ABSTRACT AN ECONOMIC ANALYSIS OF SELECTED AGRICULTURAL EXTENSION PROGRAMS IN COLOMBIA, SOUTH AMERICA BY K. James McKenzie Thecflfiective of this study was to examine the effectiveness of selected agricultural extension programs in Colombia with a view to making suggestions for improving extension work in that country. The study focused on Possibilities for improving the effectiveness of programs Within existing budget allotments because of the scarcitY 0f funds for public programs in Colombia and the relatively u“lent need for devoting public funds to other uses such as roads, education, and agricultural research. All extension activities carried out by the Corporacic’m Autonomo del Valle C191 Cauca (CVC) and the Department of Agriculture of AM1m¥uazh11367 were examined using a cost—benefit approach- EStimates of program benefits were based on interviews of PrOgram Participants; program cost estimates were constructed from records of extension service expenditures. The major products of agricultural extension work Were Considered to include technical changes in farm firms, rural households, and community activities. Changes in the managerial CdPaCity of rural people was also regarded as a Prima rY Product. Secondary products were considered to K. James McKenzie include influences on institutions relating to (but not part of) rural society (e.g. , agricultural credit), assistance provided to rural peOple in establishing careers outside of agriculture, educational assistance provided to bring rural youth up to the levels of managerial and technical competence already achieved by their parents, and influencing the orientation of agricultural research work. However, only technical changes in farm firms and rural households were reduced to monetary terms in this study. The use of the adoption distribution curve provided a useful method of aggregating the benefits of extension pro- grams directed at producing technical changes in farm firms and rural households. Benefit flow estimates were constructed by estimating distributions of adopters with and without extension influence and then subtracting the benefit flow attributable to technical changes in the latter situation fI‘Om that which was obtained in the former. In general, this produced benefit flows which increased gradually to a peak and then decreased to zero over periods of ten to thirty years, If the estimated alternative distributions had been disregarded, the estimated benefit flows attributable to “tension efforts would have been infinite in duration. The use of a distribution curve representing adoption without EXt93131011 influence permitted estimating the marginal or added Value Of extension work given that research had already been Carried out. In all cases Optimistic and pessimistic values of K. James McKenzie benefit flows were made to test the sensitivity of the calculated benefit-cost ratios. All programs were evaluated using discount rates of 0%, 4% and 8% to test for possible program ranking reversals with changes in the interest rate. Six extension programs were examined in Valle del Cauca. Benefit-cost ratios calculated with a discount rate of 4% ranged from 1.06 to 14.81. The relative size of these ratios was quite insensitive to variations in the interest rate but, because some benefit streams were estimated with a much higher level of accuracy than others, some programs were much more uncertain than others and, depending on one's decision criterion, this could cause a change in rankings. The absolute size of the benefit-cost ratios varied substan— tially with variations in the interest rate and benefit flow estimates. The total range of these in response to all sensitivity tests was 0.08 to 16.84. Income distribution effects also varied somewhat among programs. However, all factors considered, it was concluded that the CVC's fertilizer program with tomato producers and the home economics program were the most successful programs. In view of the fact that some benefits were excluded from the calculations it seemed highly likely that benefits from most CVC extension programs were substantially in excess of their extension eXpenditure although the Apiculture and the Pacific Coast programs were highly questionable. Four field districts were examined in Antioquia. Benefit-cost ratios calculated with a discount rate of 4% K. James McKenzie ranQEd from 1.39 to 3.43 for extension work in these districts. The relative size of these ratios was quite inSensitive to variations in the interest rate, and moderately insensitive to variations in cost and benefit flow estimates. HOwever, the absolute size of the ratios varied substantially With changes in all three principle components. The total range of these ratios in response to all sensitivity tests was 0.43 to 13.96. No significant difference in income distribution effects were noted among districts. No con- Clusion was reached regarding the relative success of extension programs among districts. In View of the fact that some benefits were not included in the calculations, it seemed highly likely that the benefits from extension work in Antioquia were significantly greater than extension eXpend- itures although several opportunities for improving extension effectiveness were noted. The results of this study indicate that extension agencies in Colombia could make more effective use of their resources by directing technical agricultural programs at commercial (bUt not necessarily large) farmers, directing home economics programs at families earning 100 to 500 pesos per WeEk. and focusing on changes in community activities in regions Where infrastructure is poorly-deveIOped but agricultural production can be competitive in the national market. Technical agricultural programs directed at subsis- tence Producers and/or part-time farmers have relatively low PaY'Offs in relation to costs because of the lack of new K. James McKenzie telehnology of relevance to many such producers. Home ecol'lomics programs are often not applicable to the families 0f Commercial farmers or very low income families. Material assistance such as noted in Antioquia often costs more than it is worth to recipients. However, a number of qualificat- ions must be placed on these conclusions and, in general, there does not appear to be any good substitute for develOp- ing extension management capable of detecting high pay-off Programs as they arise. If the results of this study are to be used for Purposes other than measuring extension effectiveness, two major qualifications should be placed on the benefit-cost ratios measured herein. First, no allowance has been made in the cost estimates for research eXpenditures and, in the long-run, improvements in technology must cover these as well as extension costs. Secondly, extension expenditures undoubtedly underestimate the social costs of doing extension work. AN ECONOMIC ANALYSIS OF SELECTED AGRICULTURAL EXTENSION PROGRAMS IN COLOMBIA, SOUTH AMERICA BY TH E KSISames McKenzie A THESIS Submitted to . Michigan State University 1“Partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department oangricultural Economics 1972 © COpyright by KENNETH JAMES MCKENZIE 1972 ACKNOWLEDGEMENTS The successful completion of this study depended upon the assistance provided to the author by a number of persons and institutions. While he alone must assume the responsibility for any errors or omissions, there is no doubt but what he could not have completed it alone. The Rockefeller Foundation generously provided the financial assistance necessary to support the author through- out his Ph.D. program. Dr. G.I. Trant performed his services as thesis advisor in his usual capable fashion. Dr. G.L. Johnson provided essential guidance throughout the author's entire Program and especially during the final stages of thesis Preparation . The Department of Agricultural Economics, Michigan State University provided the atmosPhere and the physical facilities for the author's work. Personnel of the Cauca Valley Corporation and the Department of Agriculture of Antioquia provided assistance in cOntacting farmers and obtaining a variety of data. Prof. Ivan Agudelo, of the University of Antioquia, assisted in °rganizing the field work in Antioquia. Aldo Patruno and u' 13 Carlos conducted interviews. ii iii A Special note of appreciation must be provided for the many Colombian farmers and their families who cooperated in the interviews conducted in the study. A number of young ladies have participated in typing wevarious drafts of this thesis but Mrs. P. Brown had the ddflousjhonor of typing most of the final manuscript. Finally, the author must make special mention of his wiflaand family for their perseverence and support through- manm entirety of his graduate program. Their c00peration and encouragement played a significant role in the successful completion of this dissertation and the program as a whole. TABLE OF CONTENTS Page IJMPOF TABLES .. ...... ... ......... .. ..... . ............ ix LIST OF FIGURES O O O ..... O O O O O O O O O O O O O O ............ O O O O O xv CMWTER I INTRODUCTION 0 O O O O I O I O O O O O O O O O O O O O O O 0000000000 1 The Status of Agricultural Extension in COlombia O O O O O O O O O O O O O O O O O I O O O O O O ....... l The Environment in Which Extension Operates O O O O O O O O O O O O O O O O O O O O O O O O O O ........ 10 The PrOblem O O O O O O I O O O O O O O C O I O O O O O O O O O O O O O O 17 The Method of Analysis .. .................. 20 II THE "OUTPUTS" OF AGRICULTURAL EXTENSION WORK . 23 Extension Goals vs. Extension "Outputs" ... 23 The Change Agent Function ........... ...... 26 Technical Changes in Agricultural Prod— uction Units .............................. 29 The Static Production Situation ........ 30 Production Under Imperfect Knowledge ... 35 Managerial Changes ........................ 39 The Management Process ........ ......... 39 Managerial Changes Defined ............. 44 Changes in Rural Consumption Units (Households) .................. ..... . 49 Changes in Community Organizations . ....... 53 Secondary Extension Functions ............. 58 III THE PROCESS WHEREBY EXTENSION "OUTPUTS" ARE ACHIEVED ......................... ....... 62 The Systems Approach ..................... 63 Extension Activities as Systems Inputs ... 65 Extension Management .................. 66 Extension Program Operation ........... 68 Representation of Extension "Outputs" .... 70 Changes in Groups of Similar Decision-Making Units ................. 70 Other Extension “Outputs" ............. 75 iv GMPTER Page Environmental Influences on Extension Work .................. .......... .. ........ 76 Physical Conditions .................... 77 Product and Input Markets .............. 80 Non-Extension Information Sources ...... 82 Institutional Framework ................ 84 Idea Sources and Extension Activity ....... 87 An Extension System Model ........... ...... 88 n7 METHODS OF MEASURING EXTENSION "OUTPUTS" ..... 98 Measuring Changes in Individual Decision—Making Units ..................... 98 Technical Changes in Farm Firms ........ 98 Technical Changes in Rural Households .. 104 Changes in Community Organizations ..... 111 Managerial Changes ................ ..... 112 Aggregating Benefits Across Decision— Making Units ....... ......... .............. 117 Estimating Aggregate Benefit Flows Due To Changes in Rural Society .......... ..... 125 Secondary Extension Outputs . .............. 133 V MEASURING EXTENSION PERFORMANCE . ............. 137 A Decision Rule for Choosing among Alternative Extension Programs ........... 138 Comparing PVB/PVC Ratios for Alternative Extension Programs Under Imperfect Knowledge ........ ...... ................... 144 Selection Of Extension Programs on a Minimum Cost Basis .................. ...... 150 Incorporating Benefit Distribution Considerations into Program Performance Assessment ................................ 153 Measuring Managerial Performance of Extension Agencies ..... . ................ 158 Intermediate Measures of Extension Performance ....................... ........ 170 VI EXTENSION ACTIVITIES OF THE CAUCA VALLEY CORPORATION ......OOOOOOOOOOOOOOO ...... 176 The Setting in which the CVC Operates ..... 176 A Brief History of CVC's Extension ActiVities ......OOOOOOOOO ........ O ........ 185 Costs for CVC's 1967 Extension Programs ... 191 Description of 1967 Extension Activities .. 197 The Livestock Program .................. 197 The Apiculture Program ................. 202 vi CHAPTER Page The Fertilizer and Farm Planning Program O......OOOOIOOIOOOOO...0.0.0.... 206 Home Economics Assistance ..... ........ . 211 BOYS. 4-H Clubs oooooooooooooooooooooooo 222 Pacific Coast Programs .. ............... 231 VII .ANALYSIS OF CVC'S 1967 EXTENSION ACTIVITIES .. 235 Program Comparison by Benefit Classes ..... 235 The Livestock Program ......... ......... 236 The Apiculture Program ................. 238 The Fertilizer and Farm Planning Program ................. ........... .... 240 Home Economics Assistance .............. 241 Boys' 4-H Clubs .......... ..... ......... 242 Pacific Coast Programs ................. 243 Calculation of PVB/PVC Ratios ........ ..... 244 The Livestock Program .... .............. 246 The Apiculture Program ................. 248 The Fertilizer Program ................. 252 Home Economics Assistance .............. 258 Boys' 4-H Clubs ............. ........... 262 Pacific Coast Programs ............ ..... 265 Program Ranking .................. ...... ... 268 Factors Affecting Differences in Program Performance ............. ...... . ........... 270 Cost Differences ....................... 271 Benefit Values per Unit of "Output" .... 274 Diffusion Effects ...................... 276 Cost Minimization ......................... 277 Possible Cost Savings by Changing Office Locations and/or Transportation Methods ................................ 278 Possible Cost Savings by Changing Extension Personnel and/or Extension Media .................................. 287 \HII EXTENSION ACTIVITIES OF THE DEPARTMENT OF AGRICULTURE OF ANTIOQUIA ............... ...... 291 Characteristics of the Department ......... 291 Activities of the Department of Agriculture 298 Costs Of 1967 Extension Programs .......... 303 The Method of Estimating Program Benefits . 308 The District of Tinajitas ................. 314 Physical Resources ....... ..... ......... 315 Existing Technology .................... 317 Communications Channels ............. . 319 Agricultural Extension Activities . ..... 320 vii CHAPTER Page Other Districts Studied . ........... ...... 329 Buenos Aires .......................... 329 Pinguro ........... ...... .............. 337 urquité ......OOCOOOOOOOOOO ......... .0. 345 IX ANALYSIS OF EXTENSION ACTIVITIES IN ANTIOQUIA ............O...‘.....I....O0...... 351 Calculation of PVB/PVC Ratios .... ........ ‘ 351 The District of Tinajitas ........ ..... 361 The District of Buenos Aires .......... 363 The District of Pinguro .. ..... . ....... 368 The District of Urquité ............... 371 All Districts .......... ...... . ...... .. 372 Program Performance Ranking by District .. 375 Factors Affecting Differences in Program Performance .............................. 378 Cost Differences ......... ....... . ..... 378 Unit Benefit Values ..... .............. 380 Diffusion Effects ......... ............ 382 Type of Program ................ ....... 383 Cost Minimization ....... ........ . ........ 387 X POSSIBILITIES FOR IMPROVING EXTENSION EFFECTIVENESS IN COLOMBIA .......... ......... 392 Extension Programs ......... .............. 392 Agricultural Production ............... 393 Home Economics ........................ 404 Community Organizations . .......... .... 410 Other Outputs ............ ............. 412 Extension Management .............. ....... 415 Problem Definition .................... 417 Collecting Information for Decisions .. 419 Analysis of Information Collected ..... 422. Decision-Making ....................... 426 Execution of Programs ................. 427 Responsibility-Bearing ................ 428 XI SUMMARY AND CONCLUSIONS ......... ............ 431 The Conceptual Structure Used for Measuring the Effectiveness of Extension Programs ....................... 432 Extension Programs Directed at Agricultural Producers ................... 433 Home Economics Programs .................. 435 Community Organizations .................. 437 Managerial Capacity of Rural PeOple . ..... 439 BIBLIOGRAPHY ....... ............ ............. 440 APPENDIX A viii THE RATIONALE FOR SELECTION OF THE DECISION RULE USED IN THIS STUDY .......... SUPPLEMENTARY DATA ON THE 1967 OPERATIONS OF COVOCO'S EXTENSION SERVICE ......OOOOOOO SUPPLEMENTARY DATA ON THE 1967 OPERATIONS OF THE EXTENSION SERVICE OF THE DEPART- MENT OF AGRICULTURE OF ANTIOQUIA .......... Page 446 459 467 LIST OF TABLES TABLE Page 1 POpulation distribution by size of center in Colombia in 1964 ............ ......... . ...... 12 2 Distribution Of peOple economically-active in agriculture in Colombia in 1964 by occupation .................... ...... ... ...... .. 13 3 Distribution of farm units, total land area, and inhabitants by size of farm in Colombia 1111959 00...... 0000000000 ...... ......... C ...... 15 4 Example of the calculation of the benefit flow resulting from a hypothetical set Of extenSj-on actiVities ......OOOOOOOOOOOOOO ....... 131 5 The relationship between management and program expenditures and related variables for an extension budget of $100,000 ............ 163 6 Estimated land use pattern for land in farm units in Valle del Cauca in 1956 ....... ........ 180 7 Number and area of farm units in Valle del Cauca in 1959 according to size and main type Of activity ....................... ........ 181 8 Distribution of population by age and education levels for rural areas of Valle del Cauca in 1964 .............................. 183 9 CVC revenues and expenditures: 1955 and 1963 ... 184 10 Revised estimate of the costs of Operating the extension service Of the CVC for the period April 1, 1967 to March 31, 1968 ... ...... 192 ll , . . Field personnel time, average daily costs, and total costs by class of field personnel .... 196 12 ‘Numbers of farmers co-Operating in the CVC 1967 fertilizer program by semester, crOp, and extension area ........................ ..... 208 ix TABLE 13 14 15 16 17 18 19 20 21 22 23 24 25 Summary of results of CVC's 1967 fertilizer program for participating farmers ............ Number of club members interviewed by level Of instruction received, level of prior knowledge of instruction received, and area of instruction .......................... Levels of utilization of the eight areas of instruction Offered to 1967 home economics Club memers 00.0.0000.........OOOOOIOOOOOIOOO Number of club members in the sample who rated the eight areas Of instruction first, second, or third in order of utility for them ....O.IO...0.0.......OOOOOOOOOIOOOOIOOOO. Numbers of boys' 4-H club members inter- viewed by level Of instruction received, level of prior knowledge of instruction received and area of instruction ............. Levels of utilization of the eight areas of instruction Offered to boys' 4-H club members in 1967 0..........OOOOOOOOOOOOOOOOOO. Number of boys' 4-H club members in the sample who rated the eight areas of instruction first, second, or third in order Of utility for them .......................... Comparison of 1967 CVC programs by benefit Classes ......OOOOOIOOOOOOOOO......OOOOOOOO... Calculated PVB/PVC ratios for the livestock program under selected sets Of assumptions ... Calculated PVB/PVC ratios for the apiculture program under selected assumptions ........... Estimated benefit flow for CVC's 1967 fertilizer program directed at tomato prOducers ......IO......OOOOOOOOOOOOOOO......O Estimated benefit flow for CVC's 1967 fertilizer program directed at corn prOduCtion OI.0.0.0...........OOOOOOOOOOOOOOOO Calculated PVB/PVC ratios for the tomato and corn sections Of CVC's 1967 fertilizer program ......OOOOOOOOOOOOOIOIO......OIOOOIOO. Page 210 217 220 222 227 229 230 237 249 251 253 256 257 xi TABLE Page 26 Calculation of average annual benefits for participants interviewed in the home economics assistance program ............ 259 27 PVB/PVC ratios for the home economics assistance program for several levels of benefits per participant, length of benefit flow, and interest rate .............. 262 28 Calculation of average annual benefits for participants in the boys' 4-H club program ... 264 29 PVB/PVC ratios for the boys' 4-H club program for several levels of benefits per participants, length of benefit flow, and interest rate ................................ 265 30 Projected costs and basic herd distributions for the buffalo project under selected assumptions O0.0.0.0.........OOOOOOOOOOOOO.... 267 31 Summary of performance measures for CVC's extension programs ........................... 269 32 Comparison of several unit measures Of total costs among 1967 CVC programs ................ 271 33 Incidence of mention Of various factors considered to limit utilization Of knowledge in the home economics assistance program ..... 274 34 Proportion of time spent in various activities by class of personnel for CVC's extension service ............................ 279 35 Costs per day of "Productive Field Personnel Time" by class of personnel for the existing transportation system and an alternative ..... 282 36 Estimated land use pattern for land in farm units in the Province Of Antioquia in 1965 ... 294 37 Average costs per day of field personnel time by class of personnel for the rural extension section of the Department Of Agriculture of Antioquia in 1967 ............. 306 3 . . 8 Total extension costs, number of families, and average extension costs per family for 1967 programs by type of family, for the six TABLE 39 40 41 42 43 44 45 46 47 48 49 50 xii districts serviced by the sectors of San Jeronimo and Santa Fe de Antioquia ........... Summary of extension participation among sample families in the four districts ........ Distribution of land among the families Of Tinajitas 0.0.0..........OOOOOOOOOO0......O Method for determining extension program priorities in the district of Tinajitas ...... Summary of results of 1967 extension programs directed at the production activities of eighteen Tinajitas families .... Results of extension programs directed at consumption activities in the district of Tinajitas in 1967 for eighteen sample families ..................................... Results Of 1967 extension programs directed at the production activities of the fourteen families interviewed in Buenos Aires .0...O.......IOOOOOOOCOOO00.0.0... ...... Results of extension programs directed at consumption activities of the fourteen sample families interviewed in Buenos Aires .. Results of 1967 extension programs directed at the production activities Of the thirteen Pinguro families interviewed ........ Results of 1967 extension programs directed at the consumption activities of the thirteen Pinguro families interviewed ........ Results of 1967 extension programs directed at the production activities Of the nine Urquité families interviewed .......... ..... .. Results Of 1967 extension activities directed at the consumption activities of the nin sample families in Urquita ........... PVB/PVC ratios for extension work in Tinajitas for several different assumptions regarding costs and benefits ................. Page 307 310 316 321 323 326 332 335 341 343 346 349 364 TABLE 51 52 53 S4 55 56 57 58 59 xiii PVB/PVC ratios for extension work in Buenos Aires for several different assumptions regarding costs and benefits ..... PVB/PVC ratios for extension work in Pinguro for several different assumptions regarding costs and benefits ................. PVB/PVC ratios for extension work in Urquité under several assumptions regarding costs and benefits I00............OOOOOOOOOOOO PVB/PVC ratios for extension programs in the four districts Studied .....OOOOOOOOOOOOOOOOOO Summary Of performance measures for the Department's extension programs .............. Costs per participating family and per resident family of extension programs examined in this study ....................... Benefits per family interviewed for the districts studied 0..........OOOOOOOOOOOOOOOO. Summary of benefits by program among 54 sample families in four districts ............ Percentage distribution Of field personnel time by type Of activity and class Of personnel for the sectors Of Santa Fe de Antioquia and San Jeronimo in 1967 ........... Breakdown Of the costs Of Operating the Extension Service Of the C.V.C. for the period April 1, 1967 to March 31, 1968 ....... Breakdown Of Fringe Benefits (Prestaciones Sociales) for C.V.C. in 1967 ................. Costs Of Operating C.V.C. vehicles in 1967 ... Program costs and number Of days of field personnel time by program for the Extension Service Of the C.V.C. during the period April 1, 1967 to March 31, 1968 .............. Results Of C.V.C.'s 1967 fertilizer program with tomato producers ........................ Page 367 370 373 374 376 379 381 384 390 460 461 461 462 463 TABLE 8-6 C-2 xiv Results of C.V.C.'s 1967 fertilizer program With corn prOducers O O O O O O O O I I I O O O O 0 Summary of Club Membership statistics for Program #4, Home Economics Assistance and Progran‘ #5, BOYS. 4"}! 0111138 0..............0 Comparison of selected extension cost and resource use factors for the Home Economics Assistance program and the Boys' 4-H clubs program. (Costs are in Colombian pesos.) Field personnel time in days ............... Estimated costs of operating the Extension Service of the Department of Agriculture OfAntioquj-ain1967 ......OOOOOOOOOOOOOOOOO Calculation of transfers to the Extension Service division of the Department of Agriculture from other departments of the Government of Antioquia .................... Calculation of administration and research costs attributable to the Extension Division of the Department of Agriculture of Antioquia in 1967 ....................... Calculation of average total costs per day of field personnel time for the various classes of extension workers employed by the Department of Agriculture .............. Estimated costs of operating 1967 extension programs in the districts serviced by the sectors of San Jeronimo and Santa Fe de Antioquia .................................. Distribution of man-days of field personnel time by office location for the Extension Service of the Secretary of Agriculture of Antioquia in 1967 .......................... Calculation of projected benefit flows from extension work in Tinajitas ........... Calculation of projected benefit flows for extension work in Buenos Aires ......... Calculation of projected benefit flows for extension work in Pinguro .............. Calculation of projected benefit flow for extension work in Urguita .............. Page 464 465 466 468 469 470 471 472 474 475 476 477 478 LIST OF FIGURES Figure Page 1 A graphical illustration of a system model ... 64 2 The extension agency component of an extension system model ....................... 66 3 Classification of adopters on the basis of a normal density function ................. 72 4 Illustration of the influence of extension efforts on the adoption of a hypothetical innovation ......OIOOOOOOOOIO......OIOOOOOOOO. 73 S A general view of an extension system model showing the relationship between extension costs and benefits ........................... 91 6 Illustration of the effects of a shift in the supply curve resulting from the adoption Ofaninnovation 00............OOOOOOOOOOOOOOO 120 7 Illustration of the effects of a shift in the demand for farm inputs or consumer's goods by the farm sector ..................... 122 8 Map of Valle del Cauca showing principal geographical areas and office locations of the CVC's extension service in 1967 .......... 178 9 Map of Antioquia showing office locations of the extension service of the Department Of Agriculture ......OOOOOOOO0.0.0.000...0.... 293 A-l Illustration of difference in project selection procedures between the constrained and unconstrained case ....................... 451 A-2 Comparison of PVB/PVC ratios as functions of interest rate for two projects ............ 453 XV CHAPTER I INTRODUCTION The problem confronted in this study was one of suggesting ways and means of improving the effectiveness of agricultural extension work in Colombia. As a prelude to defining this problem more precisely, and devising a umans of attacking it, the author attempted to familiarize himself (via existing literature) with the status of Colombian agricultural extension activities and the environment in which these are carried out. The following discussion permits the reader to join him. The Status of Agricultural Extension in Colombia In an effort to "increase the agricultural prod- uctivity of the country and the welfare of the rural peOple",1 a number of official and semi-official Colombian agencies have been permitted to take on responsibilities in the area of agricultural extension and direct technical assistance to farmers. Most of these agencies have acquired these responsibilities only in the last two decades. However, a few others were active in some form of technical assistance work prior to World War II. The institutional 1See footnote 1, Chapter II. 1 structure of Colombia is by no means a static one. A number of changes have taken place in recent years in an effort to find a really workable system. Agricultural extension is one activity which often becomes involved in such changes. In 1967 the extension service of the Federal Department of Agriculture was transferred to ICA2 and, in 1968 a fairly large expansion program was begun. This agency is supposed to serve as the co-ordinator of agricultural research, education, and extension in Colombia. Its employees report they do not know for sure how many other Colombian agencies are doing extension work in the country. Major extension programs, which are well-known, are operated by the National Coffee Federation, the Department of Agriculture in Antioquia, the Cauca Valley Corporation, ICA itself, and several producer associations. Agricultural extension activities in Colombia are complemented by work in several related fields of endeavor. Two of which bear mention here are formal agricultural education and community development. The former area includes about 15 universities with agricultural and related faculties, three agricultural normal or technical schools, about 40 agricultural high schools, and SENA (Servicio Nacional de Aprendizajé), which devotes part of its resources to agricultural courses. The universities are currently pmoducing about 550 agricultural professionals annually, of 2Instituto Colombiano Agropecuario (Colombian Agric- ultural Institute). which 70 per cent are ingenieros agronomos. About 80 per cent of these graduates are emploved bv official or semi- official agencies while the rest find jobs in the private smctor. The annual output of technical agricultural guaduates is much larger than that of professionals. A large prOportions of these people are also emploved in the pmblic sector although a significant number of the graduates cfi’SENA and the agricultural high schools find jobs in industrv and on farms as well. One of the most notable features of agricultural education in Colombia is that educational programs directed at training agricultural professional and technical personnel enrol verv few students vdth farm backgrounds. Very few rural voung peoole obtain mfificient basic education to permit them to enter such courses. As a result of this combination of circumstances nmny Colombian extension workers come from urban areas, have no universitv training, and belong to a much different social class than manv of their clients. A wide range of communitv development programs Operate in Colombia.3 Some of the agencies Operating these programs include CARE, Accion Comunal, Accion Cultural Popular, the Peace Corps, the National Coffee Federation, and the Rockefeller Foundation. Accién Comunal comes under the supervision of the Ministerio de Gobierno. By 1964 about 3For more details see "Colombian Communitv Deve10p— ment,” A Survev Report submitted bv CARE Inc. in fulfullment of an agreement with the National Coffee Federation, December 1960. 9,000 local boards had been established, about 450 full- time communitv agents were employed, and investments by local tnards were between 10 and 13 million dollars - about the same as total government investments in the States.4 Accidn Cultural Popular Operates an extensive radio training program flu'campesinos and publishes the largest circulation farm Imper in Colombia. The National Coffee Federation has been way active in community develOpment in most coffee—growing aueas of Colombia. In addition to its large agricultural mdension service it also builds roads and schools, and carries out various other such projects. In addition to Official and semi-Official activities filthe areas of agricultural extension and direct technical amsistance to farmers, a large amount of private resources aue directed along these lines in Colombia. These usuallv emanate from agricultural consulting firms, companies selling inputs to farmers, and companies buying products from farms. They are primarilv directed at commercial or larger-volume cmoducers. Agricultural consulting firms sell their service to farmers who are willing to pav for it and to farmers who are required to have it to obtain government credit. Fertilizer, chemical pest and disease control, and feed suppliers predominate among the input sellers who Offer tech- nical advice. Some Of these companies are well-established and very active in providing advice to large commercial ¥ 4Matthew D. Edel, "The Colombian Community Action Irograms, an Economic Evaluation," Unpublished Ph.D. Dissert- Mflon, Yale University, 1967, Chapter 2. producers. Throughout their recent existence Colombian agric— Lutural extension services have been regarded by some observers as inefficient and ineffective institutions. In 1358 a Mission organized by the International Bank for Ikmonstruction and DevelOpment reported the following: The Mission found the extension work of the Ministry of Agriculture and of the Departamentos disappointing in general. Rarely was there any real co-ordination of program or execution although they were working side by side; the staff of both seemed to lack an adequate appreciation of the proper approach to extension; superior guidance by higher personnel in the making and execution Of programs was insufficient; too little time was spent in the field among cultivators and too much in the Office, either because of the burden of administration or of disinclination for fieldwork, or of both com- bined; the staff of both ingeniero agronomos and Of practicos, particularly O?*Ehe latter, was far too small in numbers to be effective.5 The mission recommended that the Federal Department of Agriculture appoint a number Of technical Officers to provide high level guidance and advice for extension activities; that action be taken to reduce the high rate Of turnover of basic extension personnel; that an agricultural extension advisory board be created within the Department of Agriculture to co-ordinate the work of the various extension agencies; that the quality Of existing staff be upgraded; and that the number of personnel be markedly increased. During the next few years a number of these recommendations were implemented __ 5International Bank Of Reconstruction and Develop- ment, Agricultural DevelOpment in Colombia (Washington: I.B.R.D0 I 1956) I p. 15?. in various forms. Budgets were enlarged, staffs were increased, and the activities of Servicio Tecnico Agricola Colombiano Americano (STACA) were expanded.6 In 1961 the Commission on Higher Agricultural Emucation in Colombia made the following comment about the extension service of the Federal Department of Agriculture asit.was then operating: The organization cannot at present be said to be functioning properly. The personnel are not well trained and lack capacity. A great deal Of time is spent in writing reports. Facilities for extension are few; most of the budget is spent on salaries. Formal programs of work are theoretically developed; there are few records of farmers contacted nor estimates of the economics of the service in terms of improvement in production. Co—Ordination with DIA has been lacking.7 The Commission concluded that "more productive extension services are urgently and immediately needed in Colombia."8 Ixnumber of recommendations were made with a view to increasing the abilities of current extension workers and upgrading the educational programs by which people usually entered extension work. A study of five Colombian extension 6STACA was a co-Operative U.S. - Colombian project designed to upgrade Colombian agricultural extension work. Both field extension programs and training programs for extension workers were carried out. STACA no longer exists although many Of the extension workers who received training in this project are now employed by various agencies through- out Colombia. 7The Comission on Higher Agricultural Education, Higher Agricultural Education in Colombia, (Bogota, Colombia: A.mimeographed report by the:Commission, 1961). p. 44. (DIA was another government agency responsible for some agric- ultural programs.) 81bid., p. 91. agencies published in 1962 reached a similar conclusion and made eighteen specific recommendations designed to remedy the situation.9 A 1964 study of land settlement in Eastern Colombia concluded that: There is a great need for more effort in the field of technical assistance. The present extension workers know little about agriculture in general and even less about trOpical agriculture. Effort in the field of credit is likely to be ineffective unless means are provided for increas- ing the productive level of the farm —— that is, by providing more technical assistance. A 1966 study of the extension service of the Federal mnmrtment of Agriculture concluded: 1. That regional offices have used very changeable formats for presenting their monthly reports. 2. That there hasn't been sufficient responsibility in the presentation of these reports. 3. That extension workers have had no knowledge of some extension methods and there has been wide disparity in Opinions of how to use some Of these methods, such as movies. 9Joseph Di Franco and Roy A. Clifford, An Analytical Study of Five Extension Qrganizations in Colombia, (Turrialba, Costa Rica: Inter-American Institute of Agricultural Sciences of the O.A.S., 1962): PP. 55-60. 10Ronald L. Tinnermeier, New Land Settlement in the Eastern Lowlands OfColombia (Madison, Wisconsin: University of Wisconsin Land Tenure Center, Research Paper #13, 1964), p. 50. 11Ana Lucia Garcia de Roman, Evaluacién del Uso de Algunos Metodos Utilizados por los Extensionistas del Ministerio de AgricUltura de Colombia (Unpublished thesis for degree of Ingeniero Agronomo, Universidad de Caldas, 1955), p. 92, (translation by this writer.) m... 0.“. A 1966 USAID report suggested that: There is a great need to train more professional extension teachers at the University level. Exten- sion work must develop a stature and prestige comp- arable tO that of research or university teaching to attract competent peOple. Professional extension workers are needed for major administrative posts in existing extension agencies, for university teaching positions in the agricultural universities, for jobs in the agricult- ural normal and vocatingl schools and in the vocat- ional training centers. Other observers point out that, in spite of the limitations of poorly-qualified extension workers, lack of co-ordination among agencies, and insufficient resources to provide extension services in many municipalities, Colombian extension efforts have been, in some respects, fairly successful. A 1966 study makes the following remark: Among Latin American professional circles, the C.V.C. Agricultural Extension Service has the reputation of being the best in Latin America. Its contribution toward the development of agriculture and rural living in I38 Cauca Valley has certainly been significant.... The authors go on to recommend expansion of this service. A 1967 study characterized the extension program of the National Federation of Coffee Growers as the most successful in Colombia. 12USAID, "Agricultural Development in Colombia," a report prepared by the Rural Development Staff of USAID, Bogota, Colombia, 1966, p. 47. 13Antonio J. Posada and Jeanne Anderson Posada, The CVC: Challenge to Underdevelopment and Traditionalism (Bogota, Colombia: Tercer Mundo, 1966)T p. 119. This study suggests that there are several reasons for the greater success of this organization: First, it works only in coffee areas - this reduces the territory to about one-third of the municipalities in the country. Second, it has moved into the area of building infastructure which local governments usually fail to provide. And third, it has a stable source of income through a six per cent tax on all coffee eXport sales. Another 1967 study, reported on extension activities in the municipality of Giradota, State Of Antioquia, found that most peOple in the area were aware of what the Extension Service was doing and were favorably impressed by these activities.15 Most foreign Observers agree that continuing efforts to improve Colombian extension services are desirable and that this institution can play a much more effective role in the development Of the country than it is now doing. Admin- istrators of many Colombian extension agencies also appear to adopt this view. The problem arises, however, in deciding what changes are most apprOpriate, given the exist- ing situation. Earlier recommendations were quite general and appear to have been geared to making the Colombian extension system more like the U.S. system on the implicit assumption that what appears to have worked well in the U.S. 14Herman Felstehausen, "Fitting Agricultural Extension to Development Needs: The Colombian Problem", Research Paper #39 of the Land Tenure Center, Madison, University of Wisconsin, 1967, p. 13. 15Ivan Agudelo, "Evaluacién de Algunos Labores de Extension en el Area de Giradota, Colombia," Unpublished Master's thesis, Inter-American Institute of Agricultural Sciences, Turrialba, Costa Rica, 1967, pp. 75-77. 10 would work well in Colombia. Later studies, like that of Felstehausen, have been more specific and have served to point out why this assumption is not, in many cases, a valid one. Felstehausen emphasizes that much of the rural infa- structure, which provides an apprOpriate environment for agricultural extension work as it is carried out in the United States, is absent in many areas of Colombia. He also suggests that "neither foreigners nor Colombians know enough about what will work in Colombia."16 The Environment in Which Extension Operatesl7 As a result of its highly variable terrain, which ranges from coastal plains to snow-capped mountains, Colombia boasts a wide variety of climatic zones. Many individual states have almost as many climatic zones as the country itself. However, as the country is situated at the equator, seasonal variation in temperature within all of these zones is slight. The country's total area is about 114 million hectares of which only 27 million is divided into farm units. 16Herman Felstehausen, Op. cit., p. 19. 17More detail on this particular tOpic may be found in "Colombian Agriculture", a mimeographed report prepared by the Office of the Agricultural Attache, U.S. Embassy, Bogota, Colombia; "Notes on Recent Developments in Colombian Agric- ulture", a mimeographed report prepared by José Americo Castillo and Gerald I. Trant, Agricultural Economics Section, Faculty of Economic Sciences, University Of Valle, Cali, Colombia; and "Agricultural DevelOpment in Colombia", a report prepared by the Rural Development Staff, USAID, Bogota, Colombia. . ... I I ...,“ 4 ‘. I" '~ 11 The Eastern half of Colombia (the Llanos) is sparsely pOpulated and most agricultural land in this area is devoted to cattle raising. (About one-third of the country's permanent pasture is located in the Llanos.) It is often suggested that this area has a potential for a much more intensive type of agriculture than it now supports. The Western half Of Colombia is divided by three rugged mountain ranges but four-fifths Of the present pOpulation lives in this part of the country. In this region, sizeable areas Of good, level agricultural land are found in the Magdalena and Cauca Valleys, the Savana de Bogota, and the Magdalena Delta along the Carribean Coast. Some of the larger farms in these areas are now using a number Of improved practices including modern machinery. Much of the hilly and mountainous land in this region is also farmed and this accounts for a large meportion of the production of some crOps, particularly coffee. Colombia's gross agricultural product is increasing at about the same rate as her population. However, agric- ulture's share of gross national product is declining due to the more rapid rate Of growth in other sectors. In 1965 agriculture accounted for about 30 per cent of GNP. The principal components Of this output were coffee, cattle, corn, rice, sugar cane, plantain, and cotton, although cassava, wheat, beans, potatoes, barley, bananas, cacao, sesame, sisal, soybeans, tobacco, milk, lemons, oranges, pineapples, and various vegetables were also produced in substantial 9 no... ..., ...-1 12 quantities. Coffee is Colombia's major export commodity, but bananas, sugar, cotton, tobacco, cattle and a few other agricultural products also earn some foreign exchange. Part of the domestic demand for agricultural products is met by imports, principally from the U.S.A. However, in 1966 agricultural exports exceeded agricultural imports by 346.3 million dollars (U.S.) thereby providing foreign exchange to pay for a wide variety of non-agricultural imports. The rural - urban distribution of population in Colombia as a whole is similar to that in most Latin American countries. Table 1 gives a description Of population distribution by size of center in 1964. Most centers of less than 20,000 people have no significant industries and are essentially service centers for agricultural areas. Table l.-—Popu1ation distribution by size of center in Colombia in 1964.a . No. of % Of Size of Center People Total fibre than 20,000 inhabitants 6,322,984 36.1 Between 1,000 and 20,000 inhabitants 3,244,227 18.6 Rural areas and centers Of less than 1,000 inhabitants 7,917,297 45.3 Total 17,484,508 100.0 aDepartamento Administrativo Naciénal de Estadistica (DANE , XII Censo Nacidhal de ROblacién (Juliogls de I964), Resumen General (Bogota: Cblombia: Imprenta Naci6nal, 1967), p. 32. 13 Many of the income-earners who reside in these towns are either owners of larger farms or farm labourers. Thus, over 60 per cent of the population may be regarded as "rural". Census data indicate that 46.8 per cent of the economically-active population of Colombia was employed in agriculture in 1964. The distribution of these people among various occupations is shown in Table 2. It is evident from this data that Colombian agriculture relies very heavily on hired manual labour as an input. Moreover, in addition to the 967,826 hired laborers, a large proportion of the 1,074,410 independent workers and family helpers were probably also doing manual work. Thus, although large tractors and combines are found on the sugar plantations and larger rice farms in Colombia, probably about 85 per cent Of the people employed in agriculture in 1964 worked as hired manual laborers or as Table 2.—-Distribution of people economically-active in agriculture in Colombia in 1964 by occupation. W . NO. of % of Occupation Persons Total Employer 307,772 12.8 Independent Worker 706,649 29.4 Family Helper 367,761 15.3 Manager or Foreman 48,081 2.0 Manual Laborer (Employee) 967,826 40.2 Others 7,178 0.3 Total 2,405,267 100.0 aXII Censo Nacional de Poblacién. 14 owners or renters of small plots Of land where human labor was the principal input.18 Many Of the remaining 15 per cent appear to be involved either totally or partially, in super- vising hired laborers. As illustrated in Table 3, 57.9 per cent of the farm units in Colombia were less than four hectares in size in 1959. These farm units contained only 3.6 per cent of the total land but provided homes for 48.6 per cent of the total farm pOpulation. (Probably a large prOportion of the 1,074,410 independent workers and family helpers recorded by the 1964 Census were in this group). In the same year, 1.7 per cent of the farm units controlled 55.1 per cent of all agricultural land.19 Colombia's land distribution problem has received much attention in recent years, but the situation probably has not changed greatly since 1959. Many of these larger farms are devoted to cattle-raising.20 Some of them utilize large tracts of reasonably good agricultural land for pasture, while peasants with small holdings use parts of the steep mountain sides for crOps such as corn and potatoes. leor the 1,209,672 farm units recorded by DANE's 1959 Censo Agropecuario, 64.6% relied on human power alone while only 2.5% were classified as mechanized. The other 32.9% used various combinations of human, animal, and mechanical power. 19Recent studies by INCORA (the National agency in charge Of land reform) suggest this is a biased estimate because some Of the large holdings exist on paper but not in reality. ' 20DANE's 1959 Agricultural Census reported that 72.5% of the farm units over 30 hectares raised cattle while only 32.2% Of the farms under 30 hectares were involved in this activity. 15 .mm .m ppm oNt.m .vw a .Hmcfimmumm pumwauasz “ ammOHoo .muomom~ pwaa .Aowumsommoumm Omsmuv mOAHmpom oumm mmcowomufloamxm mp Happwomz Owuouomuao .mofiumflpmumm Op Hmcwwomz O>flumuumfispr¢ oucmfiuummmam o.ooa vm~.mmm.m o.ooa www.5mm.bm o.ooa www.mom.a H6009 m.m www.mmw H.mm ~on.mvo.ma h.H mmm.o~ mmumuown com cps» who: H.~H Hem.nm> w.w~ one.a>~.h m.m beo.moa mommuoon com I on m.mm wem.n~m.~ h.vH mm~.~mo.v m.am mom.mmm mmumuomn on I v m.mv ~mm.hma.m w.m -H.bmm m.hm www.mmw mmumuoms q smnu mama w uwQEsz w Hones—Hz w monasz when: sums Eumm mo ouwm mucmuwnmncH pond puma Hmuoa m.mmmH ca mwaOHOO :H show up when an mucmuwnmccw ppm .mmum papa Hpuou .muflca Eumm mo cowusnwuumaolu.m mange 16 The reasons for the existence of this situation are too involved to explain here.21 However, the situation itself is both a reason for investing in agricultural extension and a reason why extension's task is a very difficult one. There are many other aspects to the environment in which Colombian extension programs must be Operated. Inad- equate tranSportation networks, restrictive marketing systems, ineffective local government, serious health problems, and low education levels are the rule rather than the exception. Health problems are such that in 1960 average life expectancy for Colombia as a whole was about only 50 years compared with 69.5 in the United States.22 Other parts of the Colombian scene include a highly class-determined social structure, limiting legal codes, and the memory of the Violencia. 21For an interesting discussion of the evolution of this situation see Colombia: Social Structure and the Process of Develgpment, (Gainesville, Florida: University of Florida Press, 1967) by T. Lynn Smith. 22United Nations, El Desarrollo Economico de American Latina en la Post - Guerra, A Unitediations Report dated December, 1963, p. 612 23The "Violencia" is now apparently at an end, although a few murders and acts of bandititv associated with "Violencia" activities still occur. The "Violencia" has been described as a "peculiarly Colombian phenomenon" as its origin lay in the Colombian political system and social structure. It involved illegal murders, torture, thievery, and destruction of property. During the past 30 years, approximately 200,000 peOple lost their life in the "Violencia". Most of these were from rural areas and, as a result, a noticeable migration to cities, such as Bogota, Cali, and Medellin occurred. For further details see Chapter 13 of Colombia: The Politigal Dimension of Change (New Haven: Yale University Press, 1967) by Robert H. Dix. 17 It is within this environment that agricultural extension must function to effect what one writer calls "practical and worthwhile changes given the existing system". The Problem There are two major problem areas associated with using agricultural extension as a tool to effect such "practical and/or worthwhile changes". First, there is the question of whether more (or less) resources should be devoted to this activity. The author does not pretend to attack this problem for it would of necessity, involve an examination Of alternative and complementary programs - a task much too large for a doctoral dissertation. Second, 25 there is the problem Of how efficiently extension agencies utilize existing budgets. This involves questions Of what extension workers do and how they should pursue their tasks. It will be recalled that the recommendations of earlier studies usually included, either directly or implicitly, the expansion of extension budgets in spite of their lack of any analysis of alternative or complementary programs. The author believes this to be an error and chooses to attack the problem of extension effectiveness by examining the second problem area listed above - the efficiency of 24Stephen Enke, Economics for DevelOpment (Englewood Cliffs, N.J., Prentice-HaII Inc., 1963): p. 148T’ 25The meaning of the term "efficiently" rests on the conceptual framework developed in the first five chapters of this study and cannot be precisely defined at this point. «- ... ‘. . u I 18 extension work. This does not mean to imply that the expansion of extension budgets may not be apprOpriate or important. However, it does mean to imply that resources in Colombia are scarce and, while better results 33g be Obtained with larger budgets, they may well be Obtained at a very high Opportunity cost (e.g., fewer schools or roads). On the other hand, any gain in extension efficiency is all net gain. Unfortunately, no complete, workable, theoretical and methodological framework specifically-adapted to measuring extension efficiency in economic terms has been available to decision—makers who must choose between extension programs or select the means of carrying out these programs. Agricultural extension workers, particularly in North America, have for some time been carrying out formal evaluations of their own work.26 These evaluations have usually been directed at determining extension impact, comparing extension methods, or improving specific tech- niques. All Of these functions are useful but they are also restrictive. They do not, for example, permit the compar- ison of 4-H Club programs with farm account projects to determine if one should be expanded at the expense of the other. In recent years some economists have become 26See Chapter 33 of The Co-Operative Extension Service, (Englewood Cliffs, N.J.z Prentice-Hall Inc., 1966). edited by H.C. Sanders. 19 interested in evaluating extension efforts.27 Such activities have occurred mainly in connection with programs in under-developed countries and have been directed guimarily at obtaining rough ideas of the general circum- stances under which agricultural extension may be usefully employed as a tool in accelerating agricultural deve10pment. These approaches have, on the whole, been quite useful and extremely relevant for development planning (i.e., selecting the size of the extension budget) but they are of little use to extension personnel who must choose between alternative extension programs. The main reason why no such structure has yet been developed is that the nature of the benefits resulting from extension work is extremely complex. Agricultural extension work is essentially an educational function and the problems of objectively evaluating the products of educational . . . 29 processes in general are Widely recognized. However: 27See, for example, "A Case Study of the Economic f Impact of Technical Assistance". Unpublished Ph.D. dissert- ation, University of Chicago, 1958, by Clifton R. Wharton, Jr. 28Perhaps the most relevant and ambitious effort in the area of measuring extension effectiveness was the controlled experiment with intensive agricultural extension work carried out in Michigan in the late 1950's. For more details on this program see The Change Agent and The Process of Change, Research Bulletin 17, Michigan State University, Agricultural Experiment Station, East Lansing, Michigan, 1967 by James Neilson. However, the Michigan study was an extremely pragmatic one and several potentially important conceptual considerations were entirely omitted. 29E.A.G. Robinson and J.E. Vaizey (editors), The Economics of Education, (New York: St. Martin's Press, 1966). 20 tmcause of the critical significance of agricultural gmogress in the developing countries of the world, it is emsential that resources devoted to extension be utilized am effectively as possible. Thus, it appeared to the author that any small measure of progress in measuring extension efficiency might have far-reaching implications and, for this reason, this problem was selected as the topic of this dissertation. The Method of Analysis This problem was approached in two major steps. First, an attempt was made to organize a conceptual frame- work for analyzing the effectiveness of extension work. Second, with the aid of this conceptual framework, selected Colombian extension activities were examined with the hOpe that, by measuring their relative performance, hypotheses regarding ways for improving extension effectiveness might be advanced and the usefulness of the conceptual structure might be analyzed. The thesis is organized in the following fashion. The next chapter begins with the general goals acknowledged by extension workers and legislators who have created extension agencies and then constructs a classification of the usual functions of agricultural extension. Specific definitions of the "outputs" associated with these functions are also developed. In the third chapter, a description of the way in which these "outputs" are achieved is advanced. '1 21 This provides a basis for relating particular sets of "outputs" to specific extension activities and permits recognition of the many possible causes of variability in extension performance. In the fourth chapter, methods for measuring extension "outputs" are developed. In the fifth chapter, measures of extension "outputs" are incorporated with costs to arrive at measures of overall extension per— formance. These five chapters form the conceptual frame- work which is, to a large extent, an adaptation of some components of existing economic and communications theory to the problem at hand. Chapters VI through IX describe the empirical analysis which was restricted to a few of the extension programs in Operation in Colombia at the time the study was being made. From the point of view of locating oppor- tunities for improving extension effectiveness, this restriction has several obvious disadvantages. One of the most serious of these is that it was not possible to make estimates of costs and returns for programs which might have been carried out had the existing programs not been Operated. However, the variety of programs examined tends to compen- sate for this somewhat. It was decided to work with two agencies in different parts of Colombia, and by so doing, achieve an analysis of a rather limited sample of extension activities in the country as a whole. While this sample was by no means a random one, it was the author's opinion that some useful hypotheses and suggestions about the 22 effectiveness of agricultural extension in Colombia (and perhaps other parts of Latin America) could be drawn from it. The Cauca Valley Corporation, from whence the idea for this study originated, offered its co-Operation. The CVC's 1967 extension operations are discussed in detail in Chapters VI and VII. The Department of Agriculture of Antioquia was also approached in this regard and reacted favorably to the idea. Chapters VIII and Ix deal with extension activities carried out by this agency. Because Of data limitations and the methods of Operation employed by the two agencies, it was not possible to conduct these evaluations on exactly the same basis. CVC's extension activities were examined by program; extension activities 30 This makes in Antioquia were examined by district. comparison somewhat more difficult than if the same basis had been used on both agencies. However, it also points out the advantage of having measurement techniques which can be applied to any set of extension activities and the need to further develOp and refine the conceptual structure used in this study. Possibilities for improving extension effect— iveness are examined in detail in Chapter X while Chapter XI presents the summary and conclusions. 30Comparison of extension work by district can serve to answer the question "Where should extension work be done?" However, in this particular case, results in this regard were inconclusive. CHAPTER II THE "OUTPUTS" OF AGRICULTURAL EXTENSION WORK Extension Goals vs. Extension "Outputs" Most agricultural extension agencies suggest, either explicitly or implicitly, that they are trying to achieve one or both of the following goals: (a) improvement in the level Of well—being of farm people, (b) improvement, through gains in agricultural efficiency, in the welfare of the consumers of farm products.1 1The two general goals Of extension work cited here appear to be common throughout the world. For example, in 1966 Colombia's Minister of Agriculture, José Mejia Salazar, stated in his address to the Congreso NaciOnal that Colombia's general extension goal is "to increase the agric- ultural productivity of the country and the welfare of rural people." Another writer, J.N.A. Pender, remarks that "the objectives of rural extension are: to raise agricultural productivity, standard of living of the rural population, and rural welfare." ("Principles of Rural Extension." Methods and Programme Planning in Rural Extension. wagen- ingen, The Netherlands: H. Veeman & Zonen, 1§§6). An Indian eXpert, S.P. Bose, in a paper entitled "Aims and Methods of Agricultural Extension and their Adaption to the Human Factor in Developing Countries" and presented at the Rehovoth Con- ference on Comprehensive Planning Of Agriculture in DevelOp- ing Countries at Rehovoth, Israel in 1963 suggests that "the basic aim of agricultural extension is to bring about improvement in farming through the application Of science and technology." The Smith-Lever Act Of 1914, which created the U.S. Co-operative Extension Service, states that this agency was brought into being "in order to aid in diffusing among the people Of the United States useful and practical infor— mation on subjects relating to agriculture and home economics, and to encourage application of the same." Many extension agencies in developing countries appear to have chosen (or been assigned) their goals to some extent on the basis of U.S. or EurOpean eXperience. 23 24 These goals provide a useful but in some ways inadequate basis for guiding extension activities. These inadequacies arise for the following reasons. First, because "well-being" and "efficiency" are not easily measured, especially at the national level, workable extension Objectives must be phrased in terms of intermediate goals which are sometimes only vaguely believed to be the best instruments for realizing these two ultimate ends. Secondly, extension work can and often does produce benefits which accrue to persons other than farm people and the consumers of farm products. Thirdly, improving the well- tming of farm people and improving the welfare Of urban consumers may turn out to be competing goals and the general nmndate given extension agencies provides no means of resolving the resultant conflict. Finally, the general goals provide no guidance for extension workers who must choose among programs which benefit different groups of farm peOple or consumers. In view of the above three problems it does not appear possible, for the purposes of this study, tO consider the achievement of these goals 225 §2 as the products of extension work. Instead, it appears necessary to attempt tO Specify the different kinds of specific benefits which extension programs can produce, describe how these benefits get produced, and develOp methods for measuring these benefits. If this can be accomplished it should then be possible to compare the total costs of any particular exten- sion program.with its resultant benefits. TO facilitate this 25 sort of analysis, a classification2 of extension functions has been constructed. This is an attempt to specify the sorts of things which extension agencies do as they try to achieve the two general goals cited earlier. These functions are distinguished as separate because, at the outset Of this StUdY. it appeared that a separate analysis of their corres- 4. ponding "outputs" would be a convenient way to break the i overall problem of evaluating benefits into smaller and essentially additive parts. The functions of agricultural extension are considered to be: A. Primary Function - Acting as an Agent to Bring about Beneficial Changes in 1. Agricultural Production Units 2. Rural Households 3. Community Organizations 4. Managerial Structure of Rural People B . Secondary Functions 1. Influencing Technology Developers or Information Sources. 2. Influencing Institutional Activities Outside of Rural Society. 3. Education of Rural Youth in Accepted Practices. 4. Assisting Farm People in Evaluating Non-Farm Opportunities. A . 1though agricultural extension can serve as a 1:001 for r ' c . . o edIStrlbuting income in society as a whole, this has “Ot \ 2The ensuing classification of extension functions on the activities of extension agencies as they e ge ated in the past. These activities — as well as refereneral goals " have been determined by the frames Of e aggce'provided by the legislative bodies which created eXtensincleg' firhus. some Of the functions performed by widely.on agenCJ-es have been omitted because they are . In ess regarded as regulatory rather than extenSion functions. e ence, this Classification is a summary of what appears I generally accepted as agricultural extension education. L is based have Oper 26 been listed as an extension function per E2 because it is essentially one of the outcomes of the functions defined here. The Change Agent Function It appears that the primary function Of agricultural extension is one Of inducing beneficial changes (either immediate or delayed) in the activities of farm peOple.3 This is essentially a restatement of the mandate provided to the Co-Operative Extension Service of the United States by the Smith-Lever Act of 1914. Most extension literature stresses the change agent function Of extension work and the literature in communications and sociology abounds with references to extension services as change agents. Thus, it will be considered here that the primary function of agric- ultural extension is to act as a catalyst or change agent in rural society.4 The alert reader will note that "farm peOple" and "rural society" are not necessarily synonomous terms. 3In some countries, the U.S. being a notable example, the work Of the agricultural extension service is not con- fined to farm peOple but also involves other participants in the agricultural production-consumption chain. Such part- icipants will not be specifically considered here because no such extension activity was encountered in the two agencies examined in this study. However, the discussion in general appears to apply to work done with such participants. Indeed, such work is mentioned as a possible alternative to existing programs in Chapter X. 4For the time being the reader is asked to interpret the word "change" in its most general way. Specific . definitions of changes in rural society are given later in this chapter. an .9okl - sun 1.! 27 However, as both concepts have slightly different meanings in different countries, it is difficult, if not impossible, to distinguish them precisely. This study will be conducted on the assumption that "rural society" and "farm peOple"‘ represent approximately the same set of individuals because this appears to apply in Colombia.5 A somewhat different approach would probably be required if one were investigating extension services to a rural society where this group was significantly non-farm in nature. The activities of rural society are usually carried out in three somewhat distinct sorts of organizations: pucduction units, or farms; consumption units, or households; and various types of community organizations such as school boards, township councils, threshing gangs, and so on. While these three categories are by no means mutually— exclusive or precisely-definable they are sufficiently well- recognized and sufficiently different in nature to warrant examining the effect of extension work on each of them separ- ately. However, it can be noted that this classification arises because Of the different decision-making processes used to COpe with different problems encountered by rural people. Thus, it can be expected to break down when the problems and decision—making processes of any two Of the 5It should be pointed out that rural society in Colombia includes a large prOportion of farm labourers. These peOple are very much a part of rural society and will be considered as "farm peOple" here although they might be excluded in some definitions. 28 three types of units become heavily interdependent. Agricultural extension workers may influence the decisions Of these three types of decision—making units. However, they normally do not have any power to actually make these decisions.6 They usually must rely on the peOple who control these units to make use of the information and advice that they provide. In fact, the change agent role of extension workers fits Berlo's concept Of "communications" so closely that it may be useful to quote a rather succinct and illuminating statement Of his here. "In short we communicate to influence - to affect with intent."7 When extension workers are acting as change agents they are trying to affect or influence rural peOple so that they (the rural peOple) will make changes extension workers want them to make. The problem of objectively evaluating the change agent function Of extension work then becomes one of measuring what changes occur and objectively evaluating these changes. Con— siderable attention will be paid to this problem in Chapter IV of this thesis. Having accepted the change agent function as the rmimary duty of extension agencies, it is now necessary to develOp specific definitions Of changes in each of the three 6This refers only to a capitalistic organization of society in which extension workers are public employees and the final decisions of farm people rest with themselves. If "extension workers" included technical government advisors in a communistic society, their power might be considerably different. 7David K. Berlo, The Progess of Communication, (New York, Rinehart, and Winston, 1933), p. 12. 29 types Of decision—making units which extension agencies may influence: farm firms, rural households, and community organizations. This will be done for individual units in these categories in subsequent sections of this chapter. However, as extension agencies typically attempt to influence (i.e., produce changes in) groups of the three categories of units defined above, the results of extension programs will have to be examined at the group level. Thus, the conceptual analysis which begins with individual units in this chapter'will have to be extended to groups Of such units later in the thesis. Technical Changes in Agricultural Production Units The division of the activities of rural society into production, consumption and community action is somewhat contrived. The artificial nature Of this classification tecomes clearer if one notes that individual members of farm families often belong simultaneously to a consumption unit, a production unit, and one or more community organizations.8 It is therefore immediately Obvious that decisions regarding changes in production will be in some way related to decisions about the activities of the household and the cpmmunity organizations to which the individual belongs. The 8In a sense both farms and households are group organizations as they are typically composed of all Of the nmmbers of the farm family. However, it will be assumed tmre that both farm and household decisions can be treated as if they rested with one manager. 30 static economic theory Of production handles this inter- relationship by assuming that the entrepreneur has a given amount Of resources which he is willing to devote to production and that he then attempts to maximize the returns to these resources. Thus, this theory actually deals with a subset of man's problems on the assumption that it can be usefully isolated from the total problem set. This need not be the case. Experience would suggest that there are many situations where the individual's production problems can- not be readily isolated from the individual's consumption {noblems and his community action relationships. In these situations production theory by itself is not particularly useful. However, the need for production theory is not then eliminated - an amalgamation of production theory with consumption theory and perhaps other theories as well is required. The Static Production Situation For the purposes of this study it is only necessary to summarize the major features of static production theory as it is usually applied to agricultural firms9 and modify these slightly for the purposes at hand. It is usually assumed that the entrepreneur or controller of production for any production unit possesses 9Anyone interested in a more detailed analysis is directed to one or more of the many texts which deal with this subject. See, for example, Charles E. Ferguson' 5 The bhoclassical Theory of Production and Distribution, London emd New York. Cambridge University Press, 1939. 31 (owns)10 limited amounts of one or more resources such as time (labour), capital, land, and skills. These limits can be regarded as being determined initially for any entre- preneur by various rules of society and other such circum- stances beyond his control. The entrepreneur must then decide, subject to those rules of society he elects to abide by, how much of these resources he will devote to the activities of production, consumption, or community activities. It is generally assumed that the decision regard- ing the amounts of resources available for production is made at the beginning of each production period on the basis of expected returns in the various areas. Accordingly, upper limits of owned resources available for production in any production unit in any particular production period are assumed to be fixed. It is also generally assumed that the entrepreneur's objective in any period is to maximize the returns to this fixed bundle of resources. The returns which he realizes over and above that which he could have Obtained by selling or hiring out the resources initially on hand is given by MD 3 (1) NH = pi(t) yi(t) - Z pj(t) xj(t) i=1 j=1 10Ownership is not, in fact, a fixed concept but involves certain rights to use which vary among societies and among resources. This will be ignored here for the sake of simplicity but it is significant to note that the nature cm’ownership may be an important factor in determining resource use and hence the occurence or non-occurence of technical changes in production. 32 where yi(t) are output levels, xj(t) are input levels, pi(t) are product prices and pj(t) are input prices for period t. The pj(t) for Owned resources are their salvagell values but if more of these same resources are bought, the pj(t) of the purchased amounts must be their acquisition prices. All prices are calculated so that no changes, except for changes reflected in R(t), occur in the value of the bundle of owned resources during the period. Because of transportation and marketing charges pj(t) [salvage] is typically less than pj(t) [acquisition]. As all of the yi(t) and xj(t) are assumed to be market goods, this is a fairly straightforward situation because prices are readily available. It can become complicated somewhat by the fact that, under some circum- stances, the prices Of the xj(t) and yi(t) vary with these same variables. These variations are associated with the well-known economic phenomena of monOpOly, OligOpOly, and npnOpsony. However, agricultural firms do not Often find themselves in these circumstances and, unless stated other- wise, it will be assumed in this study that the prices of all market goods are independent Of levels of xj(t) and yi(t). Practical problems also arise in calculating pi(t) 1Ferguson's excellent text on production theory does not distinguish between acquisition prices and salvage values but, as this is accepted practice in analyzing agric- ultural production activities (see "Resource Fixity and Farm Organization" by Clark Edwards, Journal Of Farm Economics, November 1959), this modification of the "neo- classical" theory is incorporated here. 33 and pj(t) for items not bought or sold during the period but, on the whole, the situation is reasonably straight- forward. Some resources (inputs) are not readily quanti- fiable. In such cases pj(t) and xj(t) are not separately apparent but the total value of the input, vj(t), is (e.g., veterinary services) usually known. The limits of R(t) are determined by the levels of owned resources and the technical relation between the xj(t) and the yi(t). This relationship is typically assumed to be known and fixed. It is usually eXpressed as follows in the case of one output. (2) y(t) = f [x1(t),..., xj(t),..., xn(t)] Ferguson describes this production relationship as follows: A production function is a schedule (or table, or mathematical equation) showing the maximum amount Of output that can be produced from any specified set of inputs, given the existing technology, or 'state of the art.‘ In short, the production function is a catalogue of output possibilities.12 Therefore, in his View, all points underneath the production surface are also feasible. Thus, if one accepts this view — which is common in most economic texts - it is necessary to acknowledge that a good deal of technology lies beneath the surface itself. This being the case, it is quite possible for a good deal Of technical change, and hence improvement in R(t), to occur without changing the amounts of xj(t) as 12C.E. Ferguson, Microeconomic Theory (Homewood, Illinois: Richard D. Irwin Inc., 1969), p. 116. 34 they are typically defined.13 For now it can be noted that, in the situation just outlined, the entrepreneur who possesses perfect knowledge of prices and technical conditions will, if he wishes to uaximize profits, act in a very predictable fashion. Firstly, he will discard all points lying underneath the function as obviously inferior to those on it. Secondly, if perfect competition prevails in factor and product markets, he will select levels of inputs so that: (3) 3yi(t) - p (t) = (t) 5 xj(t) i pj for all i and all j except where some xj was owned at the beginning of period t. In this case it is permitted that: 8 yi(t) (4) pj(t) [salvage] s E—ifTET- ' pi(t) 3 s pj(t) [acquisition] 3 Y-(t) Where Pj(t) [salvage] > §_;§TET ° pi(t) then some Of all of the resource j would be sold or rented out until relation- 8 y-(t) ship (4) was achieved. Of course, if §-;%TET ' pi(t) < J pj(t) [acquisition] more Of the resource would be acquired 13Of course, by changing the definition of inputs to distinguish the effects Of timing, sequencing, and mode of application within the production period, one can describe a production surface with no feasible points underneath it. 35 until relationship (4) was achieved. The static production model describes the situation where the entrepreneur possesses perfect knowledge Of prices and technical relationships. However, for the purposes of this study, it is necessary to relax the assumption of perfect knowledge. Only then will it be possible to explicitly define technical changes in production units. Production Under Imperfect Knowledge In practice there are several reasons why the entre- preneur may not select the set of yj(t) and xi(t) which maximize R(t). First, he may be in such a poor state of knowledge about the total possible production relationship that he may choose deliberately, or through ignorance, not to use some inputs or produce certain outputs at all. Secondly, he may believe that certain inputs and outputs are profitable to incorporate in his Operations and he may do so, but he is not yet at the stage where he has found the Optimum total quantity combinations, rates of use, or modes of application. Thirdly, he may not have control over some of the factors affecting output - their control may lie with other persons or‘with nature.14 Finally, price information may be less than perfect and, as a result, price expectations (upon which decisions are presumably based) may not coincide with actual prices. These circumstances create a dynamic situation and l4Such phenomenon are sometimes referred to as technical externalities. In agriculture, a common example is weather, which can be regarded as a group of uncontrollable ”inputs" . 0| I . 36 nake it necessary for the entrepreneur to manage the firm's operation rather than to mechanically select the Optimum cmmbination of inputs and outputs. It is interesting to note that all of these circumstances come under the heading of imperfect knowledge - even the third one. For instance, it would be possible for the producer to maximize profits, given the level of the uncontrollable factors, if he knew what this level were going to be. The first problem area is, of course, covered in the static theory of production by the assumption that the state of the arts is fixed. In reality it is not: it is highly variable — at a price. Many resources have been devoted in this past century to discovering new inputs and outputs, new techniques for combining Old inputs, and different rates of use of combinations Of inputs over time. Indications are that this can be a highly profitable pursuit,15 even though it often involves what appear to be relatively high nominal costs. It should also be pointed out that the state of the arts - as it is called - is not necessarily the same for all producers. As will be seen later, some producers may be using new technology quite effectively while others have not yet discovered it. The second possibility is covered in the theory by the implicit assumption that optimal combinations Of inputs 15Zvi Griliches, "Research Costs and Social Returns: Hybrid Corn and Related Innovations," The Journal of Political Economy, Vol. 66, October, 1958. . .,.. . \ .... 'I n. L... ‘.. .- . . u a 37 (for any given state Of the arts) are known to all producers. This is not necessarily the case in reality. The existence of new producers, new inputs, new outputs, and new production techniques Obviously means that some producers are at times learning about the production surface. Such producers are unlikely to know it sufficiently well to select the set of yi(t) and xj(t) which maximizes R(t) for each value of t. Furthermore, one would intuitively expect the location Of economic Optima to be difficult for farmers to determine due to the complex nature Of the production relationships they face. The fact that producers may not control all of the inputs and outputs in their production units is sometimes overlooked. In crop production, rainfall, insects, diseases, and various other yield-affecting factors are determined not by the producer, but by nature. These "inputs" are not within the producer's control except by introduction Of other inputs such as irrigation, insecticides, herbicides, fung- icides, or others. These uncontrollable inputs make it necessary for farmers to base decisions on output distribut- ions rather than deterministic values. Finally, price uncertainty has always been a problem for producers trying to maximize profit. Agricultural prod- ucers Often have very good estimates of input prices but, when making production decisions, their estimates of the prices they will receive for their products are usually poor 38 because of production lags. However, they typically use these price estimates anyway, although, when they realize that they are poor estimates, they may adopt different decision models and/or criteria than they would have Other- wise. All of these problem areas become more complicated when the phenomenon of multi-period planning is introduced. Buny inputs used in agricultural production processes are not used up entirely in one period and they must be purchased with a view to the returns possible in a number of prod- uction periods. Thus, prices and production surfaces for several periods must be analyzed to make the decision to buy or not to buy. In conclusion, it seems reasonable to classify four types of changes in the set of yi(t) and xj(t) associated with the production process of an individual production unit: (1) movements to previously unused parts of the total production set in an effort to increase R(t) by utilizing inputs which were not previously being acquired by the firm. This essentially means a change from a zero level of use to a positive level of use. (ii) movements to previously unused parts of the total production set in an effort to increase R(t) by using inputs already being acquired by the firm. These changes may involve changes in total quantities, xj(t), changes in use rates, and changes in techniques of application for Old inputs and/or the introduction of new inputs. (iii) technical changes not controlled by the prod- ucer (e.g., random variations in rainfall, insect levels, etc.). 39 (iv) price-induced technical changes made by the producer. These may or may not involve new knowledge of the set Of production possibil- ities, but one would intuitively expect them to involve only changes in the quantities of inputs used per production period, not changes in production techniques or rates Of use within production periods. The first and second of these are of primary concern in this study although the third also affects R(t) and must be con- sidered in conjunction with management and managerial changes. The fourth provides a variation which is manifest in production adjustments but not in a strictly technical sense. Managerial Changes16 It is evident from previous discussion that, if any controlled technical changes are to take place in agricult- ural production units, someone must decide to make them. Thus, it is also evident why it may be profitable for individual producers to devote resources to management - the process whereby they decide whether they should or should not neke technical (and managerial) changes. But the process itself is not quite so clear. The Management Process Various descriptions of the decision-making process have appeared in scientific literature over the years. Most 16This discussion is meant to apply to managerial changes in the three types Of decision-making units disting- uished in this study. It will relate heavily to agricultural production units because the economic theory for this category of units has just been discussed. 40 cm these appear to be very similar in general concept although detailed specification varies considerably. It appears gen- erally agreed, however, that the making Of decisions is part of a much broader function which may be referred to as manage- nent.17 Johnson regards management as a "problem-solving activity of a firm [household] which controls the production [consumption] process."18 In his conception it is a separate controlling unit and not part of the production process per se, even though (as he undoubtedly would acknowledge) the management activity does utilize one of the same resources (the manager's time) as is usually used in the production process. It appears useful to adopt this concept of manage- ment here, and, in addition, to extend it to the analysis Of changes in community organizations. Johnson has described 19 the following five managerial functions: (i) Observation: This term is used "in its broadest possible sense, including the 17It is perhaps necessary at this point to dis- tinguish between management and skill. Skill may be regarded as the ability to perform some repetitive task, albeit (at times) a highly complicated one. Many Of the functions per- fOrmed by tradesmen and professional people fall into this area. Management, on the other hand, refers to decision- ueking (which is usually of a non-repetitive nature). Pro- fesional people Often find themselves performing managerial functions as well as employing skills. Both the performance of skilled tasks and managerial functions utilize the same resource, human time. 18Glenn L. Johnson, "Methodology for the Managerial Concept," The Management Input in Agriculture, a study Sponsored by the Agricultural Policy Institute, Southern Farm Management Research Committee, Farm Foundation, April 1963, p. 9. 191bid., pp. 9-10. 41 perceiving and recording Of both normative and positive experiences generated by controlled and uncontrolled events." (ii) Analysis: This function determines the meaning Of information acquired by observation. "Concepts derived by ded- uction are checked against each other for internal consistency. Concepts derived inductively from Observation are also checked against each other and against those derived from deduction for logical consist- ency. The result is the conversion Of uninterpreted 'analytic' systems into inter- preted or 'synthetic' systems." (iii) Making Decisions: "It seems that the crucial aspect of a decision to solve a problem is that the decision-maker arrive at an appropriate compromise among the relevant concepts of goodness and badness er se in view of positivistic concepts of Shit-Has been, what is, what will be, if various actions are decided upon and carried out. The decision- maker following some decision-making process within some decision-making structure and following some basis for choice (such as the minimax, maximize average expected net return, 'satisficing,‘ etc.), decides upon the best action to be taken." (iv) Execution: "Decisions are not really final until steps are taken to place them into effect." The distinction between execution of decisions and the previous three functions is somewhat analogous to the military distinction between tactical and strategic decisions. A number of tactical maneuvers must occur to implement a strategy. (v) Bearing Responsibility: As the responsibilities which the decision-maker bears affect the specifications for his choice, the responsib- ility-bearing function is an essential part of the managerial process. Recent thinking has expanded this list to six functions yinciluding "problem definition". This logically preceeds th . e‘Nmer five functions in any problem situation. It can be [L' ‘H.. n‘_ . ‘7 . 'e. a ..‘ n 42 mmmithat, like the previous classification Of human amthuties, this classification is somewhat contrived. It nevertheless appears to be a modestly useful structure upon which to rest this discussion. It may be useful to categorize decision situations inmatwo groups; those where the decision-maker believes he has perfect knowledge of the outcomes Of the various possible alumnatives, and those where the decision-maker recognizes smmzuncertainty with respect to some outcomes. In the first case, the decision-maker's observations and analysis haweled him to his beliefs and now he must merely choose the alternative with the preferred outcome. He must utilize his concepts Of "goodness" and "badness" - to paraphrase Jcmnson's terminology - to determine the preferred outcome. GiVenthese concepts, his decision is always right for him,20 “flessrus original beliefs about the various outcomes were fluse Une., unless his Observations and analysis were in error) or he has applied these concepts incorrectly. Decision situations in the second group are more CO“‘Plicated, and probably more common into the bargain. Herevtzhe decision-maker is not certain what will happen as aremflt of some or all Of the alternative actions open to h' . . . . . lm- His observations may be inconsistent With one another 0 r, f<‘iCtors beyond his control may be set at levels unknown \ 20This assumes that the individual's concepts Of . and badness reflect what is actually good and bad y a 1m. It also assumes that he is the only one affected 0n OW outcome and thus the effects Of alternative actions ers do not complicate the analysis. QOodn f0: 988 43 to him when he has to decide. Thus, unless all possible outcomes of one alternative are preferred to all possible outcomes of all other alternatives, he can make a wrong decision. If he does so, he suffers a loss in welfare in that he is worse off than he would have been if he had known the eventual state of nature at the time Of the decision and acted accordingly. To adequately describe such decision situations, stochastic or probabilistic models are necessary. Indeed, fairly complicated models are required to describe many 0f the day-to-day decisions people make.21 Many Of these are naive in that the alternative of delaying the decision is not considered to be relevant. Once such an alternative is introduced, the problem can only be analyzed with some sort of sequential decision-making model.22 This might incor- Porate searching for additional information or experiment- ation as new alternatives. Furthermore, in the previous dismission only a few possible criteria were mentioned. ActUally, many criteria are possible. In addition, managers' col"cePts of goodness and badness may lead them to attach \ 21For a discussion of such models the interested 1‘ might refer to Chapter 9 of John E. Freund, Mathemat- Statistics, (Englewood Cliffs, N.J.: Prentice-Hall Inc., York.' or Abraham Wald, Statistical Decision Functions, (New RObe' John Wiley 5. Sons Inc. , 1950T, or Howard Rai'ffa and (C rt.Schlaifer, Applied Statistical DeciSion Theory, ambridge, Massachusetts: The M.I.T. Press, 1961). “‘Ode 22An elementary discussion of sequential decision b11118 can be found in Chapter 15 of A.M. Mood & F.A. Gray- CGr' ingroduction to the Theory of Statistics (New York: aw‘Hill Book CO. , 1963) . [Eade 1031 l.- 5.; 44 different values per unit to gains or losses depending on 23 In such cases, the the size of these gains or losses. relevant valuations Of outcomes would not be given by net monetary values. Finally, there are undoubtedly a large number of problem situations that man faces which are repetitive and can be resolved by habit once an initial decision has been made. For example, most of us do not go through a complicated decision process in order to determine Whether or not to go to work every day - although there are occasions on which we may. Generally, the decision to go to work or not is an automatic one based on the original decision to take the job. Only in the case of unusual circum- Stances (e.g., a broken leg) do we break the habit. Habits can be both useful and dangerous. They are useful in that they reduce decision-making costs but they are dangerous in that , by deliberately avoiding the decision-making process, W . e may overlook advantageous alternatives. MarlaSerial Changes Defined Managerial (or cognitive) changes can be defined as Changes in the managerial process. Thus, they include new Obs"ervations, new Observation-making skills, changes in analytical procedures, changes in values, and changes in beliefs about non-normative matters. Indeed, a change in the way any one Of the six steps described earlier is \_ ( . 23Milton Friedman, Price Theory: A Provisional Text, Chlcago, Illinois: Aldine Publishing CO., 1962, pp. 68-73.) 45 performed may be regarded as a managerial change. Although this definition of managerial change is far from precise, it is sufficient to enable us to note that there are many things which can be included in this general category. Perhaps the most controversial of these are changes in values. Concepts Of goodness and badness p_e_r_ s_e_ appear to be used explicitly or implicitly in all decisions. It is known that these vary from individual to individual and from society to society. It is also known that this variation is Often a source of conflict among individuals and societies. For example, Brewster formulates the following general prOb 1 Em ; Why are people in economically backward societies so reluctant to abandon age-old techniques and latch onto the gadgets Of progress, such as higher- Ylelding seeds, more productive animal breeds, c><>ntour farming, steel-beam ploughs, pesticides, Chemical fertilizers, sanitary wells and the like? 24 while Brewster appears to overlook the possibility that such techrliques may not be profitable at the moment for many such armtars, he certainly appears to have a valid point. People in economically backward societies sometimes do seem rel“Ctant to change even when the techniques appear to be ptofitable and the people themselves are clearly aware of their existence. Sociologists, anthropologists, economists, \ 24John M. Brewster, "Traditional Social Structre as arriers to Change," in Agricultural Development and Economic 7%, edited by H.M. Southworth and Bruce F. Johnson. aca, New York: Cornell University Press, 1966) . 46 amicothers have advanced various explanations for this. These include lack Of technological creativity, lack Of achievement motivation, a strong distaste for work, lack Of an eXperimental frame of mind, the high subjective risk level associated with new techniques where the Old ones Provide a sure but low level of living, and the lack of recognition Of needs or wants. There is undoubtedly some element Of truth in all of these explanations. And, implicit in most Of them is the theory that there are value differ- ences between peOple in economically backward and economically advanced societies. Given this situation, the Cmeet-1101') arises as to whether extension workers can change values and, if they can, are they justified in doing so? u°e'v are they really improving managerial capabilities if they do?). It appears that some concepts of goodness and badness are connected with man's basic drives such as hunger, Se 0 o O I x, Self-preservation, rest, diverSion, and recognition 25 Satisfaction of such drives are generally in Society. regaI‘ded as good Egg s_e_ although in circumstances where such driVes are competitive, one or more may fall into disrepute relative to others. It appears that other concepts of good- n ess and badness have been developed from the group 0f c oncePts connected with man's basic drives. They appear to \ ex 25This is not meant to be a complete or mutually- cluSive classification of basic drives. 47 be analogous to habits which originated from some previous decision on the part of the individual. For example, as bananas satisfy man's hunger, and are viewed by most individuals as tasty, they may conveniently be regarded as <300d in decision procedures directed at choosing the comp- osition Of the family food basket. Such derived concepts of 900dness and badness may be wrong because of faulty reasoning. For example, if one eats bananas and becomes sick, he may conclude that the bananas made him sick (even though they did not) and therefore, bananas become regarded as bad p35 .53: Such concepts may also be wrong because, over time, new considerations become relevant and, even though the original conclusion was correct, it no longer holds. For example, certain foods may be regarded as bad p_e_r_ s_e because at one time they spoiled very quickly and were a source of illness. HoweVer, such an oversimplification would no longer be valid if improved methods of preservation were available to prevent rapid spoilage. Instead, it would be necessary to retreat to mcDre basic concepts Of goodness and badness to determine whether or not it would be right or wrong to actually eat the food in question. It appears that many value differences among individuals and societies may be found in derived, ratl'ler than basic values and can be attributed to faulty reasOning and the failure to re-evaluate derived concepts Of 90odness and badness under changed circumstances. This being the Case, it would appear that extension workers can and 48 should influence the derived values Of rural peOple whenever these values are in error. This conclusion appears to be in conflict with the assumption that the informed individual is the best judge Of his own welfare, an assumption which has long been popular 26 However, if an individual is making among economists. decisions on the basis of derived value concepts which have been arrived at through reasoning from basic normative concepts, his derived concepts of goodness and badness may not reflect what is actually good or bad for him. The reasoning itself may be in error. Where such a situation exists, it is possible for someone else, given superior knowledge of reality and information about the individual's basic normative concepts, to be a better judge of the individual's welfare. Unfortunately, problems arise because of the difficulty in distinguishing between basic and derived values. Nevertheless, it appears that, given basic values, derived value concepts can be acquired and changed through logical reasoning. Where such reasoning has been in error it is possible to change such concepts for the better in the sense that when the faulty reasoning is corrected, the new derived concepts will better reflect the individual's basic value concepts. In this sense, there is no conflict with the assumption that an informed individual is the best 26As the consumer's preference function is regarded as being determined by the consumer himself (see Ferguson, p. 14) and the consumer's welfare is generally regarded as being directly related to his achieved level of utility (see Ferguson, Chapter 16) the existence Of this assumption (which is often implicit) is readily apparent. 49 judge of his own welfare as the individual may lack infor- mation on either normative or positive matters and hence err in his judgement. Changes in Rural Consumption Units (Households) The economic theory Of consumer behaviour is well- known. It is assumed that each household attempts to maximize satisfaction or utility by allocating its money income among a known selection of goods and services. In addition the following assumptions are made: (a) Each consumer [household] has exact and full knowledge of all information relevant to his consumption decisions - knowledge of the goods and services available, and of their technical capacity to satisfy his wants, Of market prices, and of his money income. (b) Each consumer [household] has a preference function that (1) establishes a rank ordering among all budgets; (ii) for pairwise comparisons, indicates that A is preferred to B, B preferred to A, or that they are indifferent; (iii) for three—or—more-way comparisons, indicates that if A is preferred to (indifferent) to B and B is preferred (indifferent) to C, A must be preferred (indifferent) to C; (iv) states that a greater budget is always preferred to a smaller one. Under these assumptions utility (or satisfaction) is regarded as at least an ordinal function Of the amounts of goods and services consumed. It can be eXpressed mathematic- ally as: (S) U = h(x1,...,xn) ‘— 27Charles E. Ferguson, Microeconomic Theory (HomewOOd, Ulinois: Richard D. Irwin, Inc., 1969), p. 14. 50 where U is a measure Of utility and x1,...,xn represent quantities of goods or services. (If one so desires it is possible to regard all sources of satisfaction as being included in the set of xi, i = l,...,n. However, the analysis of non-market and/or abstract goods and services in this framework is often not particularly enlightening.)28 It can be readily shown that the consumer maximizes29 satis- faction or utility subject to a fixed income if he (or she) selects goods and services so that: 3U 3U _ ._ . . (6) 5;; / EE— — Pxi / ij i, j—l,...,n 311 j and P denotes price. Thus, for a fixed income and a given utility surface a consumer's satisfaction cannot be increased as long as condition (6) holds. 28It will be noted that, because Of its formulation, consumption theory, like the production theory described earlier, says nothing about the production of non—market goods and services, a function in which the typical household is usually heavily involved. Because of this fact, it is necessary to include these activities implicitly in this discussion. Thus, some of the technical changes in consump- tion defined later actually belong to this area and not to the classic consumption model per se. 29Actually, the condition shown applies only to cardinal utility functions as ordinal functions do not possess derivatives. However, if g is ordinal it can be regarded as a set of cardinal functions which are monotonic transformations of one another. If these are denoted h = {hk} where k = l,...,m utility will be maximized if: 3hk / 3x1 3h axj = PXi / ij 1, 3:1,...'n 3:1 (of course, m need not be finite.) 51 Corresponding to the technical changes in production units discussed earlier it is also possible to distinguish technical changes in consumption units. One crucial difference is that outputs are not measured physically and output prices are not found in the market place. Instead, utility serves as the common denominator of man's values. Technical changes in consumption units can then be classified as follows: (a) changes in the amount of utility derived from a specific set of acquired goods and services because of a change in the way they are consumed. (b) price-induced changes in the relative amounts of different goods and services consumed. (c) changes in the amounts of different goods and services consumed due to random, uncontrollable influences on the household. (d) movements along indifference curves towards (or away from) Optimal points in an effort to correct previous errors. (e) changes in the amounts of different goods and services consumed because Of the introduction of new goods or services. The first of the above technical changes is analogous to a change in the technical process of production described for production units. For example, if a housewife discovers a new way of preparing food which eliminates waste, she is 30 incurring a change of this type. A change of this nature enables her to get more utility out Of a given income if it k 30This is actually production of a non-market good huh within the framework used here, it is classified as Improved consumption efficiency. 52 can be achieved with no additional eXpense. Such changes might be adopted by a person at any income level. Price-induced changes and random, uncontrollable changes in consumption are not of major importance in this study. However, extension workers can probably provide a useful service by showing rural consumers how to adjust to {mice changes - this may require a movement to a previously- Lmknown part of the utility surface for some - and illustrating how to handle random influences on consumption (e.g., insurance against medical expenses). Movements along indifference curves in an effort to correct previous errors can be eXpected where recent adOption or price changes have occurred and consumers have not yet managed to find the Optimal point through experiment- ation. These mistakes could be expected to arise for complicated (difficult—to-understand) consumption goods. For example, it might require some time for a household to determine whether or not a new good was actually providing the expected satisfaction (e.g., vitamin pills might have to be taken for several weeks before their effects were noticed.) It can be readily seen that most of the managerial changes discussed earlier apply directly to consumption activities. It is in the consumption unit or household that development of an apprOpriate set of derived values becomes crucial. In production of marketable commodities only a few of such values are necessary because profit reduces so mnw'considerations to a single common denominator. In the 53 consumption unit every good and service which might be gmrchased has tO be analyzed by the consumer in order for rum to formulate his utility function. This problem becomes particularly complicated in a highly-develOped economy where nuny consumer goods are available. It may be relatively simple in an economy where few goods and services are sold Unconsumers. However, in the latter case, mistakes may be nmch more costly to the individual because the very nature cm such an economy dictates that most consumption is directed an satisfying basic physical wants. Thus, while changes in mwrpart of the management process may produce significant cmanges in consumption efficiency, knowledge of the values cw various consumption goods available and within the budget of the consumer appears to be of key importance. Changes in Community Organizations Although academicians have long been interested in the economics of collective human activity, there is no widely-accepted theory Of choice in this area comparable tO that which exists for either individual consumption or prod— uction activities. This is not surprising when one reflects upon the widely-varied forms Of government which have developed throughout history as mankind has attempted to find a practical solution to this same problem. Actually, both consumption and production theory are extremely relevant to the economics of community action. Their main drawbacks lie tithe difficulties of arriving at social welfare functions 54 and in creating functional markets for public goods. While it is widely-acknowledged that men banded together may Often achieve their ends more efficiently than if they act as individuals, the problem Of exactly when they do (or should) join together in common cause has confounded the best minds to tackle it. One writer states the problem as follows: When will a society composed of free and rational... individuals choose to undertake action collectively rather than privately? .... On what grounds can the individual decide that a particular activity belongs to the realm of social as Opposed to private choices? Another describes the problem as: What criteria can be employed tO judge the Optimal size and sc0pe of the governmental sector? As shall presently be seen it is those very characteristics of activities which causes them to be placed in the realm of public responsibility that also causes them to be difficult to evaluate. Thus, both these writers are essentially voicing the same question. While at first glance the rather weighty issues in which this discussion has become involved may seem to be unrelated to the tOpic of this study, it must be noted that 31James Buchanan and G. Tulloch, The Calculus of Consent, Chapter 6. 32John F. Due, Government Finance: Economics of the Egblic Sector, (Homewood,lllinois: Richard D. Irwin, Inc., 1968’, p. 4. 55 public activities involve basically the same conceptual problems of choice whether they are local, regional, national, or international in nature. Agricultural extension is unlikely to influence public bodies in the latter two areas - although U.S. extension services have had "policy" pmograms on national issues. However, extension workers can and do influence local and regional bodies both directly and, indirectly, through the individuals (who through election or cmherwise) control these bodies. Some noted writers in this area differentiate between agricultural extension and community develOpment. 33 This is not of importance here. The important thing is that extension can and does influence changes beyond the individual farm or rural household. It is this influence on public decisions which is deserving Of attention in this study. But, as no comprehensive theory is available to serve as a frame of reference, only a few isolated points regarding the nature of changes in this area can be made. Firstly, it can be noted that, in the absence of externalities, markets which serve to equate private marginal costs and marginal revenues could eliminate the necessity Of public activities. Of course, externalities do exist and market imperfections, much more complicated than the textbook cases of monOpOly and OligOpOly, do inhabit the real world. 33Arthur T. Mosher, Varieties of Extension Education and Community DevelOpment, Comparative Extension Publication No. 2 (Ithaca, New York: New York State Agricultural College at Cornell University, 1958) . 56 Market rules must be established. Non-market goods demand much more of our attention and income distributions can never likely be fixed to people's satisfaction. As man's environment changes it would seem that there would be ample opportunity for new, justifiable public activities to arise and old ones to disappear. Such changes could be influenced advantageously by extension workers. For example, if new technology made irrigation economically feasible in an area, extension workers might point out the need for legislation to ration water so that disputes might be avoided. They nught even work with local people to get an acceptable set of by-laws drawn up and passed. Secondly, it must be pointed out that many community activities just will not function on a voluntary basis. One of the problems here is what is known as the "free rider" effect. If an individual stands to benefit from a program (such as fire protection) whether he contributes or not, he will have no incentive to contribute. Other problems - such as individuals refusing to give up their prOperty at market prices for public purposes - also arise. In any event, some means of making non—Pareto decisions is almost essential for any significant amount of public activity to occur. Changing circumstances may mean that changes in the nature of these decisions is appropriate. Again, agricultural extension may perform a useful function by pointing this out to farmers and assisting them.in 57 resolving specific problems. The fact that extension agencies have representatives in many communities may mean that they may be in a position to pass the lessons learned in one area on to many others. Although the distinction is not as clear as with production units, it is possible to classify the activities cm community organizations into both production and manage— nent. That is, associated with all public action projects there are both "decision-to-act" and "action" functions. The management process here is indeed a complex one because 1mm.only must a decision be reached by many individuals, a cmmmunity decision rule must also be decided upon. Majority rule, concensus, dictatorship and others are all possible alternatives, none of which have any absolute claim to being correct. And, even when public decisions are reached, the administration of public projects is complicated by the nature of the responsibility-bearing function. Given all the problems that exist in the area of 34 public action, it is not surprising to find that a 34An excellent practical example Of the public decision-making process at the local level in Colombia was the Co-Operative Electrification Project in the municipal- ities of Sevilla and Caicedonia in the Cauca Valley. This has been well documented by James Ross in his Ph.D. dissert- ation entitled "Implications of Co—operative Rural Elect- rification for Economic and Social DevelOpment in the Department of Valle, Colombia, South America" completed in the Department of Agricultural Economics, University of Illinois, 1965. His study illustrates some of the diffic- ulties encountered in making non-Pareto better decisions, the importance of environmental factors in determining the success of public projects, the necessity of public action in such an instance, the interrelationships among innovations, and many other of the phenomena discussed here. 58 tremendous potential exists for useful extension work. How- ever, as will be seen later, the fact that possibilities for useful work exist does not mean this work can be carried out economically or that it will even be recognized as useful. The talents required to do this work and the costs Of doing it.may differ substantially from work done with agricultural {Reduction units or rural households. Secondary Extension Functions Although the principal function of extension agencies is that of a change agent, such agencies usually perform other functions which are, in general, complementary to the change agent function and are Often regarded as essential features of extension work. These other functions produce what will be referred to here, for convenience sake, as "secondary outputs." As noted earlier in this chapter, these include (a) influencing technology developers or information sources outside Of rural society, (b) influencing institut- ional activities which have a bearing on, but are not part of, rural society's activities, (c) providing vocational education for rural youth or new farmers in practices already adOpted in their areas, and (d) assisting in the adjustment of rural peOple to occupations other than farming.35 35Other secondary extension functions are also possible although these four seem to be the main ones. The others will be briefly mentioned in the sections on aggreg- ation problems in Chapter IV because they appear to come under the heading of externalities. 59 The fact that agricultural extension can exert a useful influence on technology developers and information sources is widely-acknowledged. Most extension agencies usually attempt to develOp a close working relationship with experiment stations and other information sources. This permits them to obtain the most recent research results as soon as they are available. The flow Of information which CkveIOps from these contacts is usually not a one-way affair. Research agencies often learn useful things from extension workers and, as a result, research can become more problem- oriented and more relevant to the potential users. In addition to influencing technology developers and information sources, agricultural extension services can influence other institutions as well. They appear to have influenced university curriculum significantly in the United States. Agricultural policy-makers often find them a useful source of information about the current farm situation. Credit and marketing agencies may find that extension personnel are useful sources Of information for planning pur— poses. Regional develOpment agencies and regional govern- ments may use information gleaned from interaction with extension workers. This function of agricultural extension contributes indirectly to both of the general extension goals cited earlier and perhaps some others as well. Agricultural extension can function as a training agency to bring young peOple and farmers new to an area up 60 to the technological and managerial levels already achieved by existing decision-makers. This is cited as a different function from introducing changes to rural society because it does not involve introducing changes to existing units - it involves facilitating the natural replacement cm the decision—makers of these units. Agricultural extension has undoubtedly performed this function in North America. Farm young people in both the United States and Canada have often learned things in 4-H Clubs which their fathers and mothers already knew. Indeed, local farmers and farm wives are Often used as 4-H Club leaders and, in these positions, they Often rely heavily on practical experience. (This sort of extension activity might be classified as either "influence on a community organization" or a perfor- mance of a "secondary extension function.") The function of assisting rural peOple to evaluate jobs outside of the agricultural sector is particularly relevant when the agricultural sector becomes of declining proportionate importance in society's economic structure. This phenomenon is almost always associated with economic progress. Agricultural extension, by improving the manager- ial ability and technical skills Of rural young peOple - and perhaps some Of the Older folk as well - may significantly improve their ability to take up new positions outside of agriculture, although not necessarily outside Of the local area. Perhaps the most important contributions in this area 61 have been made by 4-H leadership programs and farm manage- ment projects. However, it should be noted that extension workers sometimes emphasize the advantages of agriculture so heavily that they overlook the fact that benefits may be realized by assisting some peOple to transfer to other endeavours. Thus, this function of agricultural extension nay not be acknowlepged by extension agencies even when it is actually being performed (either consciously or other- wise). As the benefits associated with these four functions are quite difficult to analyze and the two agencies studied in this research project were not heavily involved in these activities, relatively little consideration will be given tO them throughout this study. However, they have been identified because they can, under certain conditions, be rather useful functions and thus they do represent possible avenues for useful extension work in both developed and developing countries. CHAPTER III THE PROCESS WHEREBY EXTENSION "OUTPUTS" ARE ACHIEVED Now that the "outputs" of agricultural work have been defined, it is necessary to examine the process whereby these "outputs" are produced. This will then permit relating specific sets of extension activities to their associated "outputs", a step which is essential if an Operational method 1hr measuring extension efficiency is to be devised. Included in this discussion is a general description of those factors other than extension activities themselves which influence the results of extension programs. Recognition of these will permit later analysis of some of the managerial pmoblems facing extension agencies and an understanding of environmental limitations on extension effectiveness. As agricultural extension work produces "outputs" in only a very general sense, it is not useful to apply the standard economic model of production to the analysis of extension work. In order to formulate a conceptual model which may Offer Operational possibilities for analyzing extension activities, it is necessary to turn to a more general and comprehensive technique. The technique chosen for this purpose is systems analysis. 62 63 The Systems Approach Systems analysis has recently become a popular pmrase in North America although its meaning is far from pmecise. TO persons working in engineering and the physical sciences, it refers to a group Of mathematical techniques vmich have been developed for coping with the quantitative models that are necessary for designing and controlling some cfi’the complicated real-world phenomena now being dealt with in these disciplines. A more general interpretation of systems analysis is used by persons working in government and business. To them systems analysis is usually regarded as a comprehensive approach to problem-solving wherein all significant relevant factors are considered in choosing among a group Of proposed actions. These two interpretations Of the technique are, of course, related. The latter can (and sometimes does) become reduced to a mathematical model if sufficient information is available to write the necessary equations. It is this latter interpretation of the systems approach which applies here because the phenomenon being studied is not sufficiently well-understood to permit modelling it mathematically. The typical systems model is often portrayed graph— ically in the manner shown in Figure 1. Essential charact— eristics are input, output, environment, and the endogenous components of the system itself. Inputs are those controll- able factors which influence the system (and hence the system 64 output) but are generally not influenced by the system. [mere the system influences the inputs, either through system output or through some intermediate variables, this pmenomenon is known as feedback and requires special analyt- ical techniques. Environment is also independent Of the system but can affect output; it is distinguished from input because it is exogenous and not controllable. In Figure l a dotted line shows the influence the system can exert on its environment. For analytical purposes it is Environment 1 Endogenous Inputs Components of ::Outputs the System Figure l.--A graphical illustration of a system model. desirable to define the boundaries of the system model so that this particular feedback relationship is weak or non- existent. Outputs are the relevant products or results of the system. They need not be desirable entities but usually many Of them are. For example, one Of the by-products Of many production systems is waste material which has to be disposed of at a cost. The endogenous components of any system are those internal relationships which exist between input. output, and environmental factors. In a mathematical 65 formulation they appear as equations. Variables whose values sue determined simultaneously within the system are referred to as endogenous variables. Typically, output variables will In endogenous while input and environmental variables will he exogenous. However, systems Often include several inter- nediate endogenous variables which are not of interest as final outputs. In order to approach agricultural extension in a systems framework, it is necessary to regard certain aspects ct extension work as components of the general system model. All extension activities will therefore be regarded as system inputs. The nature of these will be discussed in the next section. The "outputs" of the system were defined in Chapter II. These now have to be described in a way in which they can be associated with a particular set Of exten- sion activities (inputs) applied to the system. Environment can be divided into several component parts for easier analysis. These will be discussed in detail later in the chapter. Endogenous components Of the extension system model would describe how extension activity influences extension outputs under given environmental conditions. Extension Activities as System Inputs Extension agencies are not unlike the three types of decision-making units discussed in Chapter II in that their activities can be divided into two groups, technical and managerial. The former group may be usefully referred to as 66 "program Operation" activities for they include the actual performance Of functions which directly influence the dkcision-making units in rural society. It is the nunagerial activities which determine the nature Of the pmogram Operation activities. This relationship is shown in Figure 2. AGRICULTURAL EXTENSION a -------- Management ‘”w- Environment Idea """n’ mp Sources “‘~- Decisions ~- RURAL SOCIETY Program * Operation Figure 2.--The extension agency component Of an extension system model. Extension Management Extension management activities involve: (a) detecting problem areas in local communities. (b) assimilating information on (i) new ideas whicE might be useful to members of rural society, (ii) performance of current and past programs, and (iii) new and existing extension methods. (c) analyzing and interpreting this information as it applies to local conditions, then using it to plan (design) extension programs. The sources of these ideas include agricultural experiment stations, farms, private companies. foreign aSencies, etc. 67 (d) evaluating program prOposals and selecting promising ones. (e) administering the selected program prOposals (this includes the tactical or day-to-day decisions necessary for administration). (f) accepting responsibility for the program results. While changes in rural society are produced only through "program Operation" activities, some secondary extension outputs arise as by-products of extension managerial activ- ities. Three of the dotted lines in Figure 2 indicate the messages or Observations which influence extension managers. The feedback effect from extension to idea sources is shown by a dotted line in the Opposite direction to that which illustrates extension observations. Because of the nature of extension work it is impossible for public representatives to give extension workers complete mandates.2 Extension agencies Often have a wide range of programs which they can operate as well as a variety of ways in which each of these programs can be carried out. In order to make intelligent choices among these alternatives, they must devote a good deal of time and effort to planning and selecting programs. Because of the limited understanding of social relationships this general problem is usually reduced to one of designing a promising, feasible set of discrete alternatives. Furthermore, it is 2The fact that extension decision-making is not vested in a single person or even at a specific level in the Public decision-making chain makes the responsibility- bearing component Of management difficult to achieve. 68 usually not desirable (because of the time and costs involved) for an organization to evaluate a very large number Of programs. In doing this, it is usually necessary to use estimates of costs and/or benefits which have size- able possible errors and may therefore lead to incorrect (moices. It is also possible that highly effective programs may be overlooked. In an effort to keep managerial costs low and avoid overlooking highly effective programs, program planning often is left to field personnel who are familiar with the needs of rural society and the environment in which they and rural peOple must work. Extension Program Operation Extension program Operation may be usefully discussed under the headings of extension personnel, extension methods, and supporting services. The analysis of extension personnel is complicated by the fact that, as mentioned earlier, the same peOple who plan and choose programs Often Operate these programs in the field. Perhaps the two most significant characteristics of extension personnel for program Operation purposes are communications ability and technical competence.3 How one measures either of these characteristics is not immediately 3This assumes that extension workers are promoting changes which are advantageious to their clients. The dis- cussion of Chapter II and III indicates that, the deter- mination Of net advantage is not a simple matter and may require talents different from those embodied in communic- ations ability and technical competance. This, however, is a managerial function - not a program Operation function. 69 evident. Neither is it obvious as to how one acquires them. 1b be sure, formal training programs help and most extension agencies rely heavily on these both before and after employ- nent begins. However, experience can also produce similar results. One of the interesting features Of North American and European extension work has been the employment of well- trained extension personnel with backgrounds quite similar to those of extension clients. Thus, it is not clear to what extent their effectiveness has been determined by their advanced training or by their intuitive knowledge gained outside of their formal training program. Nevertheless, one would expect that different sorts of capabilities of extension personnel would be apprOpriate for different char- acteristics of program participants (e.g., age, education, values, etc.) and different types of programs.4 One alter- native that is often Open to extension agencies which are trying to get the most out of a fixed budget is to hire personnel with lower qualifications and thus minimize costs per day of extension work. This can be done as long as the personnel with lower qualifications produce the same results. However, one would not expect this to be the case in general and some assessment of the extra productivity balanced 4For some changes in rural society, the requirements of extension personnel may be well-established; for others this may not be the case. For example, teaching farmers how to cultivate new varieties Of rice is a very specific task with particular, established requirements. Getting farmers to 11“Prove their managerial process is not. Little is really known about the managerial process let alone about how to change it effectively; 70 against the added cost must be made. A variety of extension methods are available. These include group meetings and discussions, tours, individual farm visits, result demonstrations, method demonstrations, movies, radio and television programs, pamphlets, contests, and others. TO date no large body of information on the specific effects of these variables under specific conditions is available although many isolated studies have been carried out in this area. Unfortunately, a review Of these would lead one to believe that the appropriate method to use depends on the audience to be reached, the message to be carried, and some environmental conditions as well. Thus, extension workers will Often have to choose not only the message but the communication vehicle as well. Supporting services for extension programs include such items as transportation, materials and supplies, communications, subject matter Specialists, and so on. These are Often highly complementary to personnel time and SPecific extension methods. Thus, some programs may be infeasible without them. Representation of Extension Outputs Changes in Groups Of Similar Decision-Making Units Extension programs are typically directed at groups of similar decision-making units on the assumption that if a change is appropriate for one unit, it should be approp- rlute for others which are similar in nature. Thus, it would ‘V 71 be convenient to have some device for representing the effect of an extension program on a group Of such units.5 For some changes in rural society recent research in the area of the adoption of innovations provides at least a partial answer. This answer applies to those changes which spread through a given social system or geographical area over time and permit construction of a time distribution of adOpters.6 For any given social system7 it has been shown on numerous occasions that the number of adopters plotted against time produces a bell-shaped curve which Often approximates a normal density function.8 On the basis Of this observation, Rogers9 has classified adOpters as innov- ators, early adopters, early majority, late majority and laggards as is shown in Figure 3. 5The reasons for this will be more clearly seen in Chapter IV when aggregation problems are discussed. 61n some areas extension work may concentrate heavily on changes which cannot be usefully represented by adoption distributions. Movements towards economic Optima which differ from farm to farm or from year to year on the same farm are examples. 7A social system is defined by Rogers as a "pOpul- ation of individuals who are functionally differentiated and engaged in collective problem-solving behaviour". However, this definition is neither precise nor widely-accepted. 8Gwyn E. Jones, "The AdOption and Diffusion of Agric- ultural Practices," World Agriculgural Economics and Rural r Sociology Abstracts, Vol. 9, NO. 3, September 1967} 9E. Rogers, Diffusion of Innovations. (New York: The Free Press of Glencoe, 1962), p. 162: 72 I I o. Of I ndividuals I opting | | l I I I I ' I I I ' u I : i 34% I 34% I 16% 2.5% l Early ' Late I Laggards nnovato : Adopters : Majority : Majority : X - 20 X — o X f + 0 Time of AdOption Figure 3.--Classification Of adopters on the basis of a normal density function. It is, Of course, an easy task to convert the normal density function of Figure 3 into a normal distrib- 10 ution function. Having done this, one can readily see that the likely effect of successful extension work would be to shift such a distribution upwards and to the left —— from Ao(t) to AE(t) ——-as is shown in Figure 4. Of course, in practice it will only be possible to Observe one of these distributions in any given situation, and, as a result, the 10As a normal distribution can be completely specified if its mean and variance are known, it would be Quite legitimate to regard extension work as changing the mean adOption time and perhaps its variance as well, if the normality assumption holds. 73 detection Of the precise effect of extension work is not an easy Number of Un Adop To task. its ting Date Time Figure 4.--Illustration of the influence Of extension It i efforts on the adoption of a hypothetical innovation. 3 important to note that there are four characteristics of the adOption distribution which extension may influence. These include: (1) starting date (2) average rate of adOption (3) shape of the distribution (4) total number eventually adOpting11 its 11The normal distribution has one flaw in terms of ability to approximate observed adOptions over time. Observed adoptions begin and end - the normal curve does not. Thus, a truncated normal distribution would perhaps be a better model. The "number eventually adopting" should be interpreted as if a truncated normal mode were used. Tmu8.tme two curves in Figure 4 do not necessarily approach the same upper bound. VI 's '1 74 A priori, many possibilities exist regarding the nature Of the differences between Ao(t) and AE(t). Shape, starting date, average adoption rate, and total number Of units eventually adOpting could each be affected in several different ways. The net effect of extension influence could be, for example, to increase starting date and lower average adoption rate while leaving shape and total number adopting about the same. This might be expected to occur where innovators are in a position to change long before they would hear about the idea if only non-extension sources were available, but late majority and laggards cannot be speeded up because the group Of factors which would have held them back without extension influence also holds them back at the time extension effort is applied. On the other hand, it may be possible to speed up both starting date and average adOption rate by an extension program. AE(t) may also involve a larger total number Of units adopting a given change. This could be expected to happen where extension introduced a change to an area much sooner than it would have happened without extension and the environmental characteristics of the area are so much different then than they would have been later that the change will spread to many more units than otherwise. If extension effort does change an adOption distrib- ution, benefits per decision-making unit may be different in the extension-influenced distribution than in the alternative \. _- b l 75 distribution. In fact, extension may be able to produce benefits even if it does not change the adOption distrib- ution at all by simply making sure that changes are made correctly - not just adOpted. This would appear likely to happen where an innovation is quickly adOpted but requires considerable time to be incorporated in the technical Operation Of the adOpting units at Optimal level. It can also be noted that any one extension program may affect more than one adOption distribution. This would likely be the case where interactions are present in adOpting units which make a composite change more advant— ageous than the sum of the several component changes made separately. Other Extension "Outputs" Although adOption distribution curves provide an intuitively-appealing method of representing the net influence of extension Of specific changes in a given geographical area, it has already been noted that they may not be able to describe some sorts of changes which extension effort may inspire. If the changes made by individual decision-making units are unique to those units, adOption curves Obviously do not exist. For example, an extension agent might aid a farmer in designing a new barn - a rather unique change which would not be applicable to any other fanms (i.e., even if other farmers were building new terns. the appropriate constructions for them might all be :4 76 quite different). Extension workers can help farmers and rural housewives move closer to the Optimal combinations Of inputs or goods and services for their situation. Such results might possibly be described by adoption distribution curves but, to the author's knowledge, this has not been done. Extension agencies can aid farmers in making decisions on rejecting innovations as well as adopting them. Where innovations which are not worthwhile are being promoted by private companies, farmers may find that extension advice is quite useful because it helps them avoid losses they would have incurred otherwise. AdOption distribution curves are not likely to be useful for describing this potential result of extension work. Thus, for secondary extension "outputs", and for some changes in rural society as well, it appears to be necessary to examine all individual results or changes and evaluate these one by one on their own merits (based on the criteria outlined in Chapter III) to arrive at some measure Of total extension benefits. This may well be a tedious and cumbersome task but where the adoption distribution technique cannot be employed, no other method is readily apparent. Environmental Influences on Extension Work The results of the application of a given set of extension activities to any particular rural society will C1ePend rather vitally upon a number of environmental factors. 77 The factors of relevance will probably vary somewhat from one situation to the next. However, it is possible to distinguish four groups of environmental factors which appear to be significant in most cases. These are: (i) Physical Conditions (ii) Product and Input Markets (iii) Non-Extension Information Sources (iv) Institutional Framework Physical Conditions Physical resources include such factors as soils, climate, buildings, machinery, roads, telephone systems, electricity systems, and so on. Both natural and man-made physical resources play an important role in determining the usefulness of many changes to rural society and thus are an important set of environmental variables for the extension system model.12 If physical resources are apprOpriately priced, their effects on the usefulness of new ideas can be analyzed in terms of profitability of prOposed changes. (Of course, some Obvious technical limitations and advantages can be detected without having to incorporate prices into the analysis.) Nevertheless, the price mechanism, when it works effectively, is an immense aid in reducing the complex set of physical environmental features affecting the utility of a new idea to a more convenient indicator of that utility. Not only is the character of physical resources 12In some cases, physical conditions may be the object of change by means of an extension program (e.g-. a conservation program). In such a case, those physical factors affected would be output variables rather than enVironmental variables. in 78 important but also their variability among decision-making units may be critical. It may well be that, because of existing resource endowments, techniques which are profit- able on some farms in an area are completely irrelevant for other farms producing the same products. This would lower the extent of applicability of a given change and reduce the potential pay-Off of a given extension effort. It could also have important implications for income distrib- utions. It is fairly evident that agricultural extension has been able to achieve little improvement in farming in some areas simply because natural physical resources were so poor relative to competing areas that few or no useful changes could be promoted. Apalachia in the U.S. and parts Of the Maritimes in Canada are cases in point. As one writer notes, "the acceptance of improved farming practices is determined largely by economic considerations (advant— ages)"13 and these, in turn, are determined by physical resources and prices. Thus, it is evident that physical resources indirectly affect the relative advantage of farm people remaining in agriculture versus changing to some other occupation. Where physical conditions are severe, agricultural extension workers may find that assisting in the adjustment Of rural peOple out Of agriculture is the most beneficial function they can perform. . 13E.A. Wilkening, Acceptance Of Improved Farm Pract- 1ces in Three Coastal Plain Counties, Technical Bulletin 98, Nor Chro na Agricu tura Exper ment Station, Chapel Hill, North Carolina, 1952, p. 5. 79 The effects of physical conditions can be significant in determining the net advantage of changes in all three types Of rural decision-making units (i.e., farms, households, and community organizations), although they are perhaps most evident in the first and last Of these. In some cases, community organizations may be able tO overcome physical limitations which are not within the control of individual farmers. An example of this is the construction of drainage ditches which can serve as outlets for tile drainage on a number of farms in a given area. Sometimes individual farms and households may have to change certain physical characteristics Of their Operations before other technical changes can occur (e.g., the advent of tractors and electricity Opened up many possibilities for further technical changes). However, there are many physical conditions which are essentially uncont- rollable (e.g., climate, soil). Whether physical conditions can or cannot be changed, they are certainly going to have an influence on input-output relationships in all three types Of rural decision-making units and hence, as will be seen in Chapter IV, on benefit flows attributable to extension work. In addition to influencing the actual net advantage Of any particular technical change in rural society, past experiences with physical conditions may have influenced the beliefs of rural people to the extent that key managerial changes must occur before certain technical changes can be accepted. For example, if farmers believe that fertilizer burns plants because they have seen some plants wither under 80 other chemical treatments, it is unlikely that they will believe extension agents who promote the use of fertilizer. The physical environment in which peOple live undoubtedly influences their beliefs about both normative and non- normative matters and it is with this managerial structure which extension agents must begin their work. Another important way in which physical conditions can enter the extension system model is through their effect on the costs of doing extension work. The physical environment of an area may influence travel time per farmer, transportation costs, and/or the cost of applying other extension techniques such as mail circulars, radio, television, and so on. These effects can be so great as to completely eliminate the feasibility of using some extension techniques in some areas. Product and Input Markets The effect of market structure on extension efforts is sometimes overlooked by extension workers, particularly those not trained in economics. As societies become more developed, farmers in most communities find that their local price levels are determined as part Of national or even international market adjustments. Thus, the profitab- ility of any change in local agricultural production units, and perhaps local households and community organizations as well, is determined by influences beyond the boundaries of the local area. This has been particularly striking in 81 some less-develOped countries where shifts have been made towards specialized cash crOps for exports and away from diversified food crops for home use. Once the shift has been made, fixed assets and/or changes in technical knowledge may place the adopters in a situation of high risk. What may have looked like a big pay-Off to extension effort can suddenly turn into a disaster under such conditions. Price stability and market dependability become significant considerations as the shift to a market economy occurs. Farmers may be entering a high-risk Operation if they commit themselves to technology which requires inputs whose future availability is questionable. Extension workers may find that changes in managerial structures or changes in community organizations must occur if such risks are to be adequately analyzed or coped with by rural peOple. On the other hand, they may find that changes in the market structure itself must occur before the risks to farmers are reduced sufficiently for them to make advocated technical changes. Extension work with all three types of decision- making units is likely to be influenced by markets. Rural households are unlikely to adOpt recommended sanitary practices if the inputs required to implement these are priced beyond their incomes or are not available locally. Community organizations may find similar barriers to change as they attempt to improve roads, schools, or other community 82 projects. Thus, market structures, and particularly prices of products and inputs are important determinants of potential benefit flows from extension efforts. Also, past observations of markets may have provided rural peOple with beliefs which must be overcome if these beliefs are barriers to change and are, in fact, no longer valid. Non-Extension Information Sources In almost any rural community there are a variety of non-extension information channels linking farm peOple to information sources outside the area. The number Of these channels, the messages they carry, and the sources they tap will have an important bearing on the effectiveness of extension work. It appears as though they can both complement extension activities and substitute for them. Some Of these channels exist because Of deliberate attempts to create changes; others do not. Some may reach quite a different subset of the rural pOpulation than the subset reached by extension workers; some of them may reach essentially the same subset. The principal categories of non-extension information channels appear to be agribusiness firms and the mass media. It is fairly obvious that, in technically-advanced societies a substantial substitution relationship exists between agricultural extension activities and the efforts of agribusiness. Farm input suppliers find it to their advan- tage to promote their products vigorously and they often flat»! I 4 7'" 83 provide the farmer with useful (although perhaps one-sided) advice in so doing. (Even a fair amount Of competition in this area does not necessarily ensure the release of all of the information a farmer might find useful.) Private farm management consultants are not uncommon and where they and agribusiness substitute as a link between farmers and information sources it would appear that the potential returns to extension effort are likely to be much lower than where they are non-existent. Accordingly, it would appear appropriate to curtail extension efforts somewhat when such substitutes arise.14 The mass media, which include newspapers, radio, periodicals, and television, Often perform an educational function independent of extension work. Thus, even if extension agencies never utilize these channels of communic- ation, their possible effects may be quite different depending on the nature of the messages which the mass media carry and who these messages reach. For example, if the mass media keep farmers informed Of market conditions, the results of some extension efforts to influence produCtion technology may be quite different than if farmers are in a high risk situation because of lack Of market knowledge. Of course, extension agencies Often utilize the mass media directly and, in such cases, the character Of their messages together with the cost Of producing these messages is likely 14However, it is unlikely that non-extension infor- mation sources will invest enough in disseminating useful information, particularly in cases where diffusion occurs, as they cannot capture all Of the benefits to be gained from such an activity. 84 to be different than if some other extension method were used. Where such non-extension information sources complement extension work, the pay-Off to a given extension effort will likely be enhanced. Institutional Framework The institutional framework forms the third major set of environmental variables of relevance for the extension system model. The term "institutional framework" includes a wide variety Of formal and informal organizations and customs which provide the set of rules and public facilities essential for the resolution of conflict when men pursue individual Objectives in contact with one another.15 Per- haps the most widely-noted feature of the institutional framework in an agricultural development context is land tenure patterns. Although much research has been devoted to this tOpic, many of the effects of different types Of land tenure on technical and managerial changes are not well understood. Of course, many other factors are included in the term "institutional framework".‘ Credit arrangements, market institutions, social customs, legal facilities for 15Of course, some of these institutions may be within the control of the local community and may thus be modified by extension influence as suggested earlier. Simil— arly, extension can have a feedback effect on institutions outside of the local community. However, for most individual changes in farms, businesses and rural households, the institutional framework is rigid - changes in this framework are likely to occur only if it becomes evident that they are necessary to facilitate a significant number Of changes in individual decision-making units. ... .I, 41%|; ‘- 85 protection of the individual, the structure whereby community decisions are made, and others are of relevance. For example, the approval Of local religious leaders in some areas, is absolutely essential for change agents to be accepted by the local community. The ways in which the institutional framework can influence the extension system model are many and varied. Institutions may either thwart or facilitate changes in farm businesses and rural households. For example, credit institutions which provide rural peOple with adequate, dependable sources Of low cost funds for making technical changes Obviously have a much different effect than those which do not. Similarly, national institutions may have either a restrictive or encouraging affect on the activities of local community organizations. For example, national policies which do not allow local communities either the funds or the authority to build roads or schools will have a much different affect than those which do. As much of the institutional framework Of any society is a result of a variety of governmental activities, it is perhaps appropriate to regard agricultural extension as a component of a system of governmental activities which itself is part Of the environment in which extension must work. As has already been noted, the benefits Of individual changes are largely determined by physical cond- itions and prices but these in turn are influenced heavily by institutions. Roads, schools, market institutions, and 86 land tenure arrangements are cases in point. Thus, govern— ments are faced with the difficult task of balancing and sequencing extension expenditures with those Of other programs in order to ensure that the return to the total public budget is as high as possible. As noted at the out— set of this discussion, this general problem is beyond the scope Of this study. At this point, it is only possible to note that other governmental activities16 may create an environment where extension can be extremely productive - perhaps essential to the total success of the project. For example, government irrigation and drainage projects may change local physical conditions so greatly that a drastically different set of techniques are necessary for farmers to COpe with these. Land tenure changes may do the same. Conversely, lack Of changes in regressive tax structures, oppressive land ownership policies, restrictive trade practices, and various other institutional conditions may so limit the possibilities of useful extension work as to make the expenditure Of extension funds a complete waste. 161t would, of course, be a mistake to assume that a favourable institutional framework can be provided without an adequate tax base. Some such services are useful as both consumption and intermediate production goods. In the latter case they will likely generate further incomes; in the former they will merely use up existing ones. Many of the services available in advanced societies today have only attained their current character because incomes have increased to the levels where these services could be afforded. The services did not always generate incomes to pay for themselves. rim - 87 Idea Sources and Extension Activity There are a variety of idea sources which extension workers may draw upon to define useful changes for promoting in rural society. These include basic agricultural reSearch work in government and private industry both locally and, in some circumstances, in foreign countries; experiences Of farmers in local and distant areas; and applied or practical experiences Of extension personnel gained as they pursue their daily duties. There is often a significant lag between the time when discoveries are first made and the time when extension peOple start to promote them. The practices being promoted by extension services in most parts Of the world today were largely developed by governmental and private research expenditures 5, 10, or even 20 or more years ago. Such a lag is often necessary to allow time for ideas to be evaluated and tested under field conditions. But it may also occur because extension agencies are not keeping informed of, or recognizing, worthwhile ideas. On the other hand, extension work will become barren unless useful, relevant, new ideas are supplied to extension workers in a fashion which they can readily understand and interpret for the rural people in the communities they serve. There is little hope of extension workers achieving signif- icant results unless they are promoting changes which are "practical and worthwhile given the existing system".17 17See footnote 24, Chapter I. i l; 88 Because of the wide variability in physical conditions, prices, and institutions influencing rural societies, this means that research must Often be designed to discover relationships that apply under very localized conditions. It also means that extension personnel can, as mentioned earlier, perform a useful function, by influencing research workers along these lines. Thus, it is rather apparent that research and extension work are highly complementary in the efforts to bring about beneficial changes in rural society. This relationship probably applies to all those categories of decision-making units defined earlier although it is most commonly associated with farm firms. Thus, it is difficult to assess the benefits attributable to one or the other Of these independently. Should one be lacking, and another present, rather high returns could be attributed to the provision of the missing link. An Extension System Model It seems Obvious that, if extension workers are to induce and inspire changes in the activities of farm peOple, they (extension workers) must create changes in farm peOple themselves. These changes include improved knowledge of agricultural production techniques, home economics, farm management practices, and group activities; changes in attitudes towards new ideas and the usefulness Of community action; and new, improved abilities or skills acquired through 89 practice or Observation. Some of these changes are quickly translated into specific technical changes; others have a much slower utilization and perhaps apply to many technical changes. All such changes in farm peOple are created by out-Of-school educational programs. As one writer points out: The essence Of American extension is that it is an out-Of-school educational process: (1) working with rural people along those lines of their current interest and need which are closely related to gaining a livelihood, improving the physical level Of living, and fostering community welfare; (2) utilizing particular teaching techniques; (3) conducting with the aid of certain supporting activities; and (4) carried on with a distinct spirit Of co- Operation and mutual respect.18 It appears to be the Opinion of many informed individuals that this essence is applicable to extension programs throughout the world. Of course, the mechanics of extension work usually must differ from country to country - or perhaps more appropriately, from area to area. As Mosher points out, "We could have a study of comparative extension education in the United States alone!"19 The needs and interests of rural people vary from area to area. Because of the variations in 18Arthur T. Mosher, Varieties of Extension Education and Community DevelOpment, Comparative Extension Publication No. 2 (Ithaca, New York: New York State Agricultural College at Cornell University, 1958), p. 12. lgIbid., p. 10. r**-»-—-tn 90 environmental conditions discussed earlier, the methods by which rural people gain their livelihood, the factors which contribute to an improved level of living or community welfare, the teaching techniques which are effective, the supporting activities which are useful, and the means by which co—Operation and mutual respect are achieved may be vastly different in one area than they are in another- In order to specify the set Of endogenous relation- ships in an extension system model, one would have to amalgamate a great deal of knowledge, both conceptual and empirical, from many disciplines. This knowledge is not available at present and, even if it were, the task of assembling it in a useful form would be enormous. Thus, no attempt will be made to build such a model here. However, it may be enlightening to very briefly outline its general nature and discuss some Of the probable building blocks in such a structure. Figure 5 expands the rather naive model Of Figure 2 by incorporating the main features of the preceding discussion. The solid directional lines indicate significant, direct influences whereas the dashed directional lines show Obser- vations and/or feedback effects. A translation of "rural peOple" into"decision-making units" is shown to illustrate both the diffusion and managerial processes. The dotted lines show where extension costs and benefits occur and emphasize how these must be compared in this study in order to evaluate extension activities. 91 .muwuocon 0cm munoo cofimcouxo :oo3uon mesm:0wumaou onu one3osm Hopes Emumxm cowmcouxo am No 3mfl> fimumcwm ¢II.m ousmwb muawwcom COHMCQUXN 1w_m V _ _ _ u . momcsco Assasumnsouo macsumuqnmmuo >uwc2EEou on m0~0£0m=o= Assam Any mshwm such Ame noduusooum.~ unawmsuoo nu=o> Haunt ADV m0>w30m=o= An. muOEumm Ame oHooom Hausx nocuo.~ mCOMumecmmuo >uwc=EEou Hoe moaonwmsom Hausa Any meusm Eumm Ame ucofioomcmz.H [ll-III. mBHZD UZH¥s30msoz any muoEumm Ame mucuafioHuusm cosmcwuxm.a mqumm A<¢Dm m coHumuomo M Ecumoum _ . macamwooo ” I II ..I.. \H [I IN I I. ucvswomsm: ZOHmZWme A<¢DBADUHKU¢ MUWOU cesmcouxm monusom mon .m IIIJ xuozusmum ascoHusuwumcH .v IIIJ moousom :OMumENONCH casmcouxmlcoz .n null muoxum: usmcH one uusooum .~ uuu$ macsuwocou Housm>nm .H IIL Bzmzzomm>zm 92 One of the essential building blocks Of the extension system model would be a realistic description of the individual adOption process. The management process dis- cussed earlier includes this but, as management is concerned with much more than adoption, and as a considerable amount of work has been done with reference to the latter area, it may be worthwhile examining it. Rogers has described an individual adoption process which includes six stages: aware- ness, interest, evaluation, trial, adoption, and continued use or acceptance.20 This has come to be known as the "Americal diffusion model",21 but not all research supports its existence in the usual form. Some findings note that farmers skip one or more stages, especially the trial stage.22 A number Of attempts have been made to model the diffusion process quantitatively. These do not appear tO be the exclusive subject matter of any one discipline and the literature is numerous and diverse in nature. For example, Brown focuses on the efforts of geographers in this area in his 1966 dissertation. His concern is with innovations adOpted by individuals or decision-making units acting as single individuals and, being a geographer, he limits his activity mainly to spatial considerations. However, he 20E. Rogers, op. cit., p. 81. 21Otis Oliver-Padilla, "The Role of Values and Channel Orientations in the Diffusion and AdOption of New Ideas and Practices: A Puerto Rican Dairy Farmer's Study". Unpublished Ph.D. dissertation, Department of Communication, lfichigan State University, 1964, p. 12. ZZIbid. r“ Wm 93 points out that the "availability of the innovation to potential adopters may be quite invariant with the spatial 23 At times, pattern Of the availability Of information." knowledge of an idea precedes the facility to use it and inspires individuals to search out this facility; on other occasions the reverse is true. Brown pursues this idea and develops a model which, in essence, states that "the probability that a randomly selected resident of place i adopts the innovation in time t+l equals the probability E222 adoption did not occur by that individual before time t+l, aaaa he receives sufficient information about the innovation during time t+l, aaaa this information is accepted because he is neither socially or economically resistant to adOption, and aaaa during time t+1 he comes into contact "24 It can with at least one distributor of the innovation. be readily seen that extension workers can influence the probability that an individual receives sufficient infor- mation about an innovation but they may have little or no control over the other probabilities in the chain. It has Often been suggested that the nature of the change being promoted would be important in determining the success of a given extension program. Rogers25 suggests 23Lawrence A. Brown, "Diffusion Dynamics: A Review of the Quantitative Theory of the Spatial Diffusion of Innovation," Unpublished Ph.D. dissertation, Northwestern University, Evanston, Illinois, 1966, p. 3. 24Ibid., p. 31. 25E. Rogers, Op. cit., p. 124. F- a 94 that relative advantage, compatibility, complexity, divis- ibility, and communicability are characteristics of innov— ations which appear to be important in determining adoption rates. Relative advantage is defined as "the degree to which an innovation is superior to ideas it supersedes".26 Compatibility is "the degree to which an innovation is consistent with existing values and past eXperiences of the 27 adopters". Complexity "is the degree to which an innov- ation is relatively difficult to understand and use".28 Divisibility is "the degree to which an innovation may be 29 tried on a limited basis". Communicability is "the degree to which the results of an innovation may be diffused to others".30 Unfortunately, none of these factors are properties Of the change pa; aa. They all depend to some extent on other factors such as existing managerial structure, expected price levels, and existing income levels. Even divisibility is not entirely a property Of the change by itself. Bulk milk tanks are cited by Rogers as being indivisible but, on a very large dairy farm (say, one having 5,000 cows) they would be no more indivisible than a new variety Of corn on a very small farm (say, one acre). 26Ibid., p. 125. 27Ibid., p. 126. 281bid., p. 130. 291bid., p. 131. 30Ibid., p. 132. 95 As examples Of the many and complex endogenous relationships that may exist in an extension system, it may be useful to use two communication studies which were done in Colombia. Ramos3l took note Of the fact that rural people may form preconceived images Of the characteristics they feel extension agents should possess. (Such images might be formed on the basis Of previous experiences with persons who are regarded as sources Of information, or figures of authority in the community.) If the extension agent does not fit these preconceived images, his credib- ility: favourability and extent of contact in the local co"munity may be lower than it would have been if he did fit them. Ramos' study suggests that, in Colombia: experience, age, and social distance32 appear to be import- ant considerations influencing these factors. Thus, Colombian extension agents who do not possess these charact- eristics to the extent that rural people feel they Sho‘fld' Will likely be less effective than those who do. In terms :f the model used here, a change agent who is, a priori, ess acceptable to rural people will likely have less effect on a n adOption distribution than ones who are, a priori, more \ in Th 31Eduardo Ramos, "Client-Change Agent Relationships of Coree Colombian Villages", Unpublished M.A. thesis, Dept. runInnications, Michigan State University, 1966. In em 32Social distance has been defined as the difference grou pathetic understanding that exists between persons, 90831:: or between a person and a group. It apparently is 1e to Obtain useful measures Of this concept. 96 acceptable.33 A study carried out by CIRA34 in 1967 investigated the effectiveness of the press, radio, fliers, and loudspeakers in a diffusion campaign in the areas of Cundinamarca-Boyaca and Antioquia-Caldas. The results showed that loudspeakers alone or in combination with other techniques were the most effective means to get campesinos to take the action desired. How the cost Of loudspeakers compared with other methods and how loudspeakers would work With Other sorts of changes or in other areas Of the country was not known. The literature in the social sciences abounds with StUdies such as those already discussed. However, not only are there disagreements as to the nature Of the relationships between key variables, there is much contention as to which variables are key ones and how they are to be measured. For example, Whittenbarger35 developed and tested an "attitudes - towards _ social - change" scale. He stated that such a scale "should be a valuable tool for change agencies as it "0‘1 n . . 1d enable agencies to be appreciably more effiCient and rura 33Of course, if agents who are regarded a priori by regal: PeOple as less acceptable are cheaper than ones w o are d acceptable, this information by itself does ed a riori as not provide any basis for choice. Sobre 34Centro Interamericano de Reforma Agraria Estudio Ma Eficacia de Divesos Medios de Comunicacion en La IIIS<|;_.L2acion de una CamEaha die Divul'gacion (Bogota, Colombia: t1JtO Interamericano de Ciencias Agricolas OEA, 1967) . Chang 35Robert Lee Whittenbarger, "Attitudes Toward Social mentae in a Rural Colombian Community: An Attempt at Measure- Unive' Unpublished Master of Arts thesis, Rural Sociology, rsity Of Wisconsin, 1966. ‘ 97 effective by permitting them to make a rapid and more meaningful selection of communities among which to allocate their all too scarce time, money, and effort ..."36 As this exercise was carried out in Colombia it would have been convenient for this writer if his scale had been validated in the field. Unfortunately, it was not. 361bid., p. 2. r«--———-—---I CHAPTER IV METHODS OF MEASURING EXTENSION "OUTPUTS" In order to measure the efficiency of agricultural extension work it is necessary to measure those "outputs" °f eXtension programs which were identified in Chapter II. A5 ""111 be seen in Chapter V, choices among programs are much more readily made if all "outputs" can be reduced to a comm°n denominator. It is particularly convenient if the CoInmon denominator is the same for "outputs" as it is f . or Thus, this chapter discusses pOSSible ways "inputs". of measuring extension "outputs" in monetary terms. In order to attach monetary values to extension-inspired changes in groups of decision-making units in rural society; it is necessary to have (a) some way 0f valuing SUCh changes for individual units, and (b) a method 0f aggregating these Flues for all units affected. Thus, the first topic Of interest is the individual units themselves. Measuring Changes in Individual Decision-Making Units Tech . l“ical Changes in Farm Firms If the return to the owned resources Of the entre- Dre he‘lr in period t is given by R(t), then the change in this 98 r» -——:———‘——-. , an 99 chantity from one period to the next is given by: (1) AR(t) = R(t) - R(t—l) Now, if there are no external effects, no associated managerial changes of value, and the changesl in xj(t) and Yi(t) which give rise to AR(t) are directly inspired by extension influence, then AR(t) can be regarded as a comPlete2 monetary measure, for that production unit in penuxi t, of the results of the extension effort which prod- tmed the changes in xj(t) and yi(t). However, several problems arise in trying to utilize this measure of the results of extension effort. These are: (l) Exactly what changes in x.(t) and yi(t) are attributable to extension effort? (2) How can AR(t) be used to calculate the total value for that production unit of the extension influence which resulted in AR(t)? (3) How should managerial costs incurred by the decision-making unit be allowed for? In answering the first question it is necessary to recall that several different types of changes in xj(t) and yi(t) are possible. Extension-influenced changes include only the adoption of some innovations and some controlled adju Stments in input and/or output mix. They may be iff- lcult to distinguish in practice from random changes in {maptl 1The variables x.(t) and y.(t) were defined in lute er II. The reader will recall that, given the usual is popretation of the production function defined there, it xj(t) SSible to have changes in yi(t) without changes in rele 2"Complete" means that no other results are of Val"Ice. 100 xj(t) and yi(t) and controlled changes in these variables inspired by some other influence such as changes in price expectations or information provided by other change agents. Variations in yi(t) due to weather conditions are especially common in agriculture and, only in rare circumstances, will one find situations where price expectations are fixed and other change agents are entirely inactive. Thus, it seems aPPrOPriate to ask the question, "What happened because of extension effort that would not have happened otherwise?" This View is analogous to the marginality approach in production analyses. Other factors may have some influence on whether a given change is made or not and may actually act concurrently with extension. However, the value 0f the Change attributable to extension is the difference between that which actually results and that whiCh W0111d have resulted if the extension effort was not present. In order to resolve the second problem mentioned above it is necessary to refer to the expression for R(t) esté‘blished in Chapter II. From this, it is possible to See AR(t) is given by: (2) n n+k m AR(t) = [iglPi(t)Yi(t) + i=1§+lpi(t)yi(t) - jglem Xj(t) m+h n - j=l}|‘1+1pj(t)xj(t)] - [iglPi(t-1)yi(t-1) m \ adepti 3This is the micro equivalent of the "shift" in the On distribution described earlier. 101 where there are k new outputs and h new inputs in period t. If flu:.levels and prices of xj(t) and yi(t) remain the same hifuturwe periods, a benefit flow results with the constant pmtperdxod value being equivalent to R(t). The length of fins beruefit stream will be the period between the time when the entrepreneur makes the change in question due to amension influence and when he would have made it without mmension influence. This latter date is not readily Observable in practice and presents a difficult estimation mmblenh (Whether this is any more difficult than estimating Um useful eConomic life of something more physical in nature mmh as a hydro—electric project is a point of contention.) It . . . woulxi seem useful to try Simultaneously to estimate this da . . te afui solve the problem of whether a given change is att ' . . . rlblltable to extenSion or not by asking "When would this Cha nge have been made without extension influence?" These conclusions raise two further questions: (a) what is the appropriate length of the benefit stream if the change would never have been made at all without extension effort? (b) if the prices and levels of x.(t) and y.(t) vary over time, should the benefit stream incorporate these variations? The entrepreneur will likely4 continue to retain his impro Ved combination of X3. (t) and yi (t) until (a) he retires from b uSiness, (b) other improved technical changes are dis- COVere . . . - <1, or (c), price changes make it unprofitable for him actua a1 4Of course, if he does not correctly assess what has lnng‘, atiolr'llappened, he may discard a worthwhile technical '13—]. 102 to retain it. If price changes make it necessary for the entrepreneur to stop using the new techniques, the benefit flow stOps. If other improved technical changes are dis- covered, their value must be based on further improvements in the combination of xj(t) and yi(t) and, therefore, the benefit flow in question does not end even if such changes are made. The question of the length of the benefit flow for decision-making units which cease to exist is not as readily answered. As far as the entrepreneur is concerned, new techniques will produce benefit flows only as long as he is farming. (This may explain, to some extent, why older farmers tend to be less inclined5 to adOpt new practices than younger ones - the present values of the changes are lower for them because of the shorter expected benefit flow Period.) However, when the farm is transferred to another operatOr, he too may use the technique in question. If the new c)Perator learns from the original operator (or in some other indirect fashion from the extension effort) . the benefit flow continues. In essence, each technical innov- ation . 6 . . Produces a permanent benefit flow even if it becomes Obsolete (i.e., is supplemented by another innovation) prOV' . 1dlng it is passed on to future generations. Pr 5Herbert F. Lionberger, Adoption of New Ideas and ac . Pp. 1:Efies, (Ames, Iowa: Iowa State University Press, 1960), at so 6The permanent benefit flow may take on a zero value not me future time. This would result from price changes — tom technical obsolescence. 103 Except for the cases where innovations produce shifts in supply or demand curves and thereby affect price, prices will be independent of the technical change itself. Therefore, the appropriate benefit stream should be cal- culated using the actual prices for each period. Price- induced variations in the levels of xj(t) and Yi(t) should also be incorporated by using the actual values of these variables. Thus, price changes can easily make the bMisfits in future periods much different from that in which the change is made. Presumably, they will not likely make benefits negative as producers can be expected to discard the innovation if this situation arises. Of course, if the innovation involves the acquisition of capital assets, it may not be possible to realize sufficient funds from the sale of these to avoid a loss. However, producers may not initially8 adopt innovations at Optimal levels. They may Simply adOpt and move towards Optimum over time as they learn more about the innovation. Where this is the case the appropriate calculation of the benefits in any period W0111d again have to be based on actual prices and quantities. As aetu a1 prices and quantities for individual units are likely 7These will be discussed later in the chapter. all e 8In fact, some innovations may not be profitable at and vet! though they are promoted by extension personnel CDpted by some farmers. In this case, AR(t) would be negati Ve for adopters and zero for non-adOpters. A 104 (zostly to obtain in practice, they may be usefully approx- :imated by assuming Optimization and using average prices or Iprojected price trends. However, this practice is Obviously :inappropriate for valuing extension efforts directed at ‘assisting farmers to move closer to Optimal combinations. If the costs of additional resources which were devoted to managerial activities by the farm manager in order to make the change are not already incorporated in AR(t) they should be added.9 These may include lost earnings ‘when attending meetings, transporation expenditures, and so on. Unfortunately, it may be difficult to assess the Inanagerial costs associated with any particular change. .Also, changes in decision-making costs due to extension effort may be either positive or negative. It can be noted that even if no change occurs in the extent of decision- lmaking activities Of farmers as a result Of extension influence, these activities will likely occur at a different .Eim§_than they would have otherwise and this may well represent an added cost. 'Iechnical Changes in Rural Households The discussion of consumer theory in Chapter II implicitly suggested that, if there were no externalities, H 9Presumably all production costs of making the change have been included. This would include physical changeover costs, special input acquisition costs, etc. These and added managerial costs might be "once-and-for-all" costs and, if smflxis the case, AR(t) would take on a different value in fiuzadopting period than thereafter. The cost of replaceable caPital items would be prorated over the useful life Of assets according to standard accounting procedures. 105 ‘the differenceSin utility attributable to technical changes .in households would be appropriate measures of their worth. frhus, corresponding to the change in R(t) defined in the last section, AU(t) can be defined as: (3) AU(t) = U(t) - U(t-l) iflowever, as with agricultural production units, several problems arise in trying to employ this as a measure Of the value of the extension effort which inspired the change. These include: (a) determining what technical changes occur because of extension effort, (b) how does one handle costs Of managerial activities and/or "once-and-for-all" technical activities associated with making the change? (c) who is to judge the household's welfare or utility level? (d) regardless Of who judges welfare, how can it be measured? (e) how should changes in preferences over time be allowed for? (f) how does one use AU(t) to arrive at an estimate of the total value of the effort which inspired the change? Problems (a), (b), and (f) are essentially the same as the ones encountered in valuing technical changes in production units and can be handled in essentially the same way. Costs Of managerial activities and/or "once-and-for- all" technical activities of the decision-making unit in order to make the change are essentially capital costs which flunfld be deducted from benefits (or added to extension 106 «expenditures) in the period in which they occur. Problems (a) and (f) can be attacked simultaneously by asking, "When ‘would this change have occurred if extension effort had not been applied?" In the earlier discussion of consumer theory it was assumed that the informed individual is the best judge of his (or her) own welfare (i.e., the utility function for any consumption unit is given by the family's own preferences). This need not have been the case. The analysis could have been applied to anyone's construction Of the household's utility function providing such a construction actually reflected apprOpriate values for that household. But what is apprOpriate? It has already been argued that an individual's derived value scheme may not represent what is best for him if he is not thoroughly familiar with and able to evaluate all possible effects of the goods and services available to him. Indeed, this appears to be the rationale ‘behind the policy (which is generally-accepted in most societies) that children and mentally—ill peOple should not loe allowed free choice in a variety Of circumstances even ‘when such choices have no external effects. Of course, the ‘problem under discussion disappears if everyone concerned arrives at essentially the same utility index for a given household. However, in extension work conflicts can and often do arise between extension workers and their clients. And, where the clients are relatively uninformed about a {IV 107 particular topic10 extension workers may, in fact, "know Ibest". As some position has to be reached on this point Ihere, the author accepts the assumption that the informed individual is the best judge of his own welfare providing ‘he is well informed on 22th normative and positive matters. This position appears to be consistent with bgth the dis- cussion Of basic and desired values in Chapter II and the usual position taken by economic theorists. The problem Of measuring the utility or preferences of individuals is also a difficult one. Von Neumann and Morgenstern11 have developed an Operational utility theory and this has been applied by several other workers.12 How- ever, these applications typically involve measuring the utility Of one or two items under risk conditions. They do 10In such cases it would seem that extension activ- ities directed at managerial changes could eliminate such conflicts providing extension workers are both highly- informed about the topic and capable of effecting managerial changes. Nevertheless, where extension workers have difficulty convincing their clients Of the value of some proposed change, they may be well-advised to re-examine their own analysis of it. This applies to households, production units, and community organizations. Apparently some extension workers make re-evaluation a common practice while others adopt more of a "missionary" philOSOphy which pre- cludes such questioning of their own values. Another factor Of relevance here is the cost of creating managerial changes to eliminate the conflicts. Where this is high the best strategy may be not to promote the change which is in conflict. 11J. Von Neumann and O. Morgenstern, Theory of Games and Economic Behaviour, New York: John Wiley & Sons, 1964. 12See, for example, Albert N. Halter, "Measuring the Utility of Wealth Among Farm Managers", Unpublished Ph.D. thesis, Michigan State University, 1956. o 1 \ e c 108 not provide a basis for arriving at a utility function for all choices facing the consumer. Furthermore, they are not sufficiently well-developed to permit accurate measure— ments to be made at a low cost. Thus, trying to estimate complete individual preference surfaces is not feasible in this study. Nevertheless, the ranking Of some alternatives may be enlightening and will be used. The problem of allowing for changes in preferences over time is not independent of either of the two preceding problems. By now it is Obvious that preferences can change as new information is acquired and/or new methods of evaluating Old information are develOped. But, if an individual changes his (or her) preferences from one period to the next, should the AU(t) be evaluated with the new preference function in period t and the Old one in period t-l, or should some other method be used? Presumably the Old preferences existed because the individual was uninformed and, if he had been informed, he would have actually had the new preference function in period t-l. This being the case, the new preference function would, according to the reason— ing on the problem Of who judges the individual's welfare, be the apprOpriate one to use in both cases. This position will be adOpted here although there are some rather deep philOSOphical issues associated with this problem which have been entirely ignored. Because Of these difficulties it is not feasible to 109 use utility theory directly to measure the value of technical changes in rural households. However, it can be used indirectly in many ways. The fact that any worthwhile technical change occurring in a rural household can be regarded as an increase in utility from a given income is particularly useful in making evaluations of changes. When income is fixed, increased expenditure on some goods or services necessarily implies decreased expenditure on others. Thus, any reduction in the cost of producing a given amount of utility by making more efficient use of some goods frees funds for other uses and this freed expenditure can be used as a measure of the value Of the change. For example, if a new cooking technique produces less waste and, as a result, food eXpenditures can be reduced by 10 per cent while the family remains equally well—fed, the cost reduction could be regarded as a measure of the value of the change. Similarly, if a housekeeper becomes able to provide a service that is available in the market place, the market price of this service less the costs of producing it would be an appropriate measure of its value. This would apply whether the family had previously purchased the service or not as long as it is now produced at home. Programs directed at instructing housekeepers to diagnose minor illnesses can be valued in this way. Nutrition and health programs, which will be dis— cussed later, are examples of extension activities which are ‘H 110 Often designed to produce a higher utility at an equal or higher cost. Where higher costs are incurred, some other expenditure must be reduced and the loss in utility from these must be less than the gain from the goods acquired through the increased eXpenditure if the net gain is to be positive. In both these sorts of programs there Often exist Objective criteria for the increased welfare due to adoption. Nutritionists and medical scientists know a good deal about the loss in physical well-being which results from poorly-balanced diets and poor health practices. How- ever, there is no easy way to balance the loss in utility which will result if purchases of other goods must be reduced in order to improve diets and hygiene. Because of these difficulties it may be useful to adopt a few rules-Of-thumb for evaluating changes in rural households. These are: (1) Assume that the value of ideas rejected by rural households when they are well-understood is zero less the costs of learning about the ideas. (2) The value of technical changes which involve no additional expense on the part of the household may be usefully assessed on the basis of empirical evidence of their affects on physical well-being. (3) Established scientific evidence may indicate the potential benefits Of better nutrition and like improvements but it should not be assumed that this necessarily makes these changes worthwhile when the recommended practices do require added expenditures. (4) In communities where income or prices are . changing extension effort has greater Opportunity ';v. 'n. -. ~. « . n, _ u u I. -. \- u p’l- 111 to perform valuable services for rural house- holds than where they are constant, other things being equal and income distribution considerations being neglected. Changes in Community Organizations Most of the problems associated with valuing changes in the Operations of rural households are also associated with valuing changes in community organizations. In addition, the problem Of getting some sort of community preference function arises. (This was ignored in the case Of households because, even though a household is made up of several individuals, it usually exhibits some obvious preference pattern either through authority or consensus.) Thus, it is necessary to acknowledge that the possibility of applying any formal structure similar to that used in the case Of the farm firm is remote, and a less direct means must be devised. If a community organization produced a particular bundle of services which either could be acquired by other means or actually saves members of the community time, effort, or some monetary outlay, these related benchmarks can be used as prices. Then, it is simply a matter of aggregating returns and deducting costs to get an estimate of net benefit flows. To the extent that the community activity in question can be regarded as a result of extension effort these net benefits would be a measure Of the value of this effort. Again one would ask the question, 112 "When would this change have occurred in the absence Of extension influence?" However, as noted earlier, the effects of extension efforts directed at establishing or assisting community organizations may not be readily detected. Most of it may be related to facilitating the collective decision process and may not directly influence any specific technical change. While the possibilities for worthwhile work in this area may be quite great, little is known about how to do it let alone how to determine what changes are inspired by specific efforts. Managerial Changes As managerial changes may Often be significant consequences of extension work, it is necessary to attach some value to them if one is to reduce all of the outputs attributable to specific sets of extension activities to the chosen common denominator. In terms of the framework of analysis used in this study it seems that managerial changes, whatever their nature, can only be valuable if they eventually influence, either directly or indirectly, technical changes in farm units, rural households, or rural O O O 13 I 0 community organizations. This obtains because returns 13For the sake of completeness two other possibil- ities should also be mentioned. First, it may be that extension program participants get some measure of satis- faction (increase in welfare) out of participating in extension programs or out of simply possessing new knowledge. Such a phenomenon might be regarded as a result Of managerial changes but, perhaps more appropriately, it could be described ~- ~- 113 to owned resources in farm firms (and utility in households) can only be increased by changing the levels of the inputs (or goods) used14 and the ways in which these are employed. Direct influence can be said to arise if, as a result of a specific managerial change, different decisions about technical matters are reached than would have been otherwise. Indirect influence can be said to arise if the managerial change in question gives rise to further managerial changes which in turn influence decisions on technical matters. It is readily apparent that there is a very great potential for managerial changes. Exponential chain reactions as the consumption component of extension work. As extension legislation and policy statements typically pay little or no attention to this as a valuable output Of their work it will be ignored in the discussion Ofevaluing managerial changes. (Of course, this phenomenon is Often a quite important consideration in choosing particular extension programs to achieve other ends because some methods may be more readily accepted than others because of their consumption component.) Second, in some instances, agric- ultural extension may, by supplying information to rural society, confirm in the minds of farm peOple that their existing ways of doing things are the best available. This would possibly produce intangible benefits (non-market goods) in the form of increased assurance, confidence, and/or satisfaction for the farm peOple involved. It might also produce tangible benefits in the form Of reduced search for better ways if, in fact, this is an unprofitable activity. However, the intangible benefits appear to be essentially the same as the consumption component of extension work and will be ignored for the same reasons. The possibility of the situation existing where farmers are spending too much time searching for better ways of doing things also appears remote to this writer or, at least, self-correcting and will be ignored. 14The usual assumption that different qualities of a good are regarded as different goods is made throughout. .v 114 are obviously possible. Furthermore, even if these do not occur, a single managerial change may influence many technical decisions. For example, a soils short course might change a farmer's image of how plants acquire nutrients from the soil and thereby influence many decisions in the area of fertilization and soil care. Basic information on nutrition might influence many aspects Of food preparation and use. A farm management program might inspire people to take courses in technical agriculture which could eventually influence decisions in many diverse areas. While this Potential is great, it is also extremely difficult to analyze and it can be easily under or over valued. TO be consistent with the conceptual framework already developed one would have to say that the value Of a managerial change can be Obtained by taking the present worth of all the technical changes it influences. Such a statement has little practical meaning until ways are deViSed for detecting the technical changes which specific managerial changes actually do affect. In general these are not available. One would intuitively expect increased baSic knowledge in an area to influence technical changes in that area (e.g., a soils course would influence fertilizer decisions). However, by-product effects are Commn because of man's ability to imagine or extend his knOwledge from one area to another on a hypothetical basis. For example, hybrid sorghum was more quickly accepted than hYbl‘id corn in the U.S.A. because farmers were familiar 115 with hybridization through their use Of hybrid corn. It is even difficult to tell when a managerial change occurs. Two people attending the same extension meeting may be affected differently and may make different decisions about similar technical matters because they assimilated the same messages into their image of reality in different ways. Also, it is significant to note that, just because a managerial change occurs, its value does not necessarily have to be positive.15 A farmer may, because Of some new information, make an unwise decision he would not have other- wise, even if his information is correct. For example, if a farmer learns that variety K outyields Old variety M by 10 Per cent in field trials, he may adopt variety K on this baSiS. Whether this is wise or not depends upon whether K aetually does outperform M for him. It may not for various 16 reasons. If it doesn't then the new, correct information has produced an unwise decision. Although it seems . 15This Observation is relevant to the previous , d{Soussion on the value Of technical changes. As all tech- n$05.11 changes must have some managerial change associated With them, at least in the form of a new Observation if nothing else, there may be some question as to whether it 1? Possible to have worthwhile technical changes without Simultaneously incurring worthwhile managerial changes. However, it seems likely that many managerial changes associated with specific technical changes are Of little or no Value beyond that, although the total value of all of the managerial changes associated with many technical changes may be significantly greater than the value of all specifically- aSeociated technical changes because they may interact to lead to a variety of indirectly-associated technical changes. 16The information is correct if variety K did actually outperform variety M by 10% in field trials. The inference made by the farmer is incorrect if K does not outperform M on his farm. 116 generally-agreed that better-informed people make better éhcisions, this example indicates that the Old adage "some kmowledge is a dangerous thing" has empirical support. One managerial change which extension workers Often gnomote is increased managerial effort - "consider the farm msa.business" it is sometimes said. This may be a valuable change if the added effort required to do so Iuoduces results which outweigh the added costs associated wiulthis effort. Lost time in production activities plus (xmts of accounting services and so on may be greater than t1Kéreturns they produce. Thus, it is not always wise for fluafarmer to increase the resources he devotes to management. Another point which seems Of relevance in valuing managerial changes involves the time dimension. If a managerial change occurs at time t but is not useful (i.e., ckms not influence any decision on technical matters) until thM!t+k, the cost incurred in making it could have been lower by delaying it until time t+k and, there would have bmfllno loss in the associated benefits.17 Thus, other t1lingsibeing equal, managerial changes which produce technical Cl”hinges immediately will be preferred to those with delayed effects. On the basis Of this discussion three rules—Of-thumb 17This assumes that costs are calculated on the basis ofpositive interest rates and that there are no other fore- gone returns . 117 nay be established: (a) Managerial changes which cannot be directly linked to Specific and potentially-valuable technical changes should be discounted heavily because they may have zero value. (b) Managerial changes which will not likely lead to technical changes in the near future should be discounted as they could have been post- poned without loss Of benefits. (c) Managerial changes will have a greater value if they occur in a decision-maker with a long career Of management ahead Of him (or her) than in a person nearing the end Of his (or her) career. Bladdition, two hypotheses suggest themselves: (a) It is advisable for extension services to promote related managerial and technical changes together rather than in separate programs. (b) The benefits of technical changes are likely to be more obvious to program participants than those which may accrue to managerial changes. Aggregating Benefits Across Decision-Making Units Problems in aggregating benefits across individual dafision-making units can arise for two reasons. Firstly, anus may be considerable difficulty in arriving at a common dmxmdnator among units. This will occur where the benefits OfChanges in individual units have been measured in non- m<”mtary terms. Secondly, there may be substantial external effects for which an allowance must be made. Externalities al’ise'mainly through (a) the diffusion phenomenon, (b) changes i“the supply and demand curves for market goods and services, and(c) managerial changes. 118 The common denominator problem is evident in Several areas. Changes in utility in rural households, even if they are generally-regarded as positive, cannot be 3readily aggregated in terms Of utility because of lack of a.method of making valid inter-personal utility comparisons. Ann, even if they could, they would eventually have to be ‘translated into monetary terms in order to be aggregated ‘Vith other extension outputs. One way Of getting around ‘these problems is to assume that a given change is equally \valuable to all households and, by valuing it at the cost (bf providing an equivalent level Of utility as suggested Gearlier, a monetary aggregation can be made. However, as Vvill be seen later, there are circumstances in which this Inethod breaks down. The same reasoning applies to common <3enominator problems arising in other areas such as manager— :ial changes or changes in community organizations. In the sections of this chapter devoted to valuing (:hanges in individual farm firms, rural households, and <:ommunity organizations it was assumed that there were no saxternal effects associated with the change being valued. VVhere this assumption is correct, the total benefit flow Eattributable to changes inspired by extension effort can be <3alculated by summing the benefit flows accruing to the individual decision-making units directly influenced by extension work. ‘Where it is not, some adjustment for externalities must be made. 119 The diffusion Of ideas from decision-making units Which participate in extension programs to others which do not is essentially an externality. Representing extension outputs as shifts in adoption distributions wherever appropriate, permits explicitly accounting for this exter- nality. Thus, the appropriate method for deriving an aggregate benefit flow for the promotion Of a specific innovation requires aggregating over all decision-making Indts in the area described by a new adoption distribution - Imw.just over program participants. Providing no other. anernal effects are present and measurements Of the value cfi'the change for individual units can be made, the empropriate aggregation method is simply the summation Of efll benefit flows for all Of those individual units Specified by the shift in the adoption distribution. There also exists the possibility that, under the extension- influenced adoption pattern, managerial costs incurred by decision-making units which do not participate in extension programs will be different than they would have been under the alternative adoption pattern. Such cost differences can be accounted for by adjusting benefits accruing to these units in the same way as for adopters which participate in extension programs. Changes in supply and/or demand curves for market goods and services may result when an extension agency successfully promotes change(s) in groups Of decision-making 120 intits. These changes have implications for this study both.from.the point of view Of who receives the benefits from extension effort and how these benefits should be evaluated. The movements which may occur in the supply curves for farm products are illustrated in Figure 6. In period zero, the actual supply curve is given by SA(0) but, as an innovation spreads through the group Of adopters, it tends to move downward and towards the right ending up at SAmn in the period m. If agricultural producers do not innuease total profits from period 0 to period m, then the total per period benefits attributable to the innovation SA(O) D Snp(0) SA(m) Snp (m) h--- QA(0) QA(m) Quantity Figure 6.--Illustration of the effects of a shift in the supply curve resulting from the adoption in an innovation. a." 'e 121 after it is adopted by all producers will be given by the increased consumer's surplus18 area PA(0) a c PA(m). However, if producer's surplus is larger19 in period m than in period 0, then the total per period benefits would include both the change in consumer's surplus [area PA(0) a c PQ(m)] and the change in producer's surplus which is given lnrarea PA(m) c d PNP(m) less area PA(0) a e P (m) where NP Supply curves SNP(0) and SNP(0) represent the supply curves Vfluch would exist if all factors in the agricultural sector Were being paid their Opportunity cost. Thus, the shaded auea PNP(0) e c d PNP(m) would represent the total per Emriod benefit after period m. Of course, where the group cm decision-makers influenced by the change do not constitute a significantly large share Of the market, the shift in the supply curve can be ignored and it can be assumed that the innovation does not directly influence price. In this case, total per period benefits whenever the 18This analysis assumes that the demand curve for the product remains unchanged while the supply curve shifts, the usual net result being that larger quantities will be sold at lower real prices. The gain in consumer's surplus shown in Figure 6 assumes constant marginal utility of money. [For an explanation Of the implications Of this assumption, see William S. Vickery, Microstatics, (New York: Harcourt, Brace & World, Inc., 1964), pp. 66567.] Where markets for products approach perfect competition as is often the case in agriculture, the shift in the supply curve will automatically occur as more and more farmers become willing and able to produce at lower real prices. This, of course, applies only if there are no artificial barriers such as price or output controls in the product market. 19Presumably it would not decline although this is not an impossibility. 122 iJuiovation is fully adopted will be given by the change in producer's surplus, whatever that may be. Figure 7 shows the effects of a shift in demand for farm inputs or consumer's goods resulting from a widely- adopted innovation in the agricultural sector. A shift in demand for farm inputs could result from new uses of old inputs or from complementarity in production between new inputs and other old inputs for which new uses have been found. A shift in demand for consumer's goods could result from innovations in household Operations (e.g., the introd— uction of sewing machines might increase the demand for Cloth) or increased incomes from production innovations. 13(0) D(m) Price PA(m) ————————————— . PA(O)'—_——————-—- l A(0) QA(m) Quantity IO Figure 7.--I11ustration of the effects of a shift in the demand for farm inputs or consumer's goods by the farm sector. 123 JKlthough the demand shift shown is towards the right, this need not be the case in reality. The introduction of tractors in North America completely obliterated the derived demand curves for many inputs complementary to horses. Furthermore, in some cases entirely new demand (waves will arise and, for capital items, these may be Immm greater in early periods than later when stocks have been4accumulated. If the shift in demand corresponds to that shown in Figure 7, the benefits accruing to the Supplying sectors will be given by area PA(m) a b PA(O) . The change in producer's surplus accruing to farmers (or eappropriate and, if prices are unrelated to the adoption 0ftme innovation in question, this expression could be Qumidered as a useful approximation Of the total benefits aCcruing from the change. There is also the possibility tZhat extension effort, while creating a new adoption dis- tiribution, actually changes the order of adopters signif- 1icantly. Where this is the case the relationship between )Nt) and A(t) would differ between AE(t) and A0(t). How- eVer, as this is difficult to ascertain in practice and Fmobably would not influence the results substantially, it 21Rogers, op. cit. 129 will be ignored here. Thus, it would seem advisable to use equation (6) to calculate B(t) in this study. A rather significant problem arises in that the nature Of f {A(t)}is not readily determined. An approx- imation Of it can be constructed if the volume Of production cu'input use is related to A(t) and benefits per adopting unit are directly related to either of these factors. This is Often approximately the case in practice. For example, a new fertilizer may represent a saving of a constant amount per acre or per bushel over a limited range of 22 production. If such an intermediate measure can be found, it is possible to express B(t) as: (8) B(t) = [AE(t) - g {AE(t)} - A0(t) - g {A0(t)}] V(t) where V(t) is the benefit per unit Of the intermediate measure and g {A(t)} gives the number of units Of this measure for each period. Consider the following rather simple example. Suppose A0(t) is given by: (9) A0(t) = 0 t<5 = SOt-ZSO SStSI7 = 600 17 undone mo muses ssucoasm saves m m m e . sine as any a u .mmsus>suom GOsmswuxm no you smosumnuomas m Eosm mGHuHSmms BOsm usmmcmn on» no COHuMHSOsoo on» NO msmeoxmuu.v msnme 132 $84000 V(t). Second, a constant V(t) would have made a significant difference in B(t), especially in later periods. The benefits shown accrue to the agricultural sector but, in this example, late adopters receive lower total benefits than early adopters for two reasons - their production units are smaller and their benefits do not start as soon. For example, the first adOpter has already received $720 in benefits before the last adOpter adOpts.23 At an interest rate of 5 per cent, this would make the first adopter better off by $941.23 in period ten. Furthermore, in period ten the first adopter receives $60 whereas the last adopting 24 unit receives only $24. If there had been a product price decline and/or input price rise as a result of supply and/or demand curve shifts due to this innovation, the late adOpter would have been affected in much the same way as by the exogenous decline in V(t) used in the example. While this example is somewhat naive in several ways, it is believed that it illustrates the essential features of 23Total benefits to the first adOpter for the first nine periods are given by: 9 2 200 (.5 - .02t)= $720. t=l 24 Total "intermediate" units in the extension- influenced adOption as a function Of time are given by: 1 _ 1 2 AE(t) [200 -‘10 AE(t)] - 200 AE(t) - 10 [AE(t)] Thus, the average number Of intermediate units produced by the last adOpting group is given by: 1 2 a 200 AE(t) — I‘O‘ [AE(t)] = dt 200 - 0.2 AE(t) = 80. 133 the phenomenon under consideration. Further realism could be achieved by using an S-shaped adoption distribution and a more SOphisticated model of the firm to calculate V(t). However, in any given situation the determination of the empirical nature Of these structures would likely be quite costly. Thus, models similar to this example will be used in this study. Secondary Extension Outputs The secondary extension outputs defined in Chapter II defy easy evaluation. Thus, they too can be easily under or over valued. However, some general ideas do suggest themselves and they will be mentioned here. In the subsequent empirical analysis, the worth of secondary extension outputs will not be reduced to monetary form directly. The feedback phenomenon described earlier as "influencing technology develOpers and/or information sources outside Of rural society" will be valueless unless specific actions are taken by research agencies or other information sources to change their activities. If such actions are taken, the value Of the feedback effect would be evident in improved effectiveness of these organizations. This is not easily measured but presumably it can, over time, be detected. As the estimation Of benefit flows attributable to research requires at least as much attention as this study gives to extension, formal measurement of the extent 134 0ftme benefit flow due to the feedback effect will have to he ignored. However, some informal attention will be paid to the extent of extension-research interaction. The influence Of extension on other information sources and other institutions will be handled in the same way. When agricultural extension services function as training agencies in order to bring young men and women (and rmmcomers to an area) up to the technological and managerial levels already attained by existing decision-makers, they may perform a valuable service.25 The value of this service may be most readily estimated by calculating the cost Of performing it by some other means. Presumably, if it were not done by the extension service it would be done by the young peOple's parents or by some other agency. When tech- nological and managerial skills become complicated to the extent that they have to be learned, an investment in learning is always necessary. This naturally involves costs in terms of the time Of the learner and Of the teacher. There are Often economies Of scale and specialization to be gained by establishing some sort of formal training program to handle this function rather than letting it occur naturally. Agricultural extension may produce benefits by realizing some of these economies. It would intuitively appear that such economies would increase as the state of the arts progresses and, therefore, possible gains would be larger in developed than in underdeveloped societies (i.e., 25Raising individuals beyond these levels would come under the heading of "managerial change". 135 the level of knowledge in developed societies requires more learning). It would appear that the gains tO be realized here are relatively small, particularly when one considers that other types of organization (such as regular schooling combined with in-the-home-training) can also realize similar economies. When agricultural extension performs the function of directing rural people to jobs outside Of the agric- ultural sector it may produce benefits which are spread among many groups and, as a result, are not readily measureable. In economic terms one could view the result as being a lowering Of the wage differentials between agric- ulture and non-agricultural employment. Such differentials presumably arise because of differences in the quality of labor, lack Of information, and equalizing differences which account for differences in non-monetary benefits associated with the different types of employment (e.g., different types of degrees of risk, type Of environment, etc.). An analysis Of the extent to which extension effort actually does improve information about alternatives elsewhere relative to those in farming and/or improves the quality of labor together with some measures of wage differentials and the number Of people effected could lead to a value measure of this secondary extension output. Needless to say, it would be tOO costly to do here. Furthermore, the value of this output is probably significant only at a much more aggregate 136 level than is involved in this study. Thus, where this function is noted in the subsequent empirical analysis, no measurements of the benefits in monetary terms will be attempted. CHAPTER V MEASURING EXTENSION PERFORMANCE There are several factors other than extension "outputs" which must be considered in measuring overall extension performance or effectiveness. Extension program costs must be measured and compared to measures of extension "outputs" to assess the "net" value of the "production” component of extension work. TO usefully make such a comparison it is necessary to have some decision procedure for choosing among alternative extension programs. Further- more, tO assess the overall performance of an extension agency it is also necessary to consider "managerial" 1' Measure- efficiency as well as "production" efficiency. ment Of both "production" and "managerial" efficiency is the concern of this chapter. The first four sections of the chapter deal primarily with the former by analyzing the problem of selecting acceptable procedures for choosing the best set of extension programs that can be Operated on a fixed budget. 1It makes little sense, for example, to say that an extension agency is necessarily doing a good job if it Operates one very successful program but spends 80 per cent of its budget planning and selecting that program. The agency might have been better off to have Operated more pro- grams and done less planning. Thus, these possibilities also have to be considered in evaluating overall performance. 137 138 The procedure which is selected is Operational in that it can be applied, as is demonstrated later in this study. It is not recommended for formal application by extension agencies because it is probably too costly to implement on a formal basis (i.e., it appears to be an inefficient managerial practice). However, it is believed that it would be useful for extension decision-makers to apply this gmocedure on an informal or intuitive basis. The fifth section of the chapter deals primarily vuth ways of assessing "managerial" efficiency and the (maracteristics of the trade-Offs between "management" and "production" when the total budget is fixed. The result Of this discussion is somewhat less satisfying than that directed at "production" efficiency. The general conclusion is that theory, methodology, and empirical knowledge in this area are probably insufficient to enable one to construct a fiumal analytical structure that would be superior to the “Hes of thumb and intuitive procedures in common use. The last section Of the chapter discusses more traditional measures of extension performance and relates them to the measures used in this study. A Decision Rule for Choosing Among Alternative Extension Programs As most extension programs probably produce some FWOducts which can be readily reduced to monetary terms and (”fibers which can not, and different extension programs Often 139 have different income distribution effects as well, it might appear that the only way programs can ultimately be compared is to leave the decision entirely to the subjective judgement of extension decision-makers. However, this only avoids the basic problem which is one Of objectively choosing among a variety of possible programs. If no sat- isfactory method is available to anyone, extension decision- makers may be just as badly Off as everybody else - perhaps even worse because they will likely substitute their own ideas and values for the apprOpriate but unknown concepts, even though these ideas and values may be quite inaccurate. In view Of this situation, it would appear useful to try to select a single rule or technique for comparing extension Programs on the basis Of monetary costs and benefits with the hOpe that most, if not all, extension outputs can eventually be placed on this basis.2 If a rule for Comparing extension programs on the basis Of monetary costs and returns can be selected, and extension "outputs" can aCtually be measured in such terms, then only the problem of incorporating income redistribution effects into the \- 2As noted elsewhere in this study, the principle Of OPportunity cost often provides an excellent means of reducing seemingly unmeasureable extension "outputs" to a monetary basis. Furthermore, those individuals who say S11Ch reductions are impossible and proceed on some other basis are, indirectly, achieving the very reduction they Say is impossible. For example, the extension administrator who maintains that a 4-H Club program (which achieves Certain "intangible" benefits) is superior to extending a new wheat variety whose benefits are more easily measured dlrectly in monetary terms, is implicitly placing a minimum monetary value on the 4-H Club benefits. 140 decision procedure remains in order to objectively choose among such programs.3 Several criteria suggest themselves for selecting an apprOpriate decision procedure. Firstly, it should agree vdth the marginal productivity rules supplied by the theory of the firm should the situation arise where they cmuld both be applied. Otherwise, it would be inconsistent vfith a well-developed and widely-accepted body Of knowledge cmncerning economic decisions. Secondly, it should be designed to choose among a limited number of discrete alternative programs because this appears to be the sort cf choice which typically faces extension agencies.4 Thirdly, it should be able to account for the fact that Emoject benefits are Often not realized in the same period in which costs are incurred. Fourthly, it should be able to by-pass the requirement Of physically quantifiable inputs, Cmtputs and functional production relationships as these are, in general, unknown. Fifthly, because Of the difficulties ‘1 3This ignores the problems created by imperfect knowledge (i.e., estimates). These are dealt with in the Second section of this chapter. 4If a method for choosing among a number of discrete EHOgrams were available, many possible alterations in program esign could, technically-speaking, be evaluated as separate EWOgrams. Thus, the discreteness Of the programs is perhaps Tore important from a theoretical point of view than the limited number" condition. The latter arises from the {Hactical necessity of efficient "management". 141 usually encountered in measuring social costs,5 it would be convenient if the chosen technique could limit the extent tx>which these must be estimated. And finally, as suggested at the outset Of this study, it should be suitable for selecting the best group Of programs to be Operated with a fixed amount Of funds. This final criterion is adopted because upper limits cm extension funds are Often beyond the control of extension decision-makers. In the long-run extension agencies can be expected to influence the size of their budget, but, in the short-run, it appears that they are, at best, concerned with (a) designing a set of promising, feasible programs, and (b) operating the most effective subset Of these smograms that existing resources will permit. This is only reasonable because, as also noted at the outset of this dissertation, the effectiveness of both alternative and cmmplementary programs must be analyzed in order to object- ively decide upon extension budget allotments. Extension agency personnel cannot be expected to perform this analysis. The decision rule which meets the above criteria 5As funds for extension work are almost always {movided through taxation, and most taxes are welfare- reducing, extension expenditures would have to be adjusted flu this welfare loss if social costs were to be estimated. (\more relevant cost estimate might be the marginal return miresources in other public uses. Although this is seldom readily available in even approximate form, its use will be <flscussed later in this chapter to illustrate the basis for the decision procedure actually selected. 142 can be stated as follows:6 (1) Rank all projects on the basis of the numerical value Of the ratio of PVB to PVC for each project where 3(0) 3(1) 13(2) B(n) - PVB = ————— + . + ——————- + ... + —————— for all B(j)>O 1 (1+1) (1+1)2 (1+1)n B(0) 3(1) B(2) B(n) - PVC = + + + ... + ——————-for all B(j)0 where x* is a finite constant, then x* is said to be the probability limit of the sequence {x(n)}. 148 several multiplicative calculations, the ultimate distrib- ution Of PVB/PVC is likely to be difficult to judge unless some sort of sensitivity or simulation analysis is conducted. Unfortunately, a complete simulation is usually a complex and costly job. McKean says, ”On balance, the Monte Carlo method and the probability Of outcomes that this technique produces, may not be worthwhile in cost-benefit analysis, except in the exhaustive examination of very expensive 11 This is certainly the case if a large number prOposals." of underlying basic variables are involved and the means and variances Of these are estimated from special samples. However, if the estimates of the underlying variables are arrived at in some other fashion, as is Often the case, then the variances of these variables may be similarly Obtained. This would apply particularly where engineering and/or biological data are used. Thus, the additional cost Of the analysis is usually incurred through programming and computer time. These costs vary largely with the mathematical nature of the phenomenon being examined. Sim- ulation is definitely out of the question if a computer is not available but a sensitivity analysis can be conducted without such facilities. Thus, it is this latter approach ‘which will be used in this study. Once one has available either mathematically- calculated or simulated probability distributions or sets of 11Roland N. McKean, Efficiengy in Government throu h Systems Analysis, New York: John Wiley & Sons, 1958, p. '75. 149 sensitivity analyses, the problem Of the apprOpriate decision rule once again presents itself. A number of rules are available. A pOpular text book example is the "minimax" criterion. This requires one to minimize the maximum loss. Thus, instead of choosing projects with high expected values one would choose those which, even if things turned out rather badly, would produce very low losses. Other more SOphisticated procedures suggest that one maximize utility which is considered to be a function of both the set of possible outcomes and their set of probabilities of occurrence . (2) v = f({El,...,En} {p(El),...,p(En)}) where V = utility Ei = measure Of interest for event i: i=l,...,n p(Ei) = the probability of Bi occurring. Most Of the previously stated limitations of utility analysis (see Chapter IV) apply to the attempted implementation of this suggestion. Thus, the problem Of an apprOpriate decision rule for the case Of uncertain outcomes appears unresolved. Therefore, this study will not attempt to select a decision rule but instead, will attempt to show how different programs might or might not be chosen depending upon what decision rule is used. 150 Selection Of Extension Programs on a Minimum Cost Basis In some cases it is not necessary to consider the PVB/PVC ratio explicitly in order to make choices among different extension programs. This situation arises when several different ways Of achieving essentially the same results are available. In such cases, extension agencies can make the most effective use Of funds by selecting the program which achieves these results at minimum cost. To apply the "minimum-cost" criterion one must establish (a) what are "essentially the same results" and (b) what are the relevant costs to be minimized. One can immediately see that the PVB/PVC ratio is implicit in the "minimum-cost" criterion. "Essentially the same results" will be achieved by different programs if and only if these programs produce the same PVB value. Obviously, the same PVB value will result from any given set Of extension "outputs" with the same changes in adoption distribution(s). Thus, if several extension programs are capable Of producing a given set Of extension "outputs" with the same changes in adOption distribution(s), PVB will automatically be the same in each case and the one with the greatest PVB/PVC value will always be the one with the lowest value of PVC. Thus, the calculation of PVB can be by-passed. For example, if it is desired to change the adOption distribution of a new wheat variety in a given area, the 151 same change might be achieved by a radio campaign, a circular pamphlet, or by visiting a sample of farmers. As has been illustrated previously, the relationship between these various methods and the results Obtained will depend on the total environment in which the extension agency is working. But, should each of these three methods produce the same results in terms of the change in the adoption distribution(s), the extension agency can automatically eliminate all but the least expensive of them. In the case of extension "outputs" which cannot be analyzed in terms Of changes in adOption distribution(s), there does not currently appear to be any possibility of making minimum cost decisions except in terms Of inter- mediate goals which can be reasonably assumed to produce essentially the same final results. For example, if it is desired to provide a group of rural youth with a given 4-H Club program, then there undoubtedly exist several possible ways Of Offering essentially the same program. If the effects on program participants are no different, the least eXpensive Of these should be chosen. These different ways may simply involve different meeting schedules, travelling routes, Office locations, or other similar matters. However, there may be significant cost differences among the alternatives. Another way of looking at the cost minimization problem is to consider it as one of achieving the maximum 152 amount of some selected physical measure12 of output for a given cost. If this approach is to be useful, the value per unit Of the selected measure of output must be independent Of the method and level of producing the output. In other words, the expected value of a unit of output must either be constant or vary with output in the same fashion for all methods Of production. (This does not require output to be homogeneous. However, any heterogeneity must be random and not dependent on the method and level of production.) Thus, in the case of 4—H Club work, if the unit of output is one club member who comples a given 4-H program, the costs per club member would be minimized if the maximum number of club members possible were serviced with a fixed amount Of resources. Of course, if in the process of trying to increase the number of club members serviced with a given amount Of resources the eXpected value of the benefits per member were to change, this rule would not be applicable. The relevant costs to be minimized in applying the "minimum-cost" rule are obviously the same costs that would be included in PVC if the PVB/PVC ratio was being calculated. It was established earlier that extension expenditures could serve as a useful measure Of extension costs for comparisons Of extension programs, although they 12The selected physical measure should reflect only the results of the program which have positive or negative values. Results with zero value can be ignored. 153 do not necessarily represent the costs to society of doing extension work. However, it may often not be necessary to calculate total extension expenditures to minimize costs per unit of “output". As with any other endeavor, only the marginal or variable costs are relevant. In addition, it may be possible to omit cost calculations entirely if it is known by means of physical measures that costs will necessarily be lower with one alternative than another. It might be quite appropriate, for example, to choose among several extension programs on the basis of the amount of extension-worker time involved to produce a given output. However, this choice would rest on the assumption that other costs do not change when time is reduced. Incorporating Benefit Distribution Considerations Into Program Performance Assessment Extension work, like most government programs, produces a distribution of benefits among members of society that may be substantially different from the distribution of the costs incurred in carrying out the program. As this study is primarily concerned with the effectiveness of extension work given a fixed budget, it can readily be assumed that the distribution of the costs of funding extension work is the same for all extension programs which might be operated with the given budget. This obtains because the budget will be provided in precisely the same 154 way no matter what set of programs are operated. Thus, the distribution of such costs can be ignored in comparing programs.13 Differences in benefit distributions among programs become important when the benefits received by some individuals or groups in society are regarded as more valuable than the same benefits received by other individuals or groups. This is particularly significant in view of the fact (established earlier) that the results of some extension programs may produce negative benefits for some groups and positive benefits for others.14 Follow— ing this line of reasoning, it would seem that, in order to adequately deal with benefit distribution considerations in choosing among programs, it is necessary to know (a) the distributions of benefits from alternative programs among different individuals or groups, and (b) the appropriate weights to attach to the benefits received by each of such 13Cost distributions cannot be ignored when the method of obtaining funds or the size of the total government budget is the problem under consideration. 14It is assumed here that extension decision-makers are willing to operate non-Pareto-better programs and that both the public representatives who hire them and the society which elects and/or sanctions these public repres- entatives eXpects them to do so. It would appear that there is substantial accumulated evidence supporting this assumption. Extension decision-makers have made non-Pareto- better choices in the past and, while they may not have been aware they were doing so in some cases, it seems unlikely that they were totally ignorant of the nature of these choices. 155 individuals or groups.15 In determining (a) it seems essential to have some acceptable way of classifying members of society because the description of benefit distributions by individual is not likely to be practical. There are a number of different classifications which might be proposed for this purpose. However, it will be noted that the choice of relevant groups requires that something be known about the approp— riate weights because it would be pointless to work with such groups unless the weights were approximately the same for all individuals in each group. It also requires that, for each instance of concern, benefits be distributed approximately equally among all individuals in each group. Otherwise, such an analysis could be quite inconsistent with a similar analysis conducted on an individual basis. (For example, a program giving low income families an average annual income increase of Sx/family might be viewed in 15An alternative to this procedure would be to attempt to utilize Lorenz curves to construct an index of income distribution. This index would then be employed as a variable to be traded-off against the PVB/PVC ratio in the usual indifference curve analysis. However, such a procedure is cumbersome and is fraught with the same difficulties as the procedures suggested above. Even the assumption of income equality being prefereable to inequality does not solve these because of the difficulty in comparing crossing Lorenz curves. Gini ratios have been used to circumvent this difficulty [see James Morgan, "The Anatomy of Income Distribution", Review of Economics and Statistics, Vol. 44 (August, 1962)]. However, this can only be done because they contain theoimplicit assumption that all parts of the area under the 45 line in the usual Lorenz curve diagram should be weighted according to their size. This appears equivalent to the more direct approach suggested here. 156 quite a different light if the 5x were equally distributed among such families than if it were all given to one family in the group). The only way of classifying members of society which appears to have wide acceptance and appears close to being consistent with the above requirements is a 16 If benefit classification based on current income. recipients are divided into k classes on the basis of current income, the PVB for all projects can be recalculated as follows: k (3) PVB = i§1 L.PVB. OsL.sl 1 1 1 where L1 is the weight attached to the PVB received by the ith group (i.e., PVBi). Thus, the PVB value is adjusted for income distribution considerations.17 PVC values need not be so adjusted if they include only the funds expended 16This is by no means a completely adequate class- ification for the problem at hand. It overlooks the fact that some people may need higher income levels than others to achieve what is generally regarded as the same welfare (e.g., one person might have very high medical expenses while another has none). Furthermore, it does not completely incorporate net worth (i.e., income earning potential) although income is likely closely related to net worth. Thus, the weights one might intuitively attach to benefits received by a given class may not be the same for all members of the class. 17This analysis, like most formulations of the trad- itional trade-off between "equality" and "efficiency", hides a crucial feature of the basic problem at hand. The measure of efficiency which has been used in this study (i.e., the PVB/PVC ratio) is, in part, determined by the prices used in its calculation. These prices are partially determined by the existing distribution of income. Thus, the apparent trade-off is not being made between two distinct and independent alternatives for, if the existing income dist- ribution were different, prices could be different and the PVB would be affected accordingly. No attempt is made in this study to allow for such interactions. 157 to finance the project. However, if they included losses to some members of society as a result of implementing the project, this portion would require adjustment. Of course, the obvious problem is to determine Li‘ Presumably L is some decreasing function of current income. It can be presumed to be unity for incomes of zero and to decline to zero as incomes become very large. However, its rate of decline depends on social preferences. Thus, the determination of L1 is very similar to the problem of deter- mining a social welfare function. This has been investigated by economists on prior occasions without significant break— throughs for practical purposes.18 Nevertheless, there are some things to be gained from this analysis. First, it is obvious that programs with (unadjusted) PVB/PVC ratios less than unity may be preferred to programs with (unadjusted) PVB/PVC ratios greater than unity, once the Li have been allowed for. (Of course, if this occurs the resultant ratios for both projects would be less than unity.) Secondly, the incorporation of L1 in the PVB/PVC ratio may, in some cases, leave program rankings entirely unchanged. Thirdly, the chance of program rankings being changed by benefit distribution considerations will likely be greater the wider the difference in current incomes among benefit recipients in different programs. Because of the nature of the results of extension work it may be that a single extension program will produce 18See, for example, Kenneth Arrow, Social Choice and Individual Values, (Cowles Commission, Monograph No. 17; Wiley, 1951). 158 a benefit distribution which varies over time. For example, if an extension program was directed at promoting the adoption of a new potato variety among a group of commercial potato producers which produced most of the supply for a given market area, and extension workers succeeded in getting this variety widely adopted among producers over a four-year period, the first-year benefit distribution might be highly skewed towards high—income individuals whereas, if virtually everyone consumed potatoes, the fourth year's benefit distribution might be almost even among all income classes. In comparing extension programs the calculation of a value of PVB for each income class automatically takes such differences in benefit distributions among programs over time into account. Measuring Managerial Performance of Extension Agencies Extension agencies are faced with the problem of having to decide how much of their fixed budget they should devote to managerial activities in view of the fact that any increase in management's share, while it may increase the effectiveness of extension programs, must surely reduce the funds available for operating such programs. They also face the closely related problem of making management most effective with a given amount of resources. The nature of the relationship between these two problems will 159 . . 19 become clear as the d1scuss1on proceeds. Analysis of decisions regarding how much of the fixed budget should be devoted to management is the first tOpic of consideration. Assuming that all programs or projects have the same distribution of benefit flows among beneficiaries and that these benefit flows are measureable in monetary terms, then the relevant measure of the per- formance of an extension agency is the PVB/PVC ratio calculated on the total extension budget. This can be designated (PVB/PVC)T and is given by: PVB. J (4) (PVB/PVC)T = for k programs '0 < O u. + 3 ...: LJ. U. IIM W "M W H where M is the present value of the total expenditure on those managerial activities required to initiate and implement the k programs and PVBj and PVCj represent the present value of benefits and costs respectively associated with the jth program. Now, if all programs possess only a single cost period and all management activities associated with those programs occur in that same cost period, then 19Another function which extension managers must perform is that of supplying their superiors with infor- mation which accurately describes extension effectiveness and enables them (i.e., their superiors) to determine the extension budget simultaneously with those of competing and complementary programs. (It would seem that such infor- mation should include information on the interrelationships among extension and such other programs.) 160 the total budget for that period is: IIM W IIM w (5) C(0) = PVC. + M = 1 J j Cj(0) + M(O) j 1 It can be readily seen that, for a given budget, the larger the amount allocated to M(O), the smaller the amount available to operate programs. However, if management activities are increased for a given amount of funds devoted to programs, one would intuitively expect that program performance would normally improve - at least, up to some point. Thus, extension managers face a paradox. They can usually improve the returns to a given expenditure on programs by increasing managerial activities but, when they do, they reduce funds available for carrying out programs. It is therefore necessary to find a means of determining the appropriate compromise. To measure program performance, the ratio (PVB/PVC)p is defined as follows: k k .glpvs. j§1PVBj (6) (PVB/Wm]? = H. = c(oI-MPVC. for all possible Programs. 3 J 162 Thus, management can determine the Optimal proportion of the budget to be devoted to management activities by setting:21 d (PVB/PVC)T (11) = o d [M(0)/C(o)] Unfortunately there does not appear to be much one can say about the Optimal level of M(O)/C(O) unless one knows the nature of the functions "f" and "g". Table 5 shows the relationship between several of the key variables for a budget of $100,000. From this table it can be noted that managerial expenditure per dollar of expenditure on programs increases exponentially as the ratio of managerial expenditure to total expenditure increases linearly. This would tend to offset the diminish- ing returns to management expenditure one would expect to find in equation (7). Another factor which would partially offset the diminishing returns to management would be the declining returns to program expenditures, C(O) - M(O). For example, if (PVB/PVC)T = X when M(O)/C(O) = 0.5, then (PVB/PVC)p = 2X. Now, if (PVB/PVC)T is to increase above X when M(O)/C(O) is increased to 0.6, then (PVB/PVC)p must increase beyond 2.5x as there is now less money devoted to programs. However, at M(O)/C(O) = 0.6, there is $1.50 expended on management per dollar expended on programs as 21This is the first order condition for a maximum. 'The second order condition would also have to obtain. 163 0.H 0 000.00H 000.00H 00.0 0.0 000.0H 000.00 000.00H 00.0 0.0 000.0N 000.00 000.00H mm.m 5.0 000.0m 000.05 000.00H 00.H 0.0 000.00 000.00 000.00H 00.H 0.0 000.00 000.00 000.00H 50.0 «.0 000.00 000.0v 000.00H mv.0 m.0 000.05 000.0m 000.00H 0m.0 «.0 000.00 000.0m 000.00H HH.0 H.0 000.00 000.0H 000.00H 00.0 0 0.0 000.00H 0 0 0 000.00H 0 1002 . 1000\1002 1000\1002 loosilovo loss loco mwzufipcmaxm Emuwoum 0 wow ucmemmcmz musuflpcmmxm musuflpcmmxm pompsm musuwpcmmxm ucmfimmmcmz co pmccmmxm Emumoum ucmEmmmcmz Hmuoe cowuwomoum .000.00Hm mo ummpsn coamcwuxm am How mmHanwm> pmumHmu paw mmusufipcmmxm Emumoum can ucmemmcmE cwm3umn mwnmcowumamw OSBII.0 magma 164 Opposed to only $1.00 when M(O)/C(O) = 0.5 If the marginal rate of returns to management were constant, (PVB/PVC)p would equal 3.0x at M(O)/C(O) = 0.6 but, as marginal returns to management likely diminish, (PVB/PVC)p will probably be somewhat less and it may, or may not, pay to increase M(O)/C(O) to 0.6. As program expenditures are reduced by $10,000 when M(O)/C(O) is increased from 0.5 to 0.6, (PVB/PVC)p would likely also increase for this reason and further Offset the diminishing returns to management. Thus, it appears that the Optimal prOportion Of the budget to be devoted to management depends critically on the nature Of "f" and "g" and one should not rule out any possibility (except perhaps 0.0 and 1.0) without some knowledge of these relationships. Measurement of the effectiveness Of a given amount of resources devoted to management is the second tOpic of consideration in this section. As "f" and "g" are not known in any useful form, a fOrmal quantitative analysis of this problem is impossible. Thus, it is necessary to rely on'a rather informal approach. This involves defining what extension managers must do as they perform the various managerial functions discussed earlier. If such duties do not appear to be being carried out, this would appear to Ina evidence of unsatisfactory performance. However, this approach has one critical weakness. It does not permit hit this stage) considering the costs and availability of Obtaining personnel who would carry out such duties should the situation be encountered where they are not being 165 performed. Detecting problem areas in rural communities requires that extension decision-makers be familiar with the current environmental factors described in Chapter III. They must be familiar both with the values and goals of rural people and the non-normative aspects of local conditions. This has typically been accomplished in the past by keeping extension agents resident in the local communities, and by hiring people with considerable familiarity with local conditions or, at least, with the ability to assess the relative importance of local problems. Presumably, if these avenues are not followed, some alter- native means must be chosen. The lack of any such means appears to be a sign of managerial ineffectiveness. How- ever, a large amount Of time spent on this is also a sign of ineffectiveness in that it will reduce time available for other work. Several methods are available whereby extension agencies may assimilate the sort Of information necessary for making decisions. Universities, experimental stations, other research centers (either local or foreign) and agri- business usually serve as the sources Of new ideas. Extension agencies can keep up with these ideas by hiring people trained in technical agriculture and familiar with 22This could be provided (to some extent) by university courses in the social sciences (sociology: communications, economics, etc. 22 166 these sources and also by having them keep in touch through personal contacts, special co-Operative meetings, and similar ventures. In the U.S., university employers Often have joint research-extension appointments which serve this purpose as well as facilitating the feedback of information on current problems to researchers. The lack of contact Of extension workers with sources of new ideas also appears to be a sign of managerial ineffectiveness (unless such sources are not available). Of course, it is possible for extension workers to spend too much time in such pursuits especially if the new ideas are Of little relevance to the rural people they serve. Information on the relative success or failure of programs is important if extension workers are to learn from experience. As this information must relate to the agency, not just to individuals, it should be recorded for future reference. Cost records must be kept in a fashion which permits the costs of specific programs to be derived and cost ratios or indices useful for projecting future program costs to be calculated. Benefit flows corresponding to the same set of Specific programs should be Observed and measured. Although benefit flows may extend several years beyond program costs, extension workers can Often make some estimates of these if they are trained to do 80. Lack of appreciation of extension costs can readily lead to incorrect decisions and may be regarded as 167 a sign of managerial ineffectiveness. Information on extension methods must be gathered so that extension workers can continue to improve their ability to carry out extension programs. This is Often achieved by short courses or additional training in extension education, communications, or sociology. Lack of such in-service training would appear to indicate managerial ineffectiveness. In addition to detecting problem areas and assimilating information which would suggest solutions to these problems and ways to extend these solutions, extension agencies must be able to evaluate whether new ideas are worthwhile and whether they (the agencies) can successfully promote them among their clients. While research may have provided several solutions to a given problem, one or two of these are likely to be more worthwhile than others and, at the same time, one or two (not necessarily the same ones) may be relatively easy to extend. Distinguishing these is essential if extension workers are to plan good programs. As suggested earlier, planning extension programs is not usually regarded as an Optimization process - it is one Of designing promising, feasible alternatives. Neverthe- less, the more familiar extension peOple are with evaluating changes in rural society and the relationships between extension inputs and extension outputs, the more likely they are to come up with good feasible alternatives. Lack of such familiarity can therefore be regarded as a sign Of 168 managerial ineffectiveness. In order to cut down managerial costs, evaluations are Often made on a rather intuitive basis before programs are completely planned. Following this sort Of evaluation, a set of programs are planned which will exactly exhaust the available budget. TO some extent this practice has merit as it avoids using resources to plan programs which are rejected later. However, to the extent that incomplete information on program costs and benefit flows are available, it may lead to poor evaluations. The function of administering extension programs involves controlling eXpenditures and making the short-run (or tactical) decisions which are necessary to keep programs operating effectively. Modification of programs in Operation, provision of supporting services for personnel, and performance of accounting and communication functions are some of the more common duties of administrators. Where such duties are not being performed, one can reasonably conclude that managerial ineffectiveness exists. The function of responsibility-bearing is Of particular importance in extension work. Each individual extension worker will have his own interests and, if these do not approximately coincide with those of the agency, extension performance may be severely curtailed. This comment takes on particular significance when one realizes that extension workers may often be in a position to do things which further their own self-interest, but contribute 169 nothing to the extension functions defined earlier. The usual method of reconciling employee and agency interests is to use the price mechanism. For this to be effective, employees must have a clear understanding of what agency Objectives are and be willing to attempt to achieve these in return for their salaries. Also, there must exist some mechanism whereby salaries are curtailed if individual performance is not adequate. To avoid the "doctoring" of performance measures by extension workers who typically must make these measurements themselves, this mechanism requires careful design. Lack Of mechanism for the reward of good performance and discouragement Of poor performance can be taken to be a sign of poor managerial performance. The preceding discussion suggests several signs of managerial ineffectiveness in extension agencies. It also suggests the following conclusions: (a) Detecting problems, analyzing possible solutions, and planning programs would appear to be the function of those extension personnel who are in intimate contact with local communities as they can be expected to develOp a better intuitive feeling for such tasks than personnel who are not in such contact. (b) Broadly-trained personnel (i.e., personnel who have some training in technical agriculture and some in the social sciences) appear essential for extension field work. (This is supported by at least one empirical study.)2 However, as noted earlier, consideration of the costs and 23Alexander Grandison Warren, "A Study of Some Training Factors Associated with the Success or Failure of Co-Operative Extension Workers", Unpublished Doctor of Education thesis, Oklahoma State University, August, 1960. 170 availability of personnel who would be effective managers cannot be incorporated in the framework for assessing man- agerial performance at this stage. This will be commented on in Chapter X. Intermediate Measures of Extension Performance Extension agencies have long been interested in evaluating their own work. They have been assisted in this regard by university departments Of extension education, communications, agricultural economics, and rural sociology. Most such evaluations can be grouped into two general . 2 categories: (a) Examination Of an agency's personnel and/or organizational structure in an effort to detect ways Of changing ti so that the agency will produce more effective extension work. (b) Measurements of the specific achievements of an extension program in comparative or absolute terms, on the basis of program impact, number Of changes actually made, number Of peOple influenced or similar variables. These general classes of evaluations appear to be examining what have been referred to in this study as extension man- agement and program Operation activities, respectively. The measures of performance used in such evaluations are more easily Obtained than those used in this study but they do not provide as "complete" an assessment Of performance. 24Ivan Agudelo S., "Evaluacion de Algunos labores de extension en el area de Giradota, Colombia? Unpublished Master's thesis, Interamerican Institute of Agricultural Sc1ences, Turrialba, Costa Rica, 1967. 171 Thus, they are referred to here as "intermediate" measures. The "intermediate" measures used in the evaluation of personnel usually involve measures of personnel qualifications, attitudes, and knowledge about extension methods, Objectives, and so on.25 Criteria for evaluating organizational structure are usually ”rules-Of-thumb" established on the basis of previous experience regarding what will or will not facilitate program planning and Operation. As the earlier discussion of measuring managerial performance has indicated, such procedures are probably more useful than any formal analytical structure available at present. Of course, care needs to be taken in using guidelines developed in one part of the world to evaluate managerial performance in another. Measuring program performance is a different matter. Intermediate measures of program performance may be divided into three categories: indices of work performed by personnel, indices of changes in knowledge on the part of program participants as a result Of extension programs, and indices of change made in technical practices as a result of extension efforts.26 The first of these categories usually includes measures such as numbers of visits, phone calls, demonstrations, meetings, and/or other activities 25An example of this sort of evaluation is a 1962 study entitled, An Anal sis of Five Extension A encies in Colombia prepared By J. Di Franco and R.A. CIIfgordof I.I.C.A., Turrialba, Costa Rica. 26The study of Agudelo (see footnote 37) discusses the first of these two categories briefly and uses both of the last two to evaluate extension work in the Girardota area. 172 carried out in a given period. The second requires the questioning of program participants on specific subject matters in a manner similar to the examinations conducted in formal educational courses. The third incorporates specific information regarding when adOpters were made and why they were made. Such technical criteria can, under certain circumstances, be consistent with the economic criteria already discussed. Where this is so, it is desir- able tO use them because they can usually be measured more PVB PVC However, these intermediate criteria can, at times, cheaply than ratios. conflict with the economic criteria develOped in this study and, where this occurs, reliance on intermediate criteria can result in the choice of inefficient extension programs. For example, indices of extension work performed do not necessarily have any positive relationship to changes in rural society, let alone to §%%-ratios. Indeed, they may have a negative relationship as there may arise a tendency for field workers to concentrate on the activities being measured as Opposed to others which are not measured. This could be expected to result in a decline in product- ivity. Also, as pointed out earlier, changes in knowledge (managerial changes) are not necessarily valuable. Thus, examinations on subject matter may not measure valuable results. Independent evidence would be required to assess whether or not the knowledge gained was useful. 173 The fuzzy nature of the connection between "intermediate" criteria and economic criteria can be illustrated by the following example. Agudelo discovered that the attitude Of rural peOple towards the extension service, which is sometimes used as a measure of extension performance, was positively related to adoption of practices in the Girardota area of Colombia.27 As adoption is related to benefit flows in the manner discussed earlier, this would imply that a favorable attitude towards the extension service is positively related to benefit flows. However, it is not clear whether or not this relation is causal and, even if it were, it might be causal in either direction or involve positive feedback effects. Further— more, no measure Of rural people's attitudes can possibly incorporate the negative effect of extension costs into the analysis. Thus, one would seem to be on very shaky ground if one concluded that the efficiency of extension work was positively correlated with the attitude of rural peOple towards the extension service. (Of course, one might be correct in concluding that extension benefits were posit— ively correlated with this "intermediate" criterion of performance.) Some extension evaluation studies28 appear to be 27Agudelo, Op. cit., p. 79. 28J. Neil Randabdugh, "Evaluating Extension Educ- ation", Evaluation in Extension, prepared by the Division of Extension Research andTraining, Federal Extension Service, U.S.D.A. 174 analyzing extension work by attempting to minimize costs per unit of extension output. This procedure necessarily involves the assumption that expected value per unit of output is independent Of the method or system employed. In the case where it is desired to speed up a group Of technical changes, output might be measured in terms of number of farm contacts or number Of changes made. How- ever, such measures Of output may easily have different unit values depending on how they are achieved. If, for example, an extension service decides to reduce costs per farmer visited by having their agents concentrate on smaller areas and visit more farmers in those areas with no change in the total amount of resources employed, there appears to be a strong possibility that unit returns achieved by visiting additional farmers would be different from original-unit returns. This possibility arises because of the diffusion phenomenon and the fact that the additional farmers visited may not influence the adoption distribution in the same way as the original farmers. The use of number Of changes made by program participants as a measure of output also bears this short- coming as it completely overlooks diffusion possibilities. (Of course, in a pOpulation where diffusion is non- existent, these measures could be useful.) However, while existing evidence suggests that rates Of diffusion differ greatly among populations and that in most areas there are a few farmers from which diffusion is virtually nil, there 175 does not appear to be any evidence to suggest that there are social systems where diffusion is non-existent.) This measure of output also bears the weakness that all types of changes are considered to be of equal value. This appears to be quite an unlikely event and should be investigated before any conclusion is reached in this regard. One might be better Off to form a weighted index of changes deriving the weights from extension workers' Opinions Of profitability and likely extent of diffusion than to assume all changes of equal value. CHAPTER VI EXTENSION ACTIVITIES OF THE CAUCA VALLEY CORPORATION This chapter serves primarily as background for Chapter VII which constitutes the main analysis of CVC's 1967 programs. Most Of the data used in the subsequent analysis are presented and/or discussed in this chapter in order to provide a clear picture of both the shortcomings and potential Of this part of the study. As some considerable difficulty was encountered in acquiring similar sets of data for all 1967 extension programs, it is hOped that the verbal discussion will make some contribution towards bridging the gap created by this deficiency. The Setting in Which the CVC Operates The Cauca Valley Corporation - or the CVC as it is commonly known - is an autonomous public agency created by Colombian Legislative Decree 3110 on October 22, 1954. To some considerable extent it was patterned after the Tennessee Valley Authority of the United States. Its original PUrpose was described by one Colombian writer as follows: [The CVC was assigned] the mission of trans— forming the Upper Cauca Valley and its surrounding regions: that is, to produce more and better 176 177 quality foods at lower prices; more electric power for factories, homes, and farms; more and better transportation systems at lower costs; and in general, improved economic, social, and cultural conditions for all the inhabitants, through better and more intensive use Of natural resources. For almost fifteen years the CVC has worked in an effort to carry out this mission. It now appears to have become a permanent feature of the Colombian landscape although some of its original Objectives have been considerably modified. For example, it was originally intended to be a regional rather than departmental organization. However, its act- ivities are now confined almost entirely to the Valle del Cauca. Valle del Cauca is located in the west-central part of Colombia. Figure 8, illustrates some of its geographical characteristics. The flat valley floor accounts for a relatively small prOportion of the total land area of the region, but about one-third of the land in farm units is located there. This area also supports a large part Of the department's population. Cali, Colombia's third largest and fastest growing city, is located at the western edge of the valley floor. Several other small cities are located along the Cauca River. Buenaventura, on the west coast, is primarily a port city with relatively little industry. How- ever, about one-quarter of a million peOple live in the lAntonio J. Posada, "Sentido y Espirita de la CVC," La Neuva Economia, Vol. 1, No. 2 (April 1961), Bogota, p. 222. 178 CHOCO I I I I I ..... ‘s‘ (.’ Bhenaventura ‘J’ I‘\ 3 : l“ \ I \\ PACIFIC g I (I) 1 OCEAN w ' ,3) a ' ..7 a, j 40 fi$ I L? I A? I W “I“. I Q r” N I q I ’U I ."I 1’ 44 I 'U I (0 I "U ’ 'U I C ,’ Q) (9' a . I ,3. E ! ‘V" CAUCA Figure 8.--Map of Valle del Cauca showing principal geographical areas and Office locations of the CVC's Extension Service in 1967. 179 Buenaventura area. Their pattern of life is somewhat primitive and the area is noted for its high unemployment, lack of economic Opportunities, and serious malnutrition problems. In general, the mountains and hills are not well suited for intensive cultivation but many parts are used for this purpose by subsistence farmers. An estimated land use pattern for land in farm units in Valle del Cauca in 1956 is shown in Table 6. A large prOportion Of the coffee, plantain, and pasture is located in the mountains and hills. Sugar, corn, beans, cotton, rice, bananas, and pasture predominate in the valley. This area has an average temperature of 75 to 80 degrees Fahren- heit. The mountain areas are cooler but frost is unknown in the habited parts. Dry and rainy seasons alternate fairly regularly by three-month periods producing two rather distinct crOpping seasons in each year. Agricultural output of the Department and particularly the valley floor, has been increasing somewhat faster than for Colombia as a whole. The number and area of farm units in Valle del Cauca according to size and main type of activity as they existed in 1959 are shown in Table 7. This table illustrates the concentration of land holdings (which is prevalent in most parts of Colombia), emphasizes the importance of cash-crop farming in the region, and suggests that a large part of the commercial agricultural output of the area, particularly of the valley floor, is controlled by relatively few farmers. 180 Table 6.--Estimated land use pattern for land in farm units in Valle del Cauca in 1956a Land Use Hectares % Temporary Crops - Rice 7,800 0.7 - Barley 1,500 0.1 - Beans 13,200 1.1 - Corn 43,100 3.7 - Millet 3,700 0.3 - Cotton 7,300 0.5 - Potatoes 1,900 0.2 - Tobacco 500 - - Wheat 900 0.1 - Cassava 4,100 0.4 - Others & Fallow 40,800 3.5 Permanent Crops - Bananas 9,000 0.8 - Cacao 1,900 0.2 - Coffee 105,500 9.0 - Sugar 80,100 6.9 - Plantain 31,700 2.7 Permanent Pasture 576,100 49.3 Mountains & Forest 184,400 15.8 Other (Lands, Buildings, etc.) 54,400 4.7 Total 1,167,900 100.0 aAdapted from DANE's 1965 Encuesta Agropecuaria. (DANE is an abbreviation for Departamento Administrativo Nacional de Estadistica which is the Official agency for compiling economic and pOpulation statistics in Colombia.) 181 .0H .m .AM00H.mmosmmom OfiEosmeHmo wuflsomm .OHHm> mo >uawum>flso "manfioaoo .HHMUV .000H I mosmu amp OHHM> amp Owumsommoumd Omcmu Eoum pmummpmm H.m50.aom oao.a o.mmm.aov Hmm.m m.5~m.o- Hms.ma m.mma.mm Hm5.¢~ Hayes ~.mmo.~ 0 H.~mm.~ mm ~.05v am v.m0 mm cmconcmna ~.mm~ H m.m0m «H m.H5a Hm m.m5v o0m.a manages m.mmm.mm mas o.mmm.O5 «mm 5.5vs.ma veo.a ~.mo¢ awe manaaouo paw xooumm>wq e.Hmm.Hm mma H.mmm.mm sum 0.Hm5.m mmfl 0.H¢ om sooumm>ea cmxaz m.o¢m.maa mum 0.~0m.m0 «M5 ~.omm.m mwm 0.405 mmm 06mm m.m55.m5 m5H H.H5M.H5 mmm s.¢mv.m omm m.mam 5mm shame ~.oom.H~H «Ha m.vm¢.amfl mm0.~ v.mvm.mma mmm.0a H.~om.mm ~H~.- manaaouo mwumuowm .Oz mmumuomm .Oz mmumuomm .Oz mmwmuomm .oz 5pe>fiuom mmumuomm 00m mmHMOOOm mmumuomm mmwmuomm 0 no case who: com on cm on on a case mmmn mass ou mcflpuooom mmma cw mosmu amp m5ua>fluom «0 mHHmb Ga muwcs Emma mmwu same can Oman mo mmum paw HmnEszII.5 manna 182 Some Of the social characteristics Of the rural areas Of the department can be illustrated by Table 8 which shows the percentage distribution of the population by age and education. The high prOportion (78.3%) of the population under 34 reflects the low life eXpectancy, high birth rate, and high out-migration rate in the area. These data suggest that at least 50 per cent of the rural pOpulation are, for most purposes, illiterate and that close to 90 per cent of this group have had no more than three full years of schooling. These data also indicate that, while education levels are improving over time, the rate of improvement is slow. Data from the same source show that about 35 per cent of farm Operators have less than one year Of education, 75 per cent have less than three years, and 95 per cent have less than five years. Thus, while this group is somewhat better educated than most rural peOple, many farm Operators (perhaps 50 per cent) probably cannot read effectively. CVC's activities since its inception have included land reclamation projects, power generation, transmission, and distribution projects, reforestration programs, agric- ultural extension work, soils research, mapping and testing, and various small development projects. CVC now owns and Operates most Of the electric power sources in the Cauca Valley. The history of CVC's activities and the manner in which they have been financed are illustrated in Table 9 which summarized CVC's revenues and eXpenditures for 1955 and 183 .Hmusu mm pmumpflmcoo mwm3 Amumomnmov ummm mussoo On» MO mmfimuso mmmu< .0m«100« .mm .A500H .pwnmmwmomez "mmeOHOU .wuomomv .mosmo amp OHHm> .lamma me me ceases .c0flomanom m0 Hmcoaomz cmcmo HHx .m.mzmflcod Hw>mq soflumospm >3 msowu 00¢ aw coflumasmom 00 w mv0ma ca mosmu Hop maam> mo mmmum Hanan now mam>ma newumospm paw mmm an coflumHsmom mo cowusnanumfloun.m manna .mmwuommumo musufipcmmxm umcuo macaw pmumuuoum cmwn wpmmuam m>mn mmmcmmxm coflumuumflcflEpdn .mOmmm smanEOHou cw pmuSmmme mum mumoo .ama .m can 00H .m .lmmmfl .oncs: umoume "annsoaoo .wuosomv .smnflmcoauncmua paw ucmsmqam>mpumpso Ou mmwwaamnu .U>U .mpmmom .4 mason 0cm mpmmom .0 Oficoucdm 184 mmo.sop.aqa 0~H.mm~.a mmusuflucmaxm Hmuoa 000.00«.« 0«0.00« ucmEmHsvm mo wmmsowsm 000.««0.0H muommoum uwnuo 0 “0300 «O mmmnousm 000.000 mmflpsum HMCmemm «w«.000.0 mamuwoum amusuasowum< «00.0mv.0 mxwoz msom:MHHOOmwz musmewmm unwa so mwfipwmnsm HOH.00~.50 «05.05« muomfloum um3om 005.000.5 050.5H muomnoum coflumsmaomm page 000.000.H H05.v00 mmwpsum unquOCwmsM 000.000.5 mocmswucwmz paw soflumuwmo uomnoum 1405.000.00 Ao5H.00H0 nmmmcmmxm aceumuumacesu< 5mv.mV5.0MH ooo.oom.a macw>mm Hmuoa 000.500 mmEOOGH umsuo 000.000 mcoflusnwuusoo Hmuwmmo 000.000.0H mxuos msomsmaamomflz MOM mmocm>p¢ 000.«00.5H mufimoum AfldOHmO c0 mmumzm «00.«0« mmousOmmm pepmwu Hmnuo 500.v0a.05 mason xcmm 000.««0.0 mmfioswwd umnuo Eouw mucmuo 5H«.00 mmfiuwammfiowcsz Eouu mcofiumwwmoummd m0v.oaa.a ooo.omo.a mumum scum mucmuo a mcoflumeumoumad 000.00«.0H 000.00« COwumz Eouw musmuu 0 mcoflumwumoummd 500.0««.va mmxme nosqwq 0cm wumumm «mom 0H«.50«.5 mmoa>umm umnuo paw auwowuuomam 00 00mm 000A 000a EOUH m000a van 0000 ”mmusuflpcmmxo paw mmscw>mw O>UII.0 manna 185 1963. A heavy dependence on bank loans (mainly from international agencies) is evident. These funds have, of course, been used primarily for long-term investments such as electric power generation. The figures also show that agricultural programs have represented a relatively small proportion Of CVC's total expenditures. About 70 per cent Of the funds allocated to agricultural programs have been used to Operate the agricultural extension service. A Brief History of CVC's Extension Activities In 1956 the CVC assumed responsibility for an extension service which had previously been Operated by the Federal Ministry Of Agriculture. The Ministry of Agriculture continued to supply funds for this service and, as a result, retained some considerable control over its general policy. In the late fifties and early sixties, CVC also worked closely with STACA and adopted many extension practices used widely in the United States. During this period many extension workers received training through STACA programs and from I.I.C.A. personnel from Turrialba, Costa Rica. Prior to 1966 the CVC attempted to maintain a fairly complete, general extension service such as that Operated in the U.S.A. However, because of limited resources only about ten field Offices could be Operated. Thus, each field office had to service a relatively large area and some parts of the region were entirely without CVC's extension services. Each 186 field Office was staffed by three to five extension workers. A typical office would include one professional agricult- uralist, one assistant working primarily with boys' 4-H Clubs, one assistant working primarily with women's and girls' clubs, and one secretary. The assistants usually had some technical training but were not university graduates. Annual reports during this period usually included statistical summaries of numbers Of 4-H Clubs, club members, farm visits, results, demonstrations, movies shown, meetings held, and various other activities. Club work, farm meetings and individual farm visits were the principal extension methods in use. Subject matter specialists were employed to assist field personnel. A fund was created to aid in the financing of various projects, particularly in connection with 4-H programs. Some effort was made to ensure that the CVC's extension service complemented other such activities in the region. As the National Coffee Federation Operated a relatively comprehensive program in the mountainous and hilly areas, the CVC confined their Operations mainly to the valley floor. Furthermore, they also avoided working with producers of crOps such as cotton, rice, and sugar because these farmers already had access to technical assist— ance through their Federations. The CVC extension service has experienced what appears to be a relatively high rate of turnover of personnel 187 since 1956. A 1961 study indicated that 60 per cent of the people on the staff at that time had been with the service for less than two years.2 By actual count, 60 per cent of the people on the staff in January, 1962 were no longer on the staff in January, 1967. Some of these changes occurred because Of the number of employees declining from about 55 in 1960 to slightly less than 40 in 1967.3 However, even when this is taken into account it would appear that, for an extension service, the rate Of turnover of personnel was undesirably high. A 1961 study pointed out that many Of the professional agriculturalists did not remain in the same office for more than one year.4 Thus, while the CVC was attempting to follow North American extension practices, it apparently was unable to achieve one important feature Of the North American system - continuity of personnel. Partly because of this reduction in staff and partly because of changes in programs, the CVC has deemed it 2Jorge Ramsey and Roy A. Clifford, "Analisis del Servicio de Extension Agropecuaria de la Corpgracién Autonoma RegiOnal del Valle del Cauca," A Mimeographed Report of the I.I.C.A., Turrialba, 1961, p. 20. 3This decline took place even though total expend- itures in constant pesos (1954-55 = 100) increased from 496,000 in 1956 to 777,000 in 1967. (Expenditures were 540,879 pesos in 1956 and 2,803,578 pesos in 1967). Average real costs per employee therefore approximately doubled during this period. It would appear that this increase was due to both increases in real wages and real increases in costs of associated inputs such as transporation, materials, and supplies. 4I.I.C.A., "Informe de Visita a1 Servicio de Exten- sion de la CVC," A Mimeographed Report by a group of I.I.C.A. students, Turrialba, 1961, pp. 5-6. 188 advisable to close a number of field agencies in recent years. By 1967 only five Office locations were being maintained: Cali, Palmira, Tulué, Roldanillo, and Buena- ventura, In 1968 the Roldanillo Office was closed as part of a cvc - INCORAS agreement whereby INCORA assumed respon- sibility for a number of CVC programs (including agricult- ural extension) in this area. As this study was being conducted, the Palmira office, which was located at the I.C.A. experiment station, was also closed. The principal reason for this decision was that the Palmira extension workers all lived in Cali and it was considered more efficient for them to function out of the Cali office. Thus, at this writing, only three Office locations existed: Cali, Tulua, and Buenaventura. About 1966, a significant change took place in that part of CVC's extension Operations which were directed at farm production units. Prior to this time much effort had been directed at very small farming Operations (mggif fundia). In 1966, programs were designed to reach larger producers and some of the previous programs were discontinued. According to the head Of CVC's agricultural department, it was concluded that previous efforts were having little affect on production and there was little future for the very small producer in agriculture. Therefore, CVC administrators believed greater social returns could be achieved at the 5INCORA is Colombia's national land reform agency. 189 same cost by working with producers who controlled larger volumes of production. In examining reports and talking with CVC employees, this writer ran across a number of problems which appeared to re-occur frequently over the years during which the CVC has Operated the extension service. These included lack or inconvenience of transportation, lack Of equipment and materials, low salaries, frequent failure Of extension programs, and an excess of reports which apparently did not appreciably aid personnel in solving the previous problems. This writer is not in a position to say to what extent these comments were justified prior to 1967 but they nevertheless provided useful aids for the orientation of this research. Many CVC extension programs are designed to coincide with the agricultural rather than the calendar year. In View Of this circumstance and the way available data had been recorded, it was decided to examine the activities Of the extension service for the period April 1, 1967 to March 31, 1968. It was considered that this period approximated the agricultural year sufficiently closely so that a useful analysis could be carried out.6 During this period six major programs were Operated: a livestock program (Fomento Ganadero); an apiculture program (Fomento Apicola); a fert- ilizer and farm planning program (Fertilizaciény Planifica- gipn de Fincas); home economics assistance (Asistencia a la k 6The two fairly distinct crOpping seasons which occur annually in Valle extend from October to March and April to September, respectively. 190 Familia Campesino); boys' 4-H Clubs (EducaciOn Rural); and special Pacific Coast programs (Fomento y Extension Costa del Pacific). Several minor programs were also Operated. Field personnel time was distributed among the five offices as follows: Cali, 29.5 per cent; Palmira, 19.5 per cent; Tulua, 14.7 per cent; Roldanillo, 12.4 per cent; and Buena- ventura, 23.9 per cent. Supervision was handled from the Cali Office and involved one professional agriculturalist and one assistant as well as secretarial services. A total of six secretaries and two chauffers were employed by the extension service. In 1967 several different classes Of field personnel were employed in the extension service. Many were professional agriculturalists (ingenieros agronomos and veterinarios) with university training in agronomy and animal husbandry. (Both the agronomos and veterinarios have signif- icantly different training experiences than their North American counterparts.) However, many non-professionals were also employed. These included both male and female personnel. The former group (practicos) were usually involved with boys' 4-H Clubs and the latter group (mejoradoras de hogar) were mainly responsible for home economics assistance. Agronomos were involved in virtually every program because of a rather unique organizational structure whereby both program heads and office supervisors were designated. As the mejoradoras and pgacticos did not possess university degrees, all such reSponsibilities feel to the agronomos. 191 Costs for CVC's 1967 Extension Programs An expense statement for the major programs in the period under consideration was supplied by CVC's Accounting Division.7 However, this statement was found to be unsuit- able for the purpose of this study for two reasons. Firstly, the costs Of the minor programs which were Operated during this period were not separated from the costs Of the major programs. Secondly, personnel and transportation costs had not been allocated to the major programs correctly. There- fore, a revised cost statement was constructed (see Table 10). In the revised cost statement minor program costs have been separated from major program costs. Costs of extension personnel have been allocated among programs on the basis of time spent in these programs. Vehicle Oper- ation costs have been allocated on the basis Of the number of kilometers recorded in the respective programs. The number of kilometers recorded in each program was based on personnel time (for those who drove cars) and average number Of kilometers recorded per day. Costs of chauffers were cal- culated in the same way as those Of extension personnel. The resultant total estimates of personnel costs were essentially the same as those supplied by the Accounting Division although 7See Appendix B, Table B-1. ll92 000 00000 004 .mu0wmsmno mo m0mc0mx0 000000>muu 000900000 .00000000 m0s0000000umo .000005020 00 00000000 00:00c0 00: 00000 .00000 :00QEO0OU s0 .00cc0000a 00000 U>U can 0:05000000 0c00:9000< 0.U>U an 00000030 0000 3000 00000900000 005.050.« 00«.«00 «00.000 050.000 000.000 000.000 0«0.00« 000.050 I 00009 0:000 000.500 000.00 «00.000 000.00 0«0.000 0«0.0« 000.00 000.55 I :O0umuum0s0spd I 000.00 000.0« 000.0« 000.00 005.0 000.0 0«0.0« 505.000 c0000>u0msm 0000:00 000.050 000.00 000.000 005.«00 00«.00 0«0.0« 500.00 000.00 000.0« 00009 050 500.00 000.« 500.0« 000 000 000.0 005.« 0«~.0 005.0 090000000000x 000.0 I 000.0 I I 555 050 000 «50.0 ucmem0avm can 0000000 no mo0xab 0 00:0:000002 000.00 I 000.0 000.0 000.0 I I 00«.0 50«.0 000:00 000000 050.50 500.« 0«0.0 000.00 000.« 00«.0 550.0 005.0 500.0 00sc0000m c000 Icauxm uo 000:0 Imxm 000000>oua 000.00 I I 500.00 500.00 I 000.00 I I pmu0uuamnu no 00000 000.000 050.0 050.00 000.50 005.00 000.00 000.00 550.00 I coflumummo 00o0no> «~0.00 000.« 000.00 000.5 500.0 «0«.« 0««.« 500.5 000.0 00000090 can 00000000: noncomxm 0000:00 000.500.0 000.05« 050.00« «5«.5«« 000.000 0«0.05 500.00 000.000 50«.000 00009 050 050.0« 0«0.«« 000.« 0000 00:000000000 000.000 «00.00 00«.50 000.05 000.000 000.0« 000.00 000.05 000.00 000000C00 00:00» 0«0.500 000.«00 000.0«0 050.500 00«.00« 000.00 005.00 500.000 000.00 .00000000 new 00003 0 mo 0ccomu0 nIv U 0 0 00009 02000000 06000000 02500 0000000004 5000000 6000000 8000000 s00m0> 00:00 00000 :I0 000Eocoum 0:0:0000 5000 00:00: x0000 Iu0asm 02000 0000 0000000 .0000 0E0: 0 00n000uuom I00m¢ I0>0q 0000:00 00 00000000000 00000 .00 noun: a» 5000 .0 00000 000000 any 000 U>U 0:0 00 000>u0m c00mc0ux0 0:0 0:0000000 no mumou 050 no 00050000 0000>00II.00 00009 193 the allocation among projects differed significantly.8 Transportation costs were originally calculated by the Accounting Division using an average cost per kilometer (including cost of chauffers) of about 1.40 pesos whereas the revised costs were calculated at 0.6088 pesos9 per kilometer plus the costs of chauffers actually used by the extension service. As the costs of chauffers represented slightly more than half of CVC's total transportation costs and the extension division used chauffers considerably less than other divisions, a significant reduction in the estimate of total transportation costs resulted. PrOport— ionate allocations of transportation costs among the programs also changed significantly. Accounting data for items other than personnel and transportation costs were accepted with some pro-rating to other programs. The estimates of administration costs provided by the accounting division were also accepted. These represented the extension service's share of the total costs of CVC's administrative structure which included accounting, personnel, and trans- port divisions as well as various executive positions. The costs of extension supervision were re-estimated along with 8Estimates of the time spent in various programs were obtained from field personnel individually. Salaries paid were obtained from the payroll office. Employee benefits were calculated using a factor of 54 per cent of wages and salaries. This factor was supplied by CVC's accounting division and its make-up is shown in Appendix B, Table B-2. 9The estimate of 0.6088 pesos per kilometer represents all vehicle operation costs except the costs of chauffers. The calculation of this figure is shown in Appendix B, Table B-3. 194 extension program costs. Supervision costs include the costs of personnel employed full-time in directing the day- to-day operations of the extension service together with the associated items necessary for performing this function. Both administration costs and extension supervision costs were prorated to the various programs to obtain estimates of total program costs. The final estimates of individual program costs may be regarded as estimates of the costs of both "operating" and "managing" 1967 programs, although, in reality, some of the managerial costs attributable to 1967 programs had been incurred in previous years and some of the 1967 costs could be attributable to management activities related to future programs. Direct personnel costs (salaries and fringe benefits of field and office extension workers) in the revised cost estimate accounted for 60.5 per cent of the total costs. Administration was the next largest item accounting for 20.9 per cent of total costs. Probably 80 or 90 per cent of the administration costs were salaries and fringe benefits of office personnel. Transportation costs represented 13.0 per cent of the total. Data on personnel time for various programs were analyzed10 to arrive at estimates of the average total costs per day for various classes of personnel. This was further subdivided into "salaries and fringe 10Appendix B, Table B-4. 195 benefits" and "other costs".11 These data are shown in Table 11 along with number of days of field personnel time and estimated total costs of employing each class of personnel. Approximately 56 per cent of total costs were attributable to the activities of the agronomos and veterinarios. "Salaries and fringe benefits" varied con- siderably by class of personnel while "other costs" were approximately the same for all classes except the mejoradoras de hogar. This was accounted for by the fact that their work invOlved less travelling than the work of either the agronomos or the practicos and they used somewhat less material and supplies. As the system of allocating administrative expenses ensured that these would increase prOportionately with increases in programs and supervision costs, it was not possible to accurately distinguish variable and fixed costs. The marginal cost of employing additional personnel was probably at least equal to direct program costs because additional transportation, materials, and supplies would almost surely be required with each personnel increase. Some increased "administration" might also be required. If it is assumed that both supervision and administration are fixed while all other items are variable, variable costs would be 73 per cent of total costs. Thus, it appears that a large 11Average "salaries and fringe benefits" per day vmre calculated independently and "other costs" were calcul- Efifid by subtracting "salaries and fringe benefits" from total costs per day. 196 .iv-m manna .m xfiuammma mmmc .mumoo Hmuou Hmsuom Hmswm >Huomxm no: mooo awn mom mumou Hmuoa mumasoamo on com: cofiumnvo codmmmnmmu on» How mumoo pmuommxw on» no amuou on» mmsmomn as can no: mmoc GEsHoo macaw wmmh.m>m.m mam oqa mud ma~.m mommmHo Ham omm.mm~ mma med om oo~.H moumuno oop.mvv mom mma NHH mmn.a moofluomum oam.ma¢ HNN moa NHH mmm.a ummmm mm monOUmuoan vmm.mmw.a mmv Hma «mm wmm.m moflumcflumum> 0cm moEocoum¢ AmOmmmv >mo mom moo mom “mommmv zoo mom mwmo mo mumou mumou mumou muwmmcmm wmcaum umnEdz HmMGMMMMW Hmuoe amuoe umsuo a mmHHmHmm m >2 mumoo Hmuou HoccOmqu pawflm mo mmmHo can .mumoo >HHMU mmmum>m .mEau Hmccomumm vamwmuu.aa magma 197 part of the total costs per day would be incurred if the extension service were expanded. However, as perhaps 10 to 20 per cent of total costs are fixed - or at least do not appear to increase linearly - it would appear that a some- what larger operation would have slightly lower costs per day rather than equal or higher costs. Description of 1967 Extension Activities Because of cost restrictions on this research and the nature of the benefits being produced by the various programs it was impossible to evaluate program benefits as precisely as program costs. However, detailed program desc— riptions were obtained from extension workers and lists of recommended practices were compiled. For three programs it was possible to survey samples of the program participants to acquire information about the services received and how they were being used. For the remaining major programs it is only possible to describe their general nature and suggest possible factors influencing their relative effeCtiveness. It was necessary to ignore entirely those activities class- ified as minor programs. Livestock Program This program was Operated in both the Cali and Roldanillo - Tulua areas. Costs of Operation for 1967 were estimated at 190,832 pesos and 185,508 pesos for activities in these respective areas. Almost all extension work was 198 achieved through visiting individual farmers and advising them on their Operations. Towards this end about fifty farmers were contacted by the extensionists in the Cali area and about forty in the region between Roldanillo and Tulué. A number of the farmers contacted in both areas were not interested in the program and only about fifty of the ninety farmers contacted could be regarded as participating in the program. Thus, extension costs per participating farm were about 7330 pesos. Most of the farms in the program were medium to large in size (50 to 200 head of cattle - dairy and/or beef) and there was little doubt that size of Operat- ion would not be a severe restriction on the possible utilization of new techniques. A list Of recommendations12 made by extension personnel was Obtained from one of the agronomos working on the project and this writer accompanied him to one of the co-Operating farms. It was quickly evident that, although some measures of changes made might be Obtained, the eval- uation of these would prove to be extremely difficult. Many of the practices recommended (e.g., artificial breeding, better selection procedures, improved pastures) would produce returns only after several years of use and would require a number of years of good farm records to detect. Few measures of income or income increases were available from either the 12These recommendations covered the areas of feeding, tweeding, and herd health care. Most of the recommendations nlthe area of feeding concerned pasture improvement as Pasture constitutes the main part of the ration. 199 farmers or CVC employees although some production records were being maintained or initiated in a number of cases. Therefore, no attempt was made to interview participants in this project or evaluate actual changes made. However, the CVC had prepared a project proposal early in 1967 and this is reproduced here to acquaint readers with the project and CVC's method of project analysis. I. Description Of the Project A. Nature Of Project During the operation of this project live- stock farmers and their employees will be trained and educated in aSpects of animal health, better handling of their pastures and cattle and the latest techniques for Oper- ating livestock farms; and the credit devoted to these farms will be directed to better uses. Specific Objectives of the Project 1. To familiarize livestock farmers with modern management practices. a) Economic analysis of the livestock farm b) Pasture improvement (i) division of pasture fields (ii) analysis of soils (iii) fertilization c) Systems of livestock records (1) health - vaccination (ii) production (milk and beef) (iii) calving d) Improvement of livestock e) Specialization of the Operation 2. To increase the productivity of livestock farms. 3. Better utilization of the resources of the farms. Localization Department of Valle and the North part of Cauca. Municipalities in Valle; Dagua, Jamundi, Cali, Palmira, Candelaria, Cerrito, Guacari, Buga, San Pedro, Tulua. Municipalities in Cauca; Santander de Quilichao, Puerto 200 Tejada. The operation of the livestock program has been divided into two zones: the northern zone includes Valle from Cartago to Buga inclusive, and the southern zone services the municipalities of Valle from Cerrito to Jamundi: together with the northern part of Cauca. D. Current State of the Project13 Data for the Operation of this project are obtained about pastures from I.C.A., about soils from I.G.A.C. and CVC, from reports of the Fondo Ganadero, from the review of the Banco de la Republica, from work of the Live- stock Sections of the CVC, from livestock censuses of DANE, etc. With this information and with the visit of the French Mission assigned to C.A.R., their [the French Mission's] system will be adapted to the cond- itions of Valle. E. Stages of the Project Even though the progress of the project depends on the physical condition of the farm, and the economic situation and the attitude of the farmer, the following general stages can be established: - farm inventory - determination of problems - economic study of the farm (at beginning of the project) - agreement of criteria between owner and extensionist - work plan a) size of farm and tOpography b) soil study c) irrigation feasibility study d) pastures - seeding - improvement - subdivision and rotation e) fertilization 13I.C.A. refers to the Colombian Agricultural Institute, I.C.A.C. refers to the Institute of Livestock and Crop Producers; C.A.R. refers to a regional agency similar to the C.V.C. II. III. 201 f) animal husbandry - feeding - health and sanitation - management - selection care of calves g) livestock records h) periodic economic studies for extension project Relation and Influences of the Project with Projects of Other Organizations A. In relation to the need for human resources, advice will be available later from the French Mission now assigned to C.A.R., Bogota. Influence on projects in Operation or to be operated by other organizations. This project is complementary to those under the auspices of the Livestock Fund of Valle. Justification of the Project A. Economic Benefits 1. Production of exportable commodities - production of beef for export in fewer years - production of foreign exchange quickly 2. Production of commodities for popular consumption: milk and beef 3. Increase in productivity More profits with less investment than at present. 4. Reduction of costs as a result of higher productivity per unit of area. Social Benefits - specialization of labor - better product for the market (health) Type of Farmers Who Would Benefit — 20 medium size farms - 20 larger farms Region Which Would Benefit - appears in the localization Of the project Priority of the Project - [left blank in original document] 202 F. Technical Studies that Back the Project Different bulletins about pastures and feeding by I.C.A.; the systems of Program Planning of the French Mission; work exper- iences of the CVC. IV. Annual Cost of the Project Salaries of two Veterinarios, two Agronomos and one secretary 212,900 Fringe Benefits (Prestaciénes Sociales) 117,100 Materials and Supplies 30,000 Travelling Expenses and Transporation 197,000 Public Services 1,000 Communication Services 7,000 Printing and Publications 5,000 Office Rental 5,000 Visitors EXpenses, Meetings and Exhibits 5,000 Maintenance and Insurance 3,000 Miscellaneous 20,000 Total 603,00014 The Apiculture Program The Objective of this program was to expand honey production (both by increasing the number of honey producers and increasing output per producer) and, by so doing, to increase the incomes of farmers involved in this activity.15 Advisory work was directed at individual honey producers and at apiculture clubs which included both bee-keepers and other persons interested in learning about honey production. Clubs were sometimes Operated in co-operation with rural schools and between 10 and 40 students received instruction when the extension workers visited the school's apiary. The two 14Unpublished outline of CVC's 1967 extension pro- grams prepared by the Extension Service. Translation by the author. Costs are measured in Colombian pesos. lsIbid. 203 extension workers assigned to the project spent almost all of their time making visits to the individual producers or the clubs. They had established a specific set of demon- strations and instructions designed to show and inform interested peOple how to produce honey profitably. In addition to working with the clubs and individual bee-keepers, the field personnel of the project engaged in 16 the following activities: (1) Organizing a bee-keeper's association with 138 members in Valle and surrounding departments. (2) Teaching a course in apiculture to 12 Peace Corps Volunteers working in Valle. (3) Organizing the third National Apiculture Conference in Cali. This was attended by 222 people from various parts of Colombia. (4) Investigating, in co-Operation with the Association of Bee-keepers, the possibilities of developing an export market for honey. (5) Assisting bee-keepers in Obtaining credit from the Ca'a Agraria, 7 The total amount of credit granted during the year was 120,000 pesos. Field personnel indicated that these activities took a relatively small amount of their time but, it was difficult to tell exactly what part they had played in them. The program was Operated out of the Cali office but bee-keepers in many parts of Valle del Cauca were visited. One agronomo and one practico were employed full—time in the program. Average total costs per day of field personnel time were 453 pesos. This was about 100 pesos per day higher than -‘ 16Annual report of the apiculture project for 1967. 17The national Colombian credit agency. 204 costs would have been had they been equal to the CVC average for the classes of personnel involved.18 The high costs were largely attributable to the large amount of travel involved and the fact that a chauffer was assigned almost full-time to the program. In view of the fact that both apiculture clubs and individual producers were serviced, costs per participant are not a particularly useful measure for this project. Thus, costs per hive have been calculated instead. The total number of hives maintained by all program participants in 1967 was 418. Thus, average total costs per hive were 498 pesos. It was difficult to establish any precise set of initial conditions for this project as many of the particip- ant bee-keepers and clubs had also received assistance in previous years.- It was also impossible to get very much data from participants on income increases due to the assistance received. However, several participant bee— keepers and clubs were visited and informally interviewed. The following information is based on those interviews and talks with extension workers involved in the program. Most of the bee-keeping Operations serviced were quite small by North American standards with the number of hives per bee-keeper varying from three to twenty-five. Most operations were secondary sources of income for the 18As shown in Appendix B, Table B-4, actual costs exceeded predicted costs by 48,202 pesos. 205 producers involved, although in some cases this income was quite significant. In a few cases, hives were merely old boxes and extraction procedures were primitive. Honey was usually sold in the local neighborhood or delivered to Cali for sale. Average price received appeared to be about twelve pesos per kilogram for a good quality product. How- ever, some producers were located in areas where most sources Of nectar available were sugar cane flowers and the L, resultant product somewhat resembled molasses. This product commanded a lower price. Because the bee-keepers lived some distance from one another there was relatively little social contact among them, and little opportunity to exchange information. The bee-keeping clubs brought together people interested in bees but, in any given area, the potential honey production was such that only one or two of the members could actually become commercial producers. (The remainder of the members undoubtedly found that this training contributed to their general educational experience but it would be of little or no direct practical usefulness.) Some bee-keepers visited reported receiving technical assistance from other sources (e.g., the Coffee Federation). Some bee-keepers reported substantial increases in production and profits as a result of the assistance received; others reported that losses resulted when they followed the advice of extension personnel. Total output of honey varied 206 fronllfl to 40 kilograms of honey per colony per year depending'on weather, diseases, and cultural practices fiollowed. Thus, total revenue per colony ranged from 100 to 600 pesos. Costs of production involved mostly labor and this depended considerably on the size of the harvest and problems encountered. Some producers had increased their profits as much as 100 pesos per colony from following extension advice. However, the average increase was probably much less. The Fertilizer and Farm Planning Program This program was part of a co-operative project involving CVC, F.A.C. and a Colombian fertilizer company. During the period of 1962 to 1966 field trials were conducted on about 400 farms in Valle del Cauca in an effort to deter- mine what analyses and levels of fertilizer would be profit- able for a number of crops in different areas. Similar Programs were also Operated in other parts of Colombia. This part Of the project used the information obtained from these fiEId trials and other recent research in Valle coupled with the knowledge of fertilizers, crops, and soils of the region which CVC personnel had accumulated in their technical tralDing as agronomos or practicos. The program objectives were: (a) "to show, by means of farm trials, the worth of adequate fertilizer and [in so doing] to increase its use ~wfl-V 41? 207 (b) to teach the farmer to consider his farm operation as a business lkxcorder to achieve these objectives it was originally decided to take into account the following factors in select- ing farmers to participate in the program: (a) [participant should have] a tradition as a farmer (b) preference to be given to owner-Operators (c) adequate technical conditions for the crOp (d) cultural practices of the farmer should be acceptable (e) dedication of the farmer to his crop (f) preference for medium-sized farms (10-25 hectares) where fertilizer had not been used before (9) first make a broad selection, and later restrict it (h) keep in mind that the program is one of credit, education, and research; only a good co—Operator will permit achieving these Objectives.20 - However, because a good deal of time would have been involved in locating such farmers, many participants who did not have all these characteristics had to be selected. The number of farmers21 co—Operating by semester, extension area, and crop are Shown in Table 12. Each farm was visited approximately ten times during the growing period. Fertilizer was supplied on credit by CVC and recommendations regarding seedbed Preparation, seeding fertilization, weed control, insect COntr01, disease control, irrigations, and harvesting were made. The regular visits permitted extension workers to make sure that their recommendations had been understood although in sOme cases not all of these recommendations were followed. X T 19Unpublished outline of CVC's 1967 extension programs. ranSIated by the author. 2 (N 0Unpublished work report of the fertilizer project, 0V- 1967). Translated by the author. 21 e A number of 4-H Club members also participated in SESEntially the same program but for the purposes of this P “dY they have been considered part of the Boys' 4-H Club rogram. Costs have been adjusted accordingly. 208 Table 12.--Numbers of farmers co-Operating in the CVC 1967 fertilizer program by semester, crOp, and extension area Roldanillo Palmira Total lst Semester — Tomatoes 0 7 7 - Corn 5 9 14 2nd Semester — Tomatoes 4 11a 15 - Corn _6_ 3 9 Total 15 30 45 aThree farmers in this group co-operated in the previous semester. Average total costs per day of field personnel time were 387 pesos, about 40 pesos per day lower than they would have been had they been equal to the CVC average for the type 0f Personnel employed. Costs per day were almost the same in both extension areas. Average total costs per partici- pating farmer were 2774 pesos in the Roldanillo area and 3257 Pesos in the Palmira area. This cost differential existed because extension personnel spent more time with tomato growers than with corn growers. Average extension costs per grower for tomato and corn growers were estimated at 3832 99303 and 2395 pesos, respectively.22 In order to ascertain the results of the fertilizer pr°9ram, some participants were interviewed in detail and most of the remainder were interviewed briefly. Data on the \ f 22These estimates were constructed by solving the 01lowing equations: 4xl + 11X2 = 41673 (Costs of program in Roldanillo) 18x1 + 12X2 = 97711 (Costs of program in Palmira) T213 agreed closely with one agronomo's comment that tomato g'OWErs required about 50 per cent more time than corn growers. L 209 Participants were also obtained from CVC files. Because the CVC had maintained test plots on some farms and the Program was just being completed when this study began, it was possible to piece together fairly good estimates of what had actually happened during the program. Data were collected on technical changes made due to the program together with associated increased costs and increased returns. Where all or a large part of the crop was not har- vested because of bad weather, serious disease problems, or other causes yield increases due to the technical changes were usually presumed to be zero. This meant that the farmer suffered a loss of the additional eXpenditures for fertilizer, sprays, etc. (A total of ten corn producers and five tomato producers found themselves in this situation.) A summary of the results of the fertilizer program for program participants is shown in Table 13. Data on individual participants are given in Tables B-5 and B—6 of Appendix B. Income increases resulting from the new technology were considerably higher on the average (both per farmer and per E1253) for tomato producers than for corn producers. While both groups experienced positive average gains, nine corn producers and six tomato producers suffered losses because the added eXpenses exceeded added returns. Participating farmers were also asked to give an estimate of the length of time before they would have adopted the new techniques without extension help. In the case of 210 Table l3.--Summary of results of CVC's 1967 fertilizer program for participating farmersa —; Tomato Producers Corn Producers Yield Increase +160.2 boxes/plazab 0 to 774 boxes/plaza +252.2 kilograms/ plaza 0 to 958 kilos/plaza Price per Unit 13.66 pesos/box 7 to 24 pesos/box 1.21 pesos/kilogram 1.07 to 1.36 pesos/ kilogram if? Gross Income' 2188 pesos/plaza 306 pesos/plaza Increases 0 to 12427 pesos/ 0 to 1303 pesos/ plaza plaza Cost 254 pesos/plaza 171 pesos/plaza Increases -657 to 1304 pesos/ -631 to +1008 pesos/ plaza plaza Net Income Increase 1934 pesos/plaza —1304 to 12427 pesos /plaza 135 pesos/plaza -600 to +768 pesos/ plaza Area of Crop Grown 1.293c plazas 0.23 to 4.0 plazas 3.971 plazas 0.75 to 8.0 plazas Net Income Increases per Farm 2362 pesos -ll70 to +22820 pesos 536 pesos -3480 to +4918 pesos aThe first figure shown represents the average or mean change. All averages for tomatoes are weighted by boxes. All averages for corn are weighted by kilograms. Ranges given are ranges by farm. b acres. A plaza is a measure of area equal to approximately 1.5 cOnly l9 farms were used as 3 participated twice - if 22 observations had been used, the average would have been 1.221 plazas. _____£ 211 mmmito producers this was surprisingly short - 3.16 semesters mIEIverage. However, these estimates were considered reasonable by the agronomos associated with the project as the technology, although new to the participating farmers, was not entirely new to tomato producers in the region. Par1:icipating farmers listed neighbors and other agencies as Eilternate information sources. Information on the diffusion of information from Part:icipating farmers was also gathered. This will be useél and explained in the calculation of program benefits in Chapter VII . Hmne Economics Assistance The home economics program was directed at rural hol-Ilsewives and teenage girls throughout the department. It was Operated at all office locations. Extension personnel ineluded mainly mejordoras de hogar although agronomos and Eracticos spent a total of 162 and 217 days respectively w"Hiking on the program. These latter classes of personnel werxanminly involved in supervision and delivering the fléjcmadoras to work. Average total costs per day of field Parsonnel time were 239 pesos which was just about equal to cVC's average for such a complement of personnel. Costs per day were somewhat higher than the CVC average for Wejoradoras due to the involvement of the agronomos and EFacticos. 212 The nature Of the program was quite similar at all ofliice locations. Each mejoradora formed several clubs in mural neighborhoods (veredas) and the clubs were visited weekly. Some clubs were made up entirely of housewives; some were girls' 4-H Clubs; and others had both young girls mxi housewives as members. One mejoradora could usually seruvice four or five clubs. A total of 45.7 clubs23 were oPerated for a full year. An average of 49 days of field Personnel time was required for each club although only about hal;E of this time was actually spent at the meetings. Average club membership was 21.3 persons. Average total extension COSts were 11,777 pesos per club or 551 pesos per member. Nmuber of days per club, extension costs per day, club I“embership, and average costs per club and per member varied Slmewhat among office locations.24 For example, average ‘33sts per day of field personnel time ranged from 195 pesos 5h1 Buenaventura to 273 pesos in Cali. Average costs per Cldab member ranged from 428 pesos in Buenaventura to 706 Exesos in Cali. Some of the cost differences among office lOcations were due to differences in other duties performed 25 by the mejoradoras. The remainder was probably due to \ 23Some clubs were not operated for 12 months. , 24More details on this can be found in the tables in Appendix B. , 25For example, one mejoradora spent some time giving lnstruction in a local normal school for public school teachers. As these duties represented a small prOportion of tfie total costs of the program and were difficult to disting- u18h and analyze, they have not been separated from the club wOrk. As a result, estimated costs of the clubs may be as mPCh as 15 per cent higher than actual costs in some locations. ..unnn2-0m “-1 “a ‘s. A. 213 different work methods, different sizes of clubs and dtfiferences in the amount of time spent with each club. The clubs held their weekly get-together at some cmrvenient location in the neighborhood (e.g., a member's housse, a local store, a community building). Most members wallumucw Hones: Hmuou may .04 mamsvw mmonom mafia comm mo Hmuou mnem mNH.H 0.5 mmH.o~ m~.o om.~ mam.v mmH.v mmmum>4 H OH mm o v m I mcmunmw o H mm H o a «H Hmcssoo cpHooa o H om o H m m suumaucH HHmsm H m Hm o a a m ucmsm>oumsH caosmmsom o 8 mm o m a . coom m vH NH 0 v m m OGOflmwm H HH mH o a m m nuHmmm H HH mH H H o N mcHnuoHo umoEHa -mco sHHmsuuH> amoeHa -mco sHHmsuuH> «\z .o uson< usonm HO O Ooa3osx Howhm MO Hm>mfl Emumoum cw magmawm>¢ Ewumoum ca maanHm>¢ mcoz coHprMmmmm page no HH< umoeHa page no OHmmumco usona sHHmsunH> .wo Hw>m% .m cofluoouumcfi mo mono was .pm>flmomu cowuonuumcfl no mo .masflaoau coflposuumcfl mo Hmsmfi an umawfi>uwuafi ESE BE .8 Has: mumnEmE Dado mo Honssz||.v~ ownmh 218 2. 2 per cent). Although these measures are crude ones, it is obvious that the mejoradoras were concentrating on the introduction of new ideas and techniques to the club members. In terms of the analytical framework developed previously, most of the material taught could be viewed as either technical or managerial innovations - not just improved combinations of old inputs - and initial adOption conditions were close to zero for most of these innovations. On the other hand, where participants had previous knowledge of some of the things taught (e.g., clothing, health, hygiene, and gardens), several possibilities existed. These included: (a) innOvations in these areas were already adopted by the group to the extent that they were useful. (The income constraint was probably important here). (b) the mejoradoras were facilitating movements towards economic optima. (C) the club members were innovators or early adOpters who were increasing their knowledge in these areas and would Serve as a source of information for neighbors in the fUture. Possibilities (a) and (c), and (b) and (c) are not mutually- exclusive pairs. Two measures of utilization of instruction were 30 The first measure Col , . leCited from persons in the sample. was I I I 9 One of income increase due to increased production of Sa leable products. It should be noted that these products was 30These measures were not taken where no instruction had reCeived or where the persons interviewed indicated they atio 1"Kawn all the material taught prior to their particip- 1n the program. k. "I. 219 were seldom sold - they were usually consumed in the home. However, they were valued at what they would have cost the household if they had been purchased. The value of the purchased inputs used to produce these were deducted as costs I . 4-- and, where the opportunity cost of labor appeared to be positive, it was also deducted. Where income increases were not easily estimable or where the estimated income increases covered only some of the uses, a frequency of use was estimated. No exact definitions were constructed for the frequency categories and thus, they are not particularly precise. However, the distinction between "Almost Never" and "In Daily Life" appears to be intuitively-Obvious and particularly revealing, especially in the areas of Home Improvement and 512L631 333%. Table 15 shows the results of these attempts to mea551-Ire level of utilization. The income estimates must be Considered rather crude because they were made entirely by ' recall and some Opportunity costs may have been overlooked. SOme considerable difficulty was encountered in determining What part of 1967 income in a given area was actually due to 1967 CVC assistance. It was generally assumed that, when a club member reported knowing nothing before, the entire amo Inn; was due to new learning. However, this was not so a): where some or all of the knowledge had been acquired pre\’i<>usly. 220 .mmmfiflmfifl UGMHflmquH HHMEw .0 #Hmum Ou Emuoouo mOfiEocoom wEon on» ca omcflmo woowa30cx on» owns mam .movwc: muflso ma Hum» mom momma oonH an mEoocw no: ommmmuosfl 0:3 :mEoz on» no wmmo one a .mmcfinomfi mcfl3mm mo wufiaflnmaflm>m Mom moms coon mm: ucmEpmsnom ozm «\z «\z «\2 5mm mH HmHnmummm>o mawuumo m m on «\z o Hmcsfioo cmfloom a\z «\z «\z mmH mH suumsecH HHmsm o m mm ooH a usmEm>ouoEH meow om mH m nacha H ooom m Ha ma «\2 o wcwwohm m 5H m «\z o nuammm mH m OH mHmm mm mcHruoHo OMHH mmEHu Hm>mz HmOmomv mommmuocH wafimo Imeom umoEH¢ OHQMMSmmmz wEoosH CH mumsz mHansmmmz mmmmuocH mEoosH mcwocOfinmmxm mommmuocH amused momum>< umnEsz mEoocH wanmusmmmz cw ooumuomuoocH uoz moomHBch «O on: mo hocwdqmum ou ooummmo cofiuosuumcfl mo mmmum ucmflo ecu mo coflumuwwwu: mo mamsmq|u muonEOE nsHo moHeoeoom meow umm~ .2 ages 221 Considerable variability was noticed in the level of utilization of each category of instruction among the members. Although the 29 sample members who reported income increases in Clothing averaged a gain of 290 pesos in the year, this ranged from 30 to 970 pesos. Seven other members who received new instruction in this area reported no income increase. Most Of the families could not be expected to maintain the same level of usage in the future - some might cease entirely although others might expand usage. The rate of utilization of gardens in the future was also question- able as in some cases this had been a special club project which participants might easily discontinue without the "Special project" incentive. In order to Obtain some measure of club members' oPinion of the relative usefulness of the eight areas of instruction, members interviewed were asked to rank their first three preferences. The results of these rankings are 31‘0“"! in Table 16. It is obvious that Clothing and Food Were, in general, the first and second choices of the group. Gardens and Small Industry seemed to be rated about equally in third spot; Health and Hygiene were rated close together but slightly less popular than the preceding four; Ho . me Improvement and Accién Comunal were ObViously the 1e a"gt-“preferred areas. This ranking appeared to be reason- ab 1y Consistent with the levels of utilization given in Table 15 ¥ 222 Enable l6.--Number of club members in the sample who rated the eight areas of instruction first, second, or third in order of utility for them Not First Second Third Rated Totala :1: '5; Clothing 25 0 6 36 Health 4 18 31 Hygiene 6 21 32 1).. Food 5 17 8 10 4o ‘ Household Improvement 0 0 3 31 34 Small Industry 4 5 7 21 37 m M 0 0 1 26 27 Gardens 3 2 11 23 39 Total 40 40 40 156 276 \ a TOtals are less than 40 by number not receiving instruct- ion and number indicating they already knew all that was tanght BOY'S 4-H Clubs The Boy's 4-H Club program was directed at YOUthS and YOUng men throughout the area. No specific age limits were in effect and boys as young as 10 or 11 years were sometimes inCluded in the same club with men in their early twenties. This practice was followed so that clubs of easol'lable size could be formed in the various communities. The prOgram was operated at all office locations except “a... 223 Buenaventura. The extension personnel involved included mainly practicos although agronomos spent some time super- vising the program and helping out in special programs while mejoradoras also co-operated in this program occasionally. Average total costs per day of field personnel time were 302 pesos which was slightly lower than the CVC average for the complement Of personnel used. The nature of the program varied somewhat among Office locations depending on the interests of the Club members and the involvement of the local practico in other activities. However, each practico organized several clubs and regular meetings took place during the year. One practico could service 3 to 6 clubs depending on his other duties. A total of 16.7 clubs were operated for the whole year.31 An average of 89.2 days of field personnel time was required for each club although only about half of this time was actually spent with participants, the remainder being devoted to travelling, preparing programs, and administrative duties connected with the program. Average total extension costs were 26,950 pesos per club or 1,789 pesos per member. Number of days per club, extension costs per day, club membership and average costs per club and per member varied somewhat among Office locations.32 The clubs held their weekly meetings at some 31 . See Appendix B, Table B-6. 32See Appendix B, Table B—7. “7::— _ _ 224 location in the local community to which all could travel conveniently. Most members walked to the meetings. During their meetings instruction was provided in the following areas: Seeding corn — preparation and disinfection of the soil, methods of seeding, selection of seed, control of insects and diseases, fertilization and the use of nutrients by the plant, weeding, and others. Vegetables - preparation and disinfection of the soil, fertilization and the use of nutrients, trans- planting, irrigation, control of insects and diseases, germination test, food value of vegetables, weeding, thinning, making a hot bed, laying out a garden area, and others. Citrus Fruits - pruning, control of diseases and insects, thinning, fertilization and the use of nutrients by the plant, planting, weeding, harvesting for market, and others. Swine - feeds, feeding methods and rations, the value of feed nutrients, vaccinations, castration, control of internal and external parasites, assisting at birth of pigs, and others. Cattle - vaccination, feeds, feeding methods and rations, the value of feed nutrients, castration, con— trol of parasites, care of calves, veal prod- uction, and others. Household Improvements - fixing and painting walls and floors of houses, enlarging and improving houses, making furniture, repairing chairs, etc. Hygiene - construction and use of latrines and other waste disposal structures including garbage cans and cess pools, ventilation and disinfect- ion of houses, etc. Instructions were complemented, where possible, by demonstrations on the members' farms. However, because of the nature of the topics discussed and the resources available to members, many had little or no Opportunity to actually 225 use some of the instruction (e.g., members without pigs might observe at the farm of a neighbour who had pigs, but they would not have much of an opportunity to actually do anything). Participants in the program were usually sons of local farmers, tradesmen, or obreros. Some of the families of these youths were of quite modest means; others were reasonably well~off by local standards. The average weekly income reported by club members interviewed was 134 pesos. However, in some cases this was family income and in other cases it was the member's income. In the latter instances, family income was usually much higher and the young man had already become established "on his own". Weekly family incomes reported ranged from 30 pesos to 250 pesos per week. In those few cases where club members were from families with medium to large holdings, it was questionable whether the estimates reported were actually income estimates - they seemed to correspond more closely to weekly expenses for individual or family living. The communities in which the Boys' 4-H clubs were located were Often the same ones in which the Home Economics Clubs were located. Some families would have boys in one club and girls in another. However, while members of the most affluent families in the community seldom participated in the Home Economics Clubs, there was some representation Of this group in the Boys' 4-H Clubs. 226 In an effort to get some idea of the benefits a1:tributable to this program, 18 club members were inter- xniewed. These participants were selected at random with tfiie restriction that no two members of the sample could tnelong to the same club. As with the Home Economics Clubs, it: was decided to consider all club experience and adjust tiiis to the equivalent of 12 months' experience in the benefit-cost calculations to be made later. Table 17 shows the number of club members interviewed bu? level of instruction received, level of prior knowledge (If this instruction, and the area of instruction. One might Say that the equivalent of 33.3 per cent of the club members received no instruction whatsoever. The equivalent of 13.8 fixer cent received about half of the instruction Offered, axua the equivalent of 52.9 per cent received all of the iJIstruction offered. This was partly due to the short-term nature of a few of the clubs and the fact that some members hadlittle or no interest in some areas of instruction. The table also provides an idea as to the extent to which the instruction received was, in fact, new knowledge for those members receiving it. The conclusion is that, for the most part, the members were learning new things. The same inter- pretation applied to the data in Table 14 applies in this Situation (see p. 218). Three measures of utilization of the instruction received were collected from the persons in the sample. This .cm3OH>nmucH muonEmE mo Hones: Hmuou may .mH mHmch mmonoo mcHH zoom mo Hmuou mafia 227 ms. m~.~ m.o I m.o o.~ o.o oomuo>< I I N I I I oH Honuo H H m I H m m ocmHohm H I v I H H HH ucme>ouoEH oHonmmsom H N m I H N v mHuumo I N m I H N v mcH3m N m m I I N m muHsum mSHuHU H v OH I I N H mOHnmumom> I o o I I N v choc mcHooom HHa oHom mooz HHa HHom mooz ooamwwwmmmm omosHa Iooo sHHoouuH> umosHa Imoo sHHoouuH> «\z . H uoond usond mo mocmHzocx HOHHm mo HO>OH Eonooum cH mHanHm>< amnooum cH OHQMHHM>¢ ocoz oO>Hmowm pore mo HHa omosHa page no “HomImco ozone sHHoopuH> conwomwwww MGOHuOsnumcH wo mono can oo>HmomH coHuOsuumcH mo moowH3ocx HOHHQ mo Hm>mH .ow>HmomH cOHvosuumcH mo HO>OH an oo3mH>umucH mumnfiwz nsHU :Iv .mhom mo H02E52II.nH mHnme 228 data is presented in Table 18. Club members interviewed who received no instructiOn or indicated that they knew all that was taught have been regarded as not benefitting. The number of persons benefitting in each area have been divided into two groups - those utilizing their knowledge mainly in club activities and those already having received measurable income increases at the time the interviews were taken. In addition, the members who indicated they would likely find direct practical application for the knowledge in the future were noted. These included those who already had eXperienced measurable income increases and those who likely would in the future (i.e., those that were in a position to make good use of the knowledge). Of course, some of the participants interviewed were in a position where they were using (or would likely use) the knowledge they had gained in commercial ventures whereas others could (or would) probably only use it to further production for their own use. These two categories are distinguished in Table 18. As with the Home Economics Clubs, some difficulties were encountered in valuing the change in income due to instructions. Many of the income increases, e3pecially in the area of vegetables, were measured on the basis of eXpenditures saved or the market value of home-used products. No allowance was usually :made for additional labour costs for products produced for home-use although out-of-pocket costs were deducted. Table 19 summarizes the rankings of the eight areas of instruction made by the members interviewed on the basis 229 .uwxnms on» cH OHmm you mausoonm mo coHuoooono may ou mummmu :mHHMHOHmeeooa .coHuoEdmcoo 080: MOM cuoo ocHsoHo ..o.mv momma Hchmumo mo cOHmH>OHm on» Ou muommu ammo czoan .unmsmu mm; pony HHm oCH3ocx woMOHHm no GOHuOSHumcH ocH>Hmomu uoc mumnfioe mo Hones: man an om3OH>umucH Honesc Hmuou on» can» mmOH mH mHssm com H I N H N Hmayo I I I 4 NH NH moonsm I I I I o o ucofio>ouoEH oHonmmsom oom m N m OH MH mHuumU oo¢.m H N m mH vH mcHsm ooo.oH H H m NH mH muHsum mouuHU mow m m HH HH oH mmHnmumom> omo.H m o o o aH ouoo ocHooom mommmuocH mommwuocH mOHuH>Huo¢ mcHuHm room ocHocmHummxm weoocH Imwwwww on: £30 QMHU CH Iocmm omoru How anmusmmmz III. Hchz Honsdz mmmmuocH mEoocH ocHocmHHmoxM oususm CH mocmHsosx Hmuoa Honccd oomum>¢ Honfisz comuosuumcm mom ocHNHHHuo COHuMOHH as Hmnfisz uomuHo och On mHoxHH HOQEdz .soaH cH muonsos noHo mIv .msom on woumwmo coHuosuumcH mo mmmum unowmmcu mo cOHumNHHHus mo mHm>mHII.mH OHQMB 230 <5f the utility of these areas to them. Seeding Corn, \hegetables, Cattle, and Swine showed up very well in these rankings. As the latter two categories did not show up particularly well in the utilization measures (see Table 18) tfliis was somewhat inconsistent. Part of this appeared to hue accounted for by the fact that the cattle instruction was sometimes supplemented by a credit program with very favourable terms. This enabled some members to own a cow arnd raise a calf for one year. However, they usually had tr) sell the cow to repay the loan and the calf was Often Enald as well because of lack of sufficient land to embark on 51 cattle enterprise of even a few head. Of course, this was by no means the case with all participants in this program almd.two club members interviewed actually had (or their Table l9.--Number of Boys' 4-H Club members in the sample who rated the eight areas of instruction first, second, or third in order of utility for them. V . Not , First Second Third Rated Total \ Seeding Corn 6 5 2 14 Vegetables 4 4 8 o 16 Citrus Fruits 1 l 1 10 13 Swine l 6 1 6 14 Cattle 5 2 6 - 13 Household Improvement - — — 6 6 HYgiene - - - 12 12 Other 1 - - 1 2 ¥ Total 18 18 18 36 90 231 father had) sizeable cattle herds. The possible income :increases in these cases could easily amount to several liundred or thousand pesos annually but, because of the nature (of the enterprises there was no reasonable way to place a single-valued estimate on it. Pacific Coast Programs The programs carried out during 1967 in the Buena- ventura (Pacific Coast) area fall logically into four categories: (a) the water buffalo project (b) technical assistance to individual farmers (c) fomento and experimental work (d) home economics assistance FYDr the purposes of this study the home economics work in tilis region was included in the Home Economics Program and t1”Ills, it need not be discussed again here. The other three a-<=tivities constitute the Pacific Coast Programs. The Caperating costs of these three activities to CVC were eStimated as follows: water buffaloes, 362,646 pesos; 'technical assistance to individual farms, 22,666 pesos; f ..omento and experimental work, 68,000 pesos. Thus, about 80 Per cent of the total cost of the Pacific Coast Programs was attributable to the water buffalo project which was really an applied research experiment. The water buffaloes came to Colombia early in 1967 as part of a joint CVC - INCORA effort to test their 232 usefulness in the heavy rainfall areas of the country. In addition to the above maintenance costs for 1967 a capital expenditure of about $25,000 (U.S.)33 was made to Obtain a small basic herd (about 50 head) and fly them to Colombia. Some ongoing maintenance costs were born directly by INCORA although CVC was mainly responsible for maintenance. During 1967 and 1968 the buffaloes were kept near the Calima River close to Buenaventura under careful observation. Feeding trials and other tests were carried out. If these trials prove favourable more buffaloes may be imported in the years to come and distributed among potential users in the area. It is anticipated that they may be useful in the production of milk and meat, as well as in the provision of power for agricultural tasks. However, as the Colombian environment is somewhat different than that in which the watter buffalo now flourishes, considerable uncertainty exists about this venture. The benefits from this project will not start to occur until a larger herd has been built up (or imported) and some distribution to potential users takes place. This may take between five and ten years and even then, possible benefits appear to be quite uncertain. Technical assistance to individual farmers in the Buenaventura area was quite limited. Half a dozen small dairy farms (about 6 cows each), two small poultry farms, and 3Private conversation with the head of CVC's Agricultural Department. 233 usefulness in the heavy rainfall areas of the country. In addition to the above maintenance costs for 1967 a capital expenditure of about $25,000 (U.S.)33 was made to obtain a small basic herd (about 50 head) and fly them to Colombia. Some ongoing maintenance costs were born directly by INCORA although CVC was mainly responsible for maintenance. During 1967 and 1968 the buffaloes were kept near the Calima River close to Buenaventura under careful observation. Feeding trials and other tests were carried out. If these trials prove favourable more buffaloes may be imported in the years to come and distributed among potential users in the area. It is anticipated that they may be useful in the production of milk and meat, as well as in theprovision of power for agricultural tasks. However, as the Colombian environment is somewhat different than that in which the water buffalo now flourishes, considerable uncertainty exists about this venture. The benefits from this project will not start to occur until a larger herd has been built up (or imported) and some distribution to potential users takes place. This may take between five and ten years and even then, possible benefits appear to be quite uncertain. Technical assistance to individual farmers in the Buenaventura area was quite limited. Half a dozen small dairy farms (about 6 cows each), two small poultry farms, and 33Private conversation with the head of CVC's Agricultural Department. 234 several African palm operations were assisted. Climate ade soil conditions in the area are such that dairy and pxyultry Operations do not appear to be competitive with siJnilar Operations in the Cauca Valley. Several of these operations were visited in the course of this study and timere seemed to be little chance of them developing into comercial farms. Most were owned by non-residents as a hobby. African palm is an important crop in the area and is reasonably well-suited to local conditions. However, owners (If the plantations appeared to be getting most of their technical information from other sources. Fomento and experimental work included the prOpog- Eition of new varieties of fruits and other crOps followed by tfiie distribution of these to the natives. No effort was made txb find out exactly what effects this was having on local agriculture because of the expense that would have been involved in tracking down recipients. Most of the local residents earned their living by working on the large African palm plantations, cutting wood, or gathering wild fruits and Sugar cane. Although they might profitably use the new Varieties in the future, there did not seem to be an easy Way of measuring the extent of this usage. CHAPTER VII ANALYSIS OF CVC'S 1967 EXTENSION ACTIVITIES In this chapter an attempt is made to measure the relative success of the six major extension programs operated by CVC in 1967. The criteria used are those dis- cussed in Chapter V. As a preliminary step the programs are compared by all of the benefit categories which were described in Chapter II. Next, PVB/PVC ratios are calcul- ated for selected benefit categories on the basis of either actual data or stated assumptions. As it was impossible to estimate PVB/PVC ratios which incorporated benefits from all benefit classes, the general comparison by categories serves as a basis for judging the proportion of benefits included in the calculated ratios. Sensitivity analyses are carried out for all programs. A brief discussion of factors influencing the relative success of the various programs is also included. Program Comparison by Benefit Classes Each of the six major programs Operated by the CVC was examined in an effort to discover whether or not it was producing benefits in any, some, or all of the benefit categories described in Chapter II. The results of this 235 236 scrutiny appears in Table 20. A plus sign in any square indicates that the program named in the associated column appeared to be producing some positive benefits of the type indicated in the associated row. Two plus signs appear in what is believed to be the most important benefit category for each program. A negative sign indicates that the program was producing some losses. Some programs resulted in both losses and gains in the same category. (For example, the apiculture program produced income gains for some participant bee-keepers and income losses for others.) A number of the squares in Table 20 have question marks because there was considerable doubt as to whether the program concerned was producing positive, negative, or zero benefits in that benefit category. It should be emphasized that - with the exception of the double plus sign - the table itself says nothing at all about the size of the benefits or who received them. It merely serves as a device to describe the benefit-producing possibilities of the various programs. The large number of zeros in the table should not be interpreted as saying any- thing about the general success or failure of programs. This merely reflects that programs in general were directed at producing specific types of benefits. And, as one might expect, changes in agricultural firms and rural households dominated the picture. The Livestock Program Technical changes on the farms of extension partic- ipants appeared to be the most important source of benefits ¥ _ 2:37 .mucmoHoHuumoIcos co mucouscH qu mussooom cocweocwno consmqu ones . . ousuHsoHuo< uo OumN ooums 0+ OuoN ouoN OuoN uso o>oz Ou oHooom osHumHmm< mucouso . muompcuz mo ucUEOOMHoom oumu ~+ 0+ Ouou ouoN OHON Hausumz ocHupuHHHomm mucosa m+ m+ m+ cums mouoN NouoN mcoHusuHumsH “onuo so mocOsHucH Iowm m+ OHON oumu ~+ OuoN «Oman moousom :OHumshowcH co mucouscH ouou ouoN oumn Ouou OHON OuoN HmHuoomsm: mucmoHoHuumm ouwu oumn ouon onus OHON onus HsoHsnoos Icoz chHusNH oumn m+ N+ N+ N+ NOHON HpHuooscm: mucmoHoHuumm Icsmuo :onco x >UHcsEEOU onus ouou mouoN cums + Cums Hsochooa . u m muoHoom ouoN ouou mouwn cums Noumu OHON HpHuwoocpz mucmoHOHuusm Housm ouou OHON m+ ouoN N+ ouon HsoHsnooa Icoz mpHoa Iomsom OHON OnoN + N+ OuoN ouoN HmHummmsm! mucmoHoHuumm Hausa OHON mouoN ouoN + oumu OuoN HMHuoossmz mucmmHoHuumm come at. Sou I come o+ HaroHssos.H oIsoz mEE moososo HousuHs NouoN ++ OHON + NI+ + HmHumomsmz mussmHoHuupm :onco x IoHumd N++ + oumn I+ I++ ++ HuoHcrooa . u m masuooum mnsHU mocsumHmm< Emuooum och Epuooum smuooum oucsno poocoussH uHco ammoo mIv mOHEocoom IssHm Bush a ousUHsoHad xooumm>HH no wows msouw no onus onHusm .mmom 0&0: HONHHHuHOm Esuoouo >n munocmm mo ucoEmmommc muHuocwm mo :OHumOHuHmmmHu .mwmmmHo uHmmcon up mfimuooum U>U homH mo cOmHummEOUII.ON Hands 238 for this program. Many participants were reported as having little contact with their colleagues and diffusion was probably slow or negligible as a result. However, if it existed, it was likely positive. Because of the technical changes which were occurring and, because of the many direct contacts between participants and extension agents, one would be led to believe that future decisions on participants' farms would be favourably influenced as well. Little carry- over influence on rural households would be expected because many of the participating farmers lived in towns and their farm business was not intimately connected with their house- hold Operations. Some carry-over to community decisions might be possible but this was not considered likely. There did not appear to be any appreciable feedback to research workers in this program although there was a possibility of favour- able influence on the French Mission and the Fondo Ganadero. No other sources of benefits appeared to exist. The Apiculture Program Although some losses occurred on the farms of extension participants as a result of the apiculture program, these were probably outweighed by benefits. Diffusion effects attributable to the program were probably close to zero because of the lack of contact among producers and the limited possibilities of several producers operating in the same community. With expanded honey production there was the possibility that the nutritional content of diets in the 239 households purchasing the honey might be improved. Indeed, the extension workers included eXplanations Of the nutritional value of honey in their promotional efforts. This would be seriously limited, however, by the fact that honey, even at twelve pesos per kilogram, was a luxury in households where diets were poor. One of the most interesting features of this program was the attempt by the CVC to try to develOp the bee-keepers' association and open up international markets. However, the possibility of selling honey to other countries (namely the U.S.A.) did not seem bright, as Mexico and Argentina are large exporters of honey to the U.S. at about $0.10 U.S. per pound.1 To compete with these prices Colombian producers would have to be able to sell honey in quantity at about 4.0 pesos per kilogram. Because of low labour costs in Colombia this might be a possibility in the future (say in 5 to 15 years) but, at the present time, little enthusiasm could be generated for exports at this price. Greater develOpment of local markets seemed to be a very real possibility althOugh this was not being pursued. Indeed, this seems to be an essential step to breaking into the export market as better local marketing would serve to build up larger-volume, more efficient producers who could more easily move towards 1In 1967, Mexico exported 7,084,000 pounds of honey to the United States at an average price of 10.9¢ (U.S.) per pound. In 1968 total U.S. imports were 18,841,000 pounds at an average price of 10¢ per pound. (See U.S. Foreign Agric- ultural Trade by Commodities, Economics Research Service, U.S.D.A.) Most of the I968 imports were produced in Mexico and Argentina. 240 exporting if this became economically feasible. Because of these circumstances, community action benefits were dis- counted heavily and technical changes on the farms of individual participants were considered as being most important benefit category. The formation of apiculture clubs served to increase the numbers of people learning about honey production but, because many of these lacked either interest, time, or resources, or did not live in an area where more bee—keepers could start up an apiary, it did not necessarily increase actual benefits beyond what would have occurred if the extension workers had only worked with existing bee-keepers and sincerely-interested individuals. Thus, diffusion effects have been discounted heavily in this program. The possibilities for benefits from secondary "outputs" also appeared negligible. The Fertilizer and Farm Planning Program The fertilizer program also produced losses on some participants' farms. However, benefits from technical changes on participants' and neighbouring farms were, on the average, positive with the diffusion effect being more important. Some feedback to information sources appeared to have occurred but, as this was in no way organized, its value was considered relatively small. The frequent contact with extension personnel probably had some influence on the households of participating farmers. In some cases, wives of 241 participants became involved in the home economics programs because Of their spouse's contact with the CVC extension personnel who were Operating the fertilizer program. How- ever, all benefit categories but technical and managerial changes on farm firms were discounted heavily. Some attempt was made to determine if managerial changes had actually taken place on participant farms by asking participants if they would be more interested in learning about new technology, co-operating with other agencies, and using fertilizer on other crops as a result of their participation in this program. However, after a number of interviews it was concluded that respondents might well be answering in an affirmative manner to either (a) protect the extension agent, or (b) avoid being in the position of saying anything derogatory about anyone. Thus, the questions were not regarded as useful tests of the hypothesis. Some indirect influence on decisions regarding community activities might result from this program. However, questioning of partic- ipants did not reveal anything concrete here and, as a result, benefits in this category were considered unlikely. No benefits were detected from secondary "outputs". Home Economics Assistance Even though many girls and housewives could not use some of the instruction received, technical changes in rural luouseholds were considered to be the most important benefit «sategory for this program. Diffusion, as of the time of the 242 interviews, was minimal with only a few of the interviewed participants indicating they had carried any information beyond the club to which they belonged. Many indicated they had helped or instructed others in the club but benefits from this were already included in measures taken of club member's benefits. Because of the fairly intense nature of this program and the wide range of instruction given, it was considered that some carryover to future decisions would be likely to result. Thus, a "plus" appears in the Managerial section opposite "extension participants." Some of the instruction was directed at community action and perhaps some benefits would arise in this category. How- ever, because the club members gave this a low ranking, it has been rated as rather insignificant. With the exception of aiding mothers to educate girls in already-known practices and perhaps some indirect influence by informing girls about opportunities in other areas,2 no other benefits were apparent. Boys' 4-H Clubs Because of the relatively large response indicating that the techniques learned through the clubs were of limited direct practical application and the fact that many of the club members were relatively young with much of their 2In one case of record a 4-H Club member learned to sew in club activities and was able to get a job as a seamstress in a local town as a result. A few other similar cases were mentioned by extension personnel. 243 life ahead of them, managerial changes have been considered as more important than technical for this program. Diffusion as of the date Of the interviews was almost zero and, because of the limited continued application of techniques learned by participants, this was not anticipated to increase. However, practical application Of the techniques being taught was by no means absent and benefits in this area from both direct and diffusion effects might be substantial in relation to costs. Few other benefit categories were relevant for this program. The effects on the movement of people out of agriculture were probably negligible because of the concentrated emphasis on agricultural techniques. Perhaps some contribution may have been made in facilitating the natural replacement of managers but this would undoubt- edly be negligible as the extension agents concentrated mainly on techniques new to the area. pacific Coast Programs The highly uncertain nature of the benefits which might accrue from the buffalo project made it difficult to select a most important benefit category for these programs. On the assumption that this project would eventually produce benefits, it would seem likely that recipients of buffalos would be the ones who would benefit and, thus, technical changes in the farm units Of extension participants are regarded as the most important benefit category. Benefits in all other categories are either non-existent or 244 questionable. Calculation of PVB/PVC Ratios As pointed out earlier, the calculation of PVB/PVC ratios requires data on both the costs and benefits attrib- utable to the project being examined. The benefit flow attributable to changes inspired by extension work is determined by (among other things) the "shift" in the adoption distribution attributable to the extension agency. This "shift" is Often difficult to isolate but, in three of CVC's major programs, sufficient data were gathered to facilitate a description of it. In one of these cases - the fertilizer program - the data was easily converted to benefit-flow form. In the other two, lack of adequate price information made this sort of transformation quite difficult. However, by making some additional assumptions PVB/PVC ratios have also been calculated for these programs. For the remaining three major programs PVB/PVC ratios have been calculated by making several sets of assumptions about their total "output". In all cases, the assumptions are formally stated and the set of assumptions believed to be most credible is indicated. It should be pointed out again that the calculated ratios incorporate ggly the value of benefits attributable to technical changes in agricultural firms and rural households. All other benefit categories have been omitted because of the measurement problem. This seriously limits but does not completely destroy the usefulness of 7' 245 the results. In essentially every case the actual PVB/PVC ratio will be greater than or equal to the calculated PVB/PVC ratio (assuming the calculated ratio accurately represents the benefit-flow categories incorporated therein). The extent of this deviation is not believed to be large for the projects examined here. If one accepts the hypothesis that the value of managerial changes is closely correlated with the value of technical changes and notes that, in most cases, secondary extension "outputs" are rather inconseq- uential, the ranking of programs according to the calculated ratios would not be changed by incorporating the omitted benefit flows. In determining the "appropriate" interest rate for calculating the PVB/PVC ratios the following factors were considered. All benefit flows for the projects examined were calculated in current (1967) pesos. Thus, the extension of such estimates into the future would represent real values and a real interest rate would be required. First mortgage rates in Colombia have ranged from 10 per cent to 20 per cent in recent years.3 Industrial bond rates have ranged from 6 per cent to 18 per cent of their original price but have been selling below par so that their effective rates are in the 12 per cent to 20 per cent range.4 During the period 1960 to 1966 the six most active Colombian common 3Revista del Banco de la Republica, December, 1968. 41bid. 246 stocks produced an average annual rate of return of between 10.26 per cent and 15.21 per cent not including capital appreciation.5 Capital appreciation on these stocks has been about one per cent per year. Using the national consumer price index as an indicator, inflation in Colombia has been averaging about 12 per cent per year for the past twenty years although this rate has been some- what unsteady. By deducting the inflation rate of 12 per cent from an indicated current interest rate range of 12 per cent to 20 per cent, one arrives at a real interest rate range of zero per cent to 8 per cent. Thus, a rate of four per cent has been used with PVB/PVC ratios also calculated at the zero per cent and 8 per cent levels to test for project ranking reversals. The Livestock Program As indicated previously, no data was gathered on the actual benefit flow attributable to this project. PVB/ PVC ratios were calculated on the basis of 50 active participants in the program. Three possibilities were con- sidered with respect to the total number of producers influenced by this program: (a) Only active participants would be influenced. 5Alberto R. Musolem, "Dinero, Inflacion y Balanza de Pagos; La Experiencia de Colombia, 1950-1966", Mimeo- graphed Study, Universidad del Valle, 1968, p. 24. 247 (b) Twenty-five additional producers would be influenced in 1968. (c) Fifty additional producers would be influenced, twenty-five in 1968, twenty- five in 1969. Four possibilities were considered for the average annual benefits per producer. The first allowed for 500 pesos per producer which amounts to about 10 pesos per head of livestock carried on participating farms. The second allowed for 2500 pesos per producer per annum or 50 pesos per head of livestock. These were considered to be minimum and maximum likely benefits respectively. The last two alternatives considered were modified versions of the first two but permitted benefits to increase at a uniform rate from zero at the time of influence to the selected yearly amount in the fifth year. However, when PVB/PVC ratios for the latter two alternatives were calculated they were found to be consistently lower by the same small proportion for a given interest rate. Thus, the results for these two alternatives are not shown here. Shifts in the adoption distribution of 5, 10, and 15 years were examined. These were considered to be "minimum likely", "expected", and "maximum likely" shifts although they are, admittedly, rather rough estimates. However, in view of the impossibility of actually observing these values for this project and the fact that little research has been done to determine the length of similar such shifts, there was little alternative to using such "educated guesses". 248 The resulting PVB/PVC ratios for the various sets of assumptions are given in Table 21.6 The wide variability in these ratios indicates the wide range of performance which might reasonably be attributed to this program. Only a long-term detailed analysis of the participant farms could specify this performance more precisely. Thus, it would appear that a reasonable expected PVB/PVC ratio for this project lies between 0.8 and 4.0 with a low of 0.3 and a high of 10.0 being possible. The effects of variations in interest rate, size of the "shift" in the adoption distrib- ution, number of producers affected through diffusion, and average annual benefits per producer are all clearly illustrated in Table 21. The Apiculture Program As detailed data on income changes were not available for most participants in this program, it is not possible to calculate a single PVB/PVC ratio considered to be represent- ative of actual performance. However, such ratios have been calculated under a number of alternative assumptions in order to give some idea of the relationship between costs and pay- offs. Only technical changes in the production units of 6The calculations assume that benefits begin in the year following the one in which the extension expenditure was made (i.e., costs occur at time zero while benefits begin exactly one year later and the total annual benefit is received at year-end thereafter). Any benefits accruing during the year of the extension effort can be assumed to cover adjustment and learning costs on the part of program participants. 249 .wusn ussoomHom om.m mH.s sa.o mo.H ov.H aa.H meow» smouoHo umuoH momma o3» O>HMI>uco39+ m~.v mN.m vo.o vm.o mo.H mm.H momma cos HoumH use» mco O>HMI>ucmza+ Hm.~ mo.~ ~m.m om.o sm.o oo.o mono» o>Ha musooHoHouoo somHm oH.v hv.m no.5 no.0 mo.H mv.H whom» somumHm How» ocHsoHHOh oN.m mm.m mm.v mo.o om.o oo.H mums» spa onu m>HHImwsm38 msH vm.H mH.N mv.N am.o ev.o om.o mums» o>Hm mucmoHOHunmo quHm so.~ ao.m oa.s sm.o on.o oo.H meow» smoumHm . . . . . . m o mucmoHOHuumm mN N mo N Nm m me 0 am o mm o mumo : B muMHm one mm.H mv.H om.H hN.o om.o mm.o mums» m>Hm ppm was mwo omm owe wwo sOHuanHumHo poososHmcH coHuoOpm gnu :H mumospoum umosponm mom mOmmo oomN HOODUOHm pom mOmom oom =HMH£w=oomno>¢ mo umnfisz mo muHmocom Hmscs< MO muHmmcmm Hogans .chHumEsmmm mo mumm pmuomHom Hops: Emsmouo xooumm>HH map How mOHumn U>m\m>m pOUMHSOHMUII.HN mHnme 250 Inogram participants have been incorporated into these calculations for, as indicated earlier (see Table 20), other kmnefit categories were considered to contribute little to the success of this project. Benefit calculations were constructed on the basis of income increases per hive of 10, 50, and 100 pesos per year. Again these are "minimum likely", "expected", and "maximum likely" average increases. The estimates are based on the rather crude data Obtained from the program part- icipants interviewed. Two assumptions were made with regard to the total number of hives; one, they remain constant - two, they expand at a rate of five per cent per year. The latter assumption is perhaps more realistic as some expansion actually did occur in the 1960's. However, past experience suggests that rapid expansion of this industry is not likely unless local marketing channels improve and per capita incomes of Colombians increase rapidly. Three lengths of "shift" in the adoption distrib- ution have been examined - 5, 10, and 20 years. As with the livestock program, these are considered to bracket the actual "shift" but there is really very little basis for concluding that they actually do. The resulting PVB/PVC ratios shown in Table 21 indicate that there is little chance of this program per- forming as well as the livestock program.7 The "expected” 7A3 with the livestock program the benefit stream is assumed to begin in the year following the year in which the extension effort was expended. 251 mO.m mv.v nm.o m>.H NH.N mo.N N0.0 mO.H nH.H upom\m0mom OOH now» you Hm.H NN.N ON.O ON.O OO.H mm.H ov.O Nm.O mm.O umo>\momom Om Nm up on momuoc om.o sa.o ao.o 5H.o H~.o s~.o oo.o oH.o oH.o soms\momoo oH H H hO.H m>.N N0.0 mm.H mo.H HO.N ON.O ON.O OO.H Hmm>\m0mmo OOH Aucmmmuo O0.0 om.H HO.N no.O H0.0 OO.H O0.0 mv.O Om.O Hmom\m0mmm Om mm gammy OHO ON.O ON.O O¢.O mH.O OH.O ON.O O0.0 O0.0 OH.O usm»\m0mmm OH NO NO NO NO NO NO Nm Nv NO m>Hm mm>Hm Moo mo whom» ON whom» OH whom» m muHmmcmm Hmnfisz pOHumm uHmosmm mo numsmq .msOHumEsmmm pquOHmm smog: Emumouo ouspHDOHom any you mOHumu U>m\m>m pwumHsoHMUII.NN mHnme 252 PVB/PVC ratio appears to lie between 0.50 and 2.0 although it may turn out to be as low as 0.1 or as high as 7.0. Indeed, there appears to be a good chance that this program is not returning benefits as large as the costs that have been incurred in carrying it out. The Fertilizer Program The data collected for participants in the fertilizer program were used to estimate two benefit flows - one for the work with corn producers and one for the work with tomato producers. These benefit flows include only benefits con- sidered to be attributable to extension-induced technical changes on the farms of both participant and non—participant producers. No benefits in any other benefit category have been estimated. Table 23 shows the estimated benefit flow for the work done with tomato producers. It has been based on a benefit period of 3.158 semesters for all producers. This was the average time between actual adoption and probable adoption (if CVC's influence had not been present) as estimated by participating farmers. Where available, actual income changes have been used. Where these were not avail- able, an average increase of 1934 pesos per plaza - the average of the 19 participating farms over both semesters - was used. Although this included some farms which experienced losses due to bad weather, diseases, and other causes, it was considered to be a reasonable representation 2553 Table 23.--Estimated benefit flow for CVC's 1967 fertilizer program directed at tomato producers. Number of Number Total Classification Farmers of Value of a of Plazas Income) Period Beneficiaries Adopting Benefiting Affected Changes' Participants 7 7 6.12 5257 1967-1 N°“‘ participants 4 5.17 4441c Total 11 11 11.29 9698 participants 7 14 24.166 544088 Non- 1967’2 participants 12 15.52 349518 Total 15 26 39.68 89359 Participants - 14 18.72 36204 1968-1 ”0“" Participants 20 25.86 50013 Total 34 44.58 86217 Participants - 14 18.72 26892 1968-2 N°“' Participants 3 23 29.74 49100 Total 37 48.46 75992 Participants - 7 13.00 3978 1969-1 N°“' Participants - 19 24.57 30669 Total - 26 37.57 34647 Participants - - - - Non- 1969'2 participants - 11 14.22 10671 Total - 11 14.22 10671 Participants - - - - Non- 1970'1 Participants - 3 3.88 1187 Total - 3 3.88 1187 aAreas of tomatoes in non-participating farms have been calculated by assuming that they would have the same average as the 19 participating farms - 1.293 plazas. bBeginning in 1968, an average of 1934 pesos per plaza is used. However, to approximate an average benefit period of 3.158 semesters, benefits for farms in their fourth semester were reduced to 306 pesos per plaza. cBased on 859 pesos per plaza which was the average income increase per plaza for participating farms in that semester. dIncludes non-adopting participants. eIncome increases for farms where this information was not known were based on 2252 pesos per plaza which was the average income increase per plaza for known participating farms in that semester. 254 of the reality which all farms would face. All first semester participants said they would continue to use the new techniques in the future and were therefore considered to be "adopters". Five out of the six tomato producers who participated in the second semester and did not receive any positive benefits said they would not use the new practices in the future. The sixth attributed all of his loss to the weather and said he would continue to use the new practices. He is therefore regarded as an adopter but the other five are not. None of the neighbours who learned from these five producers were considered to be "adopters". Thus, the total number of adopting participants were estimated at fourteen. All those farmers who learned about practices from this group and used them during the semester of participation and the two semesters immediately following were also considered to be adopters. Therefore, the total number of farmers influenced was 37 and, in total, almost 50 plazas of tomatoes were estimated as being affected. The total extension costs of the work with the 19 tomato producers were estimated at 26824 pesos in the first semester and 57480 pesos in the second semester.8 The total value of income changes shown in Table 23 is considered to be the total benefit flow attributable to technical changes as the tomato producers affected were not 8The costs were based on an average cost of 3832 pesos to service a producer for one semester. 255 considered to be a large enough group to influence market price. It should be pointed out that this benefit flow assumes no change in market prices and that the producers affected continue to grow approximately the same acreage as they are now growing. Both of these assumptions are open to question. To get some estimate of the effect on PVB/PVC ratios should these not be valid, ratios were calculated with benefit flows (beyond 1967) at 75 per cent, 100 per cent and 125 per cent of the flow appearing in Table 23. As the violent price fluctuations in tomatoes mentioned earlier tended to be seasonal, this was considered to be an adequate sensitivity test. However, if low (or high) prices were to prevail for one or two years, the benefit flow for this project could be completely eliminated (or, conversely, increased to several times its estimated size). Because of the lack of data on diffusion, detailed benefit flow calculations were not possible for the fertilizer program with corn producers. However, as some diffusion was noted, it was decided to make some assumptions and construct such a flow. Table 24 shows the results of these assumptions. A three-year adoption "shift" and a total adoption group of 35 farmers is assumed. Based on the data collected from corn-producing participants (see Table B-6) an average income increase of 500 pesos/semester was used. 9The average income increase for participants in tflieir semester of participation was 535 pesos. 256 Table 24.--Estimated benefit flow for CVC's 1967 fertilizer program directed at corn producers. Period No. Of Farmers Estimated Total Value Adopting Benefiting of Income Changes 1968-1 9 9 6,087 1968-2 6 15 10,736 1969-1 5 20 10,000 1969-2 5 25 12,500 1970-1 5 30 15,000 1970-2 5 35 17,500 1971-1 30 15,000 1971-2 25 12,500 1972-1 20 10,000 1972-2 15 7,500 1973-1 10 5,000 1973-2 5 2,500 As in the tomato program this assumes that market prices and volumes for producers do not change significantly during the time period in which they are affected. However, possibilities for wide divergences from this assumption seemed very remote in this case. Again, a sensitivity analysis was conducted by calculating PVB/PVC ratios with benefit flows (beyond 1967) at 75 per cent, 100 per cent, and 125 per cent of the estimated benefit flow. Costs for work with corn producers was estimated at 33,530 pesos for the first semester and 21,555 pesos for the second.10 A summary of the calculated PVB/PVC ratios for the fertilizer and farm planning program is given in Table 25. 10The costs were based on an average cost of 2,395 leesos to service a producer for one semester. 257 Table 25.-~Ca1cu1ated PVB/PVC ratios for the tomato and corn sections of CVC's 1967 fertilizer program. Tomato Corn Program Program (a) At 75% Of Projected Benefits (after 1967) (i) Discounted at 0% 12.50 2.10 (ii) Discounted at 4% 11.93 1.89 (iii) Discounted at 8% 11.39 1.71 (b) At 100% of Projected Benefits (i) Discounted at 0% 15.54 2.80 (ii) Discounted at 4% 14.81 2.52 (iii) Discounted at 8% 14.12 2.28 (c) At 125% of Projected Benefits (after 1967) (i) Discounted at 0% 18.59 3.50 (ii) Discounted at 4% 17.69 3.15 (iii) Discounted at 8% 16.84 2.85 It is fairly obvious that both of these programs were returning discounted benefits much in excess of theircost with the tomato program producing almost spectacular results. However, it should be pointed out that, because costs and benefits were calculated on a semester basis and benefits were regarded as occurring in the semesters in which the costs were incurred, the calculation procedure might legitimately be regarded as being slightly biased in favour of this program.11 11In the author's opinion this merely reflects the fact that this was a “quick-return" program relative to most (others and that the assumptions of little or no return for 'the program year of most other programs is reasonable. 258 Home Economics Assistance Any sort of benefit-cost calculation for a project such as this must necessarily be open to a good deal of criticism - even if excellent data are used. Although the data gathered in this study were collected from individual interviews with a sample of program participants, a number of possible sources of bias and measurement error exist. There is not much one can do to correct for these and thus, the data must be regarded as rather crude. Therefore, there are some strong arguments for not making benefit-cost cal- calculations at all. Nevertheless, it has been decided that the pressing need for some sort of estimate outweighs these arguments. In other words, it is considered that this is a situation where 52mg bad information is better than 22 information. Readers may, of course, choose to disagree. In order to calculate PVB/PVC ratios an estimated benefit was constructed by assuming that: (a) benefits from diffusion would be zero. This was based on the relatively small amount of diffusion detected during the interviews, (b) average benefits per program participant could be reasonably estimated by multiplying the proportion in the sample "adopting"12 a given type of knowledge at a given level by the average benefits per "adopter" and summing this for all benefit categories (see Table 26). 12In this instance the term "adOpting" applies to those who were considered likely to make some use of knowledge acquired in the program. PrOportions were derived HuHmom mHomHo>mm .chunmocs mHsoHc >Hnmnoum eron usn O.H cmsu mmmH anmnonm unmou OHMHOOO NO.mH Nm.H OH.O NO maHanoo mHoono>om ma.OH OO.H NH.o Na nosHo mIa .nsom NO.NN OO.H HO.N NO OO.HH HO.H O0.0 NO oaHanon owmmmmmoo aH.mH mo.~ om.s as noHsocoom osom . N0.0N hN.N O0.0 NO ON.N HO.H ON.N NO Acuouc HHz umOEH< mHnmuo>mm OH.O OO.H Nm.N Nv HmNHHHuHmm Om.m OH.N OO.N NO . . O0.0H OO.HH NH.OH NO HHz unosHa Hnsosoz ao.sH mm.HH HO.aH Na Amwmwmmmwwa mm.OH om.NH cm.mH NO . . mO.m O0.0 O0.0 NO HHz unoch Honosoz ma.a oo.o oo.H as onsuHsoHoc hm.w OH.O OO.H NO Om.m nN.O OO.H NO HHZ mHnmuo>mmco OH.> ON.O OO.N NO xooumm>HH N0.0 ON.O OO.N NO moHumm U>m\m>o muommmm moHumm OHumm mumm OH pmpsHocH soHuanuumHo U>O\m>m mo u>m\m>m unnoomHo uoc muHmmcmm mEoocH mocmm mHnmnouo pmuommxm .mEmHooum sOHmcmuxm m.o>o How mmusmmme mocmEHOMHmo mo mumEEsmII.Hm mHnme 270 If benefits accruing in the Livestock program were to be passed on to consumers through a shift in supply, the income distribution effect of this program would still likely be unfavorable because the poorer classes do not purchase much meat or milk. On the basis of the data con- tained in Table 31, it would appear that a reasonable program ranking would be as follows: 1. Fertilizer Program with Tomato Producers . Home Economics Assistance . Boys' 4-H Clubs . Fertilizer Program with Corn Producers . Livestock Program . Apiculture Program \lO‘U’IIwa . Pacific Coast Programs Of course, some of the above rankings could change depend- ing upon how one weighted risk and income distribution effects against expected pVB/PVC ratios. It would be relatively easy, for example, to rank Home Economics Assist- ance first because of the large number of people benefiting. Factors Affecting Differences in Program Performance No complete explanation of the program results can be advanced here. The results themselves have not been sufficiently well-specified to attempt a complete analysis. Nevertheless, a comparison on several points may be useful. The points chosen for this purpose are: (a) cost differences, (b) unit benefit values, (c) diffusion effects. Cost Differences 271 program costs can be compared on the basis of (a) total extension costs per day of field personnel time, and (b) total extension costs per unit of output using a proxy measure of output where this is available. The extension costs per day of field personnel time as shown in Table 32 reflect differences in classes of personnel employed, amount of travel required, and intensity Table 32.--Comparison of several unit measures of total costs among 1967 CVC programs. Total Extension Costs Per Unit Total Extension Costs per Day of Field Personnel Time for Selected Proxy Measures of Output (pesos) (pesos per unit) Livestock 440 7330/farmer 93 Program 75/head of livestock Apiculture 453 4980/bee-keeper Program 93 498/colony Fertilizer and 3832/tomato Farm Planning 387 producer Program 2395/corn producer Home Economics 239 11777/c1ub 93 Assistance 551/member Boys' 4-H Clubs 302 26950/club 2E l789/member pac1f;c Coast 237 N/A rograms Other Programs 375 N/A 272 of use of materials and supplies. Variations in the amount of administration and supervision among programs are not reflected as these were included in total extension costs by pro-rating them on a prOportionate basis. Most of the differences in costs per day of field personnel time shown in Table 32 are attributable to differences in the classes of extension personnel employed. The only exception to this was the apiculture program where, as noted previously, about 100 pesos per day was attributable to higher than usual travel costs. As one might intuitively expect that the higher-paid, better-trained personnel would be capable of producing more in a day, there is little cause for con- cern in this observation by itself. However, it should be noted that the several classes of personnel employed did not represent different levels of capability in performing the same work (i.e., the agronomos did not do what the mejoradoras did - they per— formed quite different functions). Indeed, it would have been difficult for one type of personnel to perform the functions of another type. Thus, unless one examines the outputs of the various classes of personnel, it cannot be assumed that their marginal value products were, in fact, different. This is rather difficult to do given the data available. However, as the programs which ranked highest in performance employed predominantly non—professional personnel, the hypothesis that professionals were overpaid 273 relative to non-professionals does suggest itself. As no physical measure of output was available in any program, proxy measures were selected to permit cost comparisons. These measures must be used with extreme caution because they do not allow for some of the classes of benefits which perhaps existed in the various programs. (For example, using an estimate of extension costs/bee- keeper implicitly assumes that output varied with the number of bee-keepers serviced. This is not strictly true because of the possibility of benefits resulting from extension activities with the Bee-keeper's Association.) Neverthe- less, these measures can provide some means of comparison. It is obvious, for example, that the Apiculture program has to produce benefits per participant much greater than the Livestock program if it is to be justified. As the monetary volume of business per participant is much greater in the latter case, one would intuitively question the feasibility of this. (Of course, such comparisons tell only part of the story. High costs per program participant are warranted if high levels of benefits per participant can be Obtained - as in the Fertilizer program with tomato producers.) The data on costs per participant illustrate fairly well some of the rather difficult choice problems that extension agencies face. In terms of costs, two Home Economics Club members could be serviced for every Livestock farmer whose service was discontinued. Larger farm operations do not 274 cost a great deal more to service than small but, in the former case, monetary benefits are likely to be greater and income distribution effects are likely to be less favorable than in the latter. Benefit Values per Unit of "Output" A variety of factors appeared to be influencing benefit values per unit of output both within and among programs. These cannot be readily classified and must be looked at program by program. Data from Home Economics club members were gathered in an effort to determine limiting factors for this program. A s-mmary of these is given in Table 33. "Resources" (i.e., income and/or possessions) and "interests" were the most often-mentioned factors. Because of the heavy Table 33.--Incidence of mention of various factors considered to limit utilization of knowledge in the home economics assistance program. ngigtgg Number of Members Mentioning . Under- None Resources Time Interest . Instruction standing Clothing 1 29 2 2 2 Health 4 6 - l4 7 Hygiene 4 i 12 - 13 3 Food 2 29 - 6 3 Home Improvement 1 22 2 7 2 Small Industry 1 1 27 l Acci6n Comunal 1 23 0 Gardens 12 12 2 8 5 Total 26 119 8 100 23 275 emphasis on "resources" one is led to believe that this program, even though it appeared to be well-suited to low income families, might have resulted in a little higher level of utilization if it had been directed at groups with slightly higher incomes. (The lowest-income families in the sample were usually slower to adOpt recommended practices.) The heavy emphasis on "interest" indicates that the practices being extended did not offer big pay-offs to the families concerned. This is supported by data on benefit measures. Even benefits of 500 pesos per family per year could not be regarded as large in relation to family income or levels of wealth prevailing in other countries. (Many of the families interviewed seemed to realize their relative state Of poverty.) Furthermore, as most of the benefits came in non-monetary form they could not be readily exchanged for preferred goods in the market place. The low incidence of "understanding" being mentioned as a limiting factor indicated that extension personnel were getting their message across. Although similar data were not tabulated for other programs, it appeared that "resources” and "interest" would be important limiting factors in the Boys' 4-H Clubs as well. The programs directed at livestock producers, bee- keepers, tomato producers and corn growers were somewhat different, however. In these cases, size of business, risk, relevance of the technology to the Operation, and extension 276 effectiveness were all of relevance. In the apiculture program, for example, some losses were incurred because of incorrect advice. On most farms benefits would have been greater if the farmer had been in charge of a larger bus- iness. In the fertilizer program, CVC supplied fertilizer on credit but the farmer had to bear a risk he would not have had to bear if he had not used the fertilizer. (Benefits to the fertilizer program might have been much lower if CVC had not supplied the fertilizer on credit.) In all of these programs some farmers were encountered who felt that some of the recommended practices were not relevant for their Operation. However, this did not appear to be a major factor as extension workers seemed to be fairly capable of selecting the set of practices which would work on a particular farm. Nevertheless, extension personnel did not appear to have readily available a large reservoir of technology which they could usefully promote. One got the impression that while they would not recommend practices which they felt wouldn't work, they had to put forth a good deal of. effort to come up with workable sets of recommendations. Diffusion Effects Relatively little diffusion of knowledge from program participants to others in the community was dis- covered in the interviews conducted in this study. Cer- tainly some evidence was found of this in a few programs 277 but, on the whole, it was rather difficult to detect and measure. Several explanations of this might be advanced. Possibly this process is rather slow in the areas and situations examined in this study and only several surveys taken a few years apart would detect it. Perhaps there is not sufficient social contact in some communities for such a process to exist except on a very small scale. Perhaps the sort of technology which was being extended was not easily communicated. In one case (i.e., the apiculture program) diffusion seemed to be almost impossible because of the nature Of the business. Another reason for the lack of noticeable diffusion might have been that programs were not directed at natural innovators or leaders. If diffusion could have been stimulated for programs such as Home Economics Assistance and Boys' 4—H Clubs, returns for these programs might have been much higher. Without diffusion effects the Fertilizer program would have received a much lower rating than it did. Per- haps more diffusion might have been stimulated for other programs if the extension service had made greater use of the mass media, especially radio. However, only experiment- ation and further research can lead to an answer in this area. Cost Minimization This section is directed at answering the question "Could the CVC have produced the same results at 278 lower total cost in 1967?" As there were probably an infinite number of ways of producing approximately the same results, it is necessary to approach the problem under two general sets of assumptions. The first of these assumes that personnel and extension media are fixed while Office locations and transportation methods are variable. The second permits extension personnel and media to vary while considering office locations and transportation methods as fixed. Possible Cost Savings by Changing Office Locations and/or Transportation Methods The percentage distribution of field personnel time by type Of activity and class of personnel is shown in Table 34. If "productive field personnel time" is defined to exclude those activities numbered 2, 3, 6, and 7 in this table,16 it is possible to define two possibilities which would have resulted in lower costs: (1) any different system of transportation and/or Office location which would have reduced costs per ~day of "productive field personnel time" and still permitted "productive activities" to be performed as they actually were performed. (2) any saving in time spent "working with program participants" which did not adversely affect the level of "outputs" actually produced. 16As thus defined, "productive field personnel time" includes time spent performing both "management" and "program Operation" functions. As cost calculations also include costs of "management", the conclusions reached are not wrong but in possibility (1) of the first set of assumptions they actually reflect minimizing "program Oper- ation" and "management" costs together rather than "program operation" costs alone. "‘ 279 o.o0H o.OOH o.o0H o.o0H o.OOH Hmuoe N.H O.H O.H O.H 4.H uosuo .a O.H m.o s.m I O.o ochHone .o O.H I N.H o.m o.~ name no Home pom 65H» umoH .a o.~ o.N H.o I o.a mono mstHm .o m.a O.H O.H o.v o.m noncomm osHuHus .m N.N H.H O.m o.m m.O nssnooun osHscoHn .a N.N O.H O.H I O.m xuos ou mumruo mcHum>HHmo .m m.o~ a.m~ H.HN m.NN s.NH ocHHHm>one .N H.oo o.oo a.om m.ao s.mm nucooHoHonoa anemone coHs ocHxnoz .H Hmuoe moumuno mmHmOMMmm Mmmom moHumcHHmum> mOOHuomum me one monOpmuoflmz mOEocouwd .mOH>umm soncmuxm m.0>u How HmcsOmHmo mo mmmHo an mmHuH>Huom msOHum> OH usmmm mEHu mo coHuuoooumII.vm mHnma 280 Second, it is assumed that transportation systems and office locations are fixed, so that the possibility of using different combinations of personnel and communications media which might have produced essentially the results can be considered. Costs per day of "productive field personnel time" could have been reduced if all personnel had driven them- selves and chauffeurs had not been used. Where chauffeurs were used, they increased costs per day significantly (i.e., apiculture program). However, as the extension service employed only two chauffeurs essentially full-time in 1967, the total savings here would have been only about 36,000 pesos. Some additional costs might have been incurred by providing more dependable vehicles - particularly for the mejoradoras - or by having extension personnel become more involved in taking care of vehicle problems. It might also have been necessary to replace any personnel who were not willing to drive themselves. Thus, as the question of chauffeurs is obviously related to the transportation problem as a whole, it is perhaps better to consider it as such. There was only one vehicle available for every 2.6 extension field workers in 1967. Furthermore, these were almost all five years Old or Older and subject to the usual problems encountered in vehicles of that vintage. They were well-suited to the job but were Often out of order for several days or weeks at a " 281 time. Thus, some time was spent by extension personnel fixing cars, delivering others to work, and doing nothing because transportation was unavailable. There was noticeable dissatisfaction among extension personnel with the situation. Thus, the question arises "Would a better transportation system as a whole have been justified?" It appears that a regular trade-in program would have alleviated some of the problems by ensuring that not all vehicles became old simultaneously. More vehicles would have reduced unproductive time. An additional six vehicles would have provided one for every professional field worker and one for every three mejoradoras and/or pgacticos.17 This would have increased total vehicle Opeation costs by about 158,400 pesos.18 A regular trade-in program for existing vehicles would have increased capital requirements by about 30,000 pesos per car or 450,000 pesos. This would have added an annual charge of 90,000 pesos for interest and perhaps reduced repair costs by 0.01 pesos per kilometer or 30,000 pesos. Thus, with chauffeurs eliminated, the total additional costs Of this would have been higher but if some or all of the time saved could have been devoted to "productive work", costs per day 17This system is selected for comparison because the agronomos had the largest percentage of unproductive time and they also had the highest cost per day of total time. 18Annual depreciation: 5,000 pesos (Purchase price 60,000 pesos; salvage value 40,000 pesos). Interest - 20% of 50.000 = 10,000. Annual Operating Costs: 11,400 pesos (20,000 kilometers at 0.57 pesos per kilometer). Total Cost per Car = 26,400 pesos. I _ "" 282 of "productive field personnel time" might have been lower. To arrive at estimates of these costs the added costs of the suggested transport-system have been divided among classes as follows: agronomos, 85 per cent; mejoradoras, 5 per cent; praCticos, 5 per cent; obreros, 5 per cent. A comparison of actual costs per day of "productive field personnel time" of the suggested transportation system is shown in Table 35 under two assumptions - (1) all time spent in activities 3, 6 and 7 would have been changed to productive time, and (2) 70 per cent of time spent in activities 3, 6, and 7 would have been changed to productive time. The latter assumption is believed to be the more reasonable. Table 35.--Costs per day of "Productive Field Personnel Time" by class of personnel for the existing trans- portation system and an alternative. Alternative System Existing System Assumption #1 Assumption #2 Agronomos and Veterinarios 598 583 605 Mejoradoras de Hogar 292 291 293 Practicos 1Agricolas 348 342 346 Obreros 270 265 270 The results shown indicate that costs per day of "productive field personnel time” could have been slightly reduced if the first assumption was valid. If only 70 per 283 cent of unproductive time could have been converted to productive time, daily costs for agronomos would have been slightly higher. The author is inclined to believe that the slight increase in costs would have been more than compensated by better personnel morale and higher resulting productivity. However, there is some question as to whether the extension service could have adopted the suggested systems independ- ently of the rest of CVC. This restriction and the question- able availability of capital - even at 20 per cent interest - would be important considerations in deciding whether or not to change. Costs per day of "productive field personnel time" would also have been reduced if time spent travelling (activity #2) had been lower. Travelling accounted for 20.3 per cent of all field personnel time and, while field workers may have done some program planning while travelling, it was to a large extent "unproductive." There were three possible ways to reduce the amount of travelling time: (1) Operate the same programs in areas closer to the office, (2) expand the number of Office locations so that personnel would be closer to the field areas, (3) utilize extension methods which achieved the same result without as much travelling (e.g., radio). (Nae first alternative was not feasible. Areas close to the cxffice were either already being serviced or were not agric- ILLtural. Workers in some programs (e.g., Apiculture and Ixtvestock) had to go far afield to find extension 284 participants. It appears that there may have been some gains to have been made by increasing the number of office locations. As noted earlier there was a fairly high concentration of field personnel per office - this amounted to a little over seven man-years on the average. This meant that some workers were not located in the centre of the area they serviced but outside of it. However, changed office locations would probably have required more vehicles and created difficulties in getting personnel willing to live in smaller towns. Both of these problems would have entailed increased costs. In addition, some office maintenance costs (e.g., rental, secretaries, etc.) would have been higher. Furthermore, field workers from outlying Offices made frequent trips (once every week or two) to Cali for planning purposes and, if other offices had been opened, time spent making these trips would have increased. Travelling time could have been, reduced by Opening one or two more Offices for workers in the Home Economics and Boys' 4-H Club programs. Offices were already more or less Optimally-located for workers in the rest of the programs. Given these circumstances it appears that, on the whole, little or no cost reduction could have been achieved by having more offices. However, for other sets of programs this conclusion would not necessarily 19 éipply. Also, the larger number of offices with fewer x , 19As will be seen in Chapter IX, extension workers lei-n Antioquia spent 18.3% of their time travelling and, because of conditions, their average Speed was much lower 285 field workers per office might very well improve the reputation of the extension service and make it easier for rural peOple to seek out extension workers for advice. Almost no rural peOple visited CVC's field offices in 1967. Thus, while there was probably no advantage to having more field offices for 1967 programs alone, it definitely appears as if more offices with fewer field workers per Office could be Operated for the same total cost and produce more long-run benefits. Returning to the classification Of extension activities shown in Table 34, it seems appropriate to con- sider if any saving could have been made in time spent "working with program participants" without reducing the "outputs" actually produced. There were some Obvious possibilities here. Firstly, all work which resulted in zero or negative benefits might have been eliminated, if it could have been detected in advance. This was primarily a problem of whom to work with although in one case it appeared to involve who should do the work. The method of selecting farmers for the fertilizer program was an obvious attempt to select participants who had a high probability of adopting recommendations. As previously noted, the location than that of CVC. Thus, for certain sets Of programs, CVC's per cent of time spent travelling could probably be reduced significantly. If it were cut to 10% and total costs per day of field personnel time increased only 5% (for added Coffice costs, cars etc.) over and above those increases under assumption #2 of the suggested transportation system, costs £>mH mmuHmmcHB some on Home no seem on mos.om~ onsoso OHOH HNH ONmH th ONNH mmH mmH.mmH HHmmum Hm NHHEmm Hmnssz >HHEmm Hmnssz NHHEmm Hmnfisz \umou \umoo \umou wumoo uOHHumHo concmuxm mmHHHEmm mmHHHEmm mmHHHEmm Hmuoa ocmudeOHunmm some ucmOHmmm .mHsgoHucd mm mm mucmm can OEHsOumO saw no muouomm me» an om0H>Hmm muOHuumHo me man MOM NHHEmm mo mama so mamuoono consmuxm OOOH you MHHEmm umd mumoo sOHmcmuxm momnm>m pom .mmHHHEmm MO Hmnfisc .mumoo concmuxm HmuOBII.Om mHnme 308 in the district but these families do not necessarily farm or own land. (Some land in the district is owned and farmed by non-residents.) "Farm families" include all resident families with land. Participating families include those families who participated in extension programs at some time since the inception of these programs in the district. (Almost all of these families participated in 1967.) Part- icipating families are all essentially resident farm families. As can be seen, costs per participating family varied con- siderably from district to district. This variation may be either above or below the actual variation because of the method of calculating costs. Most of it appears to be due to the variation in percentage participation among farm families as costs per farm family are fairly stable. The Method of Estimating Program Benefits 16 of the participating Approximately one-third families in each of the districts of Tinajitas, Buenos Aires, Pinguro and Urquita were interviewed in an attempt to estimate the benefits attributable to 1967 extension programs in these respective areas. A list Of extension programs Offered in the districts and the names of families participating in each program were supplied by the extension service. As extension records were not available on an annual basis, a few of the families named by the local offices did not participate 16Samples were selected at random. 309 during 1967 but had received assistance at other times. Questionnaires were completed only for families who had participated in 1967 programs. Two families out of the 58 interviewed did not participate in 1967 because they did not feel they had benefited from the extension services programs in previous years. Two more indicated they had received no assistance although the extension service indicated they had. Out of the 58 interviewed nine participated in 1967 who had not participated previously. Of these, four did not know about the extension service previously and five knew about it, but either did not believe that it could help them or professed they had no time to participate. These data are presented by district in Table 39. During the interviewing it was found that a number of families did not recall having received any assistance in some of the areas in which they were named as participants by the local extension office. Discussions with extension workers revealed that families were sometimes recorded as participating in programs when they actually only attended a few meetings at which the tOpic was discussed. Therefore, where families could not recall the assistance after repeated questioning it was assumed that the assistance given was too superficial to be of any direct practical usefulness to them. In most cases considerable difficulty was encountered 310 me NH 3 OH om oHosom sH umnEOO Hmuoe me O OH O OH muommn pOm OOOH OH OOHumoHOHuumo smnssz O H O O N muowmn no: use OOOH OH OOHumoHOHuumo umaEsz N O O H H OOOH OH uOO use OOOH muommn OOHudeOHuHmd HmnEsz N H O H O muommn no NOOH OH OOHummHOHuumd HOO HmnEOz O N N N H omsmH>umuOH uOO umnssz Hmuoe wuHOguD OHOOOHO mmHHO mOOmOm mmanmOHB “KL“ “urn-“Ii L .ONOHHumHo HOOH mnu OH mmHHHEmm mHoEmm OOOEm OOHumoHoHuumd OOHmOmuxm mo mumEEOmII.mm mHamB 311 in determining what actually happened in 1967. Many projects had been started in 1965 or 1966 and were contin- uing at the time the interviews were made. Thus, while it was evident that certain practices had been learned from the extension service, it was sometimes difficult to say whether they were being used because of 1967 efforts or because of extension activities in 1965, 1966, or 1968. It was there- fore decided to estimate the results of all extension efforts from the time the extension service began working in each of the districts until the end of 1967.17 Estimates of income increases attributable to changes in production activities made as a result of extension programs were constructed for each family interviewed. In general, these estimates were made on the basis of the results achieved in 1967 and/or 1968. Gains and losses which appeared to be of a short-term nature were ignored.18 Additional amounts of inputs were valued at their cost in the local markets whether they were donated by the extension 17The problem of distinguishing results attributable to a specific year's programs could have been avoided initially by attempting to examine the total extension effort in each district. However, an attempt to avoid this approach was made because the costs of programs in the various districts in previous years were not known. As will be seen later, only rough estimates of these could be made. 18For example, one family tried raising rabbits in 1967. They received 240 pesos in revenue from this source with no cash expenses and very little extra labour. However, the rabbits destroyed their garden and were not kept around the next year. Neither the income from the rabbits nor the losses of the vegetables were considered in income estimates which appear in this study. 312 service or purchased by the farmer. Additional amounts of owned inputs were valued at opportunity costs. Products were valued at market prices to the farmer or, in those cases where they were consumed on the farm and the families would have otherwise had to purchase them, they were valued at what the families would have had to pay for them. Some wide discrepancies were noticed between the acquisition cost and salvage value of some products. All of these efforts were directed at constructing an estimate of those increased incomes which resulted from previous extension programs Egg which would continue to be realized by the participating families if the extension services were withdrawn from the community. The fact that families participating in extension programs received a variety of gifts from the extension service is a rather significant one and creates an unusual valuation problem. If these gifts had been less extensive they could have been regarded as educational materials and part of the cost of educational programs. However, they ‘were quite widespread and of considerable cost to the extension service. They included such a wide range of items as garden seeds, fruit tree seedlings, rabbits for breeding, free use of sewing machines, assistance in preparing and maintaining gardens, construction of rabbit hutches and pig pens, and repair or remodeling of houses. Indeed, 'virtually all of the efforts of the Obreros employed by the extension service were directed towards these ends. 313 Typically, the participating family would supply some materials and the extension service would send a man to do some or all of the work involved in the construction project. In many cases, it appeared as if the participating family would benefit by having a better house, a new rabbit hutch, or some other item, but would not really learn very much from the extension eXpenditure. Thus, the "product" to be valued was essentially the resultant physical facility - not the improved capability of rural people. These practices made it difficult to distinguish why people were co-operating with the extension service and whether or not they were actually learning something useful. One housewife commented that the practico came and cared for the garden but didn't teach the family anything. (The family harvested and consumed the vegetables but did not seed a garden the following year.) One family received a gift of five bottles of honey from the extension service. They sold four and one was broken accidentally. It was difficult to see how any lasting benefits could arise from such activities and thus, in such cases, it was assumed that the participating family had co-Operated for the purpose of receiving these gifts and that they gained nothing beyond the gifts themselves. Accordingly, benefits could not exceed the cost of the gifts. In the cases where some physical facilities were constructed for a family, there were Obviously lasting benefits. In some cases these were accompanied by increased 314 human capacities although it was not clear that the physical facilities were a prerequisite for the increased capabilities. For example, was it necessary for the extension service to construct rabbit-hutches in order to get families started raising rabbits? In other cases (e.g., housing improvements) no increased capabilities among participating families seemed possible. Thus, in many of these cases as well, benefits would not likely exceed the cost of the gifts. Indeed, they might be less because the cost of construction by the extension service appeared to be greater than in the local markets. (Some interesting side effects also resulted. For example, a few local tradesmen complained they were being forced out of work.) Attempts were made at the time of the interviews to get some idea of possible diffusion effects and the probable "shift" in adOption attributable to the extension service's activities. As in Valle del Cauca, considerable difficulty was encountered in obtaining such information and the analysis (see Chapter IX) relies heavily on assumptions in this regard. The District of Tinajitas19 Tinajitas is located about 35 kilometers from the Santa Fe de Antioquia Office on the Medellin - Turbo road. 19Much of the background information in this section was drawn from the "Socio-Economic Study of the District of Tinajitas" constructed by the Extension Service of the Department of Agriculture of Antioquia, July, 1967. 315 It can be reached by car from Medellin in about five hours. Its characteristics are described in some detail here to provide the reader with an example of the environment in which the extension service works. Physical Resources The principal asset of the community of Tinajitas is the land. Like many parts of Antioquia it ranges from hilly to mountainous and is badly eroded. Approximately 50 per cent of the land in the district is cultivated. The remaining area is devoted to pasture and trees, or is merely rugged, unusable terrain. In addition to the topographical problems, much of the land is low in nitrogen, phosphorous, or potash and some of it is gravelly as well. The altitude of 2300 meters is responsible for the average temperature of 17°C which varies little throughout the year. Like Valle del Cauca there are two fairly distinct rainy seasons each year; The distribution of land among the 161 families in the district is shown in Table 40. Although almost all families depend on agriculture for their livelihood, almost one-third of the families control no land at all and one- half of them control less than two hectares per family. Some of the operators of the smaller farms work for larger land owners to augment their income. Almost all farms are controlled by ownership as opposed to rental arrangements. Most farmers live on and manage their own farms. The owners 316 of larger holdings employ some of the men who have no land. Other landless families gain their living from store keeping, transporting goods, or performing other such functions. Table 40.--Distribution of land among the families of Tinajitas. -——— w Farm Size Number of % of % of (Hectares) Families Families Area 0 28 27.7 0 0 - 2 51 50.5 20.0 2 - 5 9 9.9 10.9 5 - 20 10 9.9 37.8 20 - 40 3 3.0 31.3 Total 101 100.0 100.0 Houses of very poor quality constitute the only buildings of significance in the area. Land improvements in the form of clearing and breaking form the balance of the investment in real estate. At the time the extension service conducted its socio-economic investigation there were 190 cows, 1,511 chickens, 90 pigs, 50 horses, and 9 mules in the district. These were of poor quality and local origin. They were distributed among the families in a similar fashion to the land. Machinery is almost non-existent although hand tools of various types are found in all homes. These include nechetes, hoes, and similar items. Approximately 20 per cent of the pOpulation are 317 males between the ages Of 15 and 60. This group might be regarded as the community's work force, although almost everyone is employed at least part time at one task or another. The extension service estimated that a typical family might earn 1,916 pesos20 a year in addition to the products it consumed from its own farm. Approximately one- half of this comes from crop sales and one-half from wages earned on other farms. Day wages ranged from 4.50 to 7.50 pesos in 1965. Public resources in the area are low. Electricity is available to a few houses. Telephone and other public utilities are unknown. The closest doctor is several hours away. There is a public school in the district but facilities in the school are poor and only three grades are offered. A police inspection point is maintained in the district. The Medellin - Turbo road is kept in reasonable condition but there are no roads for cars in the district itself. Transportation must be by mule or foot. Existing Technology The principal crops grown in Tinajitas are coffee, sugar cane, sisal, corn, beans, and onions. Vegetables, cassava, corn, and beans are grown mainly for consumption on the farms in the area. Coffee and fruits are grown for sale in Santa Fe de Antioquia. Coffee accounts for about 201965 pesos. 318 30 per cent of the land area in crops; corn accounts for about 20 per cent. Yields for all crops appear to be below Colombian averages. The primary source of power is the farmer himself. The same hand methods of cultivation have been used in the area for many years. Use of insecticides, fungicides, and chemical fertilizers is virtually unknown. Corn in the area typically requires ten months from planting to maturity. NO earlier maturing varieties are available for these conditions. Almost one- half of the farmers burn off the vegetation before planting. Few efforts are made to control erosion. Livestock is generally of poor quality. Production Of milk and eggs is low because of poorly balanced rations and inferior quality animals. Little is done to keep animals in good health other than feed them pasture and corn. Similar levels of technology exist in the households. Only 32 per cent of the households had latrines at the time the extension service made its study. Kitchens are often no more than a fireplace made by placing several rocks on the floor. Floors are usually of earth; walls are mostly mud and boards; roofs are largely tile made locally from clay. In general the houses are not conducive to disease control. The ultimate product of these poor living conditions and consequent disease problems is a short life expectancy. Only 10 per cent of the population were over 50 years of age at the time of the extension service's survey. 319 Communication Channels Most of the peOple in Tinajitas have some contact with outside areas through their periodic trips to the nearest town for marketing or through relatives living outside of the community. Many of the young peOple leave the district to search for better Opportunities in other areas - mainly Medellin. Data on changes in the population of the district by itself are not available but the municipios of Buritica and Giraldo in which it is located21 experienced average annual pOpulation changes of -1.2 per cent and +0.86 per cent during the period 1951 to 1964.22 Thus, it would appear that pOpulation is perhaps steady or declining in the district due to the outflow of persons noted by the extension service. At the time the socio- economic study was made, 50 per cent of the families read a weekly paper (El Campesino) and 44 per cent listened to the radio regularly. Radio Sutatenza and the Voice of .America were the most pOpular stations. Several other government agencies have carried out *various projects in the area. These include the Coffee Federation, the Caja Agraria, the Ministry of Education, the Ministry of Health, and others. However, the contact Of 21The district lies partly in one municipality and partly in the other. 22Departamento Administrativo de Planeacién, Gobierno de Antioquia, Anuario Estadistica de Antioquia, 1966, (Medellin, Colombia: An Annuai report of the province of Antioquia, 1967.) 320 local peOple with these agencies is relatively small as the projects were not extensive. The Department of Agriculture of Antioquia has tended to follow a policy of not working in areas where services similar to their own are available from other agencies. One of the reasons that Tinajitas and other districts were chosen by the Department was that the Coffee Federation did not regard them as strong coffee-growing areas and hence was not actively servicing them. The peOple of the area live in close contact with one another - pOpulation density is 232 persons per square kilometer - but social activities appear to be few. Family size is large (6.4 persons average) and sometimes one finds several families living in one house. However, peOple have a reputation for minding their own affairs and communication channels within the district seem to be rather slow as a result. Agri cultural Extension Ac tivi ties The extension service of the Department of Agric- ulinxre has been working in Tinajitas since 1965. The socio- economic study of the area was begun in that year and, while the study was being completed, a number of projects were started on various farms in the area. On the basis of the socio-economic study extension priorities for the area were determined using a point (or index) system (see Table 41). Possible programs were assigned a score of zero to five 321 HH N o o o O o e ocoo somsm .NHO OH O O O O O O O mmmmoo HHHV OH O O O O O O N musummm HOHV OH N O O O O O N OHOU HOV OH N H N O O O H mOmmm HOV ON H N O O O O O muHsum HOV ON O O O O O O m mmHnmumom> HOV ON O O O O O O O HmmHm HOV ON m O O O n O O OOHuHquz Ave ON N O O O a e O OansoO HOV ON N O O O O O s OcOHONO a sOHsas ONO on O N O O O O O NuansocH momuuou HHV Hmuos NuH mo mm: Nhqusoo momma uOmEHHmst mmmoosm NuOmuOOO NOUHuOmmE mxms ou me» no me» mumHosoo musmcm Ou mumnuo me» no OoHuHmom uOmEoo OOHHHOO musmOm mumem ommx Ou pmmO m Emuooum mHnmuo>mw IHm>mp mo usosuHs Ou umem mopmHSOOx umszm OH OH OOHmOmuxm O OH OmHo mcu MOOMH mmHuHHHomm HmoHOOoma uOmouO mumsumm OHOuHa OH O» uOmH momuoum Om pH pH mru mud mo Omu pOm uuoomcmuu .uHomuu .mmuHOMOHB mo HOHuume mnu OH mmHuHuoHuo Emumoum OOHmOmuxm OsHsHsnosmo now owns oososzI.He mHsoe 322 depending on the extent to which they met various criteria. The points assigned to each program for the various criteria were summed to arrive at a total score for each program. These total scores were then used to rank the possible programs. Programs 1 to 6 were selected as being most apprOpriate for the area. Horticulture was incorpor- ated in the nutrition programs because of the complementary relationship between fresh vegetables and better nutrition. These programs were begun early in 1966. To estimate the benefits attributable to the 1967 extension programs in Tinajitas the author selected a sample of 20 participating families. One of these families could not be contacted in three attempts. One other did not participate in 1967 although they had participated previously. Thus, eighteen families were interviewed regard- ing the results of their participation in 1967. The extension programs appear to have been modified somewhat since the socio-economic study was made and extension priorities were established. In 1967 a number of Iextension activities not included in the six high priority programs were carried out. Furthermore, several of these siac programs seemed to be receiving little or no attention. The results of the 1967 extension programs directed zit production activities for the sample of participating famnilies are summarized in Table 42. It can be noted that there are significant differences between the numbers Of families given assistance and the numbers actually benefiting 323 .HmOOOOHmm OonOmuxm On mOOmumHmmm mo mmowu mumummmm mm omouooms .Hm>m30O .muma Omes .mmmum umnuo OH pmuomHmmu mum mmmOu OuHa pmumHOOmmm mmmmmHOOH mEOOOH OOO .mOmpumo mHnmumom> OH xuoa Oqu pmumHOOOOO OHHmOmO mums mmHuH>Huom OoHuospoud snow ummH mean .mpme mOm>HOm meHu ms» um mHnm>Hmmno OHuomHHp uOO mmmmmHOOH mEOOOHn .HHO um HH .pm>Hmomu mp3 OH OmOS mOHEumumo ou mHOHmmoo OOO was OH pmHHmomu uOO was OOOmumHmmm mOu Oms3 .OOOH Omen umeuo mummO OH mOOmumHmmm meOOHpOH may pm>Hmomu m>mO One mOEOHOo oau mmmOu OH omooHOOH mmHHHEmmm H m mHHm A umooEOo H O Hoquoo o.>Huomqu mmOOmOo mEOOOH H mmmmmHo pOm .OOHOOOpm .OOHuomHHoomH OO pmuomHHoo uOO mumo H O OOHuOmHo A ImOmHB H O OOHOOO qumOOo pmn pmmm ON O OH H HH osHsuoHo ON e O N HH mummso IHOOmm «\z mOoz O O m mHuumo NO H m H O mmmm «\z msoz H H N NsoHsos «\z mOoz H H N mOH3m OOH O O N O OOHnnmm sinusoHs «OOHON O O O O HOOHO OON m HH H NH mmHnmumom> «\z mOoz H H N muHsum Hmommov pmHOOOO 0:3 mmOHuomHm mOOmumHmmO mOOmumHmmO mmOOmumHmmO OOH>HOOO mmHHHEmm How OOHOOOOO OOHHHmomm pom mOHHHmomm Om>HU OOHHOOOOHO mOOMOU mEOOOH mmHHHEmm mmHHHEmm mmHHHEmh mmHHHEmm HmOOOO momHm>O Ho Hmnssz HO Hmnssz mo umnssz HO umnssz .mmHHHsmO mmuHHmOHe OmmuOOHm mo mmHuH>Huom OOHOOOOOHO mru um pmuomHHp mamuoouo OonOmuxm OOOH no mpHOmmu mo OHOEEOOII.NO mHnt 324 through adOptions. In most cases income estimates could be regarded as reasonable projections for future years as well as for 1967 or 1968, the years on which the observations were based. Vegetable gardens were undoubtedly the most successful program. Rabbits, bees, clothing, and handi- crafts were the only others with any significant degree of success. In those cases where instruction was recalled but no changes or adoptions resulted, families Often commented that they lacked time to apply the new techniques or they couldn't afford the expense involved. This latter comment was not surprising in view of the fact that 44 per cent of the families interviewed had incomes less than 400 pesos per month, 44 per cent between 400 and 700 pesos and only 12 per cent earned more than 700 pesos monthly. The former comment was understandable as most people spent long hours tilling their own soil by hand or working for others in order to earn their living. (Average size of farm for families interviewed was 4.3 hectares, but most farms were less than 4 hectares in size.) Among those families adOpting new technology there were a variety of favorable comments as to how the extension service had improved production in the area. No income changes could be Observed for sisal which was just beginning to produce at the time the surveys were made. On the five farms where sisal had been planted, an 325 average of 1994 plants per farm were in production.23 Several farmers interviewed pointed out that sisal was not entirely new to the area but that the new variety and new methods of handling it were much better than the old ones. Of the eighteen families interviewed, six reported using credit from a bank or government agency. However, only one of these used credit in order to adOpt new technology. Usually credit was used to carry families from one harvest to the next. The results of 1967 extension programs directed at consumption activities are summarized in Table 43 for the eighteen families interviewed. Measures of the utility of this assistance could not be made and it was only possible to ask the families whether or not they benefited from the assistance. With few exceptions, the answer to this question was "yes". However, as can be seen in Table 43, almost all who answered "yes" received some form of material aid as part of the assistance. Thus, no conclusion can be drawn with regard to whether or not useful knowledge was gained. In view of the fact that, in most cases, the assistance took the form of material aid 9211, one is tempted to conclude that increased knowledge was not a significant result. In addition to the consumption activities listed in Table 43, 23This was an excellent example of the gifts being distributed by the extension service. All of the seedlings had been given to the participating families and a machine to assist in extracting the final product was also being supplied. 326 HH .omHHmomu uOO mm3 mOOmumemm may Omsk .HHm um HH .pm>HmomH mas uH Omna HHmu ou mHnHmnom OOO mm3 .OOOH Omen Hmeuo mummm OH mOOmumHmmm pmHOOHOOH may pm>Hmomu m>mr has OOEOHOO 03» mmmnu OH pmpOHOOH mmHHHEmmm H H H N OOHummuomm O O H O mumamm N N O O meme Hmumz H O H O N OH mOHH Hmumz m m O O HOOHpmz O O N O OOHOOOO H m H m H O mmOHHumH Hsmmmo mm Hammmo mm pHO OHO mmOOmumHmmO mmOOmumHmmO Emuooum OOOmumHmmO mOOmumHOOO HwHumumz HOHumumz oOHHHmomm uoz Om>Ho OOHmOmuxm OOHpHmomm OOHpumomm OOH>Hmomm OOH>Hmomm mmHHHEmm mmHHHEmm mo mews uoz #02 mo HmeEOz mo HmeEOz mOOmumHmmOImOHHHmomm mmHHHEmm i fl“ .mmHHHEmm mHmEmm OmmuOon MOO OOOH OH mmuHOpOHB mo HOHHume may OH mmHuH>Huom OoHuoEOmOoo um pmuomHHp msmnoouo OOHmOmuxm Ho muHsmmOII.mv mHnms 327 the extension service indicated it was Offering assistance in the areas of nutrition, first aid, improving literacy, and clean-up programs. However, so few families received assistance in these areas that none appeared in the sample.24 In the areas of latrines, housing, water lines, water tanks, and sewers, the assistance provided by the extension service usually involved providing an obrero to construct the facilities. Where projects were public in the sense that they served more than one family the extension service usually purchased some of the materials as well. Of course, the participating families usually contributed some labor. Families were questioned as to when they would have learned the new methods if the extension service had not been working in the area. Some did not understand these questions but most who did answered "never". Two suggested they might have learned a little by means of radio. Of the eighteen families interviewed, seven reported teaching some of their newly-acquired knowledge to neighbours who usually had already adopted some techniques or were likely to adOpt some in the coming year. Only one of the persons who extended knowledge to neighbours reported teaching more than one other person. In addition to technical changes in the areas of production and consumption, some other extension outputs were also observed in Tinajitas. Two 4-H clubs were Operated - 24Some nutritional advice was given as part of the vegetable garden program. 328 one for males and one for females. The boys' club involved 12 teenage youths while five housewives and 17 girls attended the other club. Many of the changes listed in Tables 42 and 43 resulted from club work, as, not infrequently, it was a 4-H club member that was the extension participant in a family. However, other benefits may have resulted from club work as well. Some managerial changes undoubtedly occurred as the young people worked together at club activities. The club members probably also picked up some knowledge about other regions from extension workers. This could be useful if they decided to leave the community. The most significant contribution of the extension service to changes in community activities came through the Accién Comunal programs. Fourteen of the eighteen families interviewed reported being involved in these programs which included improvements to the school, road repair and improvement, sewer construction, and water line construction. (Some of these items were already listed in Table 43.) All families involved said they benefited from these programs. The extension service and the Coffee Federation assisted in getting things organized but, what appeared to be more important to the local families, both these institutions provided Obreros to work on the jobs and money to buy the necessary materials. Because of the nature of these programs it was not possible to measure the extension service's 329 contribution or evaluate the probable benefits. However, local families were quite enthused about the extension service's activities in these areas. Other Districts Studied The districts of Buenos Aires, Pinguro, and Urquita were in many respects similar to Tinajitas although signif- icant differences in soils, climate, and people were noted among them. Detailed descriptions of these three districts can be found in their respective socio-economic studies prepared by the Department of Agriculture. Only the observed results of 1967 extension programs are summarized here. Buenos Aires The district of Buenos Aires is located about two and euH3 pmuMHoommm mums mmHuH>Huom OOHuospouo HOOH mmmea mHeHmmOd uOO was uH mOOmumHmmm pmuMOHOOH .pmHHmomu uOO mm3 OOOmumHmmm meu Ome3 meu om>Hmomu m>me has mOEOHOO ozu mmmeu OH pmosHOOH mmHHHEmmm .mmHuommumo .muHsum pOm e .HHm um HH .pm>Hmomu mm3 uH Ome3 mOHEnmump ou .OOOH Omeu umeuo mumm» OH Hm>HuHmoo OHemEOmmHo use OzoOx uOO mEOOOH OO uommmmcm mHuumo OOHumOHoom> HN Hosusoo H mmmmmHo e.>HuomHHp mmOOmeo mEOOOHHO OOHuOdemOmua OOm .OoHuooom .OoHuomHHOomu Oo omuomHHOo uOO mumoHO OOHuOOuumOOO A wmevmmm AN mHHm umOdEou HON H N O O OsHsooHo OH H H o H mummHOHUOmm «\z «\z o H H ostm «\z o H O a OOuHsom OHN O O o O muHeemm mom O O H m mmHemummm> HON H N N N muHDHh cOOH N m O m OOMMOU HOOOmoc mmmmmuOOH mEOOOH mOOmumHmmO mmOOmumHmmO mOOmumHmmO mmmmHOOH OOHOOmHumoxm OOHHHmomO OOHHHmomm uoz m Om>Ho mmuc mEOOOH mmHHHEmm mmHHHEmm mmHHHEmm mmHHHEmm OOHuOsooum HmsOOO mo umeEOz mo umeEOz Ho umeEOz HO umeasz mommm>< .mmuHO mOOmOm OH cm3mH>umuOH mmHHHEmH OH meu HO mmHuH>Huom OOHuospouo meu um pmuomuHo mamsooum OOHmOmuxm OOOH mo muHOmmOII.vO mHems 333 these in the future the family was questioned about the amount of time involved in the activity. This often was significant but no families reported leaving paying jobs or regular work on their own farm to perform the new activities.29 They were essentially employing what had previously been unproductive time in order to start a new enterprise or expand or improve an old one. Coffee and vegetable gardens were the most successful programs with some successes also being noted with rabbits. Most of the programs associated with production activities appeared to be directed at making farm families more self-sufficient whereas the results indicated that, when a farmer chose to adOpt technology in a given area of production, he was usually interested in making a commercial enterprise out of it. It was also interesting to note that the extension service's socio-economic study suggested fruit crops as being products with which some useful extension work could be done whereas, when the author visited the area, some fruits, particularly oranges, were in such abundant supply as to be worthless in local markets. Of course, good markets did exist for oranges at other times and, perhaps even in other areas of the department at the same time. Thus, unless extension workers could have 29One example of this was the family who took an active interest in vegetable gardening. Their net return per annum was 350 pesos but the farmer's wife had to devote 30 days of work to the garden. Thus, her daily return was about 12 pesos - a fairly competetive wage. 334 somehow found new marketing methods or techniques for changing the harvest period, their work would have been in vain. The usual production-oriented advice was simply irrelevant for fruit crOps on most farms. Income levels, farm size, and use of credit were similar to those noted amongst the sample of families interviewed in Tinajitas. Sixty-four per cent of the families interviewed had estimated monthly incomes of less than 400 pesos. The average farm size for those families interviewed was 6.5 hectares but only two families had more land than average. Five families had used credit from banks or government agencies. Of course, most families used local credit sources to keep them going between harvests. The results of 1967 extension work directed at the consumption activities of Buenos Aires sample families are shown in Table 45. As in the case of Tinajitas, in a size- able number of instances (55 per cent) assistance mentioned by extension personnel could not be recalled by the farm family. However, most (86 per cent) who recalled the assistance regarded it as useful. With the exceptions of the areas of nutrition, recreation, first aid, and medical assistance, all who regarded the assistance as useful had received material aid, usually for construction purposes. Again, this appeared to be a situation in which the families were so poor as to be unable to apply recommended practices in most areas of assistance. The extension service realized 335 mHeHmmOd uOO mp3 uH .omHHmomu uOO mm3 mOOmumHmmm meu Ome3 .HHO um HH .om>Hmomu mm3 uH Ome3 HHmu ou .OOOH Omeu umeuo memo» OH mOOmumHmmm pmuOOHpOH meu cm>Hmomu m>me One OOEOHOO osu mmmeu OH pmpsHOOH mmHHHEmmm 42 \N Z NHOOOOOO <2 \H z NI‘V‘Q‘V‘V‘MH d \N Z NHVQ‘OOMH st: \0 z NI‘OOV‘Q‘OO mfl'OOMNOOI-im FWMHHQ msIOmmHu momumuHH pH< umuHm OOHummuomm mema umumz OHQQOO Hmumz HMOHpmz OOHuHHqu OOHmOom mmOHuumH Hammmo mm mOOmumHmmO OOHpumomO uoz Hammmo mm mOOmumHmm< OOHpHmmmm OHO HOHumumz OOH>Hmomm uoz pHO HmHumumz OOH>Hmomm OOOmumHmmO OOHHHmomO mmHHHEmm mOOmumHmmO OOHHHmomm uoz mmHHHEmm mo HmeEOz M mOOmumHmmO Om>Ho OOHHHEOO mo umeEOz mmuO OOHumEOmOOU .mmuHO mOOmOm OH om3mH>umuOH mmHHHEmm mHoEmm Ommuusou meu mo mmHuH>Huom OOHuossmOoo um pmuomqu mfimumonm OOHmOmuxm HO muHsmmmII.Ov mHemB 336 this and concluded that the only way they could get peOple to adOpt was to bear a large share of the costs. It is hypothesized that many of those families who said they did not recall assistance may have received instruction either individually or at 4-H Club meetings but, because they had not been able to afford to adopt the new practices, they chose to “forget" about it when interviewed. Those Buenos Aires families who adopted new practices did not hesitate to inform their neighbours of their activities. Eight of the thirteen families interviewed reported passing on knowledge to their neighbours and a total of ten neighbours received some such assistance. Most participants interviewed had no idea when they might have learned about the new ideas if the extension service had not been working in their area. As many pointed out, there was just no other opportunity for them to learn anything new except, perhaps, the radio. As was the case in Tinajitas, several extension outputs were noted which did not involve technical changes in the production or consumption activities of individual family farm units. A girls' 4-H club with 17 members and a boys' 4-H club with 12 members were active throughout the year. Expanded knowledge of nutrition, health,hygiene, clothing, etc. was the main short-run result of this activity. Future technical changes in production and consumption activities might well be inspired or speeded up by means of this knowledge. As in Tinajitas, Accién Comunal was active 337 lovith ten of the fourteen families interviewed indicating 1:hey benefited from road, school, or water supply improve- rnents made through this organization. One of the extension sservice's principal functions here was to assist the local Ioeople in petitioning the Coffee Federation for grants to Inake these improvements. Of course, the extension service Vvould supply some labor and materials as well. In a 1:ypical project costs would be shared as follows: Coffee IPederation, 50 per cent; Department of Agriculture, 25 per (zent; local residents, 25 per cent. IPinguro The district of Pinguro is located in the municipal- isty of Giraldo close to the border of Buritica and only about twenty kilometers from Tinajitas. The extension Service Of the Department of Agriculture had been working 111a the area since 1963 at which time they started several Etrojects, including a school. They conducted their socio- economic study of the district in that year and, with this as a basis, began a group of more intensive and organized Programs in 1964. Because of this long record Of service iilid because the practico and mejoradora sometimes live in ‘tiie district, the extension service is quite well-known in time area. The socio-economic study of the district carried Cunt by the extension service identified a few problems Peculiar to the area but, for the most part, the situation 338 .i.s similar to that in other districts. Incomes are low, taxousing is poor, life expectancy is short, some children ¢::annot attend school because of lack of facilities, farms are small (2 to 3 hectares), soil is steep and eroded, and sax: on. The extension service made the following recommend- ations for work in the district: 30 1. Family (a) (b) (C) (d) (e) try to improve social structure by aiding in construction of houses so that young families would not be dominated by their parents with whom they often must live at present. attempt to modify existing dwellings so as to prevent the promiscuity observed in the district. promote animal production (e.g., bees, poultry, swine, cattle) to offset the protein deficit which most of the inhab- itants of the area currently experience. develop campaigns to improve the unsanitary conditions which prevail in most houses. attempt to develop small industry or handi- crafts in order to increase family incomes. 2. Agriculture (a) (b) (0) try to transform old, traditional coffee plots into higher-yielding ones along the lines recommended by the Coffee Federation (e.g., new varieties, fertilizer, disease control). develop conservation-oriented programs to encourage farmers to control erosion in the area. concentrate on improvement of vegetable gardens especially as these can be an import- ant source of better nutrition for most families. 30These comments are a precis of those contained in the Department of Agriculture's socio-economic study. 339 (d) by means of a community action program, attempt to control the arriera ant which currently prevents some crops from being grown in the area. (e) although livestock cannot be a basic industry, it is important to try to get each family to own their own cow to provide milk for them. This would require pasture improvement in the district. (f) poultry and honey enterprises could be developed to supplement incomes as a local market exists in Santa Fe de Antioquia. Both of these could be developed with technical assistance. 3. Economic (a) as credit will be required for some programs to succeed, attempts should be made to utilize credit supplied by the Caja Agraria or the rotating fund of the extension service itself. (b) by means of Accién Comunal, the roads in the district can be improved to facilitate marketing locally-produced products. (c) as poultry producers develop they should become affiliated with the Co-op in Santa Fe de Antioquia to facilitate marketing their products. 4. Community (a) to maintain the social welfare of the community a sports program should be developed. (b) by means of Accion Comunal, school facilities should be expanded. As in other districts, extension programs were Clarried out by means of visits to individual families, 4-H <21ubs, and Accién Comunal meetings. These were supplemented 1Djy’muonetary assistance for community projects and material iissistance to individual families. An extreme example of 340 'the type of assistance rendered to individual families was ‘the case in which an obrero employed by the extension :service worked for one family for 82 days. During this time lne constructed a latrine, remodelled the house, built a water tank on tOp of the house, and made a pig pen. The :Eamily supplied materials using money borrowed from the eextension service. The farmer estimated the value of the sservice rendered by the Department of Agriculture at 25 zaesos per day or 2,050 pesos in total. The entire family vvas quite pleased with the assistance they had been given. The results of 1967 extension programs directed at tflhe production activities of the thirteen families inter- ‘riewed in Pinguro are shown in Table 46. Vegetable gardens and clothing appeared to be the most successful programs Vlith some successes being noted in handicrafts, coffee, and bees. As in other districts, the income increases shown represent the results of technical changes only and do not Ikeflect the results of managerial changes in these areas. I”icnwever, given the likely prospects for progress in the area 311: is doubtful if managerial changes directly related to agricultural production will have much opportunity to pay Off in the future. The single poultry producer in the sample who is 1¢isted as possibly experiencing an income increase due to tleachnical assistance is an example of the uncertainty associated with adopting new technology and the difficulty ,1n estimating related income changes in advance. This 341 .mmHuomwumo moon» cH owuomHmmu mum mmmnu Eoum mCHuHsmmH mommmuocw mEoosH mad .muwsum new .mmmmoo .mcmoumo mHnmummm> nqu UmDMHOOmmm mumz mmstcnomu cowuosooum mmmaan .HHm um MH .om>Hoomu was pH cmz3 mcHEumumo cu mHnHmmom no: mos pH .omHHmowu uoc mus mocmumwmmm ooDMOHocH on» coca .hmmH can» umsuo munch cH mocmumHmmm powwowosH may oo>Hmomu o>mn was mGEsHoo 039 women cw omosHocH mmeHEmmm Am>HuHmom mHndESmmum van csocx uoc mEOUCH so uommmmvv mHuumo mcHumcHoom> Am mcHDGMHchmua n.mHuomHHU momcmno mEoocH ocmfim mHHm umomeoo .coHumoom .coHuomHHoomu so omuomHHoo uoc muwoxm Houucou mmmmeo. an cowuosuumcou pom boom «3 e s. v 932308 on v v v mummnowocmm oov H H H N moon a «H H N m >338 oom H N o m ocH3m «\z o H N m 338mm omm s OH H HH mmHnmummm> «\z o H m m 335 mNH m m H v common AmOmwmv mommmuocH mEoocH mocmumHmmd mmocmumHmm4 moocmumwmma mmmmuocH mcHocmHHmmxm mcHHHwomm mcHHHmomm uoz cm>Hu mend mEoocH mmHHHEmm mmHHHEmm mmHHHEmm mmHHHEmm cOHuosooum Hmsccd mo nonfisz mo umnfidz mo umnasz mo HmnEsz mmmuw>¢ .om3mH>umucH mmHHHEmw ousmcwm cmmuanu can no mmHuH>Huom coHuosooum on» an omuomuHo mfimumoum cOHmcmuxm hmmH mo muHSmmMIu.mv mHnme 342 family had borrowed 4,500 pesos from the extension service's rotating fund in order to get the poultry enterprise going. However, their hens contacted disease, production drOpped, and a number of birds died. At the time of the survey the family had not obtained sufficient revenue to repay the loan. It appeared as if they might break even and, if they did, they might try again next year. They might well be dis- inclined to try again on borrowed money even if they did manage to break even in 1967-1968 but their decision had not yet been made when they were interviewed. The results of 1967 extension programs directed at the consumption activities of the thirteen Pinguro families interviewed are shown in Table 47. Again, some families did not recall receiving the assistance extension personnel reported giving them. A close association between the use- fulness of the assistance and whether or not it was accompanied by material aid was also noted. In general, programs directed at consumption activities in Pinguro exhibited similar results to those in Tinajitas and Buenos Aires although the proportion of cases where families recalled assistance was higher (close to 70 per cent). Income levels, use of credit, and general conditions in Pinguro corresponded very closely to those noted by the extension service's socio-economic study several years earlier. Three families reported using credit frOm the Caja Agraria; three others reported using the Department's rotating credit fund. Most families interviewed had monthly 343 o H o H N maucmmHo «\z «\z «\z .«\2 H H 34 para N o N o o N coflmwuomm o N o N H m mxcma Hmum3 o H o H N m 335m 333 H e H v H m HMOHcmz H m m H o v coHuHHusz m N v H N h mcwmsom o N o N N v wmcHuumq HsmmmD mm Hsmmms mm UH« oHd mocmumHmm< mocmumHmm¢ mocmumwmmd mocmumHmm< HMHumumz HmHkumz msHHHmomm uoz mcH>Hmomm moms mCHoummmm @cHoummom mcH>Hmoom ocH>Hmomm mmHHHEwm mmHHHEmm coHumEsmcoo uoz uoz mo nonadz mo Hmnfisz mocmumHmm< mcHHHmomm mmHHHEmm J .omBmH>umucH mmHHHEmw ousmch cmmuuwnu may no mmHuH>Huom coHumEdmcoo man no omuooqu memumoum consmuxm homH mo muHsmmmss.wv mHnt 344 incomes of less than 700 pesos. As in other areas some evidence of diffusion was noted. Three families reported teaching neighbours some of the new technology they themselves had learned from the extension service. They all indicated that their neighbours had been quite interested and would likely adopt some, if not all, of the new methods. The families interviewed in Pinguro were in general unable to say when they might have learned about the new methods if the extension service had not been operating there. One suggested five years, another two years, but most said "never" or declined to answer. In contrast to Buenos Aires where some assistance might have been forthcoming from the Coffee Federation, there was little likelihood of Pinguro residents learning about new technology from any other agency. Extension outputs other than technical changes in production and consumption units were noted in Pinguro. Two 4-H clubs with 18 boys and 40 girls as members reSpectively were Operated. Community projects were carried forward by means of Accién Comunal. These involved road and school improvements as well as water line construction. As in other areas the Coffee Federation, the families of the community, and the Department of Agriculture shared the costs. However, most of these projects were completed prior to 1967 and consequently their costs were also incurred in prior years. 345 Urquité The district of Urquité is located on the main road from Medellin to the port of Turbo. It is only about a one- half hour's drive from the local extension office in San Jeronimo. However, the practico who works in the diStrict chooses to live there rather than commute. Extension work in Urquité began in 1965 when the extension service's socio- economic study was being completed. However, field advisory work was carried out mainly in 1966 and 1967. Like Buenos Aires, Urquita has somewhat more favorable conditions for coffee production than Tinajitas or Pingura. Twenty-nine per cent of the land in Urquité is devoted to coffee production with plantain and pasture accounting for another 40 per cent. Sugar cane, beans, vegetables, and waste land account for most of the remainder. In most respects the district is similar to the other three already discussed. However, as it is not so remotely located from Medellin, the inhabitants have more opportunities for off- farm employment. The results of 1967 extension programs directed at the production activities of the nine Urquita families inter- viewed are shown in Table 48. This data would be similar to those collected for other districts if it had not been for one sample family where results were somewhat spectacular. In all districts studied it was obvious that some families were able to make a good deal more use of the 346 .mmHHomouno wwonu cH omuomHmmu mum mums» Eoum mcHuHamou momwouosH oEoosH mc< can .mommoo .mcmcumm oHnmummm> nuHs cmHMHoommm mums mwsvHssomu coHuosvoum moose #0: «M3 uH .omHHmoou uoc mm3 oucmumHmmm omHMOchH on» con3 .muHaum n .HHu um MH .om>Hmomu mm3 pH cos3 ocwsuoumo on anHmmom .hmmH coca uwnuo ammo» cH mocmumHmmo UmumoHUcH any om>Hmomu m>mn has msEsHoo 03» mums» CH voosHocH mmHHHemmn Am>HuHmom mHnuESmmum use csocx uoc mEoocH co uomwmva mHunuo msHumcHoou> Hm Houucou ommmeo n.mHuuouHo woodman msoocH ocmMm soHuosmwmcMWm .coHumoom .coHuooHHoomu co nouowHHoo uoc mumoav msHucuHmmwmua Av mHHm umodsoo OHN H N o N mcHsuoHo hHm H H o H mumeOHocmm mNmm H v H m muuHsom ¢\z o H o H manan one H N N e mcHsm mNm H H o H mdouo “mayo omNm H H o H mmmuoo mmm m m o m moHnmummm> ovw H m o m muHsum Hmomomv mommouocH «BOUGH mocwumHmmd mocmumHmm¢ mosmumHmmd mmmouocH msHosmHummxm mcHHHmomm mGMHHmomm #02 m co>Hu mmud mEoocH moHHHEmm mmHHHEmm mmHHHEmm meHHEmm coHuosooum Hmsccd mo HmnEsz mo umnEsz mo uwnasz mo nonasz momuo>¢ .omsmH>uwucH mmHHHEMM wuHsvHD mch may no meuH>Huom cOHuosooum may no owuomqu mfimnmoum GOHmcouxm bomH mo muHsmmmll.me mHnma 347 extension service's assistance than others. While the families which benefited more were usually those with larger farms - but not ranches as these were not serviced - and higher incomes, this was not a hard and fast rule. Variations in benefits among farms, even those of about the same size, were usually large but not spectacular, with one exception. One family received assistance in many areas and were able to increase their annual income by almost 20,000 pesos as a result. In view of the fact that only a few other families experienced income increases of 1,000 pesos or more, this case deserves explanation. Firstly, this family was already one of the more well-to-do in the community and owned 39 hectares of land. Secondly, the family was quite large (17 members). Thirdly, the practico who worked in the area resided with the family. This combination of factors created a situation where the family was eager and able to change their operation so as to fully employ most family members. As they were relat-- ively well-off, they could afford to make the expenditures recommended by the practico. The estimates of net income increases for this family are probably too high as no opportunity cost for added labour was charged. However, no such charge was made as it was difficult to tell what family members would have been doing if they had not become more fully-employed on the farm. It should be pointed out that, in all districts where a practico resided, the family with whom he lived usually benefited to a much greater degree 348 than other families in the district. Families who "boarded" the practico (or mejoradora) usually received at least a prOportionate share of the material assistance provided by the extension service. It is also significant to note that, in Urquité, only the one outstanding family experienced income increases from programs other than vegetable gardens. While the sample is a relatively small one and, thus, this situation could have arisen by chance alone, there is the suggestion that perhaps assistance, as well as benefits, was more concentrated in this district than in others. In each of the other three districts more than 50 per cent of the sample families experiencing income increases experienced these in more than one area of production. The results of 1967 extension activities directed at the consumption activities of sample families in Urquité are shown in Table 49. These results are similar to those noted in other districts. Income levels, credit use, and general conditions in the district were similar to those described as the socio-economic study of Urquité and already noted in other districts. Two of the sample families reported using credit from government agencies. One of these was using it to try out new production methods in 1968 but, in general, credit use was restricted to enabling families to service from harvest to harvest. Average farm size for eight farms surveyed was 1.5 hectares - the other sample farm encompassed 349 «\z «\z «\z ¢\z H H dancumHo N N m H N m 6H4 umHHm o H o H o H mxsma nouns o N o N H m mmsHH nouns H N m o N m HMOHomz o e v o H m coHuHuusz H m N N H m mchsom N m v m b h mmcHuumH Hammmo mm Hummus ma oH< UHd mocmumHmmé mocmumHmm< moms mocmumHmw< mocmumHmmd HMHHmumz HMHHmumz mcHHHmuom uoz sm>Ho GOHHQESmcou msHonmmmm mcHoummmm mcH>Hmomm mcH>Hmomm mmHHHEmm mmHHHEmm uoz uoz m0 .3852 no H0852 mocmumHmmd mcHHHmomm mmHHHEmm .wuHsqu CH mmHHHEmu OHQEMm mch on» no mmeH>Huow coHumfismcoo on» an omuomuHo mmHuH>Huom GOHmcmuxm hme mo muHSmmmul.mv mHnma 350 39 hectares. Four of the nine sample farms reported teaching some of the new methods they had learned to their neigh- bours. Six neighbours received instruction in this way and several adopted some new methods. Most of the families surveyed had no idea when they might have learned about the new technology if the Department of Agriculture had not been working in the district. As in other districts, extension outputs other than technical changes in production and consumption units were noted. 4-H club membership totalled 13 boys and 25 girls in 1967. As in other districts a number of the results in production and consumption activities were attributable to 4-H activity but other benefits would likely be forthcoming from these programs as well. The Accién Comunal program was active in the district in 1967 and eight of the nine families interviewed reported participating and benefiting from this. Road improvements, water lines, and school expansion were the main projects. The extension service contributed some funds, organizational assistance, and labour, to these projects. CHAPTER IX ANALYSIS OF EXTENSION ACTIVITIES IN ANTIOQUIA Because cost and benefit data were collected by district rather than by program, the nature of the analysis of extension activities in Antioquia differs somewhat from that conducted for the CVC. No comparison of programs in different districts by benefit categories has been made. The first part of the chapter is devoted to the calculation of PVB/PVC ratios although it turns out that these do not reveal significant differences in extension effectiveness among districts. The remainder of the chapter is devoted to the discussion of possible factors affecting effectiveness among and within programs and districts, and possibilities for cost reduction. Calculation of PVB/PVC Ratios Serious problems were encountered in measuring benefit flows attributable to technical changes in con- sumption activities and community organizations. This has a particularly limiting effect on the analysis because of the heavy involvement of the Department's extension service in community activities. Problems were also found in measuring benefit flows for managerial changes and secondary 351 352 extension outputs, but, as in the case of the CVC, benefits in these categories appeared to be small or non-existent and thus, their obscurity does not present a serious drawback to the analysis. More difficulty was encountered in estimating "shifts" in adoption distributions attributable to extension activities than was found in Valle and some rather strong assumptions were necessary in order to permit calculations to be made. The benefit flows attributable to technical changes in agricultural production units are based on a twenty year uniform adoption distribution.1 It is assumed that the extension activities of the Department of Agriculture had a contracting effect on the earlier part of the distribution. However, even though this study originally attempted to isolate changes which occurred as a result of 1967 extension efforts, the changes observed appeared to have occurred, to a large extent, over a two to four year period depending on the district. This period is assumed to be a contraction of the first half of the theoretical adoption distribution and leads to the use of the total number of families particip- ating in 1967 as being "adopters" by 1968. This same assumption is made for all types of technology adopted as 1This is an intuitive estimate based on the facts that the districts are relatively small and that the adopters are probably relatively slow to change their ways because of the high risks and their limited general knowledge base. The uniform distribution was used as an approximation to the normal to facilitate calculations. It was considered that other assumptions outweighed this substantially in their effects on the final results. 353 there was little empirical basis for distinguishing among them, even though, in-reality, the adoption rates for different types of technology might be quite different. However, the method of calculation does not assume that all participating families adopted all types of technology or, in fact, that all families adopted some technology. Where it appeared that the results measured in the surveys were attributable to several years' extension efforts, it was necessary to estimate extension costs for years prior to 1967. These estimates were based on average extension costs for resident families serviced by the offices of San Jeronimo and Santa Fe de Antioquia in 1967. Only extension costs for years in which extension field work appeared to have been done were included. One may argue that the costs of the socio-economic studies of the districts should have also been included. However, they have been omitted for two reasons. Firstly, part of the total cost of operating the extension service in 1967, from which the costs for the districts were derived, represented tabulation of data for socio-economic study and therefore, some allowance has already been made for this. Secondly, the total useful- ness of the socio-economic study could not be considered exhausted in 1966 and 1967 extension programs. It would surely be of some utility in planning future programs in the districts studied or similar parts of the region. It would therefore, be inappropriate to charge the total cost of such a 354 study against the results observed in the survey. There was no good way of arriving at an estimate of the families in a given district which would eventually adopt technology of a given type. Presumably this is related to profitability, some characteristics of the new technology, and some characteristics of the potential adopters.2 However, as this relationship is not well-known, it could not be approximated here. Thus, it was decided to make the somewhat arbitrary assumption that benefits at the levels observed in each district would eventually spread to all resident families in that district and, that sufficient diffusion of technology would occur among families to increase this by 50 per cent by 1980. This seemed reasonable to the author but, as has already been noted, it is highly arbitrary. It would require, for example, that, if 20 per cent of the sample families were using a given technology on a given average scale at the time of the interviews, then either 30 per cent of all families would be using it by 1980, or there would be sufficient increase in average scale of use among those adopting to offset any lesser incidence of use. Although there is little basis for this other than intuition and the fact that the number of adopters in any given early part of a normal or uniform distribution bears ‘a constant relation to the total number of adopters, the 2See the discussion of an extension system model in Chapter III . 355 results appear to be consistent with other observations. For example, changes in vegetable gardens, which were widely adopted in all districts, can be maintained by almost any family that has a very small amount of suitable land. The production techniques are not complicated and most products are easily incorporated into existing’diets.3 Furthermore, most people already had some sort of garden even though products were few and yields were low. Bees and rabbits, which were not widely adopted prior to 1968, are not likely to become widespread because they require more careful attention and, once the products are realized, there is a good deal of incentive to sell them for cash to buy other foods (e.g., rice, panela). As local markets are limited for both honey and rabbits, and technical conditions limit the extent of honey production in a given area, the total potential for economic production appears to be less than for vegetables. Furthermore, with both these products there are risks which are not associated with vegetable gardens. For example, Peace Corps volunteers succeeded in establishing widespread rabbit production in some parts of Colombia in recent years only to have the total population of some regions wiped out by disease. Thus, at the present time it appears unrealistic to expect rabbit-growing to be a successful enterprise on more than a few farms in each 3Some exceptions to this were noted in the inter- ‘views. A few families remarked that they didn't like cuarrots and other vegetables which were new to the area. 356 district. As seven families in the samples of the four districts had adopted rabbits for an average annual income increase of 150 pesos per family, the assumption underlying the benefit flow calculations implies that, by 1980, the total benefit flow from rabbits will be 7,650 pesos. This could be obtained by 51 families experiencing the same average income increase as that observed in the sample or by a somewhat lesser number with a higher average increase. This does not seem unreasonable,4 although as will be seen later, the data collected does not permit much in the way of distinguishing significant differences among districts or areas of production. Severe limitations on the usefulness of managerial changes are likely going to be encountered in all districts. These limitations will include limited amounts of resources, low quality of resources, market disadvantages because of transportation costs, and other production disadvantages related to the lack of infrastructure. In the author's opinion, worthwhile managerial changes certainly occurred as the result of extension efforts. Their worth probably depends on much the same factors which determine returns to existing production activities in the various districts and may very well vary directly with such returns. However, because of measurement problems it has been necessary to exclude them from the benefit flows calculated here. . 4As 54 of 209 participating families were inter- Viewed, an unbiased estimate of the number of families already having adopted rabbit-raising by 1968 is 27. 357 Benefit flows attributable to most of the technical changes in consumption units were considered to be close to zero when valued at market prices. This conclusion was reached because virtually none of the techniques extended to the families in the area were directed at getting more out of existing incomes. This does not mean to say that the consumption assistance did not help people in the area. However, its market value was perhaps equal to or less than the cost of providing it. Thus, it was decided to deduct an estimate of the costs incurred in providing such assistance from extension costs for each district. As the obreros were the main cost in furnishing this aid, the costs of maintain- ing them in the various districts was used as an estimate of the cost of this assistance. Benefit flows attributable to technical changes in community organizations were not estimated because no information was available on precisely what the extension service did to further these activities - other than provide labor and materials. Because of the joint involvement of the Coffee Federation, the Department of Agriculture, and members of the community it was virtually impossible to say how much the final product depended on extension's contribution. If it had been possible to calculate PVB/PVC ratios for these activities alone, the results would have been quite interesting as it certainly appeared as though they were quite successful. However, without the 358 contributions of resources from both the Department and the Coffee Federation, few or no projects would have been completed. Benefits from influencing technology developers, information sources, or institutional activies outside of the area were considered to be zero. Extension personnel in Antioquia undoubtedly did have an opportunity to inter- act with personnel of agencies such as the Coffee Federation, INCORA, the Caja Agraria, other divisions of the Department of Agriculture, I.C.A., municipal governments, and a few non-agricultural departments of the government of Antioquia. There were possibilities for interaction among field personnel from the field sectors and personnel of the Communications and Agricultural Assistance sections of the Extension Service. This latter possibility could perhaps have resulted in improvements in the work of the three sections. Inter-action with the Coffee Federation, I.C.A., and other divisions of the Department of Agriculture may have perhaps resulted in some useful changes in the research activities being conducted by these organizations. No organized attempt was made in this study to measure the extent of such interaction or determine the nature of the ensuing changes. Casual observations indicated that such influence may have been slight or negligible. Extension workers had no control over research decisions and no organized procedure existed for recommending useful research 359 projects to research workers (even within the Department of Agriculture). Thus, while communication between entities undoubtedly occurred, it did not appear to be having any significant impact on research decision-making.5 Benefit flows resulting from this type of secondary activity were therefore considered to be zero, although perhaps erroneously.6 In the districts examined there were virtually no extension resources devoted to training young people or new producers in technical or managerial practices already in widespread use in the area. Thus, the benefit flow for this "output" was assumed to be zero in all cases. This lack of activity was probably an appropriate action for the extension service because all districts examined appeared to exhibit high out-migration rates due to relative economic conditions. Benefits from assisting in the direction of people out of agriculture might very well exist. Some of the 4-H club activities in the districts probably did contribute something along this line although there were no specific programs for achieving this purpose. On the contrary, the general policy of the extension agency appeared to be one of keeping peOple in the rural areas in order to avoid increased 5One exception to this was the participation of the Research Division in tabulating data for the Socio-Economic studies of the extension districts. 6It intuitively seems that a negative benefit flow in this area would be unlikely. However, such a possibility cannot be ruled out a priori. 360 pressure on the public facilities in Medellin where people would almost surely go if they left their home districts. Nevertheless, the weekly contacts with extension agents probably produced certain managerial changes in rural young people which may have encouraged them to leave their areas and at the same time may have made them more capable of doing so. For example, casual conversations at 4-H club meetings may have acquainted some young peOple with job possibilities in urban areas and the sorts of lives extension workers and/or their friends in urban areas enjoyed. It would require a much more intensive study than this one to measure any of these effects and, as extension programs were not directed at producing them, they have been considered to be a rather inconsequential part of the benefit flow attributable to the extension activities studied. Again this may be an erroneous conclusion. The possibility of error is illustrated by a family visited during the course of this study. This family was heavily dependent on the weekly contributions of several younger members who worked in Medellin. If the extension service had played some part in aiding or influencing these young peOple, then perhaps both the young peOple in Medellin and their parents in the district were better off as a result. For the above reasons no allowance was made for secondary extension outputs in the benefit flow calculations. It would seem that the calculated flows are likely to be underestimates of total benefit flows as a result. 361 The District of Tinajitas As this study progressed it became increasingly apparent that the results observed in Tinajitas were probably attributable to extension work carried out in both 1966 and 1967. As mentioned in Chapter VIII, some of the extension projects encountered in Tinajitas had begun as early as 1965 and this sort of assistance was still being rendered in 1967 and 1968. Only two of the eighteen families interviewed in Tinajitas were new extension participants in 1967. While it was possible to distinguish some assistance which had been completed in 1965 and/or 1966 and discard it as not being pertinent to the inquiry, it was not possible to do this with most of the assistance observed. As the socio—economic study of Tinajitas took place in 1965 and relatively few projects were begun in that year, the families interviewed who had difficulty recalling exactly when assistance was rendered and/or who had definitely received assistance over a period longer than one year, almost surely did not receive such assistance prior to 1966. However, because of the uncertainty regarding the actual costs of producing the results measured in this study it was decided to use three cost estimates. The "lower limit" was considered to be the cost of 1967 extension work alone, not including the obreros - approximately 111,000 pesos. The "upper limit" was considered to be the estimated costs of 1966 and 1967 programs excluding obreros - 362 approximately 246,000 pesos.7 The simple average of these two figures (178,500 pesos) was regarded as the "expected value". The use of the three figures permits testing the sensitivity of the ratios based on the "expected value". The average annual income increase due to extension activities directed at production activities in Tinajitas was 329 pesos. On the basis of diffusion, which was obviously present to some extent in the district, this was assumed to increase to 500 pesos per family in approximately ten years. This permitted construction of a benefit stream for changes occurring as a result of extension's influence (see Appendix Table C-7). A hypothetical alternative distribution without extension influence was assumed to begin in 1980 and to be complete by 1999. This could arise, for example, if a few families learned some of the new techniques by radio and these were gradually picked up by their neigh- bours. This would be a much delayed and slower process than the extension-influenced distribution. There is some justif- ication for assuming a longer - perhaps infinite - distribut- ion "shift" in this case but, because of the widespread use of radio in the area and the mention of it in survey reSponses, this alternative has been used. The calculation of the benefit flow attributable to extension work is shown in Appendix Table C-7 along with the present (1967) values 7 O O 0 Because these costs were incurred in different years, an interest charge of 4 per cent was levied against 1966 costs and is included in the total. 363 of these benefits discounted at 0, 4 and 8 per cent. These estimates of benefits and costs were used to calculate the PVB/PVC ratios shown in Table 50. In addition, Table 50 shows PVB/PVC ratios calculated for benefit flows of 18,000 pesos and 50,000 pesos per annum for 31 years. The former is regarded as a "lower limit" and represents 100 families (all resident families) benefiting by 500 pesos per year. This permits testing the sensitivity of the basic PVB/PVC ratios to some of the assumptions underlying the benefit flow calculations. An expected PVB/ PVC ratio of 3.43 with possible extremes of 13.96 and 0.83 are noted. The District of Buenos Aires Although the socio-economic study of Buenos Aires was carried out in 1964, most of the extension programs did not really get going until 1965. While this study originally attempted to isolate the results of work done in 1967, it appears that what actually was observed was the results of work carried out in the period 1965 to 1967. There were several reasons for this. Firstly, much of the work directed at consumption activities was carried out by obreros, but, according to information supplied by extension personnel, no obreros worked in Buenos Aires in 1967 (see Appendix Table C-6). Secondly, in many production activities, assistance had been continued for more than one year. Thirdly, the 364 Hm.N hm.m om.m ooo.om ow.H a¢.N mN.s N-o mHnms xHoamdds cH mm ooo.m«N mm.o mN.H hN.N ooo.mH NH.N mm.¢ mm.m ooo.om 0N.N m¢.m hm.m >10 mHnma xHUCmmmd CH mm oom.me wH.H bb.H NH.N ooo.mH HH.m Nm.h mm.mH ooo.om sm.m Nm.m m¢.m Nuo mHnms ancmam< cH mm coo.HHH vm.H mm.N mo.m ooo.mH mm we we HECCCM\m0mmmv Anemone msmumoum COHmCouxm mumou mumm quoomHo ou muHmmCom ConCmuxm .muHmmCmn oCm mumoo mCHUHmmmH mCoHumEsmmm quHmmmHo Hmum>mm How mmuHmmCHB CH xuos ConCmuxm How moHumu O>m\m>mnu.om oHnt 365 extension offices had provided a list of all assistance rendered to families in the sample and, as many families could not specify exactly when the assistance had been received, it would have been inappropriate to regard all changes as attributable to work done in 1967. However, it seems likely that a considerable portion of the work in Buenos Aires must have been carried out in 1967 as six out of the fourteen families interviewed had not participated at all prior to 1967. Estimates of the annual costs of 1965 and 1966 extension programs in Buenos Aires were constructed. These incorporated a deduction of 25 per cent8 from costs for each of these years to compensate for the costs of producing physical facilities such as latrines, house improvements, etc. whose benefits were regarded as equal to or less than costs. Again, because of the uncertainty surrounding the actual costs of producing the results measured by this study, three different estimates have been used in calculat- ing PVB/PVC ratios. The "lower limit" (120,000 pesos) consists of an estimate of 1967 extension costs alone. The "upper limit" includes also 1965 and 1966 extension costs (excluding obreros) and amounts to 450,000 pesos. (This includes an allowance for interest charges at 4 per cent). The simple average of these two figures is used as the 8This represents the proportion of costs attributable to obreros in the four districts studied. 366 "expected value" of costs for the work in the district. The calculations of the projected benefit flow for extension work in Buenos Aires is shown in Appendix Table C-8. This is based on benefits beginning at 465 pesos per family, which was the average for the fourteen families inter- viewed in Buenos Aires, and increasing to 700 pesos. It assumes that by 1982 benefits will be received by 124 families - the total number resident in Buenos Aires in 1965. This was five years earlier than in other districts, mainly because of the probable interest and influence of the Coffee Feder- ation in the district. The alternate distribution was assumed to reach a benefit level of 465 pesos per family, approximately half-way through its existence (i.e., by 1983). On the basis of the surveys, there was little doubt that some diffusion was occurring in Buenos Aires. However, the above assumptions do not necessarily accurately reflect the extent of this diffusion or the likely "shift" in the adoption distribution caused by extension effort. Because of the uncertainty associated with the above assumptions, PVB/PVC ratios were calculated by "lower” and "upper" limits of benefit flows as well for the "expected" level. The "lower" limit was based on 68 families benefiting at the rate of 465 pesos per year. The "upper” limit was based on 125 families benefiting at the rate of 700 pesos per year. Both were assumed to extend for 27 years, the same duration as the "expected" value stream. The various PVB/PVC ratios calculated for extension smograms in Buenos Aires are shown in Table 51. An "expected" 367 mH.N mH.m mN.m oom.hm HH.H ew.H mo.N muo mHnme xHoCmmmC CH mm ooo.omv bb.o «H.H mm.H oom.Hm om.m Ho.m mN.m oom.>m HN.H mm.N mH.s m-o mHnms chcmmm<.cH mm ooo.mmN HN.H om.H mm.N oom.Hm hm.> HN.HH mm.mH oom.hm mH.¢ oH.o mm.m muo mHan xHoCmmmm CH mm ooo.0NH hm.N mN.v mo.h oom.Hm wm wv mo HECCCM mom momwmv HmOmmmv mamumoum ConCmuxm mumou ovum quoomHo on muHmmam COHmCmuxm Illlllllllllnlllllulllrlllulllllllllllllllllll .munmaQ va mumoo mCHoummmu mCoHumEswmm quummmHo Hmum>mm How mmHHC mOCmsm CH xuo3 COHmCmuxm How moHumu O>m\m>mII.Hm mHnma 368 PVB/PVC ratio of 2.59 is noted along with possible extremes of 19.69 and 0.77. These data illustrate the uncertainty underlying the estimates of benefits and costs but, when one considers the benefits not included in estimated benefit flows, it seems evident that extension programs in Buenos Aires were very likely returning sub- stantially more than their cost. The District of Pinguro Most extension programs in Pinguro began in 1964, the year after the socio—economic study of the area was completed. Only the school and a few other community projects were carried out in 1963. As in the other districts it became apparent that the information from the surveys was probably describing the results of extension programs throughout the entire period of extension activity - not just 1967 alone. In fact, all thirteen of the families interviewed indicated that they had participated prior to 1967. Accordingly, estimates of costs for the years 1964 to 1966 were constructed using the same method as was used for Buenos Aires. The resultant total estimated cost for all years was 493,000 pesos. The cost for 1967 extension programs alone, excluding obreros was approximately 90,000 pesos. (Obreros accounted for 91,000 pesos in costs to give total estimated costs of 181,000 pesos in 1967). The simple 369 average of these two figures was used as an "expected value" of extension costs. The calculation of the projected benefit flow for extension programs in Pinguro is shown in Appendix Table C-9. This is based on benefits per family of 280 pesos in 1968 rising to 420 pesos by 1982. Average annual benefits for the thirteen families interviewed in Pinguro were 277 pesos. As with Tinajitas, the probable alternative distribution was assumed to begin in 1980 at a somewhat lower benefit level. Both distributions were assumed to include 99 families, the total number of resident families in the district. The various PVB/PVC ratios calculated for extension programs in Pinguro are shown in Table 52. The "lower limit" benefit flow of 12,500 pesos is based on 45 participating families benefiting by 277 pesos per annum. The "upper limit" of 42,000 pesos is based on 100 resident families benefiting by 420 pesos per annum. As in other districts these ratios will be in error mainly due to the fact that benefits from changes in community organizations and costs of the socio-economic study of the district have been omitted. The results are similar to those noted in Tinajitas and Buenos Aires. The "expected" PVB/PVC ratio is 1.48 with extremes of 14.47 and 0.29 being noted. 370 Hm.o om.H so.N coo.Ns mm.o Hm.o mm.H muo oHnme stcmsm< :H mm ooo.mms NN.o mv.o ms.o oom.NH «m.H mm.N NH.¢ coo.Nv mm.o mv.H mm.N muo mHnms xHocmmms cH mm oom.HmN ms.c ms.o HN.H oom.NH om.m HN.m HH.¢H ooo.Ns OO.H ms.v NH.N muo mHnme xHecmdme :H mm ooo.om mm.H sv.N Hm.v oom.NH wm we mo HECCCM Com mommmv AmOmmmv msmumoum COHmCmuxm mumoo mumm quoomHa on mUHmmam COHmCmuxm .muHHmCmn oCm mumoo mCHUHmmmH mCOHugfismmm quHmHMHc Hmum>mm How ousmCHm CH Huo3 COHmCmuxm How mOHumu U>m\m>mll.Nm «Heme 371 The District of Urquita The results of extension work observed in Urquité appeared to be attributable to both 1966 and 1967 extension programs. Very little advisory work had been carried out in 1965 as the field work for the socio-economic study was being completed that year. Thus, 1966 extension costs were estimated using the same method as had been used in other districts and an "upper limit" of costs of 298,000 pesos was arrived at. The lower limit of 205,000 pesos represented 1967 costs (excluding obreros) and the average of these was used as an "expected value" of costs. Benefit flow calculations presented a problem because of the extremely skewed distribution of benefits in the sample of families interviewed in the district. It was considered rather inappropriate to use the average of 2,463 when one family accounted for 90 per cent of the total benefits in the sample. On the other hand, to omit the family entirely and use the average of 283 pesos observed for the remaining eight families surveyed would also have been inappropriate. It was decided to use the average benefits per family surveyed in the four districts (excluding the one family with spectacular results in Urquité) as an estimate of the average benefits per participating family in this district in 1968. Thus, it was assumed that benefits would rise from 325 pesos per family in 1968 to 500 pesos per family in 1980. Calculation of a projected benefit flow on this basis is shown in Appendix Table C-10. The probable 372 alternative distribution was assumed to begin in 1980 and be complete by 1993. Calculated PVB/PVC ratios for Urquité are shown in Table 53. The "lower limit" of the benefit flow is based on 37 participating families benefiting by 325 pesos each while the upper limit is based on 70 resident families benefiting by 500 pesos each. The results are not as variable as those noted in other districts. The "expected" PVB/PVC ratio is 1.39 with possible extremes of 4.27 and 0.43 being noted. All Districts Because of the interest it might have for persons interested in returns to extension in general, PVB/PVC ratios have been calculated for extension work in the four districts as a group. The results are shown in Table 54. Both the costs and benefit flows for individual districts were summed to arrive at total costs and total benefit flows. Thus, the result represents a weighted average of the four districts. A slightly lower variability is noted in the overall PVB/PVC ratios with extremes of 11.58 and 0.55. The expected PVB/PVC ratio is 2.12 which indicates that in general, extension programs in Antioquia were likely returning more than their monetary cost. However, this does not say that they were better or worse investments than alternative government programs which might have been carried out. Nor 373 mN.H mm.H «m.N ooo.mm an.o hH.H hm.H OHIO mHnma xHUCmmm< CH mm ooo.mmN m¢.o no.0 Ho.H ooo.NH mv.H NH.N wv.m ooo.mm mm.o HN.H HN.N OHuo «Hams stcommm cH mm oom.HmN Hm.o mb.o mH.H ooo.NH Nm.H no.N NN.v ooo.mm «H.H os.H NH.N OHuo «Hams stcmmma aH mm ooo.m0N Nm.o Hm.o wv.H ooo.NH wm we mo AssCCm you mowmmv Hmommmv msmumoum COHmCmuxm mumoo ovum quoomHo ou muHmmCmm ConCmuxm .muHmman oCm mumoo mCHoummmH mCoHumEsmmm Hmsm>mm umoCs wuHsqu CH xuoz ConCouxm How mOHums U>m\m>mun.mm mHnme 374 om.H NH.N 0H.v =uHsHH amass. Hm.o m¢.H mm.N .maHm> omuommxmg ooo.Nms.H mm.o mm.o H¢.H =uHsHH umon= mm.N sm.m mo.w =uHaHH amass. mm.H NH.N Nm.m .msHm> omuommxm. oom.moo.H Hm.o HN.H HO.N .uHsHH umsng Hm.¢ mm.m mm.HH =uHsHH Comma: mm.N mo.q ms.m .msHm> smuommxm. ooo.on mm.H mm.N mm.m =uHsHH umong mm we so Hues Hmommmv Gowmcwuxm EOHH mumOU mama quoomHa son HHmmCmm ConCmuxm .omHosum muoHHumHo snow was CH msmumoum ConCmuxm How mOHumu U>m\m>mn|.vm mHnma 375 does it say that tax revenues should have been increased (or decreased) to expand (or cut back) such programs. As no welfare loss due to taxation has been included in the costs, it may be that these programs are returning less than their social costs despite such apparently favorable results. Program Performance Ranking by District A summary of the factors relevant to ranking the effectiveness of extension work in the four districts is shown in Table 55. No changes in rankings are noted with variations in the discount rate although it is evident that Pinguro and Urquita programs have equivalent "expected" PVB/ PVC ratios at 8 per cent but differ at lower rates. At rates higher than 8 per cent, Urquité programs would exhibit a higher "expected" PVB/PVC ratio than those in Pinguro. All districts exhibit considerable possible variance in PVB/PVC ratios and, as would be expected, this variabil- ity is not symmetric about the expected value. Thus, it is difficult to compare districts directly on this basis. Such a comparison is not entirely unnecessary as the districts of Pinguro and Buenos Aires, although they exhibit lower "expected" PVB/PVC ratios than Tinajitas, exhibit higher possible or "upper limit" ratios. Buenos Aires, for example, has an almost identical PVB/PVC ratio to Tinajitas at the "lower limit" and a substantially higher ratio at the "upper limit". Thus, depending on one's utility for large gains, it 'would be possible to rank Buenos Aires above Tinajitas. 376 Nm.H - mv.o mm.o Hm w HO>M = = HmmsmHmm sm.N . mm.o HN.H as wustus HN.H . Ho.H HN.N so om.m . mN.o mm.o so = = mHnmuo>mm HN.N . m¢.o mv.H we ouamch NHsmHm NH.¢H . ms.o NH.N mo NN.N . NN.o mH.H mm mHnmuo>mm . . . : : hflflmflm Hm HH I VH H mm N WG mmHflé mogmam mm.mH u mm.H mH.¢ so N mmocmno HH.m . mm.o ON.N mm exam :oHum mHnmuo>mm . . . oHsocoom -uHammuo sHsmHm Na N - mN H me n wv mmuHHmcHa -oHoom suHcsssoo om.mH . HN.N Hm.m so mumou muHmewm muommmm 0Humm oHumm mumm HOHuumHo :oHuanHuumHo o>m\m>m mo o>mxm>a uasoomHo mOHumm O>m\m>m CH mEOOCH mmCmm anmnoum omuommxm omosHocH uoz !; .msmsmoum ConCmuxm m.quEuHmmmo on» HOH mwusmmms moCmEHomHmm mo mumsssmnl.mm mHQma 377 Income distribution effects were essentially the same for all districts.9 These are described as "highly favorable" to emphasize the fact that, for the most part, extension participants were members of the low-income class. However, there did not appear to be any basis for saying work in any one district influenced income distributions more favorably than work in any other. Benefits and costs not included in PVB/PVC ratios were all of the same nature for each of the four districts. Whether or not their levels varied among districts is not possible to say here. However, in view of the fact that the activities involved were very similar it seems safe to assume that these benefit and cost levels were about the same. On the basis of this limited information it would seem that the four districts might be ranked as follows with regard to extension effectiveness. 1. Tinajitas 2. Buenos Aires 3. Pinguro 4. Urquité However, because of the large degree of uncertainty assoc- iated with the expected PVB/PVC ratios, it is not possible to say that this ranking is a very significant one. 9This disregards the single outstanding family in Urquité although they undoubtedly had a much above average income for the district. 378 Factors Affecting Differences in Program Performance Differences in program performance are again dis— cussed under the headings of (a) cost differences, (b) unit benefit values, and (c) diffusion effects. However, because the program comparison is by districts, these prove much less interesting than in Valle del Cauca and thus, a further category of "type of program" is introduced to permit exam- ining differences noted among different production and con- sumption activities. Cost Differences Extension costs per participating family and per resident family are shown in Table 56. In view of the fact that some considerable uncertainty existed regarding costs, it is difficult to draw definite conclusions from these data. However, there appears to be an inverse relationship among costs per family - both resident and participating - and "expected" PVB/PVC ratios. The causes of these apparent cost differences are not known as similar extension methods appeared to be used in all areas. It appears that, even though Tinajitas and Buenos Aires were the most remote of the four districts, costs for servicing families in these two districts were lower than in the other two. As Urquité was serviced by the San Jeronimo office perhaps differences in operating practices among offices may account for part of the 379 mMMH mmmN «hum mHmN mHm¢ mHHn muoHHumHQ HHd mNmN mmmm Nva mmmm mHmm News wanvus mom vva owmv m¢ON mNmm momHH ousmCHm mom mmNN mNmm mmnH Hch mHmo mmHHC mOCmsm mmoH mwbH mmvN vmmH mNom mmHv mmanMCHB UHEHH msHm> HHEHH uHEHH wsHm> uHEHH szoH cmuomaxw Momma Hm3oa omuowmxm Home: wHHEmm quonmm “mm mumou .mHHEmm mCHummHoHusmm mom mumou .mosum mHCu CH omCHmem mfimumoum ConCmuxm mo >HHEMH HCmUHmmu mom ch >HHEMM mCHummHoHuHmm Ham mumooul.mm mHnma 380 differences in costs. It would seem that the difference is sufficiently large to warrant concern as costs in Urquité appear to be approximately double those in Tinajitas and no corresponding difference in benefits was apparent. Unit Benefit Values As the family was chosen as the unit upon which to base benefit calculations, data relating to benefits per family have been summarized and the resultant statistics are shown in Table 57. It is obvious from the high standard errors that the differences between districts are not significant statistically at the usual confidence levels. However, as the averages were used in arriving at the PVB/ PVC ratios calculated earlier, differences between the averages undoubtedly influenced the numerical results.10 Nevertheless, it would be inapprOpriate to conclude from the sample data that substantial differences in benefits per participating family exist among districts. The unusual nature of the distribution of benefits among participating families interviewed merits comment. As noted in Table 57, eleven of the fifty-four families inter- viewed reported no measureable11 benefits at all, whereas six families reported benefits of more than 1,000 pesos per 10This was partially reflected in the wide variations in PVB/PVC ratios exhibited earlier. 11"Measureable" means capable of being reduced to monetary terms. 381 .mHHEMH mHCu mosHoxm any omHOCoc mmHuqu .uOHuumHe we» CH muHmmCmn ummanC on» CuH3 MHHEMM 0C» wUCHOCH Amy OOHOCmo mmHHqu “muoz HH HoN.¢ I o th AEmUOHuumHo HHC HH www.mH I o mam AmemHOHuumHo HHC m mes oom.H - o mmN Ankustus omHMHsono m uoz www.mH . o mes.N .mkuHsous q oov va.H I o th ousmCHm N mom NmN.H . o NHH AammuHC mocosm N OHH.H HON.¢ . o mes AmvmmuHC moamsm s mHH omN.H . o mNm mmuHHmcHa muHmmCmm Hmommmv Amommmv Anemone 0C CHH3 uouum mHHEmm hHHEmm mmHHHEmm UnmoCmum mom “on mo .02 muHHowm muHHowm mo mmCmm momsm>m .omHosum muoHuumHo was How omBmH>HmuCH mHHEmm Mom muHmmCmmlu.nm mHnme 382 year - about three times the average. The average annual benefit for families receiving benefits - excluding the one rather unusual case in Urquité - was 412 pesos or 26 per cent higher than the average with those not benefiting at all included. The significance of this sort of distribution is by no means clear. It may be that extension personnel were attempting to work with late adopters who would not likely adOpt until they saw results on their neighbours farms. It may be that recommended practices were just not applicable on some farms. Diffusion Effects There was no evidence to indicate differences in diffusion among districts. However, the decision to use the total number of resident families as the total population which would ultimately benefit obviously affected the PVB/PVC ratios somewhat. This decision rested on the fact that the principal sources of benefits required little or no land. As some of the minor benefit sources do require some land, the use of resident families may cause an overestimate of the benefit flow. Obviously, the techniques in certain areas will have a much more limited potential for diffusion than others. This applies to both numbers of families involved and scale of Operation. Changes in coffee production methods, for example, could possibly diffuse to all farm families in a 383 district such as Buenos Aires. However, the scale of operation on farms adopting after 1967 might be quite different than those adOpting before. This was certainly the case in Urquité where one family had the largest coffee area in the district and, as noted earlier, adopted some changes and benefited substantially. It is considered that the benefit flow estimates used reflect this situation adequately but only time will tell whether or not this is the case. There was the possibility that some of the new technology would not diffuse but, perhaps, would even be rejected by farmers considered in this study to be adopters. Several instances of rejection were noted in the surveys. These cases show up in the data as families recalling assistance but not experiencing income increases. For example, several families reported they would not plant vegetable gardens the coming year if the extension service did not supply the seeds. A few vegetables (e.g., carrots) were not very popular among some families because of the differences between them and the vegetables which were already standard fare in the region. Type of Program Table 58 provides a summary of the results of extension programs directed at the production activities of the 54 families interviewed. These data indicate that a few programs could be regarded as widely successful whereas 384 NH mHuumo mCHumCHoom> mHH HH mH H NN mcHnuoHo ¢0H oH mH N bH mummuoHoCmm mmN N s N w mmmm N+mNm.m NH+H N N HH auuHsom omm N m v m mCH3m «NH n mH v mH manndm mNm H o H H amouo umnuo mmv m o m m HmmHm Hem 0N Hm m ow mmHnmummm> mvv N m m oH muHsum ovm.N m h h vH mommoo Hmomomv muHmmCmm mUHmmCmm moCmumHmmC om>Hmomm h oaH CH Cqu mmHHHEmm uomuHo usmummmm mCHHHmowm mUCMHmHmmC mOCmumHmmC Emumoum How mHHEmm CHH3 mmHHHEmm CuH3 mmHHHEmm 0C» qCHHHmomm mCH>Hmowm mo make Com muHHmCmm mo Honssz mo Honssz uOC mmHHHEmm moHHHEmm mo HmsCCC mmmum>¢ mo HmnEsz .muoHuumHo HsoH CH umnEsz Hmuoe meHHEmm mHmEmm vm mCosm Emumoum an muHmemn mo mumsfismll.mm mHnms 385 many were quite limited in extent of success. Vegetable gardens and clothing fell into the former category whereas such items as swine, poultry, bees, and fruits belong to the latter. Handicrafts were relatively widely-adopted but this was one case in which future prospects looked poor because of limited markets. Coffee, although it was the principal source of cash income for many families surveyed, did not appear to be a crop where changes were being widely- adopted. Most of the farmers with small acreages were reluctant to take the short-term losses necessary to change varieties even though longer-term gains seemed sizeable. However, they might well do this when they witnessed the benefits on their neighbours' farms. As the four districts surveyed are not highly-adapted to coffee, it may be that benefits due to better technology in these districts are not as large as they might be in other regions. The differences in degree of success among types of program is not readily explained. It cannot be assumed that extension effort was constant among all program types. Extension personnel were emphasizing certain things in some districts (e.g., sisal in Tinajitas) and not in others. It does not appear that the number of families adopting in the area is very closely related to profitability as measured by average benefits per family adOpting. However, because of the small numbers involved, it would be inadvisable to place very much weight on this statement. Vegetable gardens and clothing appear to have been successful because: 386 (a) they were already essential parts of the families' lives, (b) changes could be made with little cash outlay on the parts of the families involved. (This appeared to limit the level of benefits in the clothing area.) (c) techniques were easy to learn and their benefits could be readily seen. Risk and limited resources (i.e., working capital) were likely instrumental in limiting the success of other programs such as poultry, swine, bees, etc. Adoption in programs directed at consumption activities appeared to hinge directly on whether or not material aid was supplied to the families. Accordingly, one can reasonably infer that income was the limiting factor here and that, if incomes in the region were higher, some of this technology would be adopted without provision of material aid. Of course, even if incomes were higher, adoption would not likely occur without some sort of educat- ional program to point out the advantages of the new ideas. The extension service's success in the area of community organizations seems largely attributable to provision of obreros and funds for community projects. As thirty out of the fifty-four families interviewed reported benefiting from these programs it seems reasonable to regard them as successful in some sense. Of course, as the families concerned did not bear the full costs, PVB/PVC ratios for these programs might well have been small (i.e., less than one) despite this reaction. 387 Cost Minimization Cost minimization considerations may be analysed by examining expenses in two categories: "salaries and fringe benefits of field personnel" and ”other costs". These categories accounted for approximately 52 per cent and 48 per cent respectively of the total costs of operating the extension service in 1967. The former category includes salaries of extension workers in the field, wages of obreros, and all primas, bonuses, pension allowances and similar fringe benefits for these employees. The latter category includes costs of office maintenance, transport- ation, materials and supplies, administration, the services provided to the field workers by the communications and research divisions of the Secretary of Agriculture and all other goods and services purchased in 1967. Both cost categories include expenditures for "management" activities because both administrative and field personnel were involved in making decisions regarding the nature of programs offered. Both categories also include expenditures for "program operation". It was not possible to obtain Specific cost estimates of these two classes of activities.12 However, it appeared that a 12Some of those "management" costs incurred prior to 1967 (e.g., the socio-economic studies) should have been charged to 1967 programs because they were costs of choosing 1967 programs. On the other hand, part of the 1967 "manage- ment" costs were probably incurred in planning programs which would continue in 1968 and perhaps 1969. Thus, even if a cost division between "management" and "program 388 significant proportion of total costs were devoted to "management". In the two regions studied, 14.2 per cent of total field personnel time was devoted to "office work" and 40.8 per cent was devoted to activities other than farm visits, meetings, demonstrations, tours, office work, and travel. Although the exact nature of the "other activities" is unclear it appeared that a considerable proportion was devoted to administrative duties, organizing community activities, and planning sessions while training sessions may have accounted for the remainder. Perhaps 20 to 50 per cent of "other costs" were attributable to "management" activities. A large part of administrative salaries, maintenance of administrative offices, and costs of collect- ing and processing data on field operations were clearly "management" expenses. However, tranSportation, materials and supplies, field office maintenance, and services of the research and communications sections were essentially all "program operations" costs. Thus, perhaps 20 to 40 per cent of total 1967 costs were "management" costs. There is little evidence to suggest that the extension service was not minimizing costs for the ”manage- ment" activities which were performed during 1967. The reports which had to be completed by field personnel naturally took a good deal of time. Similarly, field programs could operationf’hadfibeen made for 1967 expenses, the resultant "management" costs would not necessarily have been representative of the "management" costs of 1967 programs. Of course, even greater difficulty would have been encountered in measuring the costs of the programs actually analyzed as these extended over several years. 389 not be carried out and modified without field personnel spending some time analyzing what was happening in the various programs. The collection and processing of data and the carrying out of the various administrative functions that were part of the existing management system also were time-consuming. There probably were cheaper (less time- consuming) ways for reports to be completed, existing pro- grams to be observed, new programs to be analyzed, decisions to be reached, and decisions to be carried out. However, an analysis of these problems would have been quite costly and the possible gains relatively small. Thus, it will be assumed here that the extension service was, in a practical sense, minimizing the costs of the "management" activities it carried out during the year.13 A similar conclusion can be drawn about the "program operation” activities conducted during the year. Table 59 shows the percentage distribution of field personnel time by the type of activity and class of personnel for the Santa Fe de Antioquia and San Jeronimo offices. On the average, 18.1 per cent of personnel time was spent visiting farms. Mejoradoras de hogar spent less time in this activity than other workers because much of their work was done with groups rather than individual families and could be carried out in a 13This should not be interpreted as meaning that no improvements could be made in the management of the extension service. In the author's opinion the extension service could have operated a much better management system for the same cost. Thus, while what was done was probably done at least cost, a somewhat different system could have produced better results. 390 .mmonmo onHH 03» on» an cmHHm muuommu mHCuCoE mo mmHunEEsm scum cmusHCOHmum o.o0H o.ooH o.o0H o.oOH o.oOH o.ooH o.o0H Hmuoa m.o¢ o.v¢ m.Hm m.vm o.mv m.wv w.om mmHquHuum nonuo b.mH H.vN N.MH H.0N m.mH N.mH N.mH Hm>mue N.¢H o.HH N.Nm m.mN m.m v.h o.vN xnoz mUHmmo N.o I I I N.o «.0 m.o muses h.m N.H m.N 5.0 v.N m.MH m.o mCOHumuam ICosmo H.H O.H H.NH H.N o.N N.H o.N mmcHummz H.mH >.mH H.m m.mH m.hN N.m >.mN muHmH> snub mommmHo mMHODHHmm Hmmmm mOHHmCHkum> mmHouwmmd ummom moHHMCHHmum> HHC mOUHuomum mo oCm mOOHuomnm mm UCM mmuoomHOnmz mOEOConmw mmuooMHOnmz mQEOConmd muouomm OEHCoumh Cum nuom mHsvoHuCC ow mm mqum m.hmmH CH OEHCoumn Cum oCm MstoHqu on mm mqum mo mnouowm mCu How HwCCOmHmm mo mmmHo oCm >HH>Huom mo Oahu an mEHu HmCCOmumm onHm mo COHuanuumHo mmmquoummII.mm MHCMB 391 central location. Practicos and agronomos might have used this approach to a larger extent. The author was surprised. to find that the monthly reports indicated that the agronomos spent about the same proportion of their time on the farms as the practicos because they were not nearly as well known in the districts surveyed. Travel accounted for 18.7 per cent of field personnel time on the average with practicos in San Jeronimo being the highest of the six groups with 24.1 per cent. Summaries of data submitted by the Santa Fe de Antioquia and San Jeronimo offices indicate that on the average, field personnel in these offices were able to cover only 12.8 and 18.3 kilometers per hour respectively. This is a reflection of the road conditions in the area. Some of the practicos lived in the districts and walked to the farms. Within some of the districts, cars could not be used and travel was exclusively by foot or horseback. Thus, while travel time used up a significant portion of the working hours there did not appear to be much that could be done about it. CHAPTER X POSSIBILITIES FOR IMPROVING EXTENSION EFFECTIVENESS IN COLOMBIA This chapter will focus on the two categories of extension activities delineated in Chapter III - program operation and extension agency management. The purpose of the chapter is to suggest, on the basis of both conceptual and empirical considerations, guidelines for improving extension effectiveness or efficiency within the existing budget allotments. As indicated at the outset of the study, no consideration has been given to whether or not budget allotments should be raised or lowered. Extension Programs On the basis of the observations made in this study, there is no reason to believe that any one of the "outputs" of extension work defined in Chapter II should be entirely ignored by Colombian extension agencies. There appears to be potential for effective extension work in promoting technical changes in farm firms, rural households, and community organizations, and in promoting managerial changes in rural society in general. There also appears to be potential for effective extension work in performing the four 392 393 secondary extension functions outlined in Chapter II. In addition, there may be potential for effective extension work in the marketing of agricultural products - an area not investigated in this study because of the little current activity in it. However, in various parts of the country it will probably be appropriate for extension workers to focus on one or more particular "outputs" and exclude one or more others from consideration. This conclusion obtains because of the relative potential pay-offs for producing different "outputs" in different regions, the costs of maintaining extension personnel capable of focusing on all "outputs" in all parts of the country, and the differing effects on income distribution of programs directed at producing different "outputs" under different environmental conditions. An attempt will be made to clarify this rather general con- clusion in the following discussion. Agricultural Production This study indicates that extension programs directed at inducing changes in farm firms in Colombia can have a very high pay-off1 in relation to the costs of carrying them out if (a) the technology being promoted has a high pay-off lAs agricultural extension and agricultural research are, to a large extent, complementary, the pay-offs noted in this study need to be interpreted with care. It must be remembered that no research costs have been included in the calculations and that the returns measured must, in the long run, cover these as well. 394 for potential adOpting farmers and (b) extension programs can achieve a substantial shift in the adoption distribution among farms.2 Determinants of both of these conditions lie, to a considerable extent, outside of the jurisdiction of extension workers. Extension personnel can only select from available ideas those which they believe will achieve the first of these conditions and promote them using ways and means which they believe will achieve the second in the existing environment. It is not possible for a study such as this to recommend specific programs which would increase extension effectiveness in dealing with farm firms in Colombia. However, it is possible to suggest some general guidelines. These will be discussed with reference to the two general conditions mentioned above and income distrib- ution and extension cost considerations. However, it must be noted that there is considerable interaction among these two conditions, income distribution considerations and extension costs. Benefit flow levels resulting from very high pay-off technology adopted on a few farms may be essentially the same as those resulting from low pay-off technology adopted on many farms. Furthermore, income distribution effects and extension costs might be equivalent or significantly different in two such situations. The following general points apply to an analysis of the potential pay-off of agricultural technology for adopting 2This conclusion may have to be modified for income distribution considerations as discussed in Chapter V. 395 farmers in Colombia: (1) (ii) (iii) (iv) The potential pay-off to individual farmers will usually vary directly with the size of the operation (measured in units of output). This would imply that, other things being equal, extension workers should promote techniques applicable to larger farms. However, diffusion possibilities, income distribution considerations, or extension costs, may not "be equal” and hence negate this conclusion. Techniques of relevance on larger farms may be totally irrelevant on small or medium-sized farms and vice versa. This applies both because of risk considerations and the different environmental conditions surrounding the farming operations of different sized units. Mechanization is an obvious example of a set of techniques of relevance to large farms but largely irrelevant to small farms. However, there are others. Narrow profit margins per unit of output may make the total pay-off for a large producer quite significant but may not be significant enough to offset the risk for a small holder. This could apply to new varieties, fertilizers, insecticides and similar techniques. Of course, some techniques (e.g., new varieties) might apply to both large and small units. The potential pay-offs of particular types of tech- nology on individual farms will usually be much greater if the technology can be used on a number of products (with perhaps different treatment levels). As many Colombian farmers grow several crops, techniques such as fertilization, insect control, disease control, and so on, fall into this category. As diffusion possibilities are likely to be greater as well, while extension costs may be only slightly higher and income distribution effects are likely to be invariant, extension programs directed at promoting multiple-use technology (such as fertiliz- ation) are likely to be more effective than those directed at promoting single-use technology (such as control of insects unique to particular crops). In many cases, there may be pay-offs from foreign agricultural technology on individual farm firms in Colombia. However, because of price differences, limited input availability, and different input/ output relationships, the profitability of such technology may be vastly different than that in the country from which the technology originates. This does not necessarily mean the pay-off is lower - it 396 may be higher.3 However, it does imply that field trials may be necessary to assess profitability and risk under local conditions. (v) The merits of technology to the farmer should not be judged on the basis of expected profitability alone, especially on small operations. Some combination of profit (allowing for opportunity costs) and risk considerations seems appropriate for making such assessments, eSpecially where the new technology creates a possibility of monetary loss which had not existed before. The use of purchased inputs such as fertilizers, Sprays, feed supplements, and so on, fall into this category as, in many cases, traditional Colombian farmers produce crops or livestock with no purchased inputs and hence, in the case of crop failure or death of livestock, lose only their labor. (vi) The provision of credit for farmers with small holdings may eliminate one of the barriers to adoption on individual farms but it may not achieve the desired results if farmers are making "minimax" decisions (see (v) above). For very low income farms it may be necessary to offer rural entrepreneurs a virtually risk-free proposition to encourage adoption. Thus, it might be useful to try some programs where the credit need be repaid only if the technology produces the claimed results. (vii) It should not be assumed a_priori by extension workers that, for every community in Colombia, there exist available techniques with high pay-offs for farmers in the area. Because of the limitations to foreign agricultural technology (see (iv) above) and the limited resources which have been devoted to agricult- ural research in Colombia in the past, there may be many areas in which few pay-offs are possible. (Indeed, because of the uniqueness of conditions in some areas and the stringent character of the local environment, research designed to discover technology suitable to these areas may be out of the question.) In some parts of Colombia there may be no possibil- ities for doing effective extension work directed at promoting changes in agricultural production. (viii) The adjustment to economic optima for new or existing technology offers a possibility for effective extension work, particularly if product and input price changes 3Indeed, it may mean that some techniques which are known but regarded as uneconomic in advanced societies, have a very high pay-off in developing countries. (iX) (x) (xi) 397 are frequent, or the levels of uncontrollable factors vary frequently. While diffusion potential for this sort of change is often limited, it can sometimes be achieved at low cost and hence, it may be quite competitive with other extension work. Such low-cost extension work would likely involve publications and/ or use of the mass media to reach a large number of farmers. Even though diffusion is small or non- existent, such work can be justifiable from a social point of view if the cost of carrying it out is low in relation to benefits per farmer and, if private sources are not meeting this need. The potential pay-off of agricultural technology to individual producers is heavily dictated by the opportunities created by their environment. In Colombia, the environment dictates that many individ- uals will never have any land, or, if they do, this will be a very small parcel. It also dictates that they will not likely be involved in livestock product- ion or some other types of commercial agriculture. Thus, there seems little point in extending some types of technical agricultural knowledge to such rural people as there will not likely ever be any opportunity for them to use this knowledge. In both the CVC and the Department of Agriculture of Antioquia a number of cases were noted where instruction of this sort was given (e.g., individuals who were taught to care for cattle had little prospect of ever doing so). Many rural residents will have some forms of current endeavour which can be exploited more fully, but if not, then perhaps extension resources can be utilized more wisely in other areas than in extending tech- niques which will not likely ever be used. The introduction of new crOps or types of livestock to a given area may involve cost structures significantly different from existing ones. They may require inputs which are not currently available and they may place entrepreneurs in risk situations which are quite different from any of those which they have ever faced before. Therefore, field trials will likely be necessary to assess the merits of such technology to individual communities. The use of field trials and demonstrations (see (iv) and (x) above) by the two agencies examined in this study was quite limited. A related extension method involved supplying a number of local families with seeds, seedlings, or other similar supplies. The families in question regarded these "gifts" as the full extent of extension assistance. It is this 398 writer's hypothesis that a few trials of such technology on farms of interested persons coupled with complete demonstration and instruction in related cultural practices and complete follow- through to measure results and demonstrate them to the rest of the community would have been more appropriate. The levels of profitability of some of the techniques being promoted by "gift" of seeds, etc. were not even well established in the minds of extension workers, let alone among the recipients. (xii) The availability and price of inputs together with the marketing cost and availability of product markets are a significant cause of the difference in relevance of technology to large and small farms mentioned in (ii) above. Fertilizer is reasonably- priced and widely-used in the floor of Valle del Cauca where large farms predominate. However, it is very expensive to the minifundista in many parts of Antioquia - especially when he must spend several days carrying it home by horse or on his back. Similarly, minifundistas in remote areas find that their products command relatively low prices at the farm. While much cansbe done to improve marketing conditions in Colombia and extension agencies may even contribute significantly to such improvements, physical conditions dictate that some of such diff- erentials can never be eliminated. On the other hand, it may be appropriate in some areas as Felstehausen suggests, for extension agencies to promote roads and other public services so that such differentials can be reduced. Unless such reductions occur, producers in such areas will always be at a relative disadvantage in the market place. 4For example, in visits to different areas of .Antioquia in the course of this study, this writer noted oranges being left to rot even though these were being sold regularly to consumers in Medellin and Cali. The reason for this was that prices were currently so low as to not even cover the transportation costs to such markets. 5Harold Riley, et. a1. "Market Co—ordination in the Development of Cauca Valley Region - Colombia", Research Report No. 5, Latin American Studies Center, Michigan State University. 61bid., pp. 364-365. 7Felstehausen, op. cit. 399 The following general points apply to an analysis of the effect of extension work on the adoption distributions of technology on farm firms in Colombia. (i) This study uncovered no evidence to contradict, and (ii) (iii) some evidence to support, the general concept of adoption advanced by Rogers. Thus, it would probably be useful for Colombian extension personnel to View their clients as innovators, early adopters, late adOpters, and laggards. The possibilities of shift- ing any adoption distribution are likely to be far greater by working with innovators and early adOpters than with other groups. However, because of the variability in size of farm and environmental conditions in Colombia, it would seem appropriate to identify regional and sub-regional groups of farmers and work with innovators in each of these groups. Sub-regional groups might be usefully identified on the basis of products produced and income level of size of operation. Some technology may have a very limited potential for diffusion because there are only a small number of producers of the product(s) affected. Specialty crops (e.g., tomatoes in Valle) or products whose total output value is low (e.g., honey) are cases in point. Effectiveness of extension programs in such cases will be limited unless benefits per producer are high, extension costs per producer are low, or income distribution is much more favourably influenced than by competing programs. Some technology may have a very limited potential for diffusion because, even though there are a large number of producers of particular products which it affects, the number of these producers to which it is relevant is small in relation to the total population. This observation appears to be particularly signif- icant in Colombia because of the wide variation in size of farm and environmental conditions throughout 8For example, in Antioquia it was noted that extension workers were endeavouring to work directly with a large proportion of the families in the communities studied. Some of these families were much more receptive than others. The less receptive families could well have been waiting to see how techniques worked on neighboring farms before adopting them themselves. Attempting to get them to adopt at the same time as their more innovative neighbors was not necessarily the best use of extension resources. 400 the country. For example, new corn varieties relevant to the floor of Valle del Cauca are not relevant to corn producers in the mountainous parts of the same state even though they may be only a mile or two from the valley. (iv) Some technology may have a reasonably good potential for diffusion but may have a low pay-off for extension effort because it is being promoted through non- extension information sources. Thus, the potential for extension workers to shift adoption distributions is relatively small. In Colombia, producer assoc- iations for sugar, rice, cotton, barley, beef cattle, and other products are substituting for extension in these areas. In addition, input suppliers and/or the purchasers of some products also supply technical assistance thereby making the potential for extension workers to shift adoption distributions smaller than they would otherwise be. Of course, such situations should not be ignored entirely as extension can often "speed up" adoptions which are already occurring. (This seemed to be the case in the tomato program of the CVC). (v) Most of the non-extension information sources found in Colombia appear to compete with, rather than complement, extension work. However, in many areas such competitive information sources are absent or few, and extension workers find that, unless changes are promoted by extension agencies, the local pOpulation will be a long time finding out about them. In such cases the time "shift" in adoption attributable to extension will be substantial and may well compensate for high extension costs or low pay—offs to individual units. Such cases usually result in favorable income distribution effects as well. (vi) In the two agencies examined in this study there seemed to be relatively little extension effort directed at facilitating the diffusion process. It is this writer's hypothesis that it would be useful for extension workers to attempt to utilize innovators for focal points in local communities. Field trials, demonstrations, and meetings could be held on the farms of such individuals and, if it became known that these people were willing to share their experiences and knowledge with their neighbors, the total time for the adoption of useful ideas on farms in the area would likely be shortened. Another means of doing this would be through "awareness" programs utilizing the mass media. If people could be made aware that a few farmers in nearby communities were trying out 401 new ideas, the possibilities of diffusion from a few extension participants might expand considerably. The following general points apply to an analysis of the effect of extension work on the income distributions of Colombians as a result of changes in technology in farm firms. (i) The belief that work with high income farmers aggravates income distributions9 is not necessarily a valid one. In those cases where such farmers are producing products which are widely-consumed by low income people, shifts in supply can serve to pass these gains on to consumers and hence improve income distribution. In Colombia, products such as rice, beans, potatoes, corn, cassava, and plantain are cases in point. Conversely, if shifts in supply do not occur or, if the products concerned are consumed mainly by high income families, income distributions are likely to be aggravated. Products such as cattle, swine, poultry, milk, eggs, coffee, and perhaps sugar might well fall into this latter category although this situation is by no means clear cut as all income classes consume some of these products. (ii) Technological innovations which are sufficiently wide- spread as to cause shifts in supply and hence lower (or prevent an increase in) prices, may well have an adverse effect on some farmers. Technical changes in the production of products such as corn, coffee, and plantain which are produced on both large and small holdings in Colombia have a very great potential for cutting the incomes of low-income producers. Coffee for example, appears to be by far the best source of cash income for families in those parts of Antioquia studied. Should the Coffee Federation's technical advisory program, which is aimed at the better coffee-growing areas be instrumental in shifting the supply curve1 and lowering the price of coffee, the net effect in districts such as those examined in Antioquia would of necessity be lower incomes. In this case the benefits would be passed on to coffee consumers who, to a large extent, are high income families either in Colombia or elsewhere. 9This belief was encountered among some extension personnel by the author. 10This possibility is complicated by the fact that the price of coffee for export is not determined in a "free" international market. 402 (iii) In view of the fact that many rural residents of Colombia are employees of large and medium-sized farming operations, they can be adversely affected by agricultural extension programs directed at replacing hired labor with mechanized operations. Rice, corn, and sugar are cases in point. While it may be advantageous for owners of large farms engaged in the production of these crops to mechanize, there are some significant disadvantages for the workers who as a result, find themselves without jobs. These disadvantages are evident in increased pressures on cities by rural-urban migration patterns. (iv) Extension work directed at low income producers will not necessarily produce benefits for these farmers any sooner than if they had been ignored and extension efforts had been directed at high income producers. Observations made during the course of this study tend to support the notion that higher-income farmers are more likely to be innovators than their poorer neighbors.11 Thus, the poorer farmers, who tend to be late adopters and laggards may wait to see what the innovators do before they act, even though extension workers focus their attention on them. However, this conclusion would perhaps not apply nearly as strongly among similar groups of farms and regions as it would within these groups and regions. As noted earlier, it might well be appropriate for extension workers to classify producers into several groups on the basis of size and location and work with innovators in each group. (v) There appear to be some types of technology which are likely to have a positive effect on income dist- ribution in Colombia whether they shift supply curves or not. Improved techniques of relevance to crops which are widely consumed and produced by low-income classes are cases in point. Of these new techniques, yield-increasing varieties and improved cultural practices not involving purchased inputs are likely to be the most promising for adOption on low income farms. Unfortunately, extension workers are severely limited in what they can do in this area because few such varieties are available and little research in the area of improved cultural practices is being done. Of course, limitations in effective demand for such— commodities obviously exist so it is not possible to push this idea very far without taking such limit- ations into consideration. 11An exception to this may be the group of tradit- ionally-oriented farmers who own larger blocks of land primarily for cattle production. 403 (vii) Because of the social character of Colombia, it seems evident that extension programs directed at helping rural society may be more apprOpriately directed at making food cheaper for the consumer than at making greater profits for the producer, if situations arise where these two ends are, in fact, in conflict. This obtains because all members of rural society are food consumers while only a fraction of them are profit recipients. The following points apply to an analysis of the extension costs of promoting agricultural technology in Colombia. (i) Because of differences in environmental conditions between Colombia and developed countries such as the United States, there are significant differences in extension costs which must be allowed for in selecting extension programs. Salaries are much lower in Colombia but transportation and the costs of support- ing services are generally higher. Thus, it is apprOpriate for Colombian extension agencies to attempt to use a different mix of these resources than is found in North America or Europe.12 However, as these classes of resources tend to be complementary, care needs to be taken to ensure that efforts to reduce costs by changing the "mix" do not seriously weaken extension programs. (ii) Because of differences in environmental conditions within Colombia, there are significant differences in extension costs among regions. Lack of services, especially roads, in some rural areas makes extension costs per rural resident serviced so high as to reduce extension effectiveness considerably. The use of semi-professionals (i.e., practicos) who are willing to live in such areas andiuse public trans- portation is one means of keeping costs low. However, extension workers such as these need continuing competent technical support from qualified, practically- oriented, professionals if they are to be productive on a continuing basis. (iii) In the two agencies examined in this study, several possibilities were noted for reducing costs and/or 12This appears to be in direct conflict with the inmdication of the Commission on Higher Agricultural Educ- Iation (see footnote 7, Chapter I) that extension agencies :should not be spending most of their budget on salaries. 404 improving extension service at no added cost. These include elimination of the use of chauffers unless this serves to make one vehicle serve several extension workers to a much greater extent than it would otherwise, establishing a regular replacement program for vehicles, and reducing the amount of non—productive travelling (e.g., secretaries attending farm meetings). (Conversely, some cases were noted where attempts at cost-economy severely limited extension programs.) However, the potential for increasing extension effectiveness by focusing on costs is relatively small and it is doubtful whether either of the two agencies examined in this study were any more wasteful of public funds than government agencies in other parts of the world. Home Economics Extension programs directed at rural Colombian households seem to be capable of achieving a high pay-off in relation to costs even though diffusion is not substantial13 because many rural Colombian communities contain a sizeable group of households with a number of common interests, each of which offers some possibility for worthwhile change. Furthermore, these common interests can be serviced by relatively low cost personnel. The net result appears to be that, under these conditions, costs per household serviced can be much lower than the present value of future benefits per household. Home economics programs can then become 13There are two circumstances which appear to explain the apparent lack of diffusion of home economics practices examined in this study. Firstly, home economics programs Ivere directed at a substantial proportion of the total number «of families in the communities in question. Thus, only a few families were not informed of the practices directly. Secondly, the families who were not informed directly would likely be "late adopters” or "laggards" in any event and -thus, one would not expect diffusion to them to take place quickly . 405 quite competitive with programs directed at agricultural production units and, in some cases, offer a much greater possibility for useful extension work. Another significant factor is that such programs almost invariably have a positive effect on income distribution because most rural households belong to low, or low to medium, income families and the gains made from these programs do not usually get transferred to other groups by any direct mechanism (like shifts in the supply curves of agricultural products for example). It is also significant to note that although benefit-cost ratios may be high, there is little possibility of beneficiaries paying the cost because the benefits are ‘usually received "in-kind" and are not readily converted into money in the market place. Another point of interest is that home economics programs, of the sort examined in tfliis study, need not be restricted to farmers nor even to :rural residents. They appear to be applicable to a large group of Colombian families whose incomes are low (but not less than about 100 pesos per week). The following points appear to apply to an analysis cxf benefits per household served by home economics programs. (i) While the benefits of home economics programs may be difficult to evaluate, the application of the opportunity cost principle for such skills as the diagnosis and/or treatment of simple illnesses, and the making and/or repairing of clothes can produce rather spectacular results. These results should not be ignored - they reflect the fact that people have learned to do for themselves what would be (for them) prohibitively expensive otherwise. Thus, such measurements reflect net additions to these people's level of living. 406 (ii) Similarly, where it is possible to change the nutritional composition of the human diet and thereby improve the health of the individuals involved, the benefits are obviously significant. It is possible to place a value on this by having the beneficiaries relate this benefit to some other goods or services which can be traded in the market place. However, one of the difficulties with this approach is that the beneficiaries may not fully appreciate the biological effects of improved diets. (iii) The restricting effect of low incomes is felt in most home economics programs. The changes promoted often require corresponding changes in the family expend- iture pattern. This effect appears to be such that relatively little success can be achieved with families of income levels lower than 100 pesos per week. (Perhaps these are the late adopters and laggards who would adopt "when they can afford it" or when they see how things work out for their neighbors.) (iv) Although income level is a very restricting factor for Colombian home economics programs, extension workers do not appear to be focusing as strongly as they might on the problem of ”getting more out of existing incomes", particularly in the area of nutrition. A number of the extension participants interviewed mentioned that they knew (from home economics programs) that they should eat more meat and drink more milk in order to get a balanced diet, but they couldn't afford these foods. Perhaps more work needs to be done in the area of calculating diets (using available foods) which are acceptable to people and which come as close as possible to nutritional "requirements" on existing food expendit- ures. Few possibilities for improving diets on existing incomes were known to the extension workers encountered in this study. Furthermore, while some knowledge of nutritional ”requirements" was evident, extension personnel seemed to have little idea what would happen if such "requirements" were not met in full. (v) The benefits to be derived from vegetable gardens can be readily estimated by using market prices for the products and seeds and opportunity costs for the land and labor involved. In many cases, the results can be quite satisfactory although, for some 14This applies only to the extent that the products can substitute for purchased foods and/or usefully improve the diet of the family. Production in excess of economic levels of home consumption must be marketed and there may be no market for it. 407 families, poor land, lack of land, or distaste for the products produced make this sort of venture impractical. Nevertheless, many rural families appear to know little about growing vegetables for home use. Therefore, there appears to be consid- erable potential for usefully expanding vegetable gardens in Colombia. However, if this expansion is to occur, it would appear that extension workers themselves have to know more about the appropriate cultural practices to follow and many families have to acquire access to a small EIEt of good land. The families of many rural laborers do not currently appear to have access to enough good land for a garden. (vi) The benefits from programs directed at up-grading the abilities of rural families to make and repair clothes can be estimated by using market prices for raw materials and finished products and opportunity costs for labor.15 The results, although limited in some cases by the restricted availability of sewing machines, can be quite satisfactory. Most of the rural families interviewed in this study rated this type of program quite highly. While on the surface this sort of activity might appear to be severely limited by family income, it appeared as though the instruction offered often allowed families to make better use of their existing clothing expenditures and, for this reason, was very popular. Extension participants appeared to be willing to utilize such skills even when their diets were apparently poorly balanced,16 a fact which was somewhat disconcerting to a few extension personnel who believed nutrition was more important than clothing. (vii) The benefits from programs directed at up-grading the abilities of rural families to make home improve- ments are not readily estimated but the results, in many cases, appear to be unsatisfactory. If one uses market prices for finished products and raw materials, and charges labor at opportunity cost, then, for some of the programs examined in this study, the benefits appear to be negligible.17 15Again, this applies only to the extent of satis— fying the needs of the family - it should not be assumed that families can or will market the goods in question. 16The people involved appeared to be making informed decisions and simply preferred to be better-dressed than better-fed. 17This would not appear to apply to home improvements directed at improving sanitation. 408 Furthermore, where benefits are positive, there is often little opportunity of them being repeated on a continuing basis. (viii) The benefits from programs directed at up-grading the abilities or rural people to manufacture crafts (e.g., weaving baskets) are readily estimated by using market prices for the finished products and raw materials and charging labor at opportunity cost. The results, however, are quite variable and depend primarily on the market potential for the products. Usually this is quite low. (ix) The benefits from programs directed at supplying rural people with physical improvements to their homes on a shared cost basis defy easy measurement. However, if one were to ask the question "If the family were given the equivalent cost of the improve- ments in money, would it create these improvements?" the author hypothesizes the answer would often be "No". If this hypothesis is valid, the benefits from these programs could reasonably be regarded as less than costs. The fact that there is little learned from such programs which can be applied on a continuing basis also tends to limit benefit levels. (x) Where incomes are increasing, even slightly, new consumer goods which are not readily adOpted on existing incomes may become much more attractive to rural families. This would appear to apply to protein-rich foods such as milk, meat, etc. which, for the most part, are not economical substitutes for existing foods. The following points appear to apply to the analysis of costs per household of home economics programs: (i) The use of mejoradoras, who are semi-professionals, seems to be a significant factor in keeping costs per household low in home economics programs. Not only are their salaries lower than professional home economists but, as with the practicos, they are often willing to live in the rural communities and this tends to keep transportation costs down. (Of course, unless the me'oradoras have adequate continuing support from qualified, practically-oriented profess- ionals, one would expect them to soon exhaust their repertoire of innovations for the communities with which they work.) 409 (ii) There does not appear to be any particular cost advantage to having home economics and agricultural extension personnel working together, although in both Valle del Cauca and Antioquia it was noted that they shared the same vehicles. This obtains because home economics programs can often be success- ful with families of laborers and minifundistas where current agricultural extension programs have little to offer.18 (iii) In communities where it is difficult to get a group of twenty or so housewives with common interests who are willing and able to come to regular meetings, the costs per household serviced could very well be much higher than those noted in this study. However, it would appear to the author that there are a good number of communities in Colombia where home economics programs could be operated with groups of sufficient size to keep costs per household low. (iv) The costs per household serviced are likely to rise substantially if it is not possible for a mejoradora to visit several communities in a week. This would apply in remote areas where communities are either far apart or transportation methods are primitive. (v) One of the possible ways of reducing extension costs of home economics programs is the use of the mass media, eSpecially radio. As many rural families have radios, it would appear that some of the more straight-forward techniques or ideas might be promoted to a larger number of households at a lower cost per household than via existing methods. (vi) Those extension programs directed at supplying rural people with improved physical facilities (e.g., houses, water systems, etc.) were probably provided at a higher cost than the rural peOple could have acquired them for themselves. This obtains because of the administrative costs involved in having the extension agency perform such functions. 18Conversely, extension programs directed at agric- ‘ultural production units may often usefully concern them- sedves with firms where the operator's family has little to gain from home economics programs. (Indeed, the family may :not even reside on the farm.) Hence, offices to service such farms need not necessarily include home economics workers. 410 Community Organizations The evidence gathered in this study was not sufficient to warrant drawing any conclusions regarding the possibilities for effective extension work with community organizations in Colombia. While some apparently successful projects of this type were noted in Antioquia, it was not possible to measure the benefits of these projects and relate them to their costs. Furthermore, the actual "success" of these projects was obscured by the extensive financial aid rendered by the extension service itself. However, it is possible to distinguish the following points which are of relevance to an analysis of such extension work. (i) There are obviously many functions which community organizations could usefully perform in rural Colombia. Lack of roads, schools, medical and health facilities, electricity, telephone, marketing services, postal services and so on, are the rule rather than the exception. Provision of these could increase the level of living of many rural people and yet, for various reasons, little is being done to provide them. (ii) The most apparent barrier to effective extension work of this type is lack of funds for projects. People in rural communities are reluctant to organize because they know few funds are available locally. Once such groups are organized they usually encounter difficulty in obtaining funds from outside sources (e.g., the Coffee Federation and regional or national governments). (iii) The lack of adequate funds for public projects in rural areas is the result of several environmental factors. Firstly, it is a reflexion of the fact that resources in general are scarce in Colombia. Secondly, the extreme physical conditions in some areas make some projects very costly (e.g., cement to construct schools in remote areas must be transported long distances by horse or man). Thirdly, the existing institutional environment does not (iV) (v) (vi) (vii) (viii) 411 allow rural communities to have direct control over tax money. Extension workers can sometimes play a useful part in getting local communities organized and in helping them acquire funds from agencies in control of tax money. They can also provide tech- nical assistance once the actual projects begin. A few examples of both types of work were noted in this study. A limitation to effective extension work of this sort is the lack of appropriate training among extension people. Extension workers trained in technical agriculture or home economics are not necessarily well-suited to working with community organizations. Training in the areas of sociology, the psychology of group action, the evaluation of, public projects, the administration of public funds in Colombia, and the engineering aspects of roads, schools, etc. would be more apprOpriate. Because of the heavy emphasis on obtaining funds from the existing institutional complex, the economics of public projects has_been largely over- looked. Extension agencies are forced to concentrate on promoting projects for which funds may be available rather than trying to assess the merits of different projects on a cost-benefit basis. It can be noted that some public projects may generate enough added income in the local communities to pay for themselves if they are appropriately financed. This could apply, for example, to improved roads which would facilitate the marketing of local products. However, the lack of an institutional structure to capture such incomes and apply them to debt repayment, together with lack of financing discourages such projects. Public projects which do not generate added incomes but instead produce services which are consumed by local people cannot readily be funded locally because they would take funds away from existing incomes which are mainly devoted to food, clothing, and shelter. The use of external funds for such projects does not necessarily imply income transfers as the funds may have originated in the community (e.g., via the coffee export levy). However, it could mean that public goods are over (or under) consumed or that public productive projects are overlooked in favour of public consumptive projects. This could 412 apply to medical, health or educational facilities, the lack of which forces families to leave rural communities and deters professional extension workers from living there. (ix) Because of the severely limiting character of physical conditions in some areas of rural Colombia, the pay-off to public expenditures relative to costs may be quite low relative to other rural areas. This essentially means that extension work directed at community organizations may best be concentrated in a few promising areas, at least until extension budgets expand. The better solution for some communities may be to encourage the residents to abandon them entirely. (x) Extension personnel who become involved in work with community organizations may readily find themselves in conflict with the government which provides the funds for their existence. For example, extension workers may find that, to assist local organizations, they must first help them obtain funds or power over funds now controlled by regional or national govern- ments. Thus, it would not appear likely that Colombian extension agencies could readily develop policies to carry out programs designed to assist community organizations in more than a token fashion. Other Outputs Because of inadequacies in methodology, it was not 3possible in this study to perform any analysis regarding the «effectiveness of extension agencies in promoting managerial cflianges in rural society. However, there appeared to be a feeling among extension workers, whom the author met during tflne course of the study, that this was a significant part (Df their work. Several expressed the opinion that they were gyiving their clients something more than just the ability to grow vegetable gardens, to make clothes, or to produce ttnnatoes more effectively. None, however, suggested that 413 special, distinct programs were required to do this. This raises the hypothesis that managerial changes can be achieved just as effectively as a by-product of promoting practical knowledge as they can by separate programs for this purpose. If this hypothesis is correct, program benefits from technique-oriented programs will always be greater than those from management-oriented programs because they include the benefits from managerial changes. The possibility of agricultural extension assisting in the direction of rural people to occupations other than agriculture does not appear to have been given much attention by extension agencies in Colombia. There seems to have been an implicit, or sometimes explicit, intent to do exactly the opposite. This policy seems fallacious for several reasons. Firstly, there is much evidence to suggest that (Rolombia badly needs agricultural professionals with farm or at least rural backgrounds. While most rural youngsters .in Colombia have little opportunity to get even a high school .let alone a university education, perhaps it would be worth- vflnile to develop special programs to assist a few talented young people along these lines. Agricultural extension Imorkers could perform a useful function in informing rural Ineople about such a program and aiding in selecting the Exeople involved. Also, where extension workers find young people vmith particular talents which offer them the possibilities (3f better employment in other areas, there seems to be 414 little point in not pointing this out to the individuals concerned. As noted in Antioquia, young people who leave a rural community may be able to help both themselves and their families if they find gainful employment. The fact that Colombian cities have difficulty coping with the current influx of population from rural areas is certainly evidence of problems which must be faced and solved. How- ever, if individuals are, in fact, better off by moving to the cities, then the individuals should be informed so they can move. It should also be noted that the movement out of agriculture need not put pressures on the large cities if rural industry and outlying cities and towns are developed. 'Where rural people, particularly rural youth, are confronted by a decision to either stay in agriculture or leave, the possibility of local industrial employment may often be :mere attractive to them than that in a far-off city. If such industry develops in Colombia, agricultural extension vworkers can and should inform their clients about such I alternatives . The feedback of information on the success or :failure of new ideas and the needs of rural people to the developers of technology does not appear to be well- developed in Colombia. There appear to be possibilities ftxr improvement in this area. To test this hypothesis it nylght be useful for one agency to attempt to set up some 415 sort of procedure designed to achieve this very objective. If the costs and results of such a test procedure could be compared, and publicized, all agencies would have some basis for deciding whether or not it would be worth their while to do the same or try some modified version. Because of the existing structure of research and extension in Colombia, there appears to be little or no incentive for existing agencies to try to develop such a feedback mechanism. (This, of course, is not unique to Colombia.) Thus, such a test might be doomed to failure before it began. Extension Management The first major aspect of managerial efficiency mentioned in Chapter V was the location of the optimal jpr0portion of a fixed budget to be allocated to management. Estimated General Supervision and Administration costs for. CVC's Extension Service in 1967 totalled 699,077 pesos or 27.17 per cent of the total extension budget.19 These can, for the most part, be considered managerial costs for they were largely incurred in the making of decisions and acting to implement these. However, the costs of all field personnel time devoted to activities numbered 4, 5, 8, and 9 in Table 34 were also considered to be managerial costs. frhe portion of total field personnel costs accounted for by supervision and administration was deducted to avoid ___—— 19Of course, as the means of arriving at the estimate (of Administration costs was somewhat arbitrary, its accuracy is open to question. 416 double-counting and the result was that an additional 278,333 pesos was added to managerial costs. Total manag- erial costs were therefore estimated at 977,410 pesos or 40.0 per cent of CVC's total 1967 extension budget. As noted in Chapter VIII, between 20 and 40 per cent of the, total 1967 extension budget in Antioquia was allocated to managerial activities. As previous arguments did not develOp any criteria by which one could conclude this was too high or too low, no comment can be made as to how close this might be to optimal. However, even though it can only be regarded as a rough estimate of the costs of performing managerial functions, it illustrates the relative importance of management as a user of resources and emphasizes the need for managerial efficiency. In general, the management practices followed in both.Antioquia and the CVC, while they did serve to keep expenditures within budget limitations and detect financial jpractices of an unorthodox or unacceptable nature for a government agency, they did not serve to provide management (vith information on the effectiveness of programs, personnel, arui extension methods. Neither did they contribute much to auuquainting decision—makers with new agricultural technology cn: extension methods. The reasons for these comments will become clearer in the following discussion which is estructured in the form of the managerial process outlined 111 Chapter II. 417 Problem Definition It would be a mistake to say that the extension personnel employed in the two agencies examined in this study did not appreciate the problems of the rural people with which they worked - they obviously did.20 However, they appeared at times to have difficulty in translating their observations into clear statements of problems which they, as extension people, could resolve. Perhaps this was, in part, because so many of the difficulties facing their clientele could not be resolved by extension efforts alone. The para-professionals (i.e., practicos and Imajoradoras) seemed to have just as good, if not better, (appreciation for problems than the professionals in the two agencies studied. This appeared to be because of their «greater familiarity with local conditions. Perhaps they lmad.a.better appreciation for the values of the rural fanfilies with which they worked than the professionals who vuere usually from an urban middle-class background and did rust normally spend as much time on the farms. The two agencies examined in this study exhibited Invo quite different approaches to problem definition. The extension service in Antioquia spent considerable time and esffort constructing socio-economic studies directed at defining the problems of people in the communities in which tfliey worked: the CVC on the other hand, relied on a much more ——— 20Some individuals were, of course, much more adept it: this regard than others. 418 informal approach. The socio-economic studies did not, in the author's opinion, reveal much information which could not have been obtained by casual observation. It seemed obvious to even the casual observer, that housing, nutrition, clothing, health and education were poor for most of the families who were extension clients. However, familiarity with the characteristics of rural peOple is only one part of the information which extension personnel need to define problems which they are capable of resolving. They also need to be aware of possible alternative ways of doing things. To a large extent, the extension workers in both agencies examined relied upon their formal training for this purpose.21 This training was rather heavily oriented to technical agriculture and thus, extension workers tended to think of problems in terms (of low yields, poor animal health, excessive insect damage, etc. They did not appear to view low income as a problem which Inight, or might not, be resolved by techniques capable of ijucreasing yields and improving animal health. 21There did not appear to be much reliance upon tech- rquogy developed at Colombian research agencies. For example, tine CVC offices at Palmira did not function as a very strong lirflc with the I.C.A. experiment station as CVC employees did ruyt regard the station as a good source of useful ideas. How- ever, some interaction did exist and if I.C.A. had been con- dturting more applied research, CVC employees would have prrfloably been aware of the results. This lack of interaction anmnng'researchers and extension people perhaps precluded feexfloack of farmer's problems to researchers. It is difficult tc>2fault CVC for this as time spent consulting with research- ers would have probably reduced the time they spent in the fjrald.and perhaps would not have improved their existing programs very greatly as there appeared to be little new tech- nology being develOped of relevance to their clients. 419 Collecting Information for Decisions As this study has illustrated, the information required for assessing extension programs directed at resolv- ing problems faced by rural peOple includes: (i) information on the potential pay-off to individual farms or households. (ii) information on the potential extent and rate at which adOption can be influenced. (iii) information on the costs of operating the program. Information on the potential pay-off for individual farms or households should be available, in part, from the idea sources or the research peOple developing the technology. Of course, this may have to be adjusted for local conditions on the basis of experience and/or field trials. CVC's principal reliance was placed on the experiences of field personnel (either in the field or university) and CVC's ijn modest research program. Much of the technology being Iextended was adapted from standard North American texts. Ikn.exception to this was the fertilizer program which was leased on Colombian research. A similar situation existed :in.Antioquia where the Coffee Federation was relied upon lieavily as a source of ideas. While some cases were noted in vflnich good technical information was available and had not loeen obtained, extension personnel in both agencies appeared t1) be doing a reasonably good job of collecting available iJiformation for estimating pay-offs to individual farms or rnouseholds. The principal problem here was that this 420 information was in limited supply and often had not been tested under local conditions. Information on the potential extent and rate at which adOption can be influenced should come from previous extension experiences, knowledge of the general character- istics of the region, and communications and/or sociol- ogical studies. Unfortunately, all of these sources appeared to be lacking in both the agencies studied. There was little record of previous extension programs to rely on. Extension records indicated assistance given - £25 assistance actually used by the families concerned.22 Thus, while extension workers themselves may have been able to recall previous successes and failures, this was not available for :new members in the area. And, because of the high turnover <3f employees, most workers were relatively new. Knowledge (3f the regions in which they worked was limited by the short stay of extension workers in an area and the fact that Inany did not live in the region. As already noted, the ezxtension service in Antioquia attempted to make up for this but the socio-economic studies, but these were more descrip- txive than diagnostic in nature. Other sociological and genneral statistical information (e.g., numbers and types of farms) was not generally available. Information on the costs of Operating programs sruould be available from extension cost records. While no 7' 22As noted earlier, there were substantial divergences between these two measures. 421 two programs are exactly the same, historical data can provide coefficients which will be useful for making program cost projections. Unfortunately, in both the CVC and the extension services in Antioquia, cost data by program was not readily available. The CVC had a very good accounting system for obtaining such information but this was not being used very effectively. While cost records were maintained according to program, costs were Often not allocated to the programs to which they were attributable and wrong answers resulted. Furthermore, the program classification was not designed to distinguish programs with distinctly different benefit flows. Several alternative program classifications would have been useful - the one used by the author in this study is by no means the only or best one. The existing accounting system could have easily been adopted to a different classification and accurate allocation <3f costs. However, it would have been necessary for the extension service to define programs more precisely and require that all expenses (including field personnel time) be coded accordingly. In the author's OpiniOn this would not have involved any more work or any different reporting forms than already existed. The accounting system maintained byr'the Department of Agriculture in Antioquia was not at all conducive to the collection of good cost information on programs. Thus, there was little that could be done in this regard in that region. 422 Analysis of Information Collected The "analysis" portion of extension agency management requires the co-ordination of the three types of information mentioned earlier into some sort of decision criterion such as the benefit-cost ratio used in this study. This appeared to be particularly weak or entirely lacking in the two agencies examined in this study. Table 41 of Chapter VIII illustrates how the problem of comparing programs was approached in Antioquia. The results of this approach are not consistent with the results of the programs themselves (see Table 42). In general, weakness in the ”analysis" area seemed attributable to lack of training and experience. There was some evidence to suggest that the agronomos were weak in farm management, and this resulted in a poor ability to estimate the pay-offs to agricultural technology. An example of this problem was a study Of the fertilizer program which was carried out at the end of the first semester. In this study, estimates of income changes were made but, for several reasons, many of these did not accurately represent incomes resulting from the program. Where test plots were not available on the participating farm, regional average yields were used in place of ”control” yields. In most cases, these were not good estimates of what ‘would.have occurred on that farm in that semester without the new technology. It was noted in the study that some of the jyield.changes were partly due to bad weather and diseases. 423 But the question of the probable yield change due to the new technology under these conditions was overlooked. Furthermore, instead of calculating all added costs due to the new technology, only fertilizer costs were considered. As some farms were already using fertilizer and others also increased or decreased the use of other inputs (e.g., sprays or cultivation practices) as a result of recommend- ations made, fertilizer costs were not a good measure of the added costs incurred by adopting the new techniques. As a result of these combined errors, the estimates of income changes which were made in the study were not repres- entative of the income changes due to the program. In some cases they were too low because negative yield changes were recorded and/or added costs were overestimated. In other cases they were too high because some increased costs had ibeen overlooked and/or "control" yields were underestimated. Data from this project could have provided a more useful lease for future advisory work if test plots had been main- 'tained on all farms and the study had been directed at rmeasuring average and extreme results of the new techniques. 131 the author's Opinion, a farmer—oriented pamphlet summarizing the results as they happened would have been tremendously useful to corn and tomato growers in Valle. However, as it was, the study did not even provide a good basis for deciding whether or not the program was a success. Similarly, there was room for improvement in the 424 analysis of the effects of risk on adoption of technology. For example, several of the programs examined in this study required farmers or housewives to purchase inputs for techniques which had required no cost outlay whatsoever under traditional technology. While the expected pay-off was in many cases quite attractive, there was often some element of risk involved. In some cases, the possibility existed of losing all of the initial outlay. Extension personnel, while not unconscious of this fact, appeared to exert little effort to try to allow for it. They did not, for example, attempt to develop programs where, if such a catastrophe occurred in the first trial, the participants who suffered could be compensated for their losses. In some cases, special consideration was given on an ex pggtg basis to ‘unfortunate individuals. This included a second chance to jpurehase fertilizer on credit, or a second gift of seeds VflliCh had not produced well for the first time. However, there seemed to be more of an attempt to preserve the enrtension service's image on an ad hoc basis than a deliber- atxa, planned program to allow for risk in the adoption of new technology. The incorporation of potential rate and extent of diffusion into the analysis of extension programs appeared to be overlooked entirely. Most extension personnel did not seem to appreciate that viable knowledge could diffuse from early adOpters and innovators to their more reluctant 425 neighbors. There definitely was an appreciation of the fact that some farmers and their families were more approp- riate extension targets than others but the possibility of diffusion seemed to be forgotten or overlooked in assessing potential program benefits. The author attributed this to a lack of training in the social sciences coupled with a poor knowledge of composition of the agricultural industry in their regions. This latter difficulty could have been resolved, in part, by either better statistical information or resident extension agents who stayed long enough to become familiar with the area. However, the prospect of either of these becoming a reality is not bright. Deficiencies were noted in the analysis of income distribution considerations in comparing programs. Cases ‘were noted where extension personnel preferred to promote low pay-off technology to low-volume producers than to [aromote apparently high pay-off technology among high-volume giroducers - even though the gains in the latter case would jlikely have been passed on to consumers who also would fall :into the low income category. Other instances were noted Inhere unsuccessful attempts were made to promote techniques 'to poorer families who might well have learned of them just .as rapidly from their more aggressive neighbors. However, (extension personnel were very conscious of the importance of .income distribution as a consideration in comparing programs :and, in some instances, appeared to approach it quite objectively. 426 Decision-Making Because of the lack of comprehensive analysis, decisions among programs were often made on the basis of single-minded reasoning which completely overlooked import- ant factors. For example, in Valle, a decision was made to promote a nutrition program at the expense of a more comprehensive home economics program simply because the extension director felt malnutrition was the most serious problem in the area. Personal preferences and political considerations also appeared to influence decisions rather strongly. One might hypothesize that extension decision-makers were adOpting a mini—max strategy in that they tended to choose programs which would minimize the maximum possible threat to their own future if things did not go well. Such a strategy would result in relatively little experimentation *with new extension methods and/or radically different tech- :nology. With the exception of the buffalo program in :Buenaventura and the house-improvement program in Antioquia -— which were really not extension programs anyway - this was, in.fact, what was observed. Nevertheless, decisions were made and, as some of Ithe programs selected did appear to be quite successful, one cxould not conclude that changes in the decision-making (criteria would necessarily have resulted in a better set of programs. However, one would tend to feel that, in the Jxong run, a shift towards choosing among programs on the 427 basis of their expected benefit-cost ratios would produce better results. Execution of Programs As the author is not an extension specialist, he cannot pass judgement on how well programs were actually executed in the field. It seems safe to say, however, that this was somewhat variable because of the varying degrees of enthusiasm for essentially the same technology encountered among similar program participants. Better training in extension methods would have undoubtedly improved field work but few opportunities for training were available. Some extensionists seemed to be having difficulty communicating with their clients but, in the CVC, senior personnel appeared to be aware of this and were trying to resolve the problem. In Antioquia, senior personnel appeared to be more concerned with financial control than the qualities of field personnel. The competence of extension personnel in the technology 'which they were promoting should have been good because of their heavy background in the plant and animal sciences. And there were many cases in which it was evident that the agronomo or practico knew a good deal about insects, sprays, Icultural practices, etc. However, some of the comments made loy program participants led one to believe that some (extensionists did not possess this sort of knowledge. The responsibilities of the extension directors 428 tended to centre around financial administration and ensuring that their staff were busy. Tactical decisions regarding programs were left up to field personnel unless they infringed on these two areas. This sometimes meant that programs would suffer because expenditures which were not originally anticipated arose in the middle of their operation. Responsibility-Bearing One of the difficulties in connection with the bearing of responsibility for extension work is that extension directors and those persons in authority over them often have little or no way of measuring the results of extension programs. How is it possible to hold a person or a group of people reSponsible for the results of their efforts when one does not know what these results are? In the two agencies studied this difficulty was bypassed by keeping track of numbers of farm visits, numbers of telephone calls, numbers of meetings, etc. This told senior personnel that their employees were busy but it is doubtful if it made ‘them responsible for program results. Indeed, it may well lmave encouraged practices which were inconsistent with the 23 loearing of responsibility. It would appear to be possible, Ivith little or no extra effort, for extension workers to keep records of individuals actually adopting technology and of Ithe levels of benefits these individuals received therefrom. ——7 23The wide divergence between the numbers of people recorded as receiving assistance and those actually recalling and using the assistance in Antioquia is a case in point. 429 However, it would likely be necessary to devise some means of separating the assessment of programs from the assess- ment of individuals to ensure that unbiased records were actually kept. The high turnover of personnel tended to complicate the issue of responsibility bearing. Because a number of employees were in their jobs for relatively short periods of time, it was difficult to assess the performance of the programs in which they were involved or hold them reaponsible for the results. It is difficult to say what might be done to alleviate this situation. If the high turnover is due to poorly qualified people being hired and then either being released or finding it best to leave, then the answer would be to develop better procedures for selecting the appropriate candidates prior to employment. If the high turnover is due to poor working conditions, then the answer is to improve the conditions. Both of these factors may, in fact, be involved and changes in both areas may be necessary. The proliferation of extension agencies in Colombia further confuses the issue of who is responsible for what. If a single agency were responsible for extension work in a given region, governments could perhaps judge the perfor- znance of the agency on the aggregate trends in the region. lflowever, when so many public and private concerns are .involved in extension-type work, it is difficult to assess 430 where farmers are obtaining their information and for extension workers to plan programs which are not duplicating the efforts of other agencies. Nevertheless, there was noted, in both of the agencies studied, positive and successful efforts at trying to serve clients and areas not being covered by other agencies. One of the aspects of bearing responsibility is the responsiveness of decision-makers to the success or failure of programs. In the CVC and in Antioquia, it was noted that programs were usually changed or discontinued when it became evident that they were not successful. Field ‘workers were usually the primary force behind such changes as they were the first to notice whether their clients were using their assistance or not. However, there were a few programs which appeared to be relatively unsuccessful (e.g., the apiculture program in Valle) which had been operating for several years and had not been discontinued. Thus, some room for improvement appeared to exist in this area. CHAPTER XI SUMMARY AND CONCLUSIONS This dissertation was directed at suggesting ways and means of improving extension effectiveness in Colombia within existing budget allotments. Both the conceptual and empirical investigation showed that many of the determinants of extension effectiveness are beyond the control of extension agencies themselves. In Colombia, factors limit- ing extension effectiveness include restricted availability Of new, high pay—Off technology - particularly for mini: fundistas, poorly-developed transportation and marketing systems in some rural areas, and an institutional framework 'which provides very limited access to critical resources (such as land and capital) for many rural peOple. As long as these limitations exist, extension agencies can not do a great deal to improve the economic well-being of their clients and there may be little point in greatly expanding their budget allotments as some other studies have suggested. ‘Within existing budget allotments there are some possibil- ities for improving effectiveness which, if pursued, could allow existing agencies to perform a useful role until the factors mentioned above are altered. 431 432 The Conceptual Structure Used for Measuring the Effectiveness of Extension Programs The conceptual structure used in this study can be usefully employed to measurelthe effectiveness of extension programs directed at speeding up technical changes in farm firms and rural households. It does not perform particularly well in evaluating other types Of extension programs. This structure involved: (1) estimating the costs of the extension program directed at promoting the technology, (ii) estimating the pay-off of the technology to individual firms or households, (iii) estimating the "shifts" in the adoption curves of the technology, (iv) calculating the benefit flow from the pay-off and "shift" estimates, (v) calculating with an apprOpriate interest rate, benefit-cost ratios from the above-mentioned benefit flows and cost estimates, (vi) ranking competing programs on the basis of the benefit-cost ratio. To allow for income distribution effects, benefit flows for different income classes may have to be weighted differently. This thesis did not get beyond assigning such weights on an intuitive basis. Presumably such weights could be provided by extension directors. As the programs being evaluated were not large enough to influence supply curves for agricultural products and/or demand curves for goods purchased by the agricultural sector, no allowances lSee Chapters IV and V for a detailed description of the measurement techniques used. 433 for price changes were made in estimating pay-offs. To allow for uncertainty surrounding the above- mentioned estimates, sensitivity analyses were carried out. The cost, benefit flow and interest rate estimates were all examined in this fashion. However, in most cases only the benefit flow estimates needed to be varied. Fairly precise cost estimates were constructed and program rankings were typically rather insensitive to changes in the interest rate used. Thus, only variations in the benefit flow estimates had a major influence on program rankings. Extension Programs Directed at Agricultural Producers Agricultural extension programs examined in this study appeared to vary significantly in their effectiveness. Some, such as the Apiculture program of the CVC, appeared to be so ineffective that they should be curtailed entirely. Others, such as the CVC's Fertilizer program, appeared to be successful but could not be extended much beyond their existing level. It would appear that extension workers will have to be very selective about the programs they operate for Colombian farmers particularly in agencies such as the CVC and the Department of Agriculture in Antioquia. The principal agencies (e.g., the Coffee Federation and producer associations) appear to be already reaching many commercial farmers with much of the relevant technology now available for these producers. Much of this technology is either not relevant for smaller producers and minifundistas 434 or they will pick it up from their larger neighbors as soon as it is well-established. For example, tractor cultivation and harvesting techniques have no place on small holdings: modern livestock technology is not particularly relevant to farmers with one cow; improved corn varieties are not adapted to the climate of the mountains where most small- holders live; and small-holders in Valle del Cauca often learn from their larger neighbors about varieties and cultural practices which could be applied on their own farms. This latter process is facilitated by the minifundistas or their friends or relatives being employed on the larger farms. The high2 costs per day of extension work and the limited number of farmers which can be serviced in a given amount of an extension worker's time Often force the costs of providing technical agricultural assistance close to or beyond the relatively small gains that can be realized by minifundistas. In order to improve the effectiveness of existing agricultural extension programs, extension workers could be exposed to more farm management training so that they would be more capable of assessing the pay-Offs to individual farmers. A course in communications would perhaps serve to 'make them aware of the potential for the diffusion of tech- nology. As the diffusion process provides a natural :mechanism for amplifying the results of the extension worker's efforts, it is important for him to be aware of its existence and be able to assess diffusion possibilities for the technology he is promoting. However, for really significant 2 High relative to home economics programs. 435 improvements, extension agencies must await the develOpment of farmer-oriented research designed to produce technology which they can extend. In the meantime, they might try to do some on-the-farm research of their own with field trials and demonstrations of technology which appears to have potential for a given region but has not yet been tested. If they are to be successful in this work, they would require some training in scientific investigation and a better appreciation of the agricultural economy of the region in which they are working. Otherwise they may mistakenly test techniques with a fairly narrow potential. (The CVC's Buffalo program appeared to be an example of this.) If these sorts Of training can be acquired, it would probably be worthwhile freeing some of the agronomos and (perhaps practicos as well) from their present duties for this purpose. Home Economics Programs Home economics programs examined in this study appeared to be rather successful and could perhaps be expanded at the expense of some of the technical agricultural programs. The reasons for their success seemed to centre around the low costs per day of extension work, the relatively large number of peOple which could be serviced in a year using the "club" technique, and the nature of the programs themselves. The programs focused on the necessities of life-food, clothing, health, and shelter — and had something to offer most club members in each of these four areas. The programs appear to be applicable to a much larger group of rural families than 436 have been reached already and should also be relevant for some urban families as well. This does not necessarily call for a large increase in resources devoted to this program however, as the programs can service different communities on a rotating basis. If a club were Operated in each neighborhood once every four or five years, this might be adequate to bring many families up-to-date with new techniques applicable to them. Other families would then likely adopt what they could from their neighbors. Such programs should be directed at families earning incomes of approximately 80 to 400 pesos (1967) per week. Lower income families would likely have little interest or capability of applying the acquired knowledge; higher income families would probably already know much of the material presented or would not wish to use some of the techniques suggested. The Home economics programs could be improved if there were some subject matter specialists available to concentrate on develOping techniques to make more out of current income levels (e.g., new ways of preparing foods and reducing wastage, better ways of making and repairing clothes, up—dating field workers on matters in the health and hygiene area, etc.). In the author's Opinion this could be done on the basis of research which has already been carried out in various parts of the world. Channeling some of the extension budgets away from low pay-off agricultural extension programs to the areas of providing home economics subject matter 437 specialists and more on-the-job training for field workers in these programs would appear to be apprOpriate. Programs which are not educational in nature (e.g., the house building and repairing program in Antioquia) could also be curtailed to provide some resources for this purpose. Home economists could also extend their role slightly in some areas to handle gardening. (They were already doing this in some instances.) This would allow freeing practicos and agronomos for work with commercial agricultural producers3 as, in many cases, vegetable gardens were the only really successful aspect of their work with minifundistas. In order to achieve such changes, it would probably be necessary to teach home economics workers to drive and provide them with their own vehicles. Resources for this purpose would have to come from the above-mentioned adjustments. Community Organizations This investigation showed that, in some instances, agricultural extension workers can function as the catalyst to get community projects (e.g., roads, schools, community water services, etc.) rolling and completed. When this is possible and the projects are worthwhile, it would appear advisable for them to do 30. Indeed, in some cases, such projects would likely have a high pay-off because of the 3Some of the existing agronomos could function as subject matter specialists for HOme economics workers in the gardening area. This would involve providing them with training sessions and searching out ideas and techniques which they could use in the field. 438 scarcity of rural infrastructure in Colombia. However, care needs to be taken to ensure that worthwhile projects are chosen. (Schools, for example, are of little use if teachers are not available.) In general, it seems advisable to limit the involvement of extension peOple in community action work to communities where additions of roads, schools, etc. can convert the area into a viable economic one, capable of competing with other parts of the region or country. It seems inapprOpriate to direct these activities at communities where soils, climate, and other natural resources are so poor that the communities will eventually disappear because of out migration. Existing extension workers are at a disadvantage in trying to work with community organizations. They do not usually have any training for such endeavours and rely mainly on eXperience and intuition. It would seem advisable for agencies such as the CVC or the Department Of Agriculture of Antioquia to have a few members Of their staff specialize in this work. They could then acquire the necessary training in the evaluation of public projects, the technical aspects of getting the job done (e.g., engineering and construction), the sociology of dealing with groups of peOple, and the legal and institutional requirements of getting public projects accomplished in Colombia. These few specialists could then focus on "prime target areas" (as suggested above) on a rotating basis to get the job done. 439 Managerial Capacity of Rural PeOple Extension programs directed at changing the managerial capacity of rural people are much more difficult to evaluate than those directed at speeding up technical changes. However, from a conceptual point of view, they can be examined by considering the sorts of technical changes which the intended managerial changes might influence. As much of Colombian agriculture is oriented to crop production, it would appear that one set of managerial changes which might be extremely useful would be an improved general knowledge of crop-related phenomena such as soils, entymology, climate, plant characteristics, plant diseases, etc. A course such as this could be set up for Colombia as a whole and extension workers in the various parts of the country could use it when they found an opportunity to do so. However, in general, it may be very difficult for extension workers to bring about managerial changes in rural peOple unless they also have something of immediate practical use to offer. Thus, it may be best for extension agencies to leave this area of responsibility to the public school system. Perhaps courses such as the one suggested above could be offered by extension peOple in rural schools for both children and adults. BIBLIOGRAPHY BIBLIOGRAPHY Adams, Dale et. a1. Supervised Credit in Colombia's Agrarian Reform: An Evaluative Study, Bogota; Agarwala, A. N., and Singh, S. P. (eds. ) The Economics_ of Underdevelopment, London: OxfordU University Press, Anderson, C. Arnold, and Bowman, Mary Jean (eds.) Education and Economic Development, Chicago: Aldine Publishing Co., I965. Apocado, Santiago D. Observaciones‘y Recomendaciones Relativas a1 Programma de Credito Supervisado, STACA, Ministeria de Agricultura. Bogota} Mimeo, 1960. Arbelaez, M. Alberto. Prgjecto de Organizacion de un Servicio de Informacion de Cosechas, Bogota: Minist- eria de Agricultura. Barnett, H.G. IpnovationgtThe Basis of Cultural Change, New York: McGraw-Hill Book Co.,I§5§. Barringer, Herbert H., Blanksten, George I., and Mack, Ray- mond W. (eds.) Social Change in Develo in Areas, Cambridge, Massachusettes: Schenkman PGEIisHing Co., 1965. Boserup, Ester. The Conditions of Agricultural Growth: The Economics of Agrarian Change under Popuiition Pressure, London: George Allan & Unwin Ltd., 1965. Bryant, E. Kearl. "Communications in Economic DevelOpment," Research Paper #7 of the Land Tenure Center, Madison: The University OfWisconsin, 1965. Commission on Higher Agricultural Education. Higher Agric- ultural Education in Colombia, Bogota: Universidad Nacional as COIOmbia, . Deutschmann, Paul J. and Fals Borda, Orlando. Communication and Adaption Patterns in an Andean Village, Programa Interamericano deJ Informacion POpular & Facultad de Sociologia Universidad Nacional de Colombia, San Jose, Costa Rica, December, 1962. 440 441 . La Communicancion de las Ideas Entre los Compesinos Colombianos, Un Analysis Socio-Estadistico, Bogota} Universidad Nacional de Colombia{il . Di Franco, Joseph. "Differences Between Extension Education and Community DevelOpment," Comparative Extension Publication NO. 5, Ithaca, New York: Cornell Univer- sity, S . Di Franco, J., & Clifford, R.A. Estudio Analiticg de Cinco Or anizaciones de Extension en COlombia, Turriaiba: IICA, 1962. Dornor, Peter. "An Extension Philosophy for Farm and Home Development Work," Journal of Agricultural Economics, Vol. xxxvii (August, 1955). Druce, P.C. "Rural Extension and the Agricultural Economist," Australian Journal of Agricultural Economics, Vol. X (December, I966). Edel, Matthew D. "The Colombian Community Action Program: An Economic Evaluation." Unpublished Ph.D. dissert- ation, Yale University, 1967. Edmund de S. Brumner, Sanders, Irwin T., and Ensminger, Douglas (eds.) Farmers of the World: The Story of A ricultural Extension, New York: Colombia Univer- Slty Press, 1 5. Eicher, Carl, and Witt, Lawrence (eds.) Agriculture in Economic Development, New York: McGraw-Hill Book Co., I964. Fals Borda, Orlando. Campesinos de los_Andes: Estudio Sociologico de Saucio, BogotEE Iqueina, 1961. . Peasant Society in the_§olombian Andes: A Sociol- O icaI Stud of Saucio, Gainesville: University of— Florida Press, 1955. Federacion Nacional de Cafeteros. "Curso Sobre Materias Tecnicas y Extension Rural Cenicafe Chinchina Caldas,’ Mimqurafo, 1964 (Incora). Felstehausen, Herman. "Fitting Agricultural Extension to Development Needs: The Colombian Problem,” Research Paper #39 of the Land Tenure Centre, Madison: Univ- ersity Of—Wisconsin, 1967. Fonseca, Luiz. "Information Patterns and Practice Adoption Among Brazilian Farmers," Unpublished Ph.D. Thesis, Dept. of Communications, Univ. of Wisconsin, 1966. 442 Franco, Julio Cesar Aviler et a1. "Compendio de Legislacion AgroEecuaria y Organismos AgriEOlas de Colombia,“ A Report prepared by Servico Tecnico Agricola Colom— biano-Americano, Bogota, 1962. Griliches, Zvi. "Research Expenditures, Education, and the Aggregate Agricultural Production Function," American Economic Review, LIV (December, 1964). . "The Sources of Measured Productivity Growth: U.S. Agriculture, 1940 - 1960," Journal of Political Economy, LXXI (August, 1963). Hagen, Everett E. On The Theory of Social Change: How Economic Growth Begins, Homewood, Illinois: The Dorsey Press, Inc., 1962. Hanay, Emil B. "The Possibilities for an Economic Reorgan- ization of Minifundia in the Colombian Highlands," Unpublished Ph.D. dissertation, University of Wisconsin, in process. Havens, A. Eugene. "Education in Rural Colombia: An Invest— ment in Human Resources," Research Paper #8 of the Land Tenure Center, Madison: University of Wisconsin, 1955. . "Social Factors in Economic DevelOpment," Research Paper #5 of the Land Tenure Center, Madison: The University of Wisconsin, 1965. . "Social-Psychological Factors Associated with the Differential AdOption of New Technology by Milk Producers," Unpublished Ph.D. dissertation, Ohio State University, 1962. Hirschman, A.O. The Strategy of Economig DevelOpment, New Haven: Yale University Press, 1959. Holt, Pat M. Colombia Today and Tomorrow, New York: F.A. Praeger, Publishers, 1964. Hoselitz, Bert F. Sociological Aspects of Economic Growth, New York: The Free Press of Glencoe, 1963. Jamais, Juan F. "The Effects Of Belief System Styles on the Communication and AdOption of Farm Practices," Unpublished Ph.D. thesis, Dept. of Communications, Michigan State University, 1964. Kimball, W. James. "The Relationship Between Personal Values and the AdOption of Recommended Farm and Home Practices," Unpublished Ph.D. thesis, Dept. of Education, University of Chicago, 1960. 443 Machlup, Fritz. Production and Distribution of Knowledge in the United States. Princeton: Princeton Univer- sity Press. Maddox, James G. Technical Assistance byReligiousiAgencies in Latin America. Chicago: University Of Chicago Press. Maunder, A.H. "Improvement of Agricultural Extension Services in EurOpean Countries," United Nations (FAO) 1954. McKean, Roland N. Efficiency in Government throu h S stems Analysis. New York: John Wiley & Sons, Inc., 1958. Mellor, J.W. The Economics of Agricultural DevelOpment, Ithaca, New York: Cornell University Press, 1966. Ministro del Agricultura, Colombia. "Memoria del Ministro de Agricultura a1 Congreso Nacional?“ various years. Mosher, Arthur T. "Case Study of the Agricultural Program of ACAR in Brazil," Washington: National Planning Association. . Getting Agriculture Moving: Essentials for Devel- Opment and Modernization, New York: Frederick A. Praeger, Publishers, 1966. . Technical Co-OEeration in Latin American Agric- ulture, Chicago: The University of Chicago Press, 1957. Myren, Delbert T. "Bibliography: Communications in Agricultural DevelOpment," Mexico: The Rockefeller Foundation, 1965. Nicholls, Andrew J. "Development of the Peruvian Extension Service," Washington, D.C., Office of Foreign Agricultural Relations, 1952. Parra J, Humberto. "Relaciones Entre Investigocion y Extension en El Departamento del Valle del Cauca," Universidad Nacional, Palmira: Mecanografiado, 1961. Portacarrero, Cesas Alfonso. "Empathy and Modernization in Colombia," Unpublished Master's thesis, Department of Communications, Michigan State University, 1966. Posada, G., Alfonso and others. "El Exito Rural en Colombia," Universidad de los Andes, Bogota: Mecanografiado, 1960. 444 Prest, A.R. and Turvey, R. "Cost-Benefit Analysis: A Survey," The Economic Journal, Vol. LXXV (December, 1965. Raudabaugh, J. Neil. "Evaluation in Extension Education," Evaluation in Extension, prepared by Division of Extension Research and Training, Federal Extension Service, U.S.D.A. Robinson, E.A.C., & Vaizey, J.E. (eds.) The Economics of Education, Proceedings of a Conference held by the International Economic Association, London: Mac- Millan and Company Ltd., 1966. Rogers, Everett M. Diffusion of Innovations, New York: The Free Press of Glencoe, i962. . "Mass Media Exposure and Modernization among Colombian Peasants," The Public Opinion Quarterly, Vol. 29, Winter, 1965-66. . Social Change in Rural Society, New York: Apple- ton-Century-Crafts, 1960. Rogers, Everett M. and Herjog, William. "Functional Literacy among Colombian Peasants," Reprinted from Economic Development and Cultural Change, Vol. XIV, No. 2, January, 1966. ‘ Rogers, Everett M. and Neill, Ralph E. Achievement Motiv- ation among Colombian Peasants, East Lansing, Michigan: Michigan State University, 1966. Rojas, Angel and Gonzalo, Luis. "Elaboracion de Programmas de Extension en Relacion con el Estudio Socio- economico,f Secretaria de Agricultura, Medellin: Mecanografiado, 1964. Ross, James E. "Implications of Co-Operative Rural Electrification for Economic and Social DevelOpment in the Department of Valle, Colombia, South America," Unpublished Ph.D. dissertation, University of Illinois, 1966. Ruttan, Vernon W. "The Contribution of Technological Progress to Farm Output, 1950-75," Review of Economics and Statistics, XXXVIII (February, 1956). Sanders, H.G. (ed.) The Co-operative Extension Service, Englewood Cliffs, N.J.: Prentice-Hall Inc., 1966 (reviewed August 1966, J.F.E.) 445 Schmid, A. Allan. "Non-market Values and Efficiency of Public Investments in Water Resources," American Economic Review, Vol. LVII (May, 1967). . "Sources of Economic Growth: The Importance of Institutional Change," Journal of Farm Economics, Vol. XLV (November, 19637} Schultz, Theodore W. Economic Crises in World Agriculture, Ann Arbor: The University of Michigan Press, 1965. . The Economic Value of an Education, New York: Colombia University Press, 1963. . Transforming Traditional Agriculture, New Haven and London: Yale University Press, 1964. Seminar on Comparative Extension Education. "Establishing a Young Extension Service," Ithaca, New York: Cornell University Press, 1959. Sen, A.K. "The Choice of Agricultural Techniques in Under— develOped Countries," Economic DevelOpment and Cultural Change (April, 1959). Sill, Maurice Lucien. "Personal, Situational, and Commun- icational Factors Associated with the Farm Practice Adoption Process,"’ Unpublished Ph.D. disseration, Pennsylvania State University, Department of Sociology, 1958. Smith, T. Lynn. "The Cultural Setting of Agricultural Extension Work in Colombia," Rural Sociology, Vol. X, NO. 3, September, 1945. Southworth, Herman M. & Johnston, Bruce F. Agricultural DevelOpment and Economic Growth, Ithaca, New York: Cornell University Press, 1967. Stigler, George. "The Economics of Information," Journal of Political Economy, VOl. LXIX (June, 1961). Turk, Kenneth L. and Crowder, Loy V. (eds.) Rural Develop- ment in Tropical Latin America, Ithaca, New York: Cornell University Press, 1967. U.S. Foreign Agricultural Service. "Colombian Agriculture," Bogota: Office of the agricultural attache, 1960. A Report prepared by the Office of agricultural attache American Embassy, Bogota: 1960. Van Es, Johannes C. "Opinion Leadership in Colombian Veredas with Different Norms on Social Change," Columbus, Ohio: The Ohio State University Press, 1914. APPEND ICES APPENDIX A THE RATIONALE FOR SELECTION OF THE DECISION RULE USED IN THIS STUDY There are four rules which are often used for selecting among projects which produce a flow of benefits over time as a result of capital expenditures made in one or more periods.1 These thereby satisfy the third criterion mentioned in Chapter V. These rules do not apply to projects which are interdependent or mutually-exclusive.2 However, they are capable of dealing with a number of discrete (independent) alternative programs. Where only one time period is involved, they can be compared with the marginal productivity rules supplied by the theory of the firm and, under such circumstances, they are equivalent to these. They obviate the necessity of estimating social costs as they use only data on extension program expenditures under the assumption that these are proportional in the same way to social costs for all programs. Finally, they require no 1A. R. Prest, and R. Turvey, "Cost-Benefit Analysis: A Survey", The Economic Journal, December, 1965, p. 703. 2A project is interdependent with another project if costs and/or benefit flows from either one of the projects vary with the nature or the level of the other. Two projects are mutually-exclusive if the occurrence of one precludes the occurrence of the other. Where interdependent projects do arise it is possible to evaluate them with the rules discussed here only by combining them and considering them as one project. 446 447 data on physical input/output relationships providing expenditures and benefits in various periods can be estimated. They apply only where a starting date is given for each project or program although the terminating date of the programs may be infinite. These rules are:3 (a) The Present-Value Rule Adopt any project for which the associated stream of net benefits or net receipts, discounted at the 'appropriate' rate of interest is greater than zero.4 3(1) B(2) B(n) (1) NPV = -C(O) + + + —————— 1+1 (1+1)2 (1+i)“ (b) The Internal-Rate-of-Return Rule Adopt any project for which the internal rate of return is greater than the 'apprOpriate' rate of interest.5 B(l) B(2) B(n) (2) O = -C(O) + -————-+ —————7-+ . . . + ————-—- (1+r) (1+r) (1+r)n (c) The Annual-Net-Benefits Rule Adopt any project for which the annual net benefit is greater than zero when computed at the 'appropriate' rate of interest.5 (3)—S——+...+__S___fi_=_c(o)+§(__l)_+...+B(n)n 1+1 (1+1) 1+1 (1+1) 3The formulae used in all of these rules are Often only approximations to the "true" relationship because they assume that B(t) has no value until the end of period t and that NPV and C(0) occur at the beginning-3f period one. However, these approximations can be made arbitrarily small by considering shorter periods when the period being used is suspected of producing significant errors. 4J. Hirshleifer, James C. Haven and Jerome W. Milliman, Water Supply, Technology, andtPoliqy, (Chicago: The University O C icago Press, 1963), p. 152. In this situation the 448 (d) The Benefit-Cost-Ratio Rule Select all projects where the ratio of the present value of total benefits to the present value of total costs exceeds unity.7 211) + , , , + _2121_. 1+1 (1+1)n _ va (4) 1 _(_S—' ‘— C(O) + Sill + . . . + -§—2——_ PV 1+1 (1+1)n c where NPV = net present value at t=0 B(t) = net benefits in period t (may be negative) i = the "appropriate" rate of interest C(O) = the initial capital outlay at t=0 c(t) = Operating costs in period t8 b(t) = gross benefits in period t S = annual (periodic) net benefit r = the internal rate of return Simplifications for the above-rules can be constructed under certain circumstances. For example, if B(t) is constant B(t) 1 - C(O) . for all t=l,..., n and if n is infinite, NPV = Also, NPV = PVc - PVB under all circumstances. Finally, [ (1+1)n - 1] [ 1(1+1)n 1 S=NPV and the denominator is Often readily available from standard interest tables. "appropriate" rate of interest would be determined by the effectiveness of other government programs. 502. cit., p. 154. 6J. Hirshleifer, et. al. pp, cit., p. 155. This might also be stated as: "Select all projects where the constant annuity with the same present value as that of benefits exceeds the constant annuity (of the same duration) with the same present value as that of costs.” See Prest and Turvey, op. cit., p. 703. 7 Prest and Turvey, Op. cit., p. 703. 8c(t) does not include interest, depreciation, or any form of debt amortization. 449 All of the four rules assume that the "appropriate" rate of interest (or the internal rate of return) is the same for all periods. Should this not be the case, all except the internal-rate-of-return rule could be modified using different rates in different periods, i(l),..., i(n), although calcu- lations would be more complicated. Also, all of the above four rules are variations Of the same basic concept and will often discriminate between acceptable and unacceptable projects in exactly the same way. Rules (a) and (c) are equivalent in this regard. As B(t) = b(t) - c(t), rule (d) is also equivalent to rule (a). However, the exact value of the PV’b/PVc ratio will vary depending on how the b(t) and c(t) are defined even though the concept of B(t) remains unchanged.9 (This, of course, requires that the "error" in bj would have to equal the "error" in cj.) These definitions sometimes present serious problems and it is unwise to use the PV’b/PVc ratio for ranking several projects for this reason. However, no error will occur if the selection 9For example, suppose a project has the following characteristics: C(0) 1000 n 20 i = .05 b(t) 105 c(t) 20 where b(t) is defined as "sales" less "marketing costs" and c(t) is defined as "other annual costs." If b(t) were defined as "sales" and c(t) was defined as "marketing costs” plus "other annual costs”, the project might appear as: C(O) = 1000 n = 20 i = .05 b(t) = 130 c(t) = 45 In the first case, PVb/PV = 1.0595 and in the second, Egbégvc = 1.0380. In both cases, B(t) is the same and NPV = 450 rule is followed exactly. (That is to say, all projects with PV’b/PVC > 1 will still have PV’b/PVc > 1 no matter how their b(t) and c(t) are defined providing B(t) is always defined in the same way.) Rule (b) is equivalent to the rule (a) if a unique positive solution exists for r. That is, it will lead to the same conclusions regarding the selection or rejection of projects. However, it is not always the case that a unique positive solution10 can be found for r and, under such circumstances, it is often ambiguous and cannot be used. Even if a unique positive solution is found, the interpretation of the root requires care as, if O O. The roots of equation (2) can be positive, negative, or complex. There are always n of them but often several are identical resulting in less than n values of (1+i) for practical purposes. Imaginary roots can be ignored because they have no useful economic meaning. (In the case where all roots are imaginary there is no real interest rate for which NPV equals zero and the project could be readily detected as being worth- while or not without performing any calculation. A series of benefit periods with no costs would be an example). A negative root would imply a rate of return of less than -100% (i.e. less than no benefits) and is also of little practical interest. Multiple positive roots merely imply that there are multiple values of the interest rate which would make NPV = 0. Thus, there appear to be multiple "rates-of-return" and the rule is ambiguous. 451 unique positive root does exist. Where (1+i) is positive, NPV = f(1+i) as specified by equation (1) may be either positive, zero, or negative, depending on the parameters C(O), B(t), and the value of (1+i). Now NPV 8? O as: (5) cm) >.( B(_1)_+ B(2) + . . . + B(n)n 1+1 (1+1) (1+1) Thus, where C(O) is positive, the expression on the right hand side of (5) determines whether NPV can be zero more than once in any given problem. If the right hand side is positive for all (1+i) > O, NPV can be zero only once (i.e. the polynomial can have only one positive real root). This means that when examining a simple initial investment which produces a strictly positive benefit flow, the problem of several positive roots cannot arise. It is also Obvious that as long as several periods of only net costs are followed by periods of net benefits, the problem cannot arise. In other words, once an initial investment has been made (either in one or more periods), if the resultant net income stream is always positive, a unique answer exists.11 As phenomena of interest to economic analysts can Often be expected to exhibit this property, there seems little reason to avoid the "internal- rate-of-return" rule in such cases. Indeed, it should be 11For example, if a project is described as: C(O) = 10 B(l) = 50 B(2) = -60 then 0 = ~10 + §2—-- 60 and r = l, r = 2, are solutions. 1+: (1+r)2 This implies that the net present value, NPV, is only positive for rates of return between 100 per cent and 200 per cent or, 452 noted that where more than one positive real root exists for the polynomial in (1+i), the other rules are also affected12 and cannot be considered as means of avoiding the ambiguity problem presented by the "internal-rate-of-return" rule unless the interest rate is known rather precisely. (The level of precision required can be determined by examining the relationship between NPV and (1+i) for each project.) However, even if the interest rate is known rather precisely the implicit assumption13 of no constraints being Operative is inappropriate for this study. The difference between the constrained and unconstrained cases may be appreciated by considering Figure A-l. In the unconstrained case, it is merely necessary to decide which projects fall above or below some cut-off line determined by the return on funds in other uses. Thus, in the example shown, projects G, B, I, D, A, E, and F would be accepted and sufficient funds would be made available to operate these. In the constrained in other words, present value would be zero if the rate Of return was 100 per cent or if it was 200 per cent. However, this arises because the income stream became negative after being positive. Had the example been changed to: C(O) = 50 B(l) = 10 B(2) = 60 10 . then 0 = -50 + ———-+ __§2_2. and r = .2 is the only relevant l+r (1+r) answer. Note that the roots are 1.2 and -l in the second case. 12Where more than one positive real root exists for the polynomial in (1+1) it would be possible to find a project acceptable (NPV > O) at an interest rate of say 6 per cent, and unacceptable at a slightly lower rate. 13As the four rules cited earlier all suggest operating all projects which meet the given criterion, it is implicit that the total budget can be varied. 453 case, it is necessary to rank projects so that they may be selected in decreasing order of performance until the total 14 In the example shown in Figure A-l, budget is exhausted. the appropriate ranking would be G, B, I, D, A, E, F, C, J, K, H. If the budget was fixed at the level shown, only G, B, I, D, A, and E would be operated. Fixed Budget Selected Measure of Project Performance Cut-Off Line (3 B II D A. 13 F‘ C J KLIH ‘ Extension Expenditures Figure A—l:- Illustration of difference in project selection procedures between the constrained and uncon— strained case. The four rules previously discussed are sometimes regarded as useful for the purpose of ranking projects. In previous discussion it has been established that the use of the PVb/PVc ratio can readily lead to inconsistent rankings 14Of course, this carries the implicit assumption that there are more projects under consideration than could be carried out by the available funds. If this is not so, one merely has to Operate all projects being considered and no ranking procedure is necessary. 454 depending on how benefits and costs are distinguished. As this can be avoided by using other criteria, the PVb/PVc ratio can be discarded as a possible criterion for ranking projects. The "annual-net-benefits" rule can also be discarded because its only advantage lies in the form in which results are presented. Thus, only the "present-value" rule and the "internal-rate-of-return" rule need to be considered. The "present-value" rule cannot be applied as previously stated but a suitably modified version might be: "Rank all projects according to NPV as calculated by (l) and select from the top of the list until the total budget is exhausted". However, as it is possible for projects with low initial capital cost C(O) to have the same NPV as those with high initial capital costs even though annual net benefits are all positive, this would appear to be unacceptable. This problem can be circumvented by ranking projects according to PVB/C(O) [where PVB = NPV + C(O)]15 which is essentially a measure of returns per unit of . capital. If projects are selected according to this method of ranking, no higher returns to any unit of capital will be possible and total returns to the fixed budget will be maximized. For cases in which annual net benefits are not all positive, the rule can be extended to PVB/PVC where PVB is the present value of benefits from all 15If one allows for the possibility of net benefits being negative (i.e. net costs) in more than the initial period, this generalizes to give the formula cited in Chapter V. 455 net benefit periods and PVC is the present value of costs 16 It will be noted that this from all net cost periods. criterion appears to break down if PVC = O as the ratio then goes to infinity. However, this is only an appropriate reflexion of reality as such a project should be chosen before others with finite PVB/PVC ratios. The PVB/PVC ratio does not discriminate between two projects with PVC = 0, although there may be an obvious choice on some other basis. In practice it may be advisable to set some lower limit on the performance of marginal projects to ensure that funds are not spent if their return is low in relation to supposed returns in other uses. In such a case the same ranking procedure would still apply except that in some periods the total budget would not be exhausted. If this was not the case, projects with PVB/PVC ratios less than unity or even less than zero could be operated. (In order to incorporate the welfare loss of taxation in the lower limit, only projects with PVB/PVC ratios somewhat greater than unity should be operated.) 16It will be noted that this is not exactly the same as the "benefit-cost" ratio defined earlier. It circumvents the problem of obtaining different answers depending on how benefits and costs are defined. It implicitly assumes, however, that the total size of outlay in any period is unimportant and therefore only net benefits or outlay is relevant. (Net outlay at time zero is, of course, given by C(O).) If this assumption is not correct, another specification of the problem is required for all criteria discussed here. 456 To calculate the PVB/PVC ratio, the "appropriate" interest rate must be found and this is usually a difficult task. As noted previously, NPV (and therefore PVB) can, under certain circumstances, increase with an increase in the interest rate; thus, project rankings are by no means invariant to the interest rate selected. In fact, even when PVB is a monotonic function Of interest rate, project rankings can vary with the interest rate selected although the possibilities for changes in ranking are less in this case. The problem illustrated in Figure A-Z. Thus, as in the unconstrained case, the "appropriate" rate of interest must be known reasonably precisely. , Project A PVB PVC Project B _-—---- — I I I I I I I I -—-1_——-_ Interest Rate Figure A-2:- Comparison of PVB/PVC ratios as functions of interest rate for two projects. The internal rate of return offers a possible alter- native to this dilemma, when a unique positive solution for (1+1) exists. Here the criterion might be stated as follows: Rank all projects according to r as calculated in (2) and 457 select from the top of the list until the total budget is exhausted, or some minimum r is encountered. This ranking will be unique and, as it does not utilize an interest rate, it does not depend on knowledge of the "appropriate" rate of interest. Unfortunately, it can lead to conclusions which are inconsistent with the modified "present—value" rule. The problem can be understood clearly by examining Figure A-2 again. The points ra and rb represent the calculated internal rates of return for Projects A and B respectively. The result- ant ranking would be inconsistent with the modified "present- value" rule if the appropriate interest rate was, in fact, i (PVB) A > (PVB) (PVC) (PVC) 1 as then B. In fact, it would be inconsistent with the "present-value" rule for any "true" interest rate below iC as, at this rate project rankings reverse. The problem arises because the "internal-rate-of-return" rule compares projects by setting NPV = 0 when, in reality, one is interested in operating projects with NPV > O and this is how they ought to be compared. Another way of looking at it is to recall that the basic problem is one of comparing different benefit streams over time in relation to their respective costs. This usually cannot be done unless one has a method of appropriately weighting the benefits produced at different times. The appropriate weights are provided by the interest rate. 458 It thus seems relevant to postulate the question, "Why are the weights that apply to benefits occurring in different periods determined once an interest rate is known?" The answer to this appears to be that it is implicitly assumed that, if the receivers of the benefits in any period would prefer to have them in another, they can simply exchange them in the market by saving or dissaving. The market prices at which they can be exchanged are defined by the rate of interest. For example, $x today can be exchanged for $x-(l+i)t t periods from now. But is this assumption strictly correct for public decisions? If benefits from public projects accrue to individuals in society and these individuals can, in fact, make such an exchange if 17 it would appear that the assumption they desire to do so, is correct. It is sometimes argued that this reasoning breaks down when one generation must invest so that another may reap the benefits. However, this is not a problem of determining the appropriate exchange rate over time - it is one of making interpersonal utility comparisons. Thus, there does appear to be a good reason for using an interest rate to reduce benefits occurring in different periods to a common denominator, namely present value. 17If they cannot make such an exchange at the selected rate then, it would seem likely that they could make an exchange at some rate and the problem revolves around finding the "appropriate" rate. This problem can become quite complicated if different beneficiaries face different rates and/or the rates for saving differ from those for dissaving. APPENDIX B SUPPLEMENTARY DATA ON THE 1967 OPERATIONS OF C.V.C.'S EXTENSION SERVICE 459 460 oomum>m meu qums mo poumEHumO .muouuosq mCpromum moss» 0CD mo oumz mmmH mo umuumsq umuHu was new mumoo CoHumuumHCHEOC o .uCosuummoo mCHquouom .U.>.U on» he oOHHmmsmm Nam.mmm.N mNm.mHe enm.va mON.mmN hmm.mv¢ Nom.on mmH.omm NNm.mrm Hdsbh cmm.nmm III III III III III III oomm.hnm COHumuumHCHEUC Ham.ams mHN.mVH Nno.nm www.mv NmH.mOH mmm.ma HHH.mOH mmn.va Hmuoa osm hwm.om hmo.oN mmo vow mmv.N ons.N HNn.v vmn.v msomCuHHoumHz mmm.m smm.s III III NHH oss OON NNN.H ucmsaHsvm can mmonuo mo mowxmo pcm OOCOCOuCHoz omm.cH mmm.m III III III III OON.v NHN.m Huqum OOHuwo vwo.omm HNm.Nw me.mn mmH.Nv omm.NOH vwm.om nom.wm mhm.vNH momcwmxm mCHHHo>muH NNv.om mom.mm omv.s pov.m NmN.m vNN.N th.m OHm.m moHHmmsm oCs mHmHuwumz momcmmxm Hmuwcwo HON.bmm.H NHh.OhN mom.orH mmh.OHN th.omm va.om meo.th mmm.mmH Hmuoa nsm mbH.vN mNH.NN III III III III omo.N III moms HMCOHmm0uoum Noo.wmm HmN.om mmM.Ho wmm.MN omm.nHH vvm.Hm nmn.mm HNo.mo muHmmCOm moCHum HNo.hmm mmN.NoH NHm.mHH mvm.mmH mmN.mHN onm.mm HaN.neH NHm.nNH mmHuonm ecu moan: mumoo HOCCOmumm nauseous OOCdumHmm< Emumoum COHmH> Hmuos umnoo osHo mIv mOHEOCoom wCHCCde Esme EmuqOum Euumoum Iuomsw msouH umoo oHuHomm .mxom wear a umNHHHuuom musuHsoHQC xooumw>HH Hmuwao mo COHumHuomoo m.momH .Hm sous: 0» nomH .H HHHQC pOHuwm on» HOH .U.>.u we» mo OOH>HOm COHmCOuxm we» ocHumummo mo mumou on» mo C3ooxdemII.HIm anoe 461 Table B—2.--Breakdown of Fringe Benefits (Prestaciones Sociales) for C.V.C. in 1967. m Item % of Basic Salary Bonuses (Primas) 8 Vacations (Vacaciones) 7 Termination Bonus (Cesantias) 14 Family Allowance and SENA (Subsidio Familiar y SENA) 6 Medical (Servicio MediCo y Drogas) 7 Other Payments (Prestaciones Extralegales) lg 54 Table B-3.--Costs of Operating C.V.C. Vehicles in 1967.a Item Colombian Pesos % Gasoline and Oil 368,140 27.1 Tires and Batteries 130,416 9.6 Other Materials b 28,277 2.1 Transportation Services 5,568 0.4 Communication Servicesc 1,024 0.1 Garage Rentals 26,481 1.9 Contracted Repairs 363,566 26.7 Insurance 40,380 3.0 Replacement Parts 296,054 21.8 Washing and Greasing 30,500 2.2 Miscellaneous Expenses 5,839 0.4 Depreciationd 64,068 4.7 Total Costs e (excluding chauffers) 1,360,313 100.0 Average Costs per kilometerf 0.6088 aProvided by Accounting Division, C.V.C.; includes all C.V.C. vehicles. 1DIncludes towing charges, taxi and bus services in case of breakdown. C I O I Includes communication serVices related to breakdowns. dCalculated on the straight line basis using a rate Of 10 per cent per annum. eCosts of chauffers are included in direct project costs. fBased on 2,234,371 total kilometers. 462 .mhmo mHH new Houmuwmo uouumnoum mH>OE m moosHoCH m .mwmp HHH HOH noumummo souvenoum OH>OE m mopsHoCHo .mhmo me new umHEocoom use: wCo moosHoCHq .mhmo hm new umHEocoum HausuHCUHumm mCo mOOSHuCHn mnm.o u NM vXom.va + meh.va + NxNo.HNN + meN.mnv u u "COHumsww Conmmumou mCHBOHH0u may scum oODMHCOHMUM < men www.mnm.N mnm.msm.N cON.H mmh.H wmm.H vmm.n mEdumoum HHC OHm.wHI omm.Nmm ONm.MHm III mVN III mHo mamumoum uwtuo www.mI me.NOH www.mm III III III wMN muHsum . I . . . III III msmumoum th m Ham wmv NHm mmv oON H VHm ammou onHomm nvo.0HI HNH.omv vso.omv III ommo.H on man mnsHo uIv .maom . I . . III . oucmumHmmC men n mmN mmm OHm mmm oNHN ONom H NwH mOHEOCOom use: . mCHCCMHm Show hem.MHI ham HmH ovH.mMH III mH III Hem pCm NONHHHuHms oNo.mv mOH.omH mNH.m0N III OMN III mNN wusuHsona $9... N352 02.me III III III nmmm Seamus/H.H OIu mumoo mumoo moucwumHm< Home: moHumcmumuo> < momuowmxm Edumoum mouwubo no on Co mOOHuomum mmuoomu0woz mOEOCoumd Emumoum o o vx mx Nx Hx < mEHB HOCCOmumm mo whom we uwbEsz .mmmH .Hm nous: o» NGNH .H HHuas poHuma on» mCHusp .o.>.o mnu no O0H>uwm COHmcmuxm wnu HOu Emumoum he QEHB HOCCOmem onHm no name no Hwnfisz oCm mamou EmumoumII.va OHQMB .4(33 H mN.o No.o I «\z Ns.m NmN.N+ «H.H HHH.n+ va+ Nan.n+ v.vH m.mmN+ mmouuw>< m m I <\z HO (\2 v0.0N OO0.0v+ th.v+ Om0.0m+ OHN mom.m+ mHouoe I I I oz N mmm+ OO.N Omm.N+ OOmI me.N+ NH «HN+ e Is I I I oz v OOOI Om.O OOmI OOm+ I NH I m Im I I I oz 0 mOO.m+ OO.v ONO.NN+ oomI ONm.NN+ NH OOO.H+ N Is I I I oz N VON.O+ OO.H va.O+ I OON.O+ NH Nhh+ H Is I I I oz v OOmI O0.0 Ome Omv+ I OH I mHIm I I I oz v eOm.HI mN.O oomI oom+ I VH I «HIE I I I oz v FOH+ O0.0 OOH+ OmO+ oms+ mH om+ mHIm I n I oz v OmmI OO.H ommI Omm+ I NH I NHIm Nw I H I we» v mmmI OO.N ONH.HI ONH.H+ I ON I HHIm I v I oz N mm+ OO.m OON+ Obh+ omO.H+ O OmH+ OHIO Houmemm I I I oz w I OO.N I I I ON I m Id m I I oz H OHv.H+ OO.H OHv.H+ OHNI OON.H+ ON Oo+ m In NNv.NH+ HN.O m~m.o+ NHN+ OO0.0+ «N oov+ m In .w>ono H. uwummem CH poumpHmCoo appouH< mNm.m+ mm.O mOH.H+ omNI mmm+ HH mm+ N In «ON.M+ mm.O OOO.N+ Ohv+ OON.n+ mH OHN+ H In mv.O «H.H hm.O ¢\z HN.N mmm+ hm.O Hmn+ mwm+ OHH.H+ OH 0.00+ mmqwuo>< m w ,v (\2 OH <\z NH.O th.m+ vmm.N+ HHm.h+ NHH vmv+ mHmuos I I I ma» v va+ mm.O OOH+ I OOH+ OH NH+ O In HO I I I mm» s Nv+ NN.H ow+ OON.H+ ouN.H+ «H om+ O In I N H oz N huh+ om.H OOH.H+ vmn+ ONO.H+ ON OO+ m In HoumOEom I m N oz H mmm+ mn.O OOh+ OOm+ OOO.H+ ON Om+ v Id I m H oz N OHh.H+ OO.H OHO.H+ OO+ NON.H+ vH ONH+ N Is I I I oz N mNm.m+ mn.o mOH.H+ OMNI mmm+ HH mm+ N In N I I oz v mNm+ mm.O OmN+ Obv+ omn+ mH om+ H Id Houmq OCH uwummewm mpoudoo< mmquOCuo manm mmNmHm mEOOCH Anemone Anemone zoomedv Hmoxonv .Oz mumumwamm I30HHom mEmm DOC use cons Hum uo uoz mumou OEOUCH xom Hem omeCU Hum» 039 HmummEmm meCOOH uwCoom «EOOCH Honssz CH CH oquCu mmouu CH OOHHO OHOHM poHuom mqumHoHuusm moquC muwummswm DMZ CH mmcmso Hmuoe omcmco womu0>< Hmuoa Eouu mCHCHMOH ICOOH mo . w CMCO mumospoum Ho Honesz Hones? .muwosooum cusses nuHs Hammond HONHHHuuom OOOH m..o.>.o mo muHsmmmII.mIm «Hams I464 O HvH+ HN.H mmw+ HNN+ mHa+ Hms+ mmmsum>< «\z mm.ev OMN.O+ OOO.N+ «ON.O+ msm.m+ mHmuoa mnm+ m.H me+ OONI omm+ ON.H OOm+ mHIm mvH+ N.N «NH.H+ somI smm+ ON.H mmm+ «HIm NOI NO.H mmHI OOm+ HOH+ mm.H OOH+ MHIm mmh+ NO.H mmh+ wem+ mNm.H+ ON.H hhm+ NHIm Ne Hm+ O.N OOO+ OO¢I I mm.H I HHIm Hmm+ N.N mHm.s+ mHm.+I I ON.H I OHIs smummsmm I m.H I I I N I NNIm vaI o.N oms.mI oms.m+ I N .H I mNIa OON+ O.N OOO.N+ OOm.m+ Omm.m+ ON.H oom.v+ mNIm OHH+ sm.m mms+ NNm+ Nos.H+ GNH.H+ mmmsum>¢ <\z ON.OO OO0.0+ vHO.mH+ HON.OH+ NO.MH mm¢.OH+ mHmuoa mhv+ O.N Om>.m+ ONN.mI I ON.H I m I“ N¢m+ m.H NHO+ hvn+ Omm.H+ ON.H OOm.H+ N Is mH+ O.N hHH+ NHHI I ON.H I N Is NNm+ N.N ON0.0+ Omn.m+ OOO.N+ ON.H oom.m+ O Im mn+ O.e OOm+ OOH.N+ OO¢.N+ ON.H OOO.N+ m Im oooI O.N OON.HI OON.H+ I I I «NIC HOHI mn.o omNI mmn+ I I I MNIC mHm+ vm.N OON.H+ I. OONHH+ ON.H mOOHH+ NNIC He mmH+ O.N mem+ OMN H+ moo H+ NO.H com H+ HNIm oosI 0.4 cos.NI oov.N+ I I I ONIa “mummswm Hmv+ O.H NNm.H+ mem.N+ mmN.¢+ ON.H NHv.m+ mHIs nmI m.H omHI Nnm+ NvN+ ON.H NmH+ OHIm NHNI O.m HOOI OmN.H+ Omm+ NO.H Omm+ NHIa mNmI m.N ONm.HI ONm.H+ I OH.H I OHIO mNmHm mummHm Amommmv Amommmv Hmommmv Anemone HmoHHxv Hoe HO mEOOCH uoz mumoo QEOOCH OHHx “mm OOCMCO .Oz oHum OEOOCH umz Honssz CH mmcono CH OOCMCU mmouo OOHum pHOHM Show o . m CH OOCMCU Hmuoa CH mmCmso mmmuw>¢ Hobos II .mumuspoum :uoo suHs smumoua umNHHHuumm NONH m..o.>.o Ho muHsmmmII.wIm mHsms 465 .HCON wHozz On» How oumummo no: pHo mnsHo whens umwoxw mwmmuw>m 03u O0 Esmm NHH m.mNN m.nOv H.NOH m.OmN O.NNN.H muonEmz mo uonEsz Hmuoe nauseous uom O NH e.mH N.N mH O.NO mnsHU mo umnEsz Hmuoe n HOH O.HH OH.N N.HH H.HH nsHo use 3:335: .62 mono musucm>mcmsm m.Om m.qm O.mm mv O.HmN mmumnEm: «0 umnEsz Hmuoe O a”. . . 3. s .... m. $322 .1 £30 . w rm as me 1m“ mca Cm m on “.3308 . . B E . a masqammwml .323 be: mu03 : . nsHU mIv mnsHU . nH I I O. O muw>¢ I . m I I .. mcH cm CO: uo : s .msom ... 2 I I .1 maHschom $952 .38 O O N.N e N.OH mCsHO mo nonesz O.vH O.HN O.NN m.>N OO.HN N.HN nsHu Hum mHCmuwnEOx .O>4 NHH mNH mmm m.OmH OO.HON New mmuwnEm: mo umnEsz Hmuos ., m... m..ON m.mnH H mw. ommuw>< mocmumHmmC Hm ”EH FN Om OH: N. OCHmcm mHuHo .. . Om“ HNH No: H. mCHCCH mm mmwmwwz mOHEOCoom m. 1. m. .2 HP Sim? m. N. 5392 w O. ”OH ON ON. rm. MCHMCO cases mo wee: H in 3 NH mm on. 9:5: mm 3952 O O N.mH m HH n.mv mnsHo mo umnssz musqu>mcmsm HHmu OHHEHmm msHsa oHHHCM mCoHumooq oHumHumum mo maze Emumoum IoHom HHC .mnsHo :Iw .maom .m« summons pCm OOCMumHmm< mOHEocoom msom .e¢ Emumoum uOO mOHumHumum mHnmnwnEoz nsHo mo NHMEECOII.NIO oHnss 4€H5 .Huo< umsuo sH msHs O0 O ON 4N OH ON NN osHHHm>uus «5H9 O0 O ausucm>mcmsm OH «O NO Om HO musssHoHuumN OOH: usHs O0 O mnsHo sH OON NHH NON NNN Non mama “we mumoo sonsmuxm OIH mummo sou IOHN.V OON.O NHNNN Hom.mI HHO.O mumsmsoHHs nsHo .mNoO so: mum: m.OO N.HN N.NHH IN.NN N.Nm msHmemssomumN OHOHC Hos ON0.0H NNN.HN NNo.sN OOH.NNI omN.ON mumoo cpwmcmuxm masHO . . . O.N. m.O m OI N O m.m msHmIHmcsOmumsImHmHO Hanan: mIs .msom OOH.H HHN.N NOO.H «HO.N NON.H mumoo concwuxm HON OOO.HHH NNO.NNH «NO.HN amm.OHH ON0.00¢ mumoo Hmuos NH HN NH NN HN NH meuH>Huo¢ umsuo sH msHs O0 O ON ON ON NH OH HN msHHHm>mus msHs O0 O NO NO NO HO HO NO macadHoHuumN nuHs msHs O0 O ONH NNN NON osN OHN NNN «was nus mumoo :oncmuxm magnumHmms «\z pmN.m OONNN Omwaw HON OHN.H mumuwsoHHM nsHo O.OHI H.mm m.nN O.Hb, N.HH o.ONI msHs Hmccomuma OHOHMI. has moHsocoom INNNHOI OHHNOH OON.HH OOHHOH NHH.oH NHH.HH mumoo concmuxm N.N O.N O.H N.N N.N m.N OEHB HOCCOmHmmImHmHm HOQEOZ mfioz ONH OOH NHO OOO OOH HHO mumoo concmuxm HON omm.nv ONO.Hm ONO.NOH NON.OO OHN.NHH OH0.0NO mumoo Hsuos musqu>mcmsm HHmu MHHEHOC msHsB oHHHCd mCoHumooq Houomm no COHumHuomoo Emumoum IoHom HHN CH OHM mumouv .mmmo CH OEHu HOCCOOHOQ pHmHm .Emuooum mnsHo mIv .mzom OCH ocm Emumoum mocmumHmm< mOHEOCoom H.m0mmm COHQEOHOU Osom was COO muouomm Om: mousOmmm OCO umoo ConCmuxm omuOOHmm mo COmHHmmEOOII.OIm OHome APPENDIX C SUPPLEMENTARY DATA ON THE 1967 OPERATIONS OF THE EXTENSION SERVICE OF THE DEPARTMENT OF AGRICULTURE OF ANTIOQUIA 467 4158 Table C-l.-—Estimated costs of Operating the Extension Service of the Department of Agriculture of Antioquia in 1967.a Salaries of Employees Wages of Labourersb Christmas Bonus Vacation Bonus Miscellaneous BonusesC Purchase of Equipment Maintenance and Upkeep Travelling Expenses Materials and Supplies Printing and Publications Total Direct Costsd (4,094,717) Transfers to Extensione (1,416,915) Share of Administrationf (249,896) Share of Researchf Total Costs (5,953,017) Adjustment for Communication Services Total Rural Extension Costs Rural Communications Agric. Extension Division Assistance 1,572,532 75,680 222,700 657,686b -- 637,627 172,339 6,690 72,332 9,904 -- 10,005 10,533 696 5,604 91,085 550 4,298 158,912 1,700 40,015 108,382 -- 41,296 81,508 24,808 39,389 -- 48,446 -- 2,862,881 158,570 1,073,266 881,131 19,932 515,456 169,804 8,097 72,045 191,835 -- -- 4,105,651 186,599 1,660,767 +149,279 -l49,279 4,354,930 aCalculated from data supplied by the Accounting and Controlling Departments of the Government of Antioquia. bThis expense category is believed to include chauffer's salaries and costs of service personal because the total wages of obreros in the sectors totalled less then 350,000 pesos. CIncludes two special bonuses (Sobre remuneracion de Trabajg and Prima de Clima), and a transportation subsidy’iSubsidio de Transporte). dTotal Direct Costs do not agree with corresponding 1967 budget breakdown of projected expenses as these are actual expenses. 8See Appendix Table C-2. See Appendix Table C-3. 46%) Table C-2.--Calculation of Transfers to the Extension Service Division of the Department of Agriculture from other Departments of the Government of Antioquia.a Emilio Cuartos Agudelo, controller. Rural Communication Agricultural Extension Division Assistance Medical Services1 33,899 1,150 13,077 Life Insurance2 6,468 219 2,495 Public and Communication Services3 4,237 144 1,635 Termination Bonus (CesantiaSI‘ 204,511 6,939 78,892 Social Insurance (Segpros Sociales)5 6,690 227 2,581 Pensions (pensiones)6 323,828 10,989 124,919 Rentals (Arrendamientos)7 3,568 121 1,377 Family Allowance (Subsidio Fam111ar)8 293,693 -- 288,845 Other (Otras gastos generales)9 4,237 143 1,635 Total 881,131 19,932 515,456 1. 1.52% of Salaries and Wages 2. 0.29% of Salaries and Wages 3. 0.19% of Salaries and Wages 4. 9.17% of Salaries and Wages 5. 0.30% of Salaries and Wages 6. 14.52% of Salaries and Wages 7. 0.16% of Salaries and Wages 8. 45.3 % of Wages less Transportation Subsidy 9. 0.19% of Salaries and Wages aCalculated from the 1966 Statement Of Income and Expenses for the Department of Antioquia as contained in the Controller's Report for 1966-67, 4470 Table C-3.--Calculation of Administration and Research Costs Attributable to the Extension Division of the Department of Agriculture of Antioquia in 1967. 1. Total Direct Costs (pesos)a (a) Department of Agriculture- 9, (b) Office of the Minister of Agriculture- (c) Research Division- 2. Extension's Share of Total Direct Costs (a) Administration (Minister's Office) = (b) Research - Total Direct Cost Share of Administration Total Extensions Share of Research 767,343 x 0.25c = 111,252 395,217 731,892 395,217 249,896 731,892 35,451 767,343 191,835 (5,511,236b 9—iII—252 I I ) (for all three divisions) [calculated as in (a)] aAs transfers from other parts of the Government of Antioquia to the extension division were 34.6% of expenditures; total direct costs for the areas were calculated on this same basis (i.e. expenditures). bTotal extension expenditures plus transfers. 1.346 times CBased on proportion of time estimated by research and personnel. 471 Table C—4.-—Ca1culation of average total costs per day of field personnel time for the various classes of extension workers employed by the Department of Agriculture. Assume: y = ax1 + bx2 + cx3 + dx4 Where : y = Total Extension Costs = 4,354,930 pesos x = Total Days worked by Agronomos and Veterinarios = 2,254 1 x2 = Total Days worked by ijoradoras de Hogan = 3,312 x3 = Total Days worked by Practicos Agricolas = 5,601 x4 = Total Days worked by Obreros = 10,910 a a = (ggQ%—.x 1.5) + e = 312.5 + e = 407.9715 b = (1511' x 1.5) + e = 107.0 + e =_202.4715 c = (1370 x 1.5) + e = 107.0 + e . 202.4715 d = (30 x 1.8) + e = 54 + e = 149.4715 a, b, c, and d represent average total costs per day for various classes of personnel. Their general expression is given by: Correction Overhead (Monthly Salary x Factor ) + Costs/Day No. of Days worked per month The correction factor represents an adjustment for prestaciones sociales and was supplied by the accounting department of the Government of Antioquia. Overhead costs per day were arbitrarily assumed to be higher for higher paid workers as shown above. Solving the original equation for e provided the estimates of a, b, c, and d shown above. aSupplied by extension service. 472 Table C-5.--Estimated costs of Operating 1967 extension programs in the districts serviced by the Sectors of Saaneronimo and Santa Fe de Antioquia. Sector of San Jeronimo 1. District of El Brasil Agronomos and veterinarios 90 days 36,720 ngoradoraghde Ho ar 246 " 49,815 Practicos Agricolas 278 " 56,295 Obreros 350 " 52,325 Total Cost of Programs in E1 Brasil 195,155 2. District of Urquita Arconomis and veterinarios 110 days 44,880 Me oradoras de Ho ar 286 " 57,915 Practicos Agricolas 505 " 102,263 Obreros 364 " 51,727 Total Cost of Programs in Urquita 256,785 Total Cost of other 1967 Programs in Sector 19,766 Total Cost of all 1967 Programs in Sector 471,706 Sector Santa Fe De Antioquia 3. District of Tinajitas Agronomos and veterinarios 69 days 28,152 ijoradoras de Ho ar 138 " 27,945 Practicos AgricoIas 272 “ 55,080 Obreros 307 " 45,897 Total Cost of Programs in Tinajitas 157,074 4. District Of Pinguro Agronomos and ygterinarios 68 days 27,744 Mejoradbras de Hogar 97 " 19,643 Practicos Agrico as 206 " 41,715 OBferos 615 " 91,943 Total Cost of Programs in Pinguro 181,045 473 Table C—5 — con't 5. District of Buenos Aires Agronomos and veterinarios 31 days ngoradoras de HO ar 256 " Practicos AgricoIas 272 " Obreros - Total Cost of Programs in Buenos Aires District of El Espinal Agronomos and ypterinarios 74 days MejOradoras de HO an 41 " Practicos AgricoIas 269 " OBreros 614 " Total Cost of 1967 Programs in E1 Espinal Total Cost of 1967 Programs in Sector 12,648 51,840 55,080 119,568 30,192 8,303 54,473 91,793 184,761 642,448 474 Table C-6.--Distribution of man-days of field personnel time by office location for the Extension Service of the Secretary of Agriculture of Antioquia in 1967. Office éflggggflgg Mejggzdoras Pragpécos Obreros Total Location veterinarios Eggs; Egpgppgg Antioquia 242.5 531.5 1018.5 1536.0 3328.5 San JerOnimo 240.5 531.5 797.5 695.0 2265.6 San Pedro 502.0 478.5 784.0 1740.0 3504.5 Cisnoros 156.5 266.0 713.0 1044.0 2179.5 La UniOn 293.5 424.5 567.0 1217.0 2502.0 Santuario 177.0 562.0 730.5 164.0 1633.5 YolombO 215.5 -- 240.0 878.0 1333.5 Medellin 426.0 578.0 750.0 1895.0 3589.0 Nursery -- -- -- 1740.0 1740.0 Total 2253.5 3312.0 5600.5 10910.0 22076.0 4375 ONO.OOO OON.NHO OOH.NOO.H III III mHsuoe III OOO.OO OOO HOO OOO.H . HOH O . OOO OOO.N mwm.m wOOHOO OOO OO me.um uOO HOH OOOH OOO.H HON.O OOH.OH Ow".HO OOO OO OOO.OO owO HOH OOOH OOO.H OO0.0 OOO.OH OOO.NO OOO OO OOO.OO OOO HOH NOOH OOm.N OHH.O OOO.ON OOOOMM ONO om OOO.OO o m HOH OOOH HON.O HNO.O . On OOO ON . OO HOH OOOH NOO O OOH.OH . OH ON . HOH OOOH . OOH NN OOO.ON OOO OO OOO NON O OON.HH . OOO OO . HOH OOOH . OOH om OO0.0N com om com OOO O OOO.OH . OOO OO . HOH NOOH . oom Nn OO0.0." 0 com on com OOO O OOO.OH OOO.OO OOO. NO OO OOO OO O HOH HOOH ONO.N NOO.OH OOO.NO OOO.OH OOO OO OOO.OO OOO HOH OOOH OON.O OOO.OH OOH.NO OOO.OH OON OO OOO.OO OOO HOH OOOH OOO.OH HNO.HN OOO.OO OOM.O OON OH OOO.OO OO HOH NOOH NON.NH OON.ON OOO.OO OOO.O ONN OO OOO.OO nwO HOH OOOH OOO.OH HOO.ON OOO.OO OOO.O OON ON OOO.OO O O HOH OOOH OON.OH ONO.ON OOO.OO OO .O OOH ON OOO.OO OmO HOH OOOH man-0H OFCNQN O H N OCH ma N m HOH mama . OON OO OON.H OOO Om OOO OOO OH ONO.OO OOO.OO OO ONH OH OOO.OO O HOH NOOH OOO.ON NOO.HO OOO.OO O OOH O OOO.OO OOO HOH HOOH OON.HN OOH.NO . . OO HOH OOOH . OOO OO OOO OO OOO NOO NN OOO.NO . . HOH ONOH . OOO NO OOO OO OOO OOO NN ONO.HN . . OO ONOH . OOO OO OOO NO OOO NOO NN ONO.ON OOO.OO OOO.OO OO OO NNOH OON.HN OON.ON . . O HO ONOH . OOO NO OOO OO ONO NOO HN NOO.ON ONO.OO OOO.NO O NO ONOH ONO.ON ONN.ON . . OO OO ONOH . OON On ONO OO OOO NOO ON ONO.ON O . . ON ONOH . OO NN OON OH OHO ONO OH OOO.NN OON.ON OOO.NN O ON NNOH OOOHOH OO0.0N OOO.NN OON.ON OMM HN HNOH ONO OH HNN.OH ONO.OH OOO. NO ONOH we OO OOO 0 she: COHOCOuxm o» mumwwmwm NMMMMMCMMO OOH wMWOOOHC munmam mesdm mom OOMMOOOC OOOH 3 o . Hess nHuuus OOHOOOOO coHussHuunHa o>HussuouH<.ansmowm oz Houos OOHOosmm OwHHHsuN Oo .oz use» COHuanHumHo OOOCOOHOCHICOHmcouxO .usanuC a C u H H x 03 CoHucouxo scum uonu uHOOCoo soundnoum no COHucHCOHsUII.NIo OHnds 476 OOO.NOO ON0.0MN OON.OOH.H III III III OO0.00 OON ONH OO0.00 OON ONH OOOH HOO MOO.H OON.O OOO.HO OOO ONH OO0.00 OON ONH OOOH nOm.H ONH.O OOm.HH OON.ON OOO OHH OO0.00 OOO ONH anH NOO.N Nn0.0 OO0.0H ONH.OO OOO OOH OO0.00 OOO ONH NOOH mHN.n HOH.O OO0.0N OON.OO ONO NOH OO0.00 OOO ONH HOOH OO0.0 OOO.NH OON.Nm OO0.00 OOO OO OO0.00 OON ONH OOOH OO0.0 OO0.0H OO0.0m OON.NO OOO OO OO0.00 OON ONH OOOH OO0.0 OO0.0H OO0.00 OO0.00 OOO OO OO0.00 OOO ONH OOOH OO0.0 NOM.ON OO0.00 ONH.NO OOO ON OO0.00 OOO ONH NOOH NMO.HH ONO.nN OO0.00 OOO.Nm ONO NO OO0.00 OON ONH OOOH OOO.NH OO0.0N OO0.00 OOO.mm OOO OO OO0.00 OON ONH OOOH ON0.0H OON.ON OO0.00 OO0.0N OOO OO OO0.00 OON ONH OOOH ON0.0H OHv.Nn ONN.OO OO0.0N OOO vm OO0.00 OOO ONH mOOH ONN.OH OON.Om OOO.NO ONH.HN OOO Ov OOO.NO ONO ONH NOOH OO0.0N OO0.00 OOm.HO OO0.0H ONO Nv OON.ON OOO ONH HOOH OHN.NN OON.Om ON0.00 OO0.0H OOO On ON0.0N OOO OHH OOOH OO0.0N HO0.0m OOH.OO OOO.HH can on OO0.00 OnO NHH ONOH HON.vN mmm.Nm OO0.00 OO0.0 OOm ON ON0.00 OHO OOH ONOH OOO .ON HNO.OO OON.OO ONH.O OON OH OOO.NO OOO OOH NNOH OON.ON OO0.00 OO0.00 OO0.0 ONm NH OO0.00 OOO OOH ONOH HO0.0N OO0.00 ONO.NO OOO.H OOm O ONN.OO ONO Om ONOH OON.ON NO0.0m OOO.HO OOO.HO OOO No ONOH OO0.0N OOO.NM ON0.00 ON0.00 OOO OO MNOH OH0.0n OON.Om OOH.vv OOH.vv ONO OO NNOH OO0.0N ON0.0m OO0.00 OO0.00 OHO OO HNOH OOO.ON OO0.00 ONO.NO ONO.NO OOO ON ONOH ON0.0N mmm.Hm OO0.0n OO0.00 OOO NN OOOH OON.ON OO0.0m ONO.Hm ONO.Hn OOO OO OOOH an we OO nuHOOCOm OHHEMO use meHHeuh muHHOCOm MHHsmm use mOHHHadm new» xuos COHmCmuxm Ou Hooch muHOOCOm mo uwnssz Hsuoa OUHOOCOO mo nonssz OHAOuanuuuc OUHOOCOO COHusoHuunHO OuoCHOqu OHndooum COHuCOHuuuHo OOOCOOHNCHICOHnCauxm .nouH< uOCosm CH xuo: CoHnCouxo How nonu uHuuCun omuOOOOHm no CoHunHOOHsoII.OIU «Hana 4377 mHnsussHuuu< OuHOosom GOA .uanuumHo wumcumuH< OHnunoNO COHus AHuumHo OOOCOCHOCHICOOquuxu “HN.H OOO.H mm“.w ON0.00 MM" am HOO. NHH.O OON.O OOH.OO OO OO OOO.HO OOO H NOO. ON NH OOO. ONO O OO . ONO OO O OOO.N OOO.M ”ON.OH ONO.MM OOO OO OOM.M" ONO mm OMMM NOO.O NHO. OO.OH OO . OOO OO O OO H ON OOO. NO NOOH OHN.O OO .N OOH.HN OOO ON HO OO H O . ONO.ON OOO. ONO OOO NHO.O OH . OOO ON . OHO ON HO OO H O OH . OOO OH OOO. ONO mmHNm me O mOm MN com me Hv mm mamd O HH . ONO.OH OOO. ONO Homem HA 4 OOO ma . mom on do mm vmmH N OH . OOO OH OOO. ONO O meO@ mm O mOH on O OFN mm HC ma mad O OH . ONO HH OOO. ONO OHO.N Om . OOO HO OON Om HO mm NOOH O OH . OOO.O OOO. ONO ONO.O OH . OON mm . OON mv Hv mm HOOH O NH . ONO N OOO. ONO O ONO.O OO . OON OO ONN OO HO OO OOH O OH . OOO.O OOO. ONO O .wmoOHH ”h N mOh on N OHN mm Hfi mm QQH N ON . ONO O OOO. ONO O mNmNNH fiH O cam hm mad OH H? ma mmH N HN . OOO.O OOO. ONO N OOO.NH MO . OOH On . OOH mN Hv OO OOH O HN . ONO N OOO. ONO O MON.NH Om . OOO Nm OOH ON Hv Om OOH O HN . OOO.H OOO. ONO m mam mu: ”we .... mm ...M at... “N m be HHN.OH NO .ON ONO.HO O OOO.OO OO OOOH m OH O com. CA? N hQOOMH MN O cow GN mm mm mad N OH O ONMO 00‘ OHO.OH . OOO NN OO NO HOOH OOO.NH MMM.OH OON.ON ONman ”mm on OOOH OOO.OH HOO.NH ONH.ON OOO.ON ONO OO ONOH OOO.NH OOO.OH OOO.NH ONH.ON OOO NN ONOH OOO.HH ONO.OH OOO.OH ONH.HN OOO OO ONOH NOO.HH OO .NH ONO.OH OON OH O OO ONOH O HH ONO.NH ”WMHNH OMM OO ONOH OO ON ONO.M# wOO WM “MMM ¥HO3 GOfiflGQHXN ONO muflWQCflm >Hdfidh HOW ONMONH OM” QC ”wad Hsuos uuHuuCum no OOHOOOHO O OO OOMM HHHsu . OHOOCOO h M0 02 fiddfidk Hauoa OwHuoswmm euuoouue wfiddflfidh MO .02 HMO? .ousOCHm CH xuo: COHuCo uxo no u uson uHuuCOn souvenoum uo . CoHuoHsoHso II.OIU OHAOB £1753 O ON0.000 NOO nun nu N‘NO F'mOH I. 0° O OOO.“ MMMHO wmwum ONW.MM OOO O O3 OOO ”O O OO0.0 N OOO. OOO w s 'mmnfi H s 0 mm Om.“ OOOO.O ”mum “WWW“ MM" Om “WW”: ”WM w» OOOH NOO.O HN.HH ONO. OOO.OH ONO OO OO NOOH OOO.O OOO ON OO0.0 OOO Oh NON.O OOO.OH ON .ON OOO. ONO OO O OOOH H ON O OOO.O OOO ON ONO.O NHO.OH OO . ONO. OOO OO O OOOH O on O OOO.O OOO Oh mnNOHH omehH mNOO OOOOm th on M mama OOO.O OOO ON OOO.NH NOO.OH OOO.HO ONO.O OON ON . O O ON NOOH MO”: mum 3m a“ m “Om : Ow MH com. hm N O 0 Oh H “was “mum 2% E w: ”O.N” an“ 2 as FOMOMM vaOmH OOmumN OOOOmM Oom Oh "an“ OOO.OH OOO.OH OOO.NN OOO.OO OOO ON HOOH meOMH HOthH Ohm mN OOnONM OG’ Ox- ODO . OO . OON.ON OOO. ONO ON H HOH OH O. OH O ON OOOH . NNO.O OON HN OOO.N OOO OO NON NH H O . N ONOH OOO. OOO.OH ON OH ONO.ON OOO OO NN .NH OOO.O OOO.NH OON.O OOO NO OH MMW.MH OOO.NH ONO.OH OON.HM OHO OO MMOH H ONO.HH OOO.OH OON.OH OOO OO OH OO xuoz an wmmumw “MM mm NNOH COHOGOUXN Ou no uuwwwcwm OOO.NH own VO WOOH HQUOB %Hfl&dm Mom mNm A? NOH muHuocmm meHHEMh Om OOOH Oo No OOOOOOO OOOH nfifiz m hHHE HMUOB . uh uwm munmcmm MGHHHEME H 0 HnOuOnHNOOO OOHOOOOO :oHu: nHquHO ouncuuuH< anmno Hm o m .nuHaqu c H xuoa co anuuxo no O onO uHuoaon Oouoonoua mo :oHuuHao Hun--.OH .0 «Hana IIIHIHIHIIMIIH 5433 HHH 1293 031 iHIIIWIHHIIHIIIIH 3