‘mmmnmmminimum" x/ 22.: 3 1293 01068 3898 Ef"“‘ was. find.“ Usirerizzy This is to certify that the thesis entitled Wheat Seed Marketing in Eastern Bolivia presented by Nicholas William Minot has been accepted towards fulfillment of the requirements for , . Master 8 degree in Agricultural Econ. (GA Major professor Michael T. Weber April 29 , 1985 Date 0.7639 ' Msua an Affirmative Action/Equal Opportunity Institution MSU LIBRARIES “ \r RETURNING MATERIALS: Place in book drop to remove this checkout from your record. flfl§§_wiil be charged if book is returned after the date stamped below. m 1:0 c-2005" WHEAT SEED MARKETING IN EASTERN BOLIVIA BY Nicholas Williaa Hinot A Thesis Submitted to Michigan State University in partial fulfillaent of the requireaents for the degree of HASTER OF SCIENCE Departnent of Agricultural Economics 1985 ABSTRACT wHEAT SEED HARKETING IN EASTERN BOLIVIA BY Nicholas Williaa Hinot This study begins with an analysis of the econosic and organiza- tional aspects of seed industries in less developed countries. Both theoretical argu-ents, based on the special econonic characteristics of seed, and enpirical evidence of seed industries are cited. It is argued that public sector efforts should be focused on a few activities and should facilitate private sector participation where possible. The conclusions are illustrated with a description of the successful refer-s in a seed progras in the eastern lowlands of Bolivia. In addition. a diagnostic-prescriptive analysis of this wheat seed aarket is conducted, with particular attention to the probles of late delivery of certified seed. Using the results of a survey of the lowland wheat farners, the study evaluates the econoaic and financial feasibi- lity of a processing plant and storage facility which could provide seed earlier in the season. Given appropriate seed pricing policies, the investsent provides an acceptable rate of return. Recossendations are aade regarding equipment. capacities, plant location. and general seed aarketing policies and procedures. This work is dedicated to ey parents. the best a guy could ever hope for. 11 ACKNOWLEDGMENTS I have benefited greatly from the assistance of many people in the course of this study and during my graduate prograa in general. Among the faculty at hichigan State University, I would like to express my sincere appreciation to ny committee members: Dr. Michael Weber for his detailed reading and critique of the thesis drafts. Dr. Harold Riley for his longstanding interest in this research opportunity and his editorial and substantive contributions. and Dr. Russell Freed for providing constructive criticism from a technical science perspective. I as also grateful to Dr. Carl Eicher for his consistent support during ey grad- uate program and in ay professional endeavors. Financial support for the field research involved in this study was provided by USAID/Bolivia under the Agricultural Sector II Pro3ect contracted to Cheeonics International Consulting Division. I am grateful to Candace Conrad for her confidence in proposing as for this position. In Bolivia. I am particularly indebted to Dr. Preston Pattie, a good boss and a good friend, whose support and guidance throughout my stay in Bolivia was invaluable. Dr. Adriel Garey was instrumental in providing resources, contacts. and technical infornation necessary for the study. Also contributing in this regard were Drs. Edgar Cabrera and Juan Landivar. I would also like to thank Ing. Jorge Rosales, Ing. Adalid ROJas, Lic. Niguel Piaggio, and Ing. Humberto Wayar for assist- ance of various kinds. Lic. Gover BarJa for being a great “contraparte.” to Gualberto NOJica for his companionship on the road. and to the numer- ous Bolivian professionals and farmers for their generous cooperation in the study. Huch credit is also due to the Chemonics support staff, particularly Maggie Alborta and Rosario de MOJica. for their skills, dedication, and good cheer. And finally, I would like to thank Patricia Bonnard for her contri- butions to this study, as well as for her patience and encouragement during its preparation. iii TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES Chapter 1 INTRODUCTION 1.1 Background 1.2 Objectives of the study 1.3 Plan of the study SEED INDUSTRIES IN LESS DEVELOPED COUNTRIES 2.1 Definition and importance of seed industries 2.2 Characteristics of seed as an economic product 2.3 Components of a seed industry 2.4 Role of the public sector in seed industries 2.4.1 Plant breeding 2.4.2 Seed multiplication, drying, and processing 2.4.3 Seed marketing 2.4.4 Seed certification and legislation BOLIVIAN AGRICULTURE AND SEED INDUSTRY 3.1 3.2 3.3 Geography, climate and population Bolivian agriculture 3.2.1 Agricultural production by region 3.2.2 Agricultural development in Santa Cruz Seed industries in Bolivia 3.3.1 Overview of seed programs 3.3.2 Seed programs in Santa Cruz DESCRIPTION AND DIAGNOSIS OF THE LOWLAND WHEAT SEED MARKET 4.1 ° 4.2 4.3 Wheat in Bolivia 4.1.1 Patterns in wheat production 4.1.2 Wheat marketing and price policies 4.1.3 Wheat production in Santa Cruz Wheat seed in Santa Cruz 4.2.1 Seed saved from the previous harvest 4.2.2 Seed from other departments 4.2.3 Seed from Proyecto Abapo-Izozog 4.2.4 Seed from Proyecto Trigo of CIAT 4.2.5 Estimation of total seed consumption Diagnosis of wheat seed supply iv Page vi vii 11 15 16 19 32 35 35 38 38 4O 43 43 47 54 54 57 59 63 7O 7O 72 73 74 77 81 Chapter 5 ANNEX A ANNEX B Se Ou PROJECTIONS FOR THE LOWLAND WHEAT SEED NARKET 1 Progected supply of wheat seed 2 Pronected demand for wheat seed 5.2.1 Area of wheat production 5.2.2 Planting density for wheat production 5.2.3 Proportion of wheat seed supplied by CIAT 3 Temporal distribution of supply and demand 4 Synthesis of volume and storage proaections 5. 5. 5. 5. STRATEGIES TO INPROVE THE SUPPLY OF WHEAT SEED 1 Preliminary considerations 2 Capacity and equipment requirements 6.2.1 Seed drying 6.2.2 Seed cleaning and treatment 6.2.3 Seed storage 6.2.4 Other equipment and facilities 6.3 Financial analysis of costs and revenues 6.3.1 Revenue 6.3.2 Investment costs 6.3.3 Operating costs 6.3.4 Comparison of costs and revenues 4 Economic analysis of costs and benefits 5 Sensitivity analysis .6 Evaluation of alternate sites for the proceesing plant 7 Wheat seed price and marketing policy 6. 6. SUNNARY AND CONCLUSIONS 7.1 Conclusions regarding seed industries in less developed countries 7.2 Conclusions regarding wheat seed marketing in lowland Bolivia 7.3 Recommendations regarding wheat seed marketing in lowland Bolivia 7.4 Implications for further research : SEED AS A ”PUBLIC GOOD” : NETHODOLOGY OF WHEAT FARNER SURVEY mple selection estionnaire preparation and the interviews Interpretation of the results BIBLIOG RAPHY Page 85 85 87 87 92 98 100 100 101 105 107 108 108 108 111 114 118 122 124 128 134 134 137 138 140 141 144 144 146 150 151 Ubuflufl-fifififiw HQU‘UTDCDNPH U N eeeeeeeee HHPADDVO‘U‘ NPO 0‘0‘ 0‘00‘0‘0‘0‘0‘0’0‘ LL- .51» LIST OF TABLES Page Certified seed production in Santa Cruz 52 Bolivian wheat and flour production and imports 56 Wheat area and production in Santa Cruz 67 Frequency distribution of the cultivated area for sample farms 68 Frequency distribution of summer crop combinations 68 Frequency distribution of winter crop combinations 68 Sources of wheat seed in Santa Cruz 78 Estimated wheat area in lowland Santa Cruz, Winter 1984 80 PrOJected wheat area, seed consumption, and seed sales in 91 lowland Santa Cruz Temporal distribution of demand for wheat seed and the supply 94 of CIAT wheat seed Progected production, storage, and sales of CIAT wheat seed 97 Cumulative monthly processing capacity by type of air-screen 103 cleaner and type of seed being processed Capacity utilization of two air-screen cleaners under 104 different assumptions regarding CIAT seed sales and storage Narginal cost and revenue of storage capacity as a function 106 of storage capacity Gross revenues of seed processing plant 109 Costs of imported equipment for processing plant 110 Construction and installation costs for processing plant 112 Personnel costs of seed processing plant 113 Energy costs per metric ton of processed seed 115 Costs of seed processing plant 116 Financial analysis of seed processing plant 117 Economic analysis of seed processing plant 121 Sensitivity analysis: Financial and economic internal rates 123 of return Characteristics of alternate sites for the processing plant 127 Additional transport cost per 100 NT of seed for different 127 processing plant sites vi LIST OF FIGURES Nap of Bolivia Composition of Bolivian wheat/flour supply Bolivian producer prices and CIF import prices for wheat lap of Santa Cruz wheat and wheat seed growing areas Example of survey questionnaire Phrasing of survey questions vii Page 55 60 75 148 149 CHAPTER ONE INTRODUCTION 1.1 Background This study is based on research carried out by the author in 1984 under a contract between Chemonics International Consulting Division and the Bolivian Ninistry of Agriculture and Campesino Affairs. The prOJect, financed by the United States Agency for International Development, included a seed component with activities in several regions of Bolivia. Over the past four years, the proaect has fostered the rapid development of a seed industry to produce soybean, wheat, rice, and maize seed in the lowland area of the department of Santa Cruz. One of the tasks assigned to the author was to study the market for wheat seed in this area and to investigate the feasibility of a wheat seed processing plant and storage facility in the seed producing zone. As part of this study, the author conducted a survey of the lowland wheat farmers to determine various aspects of the demand for seed and the performance of the exist- ing seed industry. The wheat seed program in Santa Cruz is particularly noteworthy because it has facilitated the rapid expansion of wheat production in the lowlands raising the possibility that this region may reduce the sizeable imports of wheat. Furthernore, the program involves contract seed production by small-scale farmers in the highlands, thus illus- trating the potential as well as the limitations of such a strategy. Also of interest is the division of responsibilities among institu- tions in this seed program which constitutes a departure from the system used elsewhere in Bolivia. Recent reforms have attempted to create a largely self-supporting seed industry and to promote a greater role for the private sector. Following the completion of the assignment, the author attempted to place the experience of the Bolivian seed industry in a broader context by studying the patterns of seed industry organization in other less developed countries. Although there is a growing body of literature on the economics and organization of agricultural research, relatively little work has been done on the organizational aspects of seed pro- grams. The present study attempts to address this deficiency in the literature. 1.2 ObJegtives of thg study This study has two overall objectives. First, it investigates the relationship between the institutional and policy environment of a seed industry and its performance. The goal is to develop guidelines to help identify the appropriate role of the public sector in different circum- stances. In part, this involves theoretical arguments based on the technical characteristics of seed and the nature of its supply and demand. In addition, it draws on empirical evidence of the performance of seed industries in less developed countries as a function of the institutional and policy environment within which they operate. Second, the study illustrates the complexity of the economics of seed production through an analysis of the wheat seed program in Santa Cruz, Bolivia. The analysis aims to solve several practical problems faced by the program, particularly the timing of the supply of wheat seed and the shortage of processing capacity in the region. It is hoped that the practical issues considered can be placed in the context of the more general analysis of seed multiplication programs and their role in agricultural development. 1.3 th d Following this introduction (Chapter One), the economic and insti- tutional aspects of seed production and distribution are considered in Chapter Two. The special characteristics of seed as an econdmic product are discussed and the components of a seed industry described. Both theoretical and empirical evidence are then used to analyze the role of the public sector in developing an effective seed industry. The analysis of the wheat seed market in the lowlands of Santa Cruz employs the description-diagnosis-prediction-prescription format for analyzing market structure and performance (Shaffer, 19701. In general terms, Chapters Three and Four are descriptive and diagnostic, Chapter Five involves making progections, and Chapter Seven is prescriptive. Chapter Three provides background information on Bolivia, its agricultural development, and its seed industry. The experience of Bolivian seed programs is used to confirm some of the patterns discussed in Chapter Two. The chapter also provides a descriptive foundation for more detailed consideration of the wheat seed market in the following chapter. In Chapter Four, the market for wheat seed in the lowland Santa Cruz area is described. First, the national wheat situation and the patterns of wheat production in Santa Cruz are discussed. Then, lowland wheat seed consumption and wheat area are estimated. Finally, several problems with the supply of wheat seed are diagnosed, focussing on the problem of the late availability of certified seed. This chapter employs data from a survey of lowland wheat farmers conducted by the author in July 1984. The methodology of this survey is explained in Annex B. In Chapter Five, the survey data are used to make progections of the supply and demand for wheat seed. The temporal distributions of supply and demand are also estimated and used to prOJect storage requirements. Chapter Six presents various recommendations for improving the supply of wheat seed, focussing on the possibility of constructing a seed processing plant and storage facility. Capacity and equipment requirements are estimated, economic and financial feasibility studies carried out, and alternate sites for the plant are evaluated. In addi- tion, recommendations are made for seed pricing and marketing. In Chapter Eight, the results of the study are summarized and the conclusions reiterated. Lastly, the implications for further research are presented. CHAPTER TWO SEED INDUSTRIES IN LESS DEVELOPED COUNTRIES Over the last twenty years, there has been increasing awareness of the importance of agricultural technology in economic development. In part, this comes from a greater understanding of the linkages between agricultural development and more general economic development [Johnston and Nellor, 1961: Nicholle, 1964; and Reynolds, 1983]. In addition, it is now widely accepted that technology generation and diffusion play a central role in agricultural development [Jorgenson, 1961: Schultz, 1964: and Nellor, 1966]. Efforts to develop new technology have been directed toward yield-increasing biological and chemical technology in response to the successes of the "green revolution“ cultivars. The work of Hayami and Ruttan [1971] provided the historical and theoretical foundation for the idea that, where low wages prevail, agricultural development can and should pursue more labor-intensive yield-increasing technology relative to labor-saving mechanical technology. One of the most important forms of yield-increasing agricultural technology is improved cultivars. However, in order to disseminate this technology, improved seed must be produced on a scale beyond the capabi- lity of most research stations. The feasibility of seed importation is limited by the low value/bulk ratio of seed and by the location- specificity of most cultivars. Thus, the development of an indigenous seed industry is a critical component of agricultural development stra- tegies. 2.1 e niti n a m ortanc seed indus ries The traditional system, in which farmers save their own seed and exchange it among themselves. has served humankind for thousands of years and is probably still the dominant form of seed production. Seed selection in this system may be counter-productive if the poorest or latest-maturing portion of the crop is saved for seed. There are several intermediate approaches to improving the seed supply. Ordinary grain may be processed and treated for use as seed. Another approach is to contract farmers to return 1.5 to 3 bags of seed for every bag received, with the remainder being marketed as grain. Neither system, however, involves the use of specialized seed growers who can provide the attention needed for quality seed production. The development of seed industries, specialized sets of public and/or private institutions to produce and market seed, is a relatively recent phenomenon. In North America and Europe, seed industries evolved in response to the development of scientific breeding technology in the 19thicentury and the subsequent availability of new cultivars (Copeland, 1976: 214]. In many of the less developed countries, efforts to form seed industries began in the 1950s and 1960s, although only a minority can be said to have effective seed industries (Douglas, 1975]. A survey by the Food and Agriculture Organization (FAO) of 97 countries revealed that only 8x were at a “technically and/or economi- cally advanced level." In 85x of them, "improved varieties are available but seed supply is inadequate," and in the remaining 73 improved culti- vars were "not yet available" [Feistritzer, 1972: 87]. Although almost all the industrialized countries had seed industries for cereals, indus- trial crops, pasture crops, and vegetables, the seed industries in less developed nations were often limited to cereals and, less frequently, industrial craps. Of the 37 countries surveyed, all but three (Bolivia, Chad, and Niger) had "operational seed production and quality control schemes“ in wheat, maize, or rice [Feistritzer, 1978]. A follow-up study by the FAO in 1979-80 confirmed these general patterns noting that, in most cases, seed programs in less developed countries were ”fragmentary or pilot-scale operations“ [Feistritzer, 1982]. Given the existence of improved cultivars, an effective seed indus‘ try contributes to agricultural production in several ways. Yields can be raised as a result of the greater genetic potential, higher germi- nation rates, greater seed vigor, and lower incidence of disease. In addition, varietal purity contributes to uniform maturation and consist- ent product characteristics. This facilitates industrial processing (e.g. milling, canning, oil extraction, melting, ginning, and packaging) which becomes increasingly important in agricultural development. These benefits are reflected at the farm level in the significant returns that can be obtained from a relatively small investment. Nore general benefits of increased agricultural production accrue to society through lower prices of agricultural goods, the generation or saving of foreign exchange if the commodity is traded, and so on. The gains are demonstrated by numerous studies which have found high rates of return to national investment in seed improvement [see Ruttan, 1982]. In addition, the generation of improved cultivars and dissemination through seed industries has several advantages with respect to equity and food security goals. Food security has been defined as: ’the ability of food deficit countries, or regions or households within these countries, to meet target consumption levels on a year-to-year basis. [Valdes and Siamwalla, 1981: 2] By facilitating the continuous introduction of new cultivars, a seed industry contributes to food security in the following ways: 1) 2) 3) 4) 5) 4) 5) It improves the ability of farming households with food deficits to raise consumption levels to the extent that higher yields raise income earnings or home production. Because improved seed is divisible, simple to adopt, and relatively inexpensive, it is more scale-neutral than other forms of technology, benefit- ing small farmers as well as large. Although the higher earnings last only until the new technology is reflected in crop prices, the effect on home production is more lasting (Dalrymple, 1979]. The benefits of improved seed reach even the landless agricul- tural laborers to the extent that it increases labor require- ments, an important consideration in many Asian economies. Empi- rical evidence shows that the new cultivars increase or, at worst, do not decrease the demand for labor, in contrast to mechanical technology which tends to displace labor [Hayami, 1984]. Over the longer run, the lower per unit cost of production of marketed goods often results in lower consumer prices. When the commodity is a food crop, the benefits accrue disproportionately to the non-agricultural poor (i.e. food-deficit households) who spend a large portion of their income on food. It has also been argued that the use of modern cultivars with greater responsiveness to nitrogen fertilizer and other pur- chased inputs makes the supply of agricultural goods more elas- tic with respect to price. Shifts in demand have less impact on prices, thus moderating the impact on real purchasing power for both consumers and producers. Although the original "green revolution" cultivars were bred for high yields, breeding since then has placed greater emphasis on disease resistance and, more recently, drought resistance. With effective seed industries, these cultivars will help stabilize production levels, both at the household level and at the regional and national levels (see Herdt and Capule, 1983]. The existence of an effective seed industry facilitates the stockholding of seed in preparation for natural disasters when normal seed supplies may be consumed or damaged [see Feistritzer, 1975]. Lastly, an effective seed industry can contribute to the flexi- bility of the agricultural system by accelerating the diffusion of new cultivars developed in response to changing conditions. The following section discusses the special characteristics of seed as an economic product. Next, the components of a seed industry are described. Finally, in section 2.4 the critical role of the public sector in facilitating the development of a seed industry is discussed. 2.2 ghggggtegistigs 9; seed as an gcggomic product The organization of the seed industry, like any industry, is great- 1y influenced by the inherent characteristics of the product and by the nature of its supply and demand. From an economic perspective, five characteristics of seed are particularly relevant. First, for most crops, a seed “consumer" is automatically a seed “producer“ as well. Thus, seed industries must compete with farm-level production of seed. Their ability to compete naturally depends on the cost and performance of saved seed relative to purchased seed. Some crop seeds are difficult to harvest, process, or store, requiring spe- cial machinery or skills. This introduces economies of scale which allow seed industries to produce more economically than the farmer. This is the generally the case for pasture seed because of its small size. Similarly, some oilseeds and vegetable seeds are difficult to process and do not store well [Delouche et al, 1973: 675]. When saving seed is simple and inexpensive, seed industries must supply seed with sufficiently superior performance to compensate for the greater cost of purchased seed. This is often the case for grains and some grain legumes which are easy to harvest (consisting of the agricul- tural product itself) and store well. Although hybrid seed cannot be reproduced at the farm-level, the seed must still offer superior perfor- mance to non-hybrid cultivars of the same crop. For many grains, even in developed commercial agriculture, seed is purchased every three to six 10 years, when the saved seed has degenerated or new cultivars become available. A second important characteristic of seed is that the criteria which determine the value of the seed to the consumer are complex and varied. These criteria relate to the attributes of the seed itself and the genetically-determined performance of the plant. Important attri- butes of the seed are the germination rate, vigor, seed health, and cleanliness, while plant performance includes the yield under different agro-climatic conditions and cultural practices, maturation period, resistance to disease and pests, palatability and/or marketability of the product, quality and quantity of by-products such as stalks, and ease of cultivation and harvest. Because of the complexity of buyer criteria, seed production and marketing require knowledge of and respon- siveness to consumer preferences. Third, most of these components of seed quality are not readily apparent to the consumer at the time of purchase. Seed quality greatly determines the value of the seed to the consumer, but can only be evaluated directly at costs prohibitive to the individual buyer. While this is true of many products, it is a more serious obstacle given the the option of saving seed from one’s own crop. Without an effective demand for quality seed, there is no incentive to produce it. Clearly, there is a role for an impartial arbiter of seed quality, both to provide the consumer with useful information and to protect, and thus promote, the producer of high-quality seed. Fourth, the supply of and demand for seed are strongly seasonal and vary randomly from one year to the next according to the climate. This factor, in combination with the perishability of seed, highlights the 11 importance of proper storage from seed harvest to planting season. In addition, an agile marketing system is necessary to distribute the seed in a timely fashion. Seed delivered late may be as useless as undeli- vered seed. Fifth, plant breeding generally has “public good" attributes so that the resulting benefits are difficult to "capture“ by the respon- sible institution(s). This is because the development of new cultivars requires sizeable research costs, but, once developed, many are inexpen- sively reproduced. The seed can be multiplied and sold for Just the cost of reproduction. This prevents the original "inventor” from selling the seed at a price that also covers the fixed costs of breeding (Annex A covers this topic in more detail). 2.3 Components of a seed industry Seed production consists of seed multiplication, drying, proces- sing, storage, marketing, and quality control. Each component will be described briefly, emphasizing the case of non-hybrid grain seeds. Although similar to grain production, the production of seed, as a living organism, requires special treatment to avoid physiological deterioration, mechanical damage, and seed mixture [overviews of seed production are provided by Copeland, 1976: Thompson, 1979: Delouche and Potts, 1971: and Feistritzer and Kelly, 1978: and Douglas, 1980]. Seed multiplication begins with a small quantity of physically and genetically pure seed provided by a breeding unit. This is multiplied (grown and harvested) several generations to obtain sufficient volumes of commercial seed. Although the number of required generations and the names of each generation vary, the original seed is usually called 12 "breeder” (or ”pre-basic”) seed, the next “foundation“ (or “basic") seed, and then ”certified" seed. Where four of more generations are necessary, a "registered” seed generation comes between foundation and certified seed, and sometimes there are multiple generations of certi- fied seed. The multiplication factor from one generation to the next is between ten and forty for most seed-propagated crops. For seed produc- tion, more care is taken in weed and insect control, harvesting, thresh- ing, and processing, especially for the initial multiplications. Isola- tion to prevent crossing (for cross-pollinated crops), roguing to remove off-type plants, and field inspection are also important parts of seed multiplication [Delouche and Potts, 1971]. Seed is often harvested earlier (thus, at high moisture levels) than grain to reduce field deterioration and exposure to weather and pests. However, high moisture content in harvested seed accelerates its deterioration. If the percentage moisture in the seed is above 13%, it must be dried down to this level within 24 hours of harvest to prevent deterioration. Generally, this must be done with artificial drying systems, that is by forcing heated air through a bulk storage bin. Solar drying is slow and leads to reduced germination rates (Thompson, 1979]. Seed processing includes cleaning, sizing, treating, and bagging. A variety of machines are used to remove inert material, non-seed organic material, seeds of weeds and other crops, and damaged or immature seeds. The air-screen cleaner, the most common, separates according to size and shape using vibrating screens and air flow. The screens are available with different sized and shaped holes and can be changed to clean seeds of different crops. Another cleaning machine is the gravity table, a machine with a vibrating surface that separates according to the density 13 of the particles. A third commonly used machine is the indented cylinder which uses a revolving cylinder with small indentations to separate according to the length of particles. Other machines are available for separating according to texture, color, electrical properties, and so on. Different combinations of machines are used according to the type of seed, kinds of contamination, and the degree of cleanliness desired. Seed may be treated with fungi- cides and/or pesticides to disinfect and protect the seed through sto- rage and later in the soil (Thompson, 1979 and Copeland, 1976]. The storage life of seed depends on the its species and cultivar, its quality, the relative humidity, and the ambient temperature. As mentioned before, cereal and some bean seeds are good storers while soybeans, peanuts, and most vegetable seeds are poor storers. In addi- tion, Delouche et al (1973: 674] states that the quality of the seed entering storage is “one of the most important and over-looked consi- derations." The relative humidity (RH) of the air surrounding the seed affects its moisture content and thus its rate of deterioration. In addition, the infestation and growth of fungi and insects are greater in warm, humid environments. Harrington’s rule-of—thumb is that seed storage life is doubled for every 1* decrease in seed moisture and every 5.50 C. reduction in temperature (Harrington, 1959]. Cereal seeds last up to nine months at 30° C. 50x RN or 20° C. 603 RH. However, some tropical and sub~tropical environments do not meet these standards, and cereal seed storage will require dehumidification and/or air-conditioning, particularly for long-term storage (Delouche et al, 1973]. Narketing refers to the distribution of the right type of seed to 14 the right place at the right time for an "acceptable" price. Although often seen as mere logistics, successful seed marketing depends on two- way communication: disseminating information regarding seed performance, price and recommended practices and inputs: and collecting information regarding consumer preferences and farm-level problems (Delouche and Potts, 1971: 82-85]. As will be discussed later, this element is often one of the most troublesome for developing seed industries. Price, quality, acceptable cultivars, timeliness, and consumer confidence are critical problems, and deficiencies at any other stage in seed produc- tion is likely to be manifested in the marketing of the seed. Quality control consists of both field inspection and laboratory testing and is carried out at various stages of seed production. Field inspections ensure that standards are met with regard to physical pur- ity, plant health, weed and pest contamination, and so on. Seed testing involves laboratory tests of seed to determine its viability (rate of germination), vigor (speed and robustness of seedling growth), physical purity, cultivar purity, and seed health. Testing before cleaning can be useful to determine the types of contamination and the appropriate calibration of the machines. Testing before and after storage may iden- tify the degree and causes of deterioration, thus guiding improvements in facilities and procedures. Testing can provide the basis for a certification system to provide consumers with basic information regarding the seed. The simplest certi- fication ticket identifies the species and cultivar of the seed, the lot number, and the period of certification validity and affirms that mini- mum seed quality requirements have been met. Information on germination rate, inert content, other crop seed content, weed seed content, and 15 purity may also be identified, either in the form of the tested value or the standards which have been met. The ticket also provides the net weight and, for treated seed, warns against animal or human consumption (Douglas, 1980: 123-129]. Although the technical characteristics of seed determine the neces- sary components of a seed industry, they do not address the important issue of the appropriate institutional arrangements for seed industries. This topic is reviewed in the following section with special emphasis on the role of the public sector. 2.4 Role pf the public secgor in seed industrigs Clearly, the appropriate arrangement depends on the political, economic, and cultural conditions prevailing in a given country, as well as the crop in question. However, certain points can be made on the basis of the characteristics of seed mentioned in section 2.2 and on the basis of international experience in seed programs. It is worth stressing that developing a seed industry is not always feasible or desirable. There must be 1) sufficient potential volume to Justify the investments in equipment and training and 2) the possibility of offering seed with a significantly lower price and/or higher quality than the seed presently used. In practice, the existence of improved cultivars may be considered a prerequisite: If the main thrust of agricultural improvement is based on the development or introduction of high yielding varieties or hybrids of food, feed, fiber or industrial crops, a seed program is essen- tial. On the other hand, if the major improvement effort involves only land reclamation, market development, fertilizer procurement, etc., and varietal improvement is not provided for, then a seed program is of doubtful value. (Delouche and Potts, 1971: 7. Empha- sis in original.) 16 Other things being equal, the following factors can be expected to contribute to the feasibility of a self-sustaining seed industry for a given crop: 1) the existence of a new cultivar with significantly supe- rior performance: 2) a large cultivated area which allows economies of scale: 3) a high producer price so that even modest yield increases will Justify the cost of the seed: 4) geographic concentration of the crop area, large farms, and good transportation infrastructure which reduce marketing costs: 5) a good supply system for fertilizer and other com- plementary inputs: and 6) a seed which is difficult to harvest, process, or store at the farm-level given the local climate. Naturally, the development of a seed industry may be desirable to meet other goals such as nutrition, equity, or food security, but the above factors influence the degree to which a seed industry could be self-supporting. The role of the public sector in seed industries varies among regions. In the more developed countries, the role of the public sector is focused on a few activities, principally of plant breeding, seed certification, and legislation to promote private seed production and commerce. In Africa and the less developed parts of Asia and Latin America, public agencies are often directly involved in seed production, processing, and even marketing. The role of the public sector in plant breeding, seed multiplication, marketing, and certification will be discussed in turn. 2.4.1 Plant breeding The public sector accounts for a large majority of agricultural research expenditure, especially in the developing countries. It com- prises over 97x of the agricultural research expenditure in Africa and Asia and 95x in Latin America (Boyce and Evenson, 1975]. Private agri- 17 cultural research has been overwhelmingly oriented toward 1) commodity research for major export crops, funded by small groups of producers, processors, and/or exporters, 2) hybrid seed development, and 3) mecha- nical and chemical technology. The relatively small magnitude of private research expenditure is understandable in light of the “public good“ character of much agricul- tural research (see Annex A). The "product“ of plant breeding and that of research are similar in some important respects. While research and development related to new agricultural techno- logy may be very expensive, difficult and time consuming, the copying and multiplication of such technology - once available - may be easy and inexpensive. Development of crop varieties with high yield capacity, disease and/or insect resistance or some other desirable characteristics may be difficult and costly. But once such a variety is made available to farmers, seed multiplication may be an easy task for the individual farmer. Thus, the ability of the research agency to benefit from the economic gains associated with the varietal development may be limited to initial sales... [Pinstrup-Anderson, 1982: 57] In economic terms, the marginal costs of providing seed of a new cultivar to an individual consumer is less than the average cost of production. In addition, the ability to reproduce some seed relatively easily means that maintaining a monopoly, which would be able to charge a price covering average cost, is not possible. Competitors can reproduce the good and charge only the marginal cost. Thus, the market price will be less than the average cost to the inventor and the latter will not be able to cover the fixed research costs. The result is that the private sector will underinvest in research to generate new seed cultivars and that public investment can yield significant economic returns. Although agricultural research has become an accepted function of the public sector, the evidence of high rates of return to public research indicates that there is still underinvestment 18 (see Ruttan, 1982: 237 and Pinstrup-Anderson, 1982: 100]. The concentration of private agricultural research on the three areas mentioned above is understood in light of these factors. Research in industrial crops may be worthwhile to a processing and/or exporting firm if it is able to control the marketing channels. This allows the firm to capture the returns to its research through lower producer prices. A hybrid seed firm is protected by the fact that the hybrid performance occurs only in the first generation, thus preventing others from reproducing the cultivars it develops. Similarly, the inventor of new machinery or agricultural chemicals enjoys protection through patent laws. Legislation in Western Europe and, more recently, in the U.S. provides patent-like protection to plant breeders to allow them to capture more of the benefits. Although such legislation has been suc- cessful in accelerating private investment in plant breeding, there are some doubts regarding its usefulness in developing areas. ...without a patent system, the less-developed countries are still able to benefit from the public research institutes and also to use or even to ”pirate“ - quite legally - seed lines developed for the developed world. It is not clear how much new private research would be encouraged. (Barton, 1982: 1074]. Thus, such legislation would do more to impede the free flow of plant material than promote private research in most less developed nations. In summary, the domestic private sector cannot be expected to play an important role in plant breeding in most less developed countries. Possible exceptions include the more technically advanced developing countries, particularly those in which crops with hybrid potential (e.g. maize) and industrial crops are important. In any case, the importance of public plant breeding is underscored by evidence, both theoretical 19 and empirical, that public investments in plant breeding and other forms of agricultural research produce significant economic returns. 2.4.2 Seed multiplication, drying, and processing Although there is considerable variation among countries and crop, it is safe to say that predominantly public seed multiplication systems are common in less developed countriesl, particularly those in Africa. These programs tend to concentrate on staple cereal crops, principally wheat, maize, or rice, depending on the region (Feistritzer, 1978], Often basic seed production is done on research stations or state farms while certified seed production is carried out on state farms and by contract growers. Private firms are most commonly involved in producing seed for crops for which price controls are minimal and an annual market for seed is assured by the technical complexity of multiplication. Thus, private firms tend to produce hybrids seed, industrial crop seed, and, where the market is large, vegetable and forage seed, rather than cereal seed (Chahal, 1975: 268]. Hybrid maize has stimulated the development of private seed compa- nies in a number of countries, such as Zimbabwe (Eicher, 1984], Uruguay (Rosell, 1975: 571, Brazil (Popinigis, 1982: 533], India (Walker, 1980], and Thailand (Brown et al, 1984]. The most frequently cited case is that of the Kenya Seed Company (KSC), which was formed in 1956 to produce 1. Examples are Bangladesh (Ahmed and Ahmed, 1983], Indonesia (Benja- min, 1981], Turkey (Gunay, 1975], Egypt (El-Gamal, 1975], Norocco (Bes- ri, 19831, Niger (Hall, 1981], Burkina Faso (Potts et al, 1983], Nali (Giacich and Tours, 1982], Senegal (Bono and Lam, 1976], Ghana (Andrews and Vaughan, 1983], Nigeria (Joshua and Singh, 1982], Cameroon (Vaughan et al, 1980], Tanzania (USAID, 1985], Uganda (Kabeere, 1983], Zambia (Gray, 1975], and most other African countries (World Bank, 1981: 175]. 20 grass seed. In the 1960s, it multiplied, distributed, and promoted publicly bred hybrid maize seed, leading to its rapid and widespread adoption among large and small farmers. The company has come to produce 13,000 NT of maize seed and almost 10,000 NT of sunflower, wheat, bar- ley, grass, and bean seed. Although the government recently bought a majority share, the NSC continues to be run on commercial lines (USAID, 1980b: Hazelden, 1982: Douglas, 1980: 241-243: and Wortman and Cummings, 1978]. Industrial crops often provide more incentive for private seed firms because of the importance of high-quality raw materials, the technical complexity of farm-level seed production, the relative ease of seed distribution through the processing firm, and the geographic con- centration of industrial crop production compared to staple crop produc- tion (Nelson and Kuhn, 1974: 1601. For example, private seed industries serve rubber and oil palm in Nalaysia (Chin and Rafar, 1978]: tea, rubber, and coconut in Sri Lanka (Pinstrup-Anderson, 1982: 63]: and soybeans in Brazil (Popinigis, 1982: 536]. The private sector is sometimes involved in vegetable seed produc- tion because of the high value of the crop and the economies of scale in multiplication. In India, the private sector is involved “mainly in high profit margin crops such as flower seeds, vegetables, and F1 hybrids“ (Walker, 1980: 24]. The first private seed companies in Thailand, formed in the 1950s, produced vegetable seed (Brown et al, 1984]. Similarly, vegetable seed is produced by private firms in Tunisia and Algeria, while government agencies produce cereal seed (Douglas, 1980: 92]. Nore commonly, vegetable seed is imported due to its high value-bulk ratio, its geographic adaptability, and the technical sophistication required 21 for production. Although not common, private production of cereal seed is sometimes undertaken, frequently by cooperatives and farmers’ associations. Such organizations presumably have better knowledge of the demand for seed and may inspire greater trust among consumer/members than outside firms or public agencies. In a number of major developing countries, complex systems have evolved, combining diversified private seed enterprises as well as semi-autonomous and mixed organizations. Examples are Brazil (Popinigis, 1982], Colombia (Noncayo, 1975], Nexico (TiJerina, 1975], India (World Bank, 1982: 76], Thailand (Brown et al, 1984] and the Philippines (Sevilla and Guerrero, 1983: 1140]. On the basis of economic theory, the case for permanent public support of seed multiplication, drying, and processing is not as con- vincing as it is for plant breeding. Although the ”product" is reproduc- ible, a seed multiplication unit which receives subsidized breeder seed from a public sector research agency need not cover the costs of research, and so, competes on an equal basis with farm-level seed multi- plication. The empirical evidence concerning seed multiplication programs suggest several common problems. First, responsiveness and flexibility are important in seed production, but civil service regulations make it difficult to generate the incentive structure to promote this kind of institutional behavior. Even in the United States, this problem has been noted: government agencies have not been successful in producing and selling seed due to the critical timing involved in performing the necessary roguing, detasseling or inspection functions. The sea- sonal nature of the Job, with the corresponding necessity of long 22 hours and solid work peaks during these periods, is not compatible with the operation of a typical public agency. (USDA, 1967, cited in Delouche, 1978: 48] Douglas (1980: 91] makes a similar point: ...government agencies often are unable to delegate decision-making authority to lower levels. For successful seed production, manage- rial decisions must be made promptly and on the spot. Second, government seed multiplication programs generally depend on public funding. This can make the supply of seed less reliable, being subject to cut-backs in funding and conditioning any expansion of output on approval of greater budget allocations (Douglas, 1980: 91: World Bank, 1981: 60: Nelson and Kuhn, 1974: 1581. Shortages of subsidized government seed have been problems in Nali (Giacich and Tours, 1982: 5021, Turkey (Gunay, 1975: 491, Uganda (Kabeere, 1983: 1090], and Indo- nesia (Benjamin, 1981], among others. And third, seed multiplication generally requires considerable amounts of financial and managerial resources. It has been argued that direct involvement in seed multiplication diverts scarse public sector resources into activities that the private sector can perform. From this perspective, government efforts must be more focused on those few functions that must be done by the public sector such as research and seed certification. On the other hand, private seed enterprises may not develop sponta- neously, or may concentrate on a few crops as described above. Nany observers argue that private seed enterprises are preferable in the long run, but that public sector involvement is necessary at least initially: In most developing countries private enterprise in seed multiplica- tion is lacking to the detriment of the country concerned. In these countries, the government has to take over the Job of seed multi- plication. (Schoorel, 1972: 97] 23 Similarly, Increasing breeder’s seed or basic seed is the responsibility of the plant breeder when the seed program is in its early stages... Producing seed for farmers is not an activity in which government agencies normally have an inherent advantage. This step can be taken over by private producers or shared by them with public institutions. However, potential private producers have to be found, stimulated, informed, and sometimes financed. (Wortman and Cummings, 1978: 350] This “infant seed industry" argument can probably be defended on the basis of the need for simultaneous development of seed production capacity and public breeding and certification programs. Not knowing the government’s commitment to seed production, private firms face greater risks to early involvement than do public agencies. In addition to the normal risks, the seed producer faces uncertainty regarding the ability of the public agency to provide adequate supplies of high-quality foun- dation seed, regarding the honesty and credibility of the seed certifi- cation service, and regarding seed and crop price controls. Once the need for some public seed production is accepted, the question turns to strategies which minimize the recognized weaknesses of public programs and which promote seed production by private entities. One way to minimize the inherent weaknesses of public seed production programs is to contract out the actual multiplication to trained farmers. The idea that mechanized state farms can produce suffi- cient volumes more economically has not been demonstrated by experience. The India Seed Review in 1968 found that "the quantum of certified seed production required is so large that it cannot be produced on government farms alone" (cited in Delouche, 1978: 48]. A major World Bank progect in Indonesia involved seed production on 2000 hectares of a state farm as well as continued use of contract farmers. The former suffered from weak management, numerous delays, 24 funding shortages, and low yields due to leveling operations, leading one observer to note: It is now clear that the progect design considerably over- emphasized the role that [government] estate production should play in the process. With regard to the seed-growing districts (of contract farmers], it is further established that these were indis- pensable if Java’s seed requirements were to be covered. (Benjamin, 1981 : 252] As a result of this "costly lesson,“ all World Bank progects since have used contract growers instead of state farms (Russell, 1983]. Two seed specialists with wide international experience have noted: Producing seed through farmer-contractors requires concentrated effort but is one of the best educational tools available to sell new technology. Contrary to most government workers’ opinion it is also the least costly system in the long run. (Boyd and Jackson, 1981: 31 However, the drawbacks of contract seed production are the logistical complexity and the difficulty of maintaining quality control. In Bangla- desh, 20,000 contract growers are needed to produce Just 8000 hectares of certified rice seed (Ahmed and Ahmed, 1983]. With regard to strategies to promote private seed production, there are various possibilities including technical and management training, subsidized credit, technical assistance, the use of government personnel or equipment, and direct government investment (Douglas, 1980: 83-92]. However, the core of any strategy to promote seed production by private entities must be 1) to build confidence in the competence and reliabi- lity of public plant breeding and certification services and 2) to allow private seed enterprises to compete fairly with public ones. The latter entails resisting the temptation to control seed prices or sell seed at subsidized prices and making processing facilities and certification available to all on a fee basis. Chahal notes that this is frequently a problem and suggests the following approach: 25 if in the particular circumstances of a developing country, it is considered necessary that State must participate in production, marketing and distribution of commercial seed, it would be better to set up a separate organization either in the public sector or the Joint sector to be run on commercial lines. (Chahal, 1975: 267]. This strategy has been adopted in the case of the State Seed Corpora- tions in India which are "profit-responsible“ (Walker, 1982]. Stripped of special subsidies, such corporations are forced to charge prices which cover costs, thus allowing private entities to compete on a fair basis. Private seed firms often begin by multiplying foundation seed produced by public research stations. However, their position is preca- rious given the fact that their sales depend on the quality and reliabi- lity of public supplies of foundation seed. Nelson and Kuhn (1974: 1641 recommend the staged involvement of seed enterprises in foundation seed production and breeder seed maintenance. Under this scheme, private seed firms are initially involved only in certified seed production. Later, the government provides incentives and technical support for these firms to take over foundation seed multiplication and, eventually, breeder seed maintenance. In all of the relatively advanced seed industries mentioned ear- lier, the private sector has become increasingly involved in seed pro- duction and plant breeding, often following reforms designed to encou- rage private seed enterprises (Mexico, the Philippines, and India) and sometimes through the outright sale of public corporations (Brazil). In summary, there are a variety of institutional arrangements for seed production depending on local conditions and the crop. Public seed production is common among new seed industries, but experience indicates 26 that public agencies are not well suited to seed production. Moreover, private seed firms can and do enter this field once research and certi- fication agencies have demonstrated their reliability and once seed price structures reflect actual costs. Cooperatives and farmers’ asso- ciations can play an important role in seed production for non-hybrid cereals if given proper incentives. 2.4.3 Seed marketing As argued in section 2.2, seed marketing is more than the logistics of physical distribution. It also encompasses a two-way flow of informa- tion between the buyer and seller. The seller must disseminate informa- tion concerning the characteristics, price, and the time and location of availability and must gather information to determine buying patterns. Permanent direct public sector participation in the distribution of seed, like public seed production, is difficult to Justify on the grounds of economic efficiency. The “infant industry" argument seems less applicable for seed marketing since rural distribution networks generally exist for a variety of commodities. Nonetheless, the control of seed distribution by public agencies is common in less developed countries. In small or newly formed public seed programs, a division of the ministry of agriculture makes seed available to the extension service for distribution. Larger systems often involve a parastatal organization. Government control over seed distribution is particularly common in sub-Saharan Africa, where seed supply is managed by the public sector in 61x of the countries. Mixed systems account for another 28x and private distribution systems only 11%. Public distribu- tion systems are almost universal among the Sahelian nations and least common in ”middle-income" countries with a large-scale commercial agri- 27 culture sector, such as Kenya, Swaziland, and Zimbabwe (World Bank, 1981: 175]. The obJectives of this kind of intervention are varied: to circum- vent alleged exploitation by private traders (sometimes dominated by ethnic minorities), to facilitate the subsidization of improved seed to meet production goals, to meet equity objectives so that even remote farmers have access to seed, and simply to maintain public control over the means of production. However, the record of public seed distribution efforts is generally poor. As early as 1971, this problem was noted: Considering all the various aspects and components of a seed pro- gram, government involvement has been least effective and most grossly inefficient in the distribution and marketing phases. Thus, private sector participation in distribution and marketing should be encouraged and assisted in all possible ways... (Delouche and Potts, 1971: 84]. This has apparently become a consensus view among observers (see Schoorel, 1972: Feistritzer, 1972: Wortman and Cummings, 1978: 351: and Bunch, 1982]. Furthermore, it has been noted that open marketplaces in developing countries often have a wide range of consumer goods with the conspicuous exception of those goods sold exclusively by the public sector (World Bank, 1981: 65 and Hazelden, 1982: 526]. The problems of public distribution systems are similar to those of public seed production: 1) the organization and incentive structure common in public agencies does not facilitate responsive, independent decision-making, 2) market discipline is absent due to the monopsony of seed supply, and 3) public sector attempts to subsidize seed prices and to implement pan-territorial pricing introduce a number of difficulties to be discussed later. Innovation and flexibility are critical to effec- tive seed distribution systems, especially given the seasonal demand and 28 complex demand criteria discussed in section 2.2. In a review of public input delivery systems in sub-Saharan Africa, the World Bank interpreted the problem as follows: (Public agencies] have difficulties in adapting bureaucratic, financial, and administrative procedures to commercially oriented operations...pay scales and hiring and promotions procedures tend to be similar to those in government. This leads to reduced indivi- dual initiative, unwillingness to make quick and independent deci- sions, and consequent efficiency losses. The absence of competition in input supply also leads to a lack of innovation. Inputs are ordered in routine fashion without regard to location-specific requirements... Under the budgetary constraint that many countries experience, the actual amount that can thus be purchased remains far below the quantity desired at the subsidized price... Subsidi- zation ties its operation to the budget year, causing delays in procurement and untimely delivery to farmers. (World Bank, 1981: 60-61] Several examples are illustrative of these problems. Most of the examples are from African seed industries because parastatal seed dis- tribution systems are particularly common in this region and because many of the recent seed progects (and evaluation reports) have been carried out there. In Ghana, a Seed Multiplication Unit of the Ministry of Agriculture produced seed and distributed at grain prices by the extension service. A 1975 evaluation reported the following: The results have been - at best - moderately successful. The exten- sion workers usually had enough to do in regular work without the additional responsibilities as salesman and distributing agent. Some extension agents have conscientiously promoted good seed, and taken orders from many farmers, but have not been able to “deliver” because the distribution system did not get the seed in the area at the proper time - through no fault of the extension agent. As a consequence, the effectiveness of the extension agent is virtually destroyed. (Vaughan and Dougherty, 1975: 6] The report argued that the subsidized price impeded the distribution of seed through cooperatives, farmer associations, and other private chan- nels. Even the marketing parastatal was reportedly reluctant to begin seed marketing until "agreement on a reasonable profit can be reached" 29 (p. 7]. A 1983 report identified distribution as "probably the weakest link" in the program, yet the government was resisting a proposal to establish a mere 20% differential between seed and grain prices (Andrews and Vaughan, 1983: 19]. In Burkina Faso, a USAID progect supports foundation seed produc- tion by the research organizations under contract to the National Seed Service. The latter then cleans the seed and sells it to the state Regional Development Organizations for final multiplication and distri- bution. Until 1981, the commercial seed prices were actually set below those of commercial grain so that seed sales were "in reality a combina- tion of seed and grain sales" (Potts at al, 1983: 27]. Even under the new system, prices are set pan-territorially so that the relationship between seed and grain prices varies, and grain-deficit areas may have grain prices above the uniform seed price. The distribution of seed and other inputs in Zambia occurs through the cooperative movement which is dominated by the governing political party. Hence, the allocation of seeds and other inputs has been used as a form of patronage to loyal party members (Bates, 1981: 111]. In Tanzania, the Ministry of Agriculture produces foundation seed on state farms established and equipped through a ten-year USAID seed multiplication progect. TanSeed, a parastatal, produces certified seed by contracting individual farmers and the state farms and is the exclu- sive distributor of food crop seed. An evaluation report noted: Delivery of seed is often late, frequently supply and demand do not coincide, and costs are very high. Currently the marketing margin is about 300 percent which makes seed - which are not subsidized - very expensive for farmers. (USAID, 1985: 8] Part of the reason that the costs are so high is that TanSeed charges a 30 pan-territorial price and must cover the costs of transportation. An additional problem is that demand estimates have been based on optimis— tic "target" consumption figures given by District and Regional Develop- ment Offices. As a result, large volumes of seed have gone unsold, though less than 5,500 MT are produced annually (USAID, 1985]. The financial position of TanSeed is so precarious that in a recent year it was unable to pay its contract growers the agreed price. A recent report called for the government to invite the Tanganyika Farmers’ Association (TFA) to become part owner by buying part of the government’s share (the TFA is a private entity which had earlier been forced out of seed production by the government). This would provide for the needed recapitalization of TanSeed and generate "incentives for efficiency and cohesive management” (Walker and Potts, 1982: 29]. In Kenya, the distribution of hybrid maize seed by the Kenya Seed Company was relatively uncontrolled. Seed was made available to over 6,000 registered retailers who paid in cash for the seed. The dealers knew their region and its likely demand well and had an incentive to provide seed on time, without ordering in excess. Sales rose from less than 1,000 MT in 1965 to over 10,000 MT in 1975, 90x of which was in 10 kg units, presumably bought by small farmers (Hazelden, 1982: 5301. Naturally, the rapid adoption rate was due to the truly superior culti° var, a product of ten years of painstaking breeding effort (see USAID, 1980b]. The point is that the distribution system was capable of deli- vering the product in ever increasing volumes. The National Seed Corporation (NSC) of India was formed in 1961. Following its rapid growth in the 1960s with the introduction of the “green revolution“ cultivars, the NSC became overextended and quality 31 declined. As a result, it was left with large amounts of unsold seed even though the supply was less than 10% of the estimated “requirements" (Thompson, 1981: 515]. Since then, foundation seed production and seed distribution has been opened up to the private sector. According to the World Bank (1978: 76]: Initially, the companies tried to promote their products by demonstrations on farms: private retailers and cooperatives were encouraged through consignment arrangements, which left any unsold seed with the seed companies. Today, seed is being distributed in India by a network of more than 10,000 dealers throughout the country. Farmers have become quality-conscious and have developed preferences for brands of demonstrated quality. (World Bank, 1978] Some 300 private seed companies, as well as 12 state seed corporations, now produce and distribute seed in India. In summary, although government agencies are often involved in seed distribution, the typical organizational structure of public agencies is not conducive to the responsiveness, timeliness, and innovation needed for effective marketing. Furthermore, public distribution systems often institute price subsidies and pan-territorial pricing. These policies reduce the ability and incentive for both private and public entities to produce and distribute seed and, in the extreme, cause seed to be wasted in grain consumption. Some of the problems mentioned, such as the short- ages, delays, and political favoritism, are consistent with patterns identified in the literature on subsidized fertilizer and credit (see Dalrymple, 1975: Bates, 1981: Adams and Graham, 1979]. Clearly, only some of the problems of seed marketing are due to the institutional arrangement of seed distribution systems. Poor seed qua- lity, lack of truly "improved" cultivars, the shortage of trained per- sonnel, shortage of spare parts, inadequate equipment, and other factors are also critical problems, each of which may manifest itself as a 32 problem of “farmer acceptance." Nevertheless, it is argued that the institutional structure of the seed industry influences its performance in addressing these problems. 2.4.4 Seed certification and legislation As mentioned in section 2.3, consumer uncertainty regarding seed quality can be a serious problem. Not only does the consumer lose from imperfect information, but there is less incentive to produce quality seed. In economic terms, the returns to quality in the form of repeat purchases are not entirely captured by the firm, but distributed over the seed industry. Thus, in a free market the private firm is likely to underinvest in seed quality. This problem is illustrated by the situa- tion in the United States during the 19th and early 20th centuries: Although seed was sometimes available on the commercial market, the supply was sporadic, and often of questionable quality. Unscrupu- lous marketing gimmicks were sometimes used to take advantage of the customers’ inability of identify seeds and their quality. This led to skepticism of the quality of seeds bought off the farm and created a climate of distrust against the development of a legiti- mate seed industry. (Copeland, 1976: 213]. The Indian seed industry in the early 1960s was described in quite similar terms: ...the seed companies often compromised on quality and tried to market mixed, wrongly labeled, or poor quality seeds. The unscrupu- lous gimmicks of such agencies not only created a climate of dis- trust among farmers, but the main objective of achieving higher yields, by the use of quality seeds was also defeated. (Siddiqui, 1983: 1065] These problems have led to the establishment of services to provide impartial information on seed quality, usually in the form of certifica- tion. They are operated by public agencies and, in some cases, by non- profit associations and cooperatives. Since certification may be consi- dered a form of simple grading, the following passage is relevant: 33 Grades are generally believed to increase resource mobility and reduce barriers to entry. When inputs or products can be bought by description, rather than by seller reputation or buyer inspection, the market is widened...Grades may help to insure that individuals in a market receive prices for their product commensurate with its quality. Grades help prevent "the clever and unscrupulous“ from “outwitting the unwary and trusting." Second, total demand for a product may be increased... (Nichols et al, 1983: 76, 78] Seed certification usually involves both field inspections and labora- tory testing following established procedures and predetermined stan- dards for each type of contamination, seed species, and class of seed (foundation, certified, and so on) (see Delouche and Potts, 1971: 65]. It is clear that seed certification or legislation is necessary only when seed is traded. It is not necessary in traditional agricul- tural systems in which farmers save their own seed or in vertically- integrated seed-and-commodity schemes which can rely on internal quality control. Some authors argue that government seed production makes certi- fication unnecessary since quality can be maintained by decree (Potts at al, 1983]. In any case, new seed programs often involve a single agency for seed production and distribution, while quality control is main- tained internally. The establishment of an independent service to cer- tify seed quality is almost always an early step in this process of institutional specialization. Public sector involvement in seed certification is Justified more by the need for impartiality than by any external benefits to certifica- tion. Thus, there is no strong efficiency argument against charging user fees to cover the costs of certification. As argued earlier, self- supporting services avoid dependence on often unreliable public funding and improve responsiveness to market needs. In practice, public participation in both seed production and seed certification creates two related problems. First, the credibility of 34 the certification service may be damaged by real or imagined political interference from the seed production agency, hence the value of main- taining the administrative and financial independence of the certifica- tion service. Second, the seed production unit may claim exclusive rights to the work of the certification service, especially if no fee is charged, thus impeding the development of private seed enterprises. A user-fee policy, to be applied to public and private producers, allows the certification service to expand to meet the demand, permits finan- cial independence, and emphasizes its proper role as a “service" to seed producers, public and private. Seed legislation establishes the legal basis for seed certifica- tion, defines the actual standards or the process for setting them, and regulates the trade of seed. The U.S. model follows the principle of truth-in-labeling (Rollin, 1975], whereas the European model restricts commerce to a list of officially approved cultivars (Kelly, 1975]. Several authors warn against premature establishment of seed legislation or the creation of overly strict quality standards which may inhibit the development of a seed industry (see Delouche and Potts, 1971: 76-81 and Douglas, 1980: 122-133]. For example, efforts to enforce the use of certified seed by prohibiting commerce in uncertified seed are neither practical nor desireable when the seed industry is incipient. Neverthe- less, when designed in accordance with the current stage of seed indus- try development, seed legislation can play a useful role in facilitating seed commerce. CHAPTER THREE BOLIVIAN AGRICULTURE AND SEED INDUSTRY This chapter describes the basic characteristics of the Bolivian agricultural economy, focusing on seed industry development. It is intended to provide a useful context for the following chapters which deal with the problem of improving the supply of wheat seed in the low- lands of the department of Santa Cruz. 3.1 GeographyI climateI and population Bolivia is a landlocked country in South America, located between 9° and 23° S. and 58° and 70° W. Although it is within the tropics, as defined by latitude, the climate varies considerably from the perma- nently snow-capped peaks in the Andes Mountains to the tropical Amazo- nian forests. Bolivia is often divided into four geographic zones: the "Altiplano“ (high plain), the “Valle" (Valleys), the Yungas, and the eastern lowlands (see Figure 3.1). The Altiplano is a vast plain running north-south in the western part of the country, bordered by Andean mountain ranges on both sides. The plain lies about 4000 meters above sea level, while the mountain ranges reach altitudes of 6500 meters. The rainfall in this area ranges from 250 mm in the south to 650 mm in the north and the temperature is cool (8-12° C. annual mean) with overnight frosts during the winter. Lake Titicaca, located on the northern Altiplano, is the largest lake in South America and the highest navigable body of water in the world. The Valleys and Yungas are an intermediate zone on the eastern 35 36 .9 x12 ‘1 I ' ( \J ._.- 3"” 00 euoeen-. TA cauz \ 37 "'33:: _ fl VALLEYS AND YUNGAS l8.288 - D zones '6‘,“ _ TROPlCAL LOWLANDS C°L° '52“ ' ALTIPLANO (LNG- |Z,|92 ' c°°L 10.668 - 9:44 - ---------- ““9“”: 752° “ """""" VALLEYS AND YUNGAS 6096 ----------- SUB-TROPICAL Alsnt TROPICAL LOWLANDS 3")“ L smmso TROPICAL 1,524+ -—--------------—- . COBUMBA usL . 1 1 1 L 1 // // / m 0 70° 69° 65° 67° 66° 65° 64° 63° 62° 6l° 60° 59° 58° 'SOURCE: Wennergren and Whitaker, 1975 I 37 slopes of the Andes, lying between about 800 and 3000 meters. The Yungas in the north consist of steeply descending mountain valleys with high rainfall (around 1100 mm). The Valleys in the south provide a more gradual descent to the east and are drier (around 700 mm). The eastern lowlands, or "Oriente," form around 65x of the national territory. Most of this area is part of the Amazon River watershed though the climate varies considerably. Tropical rain forests receiving 2500 mm of rainfall and humid savannas are found in the north, whereas the southern part consists of semi-arid plains known as the Chaco which receives only 750 mm (Wennergren and Whitaker, 1975: 84-88]. The 1982 population of Bolivia was estimated at around 5.9 million, of which approximately 55x live in rural areas (World Bank, 1984: 218. 260]. Although the overall population density is relatively low, consi- derable population pressure does exist as a result of the geographic distribution. The Altiplano and Valley regions contain around 783 of the rural population but only 25x of the land area. The population density is over 50 persons per square kilometer in the northern Altiplano, 10 per km2 in the central Altiplano and the Valley areas, and less than one per km2 in the lowlands (Whitaker and Wennergren, 1982: 237]. This concentration of population in the highlands, particularly in the northern Altiplano, reflects pre-Columbian patterns of population distribution. The pre-Inca Aymara and the Quechua-speaking Incas farmed throughout these areas before the arrival of the Spanish, whereas the lowlands were only sparsely inhabited by scattered groups of hunter- gatherers. Aymara- and Quechua-speaking Indians still populate the rural highlands, although a majority also speak Spanish. The impact of Indian populations is much less pronounced in the lowlands. 38 Rural-urban migration has been an important force in Bolivian population distribution. The total population of Bolivia grew 2.1x annually between two most recent censuses in 1950 and 1976, but the three major cities have grown at a much faster rate. La Paz in the Altiplano grew 3.5x annually, Cochabamba in the Valleys at 3.93, and Santa Cruz in the lowlands at 7.2x (Romero, 1982: 3121. The high rate of growth of Santa Cruz highlights an important trend in rural-rural migration. It is estimated that between 200,000 and 350,000 have migrated to the lowlands around Santa Cruz, to lowland areas near Cochabamba, and to the Yungas near La Paz. This migration has been spurred by the opening of roads to these zones, population pressure on traditional highland farming areas, government colonization efforts, and the demand for labor from the growing commercial agricultural sector near Santa Cruz. The development of the Santa Cruz area will be dis- cussed in subsection 3.2.2. 3.2 Bplivian agpigulture 3.2.1 Agricultural production by region Bolivian agriculture is marked by extreme variation in the types of crops and livestock grown, the degree of mechanization and use of pur- chased inputs, and the proportion of output which is marketed. However. rough characterizations can be made for each of the four geographic zones. In the Altiplano. traditional small-scale farming dominates, with most of the holding less than 3-4 hectares. This is due to the land reform carried out following the Social Revolution of 1952 and continued population growth. Potatoes are the most important food crop, with barley, forages, quinoa (a high-altitude cereal), legumes, and vege- 39 tables also being grown. Sheep, llama, alpaca, and other livestock are raised as well, particularly in the southern Altiplano which is colder and more arid. Soil erosion, low soil fertility, overgrazing, and crop damage from hail and frost are serious problems, but the area continues to provide a precarious livelihood to many small farmers. Although the climate of the Valleys is less severe, it is also characterized by highly fragmented land tenure and depleted soil ferti- lity resulting from centuries of farming. Corn and potatoes are the staple crops, while wheat, barley, and fruits are also produced. The Yungas. with its tropical humid climate, supports the produc- tion of coca, citrus fruits, coffee, and bananas. Although the soil is relatively fertile and urban markets close, steep slopes and unreliable roads impede all but the most profitable commercial crops. As a result, this area is sparsely populated (USAID, 1980a: 24-34]. Agriculture in the Oriente varies according to the climate. The tropical rain forests in the north yield Brazil nuts, rubber, and animal skins. The humid savanna and, to a lesser extent, the Chaco have become important producers of beef cattle. The beef is flown daily to the major urban centers. The most intensively farmed part of the Oriente is the area near the city of Santa Cruz. Sugar cane, soybeans, cotton, and other crops are produced by large-scale commercial farms. Rice, yuca (cassava), and maize are important small-farm crops, both for sale and for auto-consumption (CORDECRUZ et al, 1982b: 363-373]. This region is discussed in more detail in the following subsection. 40 3.2.2 Agricultural development in Santa Cruz Only in the last 25 years has the department of Santa Cruz begun to fulfill the expectations held for the eastern lowlands. Like other nations bordering the Amazon basin, Bolivia has long seen the undeve- loped forests and savannas as an immense storehold of natural resources waiting to be exploited. The conviction that the United States’ frontier phenomenon ’ought’ to be transferable southwards into Latin America with equally dramatic results is one of the oldest and most persistent in the continent. [Fifer, 1982: 4081 Nevertheless, demographic and economic growth have been frustratingly slow or short-lived. In the 19th and early 20th century, the region experienced two successive boom-and-bust cycles. the first centered on cinchona bark, used to manufacture quinine, and the second on natural rubber (Fletcher. 1975 and Hiraoka, 1980]. With the loss of lowland territory to Brazil at the turn of the century and to Paraguay in the costly Chaco War of the 1930s. the geo- political importance of populating this region became even more evident. Further incentive was provided by the Social Revolution of 1952, which involved comprehensive land reform, expropriation of the largest mines, and the abolition of semi-feudal relations in the countryside. Economic chaos, uncertainty concerning land expropriation, and possibly greater on-farm consumption meant substantially reduced volumes of marketed food. Furthermore, although over 320,000 titles to land were granted over the following two decades, many of the plots allocated were quite small. For example, in the department of Cochabamba over 20x of the land was redistributed, but more than 8,000 families received plots of less than 1.5 hectares of arable land. Colonization of the eastern lowlands to alleviate these problems became an explicit goal (Zondag, 1968 cited 41 in Stearman, 1973: 383: Weil, 1974: and Fletcher, 1975]. Many of government-sponsored colonies founded in the 1950s and 1960s were poorly designed. In addition to high costs (from U55 1500 to over U33 3000 per family), they were located far from markets and on poor soil, the colonists were miners and others without farm experience, and the selection of crops was regulated. As a consequence, settlement loans could not be repaid and the rates of abandonment were 50-853. On the other hand, the numerous colonies which formed spontaneously tended to have lower abandonment rates and higher incomes. These colonies now account for 79x of the 65,000 lowland colonist families (see Edelman, 1967: Stearman, 1973: Henkel, 1982: and Fifer, 1982]. The government also promoted the establishment of foreign colonies, principally Japanese and Mennonite ones. Japanese and Okinawan colonists began arriving in 1953 under an agreement between the Bolivian, Japan- ese, and 0.5. governments (they had been displaced by the construction of a U.S. airbase). Currently, there are about 430 such families in four settlements northwest and northeast of the city of Santa Cruz (CORDE- CRUZ, 1982a]. Credit and technical assistance from the Japanese govern- ment have facilitated agricultural mechanization to produce soybeans and wheat. The average income of the Japanese farmers is about fives times that of Bolivian colonists (Kashiwazaki. 1983]. Similarly, the Mennonite colonists began arriving in 1954 from Paraguay. Since then. several waves have arrived from Paraguay, Mexico, and Canada. About 2,100 Mennonites families have settled in eleven colonies, most of which are southeast and northeast of Santa Cruz (CORDECRUZ, 1982a]. In spite of their social isolationism and cultural rejection of passenger vehicles and modern household appliances, they 42 are not reluctant to use heavy land-clearing equipment, tractors, and combines. They produce livestock, poultry, and dairy products as well as soybeans, maize, and wheat. Like the Japanese. they tend to have larger cultivated plots and higher average income than the nationals. Although the Japanese and Mennonite farms tend to be larger than those of the Bolivian colonists, averaging 80 ha compared to 19 ha. the largest farms are the Bolivian "agricultural enterprises." The Land Reform Act of 1953 allowed for the retention of holdings of up to 2,000 hectares by "agricultural enterprises," defined as farms with large capital investments, wage labor, and modern technology. These farms produce sugarcane, cotton, soybeans, and wheat (Fifer, 19821. Currently, most of the colonies, agricultural enterprises, and other farms are concentrated within 200 km of the city of Santa Cruz. Rice, maize, and yuca are produced by small farmers to the north of Santa Cruz, approaching the foothills of the Andes Mountains, where the rainfall is relatively high (about 2000 mm). Rice has become an import- ant cash crop following its expansion in the late 1950s, partly due to the production of the Japanese colonies. Currently, rice is cultivated on 35,000 hectares in the department by about 16.000 growers (CORDECRUZ et al, 1982b: 61]. Large-scale production of sugar cane, the most important commercial crop in the department, is also concentrated to the north of the city. The volume produced grew considerably in the late 19503 and again in the early 1970s, reaching an area of 60,000 hectares and an output of around 1.9 million metric tons of refined sugar. The cane harvest provides seasonal employment to some 15,000 workers, the majority migrating from the highlands (CORDECRUZ et al, 1982b: 459-460]. 43 Mechanized production of cotton, soybeans, and, recently, wheat is more common between Santa Cruz and the Rio Grande to the east where rainfall is lower (about 1000 mm). Cotton, only an experimental crop in the 1950s (Heath. 1959], experienced a boom in the early 19603 and again in the early 1970s. at which time its acreage expanded six fold in two years. Lower international prices and problems due to Bolivia’s viola- tion of an international trade contract have led to a decline in crop area to its 1970 level of about 7,000 hectares. Soybeans have become the latest boom crop, stimulated by high international prices in the early 1970s. It has grown from covering 1900 _ hectares in 1972 to 40,000 hectares in 1981. This has allowed import substitution of vegetable oil (CORDECRUZ et al, 1982b: 5231. Following the expansion of soybean production in Santa Cruz, wheat production has also increased rapidly in the lowlands as a winter crop. This expansion has been facilitated by the fact that it can use the same lands and machinery as soybeans. Wheat production will be considered in more detail in the next chapter. 3.3 Seed industry in Bolivia 3.3.1 Overview of seed programs Following the system employed by the FAO 1979-80 seed review, the Bolivian seed industry would be classified as having "fragmented or pilot scale operations“ (Feistritzer, 1982]. For this reason. it makes more sense to speak of various seed programs rather than the Bolivian seed industry. Much of the fragmentation is due to the varied agro- A climatic zones found in Bolivia and the weak transportation links between them. In addition, seed industry development has been slow due 44 to the predominance of small-scale semi-subsistence agriculture which limits the effective demand for purchased inputs such as seed. Further- more. agricultural research in Bolivia has been hampered by inadequate support and duplication of effort. Agricultural research and extension activities were initiated after World War II by the Servicio Agricola International (SAI), a joint U.S.- Bolivian organization. This structure was similar to other SAI units operating in most Latin American countries until the early 1960s (see Rice, 1974]. It established ten research stations in Bolivia between 1946 and 1964. Also paralleling the experience elsewhere in Latin Amer- ica, the transfer of these functions to local agencies in the early 1960s and the reduction in U.S. technical and financial support severely weakened both research and extension. As a result, some of the research stations have discontinued research, while others have limited their activities and are forced to operate as commercial farms to generate revenues (Wennergren and Whitaker, 1975: 250-254]. A 1980 USAID report noted that IBTA, the research and extension agency, has “spread its research personnel and financial resources far too thinly among too many commodities and too many experiment stations." It also observed that researchers had insufficient contact with farmers and were isolated from farm-level problems (USAID. 1980a: 70-71]. In Santa Cruz in 1975, these problems led to the creation of the Center for Agricultural and Technological Research (CIAT) which has displaced IBTA in the department. CIAT enjoys support from the Regional Development Corporation of Santa Cruz (CORDECRUZ) and other local institutions. In spite of the problems mentioned above. some improved cultivars have been released and disseminated. Cuban Yellow maize was released in 45 the 1950s and is now the principal yellow maize cultivar in the low- lands. Improved cultivars of rice are widely used in Santa Cruz. Wheat and certain types of potatoes are now dominated by cultivars introduced since 1960. The Mexican wheat cultivar, Jaral, has become widely used since its introduction in 1966. However, the use of improved cultivars does not indicate the existence of a sustained seed industry. Some cultivars have been disseminated by farmer multiplication with little or no renewal with certified seed (Wennergren and Whitaker, 1975: 114]. Attempts have been made to establish a wheat seed industry since the late 1960s when considerable attention was focused on wheat produc- tion by the newly formed National Wheat Institute, the USAID-Utah State University preject, and the Ministry of Agriculture and Campesino Affairs (MACA). Although a locally-bred high-yielding cultivar, Coposu. was originally promoted, it became clear that farmers preferred Jaral for its superior disease resistance and the premiums offered by the mills for this semi-hard cultivar (Gardner, 1968 and 1970]. A Seed Department in MACA was formed in 1970 to control the mar- keting of seed, formulate and enforce seed regulations, and promote seed production and distribution. In spite of this limited mandate, it became directly involved in seed production using both its own land and that of contracted growers. The Seed Department then processed the seed in its own plants, first in Cochabamba and later in Warnes (Santa Cruz) and Betanzos (Potosi). It also certified and distributed its own seed. Data on the volume of seed production are available for the first years. From 1970 to 1972, the volume of improved wheat seed sold rose from 177 MT to 384 MT, although in the latter two years several hundred tons of seed went unsold (Wennergren and Whitaker, 1975: 115]. 46 Data are available on the proportion of crop area “under improved varie- ties” for the period 1975-81. They show that the crops with the highest proportions of improved seed area were those depending on imported seed: soybeans, cotton, and sorghum. On the other hand, the proportion of area under improved cultivars was generally less than 7x for wheat and maize and near zero for pasture, barley, and potatoes (Morales. 1984: 5]. Furthermore, these figures probably refer to land cultivated with up to fourth generation saved seed, rather than that cultivated with certified seed. Thus, they correspond to an annual production of only 100-200 MT. In 1980, in an appendix to a report on Bolivian grain storage, a Kansas State University consultant noted the following: The capacity of this program to produce good seed needs to be increased. At present the supply cannot be met [sic]. The multiplication program is operating in the red (not self- supporting). It is a well-organized program but needs more support to increase the supply of good seed. (Hugo et al, 1981: Appendix G] It should be noted that this evaluation was based on a six-day trip accompanied by the director of the Seed Department. 0n the other hand, in conversations with the author, observers close to the seed program have contradicted this assessment. They report a variety of counter- productive policies within the Seed Department that have impeded the development of a more effective program. First, the volume of seed produced has been severely limited by the reluctance of the Seed Department to promote or even permit other enti- ties to use its processing facilities, in spite of the fact that budget allocations severely constrain the amount of seed it can directly pro- duce. The Cochabamba processing plant, the largest and best equipped in Bolivia, was, as of 1982, "still operating at less than 25% of its capacity“ (CORDECRUZ, 1982c: XXII-1). 47 A related problem is its uncooperative attitude toward private sector interest in seed production. The Seed Department recently refused to consider testing barley seed for a beer company even on a fee basis. The beer company had been interested in improving the quality of its raw material by initiating a private multiplication program. Furthermore, MACA sought external aid to construct as many as five seed processing plants in those regions without facilities, paying little attention to promoting seed production where plants already existed. In Zudafiez (Chuquisaca), USAID-financed processing equipment remained in its crates for over five years because the Seed Department would not run 100 meters of electrical lines to provide electrification for the facility. Serious problems were also reported in Santa Cruz. as described in the next subsection. 3.3.2 Seed programs in Santa Cruz Until 1981, seed production in Santa Cruz was conducted by MACA and followed the pattern established in Cochabamba and other parts of Boli- via. MACA contracted the production of seed, cleaned it with its own processing facilities, and sold it to interested farmers. Seed growers were not provided any special seed. so the system was essentially one of supplying clean grain rather than varietally pure seed. Under this system, the CIAT Experiment Station at Saavedra had no incentive to produce breeder or foundation seed. There is no information on the volumes of seed produced, but it is likely they were quite small. One source reports that no seed was pro- duced in Santa Cruz in 1980 (Certificacion, 1984a]. In part. this is due to the limited number of MACA personnel assigned to carry out this 48 function: one agronomist and two assistants. They were fully occupied by seed processing and were unable to devote much effort to field inspec- tions or seed testing (CORDECRUZ, 1982c: XXII-2]. More seriously, several people acquainted with the program indicate that there were abuses of the system by MACA personnel. Specifically, it was reported that some seed growers were told their fields did not qualify for seed production and were paid grain prices for the harvest. yet the grain was processed and sold as seed. Another alleged practice was to falsify records to show that seed prices were paid to growers when in fact they were not and that sacks were sold as grain when in fact seed prices were charged. Thus, the need for government support was exagerated for the private gain of MACA employees. With this background, it is not surprising that the seed program made little progress in convincing farmers of the usefulness of pur- chasing "certified" seed. In fact, it contributed to the poor image of MACA. already mistrusted for its affiliation with the government. Since 1981, the institutional arrangement of the Santa Cruz seed program has changed significantly under the guidance of the USAID Agricultural Sector PrOject II. This prOject involved a technical assistance contract between MACA and Chemonics International. a U.S. consulting firm. In September 1980, Chemonics assigned Dr. Adriel Garay, a U.S.-educated Peruvian seed specialist, to the MACA office in Santa Cruz. Initially. he worked to strengthen seed testing procedures. showing in the process that the seed produced by the MACA program and even the imported soybean and cotton seed was often of poor quality. In 1981. Dr. Preston Pattie, the prOject director. and Dr. Garey wrote a policy paper on seed program planning in Bolivia (Garay and 49 Pattie, 1981]. They made the following recommendations: 1) that seed production must be based on successive multiplication of varietally pure seed provided by the experiment station, 2) that greater effort be made to stimulate private sector seed production and distribution, such as by making processing facilities available on a fee basis, 3) that the construction of new processing facilities be dependent on progress in promoting seed production, 4) that, to the extent possible, MACA restrict its role to seed certification, and 5) that initial efforts to develop the seed program be regional. Although well received among the local agricultural institutions (e.g. CORDECRUZ, 1982c], there was resistance from the Seed Department. In addition to eroding its control over seed production, the proposal would promote the involvement of the private sector, an idea with little acceptance in MACA and in the leftist government as a whole. Persistent lobbying, particularly by the organized commercial agricultural sector of Santa Cruz, eventually led to the removal of the national director of the Seed Department and his replacement by someone less hostile to the proposed reforms. Several important changes were instituted in 1982. First, a Santa Cruz Regional Seed Board was formed to coordinate the seed program. establish quality control standards, and set foundation seed prices. It is composed of representatives of MACA, CORDECRUZ. CIAT, the Agricul- tural Chamber of the Oriente, and the professional association of agro- nomists (Zegarra, 1983: 77]. Second, a Regional Seed Certification Service was created as a branch of MACA, but with financial support from local institutions. This unit, composed of a small crew of young agronomists, works exclusively on seed certification and strives to "protect the good image of the 5% [new] Service with seed growers and farmers" (Zegarra, 1983: 79]. For field inspection and laboratory testing, it charges 3% of the final seed value, collecting from both public and private producers. USAID- Chemonics has worked closely with the Service, providing technical assistance and equipment and arranging training in seed technology at the International Center for Tropical Agriculture (CIAT) in Colombia. Third, the Foundation Seed Unit was also established in 1982. Affiliated with the Saavedra Research Station, this unit is responsible for the production of high-quality foundation seed. generally through contracts with private growers. And finally, the administration of the processing plant at Warnes was transferred to CIAT. To achieve this. it was argued that civil service regulations were not compatible with plant operation and that: the functions of certification and processing [would] convert MACA into "judge and defendant." In addition, the very limited resources of MACA require that it concentrate its efforts on one single activity. (CORDECRUZ, 1982c: XXII-3] The plant now processes seed as a service available to both public and private entities, charging prices which roughly cover costs. The plant’s rated capacity has been expanded to about 3500 MT/year with equipment obtained through the USAID-Chemonics pr0ject. Initially the program concentrated on two crops: soybeans and yellow maize. Soybean seed production has been carried out in both the summer (45%) and winter (55%) seasons for commercial soybean production in the summer. The program provides four Brazilian cultivars and has reduced Bolivian soybean seed imports from close to 100% to about 25% of consumption. Several private seed firms and the cooperative of Japanese farmers (CAICO) are active in soybean seed production while distribu- 51 tion is handled by ANAPO, the oilseed growers’ association. Maize seed production has also attracted the participation of a7 private seed enterprise. This firm buys foundation seed, multiplies the seed using contract growers. has it processed at the Warnes plant, and then distributes it. Two cultivars are currently offered. Later, rice and wheat seed were incorporated into the program. These are still handled by CIAT because. as yet, no private entities have begun production. Wheat seed is produced by small-scale contract growers in a highland area of Santa Cruz. The elevation (1500-2500 meters) provides a climate more suitable for seed production and permits summer wheat cultivation (December-April). The harvested seed is brought to the Warnes plant for processing and is then immediately distributed in the lowlands for winter season wheat production (May-September). The wheat seed program is described more fully in Chapter Four. Underlying the reforms instituted since 1981 are the explicit goals of generating self-financing public institutions, promoting private seed production and distribution systems. and coordinating these various institutions to provide high-quality seed to the farmer. The results of these reforms have been successful on several fronts. Certified seed production has expanded steadily, as shown in Table 3.1. In addition to the eight private seed enterprises which have been formed since 1981, the program has also stimulated interest in privately operated seed processing plants. The cooperative of Japanese farmers (CAICO) has completed construction of a processing plant which is expected to concentrate on soybean seed. CAICO has also received tech- nical assistance from the USAID-Chemonics prOJect. Similarly, a private seed enterprise just completed a processing plant for maize seed. The 52 TABLE 3.1: CERTIFIED SEED PRODUCTION IN SANTA CRUZ (HT) Soybean Wheat Rice Haize TOTAL 1980 - - - 0 1981 500 - - 100 600 1982 1300 162 200 100 1762 1983 1850 250 150 150 2400 1984 2100 396 700 800 3996 SOURCE: ANAPO for wheat seed (does not include seed stored in 1984). Chemonics International for others. construction of these plants has been stimulated by the rapid growth of the program, surpassing the processing capacity of the Warnes plant. and by the processing fees at Warnes which have been established to cover capital and operating costs. Although there are no private entities currently involved in wheat seed production, one seed company bought a small quantity of seed in 1984 to determine the feasibility of entering wheat seed multiplication. This is made possible by the price structure of the public program which roughly covers costs, thus allowing private sector competition on an equal basis. and by the availability of certification services and processing facilities for hire. In addition. the development of private seed enterprises with experience in other crops facilitates diversifi- cation into wheat. Naturally, much of the success of the seed program is due to the fact that the Santa Cruz area was "ripe“ for the development of a seed industry. It had a large-scale commercial farming sector, fairly well developed agricultural institutions including growers’ associations. and an important industrial crop (soybeans) for which the demand for pur- 53 chased seed was already established. However, it is also apparent that the reforms which explicitly promoted the participation of the private sector in seed production and distribution have been essential in real- izing this potential. Furthermore, although promoted by the large-scale commercial sector, the seed program is also benefitting small farmers, both the wheat seed growers in the highlands and small rice and maize farmers in the lowlands. CHAPTER FOUR DESCRIPTION AND DIAGNOSIS OF THE LOWLAND WHEAT SEED MARKET This chapter focuses on the wheat seed market in the lowlands of Santa Cruz, beginning with a review of the market for wheat grain in Bolivia. It is primarily descriptive and diagnostic. providing the groundwork for later chapters which generate prOJections of future trends and prescribe strategies for improving the supply of wheat seed. The information in this chapter was gathered from secondary sources. interviews with Bolivian professionals, and a survey of lowland wheat farmers conducted by the author in July 1984. Using lists of farmers who purchased certified wheat seed. the survey was designed to weight the responses proportional to the volume of seed bought. The methodology of the survey is described in Annex B. 4.1 Wheat in Bolivia In the 1930s, Bolivian wheat production constituted 60-703 of national consumption. Since that time, production has doubled but con- sumption has expanded eight-fold, so that imported wheat and flour now provide 85% of the total (see Figure 4.1). Large industrial mills, which supply the urban areas, obtain 90-95x of their wheat from imports. In contrast. most local wheat is consumed outside the cities, as flour from small-scale mills of less than 10 hp, in soups and "chicha," and as animal feed. Thus. it is not surprising that wheat policy has focused on raising national production to substitute for wheat and flour imports and to increase the local portion of wheat purchased by the large mills. 54 55 .7? 0.00.... 00m uMUEDOm so.) 000—. . mhaw 050—. mOOF Dmmw mmmw 0mm”. 13%» s 4. s L. .rJL s: A...» L. i l...l.n.l.+.fl4.vlT.lJllllillT.l+l+l+llo ................... 2:4 .. «00:3 .9530: 00.25:: . ..... ......................... ..... Hon I ..a: s. a s nnnnnnn u U .. // +00. .\ l .30: Detours ._.. .. [I . mu .m). ..... .. I\. m 7... . Ir \. c w s. /\ z . uroor .\ I c _ \HU ... 1 0 . ...... ... 0 x I U o . ........ I _ I. \\ +03 m ...... I( m. > .. . a \ I, \\ ‘ 50.1 . .. idiom m30fi\»> 1.50» ._I .m \ \/\ one n. «0053 Detour»... 02:3: \\ . M x i N. \ . m. _ ( ... . been \/I\ ‘ F8. >4113m KDOJk\._.>. 255.300 “.0 20:501200 3...? UKDOE 56 TABLE 4.1: BOLIVIAN’WHEAT AND FLOUR PRODUCTION AND IMPORTS (1000 NEW National Milled Imported Imported wheat national wheat flour TOTAL production wheat (all milled) (wheat equiv.) 1950 45.70 13.60 33.90 17.30 103.63 1951 45.70 8.60 32.50 32.90 123.89 1952 40.00 5.00 73.70 16.40 136.48 1953 31.20 3.60 71.60 11.20 118.36 1954 31.20 7.20 58.80 19.40 116.94 1955 35.00 16.70 34.40 ~21.90 99.82 1956 37.00 2.70 13.20 14.80 70.76 1957 37.50 1.90 54.70 63.00 179.70 1958 35.00 6.60 3.40 64.90 128.54 1959 35.00 3.40 14.30 71.90 149.16 1960 35.00 4.00 .60 77.60 143.38 1961 35.00 1.40 4.50 101.50 180.47 1962 40.00 3.40 21.50 94.40 192.61 1963 55.20 6.00 20.00 103.00 218.26 1964 57.90 7.70 4.70 96.10 196.07 1965 35.00 3.50 15.60 108.90 201.85 1966 41.00 4.50 15.10 107.80 205.82 1967 27.00 4.00 26.60 127.80 231.10 1968 45.00 6.80 54.20 109.90 251.84 1969 53.00 8.00 40.60 109.80 246.10 1970 44.20 6.60 45.70 116.60 251.84 1971 47.10 7.10 46.30 123.30 264.65 1972 53.80 8.00 82.80 109.80 289.10 1973 57.00 8.60 68.20 61.50 210.62 1974 62.50 9.40 85.70 96.30 281.95 1975 61.80 9.30 65.40 133.80 313.03 1976- 69.80 10.50 97.40 96.50 301.23 1977 55.60 8.30 160.00 71.50 314.91 1978 56.60 8.50 243.20 56.00 377.58 1979 67.80 10.20 204.30 27.00 309.60 1980 60.10 9.00 276.70 4.00 342.36 1981 66.60 266.40 333.00* 1982 66.00 170.00 236.00* 1983 40.30 331.30 371.60* 1984 68.50 360.00 428.50* SOURCES: MACArUSAID, 1984; Torrico, 1979; and Hugo et al, 1981. NOTE: Wheat equivalence calculated using a 72% conversion rate. * 1981-1984 totals do not include flour imports. 57 4.1.1 Patterns in wheat production The traditional wheat growing areas include the Valleys and, to a lesser extent, the Altiplano. For example, the departments of Cochabamba and Chuquisaca accounted for about 653 of the 96,000 hectares of wheat planted in 1982 [MACA-USAID, 1984]. In much of this highland producing area, wheat must compete with corn and potatoes. Except in a few areas, it is grown by small farmers who use only small amounts of purchased inputs. Fertilizer, for example, is generally reserved for potato culti- vation. However, wheat production probably benefits from residual ferti- lizer in the soil. In the lowlands of Santa Cruz, wheat production has increased markedly in the last six years. It has grown from 200 hectares in 1978 to around 8800 hectares in 1984 (see Table 4.2). Lowland wheat area now accounts for about 10x of national wheat area. In contrast to highland production, wheat in the lowlands is produced in the winter when the temperatures are low enough for this temperate climate crop. Another difference is that wheat in lowland Santa Cruz is grown on larger hold- ings (10-200 ha) and planting and harvesting are mechanized. It is frequently grown in seasonal rotation with soybeans, thus taking advan- tage of the fact that both crops use the same agricultural machinery. Historically, most of the growth in Bolivian wheat production has been the result of increases in cultivated area. Yields average 700-800 kilograms per hectare, less than half the average for Latin American which is 1630 kg/ha [International Wheat Council, 1984]. In the low- lands, average yields are around 1000 kg/ha. Perhaps the most intensive effort to promote wheat production in Bolivia occurred in the late 19603 and early 19705. The United States 58 Agency for International Development (USAID) contracted Utah State Uni- versity to provide technical assistance to the Ministry of Agriculture and Campesino Affairs (MACA). Although the first Utah State reports were pessimistic about the prospects for self-sufficiency in wheat [Dewey and HcAllister, 1966 and Gardner, 1966a,b], recommendations were made for reducing wheat and flour imports. From 1966 to 1973 a wheat preject within the Utah State contract focussed on varietal testing, fertilizer- response trials, and seed multiplication. Wheat area did not increase appreciably over this period, but yields did rise from around 550 kg/ha to around 800 kg/ha. This is probably the result of the adoption of new cultivars of wheat, in particular the hexican cultivar, Jaral, intro- duced in 1966. Two Utah State economists calculated a negative ex posts rate of return to the progect, arguing that USAID ignored indications that the sheep prOJect provided superior returns [Wennergren and Whitaker, 1977]. Studies of the returns to crop production in the highlands have generally concluded that wheat is less profitable than competing crops at prevailing prices. For example, a Kansas State University report reviewed comparative return studies from 1964, 1971. 1973. 1978, and 1980 and concluded that: ...gross less variable costs are considerably lower for wheat than for corn and potatoes. Wheat yields would have to double for returns to equal those to corn, and to equal the returns to the cultivation of potatoes wheat yield would have to be eight times greater. [Hugo et al, 1981: 17] Although conversations with farmers confirm that wheat is not a "pre- ferred" crop, the fact that presumably rational farmers continue to grow it indicates that it must be more profitable than other crops in some situations. 59 This apparent paradox is probably explained by the fact that corn and potatoes are less profitable than wheat on low-fertility soil. Thus, wheat is relegated to areas with naturally poor soil and to land which has already been depleted by corn and potato production. A common rota- tion seems to be potato-corn-wheat-fallow [Gardner, 1966a]. Although no comparative rate of return studies have been done for lowland wheat production, very few crops are grown in the winter. Soybeans for seed and sorghum for on-farm animal feed are currently the only alternate uses for the land. 4.1.2 Wheat marketing and price policies Wheat policy appears to pursue the goals of promoting wheat produc- tion to reduce imports and maintaining low consumer prices for wheat products. The history of wheat programs and regulations reflect the inherent conflict between the goals, the frequent changes in government (16 since 1964), difficulties in enforcement, and strong political pressure on the price of bread. Pricing of wheat and wheat products in Bolivia is more complex than for any other commodity. Both consumer oriented and producer orien- ted policies have been implemented during the past twenty years... The price of bread is alleged to be one of the most politically sensitive factors in the country. [Gardner, 1974: 25] Figure 4.2 compares Bolivian producer prices with various estimates of the price actually paid for wheat imports. The producer price is the average price paid to farmers at the farm gate. Although the estimates of import price vary widely, it appears that producer prices were roughly comparable to the import parity price until 1971 and again from 1976 to 1981. They were below import prices for the period in between. The comparative advantage of local wheat relative to imported wheat varies considerably among regions due to high transportation costs. 60 .nNnap ..0 as NDKONDKOU 3:0 “ramp .5 no 0031 Reacto>v .O<... ummomfiom too> Oflmp Chan Chap .94er the Ohmw mmmp mew mew r P .lllfillIT'lol’-ll+lvlu+ul| Jul Ifil w a «4 I“ 1J1 a J A J- a [a 0 04055 outs toe—t. ..ll... nh0.2v 007:... Court. .......... Ao30m “NJ. mmDOE 61 In the early 1960s, several efforts were made to promote national wheat production. Supreme Decrees established that 1) all importation of wheat grain and flour had to be authorized by the government and would be used only to cover “deficits" in production relative to predetermined levels of consumption, 2) importers were required to sell one unit of national flour for every two units of imported flour, and 3) a tax was placed on imported wheat flour equivalent to about 12x of the wholesale price that year [Bedoya, 1970: 9-11]. The first measure would only constitute import protection to the degree that consumption levels were held below what they would be with free trade, which, of course, is politically difficult. The second measure was notoriously difficult to enforce in spite of the two subsequent decrees reiterating the policy. Furthermore, contraband wheat flour from Argentina diluted the impact of all three measures [Gardner, 1970: 7-8]. In the late 1960s, the National Wheat Institute (INT) was created and a system was established to provide producer support prices. None- theless, the incentive effect was minimal. In the first place, studies indicate the wheat farmers are seldom paid the official support price. In addition, since the support price was implemented in 1968, it was never more than negligibly higher than the CIF price of Argentine wheat until 1973. [Gardner. 1974: 31] In 1973 and 1974, the "support" prices were raised several times, gen- erally lagging behind rising international prices (see Figure 4.2). International prices fell below the support price (fixed in nominal terms from 1974 to 1980) for a few years, although it is unclear if the mills or the government were effective in enforcing the support price. ... millers are anxious to avoid purchasing domestically produced wheat at higher prices and appear to be successful in exercising various pretexts to reject the offerings of the limited domestic wheat which is for sale. (USAID, 1976: 12] 62 In an attempt to enforce the support price, the Wheat Harketing Division (replacing the disbanded National Wheat Insititute) operated a system of five purchasing and storage centers. In spite of sizeable subsidies from USAID PL 480 Title III funds, the centers suffered from poor location, faulty design (e.g. bulk storage instead of sack storage facilities), low volume and consequent high overhead rates (30-60x of sales). In addition, there were strong indications of graft [Hugo et al, 1981: 29- 49]. Since then, at least three of the 616 HT storage facilities have been abandoned. Over the period 1975-1980, there was a tremendous shift toward local milling: the volume of imported flour fell 96x (from 133,800 to 4,000 HT), national flour production more than quadrupled (from 53,800 to 232,300 NT), and the volume of imported wheat also quadrupled. The volume of national wheat milled remained constant at around 9,000 to 10,000 HT, so that its participation in national milling fell from 13x to 4% (see Table 4.1 and Figure 4.1). Since 1981, the inflation rate has risen to unprecedented levels. Using 1981 as the base year (1981:100), the consumer price level in the second quarter of 1984 was 11,200 [Banco Central, 1984]. Combined with flour and bread price ceilings which can only be adjusted at great poli- tical cost, this has created downward pressure on real producer prices. Flour prices are far below market value, necessitating strict rationing of flour to the bakeries. In spite of the controls, large amounts of flour are smuggled into Peru. Although the hyperinflation makes any calculation difficult, in July 1984 the producer price for wheat was U59 108/MT at the parallel exchange rate. This is only 75s of the F.O.B. price of Argentine wheat 63 and 60: of the C.I.F. price of U.S. wheat in Rotterdam. The price that the flour mills actually pay for imported wheat is even less since they are able to buy dollars at the official exchange rate, a 15-50x discount relative to the parallel rate. In addition, the government gives them a 90-day grace period in paying for imported wheat which amounts to a sizeable discount in real terms (over 50x at the 1984 inflation rate of ZIOOX). This distortion reduces their incentive to buy national wheat and imposes a real cost on the Bolivian economy. 4.1.3 Wheat production in Santa Cruz In the 1960s, the possibility of wheat production in the lowlands of Santa Cruz was little more than an idea. In one of the first Utah State reports on wheat production, Gardner [1966b: 16] argued that: The ideal way to take Bolivia out of the wheat import market would be to open up a vast new area of productive land somewhere in Bolivia where wheat could be grown as a specialized primary crop, rather than as crop of secondary importance relegated to left over lands. Such an area would be level enough to permit irrigation... Regarding the Santa Cruz area, he mentioned that "reports, both hot and cold, regarding the potential for wheat production in this vast new agricultural area have been filtering out of that region" [p. 6]. In an annex to the same report, another author was less optimistic. In 1966, he visited a Mennonite colony that had experimented growing 100 hectares of wheat for home consumption the year before. Disease and lodging had been serious problems, and only 70 hectares had been worth harvesting. He concluded that: The results are in full agreement with what I would expect whenever wheat is forced outside its major area of adaptation... Histori- cally, such efforts have not met with success and I expect such effort will continue indefinitely with subsequent failures... In my opinion. Bolivia does not have the physical land area necessary to produce its total wheat needs. [Gardner, 1966b: 30, 31, 36] 64 In 1968. another study attempted to show that Bolivia might be able to produce 703 of its prOJected wheat needs by 1978 if it could increase yields by 124x and double wheat area. However. lowland production was not mentioned [Gardner, 1968]. Two years later, the same author reviewed the progress made since 1966 and acknowledged that "on the whole... yields have fallen below those levels that I thought possible to reach" [Gardner, 1970: 5]. He noted that: The one great unknown in the total national wheat picture is what might happen in future years in the Department of Santa Cruz... One farm near Portachuelo has grown wheat for more than five years and this year has about 100 hectares planted. Many other farmers are experimenting with wheat... The advantages of wheat production in this area are tremendous, providing disease-free. high quality varieties are grown. The farms are large and mechanization already ' prevalent. Rice seeding and harvesting equipment may be used for wheat also. Perhaps the biggest advantage of all is that wheat can be double-cropped with rice or cotton, thus greatly reducing land costs. (Gardner. 1970: 31. 321 At the time, experimental trials using ten cultivars of wheat had begun in the lowlands of Santa Cruz. One of the earliest efforts to analyze the feasibility of wheat production in the lowlands was a 1971 study which looked at the issue of water availability. It assumed seasonal rotation with cotton and. look- ing at rainfall probabilities and evapotranspiration rates. concluded that "there seems to be a reasonable probability of producing a fairly good crop about six years out of ten" [Hargreaves. 1971: 4). A 1976 USAID report on wheat policy commented on the situation and prospects for lowland wheat production: wheat production in the Oriente has. to date, been largely experi- mental...some recent research results indicate favorable prospects for wheat husbandry in the Santa Cruz region in the winter months (May-Sept.). [but] any significant increase in the crop will require some form of irrigation... If there is a role for wheat in the crop rotation. it will be that of a winter double-cropping enterprise in combination with cotton. soybeans. or perhaps summer grown cereals. [USAID. 1976: 10] 65 The report noted that the Abapo-Izozog research prOJect, located 150 kilometers south of the city of Santa Cruz. was the only major irriga- tion effort in the region. Since the proaect was only sustainable with substantial government subsidies, the report concludes that "it is unlikely...that any rapid expansion of wheat production will occur in this area" Ep. 11]. The Center for Agricultural and Technological Research (CIAT) was formed in the mid-1970s to take over research and extension functions in Santa Cruz. In 1978. it began regular agronomic testing of wheat. using 1,888 lines and cultivars from the Cochabamba research station. A new cultivar. Saguayo, was identified which performed even better than Jaral, the most common cultivvar by that time (CIAT. 1979]. In 1980. a “Departmental Plan for Wheat Production" was prepared which called for wheat area to expand to 80,000 hectares and for yields to be improved so that. by 1985. Santa Cruz would be producing more than the 1980 level of national wheat production [MACA et a1. 1980]. This plan. like a similar one proposed a decade before [Romecin. 1970]. never came close to fulfillment. A 1981 Kansas State University (KSU) report made the following observations concerning lowland wheat production: Mechanization of wheat is increasing... The Santa Cruz area is completely mechanized - large fields. large cultivation equipment. seeders, and combines...it is double cropped. being grown in the winter following cotton. maize. or soybeans. [Hugo et al, 1981: Appendix G] Citing the Departmental Plan. the KSU report was more cautious: there seems to exist an unjustified optimism, generated in certain quarters. in the belief that the Santa Cruz area can become the "breadbasket" of Bolivia through the implementation of large-scale agricultural prOJects. [Hugo et al, 1981: 69] 66 A key obstacle was believed to be the availability of seed. Seed is a problem for wheat [in the lowlands of Santa Cruz]. Due to climate. quality of seed deteriorates from harvest one season to planting time the next season. Currently. seed is being obtained from Cochabamba for the area north of Santa Cruz. [Hugo et a1. 1981: 200] A related problem is that tropical production of wheat is vulnerable to disease. primarily seedborne diseases such as Helminthosporium sativium. According to the data presented in Table 4.2. it appears that wheat area in the lowlands has expanded rapidly since 1978. It should be noted that the data are contradictory: one source assumes that departmental figures to refer to lowland production. another has 3272 MT for the 1980 lowland sales to the mills. and the MACA-USAID [1984] figures are impos- sibly low for 1983 and 1984. The lowland figures for 1982-1984 in Table 4.2 are estimates based on data collected by the author. The survey of lowland wheat farmers conducted by the author in July 1984 sheds some light on the patterns of wheat production. Among the 39 Santa Cruz farms surveyed. the area cultivated ranged from 15 to 2000 hectares with the median being 65 hectares. Differences were found among the Mennonite, Japanese. and Bolivian farmers as demonstrated in Table 4.3. The bulk of the Mennonite farms surveyed had less than 50 hectares while larger farms were more common among the Japanese and Bolivian wheat farmers. It should be noted that this sample over-represents large farms relative to the population of lowland wheat farms. A brief overview of the summer cropping combinations is shown in Table 4.4. Soybeans were grown on 32 of the 39 farms in the sample. In only three of these farms was another crop grown on a larger area than soybeans. Soybeans were more likely to be grown with rice on the Japa- nese farms and with maize on the Mennonite farms. This difference may be 67 TABLE 4.2: WHEAT AREA AND PRODUCTION IN SANTA CRUZ 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 Dept. of Santa Cruz Dept. of Santa Cruz (lowlands and V. MEsotermicos) (lowlands only) Area ‘Yield Production Area Yield Production (ha) (kg/ha) (MI‘) (ha) (kg/ha) (MT) 1675 789 1322 1990 708 1408 2388 670 1600 4170 626 2610 5520 880 4860 6430 941 6050 6285 1091 6860 6785 1010 6855 7200 810 5830 200 900. 180 8000 950 7600 1680 952 1600 8000 980 7840 1700 1600 2720 8000 980 7840 5000 1000 5000 4800 6476 561 3636 6640 8829 SOURCES: CORDECRUZ et al (1982b) for departmental figures. Hugo et al (1981) for lowland figures 1978-1980. Table 4.6 and assumption of a 75 kg/ha planting for lowland figures 1982—1984. MACArUSAID (1984) figures show departmental wheat area of 723 ha for 1983 and 1,578 be for 1984. Lowland wheat production for 1982-1984 is not included for lack of . reliable yield estimates. (I! (D TABLE 4.3: FREQUENCY DISTRIBUTION OF THE CULTIVATED AREA FOR SAMPLE FARMS Cultivated area Japanese Bolivian Mennonite TOTAL 10 - 49 ha 1 1 13 15 50 - 99 ha 3 2 7 12 100 - 499 ha 4 4 1 9 500 - 999 ha 1 1 0 2 >1000 ha 0 1 0 1 TOTAL 9 9 21 39 SOURCE: Survey results. see Annex B for methodology. TABLE 4.4: FREQUENCY DISTRIBUTION OF SUMMER CROP COMBINATIONS Crops Japanese Bolivian Mennonite TOTAL No crOp 0 2 0 2 Soy alone 4 1 1 6 Soy-Maize 1 1 14 16 Soy-Rice 2 1 0 3 Soy-Maize-Rice 2 1 4 7 No Soy 0 3 2 5 TOTAL 9 9 21 39 SOURCE: Survey results. see Annex B for methodology. TABLE 4.5: FREQUENCY DISTRIBUTION OF WINTER CROP COMBINATIONS Crops Japanese Bolivian Mennonite TOTAL Wheat alone 2 4 16 22 Wheat-sorghum 1 1 2 4 Wheat-Soy 2 2 0 4 Wheat-Sorghum-Soy 4 1 1 6 No Wheat 0 1 2 3 TOTAL 9 9 21 39 SOURCE: Survey results. see Annex B for methodology. 69 partly a result of the climate (the Japanese colonies are in a higher rainfall area than most of the Mennonite colonies) and partly the result of cultural factors. In the winter season. wheat competed with soybeans for seed and. to a minor extent. sorghum and beans (see Table 4.5). The survey reveals that the Mennonites wheat farmers were less likely to grow other winter field crops. only occasionally growing one or two hectares of beans or 5-10 hectares of sorghum on the side. The Japanese in the sample were more likely to grow wheat with sorghum and winter soybeans. though wheat still dominated. The Bolivians were in an intermediate position. The survey results also suggest that wheat production could be stimulated if wheat prices were higher. Growers were asked how many hectares they would have planted if the price of wheat had been 20% higher. Some 17 growers reported that they could not expand wheat acreage in the short run. generally because of the lack of available cleared land. and several others would not respond because of the hypothetical nature of the question. However, the average increase in wheat area. weighted according to the volume of CIAT seed purchased. was 38% over the current level. If accurate. this corresponds to a short- or medium-term supply elasticity of +1.8, higher than many estimates of agricultural supply elasticity [see Askari and Cummings. 1977]. Although this method is only speculative. the conclusion that lowland wheat production is relatively price elastic is supported by the observation that there are large tracts of land already under summer soybean cultivation which lie fallow in the winter. The availability of 70 fields and agricultural machinery for wheat production suggests that wheat area could be expanded with only moderate improvement in the economic incentives for wheat production. 4.2 Wheat seed in Santa Cruz The three principal sources of purchased wheat seed in lowland Santa Cruz have been seed purchased from the traditional highland wheat regions. Proyecto Abapo—Izozog. and Proyecto Trigo of CIAT. The wheat growers'association (ANAPO) markets all CIAT wheat seed and some seed obtained from seed programs in the highlands. In addition. some seed is saved from harvest to planting by individual farmers. Each of these will be briefly discussed and the results summarized to estimate total wheat seed consumption. 4.2.1 Seed saved from the previous harvest The sub-tropical climate in the lowland Santa Cruz area is the most serious obstacle to on-farm storage of seed from harvest to planting. The airport outside the city of Santa Cruz records an average of 1170 mm of rain annually, 24.60 C.. and 69x relative humidity. However, the period of storage. roughly from October to June. corresponds to mid- summer when the rainfall rises to its maximum of 176 mm per month, the average temperature is 27° C.. and the relative humidity averages 74% (CORDECRUZ. 1980]. These conditions do not allow "safe“ storage of cereal seed for nine months [Delouche et al, 1973], but are close enough so that it should be possible some years. The cost of saving seed is the forgone value of the grain (around Sb 650/kg) plus costs of storage. interest, and risk. Interest and risk are probably the major components and the total cost of saved seed may 71 be expected to be 15-503 above the grain price. The survey results confirm the widely held impression that the practice of saving wheat seed from one season to the next is more common among Mennonite farmers than among the Japanese and Bolivian farmers. Only one of the nine Japanese respondents reported saving wheat seed. He had stored 900 kg (253 of his seed requirements) to supplement the seed from ANAPO. The germination rate was reasonable, and saving the seed allowed him to plant earlier than was possible with the seed from ANAPO. None of the other eight Japanese respondents used saved seed or even knew of others that had done so. Similarly, only one of the nine Bolivian nationals interviewed had saved seed. In this case. the exception was the largest grower in the sample. He saved 46 MT of Jaral wheat seed in silos. relying only on ventilation during cool or dry days to preserve the seed. Along with 138 MT of seed from ANAPO. the saved seed was used to plant 2000 hec- tares of wheat. His intention was to save 184 MT of seed to meet all his needs for 1985. Other large-scale Bolivian growers, including one with sufficient storage capacity to save seed. were aware of these efforts but believed that seed storage was not worth the expense and risk. On the other hand. seven of the twenty-one Mennonites surveyed had saved a portion of their seed requirements. However, six of these seven were found among the ten respondents located in the the two colonies farthest from the city of Santa Cruz: Valle Esperanza and Nueva Holanda. There was a consensus among the respondents in Valle Esperanza that around 508 of the colonies wheat seed requirements were met by saved seed (the question was not asked at the Nueva Holanda colony). On the other hand, various Mennonites in the nearer colonies independently 72 estimated that only 0 to sx of the planted seed was saved from the previous year. 4.2.2 Seed from other departments Until a few years ago. almost all wheat seed planted in the low- lands of Santa Cruz was obtained from the traditional highland depart- ments. principally Cochabamba and Chuquisaca. where wheat is grown in the summer [Hugo et al, 1981]. Since the initiation of Proyecto Trigo, this source of seed has become less important although the trade con- tinues through both formal and informal channels. Informal channels consist of shipments from the highlands arranged by Santa Cruz merchants. truckers. and some Mennonite colonies. For example. one agricultural supply store (Servagro) sold about 45 MT of wheat seed in 1982 and a similar amount in 1983. In addition. survey respondents reported that several Mennonite colonies had arranged the purchase of 10 “truckloads" (60-90 MT) from the highlands in 1983 and at least two more this year (see Table 4.6). This seed is frequently Just wheat grain bought in highland markets. though the Mennonite colonies generally clean it using home-made air-separators. It is purchased at grain prices. but transportation from the highlands adds 25-358 to this price (it is a 10-16 hour trip descending 2000 meters over mostly unpaved road). Among those surveyed, three farmers, all Mennonites, had bought Cochabamba seed from truckers or merchants. More formal marketing is organized by the growers’ association (ANAPO) which has purchased wheat seed from MACA in Cochabamba, Potosi, Chuquisaca. and TariJa (see Table 4.6). In previous years, these pur- chases were arranged to supplement overall production by Proyecto Trigo. This year, the purchases were used to supply farmers wanting to plant 73 early when CIAT seed was not yet available. The quality of this seed is often not much better than the grain referred to above. Varietal mixture and low germination have been problems. leading to the rejection of some 10 MT of “seed” in 1984. Although in previous years. this seed has been sold at a 15x discount below the price of CIAT wheat seed [ANAPO, 1982]. in 1984 it was sold at the same price (around double the grain price). Growers were presumably willing to buy this inferior seed at the same price because it was available earlier in the season. 4.2.3 Seed from Proyecto Abapo-Izozog Another source of wheat seed is the Proyecto Abapo-Izozog. This progect is administered by the Ministry of Defense with financing from the Inter-American Development Bank. It consists of a varietal testing program and mechanized agricultural production. much of it irrigated. by several hundred colonist in an isolated arid region 100 kilometers south of the city of Santa Cruz. The progect also produces wheat to sell to the mills in Santa Cruz using wheat seed of promising cultivars produced by the progect [USAID. 1980a: 59]. Any excess seed is sold to farmers outside the prOJect. generally Mennonite farmers also growing in this arid region. The wheat seed is cleaned with an air-screen cleaner owned by the progect, but is neither tested nor certified. Although CORDECRUZ [1982c] recommended that the preject specialize in seed production, the prOJect technical personnel prefer to produce wheat and sell only excess seed. As a result. the supply has been quite irregular (see Table 4.6). In addition. the seed is neither varietally pure nor thoroughly cleaned. On the other hand, it is available early in the planting season and it is less expensive 74 (about 40% above grain prices) than CIAT seed. In addition, it provides different experimental cultivars. generally ones adapted to the semi- arid conditions which prevail further to the south of the city of Santa Cruz. 4.2.4 Seed from Proyecto Trigo of CIAT The CIAT wheat seed progect. Proyecto Trigo, was initiated with financing from CORDECRUZ in the 1981-82 season. More recently, it received assistance from the PL 480 Title III program. Seed production is accomplished through contracts with farmers in the provinces of Vallegrande and Manual Maria Caballero in the department of Santa Cruz. These provinces are in a highland region known as the Valles Meso- termicos (Mesothermal Valleys), where the altitude (1500-2500 meters) allows summer wheat production (see Figure 4.3). The seed growers are small farmers averaging less that two hectares of wheat per family and located in twelve zones in the two provinces. Some of the zones (e.g. Comarapa and Vallegrande) are relatively flat and situated near major roads. but others (e.g. Higuera and Pampa Negra) are hilly and isolated, some lacking even unpaved road connection. CIAT provides these farmers with seed, technical assistance. and machinery for harvest and threshing where terrain and roads permit. The small size of the farms and the fact that not all the harvest can be collected by CIAT mean that large numbers of growers must be contracted, over 500 this year. Since Proyecto Trigo has only two field agents and the collaboration of two extension agents, it is necessary to hire farmer-cooperators to help in the peak work periods. planting and harvest. Similarly. the agent from the Certification Service hired seven 75 auto @5395 NNVRMs comm sums: 3.3 @533... 2355 some: 252 @2505 3mm ...(wzz 52 :32: ~33 Ezm>ssm ..E San a0 Me: page: 5? . xaé magma. mam. mas. B maa 6a... ma? 9. 625mm... Baum 94an onm acoumcoo mom—56mm comm ommmcohca no 53.6095 A3305 0: magnum 33 one 5.6.6 as. 8:38 38m: 6239...: 68.8 8 a: 530.6 mas 358.... 83893. cool u”Balsam aaa aav aaa aaq aaa aav aaa aav aaa ave aaa scauoaaoao zoo «as aoa aaa aaa aaa aoa oaa aaa «Ha aaa «as aaa acauomnoao asaaaz ooas acaa ooaa ooaa ooaa aoaa aaa vaa aaa aaa oaa acauomaoao some use mafia... 68m .28 aamm ease an as. as. an. as. as. as. as. aa. aa. aa. aa. aa. amaaoosm cesspoooaa aaa aaa aaa aaa aaa aaa sea aaa aaa aaa aaa acauoaaoao zoo aaa aaa aaa aaa ave aaa aaa aaa flea aaa aaa aaa coauomnoso sauna: aaas aaaa aaaa ave“ aaas asaa aaaa aaa aaa aaa aaa acauomnoso case 3.5 mommzomom Ommm comm ommmnowdm 3 aa. aa. aa. aa. aa. aa. aa. aa. aa. aa. aa. aa. amasodsm ceauaodosm aaa . Haa aaa Haa aha aaa aaa Haa aaa aaa .aaa cospomnoso zoo mace oaa aaa aoa Haa aaa aoa aaa ace ass aaa Haa coauomnomo sauna: oaaa oaaa oaaa aaoa Haas aaaa aaaa aaaa aooa aaa aaa coauomnoso nae: 2.2. 5305900 comm aaaa aaaa aaaa aaaa aaaa aaaa aaaa aaaa aaaa aaaa aaaa coauoaaoao zoo aaaaa aaoaa aaaaa aaoaa aaaaa asses avaoa aaaos fiaaa aaaa aaaa aaaa casuomaoao sauna: cocoa cocoa cocoa aooaa aavaa aaaoa amass aaaas aaaaa aaaaa mance coauomnoao gas: .as. mama yams: comm «mma mmmfi Nmmfi Hmmn ommfi mama mama hmmfi 0mm“ mama «can NED <25 92% 2H Em 5mm 92¢ .ZOHEmZOO am .552 BE Emhdmm "HA.“ “Human. 92 other sources of seed may decline as well. particularly grain purchases from the highlands. the prejections in Table 5.1 assume that the pro- portion supplied by CIAT rises to only 723. subject to the maximum of 1200 MT estimated in section 5.1. 5.3 Temporal distribution pf supply and gemand While timeliness is clearly a problem in the supply of CIAT wheat seed. more specific information is needed in order resolve the problem. It is necessary to quantify the demand for "early seed" in the lowlands. In this section, progections are made of the proportion of annual CIAT seed production which would have to be stored until the following year to meet this demand. This is done by comparing the temporal distribution of the supply and the demand of CIAT wheat seed. From the results of the survey. it is possible to construct two indicies of the temporal distribution of demand. that is the proportion of seed which would be planted each week of the planting season if seed were available. The first index is derived from the “ideal planting date" reported by the wheat farmers in the survey. If a range of dates was given. the midpoint date was used and the results were weighted according to the volume of seed purchased. The resulting distribution is expressed as a percentage of the total demand and presented as Alterna- tive A in Table 5.2. The second index is based on the actual date of planting in 1984 among those respondents who were not delayed in planting by the availa- bility of seed. Some of these managed to obtain CIAT seed early, others used other sources of seed such as saved seed. and still another group could not plant until late in the season due to weather and other factors. This temporal distribution was calculated as above and is 93 presented as Alternative B in Table 5.2. In evaluating these two indicies of the distribution of seed demand. it should be noted that the “actual date" index reflects the demand that particular year (1984) when the lowland planting was some- what delayed by the weather. Thus, it does not reflect the long-run expectations of farmers as does the "ideal date” index. On the other hand. the "ideal date“ index probably includes some assumption of ideal weather which would allow them to plant early. For the latter reason. Alternative B is considered more realistic than Alternative A. More information on this topic will be available after the 1985 lowland planting season when the seed stored from 1984 will allow farmers to plant as early as desired. The temporal distribution of the supply of CIAT wheat seed can be estimated from the dates of arrival of the seed at the Warnes Processing Plant in 1983 and 1984. the only years for which such data are avail- able. In order to take into account the time between the arrival of the seed at Warnes and its availability at the farm, two weeks were added to these dates. For several reasons it is believed that the 1984 data are more representative of future trends than those for 1983. In addition to the more typical weather. the supply distribution for 1983 was truncated after mid-July when the prOJect stopped accepting seed. In the future. the progect will follow the 1984 policy of guaranteeing purchase from all registered seed farmers. Thus. the average distribution of supply was estimated giving two thirds weight to 1984 and one third to 1983. It is displayed in Table 5.2, again expressed as percentages of total supply. 94 TABLE 5.2: TEMPORAL DISTRIBUTION OF THE: DEMAND FOR WHEAT SEED AND TEE SUPPLY OF CIAT WHEAT SEED (% of annual supply or demand) May June July TOTAL 1 2 3 4 1 2 3 4 1 2+ Supply at farm level, 0 1 3 17 13 38 9 5 12 2 100 1983 Supply at farm level, 0 O 1 8 10 10 9 21 15 26 100 1984 Supply at farm level, 0 O 2 11 11 19 9 16 14 18 100 weighed average ALTERNATIVE A: Demand according to 1 10 15 20 28 23 1 1 1 0 100 "ideal" planting date Withdraw from storage 1 10 13 9 17 4 0 0 0 0 54 Add to storage 0 0 0 0 0 0 8 15 13 18 54 ALTERNATIVE 3: Demand according to 0 6 17 17 17 11 6 6 3 17 100 actual planting date Withdraw from storage 0 6 15 6 6 O 0 0 0 0 33 Add to storage 0 0 0 0 0 8 3 10 11 1 33 SOURCES: The Simply data from arrival of seed at the Warnes plant plus 15 days to be available for planting. Two alternative mea- sures of demand from the survey of wheat farmers. The differ- ence between supply and demand indicates the amount to be stored (if excess supply) or withdrawn from storage (if excess demand). See text for more detail. 95 The differences between supply and demand each week reflect the changes in the level of stocks necessary to equate the two. Excess demand early in the season indicates the need to withdraw seed from storage. while excess supply late in the season is used to replenish the stocks. The sum of either represents the proportion of the total which would be saved from one year to the next to balance weekly supply and demand. If the demand is distributed as in Alternative A. then 543 of the total production could be stored from one year to the next. If Alterna- tive B is a more realistic description of the demand, implying a later demand pattern. the gap between supply and demand is smaller. consti- tuting 333 of the total. Given the greater confidence placed on Alterna- tive B, it can be affirmed 333 is a reasonable, if conservative. esti- mate of the proportion of seed which should be saved to meet the demand for "early seed." It is worth noting that the indicies of demand do not incorporate any price differential between early and late seed. Given the possibi- lity that early seed would carry a premium. it is relevant to ask if farmers would actually pay higher prices for early seed. Among the survey respondents. fully 613 (weighted by volume of CIAT seed pur- chased) reported being willing to pay 153 more for earlier wheat seed. while 313 said they would pay 303 more for such seed. Although answers to hypothetical questions are not very reliable. the results demonstrate that timely delivery of seed has a significant economic value to the lowland wheat farmers. 96 5.4 §ynthesis of volume and storage prOJections Table 5.1 provides high. medium, and low prOJections of CIAT seed sales under the assumption that wheat seed can be made available early in the lowland planting season. In Table 5.2, it was estimated that at least 333 of the annual volume of seed production would have to be stored to meet this demand for “early seed." In Table 5.3. these progec- tions are combined to estimate the volumes to be produced. stored. and sold each year. In addition to the results assuming Alternative B (333 of the seed is stored). the table includes for the purpose of comparison the results assuming Alternative B (543 stored). The "sales" in Table 5.3 are equal to the corresponding figures in Table 5.1. “Direct sale" refers to seed produced and marketed in the lowlands the same year. as is currently done. It is 463 of the sales in Alternative A and 673 in Alternative B. ”Stored seed" refers to the volume of seed put into storage that year for sale the following year. Thus, it is calculated as a percentage (333 or 543) of the sales the following year. And finally, “production" is the sum of direct sales and seed produced for storage. All figures are expressed as net weights of wheat seed, that is. after drying and cleaning. According to Table 5.3. CIAT wheat seed production in 1990 may reach 1200 MT, the estimated maximum production in the Valles Meso- termicos. or it may be as low as 483 MT. The estimates of the volume of seed stored range from a high of 648 MT. if production reaches its maximum and 543 is stored each year. to a low of 159 MT. The medium progection of Alternative B, believed to be the most probable of the six scenarios, show CIAT wheat seed production rising to 492 MT by 1990 and the volume of stored seed reaching 252 MT. 97 .umm>wo£ no mmo> cw oeumaa owed ooLOuw .mHHMuoo saw some new .a.w com H.w defines Hwmwmoow wwv wwv wwv wwv wwv wwv wwv wwv wwv wvv wav wow mmmow wwv .wwv wwv wwv wwv wwv wwv wwv oav owv waw ovw ceauoooomm owH owH owH owH own own owH own own wwH wvH wwa concuw vaw vaw vaw vaw vaw vaw vaw vaw on voa owe wow mmamo acumen coauomhomo.zoq vww won oww aww haw wow oww www vnw wav vwv wow mmfimw vvh wan www www www flaw www www waw wwv oww ovw coauoooomm awa ava wwa vaa waa boa ooH Ho“ vwn on" wwa wwa nonpuw aov whv wwv hwv Oav vov oww vww vvw haw HwH wow mamm womuao coHHOOHOAQ soaps: OOaH ooaH ooan oOau oOaH wofia owo vww wow www owv wow mmamw OOaa ooan OOaH oOaH oOaH wwa” woon www fivh haw hfiv ovw coHuooooum wow wow wow wow wow wow vww wHw wwa oaa Hoe wwa concuw vow vow vow vow vow oww avw oww wwv www haa wow mama uoomao COHuomhowm coax concum ma poem mo xww uw o>aumcumuaw wwv wwv wwv wwv wwv wwv wwv wwv wwv wvv wav wow moflmw wwv wwv wwv wwv wwv wwv wwv wwv vhv vwv oov ovw ceauoooomm Hwa awa awa awa Hwa awa awa Hwa Hwa owa owa wwa concuw aaa aaa aaa aaa aaa aaa aaa aaa wfia voa ohH wow same wommao coHuomnomm ZQH vww wow oww aww haw wow oww www vfiw wav vwv wow moamw oww Haw wow www fivw wHw wow oww www wov wwv ovw caeuuooomm afiv wow wa www aww oww waw wHw flow wwa wwa wwa oom0um www waw aflw oow owa wwa wwa wwa wwa hHa awe wow mamm uoomwo cofiuoonomo soaps: oOaa OOaH 00a~ ooan oOaH wo- owo vww wow wow owv wow moaow OOaH 00afl OOaH ooaH ooan wwa“ hwoa woo one «vw oww ovw coauoooowm wvw wvw wvw wvw wvw wvw wow wHw owv www an wwa oomOHw aww aww aww awn aww how va vww oaw wwa aaa wow mmamm uommfio 83888 63a honeys ma comm mo wvw no o>aumcmmua< wooH vooa moon aoon Hood ooon owoa wwon hwoa won won vwou Ant: ommm hcmzzaaflHo MC mmq age Hmaoaaoa aaaaa aaaaa aaaaa aaava ooaaa aaaoa aaaaa aaaaa aaaaa voaaa Ana coca. mazmsmm amaze aa aa aa aa aa aa aa aa aa aa Ana oooac a: sad mocm>am aaa aaa vaa aaa aoa aaa Has «as one 0 Less maaaoum :a aesao> aaa aaa aaa was was was aaa aaa was aaa Ana coca. a: sad msnm>mm aaa aaa aaa aaa aaa aaa aaa aaa aaa aav Less ommamooso messb> mom." voon wmoa Noon amoa coon mwma wwmn hwmu mama gm OZHmmeOE Qmmm m0 WES mmommu ”v.0 mama? 110 TABLE 6. 5: COSTS OF IMPORTED EQUIPMENT FOR PROCESSING PLANT Unit Description Quantity price Cost (055) (USS) I. Plant equipment Generator (10 KW) 1 8515 8515 18-foot elevators 2 2500 5000 Blower/heater 1 3000 3000 Two-wheeled trolleys 2 175 350 Air-screen machine A-334 1 10000 10000 Chemical treater 1 6500 6500 Weighing and begging machine 1 750 750 Bag sewing machine 1 900 900 Subtotal 35015 II. Laboratory equipment Aluminum scoops 2 10 20 . Testing screens 10 15 150 Humidity tester 1 300 300 Higrothermograph 1 735 735 Subtotal 1205 Purchase value 36220 Packaging, transportation, and insurance (40% of the purchase value) 14488 TOTAL (USS) 50708 TOTAL (Sb 1000 at official rate of Sb 2000/USS) 101416 TOTAL (Sb 1000 at parallel rate of Sb 3250/USS) 164801 SOURCES: Equipment list from discussions with project technical staff and prices from project documents. 111 able at the site. This constitutes a conservative asausption in that power free a utility would probably be less expensive. Although Table 6.5 shows conversion of the dollar cost to local currency value using both the official and parallel exchange rate, only the forser is used in this financial analysis. This is the rate that would be paid to inport the equipaent and thus reflects the private cost. The parallel rate is used in the econoaic analysis presented in section 6.4. Rough estimates of the construction and installation costs are aade in Table 6.6 by using the expected dinensions of different cosponents of the plant and unit prices. These were estimated by the progect technical personnel with experience in the design and construction of seed pro- cessing plants, both in Bolivia and elsewhere. Soae laboratory equipaent was also included as a local currency investaent itea. 6.3.3 Operating costs Operating costs include personnel, energy, saintenance and repair, and miscellaneous costs. Personnel costs are calculated assuaing the need for a full-tine plant sanager and nightwatchsan, as well as varying use of laborers. Although the Uarnes Plant often operates with only three workers per shift, four are assused in this case because of the lower level of autosation. For example, no horizontal conveyor belts like those at Warnes are planned. Eight person-aonths are assuaed for the two aonths of double-shift processing of seed for direct sale. Four workers handle the processing of seed for storage for as nany aonths as the voluee deaands (see Table 6.1). One worker reeains during the aonths without processing activity for cleaning and aaintenance. Lastly, three person-months are included to handle the sale of stored seed in April. The results are presented in Table 6.7 and sunsarized in Table 6.9. 112 TABLE 6.6: CONSTRUCTION'AND INSTALATION COSTS FOR PROCESSING PLANT Uhit Dimensions price Cost (m) (m 2) (1000 Sb)(1000 Sb) A. Building area Receiving and drying 14x10 140 Processing 10x10 100 Storage area (225 MT) 10x18 180 Subtotal 420 315 132300 8. Covered outdoor areas Receiving area 5x3 15 Dispatching area 5x3 15 Subtotal 30 140 4200 C. Offices 4x6 24 525 12600 D. Unpaved access road 500 12 6000 8. Security fence 180 7.5 1350 F. Installation Electricity and telephone 15% of A + C 14490 Sanitation and.water 5% of A + C 7245 Special 2% of A 2646 G. Equipment Laboratory equipment 600 Feed chute 5000 Equipment instalation 1200 Subtotal 6800 H. Other 10% of A - G 18763 TOTAL CONSTRUCTION AND INSTALLATION COST 206394 SOURCES: 8ased.on discussions with and estimates by project technical staff. 113 .H.@ com m.m moans? :« >uaommmo >H£ucoe new mmez~o> poem may ease ooumfisofimu mo“ Immooonm mo mm>u some p0m ownfisvmn mafia .mEau moaCamewu may meanso amino: moo com .pmom no mmmm ecu moanop mnmxuoz moans .pOAOum on On comm no mcammmoonm umanmnmflmcau may mcfihso mnmxpoz Adam .mamm woouuc no“ 00mm uo maammmooum uanmImHosoU menudo msmxnoz unmfim mesmmm omam >039 .dmezoum392mwc com ummmcme unwam may no acme>oHoEo QEquHHSM mesmmm «COHHMHsUHmo omega .mumoo gonna uwzuo can >uwhsomm Howoom no“ xh.oa moan mpmzuo 9b you 86.6: am cam 38%... Ema m5 you 08.08 am so mmimawm 35:2. 9: no woman Hmeomaom wmmm mmmm mmmm Comm hnmm mwvm onvm Hmvm mmmm vomm Ovma ovmu ova Deon ovma ovnfi Ovna ownn ova owmfl mwmv wvmv meow 000v hove mmwv oovv “haw mvmv vmow ofimm onm OHmm Ommm oHnm oamm Onmm Guam Guam Damn mmmn vmmn mama mama Hmmfi ommn mmmu mmmn Roma mama Q¢HOB 858m: 26;. mumxnoz summons acmfim Aooon va BZfidm DZHmmmOQmm Qmwm ho mBmOO qwzzommmm ”h.© mum<fi 114 The energy costs per setric ton of processed seed are detersined in Table 6.8. As aentioned, it is assuaed that the plant will have to generate its own electrical power to operate the aachinery. In this table, “usage“ is defined as the energy require-ants per aetric ton of seed processed. It is the product of the power capacity of the unit and the nuaber of hours of operation per ton of processed seed. A 10 KO generator appears to be sufficient to seat the electrical needs of the plant. Energy costs are susaarized in Table 6.9. Haintenance costs are roughly calculated at 2% of the construction and installation cost and 5% of the CIF cost of the imported aquipaent. Although saintenance on soae itess such as the generator will probably be sore than 5%, aaintenanca for others such as the air-screen cleaner and the platfors balance will likely be less. The costs of other itess not included or not foreseen is assuaed to be 10% of the rest of the operating costs. These costs are included in Table 6.9. 6.3.4 Coaparison of costs and revenues Operating and investaent costs are suasarizsd in Table 6.9. Also included in the table are the aatiaated costs for chasicals to treat the seed, as described in subsection 6.2.2. However, these costs are charged to the institutions having the seed processed apart fros the noraal pro- cessing fees. This ites is not counted here with costs, nor with the gross revenues in Table 6.4. Table 6.10 cosbines the gross revenues free Table 6.4 and the costs free Table 6.9. The difference between the two is the net revenue, which is negative the year the plant is constructed and positive thereafter. One aeasure of prOJect worth is the net present value: the sun of the 115 TABLES.8:ENERGYCOSTSPERMETRICTONOFPROCESSEDSEED Electricity Capacity Operation USage (hp) (kW) (hr/Mr) (kwh/M Blower 9.00 6.71 .94 6.31 Elevator for uncleaned seed .75 .56 1.40 .78 Air-screen cleaner 3.00 2.24 1.40 3.13 Elevator for cleaned seed .75 .56 1.00 ' .56 Chemical treater 1.00 .75 .25 .19 Elevator for treated seed .75 .56 1.40 .78 Bag sewing machine .05 .04 .10 .00 Electrical consumption 15.30 11.41 11.76 Diesel USage Operation Usage Cost (l/hr) (hr/MT) (l/Ml‘) (Sb/Mr) Heater for seed drier 7.90 1.00 7.90 5530 Generator (10 KW) 7.00 1.40 9.80 6860 Energy cost per metric ton 12390 SOURCES: Equipnent specifications and technical relationships from technical project staff. Diesel price of Sb 700/liter which includes transportation of fuel to region. NOTE: An electrical capacity of 10 KW’would'be sufficient since not all the equipment is operated simultaneously. 116 .v.o magma ca mmscm>uu no COaumHoono may ea no: .mp0: HouOu one a“ pmooHOCa yo: man >m£a .mcwmmmoonm MOM may no when no museumsu may >9 puma mam mumoo HmoHewzo ”whoz .96 £685 06 838. "88:8 mmhmm mmhmw woman mmnmm Hmvmu momma. mnflmm «momm «worm homvmm o o o o o o o o o mflvfiofi o o o o o o o o o flammow Nflmw hamm naom comm mmnw whom comm mflmm flown vaw mooflm m00~m ammon mvfiom mwvmw mhhmv «meow vahw moumw Nomav mmnm mmHm mmmm mmnm mmam mmnm mono mmfim mmnm mmam Nbflh «bah Omar mwmh mean omON Homo econ Nvmm NNom mmmm mmmm mmnm omnm hamm mmwm omvm HNvm mmwm «Cam mama vmmfl mama mama Hmmfi ommd mama mama Noam coma mBmOO A4909 unwemwsvw.pmunogeH co«uoshumcoo mumoo acmeumm>cH hmzuo “poops pony mascaemzo mocmcmucfimz >mpmcm Huccomumm mumoo mcfiumuomo Anm OOOnV Bz ucwmmsm umz on um oeucsoomfio mscm>mh umz mscm>mn umz mumoo HMHOB uscm>mn mmonu Aooo~ Dmv Bz¢qm QZHmmmoomm fiflflm MO mHm>uaaom coaasea a oOth hammw mowwm mmoom ooohm ovmov mowmv eooov ofiohw «moomol oammo ofimmo oohmo ooONo moooo Chemo hhoho Ohmoo oommo emoomMI whwwm whwwm momma mmoNo woman Hmham oVan mnoum OmOnm mmwuow honom hmfiom Nomom monwm ooomm ammom NNNmm mohho omwmm vomoo omma vmma mmmfl Nmmfi Hmmn ommfl mama mama hmmn ommfl canyon mo mums HmcumucH mums uczoomfio was no“: m5~m> “common uoz xofl um amocsoomao msco>mp umz mscm>mp umz mumoo HouOH mscm>mn mmouw .oooH no. azaom oszmmoomm ammo mo mummuaz< onozoom ”HH.m mamas 122 6.5 iv a al s s Since all progections involve some uncertainty, it is helpful to deteraine the sensitivity of the results to changes in the assumptions regarding prices, volumes, the life-span of the investsent, and so on. Thus, the internal rates of return for both the financial and economic analysis were calculated under a variety of altered assumptions, and the results presented in Table 6.12. If the life-span of the progect is reduced from ten years to eight years, the economic rate of return drops below the lowest estimates of the social opportunity cost of capital, but the investaent resains profitable from a private point of view. If the life-span is twelve years, a clear possibility, both rates increase with the economic rate of return rising to 13%. With respect to the volume of seed, if the high prOJection from Table 6.5 is used and the processing capacity turns out to be 10% greater than estimated in Table 6.1, the rates of return rise somewhat. This increase in profitability is modest because the storage capacity and processing capacity are still limiting factors. On the other hand, if the low prOJection is assumed along with a 10% reduction in capacity (though this is not constraining), the rates of return are significantly lower, but the investment is still financially profitable. Referring again to Table 5.3, if 54% of the seed production each year is saved until the following planting season (Alternative A) and the medium proaection is accurate, then the rates of return diminish only marginally. Storage capacity is a limiting factor from the second year, but the processing capacity for direct sale is not as constraining because more of the seed is being stored. If these volume and storage 123 TABLE 6.12: SENSITIVITY.ANALYSIS: FINANCIAL AND ECONOMIC INTERNAL RATES OF RETURN UNDER VARYING ASSUMPTIONS Financial internal Economic internal rate of return rate of return Original assumptions 18% 10% Life span of project 8 years 13% 5% 12 years 20% 13% volume of seed High projection, 10% higher 24% 15% processing capacity Low projection, 10% lower 11% 4% processing capacity Proportion of stored seed 54% (Alternative A) 15% 8% Fee for seed processing 10% higher 22% 14% 10% lower 13% 6% Fee for seed storage as % of retail value Nb fee 11% 4% 2.5% fee 14% 7% 7.5% fee 21% 13% 10% fee 24% . 16% 15% fee 29% 21% SOURCE: Tables 6.4 through 6.11. High and low volume projections and change in the proportion of stored seed from Table 5.3. Note that the same processing equipment and .storage capacity are assumed. Economic rate of return is conservative and should be considered the lower limit. 124 assumptions were combined with a larger storage facility, the rates of return would be higher than under the original assumptions in this study because of the added revenue from seed storage. The fee for processing has a modest effect on the financial and economic return. It should be noted that with a 10% reduction in pro- cessing fees, the economic rate of return drops to an unacceptable level. And finally, various levels of storage fees, expressed as a percen- tage of the retail value of the seed, are tested. In general, the results can be summarized by saying that the financial return remains acceptable even without a storage fee, but that the economic return is- below the opportunity cost of capital for all fees less than that ori- ginally assumed, 5%. Naturally, any reduction of the premium on early seed below that assumed here (20%) would have to be reflected in the storage fee. It is worth noting that in order to maximize the return, the capacity of the storage facility should be larger (smaller) than 225 RT if the premium on early seed is greater (less) than 20%. Finally, it should be reiterated that the economic rate of return was very conservatively calculated, adjusting only the rate of exchange used to value the imported equipment. If other shadow prices were adopted, as mentioned in section 6.3, then the economic rates of return would be higher than calculated in Tables 6.11 and 6.12. 6.6 va uati o alternate s as for the rocessin lant The criteria for the location of the plant should include the availability of labor, electrical service, and water, the suitability of the climate for seed storage, and the costs of transporting the seed from the seed growing zones and to the lowlands. A site somewhere in the 125 Valles hesotermicoa is indicated by the relatively cool climate and the fact that little additional transportation would be required. In fact, these costs may be lower due to the fact that drying and processing reduce seed weight by about 10%. Four population centers are considered in this evaluation: Vallegrande, Kataral, San Isidro, and Comarapa. With respect to the availability of labor, the two largest towns in the Valles Mesotermicos are Vallegrande (5700 inhabitants) and Comarapa (2600 inhabitants), while San Isidro and Hataral have somewhat less than one thousand each (see Table 6.13). Although these levels of population may not be constraints in themselves, a tomato processing plant being constructed by CORDECRUZ in San Isidro will open in 1985, employing around 30 people. This may create a short-run labor shortage. Regarding the availability of electrical service, Vallegrande and Comarapa have systems with capacities of 380 KW and 300 KW, respect- ively, while the other two have smaller systems (see Table 6.13). Small- scale hydro-electrical progects by the Italian government will increase the capacity of each by 200 KW within a few years. The project in Comarapa, which involves connecting San Isidro to its electrical system, is more advanced and should be in operation in 1985. Although this will provide surplus electrical capacity during the daytime, there will be little or no surplus in the evening, needed for the months when double- shift operations must occur. For this reason, a generator may be neces— sary even in Comarapa and San Isidro. The operating costs of the plant would be somewhat lower with this service, but the need for a generator is probably not avoided. All four centers have municipal water systems. The system in Kata- ral is run by diesel pump and is consequently more expensive than the 126 others. However, the plant does not use water in the industrial pro- cesses, so that this factor does not affect costs in a significant way. heteorological data are not available for all four centers, but the information available does not indicate any major differences between the sites in terms of their suitability for seed storage. The only significant difference is that Hataral appears to have a lower rainfall than the two other sites for which data are available (see Table 6.13). Perhaps the most important impact of site location on the costs of operating the plant concern the costs of assembly and distribution of wheat seed. Table 6.14 estimates the additional transportation costs per 100 RT of seed for each possible plant site. The branched road system (see Figure 4.3) means that locating the plant on any branch will require seed produced along other branches to detour from its route in order to be dried and processed. This additional distance which seed must be transported, twice the length of the detour, is listed for each combination of plant site and production zone. The proportions of seed coming from each production zone are based on the pattern this year. The cost of transportation was approximately Sb 10/quintal/kilometer or Sb 217/KT/kilometer. The cost is the product of the additional distance, the volume, and the transportation cost. Note that locating the plant in Hataral, being located along the route to the lowlands from each production zone, involves no additional transportation. On the other hand, the other three sites would involve additional transportation costs-of 3b 690,000 to 5b 1,484,000 for each 100 KT of seed processed at the plant. Assuming only 400 HT of seed are processed each year for ten years and using a discount rate of 10%, the net present value of these additional costs would be around 5b 18 mil- 127 TABLE 6.13: CHARACTERISTICS OF ALTERNATE SITES FOR THE PROCESSING PLANT Comarapa S. Isidro Mataral vallegrande Population 2597 <1000 <1000 5715 Altitude (meters) 1814 1550 1400 1980 Rainfall (mm/yr) 576 - 501 695 Temperature (00.) 17.8 - - 16.8 Relative humidity (%) 68% - - 70% Electrical capacity Current (kw) 300 - 30 380 Projected (kw) 500 500 30 580 water system Gravity Gravity Pump Gravity Transport costs High Medium Low High SOURCES: CORDECRUZ et al (1982c), CORDECRUZ (1980), and discussions with CORDECRUZ personnel. Note that the projected capacities of Comarapa and San Isidro are of the same system since the two are to be connected by electrical lines. TABLE 6.14: ADDITIONAL TRANSPORT COSTS PER 100 MT OF SEED FOR DIFFERENT PROCESSING PLANT SITES Plant site Seed production Additional Vblume Cost zones distance»(km) (MT) (1000 Sb) Comarapa Vallegrande, Moromoro 100 54 1172 Jague, Tablacucho 60 24 312 Comarapa O 22 0 Total cost 1484 San Isidro Vallegrande, Mbromoro 50 54 586 Jague, Tablacucho 20 24 104 Comarapa O 22 0 Total cost 690 Mataral vallegrande, Mbromoro O 54 O Jague, Tablacucho ‘0 24 O Comarapa O 22 0 Total cost 0 vallegrande vallegrande O 30 O Mbromoro 4O 24 208 Jague, Tablacucho 100 24 520 Comarapa 100 22 477 Total cost 1205 SOURCES: Based on the proportions of seed from each zone this year, and an estimated transportation cost of Sb 217/MT/km at June 1984 prices. the that one kilometer of detour generally implies two kilometers of additional distance. 128 lion to Ob 40 million. By means of comparison, the price of the genera- tor is about 8b 17 million at the official exchange rate. In summary, lateral has a clear advantage with respect to the costs of transportation which is perhaps the most important site-related factor. In addition, its central location would facilitate publicity about the wheat seed preject. The disadvantage of this site is the inferior electrical service which makes the installation of a generator necessary. On the other hand, it is very possible that generators would have to be installed in the other three sites as well in order to be able to operate in the evening. This issue deserves more research, particularly regarding the availability of surplus electrical capacity and the relative costs of generating electricity and using municipal electricity. 6.7 Whggt aggg 2:19; and markgtigg 3911;! In this section, recommendations are made for the improvement of wheat seed pricing and marketing policies to complement the proposed investments. Seed pricing, for both contract seed growers and seed consumers, is considered first. The price paid to contract growers is initially set according to CIAT estimates of the costs of production plus a 20% incentive premium. However, there have been some problems with this system. As mentioned earlier, in 1983 many seed growers found it profitable to sell their “seed” as mere grain. In 1984, by contrast, the newly formed seed growers’ association refused to accept the price initially offered and a higher price was reached through negotiations. The sale price of the wheat seed is fixed before the winter plant- 129 ing season by CIAT in cooperation with AKAPO representing the lowland wheat growers. The criteria used are the costs of producing the seed, principally the price paid to the seed growers and the processing fees, and the relation between the seed and grain prices. The practice of fixing prices according to the ”cost of production" is widespread in Bolivia, not just in this seed program, but in national agricultural price policy. Although useful in some contexts, there are several weaknesses of this system. First, the costs of production vary among producers and are not easy to calculate even for a representative producer. The use of family labor and fixed assets such as land are difficult to value. Second, the hyperinflation in Bolivia makes any cost estimate out of data within a month. And third, the cost of produc- tion does not always reflect a price that buyers are willing to pay or that producers would accept. In general terms, it is recommended that prices be set taking into account a broader set of considerations than merely the ”cost of produc- tion.” Prices set by the institutions of the seed program should incor- porate the likely supply (demand) of seed which can be bought (sold) at a given price. In addition, pricing arrangements must be realistic given the high and unpredictable nature of inflation in Bolivia. All prices used in the agreement with contract seed growers should be set relative to the grain price or simply in kind using the seed as a medium of exchange. For example, the 1984 agreement established that seed growers would receive seed to plant and only repay the ”loan“ at harvest either in kind or at an amount fixed in pesos. When the real value of the cash price fall 75% in real terms, CIAT revoked the second option causing 130 considerable resentment among the seed growers. Repayment for inputs and the collection price for seed itself should be set relative to wheat grain prices. Furthermore, the grain price referred to should be one of relevance to the other party: the mill price in Santa Cruz is less relevant to seed growers than the price in highland markets. With respect to the sale of wheat seed in the lowland, a central problem is the use of a single fixed price all season. This generates excess demand early in the lowland planting season and excess supply toward the end. This creates the potential for favoritism, or at least arbitrariness, in the allocation of the early seed. Host importantly, it does not offer any incentives to find ways to delay lowland planting or provide early seed from other sources, such as stored seed. An alternative approach would be to charge a premium for seed which has been stored from the previous year. This would be in line with the strategy of pricing according to the "cost of production,” yet it raises a problem. As can be appreciated in Table 5.2, the sale of stored seed and seed harvested that month overlap for a period of three to four weeks. The simultaneous sale of stored seed and lower-priced "fresh” seed is clearly unfair, as well as creating the same kind of temptations mentioned above. The third approach, that of charging a premium for early seed regardless of its origin, seems the most reasonable. In this way, the allocation of early seed would follow the criteria of economic “urgency" as reflected in the price lowland farmers are willing to pay. The smaller and more frequent the price changes, the smaller tendency of sales to fluctuate, falling Just before a price change and rising just after one. On the other hand, the frequency of changes is limited by the 131 associated administrative problems. It is recommended that the sale price of wheat seed in the lowlands drop each week on a fixed day and by a fixed, round-number amount in order to keep the system as clear as possible. Furthermore, the price should fall by at least 20% from mid-hay to mid-June in such a way so that the average price is approximately equal to that charged this year. Another advantage to such a price structure is that it would make a similar system possible for the seed growers in the highlands. Because the administrative problems are greater here, perhaps only two or three adjustments in price would be feasible. This would, for the first time, provide an economic incentive for the early harvesting and delivery of wheat seed. Another marketing issue to resolve is the actual disbursement procedure. Currently, farmers obtain permits to buy a specific quantity of wheat seed, usually no more than a truckload, from the ANAPO head- quarters in the city of Santa Cruz. The seed is actually obtained at the Wernes Plant, about 20 minutes north of the city of Santa Cruz. Because the seed is continually arriving, being processed, and being sold it is difficult to monitor the availability of seed at Warnes. Heny farmers are forced to make several trips to the plant, perhaps staying in the city for up to a week waiting for seed. There are various alternative procedures for marketing the seed coming from the new processing plant in the Valles Hesotermicos. One approach would be to allocate seed among buyers at the new plant and have it delivered directly to the farm or colony. This alternative is the least costly in terms of transportation and handling but the most complex in terms of the administrative and communication requirements. 132 Coordinating the dispatch of seed from the plant with the purchases made in Santa Cruz would be impractical except perhaps in the case of bulk orders from a colony, cooperative (such as CAICO), or one of the largest farmers. However, it is likely that farmers would resist buying seed never having seen it. A second alternative is to transport the seed to a trenshipment point in the city of Santa Cruz, an small storage area, perhaps rented by AHAPO or CIAT. The seed would be sold from this place. This system is convenient for farmers lodged in the city waiting for seed and reduces transportation expenses for the seed going to the farms south of the city. However, it requires the establishment of a separate short-term storage area and a distinct record-keeping system from that at the Wernes Plant which may continue processing some wheat seed. Unless a suitable site is already available, this alternative would probably be impractical. The third alternative, that of continuing to use the Wernes Plant as a retail outlet, is perhaps the most convenient, at least in the short run. Although the simplest for CIAT, it is somewhat inefficient in requiring seed to be delivered north of the city for sale only to have some brought to farms south of the city. One way to combine the strengths of each of these systems would be to make some seed available at the new plant and other seed available from the Wernes plant. The price difference would correspond to the transportation and handling costs of delivery to the Wernes Plant. In this way, if a colony or large farm finds it more cost-effective to purchase direct from the plant, this is possible. Even more flexibility could be achieved with the abolition of the 133 rule that only farmers can buy CIAT seed. If a merchant has an unused storage area and feels farmers might pay for the convenience of buying seed in the city, he can experiment at his own risk. Another possibility is that seed might be brought to the more distant colonies for resale by merchants or members of the colonies. This diminishes the sizeable transaction cost that farmers in these distant colonies face. This proposal would be resisted by many working in the agricultural sector because of the danger of ”exploitation.“ However, attempts to establish “fair" mark-ups would eliminate the very flexibility sought: only relatively simple marketing which can be covered by the allowable mark-up will be undertaken. A better method might be to try to ensure that publicity about the price charged by CIAT at the Wernes Plant exists wherever the seed is sold. Perhaps seed merchants could be “licensed" by CIAT and required to post announcements of Wernes prices (furthermore the Wernes price could be listed on the certification ticket itself). This would be easy to monitor among the few agricultural supply merchants in the city. In the colonies, the colony administrators could provide effective enforcement, though abuse seems unlikely among the Hennonites in any case. In summary, it is recommended that CIAT continue selling wheat seed from the Wernes Plant, but also sell it from the new plant in the Valles Mesotermicos at a lower price. Further, it is recommended that the seed be available to selected merchants for resale in the city or to farms, ideally with some system to ensure that buyers are aware of the Wernes price and are therefore making an informed choice. CHAPTER SEVEN SUHHARY AND CONCLUSIONS 7.1 nc sions a din se in t s n l d v ad un a With the recognition of the importance of agricultural technology in economic development has come a growing body of literature on the economic and organizational aspects of agronomic research. Huch less attention has been paid to the problem of developing effective seed industries necessary to multiply and disseminate the new cultivars. This is unfortunate given the importance of the stage in the process of technology generation and diffusion and the serious deficiencies in most seed industries in the less developed countries. The present study attempts to address this issue. Chapter Two of this study identifies several important characteris- tics of seed as an economic product and relates them to the institu- tional requirements of seed industries, focussing on the role of the public sector. With regard to the characteristics of seed, several conclusions were reached: 1) Seed consumers are also seed producers so that a seed industry must offer seed with higher quality or lower price than the farmer’s own saved seed. The ability to do this varies with the climate and crop. Seed that is difficult to store, process, or clean gives a natural advantage to a specialized seed industry. 2) The determinants of quality are complex and vary among farmers. Thus, knowledge of and responsiveness to farmer demand are crucial to successful seed marketing. 3) Seed quality is not readily apparent to the consumer. In the absence of a trusted source of information, the demand for even high-quality seed will be minimal. 134 135 4) The supply and demand for seed are strongly seasonal and vary from year to year and seed is perishable. This highlights the importance of careful seed handling and agile, timely marketing. 5) And finally, the returns to most forms of plant breeding can not be captured by the research institutions. Thus, breeding has public good attributes. With these institutional requirements for effective seed production and distribution, the role of the public sector in each component of a seed industry was considered. As mentioned above, plant breeding has public good attributes, thus justifying public investment. This theore- tical argument is confirmed by the high rates of return which have been estimated for plant breeding. Private breeding has been stimulated by the development of hybrids and by “breeders’ rights“ legislation in some - countries, but for most less developed countries, the public sector must continue to play a dominant role. In the case of seed production, there is less Justification for permanent direct government participation, nor are public agencies well suited for this kind of activity. Nonetheless, public seed production is common in less developed countries, particularly in Africa. Although the "infant industry“ argument is plausible, seed policies often inhibit the development of private seed enterprises. The most effective role for the public sector is to stimulate private involvement in seed production through the l) elimination of subsidized seed prices, 2) the provision of processing and certification services available to all on a fee basis, and 3) the development of a research program which can generate new cultivars adapted to local conditions and provide reliable supplies of quality foundation seed. Additional assistance to private seed enter- prises may take the form of technical assistance, credit, and technical and management training. 136 For various reasons, seed production by private firms is more likely for industrial crops, hybrids, vegetables, and pasture crops than for the major cereals. In the latter case, farmers’ associations and cooperatives may be a more feasible alternative to government seed production. Where public seed production is necessary, experience has demonstrated that the use of contracted seed growers is less costly in the long run than the use of mechanized state farms. Compared to other components of a seed industry, public seed mar- keting is the least Justifiabla in terms of economic theory and has the worst record. Seed marketing requires a degree of responsiveness and flexible marketing system rarely found in public agencies due to the incentive structure (salary rather than commission), centralization of decision-making power, and civil service regulations. Insufficient sup- ply, lateness of delivery, high costs, and even political bias in dis- tribution have been problems. Subsidized seed prices and pan-territorial price exacerbate these problems and may cause seed to be consumed as grain. Seed certification and other efforts to regulate seed marketing are clearly necessary public functions. The non-transparency of seed qual- ity, combined with the option of saving seed, mean that impartial infor- mation on seed quality will greatly facilitate the development of a seed industry. Clearly, the service must be independent and available to all seed producers, public and private, on a fee basis. In Chapter Three, the description of the experience of the the certified seed program in the Santa Cruz region of Bolivia confirms some of the generalizations made above. The reforms of 1982 weakened the public sector monopoly on seed production and allocated different func- 137 tions among HACA, the local research organization (CIAT), and the growers’ associations. They also established “realistic“ prices for seed and made certification and processing available to public and private entities. In addition to rapid growth in seed production, the reforms have simulated the creation of private seed enterprises and private processing facilities. Wheat seed is produced by CIAT using contract growers, but at least one seed enterprise is considering seed production for this crop. 7-2 WWW Partly as a result of the availability of quality seed, lowland wheat area has grown to around 8,800 hectares in just a few years. The survey carried out by the author seems to indicate that CIAT certified seed has accounted for 45-60% of wheat seed consumption, saved seed an additional 9-14%, and other uncertified seed the remainder. The main shortcomings of CIAT seed from the point of view of the lowland wheat farmer are its unavailability early in the planting season. In spite of some complaints about the high price of CIAT seed, a majority of those interviewed said that the lateness of CIAT seed had delayed their plant- ing and claimed they would pay 15% more for seed available earlier. The prOJections made indicated that wheat area in the lowlands may grow to 13,600 he by 1995, with high and low prOJections being 8,800 ha and 30,000 ha, respectively. If the problem of lateness can be solved, the annual demand for CIAT seed could easily reach 730 HT by that time. A comparison of the temporal distribution of wheat seed supply and demand showed that around 33% (and possibly up to 54%) of the annual production would have to be stored from one year to the next in order to satisfy the early demand for seed. 7.3 138 mme da s r rdin w a e m rket n as tern 80 vi Using the projections referred to above, a feasibility study of a processing plant and storage facility in the highland seed growing region was carried out using June 1984 local prices. Both financial and economic internal rates of return were calculated. On the basis of this analysis, the following recommendations were made regarding equipment and capacities. 1) The drying capacity should be sufficient for to dry 100% of the 2) 3) 4) annual production of wheat seed in two months. Given the prOjections made here, a three-chambered drier with a total capacity of 24 HT with forced heated air flow should be adequate. This would dry about 1.0 HT/hr. The processing plant should have an air-screen cleaner, a chemical treater, a balance platform beggar, an industrial sewing machine, and two elevators. Of the two models of air- screen cleaner, the Crippen A-334 seems more appropriate to the needs of the plant. Its capacity is approximately 0.7 HT/hr. Assuming a 20% price premium on early seed, a 5% storage fee, and a ten-year life span of the plant, the storage facility should have capacity sufficient for its seventh year of opera- tion. Given the pr0jections made here, this would indicate the need for 225 HT of storage capacity. It is assumed that twelve month wheat seed storage in this region is safe. This assumption is currently being tested. Basic laboratory equipment and space for a laboratory are needed. In addition, a small office would be included in the plant design. Given these equipment and capacity recommendations, the prOJections of demand made earlier, the assumption of a 5% storage fee, and the processing fees currently charged in Santa Cruz, the financial rate of return to the plant is 17.7%. The economic rate of return is 10.3%. The latter was conservatively estimated in that only the exchange rate was adjusted to reflect true scarcity value. Harket prices were not adjusted to reflect the benefits to farmers, consumers, and plant workers as a result of the plant. Thus, it appears that investment in the plant would 139 provide a substantial financial return and an adequate economic return. A sensitivity analysis was used to estimate the impact of changes in assumptions on these rates of return. For example, a 12-year life span would raise the economic rate of return to 13%, but the absence of a premium for early seed would result in a unacceptably low economic rate of return. An evaluation of alternate sites for the plant indicated that electricity and transportation costs were the most significant factors in plant location. Hataral has a clear advantage in terms of transporta- tion costs, but more research is needed to determine if the inferiority of the public services, particularly electricity, makes this an undesir- able site. Huch depends on whether a generator would be needed at the other sites as well. And finally, several recommendations were made regarding the pric- ing and marketing of CIAT wheat seed. First, it was suggested that the price schedule of wheat seed involve a downward staircase pattern with fixed adjustments each week. This would provide incentives for storage by CIAT and others, encourage later planting among those that are able to do so, and allow a similar (but simplified) scale to be established in the highland seed growing area. Second, it was recommended that non-farmers be permitted to buy moderate quantities of wheat seed. The price of resale would not be controlled, but notice of the initial price could be included on the certification ticket to ensure that the eventual buyer is making an informed decision. This would facilitate the marketing of CIAT wheat seed by private traders in the more distant colonies. 140 7.4 Impligatiogg for fugther gesearch Huch of the available literature on seed programs in developing countries is oriented toward the technical issues in individual pro- grams. There is a need for.more comparative studies and for more atten- tion to the economic and institution-building aspects of seed programs. There are several areas of particular interest: Adoption of improved varieties: What is the relative importance of price, availability, performance, and other factors in seed pur- chasing decisions? Seed programs serving small farmers: What is the experience with such programs? What special measures are important in serving small farmers? Effective public seed production: Where public seed production is considered necessary, what institutional forms, pricing structures, and personnel arrangements have proved useful in dealing with the special requirements posed by seed production? How can such systems take advantage of private marketing channels? Types of public assistance: What forms of public assistance have been most effective in promoting private sector participation? What is the experience with mixed or joint enterprises and what problems arise? Seed industry reform: Various countries have undergone privatiza- tion in seed production. How is this accomplished and what have the results been? Promoting private cereal seed production: Are the obstacles to private cereal seed production related to controlled producer prices, subsidized public programs, and other policy variables? Or are the problems more fundamental such as the ease of on-farm saving of seed, limited effective demand, and geographic disper- sion? What is the experience with other private entities such as cooperatives and farmers’ associations? Historical perspective: Looking at the more advanced seed indus- tries in the developing countries (Brazil, Colombia, Thailand, the Philippines, India, and so on), how has the private sector become involved over time? Research on these and other related topics would contribute to the design of systems for technology generation and diffusion in less deve- loped countries. ANNEXES ANNEX A SEED AS A "PUBLIC GOOD" Although frequently used in economic literature, the term “public good" is variously defined. According to one definition, a public good has the property that "if more of it is provided for one member of a group of people, more is necessarily provided for all members" (Sugden and Williams, 1978: 159], examples being national defense and flood control. It is also defined as a good “whose availability to a single consumer is not reduced by having made it available to another consumer” [Pearse and Nash, 1981: 120], such as information or a half-empty train. These contrasting definitions correspond to Schmid’s distinction between “high exclusion cost“ and “joint-impact" goods. In the latter case, "even though people could be excluded from use of the output, this would be a mistake as along there is someone who places any value on the product, since it costs nothing to allow another user" [Schmid, 1983: 10]. Thus, in order to ensure allocative effi- ciency, the price would have to be zero, preventing the producer from recovering the costs of production. A monopolistic producer of a joint-impact good may be able to cover costs, but only by raising the price above marginal cost which results in a non-Pareto optimal level of consumption. This is because 1) rational consumers equate the ratios of marginal utility to price for different goods, and 2) since the price of the joint-impact good is greater than the marginal cost, the ratio of marginal utility to margi- nal cost is greater for this good than for ordinary goods. Therefore, 141 142 allocating more resources to supplying the joint-impact good could increase the utility of some without decreasing the utility of any. In a competetive unregulated market, the problem is different. Competition drives the price down to the marginal cost which, being below the average cost, does not allow producers to cover costs even if the social benefit justifies the cost. In the case of information, seed, and inventions, there are often high fixed costs of developing the good, but once developed it can be produced relatively cheaply. Thus, the average costs of production are greater than the marginal costs, making it a joint-impact good. Further- more, in the case of relatively simple inventions, most non-hybrid seed, and all information, it is relatively easy for other firms to to dupli- cate the “product.“ This fact introduces high exclusion costs which make monopoly impossible. Duplicating firms will reduce the price of the good to their marginal cost of production which does not allow the original producer to cover its average costs (which include the costs of research and development). The point is not that the researcher/inventor/breeder should always be able to cover costs, since some joint-impact goods are not worth producing. Rather, the point is that, in a private unregulated market, even some goods whose social benefits would exceed the costs of develop- ment and production will not be created because the original producer captures only a small portion of the benefits. The incentives of this producer are reduced because of the gains captured by those duplicating the good. The result is that potential researchers!inventors/breeders who are aware of this pattern will not enter the market, and those that do will produce a sub-optimal level of these kinds of goods. 143 There are two important policy implications. First, the public sector can play an important role in creating legal-economic institu- tions which allow producers of this type of good to capture a greater portion of the returns and thus to cover costs when the product is indeed socially useful. This is the idea behind patent legislation and breeders’ rights legislation. Second, in cases where such protection is not practical, there is an opportunity for the public sector to make investments with signifi- cant social returns. Because public agencies do not have to cover costs, they can provide goods (such as information and plant cultivars) whose benefits, though not captured by the agency itself, justify the invest- ment. ANNEX B NETHODOLOGY OF WHEAT FARHER SURVEY The survey of lowland wheat farmers was conducted to provide infor- mation concerning 1) the prevalence of the problem of late seed, 2) the temporal aspects of supply and demand, 3) current levels of seed con- sumption and cultivated wheat area, 4) the proportion of wheat seed supplied by the CIAT seed program, and 5) factors influencing demand for wheat seed. ngple Salecgiog It was decided to draw a sample of approximately forty wheat far- mers in the Santa Cruz lowland area. Unfortunately, no complete and usable list of wheat farmers could be obtained. The two flour mills in the city of Santa Cruz had lists of the farmers from whom they purchased wheat. However, these lists lacked identification of the farmers’ colon- ies or towns making it impossible to locate them. The oil-seed growers’ association (ANAPO), which also represents wheat growers, had a member- ship list, but this list had the same problem and included many soybean growers who do not produce wheat. It was necessary to draw the sample from wheat seed sales lists from 1983 and 1984 provided by ANAPO. As described in Chapter 4, ANAPO markets all of the CIAT seed and some seed obtained from other programs in the highlands. The 1983 list includes both CIAT seed sales and the small volume of non-CIAT seed while the 1984 list appears to include only the former. CIAT seed constituted 97% of the sales included on the 144 145 lists. Names selected from the buyer lists for both years in order to include farmers who obtained seed from other sources one or the other year. The list for 1984 was current as of July 5, thus excluding late- season sales accounting for about 30% of the 1984 total. However, over 90% of these late-season sales went to the largest wheat farmer, who was included in the survey. Since the purpose of the survey was to study the nature of the demand for wheat seed, the sample selection was done in a manner such that the probability that a given farmer would be chosen was directly proportional to the sum of volumes of CIAT seed purchased by that farmer in 1983 and 1984 (reference to the 3% non-CIAT seed will be omitted for the sake of brevity). Greater representation is given to the larger buyers to reflect the fact that their decisions and preferences have larger influence on the demand for this wheat seed. The method involved assembling a list of all of the seed buyers in one list with the individual volumes purchased. Then a list of the cumu- lative volume of seed purchased was prepared so that each name corres- ponded to a range of the total amount purchased. For example, a grower who bought 10 46-kilogram bags might correspond to the 4538th through 4547th bags of wheat seed. Next, bags were sampled randomly from the total number and the name of the grower corresponding to each sampled bag was selected for the survey. In other words, if the 4541at bag were sampled, the hypothetical grower mentioned would be selected for the survey. It was determined that selecting 64 bags would yield approximately 40 names of growers, since the larger buyers of seed would be selected more than once. Begs were sampled from the cumulative list at intervals 146 of 193 bags (the total volume of seed sold divided by 64), starting from a bag randomly chosen from the first 193 bags. Since each farmer’s seed purchases were grouped together on the list, this procedure guaranteed that all farmers who purchased 193 bags or more would be included in the sample, while the probability of smaller buyers being selected was proportional to the volume purchased. In mathematical terms: 31/193 if X1 < 193 P1 = 1 if X1 8) 193 where P1 is the probability of the ith grower being selected and X1 is the number of 46-kg bags of wheat seed bought by the 1th grower from CIAT in 1983 and 1984. This procedure resulted in the selection of 25 names of Nennonite farmers and 10 names of Bolivian nationals. The Japanese farmers coordi- nated their purchases through their cooperative, CAICO, so that the ANAPO lists identified the buyer only as "CAICO." Nevertheless, nine bags of seed bought by CAICO were selected by this procedure. A list of the Japanese growers and the volumes purchased by each in 1984 were provided by CAICO, permitting the selection of nine names through the same procedure described above. Thus, 44 names of wheat farmers were selected to be interviewed. Qgestionnaire prepaggtign and the intervigws The survey questionnaire was prepared by the author in consultation with technical personnel from ANAPO and the Certification Service. It included closed-ended questions regarding the area and yield of summer and winter crops, previous experience and future prOJections for wheat production, the preferred and actual dates of planting, planting den- 147 sity, sources of seed, varietal preference, hypothetical supply response to wheat price increases, and hypothetical demand response to seed available at an earlier date (see Figure B-l). In most interviews, open- ended questions were also asked to obtain additional information regard- ing the practices of neighbors, degree of satisfaction with seed from different sources, and so on. The interviews were conducted during July 10-21 1984, corresponding to the and of the winter planting of wheat. The author conducted all interviews with the exception of three carried out by a Bolivian member of Chemonics’ Santa Cruz team. Of the names selected, all nine of the Japanese and nine of the ten nationals were interviewed (we were informed that the missing one had not planted wheat either year). Among the Nennonites, two of the names corresponded to the same grower. Of the 24 remaining Nennonites, only 18 were reached. Among those not reached, one lived four hours further than any other, another did not speak either Spanish or English, and the rest could not be found at their houses, in spite of various attempts. For three of these six missing Nennonites, neighbors who grew wheat were chosen as substitutes. Thus, 39 interviews were conducted as part of the random sample. In addition, a number of other "key“ growers were interviewed, including several Hennonite colony leaders and a number of growers that had saved sizable sums of seed from one year to the next. The responses of these farmers were not tabulated with the sampled farmers, but were used to gain a qualitative understanding of wheat seed use. 148 FIGURE B-l: EXAMPLE OF SURVEY QUESTIONNAIRE CHE-WICS INTERNATIONAL en oooperacion con CIAT y ANAPO mm DE TRIM (Julio 1984, Sta Cruz) Nanbre Mierbro de ANAPO: si Iocalizacion . no m (hiltivado en verano 1983—84 Soya k’aiz Arroz TOTAL (hiltivado en invierno 1984 Says Trigo TOTAL WEN DE TRIGO U3) DE SIMILLA DE TRICO Densidad de siembra Sanil la de trigo sambrado en 1984 Fuente Var Precio Problems etc. Semilla de trigo sambrada en el pasado ‘uente Var Problems etc Qiltivado 1983 Cultivatb 1982 Sanbrara 1985 Trigo que se senbraria con un precio 20% mayor Fecha de inicio de siambra Fecha determinada por el clima disponibilidad de semilla disponibilidad de equipo Fecha ideal para Siambra I I lMLUb 2 leyenda 00 no oolonia/ooop S=Saguayo 200 oolonia/ooop J=Jaral 10 no sabe Q=Quinori 20 CIAT-ANAPO 30 ANAPO 40 50 omerciante 1 [mifomidad 60 transportista Fundimicnto 70 agricultor 3 mfennedades 1 Santa Cruz 4 Costa 2 Vallegrande 5 Disponible tarde 3 Catarapa 6 Dificil obtener 4 Shore 7 5 Cochabamba 8 Variedad preferida y por qué m manta semilla guardara para 1985 i I Para cubrir costos de almacenamiento pagaria 15% mas para samilla lista 15 abril si no pagaria 30% mas para semilla lista 15 abril Encuestadores: Lic. Nicholas Minot ch. Dexter Vargas sino 149 FIGURE B-2: PHRASING 0F SURVEY QUESTIONS I am working with ANAPO to identify problems with the supply of wheat seed, particularly since some farmers have complained about its lateness. I would like to ask you several questions about your wheat crop and the wheat seed you use. How many hectares of soybeans do you have planted now in the winter? Of wheat? Of other crops? How many hectares of soybeans (maize, rice, other crops) did you have planted last summer? What yield did you obtain? How many hectares of wheat did you plant last year [1983]? What yield did you obtain? How many hectares of wheat did you plant the year before [1982]? What yield did you obtain? How many hectares of wheat do you plan to plant next year if the weather and prices [in dollars) are the about same as they are this year? Returning to this years wheat crop, around what date did you begin planting wheat this winter? Was that date determined by the weather, the availability of seed, availability of machinery, or some other factor? What would consider the best or ideal planting date for wheat? How much wheat seed do you use per hectare? Where did you obtain your seed this year? What cultivar was it? How much did it cost? What was your impression of this seed, what problems or advantages did it have? And in previous years? What cultivar do you prefer? Why? How much seed, if any, do you plan to save to be planted next winter? Now I would like to ask two hypothetical questions. First, how much wheat would you plant if, other things being equal, the price of wheat were 20% higher (or X pesos per quintal)? Second, would you pay a price 15% higher for CIAT seed available in April or whenever you needed it? Would you pay 30% more for this seed? [In some cases, further questions were asked about the respondent’s impression of seed use in his colony/region, the techniques used for saving seed, problems with the supply of CIAT seed, and so on.) 150 n r ta of the resu ts Huch care must be taken in interpreting the results of the survey for two reasons. First, as mentioned earlier, the sample was selected from a list of farmers who bought CIAT seed. Thus, it is representative of wheat farmers who purchase seed from CIAT but not necessarily of all wheat farmers in the lowlands of Santa Cruz. Second, as described above, there was a deliberate bias in the sampling procedure. Simple averages of the responses of the sampled growers will not reflect the “average CIAT wheat seed customer,“ but rather will be biased toward the larger CIAT seed buyer who are found in . the sample to a disproportionate degree. Nor do simple averages repre- sent growers in proportion to the volumes purchased since all growers who bought more than 193 bags of wheat seed would be included and weighted equally. Weights are needed to ensure that growers are repre- sented in proportion to the volumes purchased. Although not intuitively obvious, it can be shown that using the actual volumes of seed purchased as weights would cause large buyers to be over-represented, even relative to their seed purchases. On the other hand, if the number of bags sampled of each grower is used as the weight, then each is represented proportional to the volume of CIAT wheat seed 1983 and 1984. This will be referred to as weighting accord- ing to the volume of seed bought from CIAT even though the actual weights used are different. In spite of these complications, the survey reached the buyers of a large majority of the seed sold in 1983 and 1984. The 39 growers actu- ally contacted and tabulated in the quantitative results accounted for 91% of the CIAT seed purchases. BIBLIOGRAPHY BIBLIOGRAPHY Adams, D.W. and D.H. Graham. (1984). “A critique of traditional agricultural credit prejects and policies." In C.H. Eicher and J.H. Staatz. Aggiculguggl Development in the Third Wogl . Johns Hopkins University Press, Baltimore: 313-328. Ahmed, O.A. and N. Ahmed. (1983). “Production of high quality seed for improvement crops in Bangladesh." eed Sgigngg gag Igghnolggy 11: 1079-1086. Andrews, C.H. and C.E. Vaughan. (1983). “Re-design of seed component of HIDAS II in Ghana.” Seed Technology Laboratory, Hississippi State University, State College. TA 83-1. Askari, H. and J.T. Cummings. (1977). “Estimating agricultural supply response with the Nerlove Nodal: A survey.” Ingggggtigngl_§gggggig. Rgview 18 (2) (June): 257-292. Asociacion Nacional de Productores de Oleaginosas (ANAPO). (1982). “Semilla de trigo entregada a los productorea para la siembra, Invierno ’82.” Santa Cruz, Bolivia. ________ . (1983). ’Entregas semilla de trigo a productorea, Invierno ’83.” Santa Cruz, Bolivia. Benco Central de Bolivia. (1984). Indices de Pgeciog a; por Mayor. Division Tecnica, Departamento de Cuentas Nacionales, La Paz, Bolivia. Barton, J. (1982). ”The international breeder’s rights system and crop plant innovation." §ciegce 216 (4) (June): 1071-1075. Bates, R.H. (1981). Harkets and States in Tropical Africa: the Political Bgsis of Agricultural Policies. University of California Press, Berkeley. Bedoya, R. (1970). “Aspectos de la industrializacion y commercializacion del trigo en Bolivia.“ iclo gs Confergncias - Simposium sobre Produccion dg Trigo en Bolivia. 18-22 Hay. Sociedad de Ingenieros Agronomos de Bolivia. Instituto Nacional del Trigo. Benjamin, H.P. (1981). Investment PrOJects in Agriculture. Longman, Essex, U.K. 151 152 Basri, H. (1983). “Seed-borne diseases of wheat and barley in Horocco in relation to seed certifiction programmes." §eed Scienpe and Tecppology 11: 1103-1113. Bono, H. and H. Lam. (1976). "La production de samences selectionnees d’especes vivriares au Senegal." gronomie Tropical 31 (2): 170- 178. . Boyd, A.H. and R.I. Jackson. (1981). “Consulting visit for the Casamanche Regional Development PrOJect 685-0205." Seed Technology Laboratory, Hississippi State University, State College. TA 81-1. Boyce, K. and R.E. Evenson. (1975). National and International Agricul- tural Rgsearch and Extension Programs. Agricultural Development Council, New York. Bunch, H.D. (1982). "PrOJect planning and implementation of seed produc- tion for market-oriented farming." In Seeds. Proceedings of the EAO/SIDA Technical Conference on Improved Seed Production. 2-6 June 1981. Nairobi. Food and Agriculture Organization, Rome: 476-483. Brown, L. et al. (1984). "Private sector development in the Thai seed industry." A.I.D. Special Study. United States Agency for Interna- tional Development, Washington, D.C. Centro de Investigacion Agricola Tropical (CIAT). (1979). Informs Anual 1978-9. Santa Cruz, Bolivia. ________ . (1982). “Programa de multiplicacion de semilla de trigo en los Valles Hesotermicos del Departamento de Santa Cruz.“ Santa Cruz, Bolivia. ________ . (1983). "Produccion de semilla de trigo en los Valles Heso- termicos del Departamento de Santa Cruz.“ Santa Cruz, Bolivia. Certificacion, Servicio Regional de. (1984a). “Plan quinquenal de semi- llas." Chapter in forthcoming five-year agricultural plan for department of Santa Cruz. Santa Cruz, Bolivia. ________ . (1984b). "Cuadro de semilleristas de las zones de Vallegrande y Comarapa.“ Santa Cruz, Bolivia. Chahal, G.C.N. (1975). “The framework of governmental decision-making - Free market economy with official planning." In W.P. Feistritzer and H. Redl (eds.). The Role of Segd Science and Technology in Agricultural Development. Proceedings of an International Seed Symposium. 1-6 October 1973. Vienna. Food and Agriculture Organization, Rome: 261-268. Chin, H.F. and K. Refer. (1978). "Development of seed technology in Nalaysia." eed Science and Technology 6: 1071-1077. Copeland, L.O. (1976). Principles of Seed Science and Technology. Burgess, Hinneapolis. 153 Corporacion de Desarrollo de Santa Cruz (CORDECRUZ). (1980). "Conpendio de datos meteorologicos del Depertamento de Santa Cruz." Senta Cruz, Bolivia. ________ . (1982a). "Evaluacion de los sistemes de colonizacion agraria en el Depertamento de Santa Cruz.“ Santa Cruz, Bolivia. ________ . (1982b). "Proyecto: elmacenemiento y comercielizecion de gre- nos." Santa Cruz, Bolivia. ________ . (1982c). “Estudio de fectibilided de elmacenemiento y comer- cialization de granos. Tomo II: Apendices.“ Santa Cruz, Bolivia. Corporacion de Desarrollo de Santa Cruz (CORDECRUZ) et al. (1982a). Diagnostico Agropecuario del Depagtanento de Santa Cruz: quo I. Santa Cruz, Bolivia. ________ . (1982b). Diagnostigo Agpopecuario dgl Depertamento dg Sent; Cruz: Tomo II. Santa Cruz, Bolivia. Dalrymple, D. (1975). Evaluating Ferpiliger Subsidies in Developing Countries. United States Agency for International Development, Washington, D.C. ________ . (1979). "The adoption of high-yielding grain varieties in developing nations." gricultural History 53 (4) (October): 704- 726. Delouche, J.C. (1975). ”Seed processing and storage.“ In W.P. Feistrit- zer and H. Redl (eds.).The Role of Seed Science and Tecpnology in Agricultural Development. Pppceedings ofppn International Seed Synposium. 1-6 October 1973. Vienna. Food and Agriculture Organization, Rome: 108-124. ________ . (1978). "Preparation of seed progress." In W.P. Feistritzer and A.F. Kelly. (1978). Improvgd Seed Production. Food and Agricul- ture Organization, Rose. Delouche, J.C. and H.C. Potts. (1971). Seed Program Development. Seed Technology Laboratory, Hississippi State University, State College. Delouche, J.C. et a1. (1973). “Storage of seed in sub-tropical and trop- ical regions.“ ged Science and Technology 1: 671-700. Dewey, U.G. and D.R. HcAllister. (1966). “An agronoic survey of the pos- sibilities for increasing wheat production in Bolivia.“ Utah State University Series 6/66. USU/USAID/Bolivia. Douglas, J.E. (1975). "International seed production and technology developnent.“ eed Science and Technology 3: 387-398. ________ . (1980). Supeessful Seed Programs: A Planning and Henegement Guide. Westview Press, Boulder, Co. 154 Edelmann, A.T. (1967). “Colonization in Bolivia: Problems and pros- pects.“ Lpter-Aegpicgp Econoeic Affairs 20: 39-54. Eicher, C.H. (1984). “International technology transfer and the African fareer: Theory and practice." working Paper 3184. Department of Land Hanageeent, University of Zimbabwe, Harare. El-Gamal, A.Y. (1975). "Seed industry development in Egypt." In R. Spitz (ed.). se tud' s on Seed ndust D velo men f t quppries. Food and Agriculture Organization, Rome. Feistritzer, W.P. (1972). "National and international cOOperation to promote the use of better seed." roceedipgs pf the Internetional Seed Testing Association 37 (1): 83-91. ________ . (1975). "The role of seed technology for agricultural develop- ment.“ eed Science and Technglogy 3: 415-420. . (1978). "The world seed situation.“ In W.P. Feistritzer and A.F. Kelly. (1978). mproved §eeg Production. Food and Agriculture Organization, Rome. ________ . (1982). “World seed production and supply." In Sgpdg. Egpgged- in s of the FAO/ IDA Techni al C nfe en 1 v d S P o u - tion. 2-6 June 1981. Nairobi. Food and Agriculture Organization, Rome: 51-59. Fifer, J.V. (1982). “The search for a series of small successes: Frontiers of settlement in eastern Bolivia.“ ourpa; of Latin American Studies 14 (2): 407-432. Fletcher, G.R. (1975). ”Santa Cruz: A study of economic growth in eastern Bolivia.“ nter-Americen Egonomic Affairs 28: 23-41. Food and Agriculture Organization. (various). FAO Trade Yearbook. Rome. Garey, A. and P Pattie. (1981). “Lineenientos para la planificecion del programa de semillas en Bolivia." Chemonics International Consul- ting Division, La Paz, Bolivia. Gardner, B.D. (1966a). "The economics of an increase in wheat production in Bolivia.“ Utah State University Series 13/66. USU/USAID/Bolivia. ________ . (1966b). "Wheat report - Bolivia.“ Utah State University/ USAID/Bolivia. ________ . (1968). "Prospects for a successful wheat progran in Bolivia." Utah State University Series 2/69. USU/USAID/Bolivie. . (1970). "A reevaluation of the wheat program after two years." Utah State University Series 2/71. USU/USAID/Bolivia. ________ . (1974). “Agricultural price policy in Bolivia.” Utah State University Series 8/75. USU/USAID/Bolivie. 155 Giaciche, W. and O. Toure. (1982). "Seed farms in Heli." In Seeds. Proceedipgs of phe FAO/SIDA Technicpl Conference op Isproved Seed Production. 2-6 June 1981. Nairobi. Food and Agriculture Organiza- tion, Rose: 496-509. Gibler, J.H. (1971). “The Bolivian wheat program, a review and evaluation." Utah State University Series 2/72. USU/USAID/Bolivie. Gittenger, P.J. (1982). Economic Analysis of Agricultpral Progects. Johns Hopkins University Press, Baltimore. Gregg, D.R. et al. (1980). ”Inter-relationships of the essential activities of a stable efficient seed industry." Seed Sciencp and Iggppplpgy_8: 207-227. Gray, D.H. (1975). ”Seed industry developnent in Zambia." In R. Spitz (ed.). Case Studies on Sggd Industry Development of Eight §elgcted Countries. Food and Agriculture Organization, Roee. Gunay, B. (1975). “Seed industry development in Turkey.” In R. Spitz (ed.). Case Studies on eed Indust D velo me t of i ht Selected Countries. Food and Agriculture Organization, Rose. Hall, U.E. (1981). ”Development of the seed multiplication progreene in Niger." eed Science and Technolpgy 9: $07-$11. Hargreeves, C.H. (1971). "Adequacy of rainfall for producing wheat following cotton near Santa Cruz, Bolivia." Utah State University Series 14/71. USU/USAID/Bolivia. Harrington, J.F. (1959). “Drying, storing, and packaging seeds to sein- tain germination and vigor." In ProceedingsI 1959 Short Course for Seedsmen. Seed Technology Laboratory, Hississippi State University, State College: 89-107. Heyaei, Y. (1984). "Assessment of the green revolution.“ In C.H. Eicher and J.H. Staatz. Agriculpural Development in the Third Uorl . Johns Hopkins University Press, Baltimore: 389-396. Hayami, Y. and v.w. Ruttan. (1971). Agricultural Development: An Inter- national Perspective. Johns Hopkins University Press, Baltimore. Hazelden, E.J.R. (1982). "Kenya Seed Company." In Seeds. Proceedings of the FAO/SIDA Technical Conference on Improved Seed Production. 2-6 June 1981. Nairobi. Food and Agriculture Organization, Rome: 522-531. Heath, D.B. (1959). “Commercial agriculture and land reform in the Bolivian oriente." Inter-Anerican Economic Affairs 13 (2): 35-45. Henkel, R. (1982) "The move to the Oriente: Colonization and environ- mental impact." in J.R. Ladean. Modern-Day Bolivia: Legacy of the Revolution and Prospects for the Future. Center for Latin Anericen Studies, Arizona State University, Tenple: 277-299. 156 Herdt, R.W. and C. Capule. (1983). Adpptipp. §prggg. apg Ppodugpipn Impagp pf Hggegp Rigp Vagigties in Asia. International Rice Research Institute, Hanila. Hiraoke, H. (1980). ”Settlesent and developsent of the Upper Amazon: The east Bolivian exasple.” our a ve o r s 14 (April): 327-347. Hugo, C. et al. (1981). "Produccion de trigo en Bolivia y problesas de sercadeo relacionados.“ Prepared for USAID by the Food and Feed Grain Institute, Kansas State University. International Wheat Council (IWC). (1984). pr;g_flpggp_§§gpippigp, International Wheat Council, London. Johnston, B.F. and J.W. Hellor. (1961). “The role of agriculture in econosic development." sggiggn Eggnosic Review 51 (4): 566-593. Jorgenson, D.W. (1961). ”The development of a dual economy.“ Egppppig, Jppgpgl 71 (June): 309-334. Joshua, A. and A. Singh. (1982). ”National Seed Service, Nigeria.“ In s. P eedin s of he I DA T ni ere Ispgpvpd Sgeg Epoducpiop. 2-6 June 1981. Nairobi. Food and Agriculture Organization, Rose: 510-521. Kabeere, F. (1983). ”Difficulties encountered in the production of high quality seeds in Uganda.” 522g Sgience egg Tpggpplggy 11: 1087- 1092. Kashiwazeki, H. (1983). "Agricultural practices and household organiza- tion in a Japanese pioneer community of lowland Bolivia.” Husen Egplpgy_11 (3): 283-319. Kelly, A.F. (1975). ”Recent organizational developsents and future trends in variety and seed control - Cosprehensive regulatory scheses.“ In W.P. Feistritzer and H. Redl (eds.). Th9 Role of §eed a e and Technolo in A ricu t rel eve sent. P in s of pp Iptggpgpignal §ggd Sysppsium. 1-6 October 1973. Vienna. Food and Agriculture Organization, Rose. Ladman, J.R. (ed.) (1982). Hpgern-Day Bolivia: ngacy of the Rgvolution gpg Epospggts {pp phg Futpre. Center for Latin American Studies, Arizona State University, Tesple. Hinisterio de Asuntos Caspesinos y Agropecuerios (HACA) et al. (1980). ”Plan depertamental de fosento triguero." Santa Cruz, Bolivia. HACA-USAID. (1984). Estudio de Prognostico Agropecuerio. HACA-USAID, La Paz, Bolivia. Hellor, J. (1966). The Ecpnomicp of Agriculture; Developsent. Cornell University Press, Ithaca. 157 Honcayo, G. (1975). “Seed industry developsent in Colombia.“ In R. Spitz (ed.). Qase Spudies on Sped Industry Development of Eight §glec§ed Counpries. Food and Agriculture Organization, Rome. Horales, D. (1984). "Situacion actual semillas en Bolivia." Repprt: §ggonp Roundtgble Heetings on Seed. 2-7 September 1984. HACA/USAID/ Regional Seed Board, Santa Cruz, Bolivia. Nelson, P.H. and A. Kuhn. (1974). "The potential contribution of private foreign seed firss to the establishment of national seed industries in developing countries." Zeitschrift Fur Auslandische Lpndwiptschpft 13 (2): 156-168. Nicholls, W.H. (1964). "“The place of agriculture in economic develop- sent." In C.H. Eicher and L.W. Witt (eds.). A ricul u c n s c Development. HcGrew-Hill, New York: 11-44. Nichols, J.P. et al. (1983). “Food and agricultural commodity grading." In W.J. Armbruster et al. (eds). Federal Harketing Progress in . Agriculture: Issues pnd Options. Interstate, Danville, Ill: 59-90. Pearse, D.W. and C.A. Nash. (1981). he oci re a o ro s: A Tax; in Cost-Bpnefit Analysis. John Wiley and Sons, New York. Pinstrup-Anderson, P. (1982). Agriculturpl Research and Technology in Econpsic Develppsent. Longman, New York. Popinigis, F. (1982). "The Brazilian seed programme.” In Spgds. ngceedings of the {Ag/SIDp Ipphnical Congppencg op Improvgg §gpg Egodugtion. 2-6 June 1981. Nairobi. Food and Agriculture Organization, Rome: 532-553. Potts, H.C. et al. (1983). “Hid-prOJect evaluation of the Foundation Seed Production Project (686-0245) in Upper Volta.“ Seed Technology Laboratory, Hississippi State University, State College. TA 83-05. Reynolds, L.G. (1983). "The spread of econosic growth to the Third World: 1850-1980." Journal of Econosic Literature 21 (3) (Septem- ber): 941-980. Rice, E.B. (1974). Extension in the Andes. H.I.T. Press, Cambridge, Ha. Rollin, S.F. (1975). "Recent organizational developsents and future trends in variety and seed control - Truthful labelling concept.“ In W.P. Feistritzer and H. Redl (eds.). The Role of Seed Science and Technology in Agricultural Developsent. Proceedings of an Inpernational Seed Symposius. 1-6 October 1973. Vienna. Food and Agriculture Organization, Rome. Romecin, E. (1970). “Produccion triguera mecanizada.“ Ciclo de Conferen- cias - Symposius sobre Produccion de Trigo en Bolivia. 18-22 Hay 1970. La Paz, Bolivia. Instituto Nacional de Trigo. 158 Romero, P.S. (1982). “The role of the state in the rural-urban configura- tion.” In J.R. Ladmen, (ed.). appern-Dgy Spllvia: Legagy pf phg Rgvglppiop ppg Prospeptp fog phe Suture. Center for Latin Aserican Studies, Arizona State University, Temple: 301-317. Rosell, C. (1975). “Seed industry development in Uruguay.“ In R. Spitz (ed.). Ce u es n Seed In us D vs 0 sent 51 ht e ec Spuntgles. Food and Agriculture Organization, Rose. Russell, J.F.A. (1983). ”Potential assistance from the World Bank for seed sector development with especial reference to Latin America and the Caribbean.“ Presented at the Workshop on the Development of the Latin Asericen and Caribbean Seed Sector and PrOJected Needs. 25-29 July, 1983. CIAT, Cali, Colombia. Ruttan, V.W. (1982). Agrigulturgl Bepgprch Pollgy. University of Hinne- sota Press, Hinneapolis. Schoorel, A.F. (1972). ”Introduction and use of good seed of the right cultivars in developing countries.“ rppegglpgs of ppp Interna- glpnel Seed Igsting Apsociatlon 37 (1): 93-101. Schmid, A.A. (1983). Political Scopppy pf Pppllp lnveppsppp. Unpublished sanuscript. Schultz. 1w. <19“)- Wm- Yale Univer- sity Press, New Haven. Seville, E.P. and F.C. Guerrero. (1983). "Production of quality seed in the Philippines.“ Seed Scippce ppd Techpology 11: 1139-1143. Shaffer, J.D. (1970). "On the concept of subsector studies.“ Technical Sesinar on Subsector Hodelling of the Food and Agricultural Indus- try. Harch 30, 1970. Department of Agricultural Economics, Univer- sity of Florida, Gainesville. Siddiqui, N.R. (1983). ”Production of healthy seed in India.“ Spgd Sclpnge ang Tephnolpgy 11: 1063-1070. Spitz, R. (ed.). (1975). Case Studiep on Seep Industry Development of Eight Selected Countries. Food and Agriculture Organization, Rome. Stearman, A.H. (1973). “Colonization in eastern Bolivia: Problems and prospects." usan Organlzatiop 32 (3): 285-293. ________ . (1979). “Higrantes andinos en el oriente boliviano: El caso de Santa Cruz.“ merica Inpigenp 39 (2): 381-400. Sugden, R. and A. Williams. (1978). The Pginciples of Practipal Cost- geneglp Analysis. Oxford University Press, Oxford. Thospson, J.R. (1979). An lnproductiop go Seeg Technology. John Wiley and Sons, New York. 159 ________ . (1981). ”Business managesent - An aspect of seed techno- logy.” eed Sglpnce ang nghnolggy 9: 513-517. TiJerina H., A. (1975). ”Seed industry development in Hexico." In R. Spitz (ed.). Caps Sgygipp pp Sepg lpguppyy Develppsgnp of Elgpp Splppppp Countrlgs. Food and Agriculture Organization, Rome. Torrico, J.I. (1979). ”El trigo en Bolivia y el sundo.“ Prepared for USAID. La Paz, Bolivia. United States Agency for International Developsent. (1976). ““Toward a national wheat strategy for Bolivia.“ USAID, La Paz, Bolivia. ________ . (1980a). ”Bolivia - Agricultural sector report.“ USAID, La Paz, Bolivia. _______ . (1980b). ”Kitale Haize: The limits of success.“ A.I.D. Proaect Impact Evaluation Report No. 2. United States Agency for Interna- tional Development, Washington, D.C. ________ . (1985). “Tanzania seed multiplication." A.I.D. PrOJect Impact Evaluation Report No. 55. United States Agency for Interna- tional DeveIOpment, Washington, D.C. Valdes, A. and A. Siaswalle. (1981). ”Introduction.“ In Valdes, A. (ed.). Food Spppripy for Sevgloping Countries. Westview Press, Boulder, Co. Vaughan, C.E. (1972). "Review of the Rice Seed Program in Peru." Seed Technology Laboratory, Hississippi State University, State College. TA 72-13. Vaughan, C.E. and C.H. Dougherty. (1975). ”Ghana National Seed Progras Facilities.“ Seed Technology Laboratory, Hississippi State Univer- sity, State College. TA 75-08. Vaughan, C.E., et al. (1980). "Phase II, North Cameroon Seed PrOJect.“ Seed Technology Laboratory, Hississippi State University, State College. TA 80-11. Walker, D.D. and H.C. Potts. (1982). “Report of Tanzania Seed Pr03ect mission (Pronect 621-0092).“ Sponsored by the United States Agency for International Development, Industry Council for Development, Hississippi State University, and Tanzanian Hinistry of Agricul- ture. Walker, J.T. (1980). “Philosophies affecting the spread and development of seed production in the world.“ In P.D. Hevvlethwaite (ed.). Seed Pppducpion. Butterworths, London. Wayar, H. (1984). "Produccion de sesillas de cereales." Repppt: Second Rpupdtaple paetlpgs on Segg. 2-7 Septesber 1984. HACA/USAID/ Regional Seed Board, Santa Cruz, Bolivia. 160 Wail, T.E. (1974). pzpa flandpook go; Spllvip. Government Printing Office, Washington, D.C. Wennergren, E.B. and H.D. Whitaker. (1975). [he Spatps of Bolivlan pgplgplture. Praeger, New York. . (1977). ”Social return to U.S. technical assistance in Bolivian agriculture: The case of sheep and wheat.” pmeplgap Jppgpgl pf Agplculpugal Economips 59 (August): 565-569. Whitaker, H.D. and E.B. Wennergren. (1982). “Bolivia’s agriculture since 1960: Assessment and prognosis." In J .R. Ladsan (ed.). Hodern-Qay Spllvia: Lpgpcy pf the Revolution pnd Prospects go; the Futppe. Center for Latin American Studies, Arizona State University, Tesple: 233-254. World Bank. (1981). Accelerated Developsent in Sub-Saharan Africa. World Bank, Washington, D.C. ........ - “982)- W- World Bank. Washington Wortsan, S. and R.W. Cummings Jr. (1978). T9 Feed thls prlg; Ihe W- Johns Hopkins University Press. Baltimore. Zegarra, R. (1983). ”Actividedes interinstitucionales y de cooperecion a nivel nacional para el desarrollo del programa de semillas en el area de Santa Cruz - Bolivia.“ Report - First Roundtable Hastings on Seeds. August 1983: 76-82.