHOHOS LéIitlII -. ?.Il:i ovufl2~fl3 .ouuUIQB oOHHuQB oQMHUQE omnfldfli ofiflnuH? I; .IIi. ..ll -lIIliII II: I eCrl wWIIQF :.. oOQuflZHCwE owmukcwl oEWNQ2 .fl n 2 H CO Ecaflu$>§0 .CHFCF .--II .I -I II I - 211.. .| LL2>NC§> AUII+UI+DZI ..x. CRLNCZ Iv. _d> FhC\F§Q H FCU @JUC\FC*A{¥.F ZCFCHZCFU kualFIkuFEQ axkcnm+.~a\.fiaselkueun CUU*w&0\QZIuFIF Uk>luhknfl2> .ukksnc1+aZIV\UImte> u00\FZCFCuuFF quCkIScheancer. FY42CFIF2 e" n 2 H CO 2UU0H§§§O cCHFCF 5-.- I. - -.I.--. l I I - I - ..I.... -- Fr.¢>HC...;> AUII+CI+OZI v* cabu0244q> FUC\F§Q H FCU @JUC\FC*A(XVF ZCFCHZCFU FuQIFIknFEu A.koom+.fic\.fistk1kueun CUU*MGG\QZINFIF UF>IuFFuQZ> rukk*nUI+GZIV\UInCF> uOO\FZCFCuuFF V.<.£CFI:.. H WZCF uaI * uH> u nazce nHI.WInu> u n u QZHZCF Imsa+umr3nwm3 UHFD+BFUQHOFQ UFUQ+UIFQHUFQ .nuuc.orc.orusm.uren.roa a. rmskDEUhpmmkpmseIuoau.4446 caoa .nmua.oc.onsma “FOP eueun.uaxe.umm3.sanctum.aua.xe some 4440 .onaeu .cucbu coca ca cc Im.mwru bonehv ahazhkzou boo.cco.oeelzeec*mo. quoeoaua ooo.roo.mooxzoec*mo.«1420+caua. .OHGHU -IIII-::.. Mmmweu .ouCHU mEZHFZCU e00 CF CC “on no moubfiu Moummu UCOeNGHU moououu mnzaezouI moo.moo.moo.mournze.ua eowmbwmeok: oOnQEw t:MhMMbu. Crag C¢- cusp CFHF cans ccap sac“ V~C~ «"0— (col FCC {CO U00 600 W00 «00 0440 *-FC I OI u- 01 (curl U440 * he I or u UI aqua .! II: u~<4a * F0 + QHI n ma: C(s~.I na<4a * yo I or: n or: cool mb MaebeoaDH.deDsm00M.mddflrmZCJQPFDUwDI 4U40I Cod" .mx.so.m4wo.n<4wo 44<4mc 44<4m0 44 UZCF cud". OI * fl> 020% can” I; OZI # ZOUQ * QZ> u OZZCF Cra— n.ka*~u&vnxw*UUV*GU*DUFuUUGm «any I|I||U03*DUFI aim mamfillllfII- AFCF*~qu&XM*UU*DUF + UFQ u UFO Rama .| I.r II-II .llIlilI I I“OIF+EIF~*IQU+FIF*FIQUHDUF CAN“ Ak0h*uuavuxw*§U*DZF+ UkunUka «CNL .OIk+ZIk~*mu+oou*kIhuDSk acma C§fl+ Chan CIF «(aw F20 + Fun PIP CCK~ sen + sun ",2Ie coma : 01.40 * LIC I OI N 01 CCKL- I- 0440 * PC I UI u 01 aqua I 0~<40 * F0 + 0~I u 0~I Cue“ 0~<40 * he I UII u UII mesa hDMehbemFDHendunemOflw.mKJUPVECUQVFDuwD 44<4mo 4440 Crew IIIIIIIIIhhM.wFDIHthee4m0.u2aomu.nza<4a.nzlo4aceo4wo 44<4mo 4440 crow IIIIIIIIIIIIIIII:IIeHywho.a4mo .u_omu.na<4a.ur44as ><4wo 44 n 020k C(H~ 0I * 0> u 020k can" I I- I I III . OZI a. 200.0 * 02> n 0270a. Crow A“FCF*~u0v0xw*UuV*GU*3UFNUUGm snap I,IIIIIU03*DUFH02m-rn«~ ak0k*~uaa0xw*uu*30k + Uk0 n Uk0 new" III II I- IIII. ,AQIF+ZIFC*IQU+kIP*kIQUuDUF cum“ I AFOF*Huav0xw*EU*DZH+ UH0HUP0 QCN~ l I .OIF+ZI%~*mU+OQU*FIFHDEP «CNLII CSO+ €00" CIF “(my sin + eunamIeIeIccmwz.IIIiI $20 + 200 u 51% CC- AFCF*H00v0XU*UUU*F~U*AQCFMIF0nu caraI-Iizl- cu0\020uccm rac~ II |I-f|. -II -- .| I- I II - F00\FZOHFCW newfi AEFMZ<+§Q§OIFFZCV*FC+CEQHC§0 ~0T~ IIIIIIII II -[ I - | I- “uFUZ<+§.cuan.o:uc.oumuue.a<$>.c.oe<.om<.e4moar.m oa<.oaam.oso.oar.eo.oeza.oeuu.ona<.om<.omo.ozo.mo.oo.mo.mo.do.o..oa emrmOWw.qu>Hmwbbmmp.Cm®.H4K>soau.cfinecu0eshzme.bzm.aumVoozmo 44u0<04<> m0*OF0 n 200 .ow0k0a.cew4.0F0.u~ .00h0lk20+20k00*ko+0k0uoe0 IIIIII-|III.|I!I.I;I - II I . . CQGHCQC C00n00k0 I: III-IIIII IIFCQHICFICC C0On00k0 II , K0m4.0F 00 0000.00.00000_ F20+0F0*000uc00 -I- 0F0*FZ~0uF20 NW Akok*40000xm*03*$05w +m00+04000+04o00+0ZH000+uo00+ 0—00n20k0 1 I .I. I III II I -ImwfiwlunbmflafiflowfiafivfibmfimuwaIwumsgam- .- .Fake04o00.04DX0.04DM0.040_3 .04owe04050._00.04040v 00m 4440 I !iI-IIbFDF.0Z~OQG.02—OX0.QZ~Dw0.0Z~D~EP0Z~DmcuznOEU.~00.0ZHD40. 00m 4440 .FOF .0000.uoxa.uoma.uo_3 .uowouozuouuaeuc4av 00m 4440 IIIIIIIIIIIIIIIIII .FCF .m0flemX0emwaemn3.mmem20.anaewo4avoom 4440 “FOP .0~00.0~x0.0~m0.0-3 .0uw.0_50.~00.01440. 00m 4440 .I- , IIII II III. .cIn-ne-wna, ocu040X0 oCu04000 coca oe co mos*m”0*.aoemIeuqocunqoma rum*m~o*.aoemI»noPUQQ.kIUOo%Uuk0F.H4E>ok40uaom44>okd4.Pm4ok4m0am .N WI+uI+Q~ I+Q.m..pmo..omm.«44>.»am.+zn.+uu.F420».»zmuwum.oozon4040Icqow -IIao.u.zzm.zum.zu>.zuaa.zluo.zuuuu».z<£>.zquua.m4<>.z¢4.5mq.24m0am .N a¢¢Uo2mfluoEuflfluo£mfloium 2~4.ZQGM.ZQC.EQE.FO.ZFEG.ZFu&.ZQ~4.ZN4ozuOoZSOomOo¢O.MC.KOo"Ooooocfi IIIIIIEIQN.momw‘fid4>wm.oomoo‘CmmoN44>wan.Eifloiua.24zorozzmtzumVUOEMC 4440 CFON.II 130 EZ>*AEZm+Eum.+CZ>*.CS¢+CEW~+FE>#APEm+FumvnmEC (VCR couEEm noCRWUwZEmPLFIITI HOCCCCOCCCmflEéQawCOCCCCCoom402200k~ mCCOII .cuczm ..c.m4.ozm.u_ II I; II Ii.cu+zm e.c.w4.kzm.u~ dzumnzum.~zum.p4.zum.u~ :IIIecummcum,"cum.+4.oum.u4 «Fumupum.~pum.p4.hum.u~ quun+wpwEEnxfiawu55mIrran. .z<20ko\.zommaosum.Cumm .Q44Q.&~I.QZI.O&Q.Ezu.kz&.m3.k~uoaa.m3.aa.hzm.0144Q.&4I.UII.CUmm a.kua.zuu.u44>.a40kma.Uka.kum.u44a.QZ~I.UI.zoua.zam.hqm.k.Hofiszaa UIC *IUCIMIUCF ak**.0ua+hzha+.~..\Faucaa + huaak a hung» 044>\ Fauoaa ufl40kua Ukalaaukwuoaa. thFkDZ4nflbuxwk02444>+n4044>+04Q*GIF+E4Q*EIF+F4Q*FIFuU44> F4Q u C4Q 24QIHIMMUIWIE4fl u 540 F4Q * 0&3 + F40 u F40 UIOEu+002u+EQ§u+OEun0§u 000mlzkxm I 500 I a» n 00 . , 000+ aka u an mm? + «GB + "as n H? hbmhnkbkkwunaknwwwflswpnzuu«m3 m0: I EQSM + aim Ifiazm 0&0 J,A00kl uaov u «m3 .muml.~.* AUQHI 0&0. u 00a UHF+IQXF+UQXFHGF X4F * 0&0 u 00% uuuflPukkklmflkbEEmHEuMHbém%Dumwhim?»um0«0&0 20u0*aOIF+SIkv+F0k0¥kIF+0FQu0FO 2003*.CIF+§IF.+%0c3*ka+a2uuQEu 00FO+0FQNUFQ I ow + F n F EU>+00>+F0> + u Iuaka ! I! I I‘I II|| D II- Illll III Ii-.IIIIII.IIIII-lIIIIIIIITIIIIIIIIIILIIIII. III.III-.5k~I (can .CUOO cunn ~4nn (dun Erna «(KAI “(mm (~00 FCGO Cofln (den (run Cqun CufiuI: ccfim firfim (ran (awn CCHoI nPOPolimkaquEUka35FkflkkFDbkflRPkUPkUFkIPLFm44flIud4U ! I: AFOFoQZonFQoQ4IUVQ2:30.Hx zo_mzm2~o A>EEDQI.M .uuHoI.4442m.44>.w~4m4k Zohkuzau czu IIII; .2uDFua Ill.ltu.i ll!‘l!.llulf| III .111" III I I IIIIII. 'I III. II I). I I1! kam * FUEIG u 0kxua r .I I : I In} * FUmIH u uéuo .FOF*~F4ua.axu* ”Oku * Fumla n whom 44 .FOF.UF&&.Uwaa.aEma.“3.0Fxm .HOFU.HF4ua.Fqua0 com wzflkacaaam czu .II- I ZQDFuG .C+4v + 402m n 405m b - am + my + UFQ UFQ AFDF*~F4ua.axm* IHOFU * 0 I3 * 0 n 0&0 * Fowl u nFahthUuapnxu*n~Okv:fiInanI*.kuml u a3*a&a*kum1 u C. ( ll m * hum: n UFOF are» I.uuua I yok0k.4azw .QUFa.m.omm34 .u.av aoau mzfikaoamsm 02w mszfikzcu r uzzhkzcuIM fl 1 I HRWCXU‘II‘Q mF A20+uflv\fifi*u0uflfi F .420».C¥.nuuHQ.C4445m.C44>va4mVMH4G4FHOE novm44>.no.c44>.novd44> ZO~mZuEHC .440.ur IIIIIiIIIIIIIEi. - aw.kflawoma.uQOFKMXhimwumkU>w03fiwIUCb0uu0FkKZ>PKUuflwm44>rd4wn4meCCN m.d4.uamoao.amoko.kzc.kuaoQ~4.N4ouo.zo.mo.4oomo.meo~C.CM.Ouu_C.C44~ 42¢.O44>.u¥.muunO.m4442m.m44>.mmofu.ufl.4zok.5m.umvoohwo wzukDOUGDU CZU 202FUD AyvflkDOOU + QFDOQ n QFZCU 0 II I-I ,0a4 u ZHQ H «4m0\AUE4IZ~flv +UUI4 u A—VQFDOQU a~v GFDOQU u 0K4 M .— u ~ a CC .kO~.P404u\UZ~fl u Zwa FG— .« u 1 m CC I.--III.IIIIIIIIICoDIIn DFDCIQ ako*ayvk404u.\.k0~.k404u*4mc u H4uc auvakzcau ZO~mZuE~C .y.ko.kc~.4mc.aF:cau.ak30a.azaawho4uc wzhkaooaam 02m ZGDPwa IInyVUFDOQUHUFDCfl 0m4nZHa b m4MC\a0m4lZ~&v+0m4na«VQPDOGU a~0QFDOQUn0E4 ¥.~uu F CC Ozwauzna akcwpy~F44wQ mznkDCOmDm CZIu ZUDFEO a~.44>+uu~C\auu~C*.filkaqo4ul$DC.*..hv44>la~+~044>vuwu4m4k -yuH,n~+¥.Cwo_vu~ ... 2133 i134 II.I|II |I|l lullll' I..Il'l.llllllllln I .. .. ||.II‘I Ill I I IIi. ccomofimomo230. CFE ZQDLIIwQ 0&&*10Cn0> 0uu0F*ECnIUC .~w~4m<»*qz>* 0uu0k*ECuIOC n4ZOF.fi..OOMH.oOCmH.m44>.mn4m4k*4§>*40ufl*u0ufc 6k§alzmlzc12mv*ko+zauza .kualumluCIum.*ko+uanua It: I .I;IIcm-0+IC®nc.Ou.xyxqu .uamluuamv*nu0\hcv+uamuuou flamlaaw.*.mC\kC.+aun0u AEQI¢<.*.¢C\FC.+ECHEC I 00*5an44 auolm4.*amC\ka+uaan uao*uaum4 00*440uu00 I umuim «4*Uouua .m no I.I-IIIIIII-: I .III--- - I. I---I-I;-I-I.-IIIIIII IIIIII II.III.II I II-IuMuUF*zcquc an “420%."..oomu.coon“.m44>.w~4m4k*42>*40ua*uauzc cn II3+Ea12m4£cnzm0*»o+2auza o" .kuatumluOIum.*ko+uauua a” cm-b% CflhbmbmoyxywhrInrw .uawluaam.*.m0\hcv+uamuuou Hr“ .amnflaw~*.mc\kc0+awuau r" IIIII IIIIII- I - II- I I-- III III-II!.IIIIIIII-I .zcu¢<.*.¢o\»c.+zcusc cu -HwII:-IIIII- I -IIIIII.IIII-I-IIIIII I- -II I IIIII: : ao*zau<< m4, - I q, .uonm<.*.mo\ko.+uouuo q. 1 uaowuaumq. m—IIIIIIIIII-I. a0*440uuac no" II umuzquH-I «4*r.uum .4 I II- I I I- - I I I- I I IIII-II- IIIIRn-MEISIIIEamxIEquuquS «Cu4w00m 0 .DP-OFCOIC —On4woam b III: I I I I I I - I I - I I- I IIIIIIIIIIIIIIoIKIKIIquISDI-:--oI II--- -- .a—4laq.*.m.\ho.+a~ puhuu “In. Id. d .D' moo.vofi.~ soo.ooo.o ao0.soo,fi mnoommo.m mno+oon.fi mna.5mfi.m oco+ocm.o Huu+auo.a m¢o+mmc.a Neo+0vv.w mooconfi.m oas.ooo.o floo+aoo.a «no.0wn.fl amassed.“ «00-nnfi.m amoroooao Hooonoo.a moo+umfi.a douoflv~.o Noosnofl.m ono.ncn.n Hoo.aou.fi fico.m¢m.o “Co.moo.n monumcfi.m uncooao.o “om.moo.fi «noomnm.h fion+omm.c «mononfi.m 16.2w arm no: 61 urxu . b~mcvmtu P0111“ J'HNPJFUJJ . I Ncncswrfi t.) c» Fl (.1 c‘) r) (7* c; . \ t 6 4 2 6 “‘7‘- I N 5; JP QAVMC c'qmr u.Wmc com¢m. .\ ’e “i o ”(.3 v! P" (.573 ‘1) 5’ 9'1 1‘4 c.n» u.v« o.na c.~d o.o .l38 BIBLIOGRAPHY Published Works * . A Estrutura Amie Brasileira: Dados Prolininares. Volume 1. Rio do Janoiro, Brasil: Instituto Brasileiro de Reforms Agraria (IBRA), ca 1968. * . A Industria De Curtunes Do Nordeste. Fortaleza. Ceara. "‘" Burasills—IBanco" do Nordest'e do Brasil S/A, Departuento de Estudos Economices do Nordeste (ME) , 196+. . A Sudeno E A mgcuaria Nordestina.’ Recife, Brasil: Superintendencia do Desenvolvimento do Nordesto, Departanento de Agriculture. o Abastecinento, Divisao de Documentacao, 1968. . Ancar-Ceara '68: Roaltorio Do Atividades. Fortaleza, Brasil: Servicio do Extensao Rural do Ceara. 1969. . Contribuicao Ao Estudo Das Plantas Alinentares. Recife, Brasil: Volume I. Superintendencia do Desenvolvimento do Nordeste, Divisao do Documentacao, 1967. * . Di ostioo Socio-Econoaico Da Zona Fisi ica De Baturite. Value I: "Aspectos Fisiograficos." Fortaleza, Estado do Ceara: Superintendencia do Desenvolvilento Economico e Cultural (SUDEC) , 1967. * . Manual De Estatisticas Basicas Do Nordoste. Fortaleza. Ceara, Brasil: Banoo do Nordoste do Brasil 8 A, Departe- nento do Estudos Economicso do Nordoste (BN3 E) , July, 1964. . 0 Crodito final No BNB. Fortaleza. Brasil: Banco do Brasil 87A, Departaunto Rural, 1969. . Pecuria Bovine De Corte Do Nordeste. Fortaleza. Ceara, Brasil: BNB ME. 19 5. . Plano Dirotor: Qu__a_trienio 1268-2 . Fortaleza, Brasil: Servicio Extensao Rural do Ceara, 1968. . JPro coes De Oferta E Demands De Productos icolas Para 0 Brasil. Volume I: "Fundacao Getulio Vargas.” Instituto Brasilorio do Econonia, Contro do Estudos Agricolas, September, 1966. 140 1’41 . Relatorio Anual. Recife, Brasil: Secretaria Do Agriculture. doéstado do Pernambuco. Instituto do Pesquisas Agronomicas, 19 . * . Sn imento De Generos Alinenticos Para A Cidade De Fortaleza, Fortaleza, Ceara, Brasil: Banco do Nordeste do Brasil 87A Deartanento do Estudos Economicos do Nordeste, December, 19 . Alvos do Andrade, F. Istrutina Ma I 0 Desenvolvinento Magen- aria Do Nordest______g. Paper presented to the Fifth Congress of Brazilian Agronomy, Recife. Brazil, October, 1967. Barbosa da Cruz , Waldo-1r and Ferreira do Mole, Francisco de Aseis. Estudo Geguinico Preliminar Dos Aguas Snbterraneas Do Nordeste Do Brasil. Recife, Brasil: Superintedencia do Desenvolvimento do Nordeste (SUDRIE) , Divisao do Hidro- geologia, 1968. *Burke, Jolm M. and Martins, Carlos B. ”Preliminary Design of a Beef Products Industry.” Project Brasil: Feasibility Studies and Prelimina Dos . Morris Asinow, Director. Report No. 3- , University of Ceara. Brazil and the University of California, Los Angeles, August, 1963. Dawson, Richard E. "Simulation in the Social Sciences.” Sinulation in Social Science: Readings. Harold Gnetzkow, ed. Englewood Cliffs, N.J.: Prentice Hall. 1967. Dicker-an, Alan R. The Economic Structure and Maia of a Renew System in Northeast Brazil. Fortaleza, Ceara. Brasil. University of Arizona in co-‘operation with U.S.A.I.D. and the Federal University of Ceara. 1968. Hinnan, H. R. and Hutton, R. F. A Gegoral Simulation Model for Fara Firns. Washington, D.C.: USDA, Agricultural Economics Research, Value 22. Nunbor 3 (July, 1970). J ohnson. Glenn L. "Discussion of Macro Sinulation Models.” Anerican Journal of :Aggicultmral Economics, Volume 52 (May. 1970}, Pp. 2 .2 e Johnson, Glenn 1... Deans, R.. Halter. A. 11.. Hayenga, M. I... Kellogg, E. , Manetsch, T. J.. and Pamni. K. A Simulation Model of the Nigerian Rural Econog: Phase I - The Northern Nigerian Beef Industgz. Progress Report to the Agency for International Development. East Lansing: Michigan State University , April 26, 1968. 192 Llewellyn, Robert N. Fordyge: An Industrial Dyngggcs Simulator. Raleigh, North Carolina: Typing Service, 19 5. Maior, Joel Souto. Ground Water in Northeast Brazil. Serie Hidro- geologia No. 21. Recife, Pernambuco, Brazil: Superin- tendencia do Desenvolvimento do Nordeste, Departanonto do Recursos Naturais, Divisao do Hidrogeologia, Division do Documentacao, August, 1969. Malta da Costa, F. and Bezerra, A. A. Proljeto Piloto Da Barre. Do B Bebedouro Plano De Producao Animal. Rome: FAO of the United Nations {undated5. Manetsch, T. J. Computer Simulation Analysis of a Program for Modernizing Cotton Production in Northeast Brazil. East Lansing: Division of Engineering Research, Michigan State University, August, 1968. Naylor, Thomas H. "Policy Simulation Experiments with Macro-econo- metric Models: The State of the Art." American Journal of'Agricultural Economics, Volume 52 (May, 1970;, pp. 2 3-278. *Pan American Union. Inventory of Information Basic to the Plannigg of Agricultural Development in Latin America-éBragil. Washington, D.C.: Interquerican Committee for Agricul- tural DeveloPment, 196“. *Pan American Union. Land Tenure Conditions and Socio-Econonic Development of the Agricultural Sector: Brazil. Washington, D.C.: Intorquerican Committee for'Agricul- tural Development, 1966. Quinn, L. R. Beef Production of Six Tro ical Grasses. New York: Ibec Research Institute, Bulletin No. 23, 1963. Quinn, L. R., Mott, G. 0., and Bisschoff, N. V. A. Fertilization of Colonial Guinea.Grass Pastures and Beef Production with Zebu Steers. New York: Ibec Research Institute, Bulletin No. 29, 1961. *Robock, Stefan H. Brazil's Developing Northeast. Hashington, D.C.: Brookings Institution, 1963. *United Nations. Livestock in Latin America: Status, Problems and Prospgcts. Volume II: Brazil. New York: Food and Agriculture Organization, 1965. Viana, 0. J. "Sobre A Compasicao Botanico E Producao Dos Pastos Nativos Cearenses." Bulletin of the Agronomy Society of Ceara, Volume 6 (July, 1965), pp. 29-38. lhj Unpublished Material Allen, Cleveland James. FAO Animal Production Officer. Recife, Brazil. Personal communication. September, 1969. Anderson, Raymond. Livestock Specialist, University of Arizona Project to the School of Agronomy, University of Ceara, Fortaleza, Brazil. Personal communication. September, 1969. Bezerra, Eduardo and others. Banco do Nordeste do Brasil, S/A Departamonto do Estudos Economicos do Nordeste, Fortaleza, Brazil. A series of personal communica- tions. September, 1969. Bowen, Elbert B. and others. Food and Agricultural Officer, Division of Agriculture and Rural Devo10pment, U.S.A.I.D. A series of personal communications. September, 1969. Dixon, J. J. G. ”An Economic Analysis of Range Improvement in the Cattle Breeding Area of Buenos Aires Province.” Unpub- lished Ph.D. dissertation. East Lansing: Michigan State University, 1969. Faculty and their counterparts from the University of Arizona Project at the Escola do Agronomia da Universidade Federal do Ceara, Dopartamento do Zootecnia. Fortaleza, Brasil. A joint interview. September, 1969. Johnson, Lawrence A. Associate Professor of Dairy Science, Michigan State University. A series of personal communications throughout 1969. Manes, Charles. Livestock Specialist, University of Arizona Project to the School of Agronomy, University of Ceara. Personal communication. September, 1969. Kilpatrick, Dr. Henry Mike, Agronomist, U.S. Agency for International Devo10pmont/Ibec Research Institute (IRI), and Dr. Ruy Carualheira Uanderley, IPEANE, Chefe do Zootecnia. A personal communication. September, 1969. Larson, Donald H. "A Diagnosis of Product and Factor Market Coordination in the Bean Industry of’Northeast Brazil.” Unpublished Ph.D. dissertation. East Lansing: Michigan State University, 1968. lhh Manetsch, Thomas J., Associate Professor, Electrical Engineering and System Science, Michigan State University. A series of personal communications from September, 1968 through June, 1970. Page, Gloria. An untitled, unpublished and undated paper describing a testing procedure for use on the Nigerian Model, Macro Model II. East Lansing: Michigan State University. Technical Staff of the Superintendancy for the Development of the Northeast, Division of Agriculture and Supply, Recife, Brazil. A series of interviews. September, 1969. Valente, Clinton Saboia. EngenheiquAgronomo. Fortaleza, Brasil: Servico Do Extensao Rural do Ceara. Personal inter- view. September, 1969. Webb, Thomas R. “A Systems Model for Market Development Planning: Northeast Brazil.” Unpublished Ph.D. dissertation. Michigan State University, 1969. * Denotes availability at Michigan State University, East Lansing, ”1011183110 ‘ll‘llllll. 3