Agenda and conference papers. The Rockefeller Foundation. April 29-30, 1975

Material Information

Agenda and conference papers. The Rockefeller Foundation. April 29-30, 1975
Conference on Social Science Research in Rural Development


Subjects / Keywords:
Farming ( LCSH )
Agriculture ( LCSH )
Farm life ( LCSH )


Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services ( with any additional information they can provide.

Full Text
APRIL 29-30, 1975

Draft -4/75
This paper represents a preliminary exploration of the problems and
advantages of the direct use of social scientists and social science research in private, national, and international agencies dealing with rural development. It concentrates particularly on the non-economists: anthropologists, human geographers, political scientists, and sociologists. The author is
-an anthropologists with sixteen months of ifieldwork!I in a U.S. private non-profit agency, six months in a kenyan joint university and government project, and nine years of exposure to academic anthropology and sociology.
Below, the changing organization of rural development programs is
described as background to social science involvement. The complementarity of social to physical science is noted. A distinction is made between posthoc evaluation of programs and program design and monitoring, and the types of contributions social scientists can make to the latter two are outlined. The effects of the research processes used by academics, fully employed non-academics, and part-time consultants are discussed.
Difficulties of agencies in dealing with social scientists include
poor identification of disciplinary specializations, lack of recognition of social science inventions as such, and a cultural antipathy to the whole idea of professional intervention in social systems. Difficulties of social scientists in dealing with agencies include lack of freedom to complete
N ~ie~cw~'k

investigations to their satisfaction, the need to undertake additional administrative and political tasks if their recommendations are to be accepted and implemented, and the comparison with a more secure, leisurely, and prestigious life in academia potentially open to them. Recommendations are made to improve the use of social scientists in action programs.
'Int eratinE" Development
The breadth of rural development programs as defined by governments
and international agencies has probably always been in dispute. Essentially', "integration" of rural development program implies separate administrators Sithin, or parallel to, existing government offices or ministries. These units have the same assigned roles as the ministries of health, education, industry, etc., but deal only with rural areas and interact among themselves as a kind of shadow cabinet. Sometimes this parallel body will be a mission operating at the community level, sometimes a foreign aid agency staff at a regional or national level, others a separate ministry within the national government. Extant ministries or local authorities naturally resent the implication that they cannot deal with rural health, education, etc., themselves. Planners and tax-payers alike wonder whether the additional bureaucracy is not more costly than it is effective.
The 1950's solution to this problem was the "community development worker", born out of the needs and ideologies of post-independence India (cf. Mayer et al. 1958;Sanders et al. 1958). Although the worker's task was stated primarily in terms of community organization and motivation, in order to carry out his task he had to judge local development potentials and draw in external aid where local resources were insufficient. The community development worker concept came into disrepute on the international level primarily because the typically isolated, undereducated, often urban-bred

worker could not cope with such a task without massive backup support that never materialized.
AS disillusionment with the community worker concept grew in the 1960's international development ideology shifted to an industrial-urban mode, with secondary non-integrated rural projects planned from the national level. During the project-specific sectoral period the agronomists achieved major breakthroughs and the Green Revolution began to spread. The widespread input of effective new technology into extant social-systems led to highly visible increases in social inequalities in many of these nations (UNDP 1974). Social scientists from academic (e.g. Frankel 1969) and government-consultancy (e.g. Wharton 1969) positions pointed out the implications of these changes in journals widely read by political elites, and increasing rural revolts and overcrowded cities reinforced their warnings. As a result, international assistance ideology came back again to "integrated" rural development, looking for an effective methodology to anticipate and modify negative effects of sectoral programs.
In the abstract, "rural development" is usually defined as the process of increase in welfare and productivity of all (or poor) people outside metropolitan areas and cities*. For agency X in country Y, rural development usually meis continuation of older sectoral emphases (agriculture,livestock, nutrition, sanitation) with the addition of evaluation criteria specifying the range of effects on a large percentage of the population. This is a rational proceeding, given the strengths of present agency personnel and the importance of the specific goals. But it can lead to situations in which all the integration is in posthoc evaluations and none in the design and management of the projects.
The social scientists have already made a major contribution to bringing
*Towns are usually included as an integral part of rural society and economy.

about the shift in emphasis back to integrated development. They are highly competent in posthoc evaluation, both as consultant-employees of international assistance and. local government agencies and as independently financed academic researchers. But very few are interested in the program-designd and monitoring/ management aspects that are so necessary to the implementation of the integrated rural development ideology. They should have the ability to contribute to the solution of the methodological problems of "integrating" development projects under given resource constraints.
Aside from the economists, very few social scientists are now working on these problems. This is a joint failure of the agencies that should employ such people and of the social-science disciplines that train members**to research and write independently of non-academic institutions. The Potential for Social Science
By definition /ocial scientists deal with the analysis and modification of human organizations, and /hysical scientists deal with the analysis and recombination of non-human elements. When the latter develop a new technique for rice production or cloth making or water purification, varying information is spread to organizations in government, business, religion, philanthropyi communities and neighborhoods, who then use, abuse, or ignore the technique to serve their perceived interests. Social scientists sometimes modify the organizational environments to make better use of the technique or to protect against its negative effects; more rarely they work with the physical scientists to build such inputs directly into the new technique as it is evolved. They can do much more in these areas.
Agencies which plan to introduce a technology (package of techniques)
into a particular locale for a "rural development" purpose (improving the welfare
**Some masters' programs have recently been developed to train applied researchers. A few faculty in some PhD programs communicate such
training to their students, especially but hot exclusively in economics.

and productivity of the rural majority) should include "political", "economic", and "cultural" factors within their analysis. Such categorization of interrelated social factors does not imply that the analysis and modification of plans will best be done by a political scientist, economist, and anthropologist working as a team. Depending on the problem and agency resources, it might be done by a single sociologist or anthropologist with competence in the three areas and knowledge of the locale. Usually important questions of power, profit, and value include:
"Politica! besides the agency, what organizations from the nation state to the family have a stake in the new technology and its short-term or obvious effects on the
locale? what are they likely to want of it and how powerful
are they? tan some modification in the technology or its
distribution system (including information) improve the chances
of its being accepted and used for the purpose the agency intends, or should the agency wait until a power shift has
occured in the country or region?
"Economil what effects will the technology have on the total supply and distribution of jobs and capital across
regions and classes? will it be economically profitable to
the average individual under local taxation, marketing,
communications, social and religious arrangements? are there
enough trained people to teach or otherwise administer the technology? will it help or harm national economic goals?
"Cultural": what extant local organizational arrangements and
preferences will the technology force people to give up if
it is accepted? which are they likely to regard as the most

serious deprivations? what changes in the technology,
distribution system, or other extant organizations could
soften or avert such deprivation?
Social Scientists_inA encies
Why are there so few social scientists at the professional level working in or for international assistance agencies and national ministries? Anthropologists and sociologists worked in the British colonial services during the period of expanding empire, and applied their skills to colonial occupation of Oceania and to military and civilian morale problems during World War II. Rising demands for university education ( and McCarthy-ite pressures in the U.S.) sent them back to academia to teach. Present development planning teams seldom include a non-economist social scientist.
Lack of a common technical language, scientific individualism, administrators' fear of higher status employees and scientists' fear of being trapped into actions discordant with their personal philosophies., are all side issues. Physicists, engineers, and agronomists face these as well, and yet many work for governments and private industry. Given adequate motivation by scientist and agency, these issues can be overcome.
Lack of disciplinary definition is a more central problem. If an agency hires a plant botanist or a sanitary engineer, they have a high probability of getting the skill they had expected. A social scientist is not so easy to type. An economist may concentrate on political factors (e.g. deJanvry's (197th) resolution of the rural poverty syndrome through revision of national political strategies), a political scientist on cultural ones (e.g. LaPorte's and Metlay's (1975) exploration of the U.S. public's evaluation of science and technology), and an anthropologist on economic ones (e.g. Smith's (1972) analysis of Guatemalan rural market systems) A Scot legal anthropologist who had worked in Lesotho on court systems was hired by an engineering firm to analyze community cultural

factors for a SVifziland agricultural development plan (Hamnet 1973). He Amented that their hiring procedures were illogical but how were they to know that an anthropologist/ was not necessarily expert in general community social organization? The social sciences continue a fertile inter-disciplinary exchange through such specializations but they make it very difficult for an inexperienced employer to identify the skill or person they want. One reason for the present popularity of economists on development teams may well be that from the 1930's until very recently they had reduced themselves to .a very few easily identifiable and effective analytic methodologies and were thU3 more predictable to outsiders. Recent political-cultural upsets on the world scene (OPEC in particular) have limited the effectiveness of the prevalent methods and may lead to a broadening and concomitant unpredictability.
A related distinction of the social sciences is that their products are
seldom (if ever) identified as such when they are applied to reality. An automobile, an atom bomb, or even a new wheat variety is attributed by the public to the application of a particular science. The carefully formulated balance of local and British law in colonial African courts, the "war on poverty", and the SALT talks are attributed to (lay) politicians and administrations. Such attribution is correct, it is correct to say that the wheat variety is also due to
the industrial organization that managed the research and the replication and distribution of the product; but a great deal of social science theory and applied analysis went into each achievement. This lack of recognition may be tied into a more general cultural distaste for the idea that some people are more qualified than others to tinker with social institutions. The ignorance of past social science achievement results in an equivalent ignorance of potential social science contributions, and agencies do not see a need to consult or hire them. Economists, again, are better off at present because one of their number I John Maynard Keynes was brilliant and flamboyant enough at a time of international crisis to attach his name to the resultant solution. Even there

economists seem to be less connected to the New Deal in popular writing than is Roosevelt, while the space program is more connected to physicists and engineers than to ennedy or his successors.
The crucial difficulty, in all probability, is public distaste for the idea of a specialist designing social systems. This distaste is a cultural trait found in the West and in Western-trained elites in the LDC's it may be universal. It is behind the depth of public to B.F. Skinner's Walden and +0 George Orwell's 1984. It is so pervasive that even most social and behavioral scientists who have chosen such specializations for careers justify themselves by saying their sciences are too imperfect for them to develop the degree of control and prediction they are working towards. The "layman" fears that the social scientist will have different ideas on the ideal shape of an organization and the power to implement them over his objections or without his knowledge. Social scientists thus are hired by one group to exercise their skills on another group the first considers hostile or inferior. Hospitals hire them to study and predict/control patients, housing authorities to study slum dwellers, minority organizations to study government agencies and political bodies.
Y'e- fkI i-lie, ILe rt' d cW I
Rarely is the job phrased in terms of studying or modifying the "patron,groupt as with the clientele as in the mismatching of agency goals for their community development workers with national resources and priorities in the 1950's. Organizational criticism and revision of the social scientists "patron" group is a tricky business. Failure to do it cuts his usefulness to the group and increases his frustration with applied research.
Social Science Roles in Rural Development Processes
Social science roles span two continua, one of immediacy of impact and
another of scope of analysis. The scientist's research may be fed directly into the process of choosing rural development strategies, or it may filter into this
- 8 -

arena through a series of screens --publication, teaching future politicians and~administrators, teaching future voters. The former may be designated "applied" and the latter "academic", In the sense that the ideal academic researcher is an independent, uninvolved observer. Again, a researcher may concentrate on a single project or on a human organization model that will predict the impact of a wide variety of projects for a given typeof society. An example of project analysis is Paul Kaplan's (1974) test of the effect of resident community development workers on long-term village progress in the Philippines; of modelling, Carl Gotsch's (1974) analysis of the interactions of rural-national political and economic systems and their effects on the introduction of integrated and sectoral rural development programs.
In the, extreme case, the "academic" researcher builds his theory and hypotheses, collects his data, and writes his reports with only his academic colleagues in mind. He may alter his publications slightly to attempt to avert harm to the people he studied. Known in the more industrialized countries as "academic freedom," this (extreme) case is becoming known in the less-developed countries as "academic imperialism," and requirements for research visas are being tightened to force some reciprocal benefit to the countries of the people studied. Research produced in this. fashion goes into scholarly journals and books, and occasionally builds to a new social theory which influences studetits who later become public officials and community leaders. More rarely, a publication comes to the attention of an official or other development program agent w6at
who is an energetic reader-and 10 adapt to his program needs.
On the other extreme, the "applied" researcher works with officials of the government, religious or minority organization, international agency, or other action group. He discusses and comes to an agreement on goals, and studies h is employers and co-workers to determine the range of means they can accept to attain those goals. At the same time he tries to do all the work done by the academic researcher of the same problem. Finally he combine s his conclusions on

the two study groups -the agency and the "target" population -and draws up recommendations which form the basis for a discussion with agency personnel that may continue for months or years before both parties are satisfied and the changes in procedure fully implemented.
Although the majority of studies lie between these two-extremes, there are more examples of the "extreme academic" case than are usually found at the end of a continum. One increasingly common variant is the case in which the researcher tries to make his analysis relevant to a non-academic group by tacking on a list of problems uncovered or even proposed solutions. Moving toward the applied end are the scholars who evaluate programs (consult) occasionally for an agency-and the ones who contract themselves to work in agency programs for a year or more while retaining university links and security. Contrasted to all
*the benefits that have been built up for U.S. tenured faculty since World War II, it is apparent that agency employment is not secure. Many of the best applied social scientists are drawn back to academic work by professional colleagues who cannot believe applied research is either-good science or a service to society.
The ethic of "academic freedom" consists of a positive freedom (to follow one's interest into whatever inquiry) and the lack of a constraint (being compelled by contract to carry out research or write reports that may be morally distasteful or uninteresting). The former is definitely curtailed in agency employment and consultation: at least for major projects, the. researcher must justify his expenditureof time and supplies to his employers. To a lesser extent, the academic is subject to the same pressures through grant proposals,
*university promotion and tenure fights, and teaching commitments. But'.the older
-and more prestigious academics are rewarded by increased freedom, while the equivalent applied researcher is pressed into more and more commitments to projects important to the agency but not necessarily central to him. The problem of constraint arises only to the extent that the researcher has chosen an agency
with which he disagrees on basic goals and is unwilling or economically unable to forfeit his job to seek another.

At least superficially, project analysis seems to be most suited to
applied research and model-building to academic. Agencies often press for fast results to specific problems, and academic social scientists are expected to formulate and test general theories. Models are also more definite expressions of final goals (participatory democracy or centralism, agriculturally or industrially based economy) and commit agencies to widespread modifications in ongoing programs, and so are more difficult to get accepted. Researchers at the Institute for Development Studies in Nairobi worked for six years to get-.a more decentralized planning model tailored to the national situation and accepted into its bureaucracy (Leonard 1973), but a new technique for selecting and training less progressive farmers in modern agriculture was incorporated into the extension system within two (Roling et al. 1974). Academic models can be refined. without such strong pressure to conform to current values, and, over time, through publication and teaching, can lead to revisions in public goals that allow them to be applied. Application of a model to use by a particular agency for a par*ticular locale can be eased when a social science researcher that is familiar with the model's constraints and ambiguities is acting as consultant. But such advice is most useful when the researcher also has a knowledge of agency constraints and potentials that is best gained by full-time employment with the agency.
Most of academic as well as applied reporting (and research) is actually project-oriented rather than model-building. The chances of academic project evaluation and recommendations getting into the hands of the community or agency leaders in a position to do something about them are very small. Officials and politicians below the national level in most countries are highly unlikely to have access to the publications or to be able to interpret the scholarly language. Academic project reporting goes into the process of theory building within the scholarly community, but does not effect the course of rural life in the area where the study was undertaken.
- 11 -

There is a need for a spectrum of academic through applied social science researchers. Presently academic researchers are in the great. majority. Agency and academic organizations must be modified so as to increase the number and quality of applied researchers as well. A few suggestions towards this end follow.
Disciplinary definition: It is argued above that lack of clear disciplinary boundaries is an advantage to social science training. But once trained, movement into the applied field requires that social scientists learn to present themselves in such a way that others will know what to expect oT them. Exposure to "laymen" during graduate training helps. One useful device is the graduate research project in collaboration with a local community organization, under its guidance/and control, now occasionally being used in a systematic way (e.g. Leacock 1974). Agencies in return have to learn that in social science identification the sub-field is'sometimes more important than the discipline. Personal summaries of training and interests may be more useful than disciplinary identification. Prestige,research freedom, and training applied research: National and regional research institutes with strong or sole interest in rural development problems, financed by a mixture of government, international, and private contracts and grants, could be used to train graduate-level apprentice researchers as well as development program-personnel. This dual-function applied/research and training institute would parallel the university at the graduate level and thus could be structured to provide stronger security and professional status for its staff helping to solve the problem of holding good researchers in the applied field. Most of the funding wouldbe tied to projects, but a small amount of core support (say 20-25%) would allow the group to free some members for projects or refinements which were important to them as social scientists in their broad study of rural devel- 12 -

opment problems. More general social science institutes within universities (CAEN at Fortaleza, Brazil; Institute for Development Studies at Nairobi, Kenya; etc.) are moving in this direction, bl4t few have moved to formalize training and research distinct from regular academic departments. Without this step the institute personnel are under pressure from academic colleagues to return to full-time teaching, and students continue to be trained primarily by academicians with little attention to applied needs. Since social science research work in rural areas presents some special problems of data collection and covers a different range of problems from urban or national development research, it would be worthwhile to establish a few specifically rural-oriented regional institutes around the world.
Time constraints in agency research: This is especially a problem
with model application which takes a longer time period than project evaluation. Few social scientists are now employed by agencies for long enough to undertake .initiation or basic revision of development programs. Most of these few are economists. Career positions for social scientists from all fields who have shown an ability to grapple with applied research in their graduate or sabbatical work should be increased within international and national agencies. In countries where native-born social scientists are in short supply, this should be secondary to the establishment of the research-training institutes mentioned above, but should be a publicly recognized goal for future expansion.
Application of academic project research: As stated above, much of academic project reporting is lost to the projects studieJ. One way around this is for grant-making agencies to institute similar requirements to the research clearances required recently in many less-developed countries. Only research which was done with the concurrence of a specific development agency (whether public or. private) would be financed, and research reports would have to be returned to them. This would cut out a lot of possibly useful research that was in some way hostile to the prevailing political elite of a country. Another possibility is the funding of a research bulletin which simplifies and disseminates
- 13 -

academic research to a network of rural development agency personnel, either world-wide or regional. Possibly grant-making agencies could prepare a precis during evaluation of past research grants to send to the American Council on Education's Network Bulletin, if this new publication could be expanded to cope with the influx, or to a modified UNRISD bulletin series. Proliferation of networks covering the same regions is a threat to their purpose, with projects sponsored by one agency or from one university going to one place and those by another to a second. Agency agreement should be reached on a single network center at least for each continent.
Social scientists have much to add to the present emphasis on integration, or brfljdening the effects of sectoral programs. At least one discipline anthr was effectively born out of development programs. It is time for the disciplines again to normalize their relations with such programs.
a) Francine R. Frankel, "India's New Strategy of Agricultural Development: Political
Costs of Agrarian'Modernization", Journal of Asian Studies 28:693-710. 1969 b) Carl Gotsch, "Economics, Institutions and Employment Generation in Rural Areas"
In Edgar 0. Edwards, ed., Employment in Developing Nations: Report on a
Ford Foundation Study. Columbia University Press. 1974
e) Ian Hamnett, "The Role of the Sociologist in Local Planning", Journal of
Development Studies 9 (4):493-507. 1973
d) Paul Kaplan, "The Impact of Planned Change and the Testing of a New Framework
for Community Development", Journal of Developing Areas 8 (3). 1974
e) Todd R. La Porte and Daniel Metlay, "Technology Observed: Attitudes of a Wary
Public", Science 188:121-127. 11 April 1975
f) Eleanor Leacock, et al., Training Programs for New Opportunities in Applied
Anthropology: a Symposium Sponsored by the Society for Applied
Anthropology. Boston, March 1974. publication by American Anthropological
Association, Washington D.C.
g) David K. Leonard, ed., Rural Administration in Kenya. East African Literature
Bureau, Nairobi. 1973
h) Albert Mayer, et al., Pilot Project, India: The story of Rural Development at
Etawah, Uttar Pradesh. Greenwood Press (Westport, Conn.) 1958 (reprinted
- 14 -

i) Niel$ Ruling, Joseph Ascroft, and Fred wa Chege, "Innovation and Equity in
Rural Development". presented at the Eighth World Congress of Sociologj,
Toronto, 1974.
J) Rural Development Network Bulletin. Overseas Liaison Committee (American Council
on Education). begun fall 1974. (Washington, D.C.)
k) Irwin T. Sanders, et al., "Community Development". special issue of Rural Sociology
23(1)1. 1958.
1) Carol Smith, "Market Articulation and Economic Stratification in Western
Guatemala". Food Research Institute Studies 11(2):203-233. 1972
m) UND, Global Research on Social and Economic Implications of Large-Scale Introduction of New Varieties of Food Grains: Summary of Conclusions by United
Nations Research Institute for Social Development, Geneva. GLO/71/002. 1974
n) Clifton R. Wharton, Jr., "The Green Revolution: Cornucopia or Pan4ora's Box?",
Foreign Affairs 47:464-476. 1969
- 15 -

April 9, 1975
Ralph W. Cummings, Jr.
The Puebla Project is a rural development program which deserves detailed study for several reasons. The project is aimed specifically at
small farmers. It has managed to achieve the goals originally specified.
It has had a direct influence in stimulating the establishment of similar
projects in other states in Mexico and in several other Latin American
countries as well as having had an indirect influence in the implementation
of programs in Asia and Africa. It has a well-documented history over an
eight-year period. As a result, it should cause students and practitioners
to reassess what is really meant by rural development and what is needed
to bring it about.
This paper has two interrelated purposes. The first is to explain
the Puebla Project as a rural development project. Some observers argue
that because the Puebla Project focused on agricultural production and did
not initially include specific activities in such other areas as health,
* This paper is prepared for a meeting on Social Science Research in Rural Development to be held at The Rockefeller Foundation, April 29-30, 1975. The paper borrows heavily from five trips which the author has made to Mexico since September 1972 as well as several recent excellent accounts of the Puebla Project including: (1) The Plan Puebla Staff, The Puebla Project: Seven Years of Experience (1967-1973): Analysis of a program
to increase crop production in rainfed areas of small, subsistence farmers.
Mexico, CIMMYT, 1975 (this report contains a bibliography of 22 theses and 22 articles and reports written on the Puebla Project); (2) Leobardo Jimenez S. and Reggie J. Laird, "Mexico: The Puebla Project A program to increase crop production by small, subsistence farmers in rainfed areas," appearing in Strategies for Agricultural Education in Developing Countries, The Rockefeller Foundation Working Papers, December 1974; (3)
Heliodoro Diaz-Cisneros, "An Institutional Analysis of a Rural Development
Project: A Case of the Puebla Project in Mexico," a thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the University of Wisconsin, 1974. Sterling Wortman, Reggie J. Laird, G. Edward Schuh, Donald Winkelmann, Edgardo Moscardi, and Herman Felstehausen read an earlier draft and made perceptive comments. However, the views expressed in the paper are the author's alone and should not be blamed on the others cited above.

education, and off-farm employment, it should be considered as an agricultural development project rather than a rural development project in the broader sense.*
Rural development is an especially difficult concept to define. What people want from life varies in specifics from area to area and from person to person. However, generally every person, at the minimum, needs acceptable levels of food, health, shelter, and clothing. No less important than these, he also needs less easily defined factors that in sum contribute to his self-respect, identity, satisfaction, and spiritual wellbeing. His notion of "acceptable levels" of wants and needs will, over time, very likely alter with the rising expectations that accompany development. By and large, however, it may be argued that attainment of these goals as they evolve is contingent directly or indirectly on the growth and distribution of family income, which becomes a significant objective. Rural development includes achievement of this broader range of objectives by a wide group of rural people on a sustained basis.
Some activity or activities must be initiated to increase income. In rural areas, the most obvious income-generating mechanism is agriculture. In some cases, agricultural change can result in a limited range of benefits for a limited number of people. However, a project whose initial conception, at least as related to inputs, would appear to place it in the agricultural development category, can produce a spread of benefits, indirect as well as direct, such that the ultimate results are much the same as, if not grea ter than, those achieved by projects which begin explicitly as
*Arthur T. Mosher, "Projects of Integrated Rural Development," A/D/C reprint, The Agricultural Development Council, New York, N.Y., December

recognized rural development projects. There too often is a tendency to
attempt to force inclusion of such activities as public health and primary
education from the beginning when a simpler focus initially, followed
several years afterward by the addition of other activities, might be more
appropriate. The Puebla Project is a good example of an "agricultural
development" project which really is a "rural development" project.
The second purpose is to describe the contribution which social scientists made to the implementation of the project. The institutional constraints were probably at least as important as the technical constraints
in preventing the majority of the campesinos from using services in the area. The team members with social science training had the particular
task of identifying these institutional constraints and attempting to work
outmethods of overcoming them. This input, appearing in several places
throughout the project, undoubtedly made a significant contribution to the
success of the project.
The paper will be divided into six sections: first, the origin of
the Puebla Project and its objectives; second, its various activities as
they evolved; third, training; fourth, what was accomplished; fifth, implications which might be applied to other areas of the world; finally, some
evaluative observations.
Origin and Objectives
SBy 1967, scie ntists at the International Center for the Improvement of
Maize and Wheat (CIMMVYT) were struck by the fact improved maize varieties and
agronomic production practices which had been developed on the research
stations were not being utilized widely by farmers in Mexico. In particular,

- 4
small farmers did not seem to be picking up the new technology readily. The question was, why? The scientists decided to determine if, with available varieties, yields in the region could be raised substantially; if so, whether with the available varieties new practices would be more profitable to use; and, if so, to what extent farmers would change and what factors would influence their decisions. Later after the Puebla Project had met with some success, two more objectives were added:*
a) To develop an efficient methodology for promoting a rapid increase in maize production.
b) To train national leaders for maize promotion programs and to assist them in initiating and operating their programs.
The project area chosen is in the state of Puebla, about two hours drive from Mexico City. In the area approximately 43,000 families on roughly 116,000 hectares of cropland (about 2-1/2 ha per family), about three-quarters in maize, farmed land under rained conditions. Average maize yields were only 1.3 tons per hectare and static. There was a good system of roads for access to the area. Agricultural service agencies, including credit, fertilizer supply, crop insurance, and price support, were already existing although not widely serving the small farmers. A land reform program had been carried out previously.
The project was initiated as a learning process (not a demonstration project). It tested a planned approach, based initially on what was known,
In addition the project staff itself added an internal challenge to themselves, i.e., to double the yields of maize within a five-year period.
This was not a formal objective of the project although its achievement,
of course, would have been welcomed.

with built-in mechanisms for learning and subsequent adjustments. CINMYT maintained overall control and coordination, but at the same time it tried to involve the appropriate Mexican institutions as collaborators.
The professional team consisted of ten to eleven professionals with university degrees. The composition of the team changed somewhat throughout the project. For example, a corn breeder was involved during the early period, while a technical assistance agent was not added till the second year. However, basically the team has consisted of a coordinator, four specialists in agronomic research (including a maize varietal improvement specialist), an evaluation specialist, and five technical assistance agents. A staff of approximately twenty-five local farmers, selected and trained for specific jobs, has complemented the professional staff. Several staff members of CIMMYT and the Graduate College at Chapingo provided technical assistance to the professional project staff. The project staff were paid salaries competitive with other civil service positions. There were strong professional incentives in the form of satisfaction of seeing the results of the work in the project area and its spread to other states in Mexico and other countries in Latin America.
The total cost of the project for the first seven years was $925,o145, of which $559,851 was provided by The Rockefeller Foundation, $332,737 by CIMMYT, and $32,1457 by the Graduate College and various institutions in Puebla. This level of financing was generally adequate to permit the level of activity which was desirable. Administrative procedures were decided internally and "red tape" was minimized.
The cost of the project, the staffing requirements, and the method of operation were designed in such a way that it wo Iuld be practical for the

government to assume full responsibility for the project as well as extend it to other areas if it were successful.
The Project Components
The basic components of the Puebla Project can be divided into five main activity areas, some with subactivities:
a. Production research: local production research emphasizing maize production practices (including soils, planting density, fertilizers, and sowing date) and variety testing.
b. Technical assistance: communication of agronomic information to farmers, agricultural leaders, agricultural institutions, and government officials.
c. Coordination of all activities (including the above.) impinging directly on maize production:
(1) Production credit.
(2) Agronomic inputs.
(3) Relationships between input costs and crop values.
(4) Markets.
(5) Crop insurance.
d. Institutional development ,e. Socioeconomic evaluation
We will review activities in each of these main areas. Production Research
The project area on which maize is grown is located between 2,150 and
2,700 meters above sea level in a valley bordered by three statuesque volcanoes.

The soils are variable in quality. The climate is temperate with frosts occurring mostly during the winter months (but occurring in all months except July). Frequent hailstorms occur in late summer and early fall. Average rainfall varies from 777 to 863 millimeters, 94 percent of which falls between April and October which is the maize growing season.
The initial task of the project staff was to discover technology
suitable to help farmers increase their yields under these physical conditions. They first reviewed the information available, both from previous research efforts as well as by interviewing farmers, to establish research priorities. They then initiated field tests of a number of varieties with different levels of management with respect to fertilizer use and timing, plant population, and weed control to establish the recommended levels of production inputs. At the sane time they initiated a breeding program to attempt to develop varieties better suited to the environmental conditions than those then available.
At the end of the first year of field tests it was found that none of the improved maize varieties or hybrids performed appreciably better than certain unique local varieties which had been selected by farmers in an area irrigated with sewage waters from Puebla.
Thereafter, project research has focused on deriving recommendations for varieties of maize already in use, taking into account variability in soil morphology, planting dates, elevation above sea level, and moisture availability. In 1967, the first year, 27 trials were located over the project area (ecological knowledge by which to insure optimal location of the trials was not then developed). Based on these results, the project came out w ith new recommendations that increased the rate of nitrogen

(from 80 kg/ha to 130 kg/ha) and increased the plant population (from 40,000 plants/ha to 50,000 plants/ha).
Beginning in 1968, planting date, timing of nitrogen application, and plant population were additional variables studied in the experiments. A soil morphologist spent ten days interpreting soil differences. High plant density was found to be very important to obtaining high yields under good conditions (but mayresult in lower yields if weather is not favorable). Recommendations were changed for a few areas where soil conditions were limiting. Recommendations were also made for different planting dates because the length of the growing season was found to have a very important influence on yield. Drought and hail were found to be the major sources of risk. Late frost could also be a factor. In all, recommended packages of production practices were developed for 16 producing conditions in the area. Experimental results indicated that project recommendations could increase yields by an average of 1.75 tons/ha on some areas and an average of over 1 ton/ha over the whole area, compared to the traditional practices, and by an average of over 0.5 ton/ha above yields achieved with the previous recommendations.
The advisors in research were sensitive to social science considerations in deriving the recommendations. Each of the 16 recommendations was made at two levels, one to reflect limited capital and the other to reflect unlimited capital.* Care was taken in making these recommendations
*The limited capital recommendations were selected intuitively rather than based on rigorous experimental criteria. On the average, they called for lower levels of nitrogen and phosphorus and lower plant populations than
previously recommended by the National Agricultural Research Institute
(INIA). However, the limited capital recommendations were adjusted specifically to the conditions of each of the 16 producing systems. In practice, the limited capital recommendations corresponded closely to the
factor combinations that maximized the rate of return on fertilizer.

to ensure that the farmers had a good chance of making at least as much income with the new recommendations as with previous practices in an unfavorable year.
Recent research has concentrated more attention on combinations of corn and beans which is an important cropping system in the project area. When beans are associated with maize, maize yields drop. However, because the bean price is five times (currently) that of maize, the farmer growing beans with his maize can end up with increased income. Fertilization with chicken manure increases yields of maize and beans (especially); it also increases protein content of maize and beans. The maize stalks and bean vines provide useful forage for animal feed. Maize and beans have complementary amino acid balances. Thus, the combination significantly increases both the calories and protein available from an acre compared to previous practices. Furthermore, because the combination must be harvested by hand, it also increases employment.
It is not possible to overemphasize the contribution that this strong research component, which is absent in most projects, made to the success of the Puebla Project.
Technical Assistance
Recognizing that identification of superior technology applicable to
specific environmental conditions is only part of the job, the next and concurrent task of the project staff was to develop a mechanism to reach as many of the farmers as possible in order that they would adopt the recommended technologies to increase on farm yields and incomes.
The estimated 43,000 farm operators had an average of 5.9 members per family. Average farm size was 2.7 cultivated hectares (90 percent of farms

with less than 5 h ectares), consisting of' 3.8 separate parcels. Almost all farmers (over 99 percent) owned their own land or were ejiditarios, farmers who have free possession of their land for life.
The 157 villages of the area are connected by roads which are passable during most of the year. Almost four-fifths of the farmers could read or write, although most only at minimal levels (average school attendance was 2.24 years). Electricity reached 63 percent of farmers, 114 percent had potable water, 61 percent had a radio, and 13 percent had a television set.* In 1967 almost all of the farmers knew about chemical fertilizers and approximately two-thirds were using fertilizers, although often at low levels. Over half of the farmers knew of hybrid maize but less than one percent were using it in 1967.
The agrarian reforms of the 1920's hiad been accompanied by political and organizational changes at the village level. However, many promises were unfilled and many previous private efforts had resulted in frustration. The villagers had become very competitive among themselves. Furthermore, they were generally suspicious of outsiders and especially mistrustful of government and organizations associated with it. After several decades of experience, the campesinos had-come to the-conclusion that they were better off without government services. Often inputs and cash supplied did not arrive on time. Because the inputs and credit were late, the cash credit was often expended for nonproductive purposes and therefore the campesinos had difficulty in repaying it. The inputs received often were not suited to the ecological conditions in which they were to be used. Corruption
*The last two figures are from 1970 data.

among field agents distributing the inputs and credit made their use even less attractive.
The problem of mistrust had to be endured and overcome since the Puebla Project had no alternative in the sense that eventually the government had to take over the program. For this reason, the staff felt that they had to use face-to-face communication at the beginning in order to establish rapport with the campesinos. They began using films, radio, and printed materials only by 1.969. Continuity of staff personnel in this early period, was especially important in establishing trust.
First, through repeated visits to the villages the staff studied the recognized communications systems already in existence, including both the formal system operating through the power structure and the informal systems operating through friends and neighbors. Direct information was given to everyone, including women and children, who attended the meetings held in various villages. Several things were accomplished in this process; first, the farmers felt involved; second, the staff were able to-identify what the people wanted and what problems were most important to the community; and third, the community leaders were identified. It became obvious that participation had to be on a voluntary basis to have lasting effect.
From this process, the key role of credit as a motivating force was identified. The rate of loan repayments had been very low and banks were reluctant to lend. It was virtually impossible for an individual alone to get credit and, at the low level of resources owned by the campesinos, without credit the campesinos were unable to finance the improvements needed to raise yields. Therefore, the project leaders made known to the

farmers that credit, fertilizer, seeds, and assistance were available to those who participated in the program at their own choice.
Based on their investigations, the project staff realized that the
campesinos would not voluntarily use the services provided by the government agencies. Furthermore, the public banks were not prepared in 1968 to participate in the program. Therefore, an alternative was sought to permit those campesinos who were willing to use the new technology, but who did not have the resources necessary to purchase the required inputs, to get credit. An arrangement was finally worked out with a private fertilizer dealer. In 1968, the first year, 103 farmers volunteered to participate in the project. Eighty-five of them required credit; the remainder had adequate financial resources but wanted technical assistance. In that first year, yields of the participants averaged 4.0 tons/ha compared to
2.1 tons/ha for the entire project area.*
In.1969 three major changes were made. First, the project area was
divided into four (and eventually into five) zones. A technical assistance agent was assigned to each zone (actually only four of the five zones had technical assistance agents in the first year). Each technical assistance agent had one or two nonprofessional assistants, local trained farmers, who assisted in the work with the farmers. Second, in 1969 the number of participants rose to 2,561 and the official credit banks joined the fertilizer distributor in offering credit to the farmers. Third, because the number of farmers had become too large for individual technical assistance agents, the participating farmers were organized into groups.
*No comparison of the yields attained by "participants" and "nonparticipants" prior to the start of the project is available.

- 13
Although the need for some form of organization which would permit the
technical assistance agents to work with larger numbers of farmers was
recognized from the beginning, the specific form or forms it would take was
0 not clear. The planners decided to let the eventual organization evolve
out of the experience of the campesinos as the needs for cooperative action became obvious to them. The initial participants recognized the weaknesses of individual action. At an early stage, they spontaneously suggested that
cooperation would have advantages in arranging transport for fertilizer.
'They organized in groups for this purpose. As recognition grew that group
action was useful for other purposes also, the actions of the groups expanded to meet these broader needs.
The makeup and specific rules of operation of the groups differ somewhat depending on with which of the credit institutions they are associated.
However, the basic principles are similar. A combination of responsibility
and social pressure is the key element in making the group an effective
element of change. In the case of solidarity groups, only one member has
to have a clear land title to qualify for loans from the Agricultural Bank.
The group also generally takes responsibility for repayment of credit by individual members. It works out solutions if individual members can not
or will not repay. In some cases a group will take over payments until
the member can assume his load. In other cases, individuals are excluded
from future partic ipation. Some members of communities are restricted from groups initially because they are not considered to be responsible
The groups serve to reduce transaction costs of borrowing. Finally, the
groups, made up of many men rather than one individual, can exert pressures
on the institutions to obtain rightful access to services. They have

received strong political support from the President of Mexico who appreciates the need for organization to achieve desired goals.
For various reasons including the lack of managerial expertise among
the rural leaders which has contributed to a history of corruption, cooperatives do not extend in any strength to the rural areas of Mexico. However, it is felt that a few groups are prepared for a more formal and sophisticated type of organization now.
As the project gained momentum, the staff has come to learn that the farmers are better prepared to follow than the staff sometimes is to lead. Coordination of Activities for Favorable Change in Agricultural Institutions
The participation of agricultural service institutions to supply production inputs including chemical fertilizers, to provide credit, and to ensure a reliable market for outputs is essential to the success of any rural development program. In the case of Puebla, the necessary institutions did already exist, but were not fulfilling the needs of the small farmers. Initially they resisted becoming directly involved in the project. The third component of the strategy, therefore, was to involve them gradually, using favorable program experience and pressures from the farmer groups to achieve change.
Four main credit agencies operate in the area with differences in interest rates charged, requirements to qualify for loans, and other procedures. These practices range from 9 percent yearly interest with no individual security (but a government endorsement guaranteeing the overall operation and a requirement that farmers receive technical assistance from the Puebla Project staff) of the private fertilizer dealer to 10.5 percent interest plus 1 percent service charge (the total actually comes to almost 13 percent

because of administrative procedures) with group guarantees (which can be supported by land title of a single member) and a requirement to purchase crop insurance (required by all official agencies) of the official Agricultural Bank. One official bank deals primarily with the ejidarios,.taking into account the special conditions under which they operate. A third official bank is gradually withdrawing because of other interests.
The amount of agricultural credit distributed to the campesinos in
the project area increased from 1,330,598 pesos in 1968 to over 12,500,000 pesos in 1972. The rate of loan repayment has also increased significantly. In 1972 the private dealer (with 15.8 percent of credit) reported 98.5%; the Agricultural Bank (with 39.3 percent) 94%; the Ejido Bank (with 40.9 percent) 90%; and the Banco Agropecuario del Sur (with a declining 4.1 percent of credit) only 50% repayment.
Problems still remain, however. The private dealer, who is limited to
interest rates lower than the subsidized official banks, is not making enough profit to give him incentive to expand. Generally, loan processing is often lengthy. Personnel turnover is high. Also, legal and communications (for example, the farmers do not understand the crop insurance) problems exist.
Fertilizer policy was changed in 1971. Previously, individual distributors purchased fertilizer from the producer for sale, at whatever price they could get to whomever they favored, in the villages. Now fertilizer is distributed in the town of Puebla and at two other points directly to farmers at a single (unsubsidized) official price. Usually the fertilizer is purchased by the groups who then arrange for transport to the villages.

- 16
There has been a significant change in the type of fertilizer used in the project area during the past few years. Initially 10-8-4 (which identifies percentages of nitrogen, phosphate, and potash in a mixture) was distributed. However, the project staff discovered through their experiments that there was no production response to potash in four of the five zones and that in one zone, there was response only to nitrogen. As a result of communicating this information to the farmer groups who applied pressure on the fertilizer distributor, ammonium sulfate (20.5 percent N) and superphosphate (20 percent P 205 ) are now the primary fertilizers used in the area.
Floor prices are enforced by the government price control agency (Conosupo) which purchases maize for announced prices at 14 warehouses over the area. No quality discount is given up to 14 percent moisture. This agency has greatly simplified procedures but the farmers still think the dealings are too complicated.*
Crop insurance, which is obligatory to all who receive official credit, has been a consistent problem. Insurance can be denied if the campesinos do not carry out recommended practices. Furthermore, the campesinos have felt that they already insure themselves by planting in different parcels and at different times. The criterion for payment of claims has been the source of repeated dissatisfaction. As a result of the pressure from the
* Until 1973 the floor price was above world prices. However, in practice
the bulk of the maize was sold by the farmers at below-Conosupo prices.
Conosupo often would not purchase small lots, would discount for quality, and sometimes, when storage space was not available, would not make purchases at all. It is reported that this situation has improved, in part
because of the existence of the groups but more importantly because the
world price is above the floor price.

- 17
groups, new procedures for reimbursement have been worked out whereby the farmers stand a better chance of receiving compensation when damage is limited to one or two parcels. The premiums paid by organized farmers using the project recommendations will be reduced by half in future years.
The Puebla Project staff played important catalytic roles in bringing the campesinos and the service organizations into closer contact and better understanding. They persuaded the campesinos, who had had little .previous experience-with banks, to follow the rules of the banks. They persuaded the banks, who were not set up to deal efficiently with many small-units, to simplify their procedures and to pay their field people better. They urged upon the fertilizer company the importance of supplying materials on time.
The project had very fortunate leadership in that it has had three
excellent coordinators. These coordinators have provided both intellectual as well as administrative leadership. Coordination has been provided through weekly meetings, often'held at night, of the entire technical staff to discuss problems as they arise. Staff were chosen, in part, on-their ability to work together. The relatively small number of staff, who were both well trained' and highly motivated, undoubtedly has facilitated the coordination of activities, including the ability to change programs based on the feedback provided by the evaluation unit.

- 18
Institutional Development*
Improvement in institutional performance was an essential component of the project which was discovered within the project experience. The leadership of the project came from persons who were aware of the institutional elements through previous experience in rural areas of Mexico and who had struggled with organizational issues long before they undertook the Puebla effort. As a result the technicians consciously and deliberately helped the campesinos to organize. Furthermore, the organized groups, along with their advisors, recognized that their task was not just to grow more maize but also to overhaul the banking system and improve services. The campesinos themselves insisted on some of the institutional changes. The groups were able to apply pressures to change the credit system because they represented a certain collective discipline which does not exist without organization. The technical assistance agents were not trained in this area and at one time in the project, displayed their initial lack of organizational competence by actually suspending critical parts of the program only to reestablish the activities later. The first campesinos to organize were those who had ejido experience. Without this model, the project might not have gotten its ideas together quickly enough or successful enough for the groups to succeed. Together the campesinos and the technical assistance agents learned as they went along "by doing." Still, the communicators are only beginning to make headway in promoting more advanced campesino organizations that can
* This component, not recognized explicitly in the original project plan but in fact implicit as the project progressed, is included at the suggestion of Herman Felstehausen. The discussion has benefited from correspondence with Felstehausen and the thesis, cited previously, of Heliodoro Diaz.

attack greater problems. The institutional development component is very important in distinguishing the Puebla Project from numerous other projects which have been tried throughout the world. Socioeconomic Evaluation
It was agreed to establish an evaluation unit within the project.
The team began by checking Mexican government data; it was decided these data were not sufficiently accurate so the staff collected its own. They began by using aerial photos; next identified 100 hectare segments and the parcels within each; and then took a 12 percent sample of owners in each segment (25 segments in all). They interviewed each person selected from a random sample of 10 within each segment; 251 questionnaires were taken in 1967-68. A comprehensive list of questions was included. A second survey was carried out in 1971. One limitation of the first two surveys was that they included only those who farmed land; landless laborers were not included. A third survey covering a wider coverage of questions and people may be carried out in the summer of 1975.
The surveys have played a useful role in measuring changes (which are reported in the next section) during the first four years of the project. However, the surveys carried out by evaluation unit did not play the feedback role which was originally envisaged. Two reasons probably accounted most for this.
First, formal zed social science surveys, especially if they cover a statistically designed sample and include a broad range of variables needed to give a comprehensive evaluation of change, tend to be both expensive and time-consuming. Often the results are not available quickly enough to influence operations. Second, surveys which are designed to evaluate change

often do not contain the information identifying obstacles limiting farmer use of new technology which would assist in developing means to overcome such obstacles. To provide such information requires skill in design, an understanding of the general problem, and quick analysis. However, the surveys have been useful in evaluating change over longer periods of time as a measure of project progress.
Special studies of less sophisticated design which were more quickly compiled were carried out in 1968 and 1973 to determine the level of performance of the service organizations and to identify the components of the fertilizer distribution network. These studies, especially the first, did have some role in making the project staff more aware of limiting factors. Finally, because the staff anticipated that the farmers would object to the staff harvesting ears in the field, the evaluationstaff derived an indirect method of measuring maize yields. This was very useful in measuring the year-to-year results of recommendations under the prevailing conditions.
It should be reemphasized that while the survey studies made a relatively limited contribution to the project implementation, evaluation and feedback was built into the project on a continuing basis. The staff utilized direct personal observation and frequent contact with farmers and service institution personnel to derive their data. The weekly informal staff meetings (especially active during the early stages of the project) provided an effective forum to display and reflect on the observations and make the indicated changes.

A training component of the project was built on the premise-that all disciplines must contribute cooperatively to achievement of common goals, that capacities for judgment and professional competence are the key ingredients needed to insure transferability of experience from one project to another, and that a pilot level focus is the best way of learning the "nuts and bolts" of what makes a project succeed or fail.
The basic course lasts 6 to 9 months. Three areas of training
production research (emphasizing research for generating packages of crop production practices), coordination and technical assistance, and evaluation were emphasized. The project staff prepared a training manual for each of the three areas of specialization (retaining some elements of all three areas in each manual).
Plan Puebla is now also participating in a training program in which course work and degrees are granted from Chapingo with thesis research work in Puebla.
Sixty-six persons were trained during 1967-73 from Mexico (41) and
four other Latin American countries, i.e., Colombia (16), Peru (5), Ecuador (2), and Honduras (2). Twenty-two of these trainees also carried out academic programs at the Graduate College at Chapingo. The trainees are presently participating in the operation of ten regional production programs in four countries.
The "official" definition of a participant in the Plan Puebla program is a farmer who is in a credit list. Based on this definition the following table shows the growth in direct Plan Puebla coverage during the first six years.

Estimates of Plan Puebla Participation 1968 1969 1970 1971 1972 1973
Number of Cooperators 103 2,561 4,833 5,240 6,202 7,194
Percent of Total* 0.2 5.9 11.1 12.1 14.3 16.6
Hectares in Plan 76 5,838 12,601 14,438 17,533 20,604
Percent of Total** 0.1 7.3 15.8 18.0 21.9 25.8
It must be recognized that this is a very restrictive definition of participation. Not all farmers on the credit lists are necessarily applying the recommended technology and many other farmers, some of whom do not require credit, undoubtedly are following the project recommendations. Farmers are cautious and experiment themselves. They usually only partially adopt recommended production practices, they will adopt some parts of recommendations more readily than others, and they tend to use new technologies initially on only a portion of their lands. In order to get a clearer picture of the adoption process, farmers in the area were divided into the following categories for three practices.
Levels of Adoption
N kg/ha P 20 5kg/ha Density plants/ha
Low 0 -50 0 -20 0- 30,000
Med 51 80 21 30 30,000 40,000
High 8o+ 30+ 40,000+
The upper limits for the "low" levels of adoption correspond approximately to what the better farmers were using in 1967. The lower limits of
*Based on a total of 43,300 farmers.
*Based on a total of 80,000 ha of maize.

the "high" levels of adoption correspond to the lowest rates of the inputs that are recommended presently in the area.
Analysis reveals a significant movement of farmers during the project period from "low" to "high" levels of adoption in each category.
Levels of Adoption of Each Recommendation
% of all farmers in area 1967 1968 1970 1972
High 7 33 45
Int 11 14 14
Low 82 53 41
P 2 0 5
High 24 38 44
Int 8 9 9
Low 69 52 47
Plant Density
High 14 25 39
Int 35 31 34
Low 51 44 27
Average amounts of nitrogen, phosphorus, and plants used in maize
plantings increased by 129, 93, and 10 percents respectively from 1967 to 1972.* There appear to be several possible reasons for the relatively small apparent change in plant densities including uncertainties of availability of fertilizer (which is needed to support high densities) at. planting time, the belief that corn will withstand drought (which is a major concern of farmers) better at lower stands, and the farmers' desire to increase
It should be noted that not all of this increase in fertilizer use was due
to the project. Between 1967 and 1971, the actual maize/fertilizer price
ratios for nitrogen and phosphates declined by 20 percent, thereby in
itself giving a greater economic incentive to use more chemical fertilizer.

production of large ears which they feel is best achieved with relatively low plant densities. The above table would suggest that the changes in plant densities might be greater than the averages suggest.
The following data tend to confirm the observation that the farmers
themselves experimented with recommendations, often adopting them in stages rather than as a complete package.
Level of Adoption N=200 N=200
of Three Practices All Farmers Farms in Credit List
High all 3 10 20
High for 2; intermediate for 1 11 28
High for 2; low for 1 19 27
High for 1; intermediate for 2 6
High for 1; intermediate for 1;
low for 1 L5 8
Total 58 89
One important implication of these data is that it is very difficult to analyze participation quantitatively, at least in any simple manner. Since the only measurable definition of participation was not meaningful, comparisons of performance between "participants" and "nonparticipants" generally were not carried out.
The main purpose of the project was to increase maize yields. Maize yields in Puebla (over whole area) are estimated to have risen as follows:

- 25
Average Maize Yields in Puebla
Year General Average kg/ha
1967 1,330
1968 2,14o
1969 1,832
1970 1,962
1971 1,927
1972 2,499
1967 was a poor rainfall year. 1968 and 1972 were considered roughly
comparable weather years. (1968 was actually considered to be slightly
better.) Two different methods were used to "deweatherize" the production
data to estimate the influence of new practices in increasing yields during
the 1968 to 1972 period.* One method indicated a yield increase of 24.2
* The first method was to utilize the average yields of 8 to 12 experimental plots using practices approximating those of the farmers of the area (50 kg/ha of N, 25 kg/ha of P 0 and 30,000 plants/ha) in each of the years from 1968 to 1972. The second as to utilize the average yields of the upper third of the farmers in the credit lists each year who were assumed to be using the recommended practices accurately and reasonably constantly over the years. In both cases, the average yield of all farmers was then deweatherized to give an estimate of the increase attributable to the new technology. Each method has deficiencies. In the first, the number of experiments was not large enough to adequately sample the area nor were they distributed over the project area in a way to give proper weight to the 16 producing systems. (They were not designed for this purpose.) In the second, closer supervision may have led to higher levels of usage of the new technology in the first year, 1968. Also, the upper third of farmers in credit lists does not provide a representative sample of the producing conditions in the Puebla area. The two methods gave similar weather factors for 1969 through 1971 but differed considerably for 1972. A rough averaging of the two methods gives an annual yield increase of 7.5 percent.
CIMMYT also made calculations by comparing the average production of the last two years, 1972 and 1973, with the average of the first two years, 1967 and 1968. They concluded that average yields rose from 1.7 tons/ha to 2.5 tons/ha, an increase of 47 percent over 7 years or 6.7 percent per year.

- 26
percent; the other an increase of 4+1.9 percent between 1968 and 1972. The
Plan Puebla staff feel that a 30 percent yield increase between 1968 and
1972 is a reasonable compromise.
The following data indicate how these production gains were translated
into increased income for the farm families of Puebla.
Changes in Income and Welfare*
Average Total (cash & imputed) Net Income of Farm Families (sample from total people of area who farm land; U.S. dollars adjusted for inflation)
1967 1970
Total family income us$ 666.80 us$ 825.52
% from crops 30.4+ 35.5
($ 202.57) ($ 293.06)
% from animals 28.14 30.0
($ 189.37) ($ 24+7.66)
.from off-farm activities 4o0.7 27.7
($ 271.53) ($ 228.67)
% from other activities 0.5 6.8
($ 3.33) ($ 56.13)
*The data on changes in income and welfare should be used with caution for three reasons. First, the samples for 1967 and 1970 are not strictly comparable. It is felt that the direction of change, although not necessarily
the absolute amounts, is a reasonably accurate indicator. Second, the
breadth and depth of the project progress has not been fully reflected in
*the data collected. The survey data do not pick up income by nonfarmers. This is especially important because towns like Puebla generate a lot of
economic activity which must contribute significantly to the welfare of
the rural people, especially those who do not farm land. Furthermore, the
data as presented thus far do not give a complete picture regarding comparative benefits of participants compared to nonparticipants. It was
agreed that the available survey data, while being useful for some purposes,
still left many unanswered questions in part because the questions were not asked when the surveys were designed. In order to improve the situation, a third survey is planned for the summer of 1975 which hopefully will give a more comprehensive and more accurate picture of change in the project
area and provide answers to many questions. Third, obviously, not all the income gains in the project area can be attributed directly to the Puebla
Project. Maize yields nationally have been rising at between one to two
percent per year. Per capita income in Mexico has been rising at about
three percent per year although these gains have not generally been extending to the dryland, small farmer areas.

- 2T
The income from crop production is mostly from maize (61.5 percent in 1970). Net income from crop production was estimated to have increased by 44.7 percent (from $202-71 to $293.06) over this three-year period (it should be noted that 1967 was a poor year for maize production while 1970 was below average).
These data indicate a surprisingly high percentage (at least it was surprising to many of the reviewers who did not know the area well).of income from animal production. The reason for this apparent result is that a large part (66.5 percent) of the gross change in animal income comes from milk and somehow four large dairy producers (who represented 81 percent of gross income from milk) got into the sample. Still, animals are significant. They are used primarily for savings and income rather than for consumption. Animals eat corn stalks, bean vines, and grass (pigs and chickens will eat grain) so the cost of maintaining animals is not high. The rural banking system for savings is not very well developed. Squirreling money at home loses value to inflation. It would appear that income from animals could be increased much more with new technical assistance. Beef, milk, and swine seem to pose best opportunities. Marketing poses a big problem for eggs (although it is not obvious why egg marketing is any more difficult than milk marketing).
Limited inquiries do not indicate that the income from off-farm activities in the rural area around Puebla would have been expected to decline; quite the contrary. The apparent reason for this result in the reported data is that survey figures account only for those who farm land. People who work in brick factories, etc. (which seems to be increasing considerably) may be nonland owners. Therefore, the growth in incomes from this activity probably

was not picked up by the survey. One other reason that the percentage off-farm income for farm families appears to be falling is that on-farm income opportunities may be even more attractive with new agricultural techniques. The campesino prefers to farm his own land rather than work in the city or in brickyards if his income from farming is reasonably close to what he can earn elsewhere.
Distribution of Annual Family Income* Ranges in Income (US$) % of families in range
1967 1972
$400 or less 55.8 43.54oi-6oo 12.3 20.1 +
601-1000 16.3 18.0 +
1001-2000 10.0 11.3 +
2000 + 5.6 7.1 +
These data suggest that the increases in farm family income appears to have been accompanied by a desirable distribution of income. More information is needed before a positive conclusion can be made on this.
Information on employment changes is meager. It appears that the total number of days of off-farm work for farm families remained nearly constant during the three-year period. (However, due to increases in popu* nation, the average number of days worked off-farm per worker decreased.) There is no reliable information on changes in on-farm employment. The labor "requirements" for the traditional and recommended maize production practices are estimated as follows:
*Family income in 1970 was adjusted to 1967 prices.

Labor Requirement for the Production of' Maize
traditional planting planting using recon.
1 hectare for one crop season (man/days) tech (man/days)
land preparation 9.1 9.1
planting 4.3 7.7
cultivations 8.6 10.2
harvest 18.6 25.7
Total 0652.7
it would be interesting to know on what the increased disposable income is being spent. This information has not been collected comprehensively as yet. There is, however, some information on changes in food consumption.
Consumption of Several Foods by One Member of Family
Every 1-3 days Every 4-7 days Never
1967 1970 1967 1970 1967 1970
Fish 0.8 0.8 3.2 11.3 13.9 4.2
Beef or pork 8.4 9.6 43.0 43.9 3.2 2.9
Milk 29.1 27.6 7.6 7.9 38.2 43.1
Chicken .14 1.7 5.6 14.6 17.5 12.5.
Eggs 29.1 59.4 25.9 32.6 9.2 2.5
Wheatbread 33.5 38.5 35.4 30.5 8.4 13.8
Fruit 11.6 30.5 32.7 37.2 9.0 5.0
Vegetables 14,4 34.3 31.5 38.5 12.0 9.2
Rice 16.8 30.6 44.2 46.9 4.4 5.0
Although the time period is short, these data would suggest that diet (especially eggs, fruit, and vegetables) has improved. Changes in relative prices as well as higher incomes probably had some influence on this pattern.
Improvements in farm homes between 1967 and 1970 were reported by the
following percentages of farmers: 5.1 percent of the sample of farm families changed the floor (from earth to concrete, brick, or mosaic), 13.4 percent added another room, 6.7 percent painted the walls, and 4.2 percent repaired the roof.

- 30
There are also some encouraging signs of improvement in basic rural
facilities. These nonagricultural changes, of course, cannot be attributed to the Puebla Project.
Percentage of Families Who Have Access to: 1967 1970
Electricity 63 77
Potable water 14 21
Plumbing 6 6
No data were collected in the first two surveys which measure degree of change in such other areas as education or health. An anecdotal bit of information is revealing however: one campesino reported that the increased maize yields and incomes gave the townspeople enough money to improve the road from their village. This permitted easier exit of village products and, as a result, a brick industry was begun which now employs half of the labor force in the village. One of the early uses of the bricks was to build a school for the village to which the state supplied a teacher. This apparently was not an isolated incident.
The Puebla Project staff did attempt to evaluate changes in attitudes of the farmers. Forty-four percent of the sample of all farmers in the area stated that they had increased their maize production between 1967 and 1970, and four-fifths of these attributed their success either directly or indirectly to the Puebla Project. This, in turn, had a favorable influence on their attitudes to farming. In response to the question of what activity they would engage in if they suddenly were to receive a greater amount of income than they were currently getting, the number who said they would continue to farm

- 31
and improve their production practices increased from 53 percent in 1967 to 73 percent in 1970.
Perhaps just as importantly, based on the credibility established by the agricultural progress of the Puebla Project, there is growing pressure for the project staff to move into new areas such as animal production, horticulture, education, and health services. Long-term loans are being serviced through groups to drill deepwells for irrigation in order to grow higher income crops like alfalfa and vegetables. Another group has cooperatively purchased a tractor. The groups are providing a means by which the campesinos can increasingly use their newly discovered cooperative power to pressure institutions for constructive changes in services beneficial to them.
Benefit/cost analysis was carried out by project staff using alternative assumptions regarding who benefits, how much, for how long, and under what opportunity and costs of resources. Counting only the increase in income from maize production of participants in credit lists, i.e., the direct benefits, the benefit/cost ratio directly attributable to the project (discounted at 14 percent interest rate) was 2.54. When additional account was taken of nonparticipants in credit lists who also benefited from the project, i.e., the indirect benefits, the benefit/cost ratio was raised to 4.03.* No calculations were made to include the intangible benefits accruing to the farmers. It is recognized that benefit/cost analysis is only a partial
Alternative calculations also were made using several labor cost assumptions. It was concluded that the most reasonable assumption was existence of seasonal unemployment, i.e., zero opportunity cost, except at harvest. Based on this assumption, the "most plausible" benefit/cost ratio is

measure of success. However, even acknowledging the limitations of this analysis, this magnitude of return on investment from the project aoes appear to be an impressive achievement.
Perhaps the ultimate test of the influence of the project is the extent to which it has been imitated in other areas. In March 1974+, financial support was assumed by the Mexican Ministry of Agriculture. The Mexican states of Mexico and Tlaxcala, as well as the countries of Colombia, Peru, and Honduras have adopted similar projects. The Puebla Project also has had indirect influences on rural development projects in the Philippines (Masagana 99) and in Nigeria (near Ibadan).
Implications for Other Projects
Technology for maize production in this area did not automatically flow to the farmers as it did with wheat in northern Mexico or in India. This situation is representative of a large part of the low-income world. The project was initiated as a learning process. Many specific lessons were learned. The more general implications of Plan Puebla for similar situations elsewhere include the following:
1. Feasibility studies including both statistically designed surveys as
well as in-depth interviews and personal observations should be carried
out at the very first stage to identify the major activity components in the region, the available technology, the agro-climatic conditions,
and the socioeconomic structure of the people and to provide a baseline
survey against which progress can be measured.

2. Based on these studies, clearly defined goals should be established
against which progress can be measured.
a. The goals should be as simple in focus as feasible.
b. A time phasing for activities and progress should be worked out.
3. Budget considerations should include:
a. A budget that is adequate both in amount and ease of administration.
b. If using external resources, the program should be funded at a level
which the government eventually can assume.
4. Staffing considerations should include:
a. A capable, highly motivated, well-traihed, interdisciplinary, analytical staff covering production research, technical assistance to the farmers, and socioeconomic evaluation organized into a single team
(there should be no escape from team responsibility for achieving
the goals of the project).
b. Incentives to make the work professionally and personally rewarding
to the staff.
c. Strong leadership.
d. Continuity of operation assured far enough in advance so that a
reasonable time is provided for achieving the objectives.
All possible efforts should be made to bring in and coordinate all the activities impinging directly on crop production, specifically of small
farmers, including the following:

a. Production research which is related to the national research program
(rather than an autonomous effort) and which includes on-farm research on a wide range of soil, fertilizer, and climatic conditions
within the project area and development of management techniques
which can promise the farmer substantially higher yields and profits
than he is currently obtaining.
b. Effective communication of agronomic information to farmers, agricultural leaders, agricultural institutions, and government officials the technical assistance agents must have an adequate knowledge
of technology and have confidence and be expert themselves in the
practices they are recommending.
c. Adequate production credit at reasonable rates of interest with
reasonable procedures.
d. Easily accessible and adequate amounts of agronomic inputs when
e. Accessible markets with stable crop prices.
f. Favorable relationships between input costs and crop values.
Chances for success, including longer-term continuity of program, are
strengthened if these activities are built into the ongoing governmental
process from the beginning, i.e., if political support is achieved.
6. Some process must be established to multiply the technical assistance
capability of a limited number of trained people to a large number of
participants. In the case of the Puebla Project, farmer groups served
very effectively in this objective and, in addition, served as a very effective "grass roots" force in accelerating change in practices of
the institutions serving agriculture.

- 35
7. Certain functions such as some aspects of training can be centralized while others such as field experiments and demonstrations must be carried
out close to the farmer, preferably on farms of participants. Dividing the area into zones may be helpful to establish areas of responsibility
for technicians. The criteria for selecting zone size should include
nature of soils, other climatic factors, social and political organization of the population, degree of mobility of the team, and amount of
financial resources available. However, the division of responsibilities
is carried out, the researcher must not be insulated from the farmer,
i.e., should not be separated from seeing the field results of his work.
8. Continuing in-house evaluation is extremely important to provide feedback so that project activities are responsive to changes within the
project in pursuit of the designated goals (or so that goals can be adjusted if needed).
Concluding Observations
The Puebla Project was initiated as an experiment in 1967 with limited, but important, goals. The project followed a planned approach based initially on what was already known with built-in mechanisms for learning and subsequent readjustment. A methodology for promoting a substantial increase in maize production (the first objective in the original project statement) was developed. The new practices did not represent radical changes from'what farmers were doing previously; the new recommendations utilized the same seeds but with more nitrogen and phosphate fertilizer and higher plant populations. Average maize yields for the whole project area were estimated to have increased by about 30 percent by 1970. Welfare in the area, as measured by

- 36
increase in income and such other evidence as increases in number of families who have access to electricity and potable water, appears to have been improved. The benefit/cost ratio for the project was very high.
The Mexican Government took over the financing of Plan Puebla after.. Rockefeller Foundation and CIMMYT support was terminated in March of 1974. In large part, based on the example of the project, maize promotion programs have spread to other states of Mexico and to other countries in Latin America. The Puebla Project has trained leaders and staff of these programs (the second objective in the original project statement) and has assisted them in initiating and operating their programs.
But somehow as a pioneer project achieves success, the objectives
change during the course of the project and people expect even more than the revised objectives. One question asked frequently is: "If the Puebla Project is so good, why is-the number of farmers in credit lists-after six years 16 percent of farmers and 26 percent of maize area so low'?"
First, it must be reiterated that the participation figures cited above. refer to farmers who officially signed up for credit under the program. As the data at the beginning of the section on "Accomplishments" suggest, the influence of the Puebla Project recommendations was much more pervasive than this restrictive definition of participation indicates. However, this clarification notwithstanding, the following factors, identified by the project staff as being most influential to the farmers in deciding wheth6: they adopt or not, appear to provide part of the answer to this question.

a. Adequacy of recommendations. Theoretically a project could have anywhere from a single recommendation for the whole area to separate recommendations for each farmer's field. Under fully irrigated agriculture, the former might be feasible. Farmers will make their own adaptations
to suit individual tastes anyway. However, in rained agriculture,
particularly when the area has a variety of soil conditions and when
planting dates vary, more than one recommendation must be offered. The project staff feel that 16 different recommendations, influenced mostly
by different planting dates but also by soil conditions, is optimal
for the area.
The project started with only one recommendation the first year. The
staff do not think this hurt the credibility of the project. The
initial number of participants was small and they received individual
attention. The package did not incorporate new elements but only
suggested more appropriate combinations of traditional elements the same seed, more fertilizer, and higher plant populations. Fertilizer
use had already been prevalent; 80 percent of the farmers reported
having used chemical fertilizer previous to the project although sometimes in relatively small dosages. The project recommendations were
based on careful field testing and the changes from traditional practices were introduced gradually in order to allow the farmers to
adjust. Perhaps the best rule would be that if one has to start with
only one recommendation, then it is probably better to station the conservative side, yet with enough change to show noticeable yield
and income increase. Or possibly it would be better to recommend two
levels, both conservative, depending on the capital available to the

- 38
b. Communication of recommendations. A second aspect of the problem is
that while the recommendations may be adequate, communication of the
recommendations may not be adequate. The project area was divided into
five zones, an extension man for each, which was a compromise between
the resources available and the desire to operate as closely to the
farmer as possible to transfer information and assist farmers in using
the information. The unanswered question is whether dissemination could
have been faster if more and smaller zones with more extension men were
used, i.e., could the benefit/cost ratio have been maintained? One possibility would be to subdivide the zones into subzones under which subprofessionals (similar to paramedicals in health care) could carry out closer farmer contact. To some degree, the 25 subprofessionals played
this role. Responsibilities could be allocated by concentrating the team of specialists in agronomic research at a central location, the
trained technical assistance agents at the zone headquarters, and the subprofessionals moving through the villages in personal contact with
campesinos. However, a strong two-way relationship between applied research and extension whether carried out by the same or separate
persons is very important. It is undesirable to insulate the various functions. When there are a limited number of professionals, the question of how best to utilize the trained manpower becomes a key decision.
It is agreed that five technical assistance agents is a minimum number
to serve 43,000 farm families; 100 agents is probably too many.
Aside from the number of technical assistance agents, another question arises about the quality of their training. Project staff acknowledge
that this is an area of which weaknesses do exist. There was little

- 39
to go on in trying-to prepare the men for their jobs and to provide
continuing assistance after they were actually on the job. The people
involved in technical assistance feel that they do not have the same
kind of backup as their colleagues in research work; the reason is that
the technology in research is very well worked out, whereas in the
case of technical assistance the basic methodology is not as well known.
The new technical assistance people who came into the project learned from the people who had been there before. They had relatively little
in the way of theory to guide them.
c. Agronomic risk. The assumption on which the project was organized was
that the farmer could increase maize yields as a low risk means of increasing income and security. While higher incomes are undoubtedly desired by the campesinos, the decision-making process is undoubtedly
more complex than this simple, although convenient, assumption suggests.
For historical and social reasons, the campesinos are highly individualistic and suspicious of others who do not belong to the same extended family.
The role of the wife in the decision-making is quite significant in many
subtle, although not adequately understood (by outsiders), ways. The
campesinos have very limited financial resources upon which to fall
back on in the case of failure. Many aspects of the so-called "culture
* of poverty" probably characterize these people.
In particular, aversion to risk may be an important component of the
decision-making process of the farmers. Drought, disease, pests, frost,
nonavailability of credit and production inputs, limited access to

- 4o
markets, poor health, etc., are some of the sources of risk which must be considered. Under rainfed conditions, the amount and distribution
of rainfall is the biggest source of agronomic risk. Limiting variability
in yield may be just as or even more important to farmers than achieving high levels of yield; if so, then farmers may tend to be conservative in their adoption of new practices until they are assured that their yields
will be relatively stable.*
The Puebla Project staff feel that, while there are instances where farmers
will reduce their net incomes by changing from traditional to the new
technology, farmers generally assume less risk by using the project recommendations. Their data show:
1) For average or better years, there is a high probability of an attractive-net income from using either technology. The expected net income is nearly twice as large with project recommendations as with the traditional practices.
2) For less favorable years the value of net income will be equal to or
less than 0.5 tons/ha of maize in many cases. The probability of these
low incomes is much higher with the traditional than with the recommended technology.
3) For the least favorable years net incomes less than zero can be expected. The probability of net losses is similar for the two technologies.
* Edgardo Moscardi is carrying out research, "A Behavioral Model for Decision under Risk among Small-holding Farmers," which should have very interesting implications for future projects.

- 41
d. Availability of credit and inputs. During the early period of the project,
services were inefficiently operated, i.e., slow processing of credit,
fertilizer available late or not at all, etc. Over the life of the project, some innovations in the delivery of inputs which were used in the other
projects were not tried in the Puebla Project. For example, Itminikits"
which have been used successfully in several countries, have not been
tried in Puebla. Furthermore, the farmer is responsible for carrying his
fertilizer from the central town back to his village. It does not take
many bad experiences to convince a farmer that he is better off by staying
with his traditional methods which, although not very productive, are mostly under his control. I have not seen any studies which indicate
whether the interest rates and fertilizer prices are at the most desirable
levels from the viewpoints of the farmers and the fertilizer dealer.
e. Crop insurance. This is a service which obviously was not understood and
possibly was not administered properly. It has a legitimate function yet apparently was not popular with the farmers who saw it as an unnecessary
f. Organizing the farmers into groups. Sheer numbers of participants relative
to extension agents forced some sort of grouping arrangement. Groups are a convenient means for a limited number of technical assistance agents to
spread themselves among a larger number of farmers. The project staff
also used these organizations as a means of promoting change in the service institutions a group of farmers has much more influence than an individual
in getting changes in credit.

However, some people have feared that the formation of groups in which
membership is a prerequisite to getting credit and otherservices could
serve as a means of excluding as well as including participants. The Plan
Puebla staff do not think this is a big negative factor. Their studies
indicate that the major reason that there are not more participants in the
credit groups is that the operating conditions of the credit institutions
are not sufficiently attractive to the majority of the farmers, i.e., they
apprehend a series of problems related to the policies that regulate the
authorization of institutional credit.
It is also possible that the groups can serve to give an unfair advantage
to the stronger farmers. Groups are probably composed of relatively
homogenous memberships. The bigger groups of wealthier farmers would tend to associate together and could probably have advantages in being
serviced first at the banks, have privileged access to fertilizer which
often runs short at the crop season, etc. Actions such as these or
systematic exclusion of the less responsible members of the community could contribute to faster disintegration of the traditional society.
No evidence was presented to indicate that these are actual problems.
The alternative to groups is not obvious in the Puebla situation. However, the total organization of the village in all its phases, including
an examination of what motivates different factions, must be examined closely in determining how to influence total community participation.
g. Labor shortages. It was noted that the new recommendations require more
labor, 30 percent more, than traditional practices. The additional labor

requirement competes directly with the opportunities for employment
off-farm in such places as the town of Puebla. Although possibly limiting the spread of new agricultural practices,* such a tight labor market should work to the total economic advantage of rural laborers by bidding
up wages.
h. Farm size. It must be admitted that many one-half hectare farmers were
just not willing to go into debt in order to accept the risks necessary to achieve the gains which the Puebla Project promised. Is it realistic to expect any credit program to include even more than half of the very
small farmers?
I stated at the beginning that this paper had two interrelated purposes. The first was to explain the Puebla Project as a rural development project. In intent, the Puebla Project was established neither as a rural development project nor even an agricultural development project; it was initiated as a learning process testing a planned approach, based initially on what was known, with built-in mechanisms for evaluation and subsequent readjustment and this is the basis by which it should be evaluated. However, in fact, it went beyond this. Through applied research and feedback, technical credibility in agriculture was established. Project leadership and farmer participation promoted the spread of the technology by inducing the relevant institutions to become more responsive to farmer needs. Welfare was increased
This hypothesis is being investigated by Manuel Villa-Issa in his study
"Labor as a Constraint to the Adoption of New Production Technology: The
Case of the Puebla Project."

- 44
throughout the area. Could a project which was initially designed as a 11 rural development" project have accomplished as much of the rural development objectives of improving the standard of living during the existence of the project and laying the groundwork for continued future improvements as did the Puebla Project with its seemingly narrowly defined initial thrust?
As to the second purpose of the paper, I think it is clear that social scientists did play an important role, working jointly with technical agricultural research and production specialists in the project in many ways. Essentially three types of research approaches were used: 1) a technocratic approach of introducing changes on a trial basis and evaluating results, 2) socioeconomic surveys, and 3) continual informal observation and interaction with the farmers and service agencies. Each of these approaches has its role and, when properly carried out, can contribute very positively to project success. However, it is also obvious from this experience that there are still numerous questions to be answered and that improvements must be made in social science research approaches in order to provide the needed answers. For example, it would have been very valuable to have introduced more rigor into the evaluations, especially the feeding back of hypotheses to be tested in carrying out specific activities. Methods for evaluating institutional change are particularly poorly developed.
In discussing the role of social science research in the Puebla Project, or any rural development project, one must be relaxed regarding what is considered social science research and one must do the best he can with the resources which are available. The important consideration is that

social science research however defined and however carried out should be viewed as a means to an end rather than the end in and of itself, i.e., the objective is to bring social science considerations into the decisionmaking process of the project in order to increase the success of the project in achieving desired goals. Social science research did make this contribution, effectively, in the Puebla Project.

Preliminary version
comments are welcome
March 1975
Santiago I. Friedmann **
In this paper, I first describe in broad terms the institutional infrastructure for research on rural development and the past experience on, and present opportunities for, relating research with-policy and action in Mexico.
In spite of the limitations of the analysis it seems fair to assert that in Mexico, as it has been the case most other countries, the contribution of the, social scientists to rural development has been minimal. This has been caused, on the one hand, by the belief among doers thatthe problem was essentiallytechnical and rather simple and that they knew all the answers, and, on the other hand, by the inadequacy of the concepts, theories and methods of the social sciences for dealing with the rural poverty sector. These theories have mainly been developed in and for a modern urban-industrial setting. There are several elements in the Mexican scene which give grounds for hope that a promising opportunity exists for finding and demonstrating the ways by which social sciences can contribute effectively to rural development: (i) general recognition of the limitations and failures of previous approaches, including land distribution, which has led to an awareness that there are no easy solutions to rural poverty,
(ii) existence. of a basic institutional capacity for training and research in the social sciences which could expand and become relevant to the problems of rural development, (iii) the history and variety of past actions coupled. with the magnitude of the resources presently committed to rural development Prepared for a conference on this subject the Rockefeller Foundation
in New York, April 29-30., 1975.
* Project Specialist, Ford Foundation, and Visiting Professor at the Center for
Agricultural Economics, Postgraduate College, National School of Agriculture,
Chapingo, Mexico.

that provide a unique "laboratory situation" on which to build social science relevant to the rural poverty sector. Because of these factors in Mexico the way is open for a new start by social scientists, policy makers., program managers and citizens. In section 3, I indicate the changes in subjects and style that I consider necessary and I try to demonstrate the need for complementing the distant and critical observer, excellent in denouncing shortcomings but poor in providing solutions, by a participatory and committed social scientist.
The last section is devoted to a selective discussion of the role of external agencies. General principles are proposed for-guiding foreign participation in rural affairs which are intrinsically national in character and decision. I refer, in particular to the role of external agencies in cooperative efforts aimed at development of national and international networks by which multidisciplinary capabilities for policy and action research are enhanced and a wider utilization of research results is insured. It is asserted that foreign support should be focussed on the development of the lozal decision makers and academic community's capacity for: (i) identifying the problems to whose solution social sciences research could contribute and, (ii) designing and implementing the necessary research projects including the definition of the need for foreign resources.
Following my own -inclination and the letter of invitation for this
conference, throughout this paper "rural development is being defined in terms of improvement of both productivity and well-being of rural lower income peoples."

1. The Infrastructure and Environment for Social Research in Rural Development in Mexico.
This section suffers from limitations on time that I could devote to data collection for the paper and the absence of any systematic inventory of research on rural development in Mexico.
In broad terms the Mexican situation is well illustrated by the papers of Gelia Castillo and Rolf J. Luders presented at the Bellagio Conference on the Financing of Social Science Research for Development, February 12-16, 1974 (See references 2 and6).
In Mexico the scarce senior social scientists are subject to the same
excessive demands in the universities from teaching, consulting and administration and in government for execution and day-to-day decision making. One finds very little long-term and intensive commitment to a research subject and few means of communication with other scholars in the country, while at the same time enjoying full membership of the international jet-set. There exists no institution like the Filipino Community Development Research Council. The National Council for Science and Technology (CONACYT) has just established a special group for rural development research as part of an effort to develop the national plan for science and technology. It is expected that as a result of this effort some inventory of research on rural development will become available in the near future.
Social science research is carried out in private and public universities, research institutes with or without graduate training programs, government ministries, public corporations and a handful of private development organizations.
Higher education has been growing very rapidly in recent decades, not
only in terms of larger admissions but also through developing graduate training

programs often with substantial external financial support and participation of foreign professors. Now, master programs are offered in every social science discipline, and Ph.D. programs in most. They are increasingly based on full2/
time Mexican faculty trained in the U.S. or Europe. University research in the social sciences is handled mainly by graduate divisions and research institutes of the National University, the National School of Agriculture in Chapingo, the Instituto Polit&cnico Nacional and four or five state universities. In 19 74 the Center for Research and Training in Economics (CIDE) was founded with the purpose of catering to the needs of the public sector. Ihe main private universities/in the field of social sciences are : El Colegio de .Mexico and the Iberoamericana in Mexico City, the Monterrey Institute of Technology
(ITESM) and the Universidad de las Americas in Puebla .
In most of these institutions, economics is the strongest discipline but
Chapingo and ITESM have adequate and improving master's programs in agricultural economics; El Colegio de Mexico is strong in demography and sociology; the Iberoamericana has a Ph.D. in anthropology and Chapingo has also a master's program in rural sociology. CIDE is planning research on the public institutions eemparingsthe supporting system for agriculture, focussed'on the traditional sector.' The focus of both training and research has been outside the rural poverty sector. The main exception comes from anthropology where many studies of Indian communities have contributed to documentation of the situation but r~egre--tably have provided no operational guides for policy or action aiming at improvement of such communities. The rural situation has been discussed mainly in terms of its relation to the rest of the society and mostly from a theoretical or ideological framework (class structure, duality, dependency, resources I/ There are more than students enrolled in graduate programs in social
science in Mexican institutions.
2/ In the four years 1971/1974 approximately 450 social scientists have
received scholarships for graduate study abroad. 0 t e. .
See" 4W1- f9 fzr dat on ollent'ade.

transfered out of agriculture). Little empirical work has been done to help in defining the whats, where, when and hows for rural development programs.
In the case of non-training research institutes one can mention the Mexican Institute of Social Studies (IMES), the Center for Agrarian Studies (CDIA) and the Center for Educational Studies (CEE). These institutions are 1/
staffed with capable professionals ; IMES has developed proficiency in population studies but only about 10% of its research refers to the rural poverty sector. CDIA has focussed on studies of the land tenure situation, serving as a basis for regional research and as a host for foreign people working in this field. More recently CDIA has become interested in the study of organizational structures for marketing, production, financing, etc., within ejidos and small holders, and in studies in support of a National Agrarian Training Program to be implemented by the newly created Ministry 2/
for the Agrarian Reform with technical assistance from FA0 and UNDP. Both IMES and CDIA are committed to problem-oriented research and keep close ties with developing institutions. The CEE. has kept track of development in education and does evaluative studies of educational programs.
In the case of government institutions one finds well staffed groups
doing study and analysis of the operations of the federal government agencies and public corporations. Again these groups are stronger in economics. The recently created Coordinating Commission for the Agricultural Sector (COCOSA) has been ab-le-to put together information and analysis done by technical groups within its member institutions. This effort is contributing to a better identification of short and medium term problems in the agricultural sector and to formulation of policies which reflect a better understanding of the implica1/ Total full time professional staff of IMES, CDIA and CEE numbers 27. ZI/ FAO/UNDP project involves 18 social scientists half of which are foreign

tions for the backward sub-sector in agriculture. COCOSA has also promoted meetings of members of academic and research institutions with people responsible for policy and action in areas related to rural development. These discussions resulted in preparation of an outline of a comprehensive research program for rural development. Similar results have been obtained from seminars held on specific issues of rural development like organization, agrarian reform financing, irrigation, etc., with the participation of personnel of related institutions and social scientists. The personnel from universities and research institutes participating in these meetings have been able to identify more clearly relevant research questions and as a consequence new research is being organized along these lines. Government efforts will be implemented through a newly created institution, the Center for Research on Rural Develop1/
ment (CIDER). A group of the major academic and research institutions is trying to coordinate research and is looking 'or the necessary funding.
Private institutions doing work in rural development (like the Mexican Foundation for Development) have a small group of social scientists engaged in evaluation research on their operations. Services of research institutes are also contracted for this purpose.
To complete the picture one should mention the existence-of several
foreign research groups supported by UiN agencies, ECLA, FAO UNDP, O OAS, ILO, etc., which are concerned with social research on rural development, mainly on questions of employment and on the effects of investment in physical and social infrastructure. The Economic Division at CIM[YT together with the graduate division in Chapingo have been studying the impacts of the Green Revolution on the rural poor and trying to identify means whereby the benefits of new agricultural technologies accrue to disadvantaged groups.
In general, one can say: (i) in the past very little social science
1/ The planned staff of CIDER is 36 professionals

research has been done that is relevant for policy formulation and program design and implementation for rural development; (ii) at present the universities are capable of training the human resources for these activities. Their graduate divisions, other research institutions and government agencies are staffed with hundreds of people with graduate training in the social sciences whose past experience in other areas of research could be developed to proper respond to the challenges and opportunities that are discussed in the following section; (iii) both policy makers and managers in government programs are becoming aware of the need for policy and action research and are showing interest in the social scientists' involvement in their activities. There is evidence that the latter are willing to respond positively.
2. Challenges and Opportunities in Mexico
Mexico, among Latin American countries, has special characteristicsboth in terms of challenges and opportunities for social science research. In most Latin American countries the little rural development research that exists ("there is much less on R.D. than the subject merits." Luders, pg. 198) is related to agrarian reform programs. The Mexican Revolution which resulted in elimination of latifundia; the stagnation of agriculture in the last decade and the concentration of poverty in the rural areas have been confronting Mexico 1/
with the question of what is needed besides and beyond land distribution.
Answers have been coming from many sources in many ways: federal investments in krigation and Other rural infrastructure; attempts to bring the Green Revolution to small farmers in rainfed areas (Plan Puebla); industrial descentralization; marketing services; socio-political and economic organizations
17 Gont.......... iun ..... dcvU- e1 rp tnsdon' t .lgnd themselves
t-aacyL it zckgia4(7.

of ejidatarios and small farmers, concientizaci6n and education and so forth. Thege activities have been promoted by federal state and local governments, international private and public organizations, national private groups and individuals. (Several interesting community projects are being carried out individually by professors of the National University, the Agricultural University in Chapingo and others).
Most of these projects have been initiated as a result of the push of individuals who had a quasi Mesianic belief in their ability to provide a panacea to rural problems and generally the same individuals have been responsible for their implementation. This explains two common features of rural development programs: (i) the programs are often discontinued before. completion when their promoters lose interest or power, 'and (ii) little or no documentation exists to permit their evaluation on the basis of knowledge of what was done in the program and what happened as a result of such a program. Frequently they are also paternalistic: objectives and actions are chosen by the outsider without consultation or participation of the expected beneficiaries.
The present administration is giving top priority to theamelioration of rural poverty. Massive outlays are being made to provide the rural communities with physical infrastructure and social services (roads, telecommunications, employment, formal and vocational education, potable water and sewage, health centers, etc). Institutional means for descentralizing and coordinating these. efforts and insuring the participation of the rural people in project selection
-and execution are being developed. At-the same time steps are being taken to carry out research to support the design and implementation of the action programs and to learn from them. An important part of the investmentsbeing made under the-aegis of the Integrated Rural Development Program (PIDER), partially supported by the World Bank and the Inter-American Development Bank, which is expected to reach about one million mv families over the next five

years (more than one tenth of the Mexican population). A more modest effort
is the national program for development of rain-fed agriculture (PRONDAAT)
which aims to promote technical change among small producers (with less
than five has. of cultivated land) in non-irrigated lands using the experience
of Plan Puebla and other projects. It is expected that this program will
reach more than 250,000 families in the next five years. In both these 1/
programs, research groups are being assembled to provide supporting research.
In summary, the persistence of agricultural stagnation and rural poverty;
the wealth of experience with the most varied instruments for promoting rural
development, which at present remains largely untapped by systematic research;
and the strong national commitment to rural development expressed in a massive
allocation of resources to action, poses a challenge and provides an opportunity
for social scientists to go beyond denounciation and to demonstrate their
potential for contributing concretely to effective rural development programs.
1/ Supporting research is a set of activities of study and analysis that are
carried out in a problem solving context with the purpose of: (i) utilizing,
at the initiation ol the problem solving process existing knowledge in the
development, evaluation (ex-ante evaluation) and selection of the alternative to be pursued; (ii) (program monitoring) verifying during the implementation that activities are executed as planned; (iii) (feedback analysis) utilizing
incoming information about the actual effects of actions already executed and the actual values of non-controllable intervening factors in deciding
if and how to redesign the ongoing alternative and/or to redefine its objectives, i.e., deciding ways of utilization of currently available
resources in future actions; and, (iv) (ex-post evaluation or validation)
at the end of the process: (a) verifying if, as a consequence of the actions implemented during the process, its goals were reached; (b) to observe nonanticipated results; and, (c) to check the extent by which the fulfillment
of goals and the non-anticipated results influence the solutio- of the
total problem.
The results of the ex-post evaluation and the feedback analysis are used
in the definition of the sub-problems and for the selection and evaluation
of the alternatives to be considered in the next stage of the problem
solving process (Friedmann, pg 4)..

Paraphrasing Gelia Castillo (pg. 185), one could say: the Mexican
social scientist has come of age. Hlie is respected and his advice is sought.... Also, one is tempted to repeat her call for humility -- "For his long term survival he should not bungle this golden opportunity t4 overselling what he cannot deliver."
It is my feeling that for the social scientist to succeed in responding to the challenge and the opportunity, a question of style is important. In the following section, I stress the importance of close interaction of social scientists, program managers and beneficiaries through direct involvement of the former in the program, in what one could call participatory supporting research.
3. The Uses and Style of Social Scientists in Rural Development.
Traditionally the social scientist has been used in rural development programs in two capacities: (i) to help in justifying and or in implementing a project that has been selected and designed without his participation, and
(ii) to perform ex-ante and ex-post cost-benefit analysis in order to comply with regulations of the budget office or of some international institution supporting the program.
The situation is very much the same of most-multi-disciplinary efforts in which members of several disciplines are invited or requested to contribute their skills within a conceptual framework and a methodology defined by.a dominant discipline. It is not surprising that the members of the "minor" disciplines become desinterested and only few, which are frequently not the best qualified, continue to participateand the members of the dominant discipline are reassured in their initial skepticism about the relevance of other disciplines.

It is my conviction that opportunities exist for a much more comprehensive use of social scientist's abilities for rural development. Gelia Castillo (see Ref. 2; especially pgs. 159-163) has much to say that is pertinent to the Mexican situation on the nature of social scientists' work and on the means of communicating results to policy makers. I will only stress the need to study the processes of technical innovation, the nature and changes of societal structures internal to the rural communities and of the external linkages with institutions of the supporting system and with the markets fbr labor, inputs and outputs. The differential impact of the changes in rural situation on small farmers, landless peasants, women and children also should be studied. Most of this section will be devoted to discuss questions of style. For the full exploitation of the potentialities of social sciences in development. changes are necessary both in the program managers' and in the social scientists' view of the latters role in rural development activities. On the one hand, the social scientist should not be conceived as an element alien to the program, observing it as any research topic and having no responsibility for what happens with the program. On the contrary, he should be a part of Che professional team in charge of the program, adding to knowledge of the technical processes involved in rural development actions (i.e., use of fertilizer, irrigation works, sanitation works, etc.) an understanding of the social processes that are essential for rural development to occur. This implies
1/ I have claimed elsewhere (Friedmann, pp. 8 and 9) that this is not just a
- matter of expedience. (Uma-Lele in Ref. X presents a conclusive case for
the operational importance of participation in rural development programs). In
a seminar on supporting research given to the social scientists in the staff
of PRONDAAT, I made explicit my value judgements by defining development "as
a process by which a society changes in multiple ways striving to realize its
own 2rdlecto de civilizaci6n.' This results from its own past and its interpretatibn of the experiencesOf other societies. Development is a process in
which human potential is progressively realized by means of individual and social learning process that improves the ability to adjust to and modify
the environment and to more adequately utilize& the natural and social
characteristics of the environment., Development creates the conditions that allow the human being, as an individual and as a member of a social group to
identify and manifest his needs, and to satisfy them by developing and applying his abilities."

also changes on the subjects of his concern, he will no longer be
limited to measuring economic gains and costs associated with an irrigation project or advising on the design of a "concientizaci6n" program to make the local people accept a new technology. Instead, or besides, he will try to understand the decision-making process and the aspirations of the potential beneficiaries so as to utilize the resources of. the program in a way that will maximize the improvement in their well-being as measureHe will also study-the motivations of the staff and the rules and procedures of the institutions surrounding the rural communities trying to find there reasons for the failure or weaknesses-of rural development programs instead of seeking such reasons solely in the "traditional" rural culture.
For the social scientist input to be useful it is necessary that he and the program managers and beneficiaries become educated on each others worlds so that the social scientists'sactivity of study and analysis refer to aspects that are relevant to the rural development process and his results can-be utilized by managers and beneficiaries of rural development programs. These conditions can only be met if a large fraction of social scientists concerned with R.D. spend time doing participatory supporting research for R.D. policy formulation and program design and implementation. In the rest of the section I try to argue the case for advocating a participatory role for the social scientist in R.D. programs which will provide for mutual educational interaction.
a. The education of the social scientist.. No social science discipline
provides the concepts and theories capable of dealing with the complexities of a 1/
concrete rural situation. In order for the social scientist to isolate a relevant part of the problem, amenable to the tools of his discipline, taking I/ These in general are based in generalizations, abstractions from a coher-.
urban industrial society. The main exception are the anthropologists; but
most of them have been more concerned with providing a static picture of ,,
Av:pecific situation than with identifying mechanisms for supporting develop-/
mental changes.....

due consideration of the boundery conditions, he must become inbedded in the concrete rural reality. For this purpose the insights that program officers enjoy thanks to their close exposure to reality can be of great help. In fact, I would claim that only by matching the detailed, anecdotic, varied knowledge of the program managers with the concepts, theories, and scientific methodology of the social scientists, one can hope to gain the necessary under-. standing of the processes of change within the rural societies. Program managers are in general, strongly motivated people who work long hours in order to alleviate rural poverty. They have no patience with uncommitted professors or students interested in studying. the "rural guinea pigs" and whom he perceives as the most recent addition to the zoo of rural exploiters,. in this case taking time, food, shelter, and adding frustration to the rural poor., for the purpose of completing a study that will promote their careers or allow them to get a degree. Additionally, the interaction with beneficiaries is necessary in order to understand their social and economic conditions, aspirations and their perceptions of the R.D. program. The initial attitude of the beneficiary towards the social scientists will be very similar to that just described for the program officer. Only the social scientist who brings to the program contributions perceived by its staff and beneficiaries to be at least as equal to what he takes from them can he hope to receive the additional information and insights necessary to start developing the concepts, theories and methods of his discipline to. a point where it can cope with the conditions in this new environment.
b. The development of program staff and beneficiaries.. By participating in the program the social scientist will not only be educating himself as described above, but he will also contribute to the personal development of the rest of the staff in the program and of the beneficiaries. In order for the results of social science studies on rural development to be fully utilized,

it is necessary to complement the technical knowledge and the missionary zeal of most program personnel, with a basic understanding, and a basic analytical capability bearing on the social elements. This should allow them to process the anecdotic and scattered information they capture on their daily errands in order to (i) identify and implement the adjustments in the program activities that are necessary to eliminate bottlenecks, (ii)
*correct unexpected and unwanted effects, and (iii) take advantage.of unanticipated opportunities. The interaction between the social scientist and the program officers also provides the opportunity to instill in the latter a scientific style. This will on the one hand, help him to consider experimentally the instruments and strategies of the program: not as given and good in themselves, but as means to achieve an end whose efficiency and adequacy should be validated by actual experience. On the other hand, he will recognize the need to define the goals of the program not on the basis of his own perceptions and values, but in response to needs and aspirations of the rural people. This leads us to stress the role of the social scientist in contributing to the process of these people's social learning which, I believe, is the essence of development (See footnote on page 11).
4. On. the Role of External Agencies.
I will open this section by transcribing from Friedmann and Lomnitz
some general remarks on the role of international cooperation in transfening "Intellectual Technology".
Then I will add some remarks on the specific issues of this conference.
(I. General remarks on "The Role of International Cooperation".
The process of development which takes place within a country should reflect the particular values, and the cultural and institutional traditions of that country. Consequently the objectives and measures employed in devel

programs should emerge from a decision making process dominated by individuals and institutions directly affected. However, virtually all of the decisions involved possessed both political or value aspects and scientific and technical aspects. Public and private institutions of the richer countries may contribute profoundly in the generation and diffusion of the needed scientific and technical knowledge without intruding upon the process of making the basic political and value decisions involved.
The external aid granted for the support of specific programs in a particular country should be applied for and oriented by the receiving country and should conform to their priority needs.
SThe experience of the governments of the advanced countries, in the
mobilization of their scientific and technical resources may be used to encourage in those same countries basic scientific research and the technological developments orientated towards the satisfaction of the development needs of Latin America. The universities of the more advanced countries can offer programs intended for the training of their own nationals and of Latin America'in order to perform the needed tasks. The multilateral institutions and foundations can support the development in Latin America of the research capacity for looking after their national problems. They can also contribute to the financing of studies and to the application of development programs in which the possibility of different instruments of change can be explored and which allow the drawing up of more suitable strategies for their large scale application. The transnational corporations can include through their local operations in Latin America research and development activities specifically orientated toward the economic production of goods and services which satisfy the needs of the poorest sectors and toward the design of production procedures and techniques which utilize human resources and materials available in the said sectors.

Finally, the academic and professional associations can contribute with
their meetings and their editorial policies to focus the attention and to.induce their members to work more effectively and more frequently on the problems of development as understood in the terms used by Mahbub U1 Haq '... the problem of development should be defined as a selective attack on the worst forms of'(Friedmann and Lomnitz, p. 72 and 73).
1. Some Specific Remarks.
Again I must refer to the comments by Castillo and Luders who stress the importance of foreign inputs in establishing standards and introducing new techniques; making possible research that otherwise would not be supported because of its high risk or potentially and undesirable political implications; promoting communications among disciplines, etc. It seems to me that the case made on section 2 show that Mexico offers a unique opportunity to external agencies interested i rural development for (i) helping in learning and applying the lessons from the variety of actions underway, and (ii) making these lessons available to other nations.
I concur with Castillo and Luders in assigning-great potential to loose
research networks which-is considered as a natural "turf" for external agencies. From my previous comments it should also be clear that I agree with.Castillo's statement:"The actual question therefore is, who decides the subject and object of the research network".
In closing I will extract from a memo by Michael Nelson and myself
reacting to an initiative to establish a Latin America Research Network (IARN). "In our view the majority of the potential advantages and disadvantages are already on the table, the next step is a decision on how to create a mechanism, which is essentially Latin, to address the various questions and generate a proposal to the FF. We will return to this point later..." "We have two models in mind: (i) the network would be set up with a permanent headquarters

and a wide-ranging charter to identify and promote research in any aspect of rural development, seek collaborating scholars and institutions, and carry out fund-raising activities from an array of national and international donor organizations... (ii) The network would be set up around a single research topic (or set of topics). The core group would be selected specifically to address this topic(s), and the most appropriate location would depend on the topic ... "The second model would respond to one of Yiv. concerns that the network 'self-destruct' after 3 to 5 years if it. is not responding to the needs and concerns of the institutions and countries which expect to benefit from the activity. Our position is that the second model is vastly superior to the first in the initial stages of setting up the network. If the system responds well one has two alternatives in the second phase -- set up the core group in a permanent location or -- reconstitute and relocate the core group in accordance with each new set of research topics agreed upon. One might even contemplate more than one core graup. This latter aspect is in line with suggestions made in earlier discussions that a family networks may evolve. However, in the first stage we feel that the second model avoids many of the problems raised by -- excessive bureaucratization, creation
of a new center with infinite life expectancy, siphoning off of scarce Latin American talent which is needed to resolve national problems, and the undue attraction of international salary scales.
"On the question of 'scope' it is our view that the network, if it is to make any real contribution where the final product is not merely the stmmnation of the individual pieces, must be mutlidisciplinary and under no circumstance should the start be only in the field of agricultural economics. An effort should be made to follow, and build on, Ernie Sprague's initiative at CIUYT for interdisciplinary team training tied into problem-oriented research. Tiv4

This suggests that the core team be interdisciplinary and that similar criteria be applied in arriving at which institutions and individuals within institutions, will be involved,.
'Participation from "without the region" implies: (i) that considerable training of Latin Americans will continue to be done outside the region and therefore it is important to have foreigners in foreign institutions involved in research on Latin America, and (ii)that for purposes of education, research and policy in countries outside the region, it is important to have scholars from these countries working with Latins on problem-oriented research in the region itself. The pros and cons need some airing.
'With respect to research topics, following the second model suggested above, the objectives should be highly specific with clear identificati,, of those countries with a common element of interest and how those countries may benefit-from complementary research and experience, inter-country comparisons, or adoption of comparable methodologies. We are assuming that LARN will address as a principal focus, electedd aspects of rural poverty in Latin America. On this point.i has expressed doubts on how the regional research effort eventually will benefit the rural poor. Taking this as an ultimate goal, he asks what the alternative opportunities ar4 for use of FF funds. Answers to these questions must be formulated in the process of generating the LAN proposal to the FF. It would seem essential that government decision-makers play a role in selection of the research topics and methodology; as the prime users, it is to be hoped that such people have confidence in the findings. In principle the widest possible range of institutions would have an opportunity to participate in the selection of research topics, collaborating institutions and individuals, and the core group. However, this democratic approach cannot be adopted at the outset. Which brings us to our final comment on the LARN proposal.

The Next Step:
'It appears that the crucial questions to be resolved in advancing the proposal is -- by whom and how, will this research network be set in motion? Since this is to be a Latin American network, it might be appropriate to start by putting the various questions raised to the Latin American clientele of the network. Questions such as:
- mechanisms for offering wide opportunity for participation in research,
and in decisions on collaborating institutions, staffing,.salaries,
and research topics.
translation of social science research into action
- expected benefits to Tnstitutions and countries from participation,
leading to criteria for judging the usefulness of the network
- need for a permanent core group
composition, size and nationality (Latin American vs. non-Latin
American) of the core group
- location of a core group
type of or need for association with non-Latin American scholars
- priority research topics of international interest
- funding requirements and sources of funds
the problem of loss of professionals from a national to an international program
- salary scales
- procedures for: phasing development, and on-going evaluation of
the program.
'In order to get started along such a line it would be necessary to
form an ad-hoc committee. "Plan Valdes" is a perfect example of the effective functioning of such a cormnittee at the national level. Similar committees could be formed elsewhere. However, at some stage it would be necessary to bring together what we consider to be a "representative" group of Latin Americans to review the type of questions listed above. It is quite evident that such a group could not represent all institutions from all countries.

It should represent both government and academic institutions and may even include someone from the private banking or agro-industrial sector.
..."Members of this ad hoc committee could be provided with all the
background materials on the IARN proposal and be requested to review it with institutions which may have potential interest. It is to be hoped that each member would synthesize the views expressed. At that point the committee chairman should put together a discussion paper reviewing alternative proposals, the principal problems to be resolved, and the various reactions to the problems posed. The committee would then be called together to consider this document plus any other models a member cares to present. (Observers could be invited from the FF,. other potential donor agencies, international organizations; or individual specialists-could be invited for their professional contribution). The expectation is that this committee would address issues such as those listed above. If a proposal results, the Foundation could raise certain ideas for further consideration at that time.
It will be important to build on the spirit of Alberto Valdez' lead, and assure that the initiative for IARN remains with the Latins."
The last sentence summarizes the message. International agencies should help in building basic capacities for training and research and should support the evolution of the local social sciences communities and decision makers towards self-determination.
Since September 1972 1 have been in Mexico. For two years I benefitted from a close association with Bill Lord, also a visiting professor at the Centro de Economia Agricola of the Graduate Division of the National School of Agriculture of Chapingo, and with the students and local faculty. I have also had the good fortune of being given the opportunity of working in close

contact with Eduardo Dominguez, his staff and the peasants in subregion IV of the Plan Maiz with Heliodoro Diaz and his staff at PRONDAAT and with Jos6 Silos and his staff at the Coordinating Commission for the Agricultural Sector. Their insights and frequent exchanges with Teresa
now at El Colegio de Mexico)
Rend6n/and Mike elson provided a general framework for this paper. Mike Nelson and Barry Schuman of the Mexico office of the Ford Foundation made very valuable comments to a first draft of the paper. Carlota Barrientos and Olga Cdrdenas of the local Ford Foundation staff were very helpful in gathering the data for the first section. Carlota also contributed with very useful comments and insights during the actual writing of this paper.

1. Ballesteros, Juan, "Balance de la Reforma Agraria Mexicana", Economia
Politica Vol. IV No. 3, 1972, p. 47-63.
2. Castillo Gelia, "Social Science Research; The Philippines Experiment" in
The Social Sciences and Development, World Bank, 1974, p. 141-190.
3. Friedmann, Santiago I. "What is Rural Development", paper presented to
IBRD, Development Research Center workshop.on Rural Development,
Washington, D.C., Jan. 6-8, 1975.
4. Friedmann, Santiago I and Larissa Lomnitz, "Development of Latin
America: The Rural Poor. Can the Haves Help?, ITCC Review,
Spring, 1974.
5. Lele, Uma. "A Conceptual Framework for Rural Development," prepared for
Development from Below Field Trip Workshop. Oct. 12-20, Addis Abeha,
6. Luders, Rolf J. "Social Science, Research in Latin America." The Social
Sciences and Development. World Bank 1974. p. 191-213.

Uma Lele
, A paper prepared for presentation at the meetings on Social sciences
Research. in Rural Development, organized by the' Rockefeller foundation,
to be held in New York during April 29-30, 1975.
1/ The views expressed in this paper are those of the author ajdd do not
necessarily represent the views of IBRD.

Uma Lele
The Background
This paper is based on the African Rural Development Stidy (ARDS) carried out in IBRD during 1972-74 The paper complements und presupposes the knowledge of the previous paper also based on the stidy; which concentrated on lessons of the African experience for designing rural
development programs 2/
To appreciate fully the observations made in this paper about social science research, it is useful to know the factors that prompted the study, and the approach adopted in the analysis.
The origin of the study is attributable to two interrel ted factors:
(1) the very substantial interest of the East and West Africa regions of
the IBD in finding ways of designing rural development projects which will effectively "reach" large numbers of low income rural people, given the constraints of finances, trained manpower and institutions encountered in rural development in Africa; and (2) the realization that the knowledge as how to bring about the development of the subsistence rur]. sector is very limited.
The study was, therefore, aimed at (1) asessing the casting state
1/ The main report on the Study is forthcoming as a bopk. See: Uma Lele,
Design of Rural Development: Analysis of Programs and ProjEcts in
Africa, Johns Hopkins Press (July 1795).
2/ See: Uma Lele, "Designing Rural Development Programs: Less(ns from Fast
Experience in Africa", a paper presented at the Second Inte rnational
Seminar on Change in Agriculture, organized by Reading University and the Overseas Development Institute, Reading, England, September 9-19,

of the knowledge as to the factors that"promote (and inhibit) participation of the lowest income groups in rural development programs, (2) to point out areas that need further investigation, and (3) wherever possible, to provide guidelines on the basis of analysis of past experiences for design and implerientation of the World Bank's future rural development programs in Africa 1/.
Three areas of concern were predominant in initiating the study. First, there was an explicit interest in promoting participation of the lowest income groups. Second, development was being viewed mo 'e broadly from the.perspective of improved welfare as well as of increased agricultural productivity. Consequently, there was a keen interest in identifying ways of establishing priorities and of time-phasing between and among -oth productive and social service activities. Third, there was an overt recognition of the financial, manpower, and institutional constraints encountered in rural development, and hence of the need to find ways of making maximut use of the existing rural potential with regard to these resources, and of augmenting them overtime.
The focus of the study was, therefore, consciously operational. The purpose was to examine the factors which influenced the choice of interventions in planning past programs, including the knowledge that existed in the course of planning and implementation of these programs about the actual constraints and potentials, and to assess the extent to which the programs were able to deal with circumstances-including with adverse f fects of overall
policies and institutions which were initially- not foreseen. One of the rain objectives of the ARDS was to investigate whether and how the projects
1/ See: Uma Lele, Design of Rural Development for description and detailed analyses of the projects analyzed.

reviewed, and similar other projects should have been designed, had there been more concern with broad participation. overall welfare and with the utilization of local financial and institutional resources, and equally important, had Che knowledge of economic, physical, technological and institutional factors required for designing and implementing such
programs been better.
The Analytical Approach
Due to the severe limitations of the existing conceptual framework, discussed later in this paper, generalizations were
drawn on the basis of empirical analysis. For comparative evaluations, seventeen sets of rural development projects and programs involving participation of .a number of multilateral, bilateral and national agencies, were selected from various parts of sub-Saharan Africa to represent diversity in design and implementation as well as in the environment in which they are situated. The reviews were carried out with a view to raising a consistent set of questions about these programs and projects, particularly in regard to their design and implementation, so as to derive lessons for designing future rural development projects
in Africa. In addition, rural sector surveys were conducted in Kenya and Tanzania to analyze overall development policies in these two countries and to develop the Bank's lending strategies.
All the reviews were based on data that had already been collected by a number of agencies. With a few exceptions no systematic field surveys to generate additional quantitative data were conducted. However, in all cases analyses of existing data were combined with very extensive field investigations. These consisted of (a) interviews with persons who

had considerable experience in design, implementation, supervision .and evaluation of the specific programs reviewed or similar other programs,
(b) interviews with the rural people who were being "reached" through programs and (c) searches for additional sources of datp related to programs reviewed from other donor, governmTent and research agencies. Further, very substantial literature on African rural development was reviewed to support or oppose the observations made on the basis of research results, field investigations and interviews.
The study, therefore, provides an unusual vantage point front which to (i) enumerate some of the. major questions currently faced in ,making policy and operational decisions in rural development (ii) the amenability of the social sciences to carry out the necessary anal-ycis f,)r these decisions (iii) the influence that research has had in influencing policy and operations (iv) the areas in which the knowledge gap remains substantial and (v) the implications of these various factors f-r future research. It is important to emphasize that this paper is not intended
either as a survey of literature jhich has been drawn upon extensively in the main report of the study) or of research methods available for analyzing specific problems, but rather as a way of examinin; the interaction of social science research with policy and operational decisions in rural development, and of exploring the factors that may increase the effectiveness of such interaction.
Major Issues Faced in Rural Development
One of the most basic issues currently faced in rural development is, whether and what set of objective criteria can be develcped for use

under a diverse set of conditions (i) to establish priorities in overall allocation of resources (ii) to determine sequencing and pharing of actjvitie c over time and (iii) to choose policy and institut onal interventions that will maximize the effectiveness of the resources allocated, if the given objective function is to improve welfare of certain socioeconomic classes, regions and sectors, by a given amount within a specified time period. For instance, under what conditions should development of agricultural technology and extension services receive priority over health services and literacy, if the objective is to increase welfare of certain socio-economic groups within a given time. frame? Or do thesn activities always have to be complimentary? And with what activities d such complimentaries end?
Further, what, if any, kinds of rural social structures and political and administrative institutions are particularly conducive t) increasing incomes of the lowest income groups? And relatedly, if certAin types of socio-political structures and asset distribution seem less effective in realization of the objectives than are others (as for instance was the case in Ethiopia in comparison with Tanzania until recently), should there be an effort to bring about the necessary structural changes? Alternatively, in allocating resources, should priorities be given to those social structures that .are more conducive to equalitarian distribution or to the others because these latter provide little hope for the improveme of the wellbeing of the lower income groups? And once resources are allocated to countries, regions, sectors and activities, what kinds of overall pricing, manpower, credit and technology policies are necessary for increasing incomes of the lowest income groups? Further, are institutions that, are on the surface

participatory such as Ujamaa necessarily effective in incieasi-ng incomes of the target groups? iAnd if so, in what time period? If not, what is the nature of trade-offs between political participation and growth?
These are, of course, highly complex questions involving very important ethical and socio-politicalvalues. No objective criteria exist by which to arrive at solutions to any of the above quEstions, even including those related to establishment of priorities betweEn agriculture and health that on the surface seen more easily amenable to ttraitforward economic analysis.
Our knowledge about these questions is based mainly o! subjective judgements derived from existing analyses of a number of pariial relationships, that almost always assume the basic economic and socio-political structures as given. The absence of an objective set of criteria for solution of these problems results from a number of factors. No rigorous conceptual framework exists for analysis of the relative importance of the factors that explain the phenomenon of rural poverty. The multiplicity 0o' such factors, the diversity in their importance among various rural environment ts, and the substantial interaction among them has made the development of such a conceptual framework difficult. An interdisciplinary approach that can
simultaneously take into account the economic, technological and sociopolitical factors also seems essential for such analysis. 0i r knowledge as to how to measure many of these variables is poor even whc:n their impact is known to be substantial as for instance in analyzing the 'effect' of the various aspects of Ujamaa policy on incentives, and welfare. Consequently, the data on many necessary variables are poor, as are the

methods available for their analyses..' Third, the supply of ,:ocial scientists who have both the ability-to capture the complexities.of the societies that they have to analyze,and to translate the observed relationships into a cohesive conceptual framework is extremely limited, even if there existed a useful dialogue between the theoreticians and the empirists, and even if they together were inclined to undertake. research that may involve prescriptions related to basic structural relationships. Both these conditions are. frequently absent. /
And finally of course, policy makers are rarely interested in
receiving solutions that touch upon the very basic fabric of the society. By the same token, if radical changes seen necessary in structLural relations,
1/ Social scientists of several disciplines have shown reluctance to undertake an explicit prescriptive function,and have instead confined themselves to analyzingand explaining the nature of relationships between their variables and development. Economists have taken a .nich more active role in this regard. However, until recently they have arr.-ved at prescriptions in isolation of the sociopolitical relationships, assumed that the solutions prescribed by them had few serious socioj olitical consequences or that even if the consequences were undesir;ble in the short run, as for instance the worsening of the distributioz- of income
with growth, that these effects waud somehow take care of themselves in the long run, either through an improvement in the relative income distribution or at least through improvement in the absolute incomes of the lowest income groups.

ns through the land reform and the villdgization currently underway in thiopia and Tanzania, rarely are policy maker,, willing to await a scientific analysis of the problem before arriving at such policy decision Social scientists.cani of course exercise useful influence on implementation of such policies as for instance by providing assistance in assessing how much land of what potential should be distributed to whom and under what legal and social arrangements both in Ethiopia and Tanzania. However, policy makers have frequently taken the view that such radical changes should not await formulation and implementation, if the resea:*ch results are not already in hand, particularly, given the existing state, both of the conceptual and empirical research in social sciences. Th:s view has very important implications for the nature of research to be co(hducted if social scientists are to be effective in improving the quality of planning and implementation.
To overcome the deficiencies in research it would seem that social science research may be aimed in two directions. First, social scientists may strive to improve the overall conceptual framework available for developing criteria for identification of complementarities and priorities;, and for sequencing and time phasing of activities. It is important to emphasize, however, that if such line of enquiry is to be 6f genuine operational significance as distinct from being an esoteric exercise that mainly leads to generating employment among fc]llow social scientists, considerable attention must be devoted to the trade-offs between si;implicity of ass umptior.s: and the empirical applicability of the conceptual framework ba.;ed on such assumptions, to the propriety of the choice of.variables and to their interactions, to the simplicity of methods for empirical testing of hypotheses, to the use. of critical minimum data and to the vays of cc!] reliable data promptly, given the shortage of trained mar.powTer and ir.ntLtutions within which such analyses must be carried out in tackling most oreraiional problems.

This is a tall order, and the possibility of i'1 ulfill1 1:V111 !t T tter only of conjecture.
The fact that at present social scientists have no ctematic criteria for anialy,-ing the 'big' questions of overall p i:ri ies, and choice of policies and institutions is, however, not to suggest that they cannot play an important role in improving the basis of these decisions in yet another and probably a far more effective .sy, namely through the microanalytical approach. Such analysis may be 'artial in the sense of examining a limited number of important ecenomi,, technological and administrative factors and the interactions them as they affect the realization of the rural development objectives within g:.ven sociopolitical frameworks (an approach followed in the ARDS), o. :-t may be partial in an even more restricted sense of analyzing only a particular. constraint, such as agricultural technology, or pricing policy.y, that appears to exercise an overwhelming influence on rural welfare. Par-.ial analysis can be effective despite the multiplicity of factors, because many such factors can be isolated, whose effect although important in realizationn of the objectives of rural development, may not be radically different among a variety of environments.
If rural development planning and ixnplementation is t be of a
participatory nature that draws on the 1Mowledre, experience and wishes of the local people, leaders, and administrators rather than. elitist, that relies on economists or planers from external agencies, then also of course, it is necessary to emphasize this second line of enquiry in sociall science research, to disseminate the knowledge of the processes of riral development broadly to such decision makers and thus to improve the quality of planning

and implementation of the programs.
A substantial, although highly fragmenta y, body of such micro research and information already exists on African rural development among various donor agencies, national governments and research institutes. The existing analysis and information may be classified into two interrelated categories:
I. A. Evaluation of past policy interventions to assess tie extent to which
1. they were effective in realization of the implicit or the stated
objectives of the national governments;
2. to assess the unforeseen consequences of these pclicies; and
3. to judge the efficacy of such policy interventions in realization
of the stated objectives in comparison with other alternatives _.
B. Evaluation of the effectiveness of institutional interventions: In
such research effectiveness is assessed by certain stated criteria
e.g., the effectiveness of decentraliaticn in Tan~ania rnd Kenya
in increasing: administrative efficiency, in improving the quality of planning and implementation, in local participation,. in resource
mobilization, etc.,.or the effectiveness and efficiency of the
cooperative societies in carrying out: certain credi. and marketing functions, in increasing local self-reliance, in augienting resource
1/ Numerous examples can be cited of this type of research coducted in social
sciences on rural Africa, as for instance the examination >f the magnitude
of the resources allocated to and actually expended in the rural sectors
through national planning in various countries, and cf the types of activities
to which they have been allocated, of the effectiveness of the specific
strategies adopted in the use of those resources, e.g., th, improvement approach
and the villagization approach followed in Tanzania in the '1Os and the '60s,
or of the effect of certain agricultural price policies on income stabilization,
resource mobilization or on production increase. See: Uma Kele Design of Rural
Development, for discussion of information on specific ishies.

mobilization, etc.
The evaluation of policies and institutions may be a posthumous exercise, undertaken with a view to improve the policy form nation and implementation generally in similar situations through increasing the insights developed from analysis of specific interventions, or it may be undertaken on a regular basis, as a device for identifying the constraints to realization of stated objectives, and for enabling the necessary.improvements in the course of implementation, to maximize their effectiveness Nuch of the research carried out by the Institute of Development Studies for
the government of Kenya on the Special Rural Development Programs at least in principle was meant to fall in this latter category. II. Description and analysis of the existing systems is the second type of research available on African rural development. The objective of such research is to -understand the basic. structure of the various traditional systems, to explain how they function, to examine their potential for realization of rural development objectives, to identify constraints and to derive implications for policies, planning of programs and for choice of institutions.
The existing research has numerous shortcomings, in terns of coverage of issues, data, methods of analysis and concepts, that are already well recognized. A substantial scope thus exists for carrying out research on a variety of aspects of rural development to overcome these shortcomings and to improve our knowledge 1. However, despite the shortcoming s and the highly fragmenta7ry nature, the existing information and analysis already provide a
1/ See: Uma Lele, Design of Rural Development, for detailed Ciscussion of
these issues.

- 12
sufficient basis for making intelligent judgemcnts as to the priorities in policies and planning.
The Iplication of the existing Social Science research ror further Action and Research in rural development in Africa
The existing analysis reinforces that the bu3k of the rut-al population in Africa is poor and, unlike in many Iatin American countries, this poverty is not confined to a few regions or clasies. It, is almost a lw ys spread over the entire rural sector. Consequently,. "targ-et groups" in Afr;ca are large relative to the financial resources and in particular the trai ied manpower and the institutional capability frequenLly available for levelopment.. Therefore, if the emphasis in rural development is to be on ma ss participation and on the vJiability of the process of rural development, in A rica rural development programs have to be viewed as part of a continuous, dynamic process, rather than as an 'extensive', or a 'maxirms' vs. a 'minimum' effort. The emphasis on mass participation also means that a sequential ap )-roach is frequently necessary in planning and implementing a rural development strategy, involving establishment of clear priorities and time phasing of activiti s.
Given the low productivity of the sub sistence rural sect)rj for a
variety of reasons analyzed in the ARDS, in many cases an initial emphasis on broadbased increase in productivity through a certain minimmin number of interventions and level of institutional develoTment to carry nut those interventions. appears to be a far more effective w:ay of ensuring viability of mass participation than the substantial initial concentration of resources through a large number of activities in a few regions.

Some constraints to improvement of productivity, such as
technology that is profitable at the farm level, effective extension and adequate and timely supply of inputs, a network of feeder roads and trained manpower are common to much of the subsistence agriculture in \frica, whereas others, such as ill-health caused by malaria, as in the squatter settlements in. Kenya, or the inadequate incentive system and organization as in Ujamaa villages in Tanzania are location specific. Therefore, no unique blueprint for planning is possible. The establishment of priorities needs a combination of a few critical general interventions such as those listed above, applicable to several regions, along with emphasis on development of capability of the regional administrations to identify and ameliorate additional constraints specific to individual regions. However, in all cases, including the cases
-of priorities established at the national level, the effectiveness of implementation depends on the efficiency and the coordinating ability of the regional administrations, and on the general institutional development at the regional and local level. This is why a regional focus in in titutional development is necessary from the outset even when a broad coverage of services is being aimed at. 'Ibis is particularly so if more 'omplex programs involving a number of sectors are to be planned and implementd by regional administrations over time.
Frequently the first step has to be the improvement of %he regional administrative capability for effective planning and implementation of programs directed at only a few productive activities, including food crops. Attention to food crop production often seems critical since a majority of the low income population derives its livelihood from this activity.

Because the existing methods of analysis in social sciences do not offer a basis for improvement in the establishment of priorities over the intelligent judgements arrived at on the basis of past experience, it
seems desirable to allocate the limited research capabilities to improvement of the effectiveness of the implementation of the actual priorities and for identification of, and planning and implementation of, interventions for removal of the location specific constraints.
To achieve this end, three types of research seems in order.
1. Analysis of Overall Policies
The past experience indicates that if the priority activities are to have a significant impact on the lowest income groups, the particular country's existing sectoral policies and plans, as .well as the indigenous .institutions available for rural development have to be examined explicitly to asses. the extent to which these effectively "reach" the lowest income groups in the rural areas. Such analysis will allow over't recogniticn of the existing government policies which are inconsistent with the goals of rural development.
This information is necessary either to develop possible ways of bringing about the necessary changes in policies and of ensuring their effective implementation or alternatively to modify design of programs to cope with the existing policies. For instance, pricing policies, particularly in regard to food crops, have frequently adversely affected both the distribution of benefits to the
lowest income groups and the incentives for production beyond the very minhnim domestic needs. In such cases a change in pricing policies is necessary for improving the benefits of a spec afi.c rnu'al d1evelonTment program. Alternftively a more modest increase in market sirlusn. reCdS to be planned for uch questions were not analyzed sufficiently prior to IaInning and

inncrlementtion of many of the past program. 'T rir etielie has re_ultedl in numerou,' IifficIlties during the course of implEi nt:tion. TI has, of course, also meant only a very limited imp ict on the target population. II. Research Related to Plannin and .lementation of Sp:ecific Activities.
The experience indicates that priority activities themselves require substantially greater planning if their effectiveness is to be maximized. To achieve this objective research needs to be directed towards resolution of a number of pertinent questions. For example, w-hich particular technologies are actually profitable at the farm level, how well do they fit into the existing cropping.pattern, and, therefore, ,which specific crops are to be selected for promotion? Do existing marketing systems serve low income farmers in the program areas effectively? Given the social ties that often exist between peasants and traders and given the extreme scarcity of trained
manpower available to implement market interventions, will a new .marketing system which may seem desirable in principle actually benefit the lowest income groups? Or will it only aggravate the tensions between the cultivators and the merchants, with adverse effects on the cultivators as has frequently been the case in the past programs? In sum, what really are the constraints to increasing agricultural productivity in a specific case and what steps are necessary for effective implementation of the interventions if these critical constraints are to be ameliorated? How to develop the necessary implementing capability? And what is Lhe time horizon implicit in the results to be expected from the program? To answer such questions planning of programs will of course require considerably greater and direct support of operationally oriented research than has been the case in the past. In a few cases, where an effective technical and institutional capability to coordinate and deliver such services already exists, the first phase of research and development outlined above may of course be skipped.