Cover Image
close this book Laying the foundation - The institution of knowledge in developing countries
close this folder Part I The institution of knowledge
View the document Chapter 1 Institutions and their context
View the document Chapter 2 Institutions of knowledge as formal organizations: an initial balance

Chapter 2 Institutions of knowledge as formal organizations: an initial balance

Hernando Gómez

It is hard to dispute the notion that teaching people to fish will be more useful to them in the long run than giving them fish. Likewise, building a solid research capacity in Latin America will be more beneficial to the region over the long term than supporting individual research projects.

Yet, national and international donor agencies have long been biased toward funding individual research projects in developing countries rather than the institutions that conduct them. There are two reasons for this The first is that we know, or think we know, more about designing and evaluating research projects than we do about building and strengthening research institutions. The second is that we feel we can exert more control over the success of a research project than over the success of a research institution, whether or not this is in fact the case.

In other words, doing the appropriate things is more adequate than doing things appropriately. This consideration would be sufficient for going back to the theme of institutions; however, our knowledge of them is more limited than our knowledge and skills on the management of single research projects.


Research on Institutions: The State of the Art

The capacity to promote the development of research institutions requires a thorough understanding of the institutions themselves. Granted, we do know some things about institutions, which we also refer to as complex organizations. In fact, we have many insights and many convincing (if partial) explanations.

The literature on institutions covers many different types of institutions and employs theoretical perspectives from a wide range of disciplines. The institutions examined include public firms, government agencies, nonprofit organizations, community institutions, and social movements with a certain degree of rational bureaucratic organization. There is also an extensive anthropological and historical literature on institutions as "recurrent or internalized interaction patterns." (We refer here to literature on the institutionalization of science, which is closer to the anthropological definition of institutions and which is helpful in identifying the contextual or macro factors that determine the success or failure of research institutions.)

At the theoretical level, we can distinguish between approaches that characterize the organization-as-machine versus the organization-asculture, as well as perspectives that place an emphasis on explanation versus an emphasis on management. Disciplines range from the formal legal school (of Comte and Maine) and the industrial engineering school (of Taylor) to business administration, public administration, and social science perspectives, as well as to cybernetics, systems, and chaos theories.

Many of the central concepts of institution-building theory, although understood by most researchers, are still somewhat vague. For example, the terms "institution" (or organization) and "building" (or development) can be understood in different ways, and what seem to be minor nuances may be magnified when used in lengthy, complex, and self-referring texts. This ambiguity is exacerbated by the lack of definitional precision (in epistemology, methodology, purpose, and specific subject matter) that characterizes so much "soft" social research.

The literature on institutions also tends to rely heavily on paradigms. Some of the most influential paradigms used in the description of complex organizations include technological division of labour, hierarchical division of labour, information networks and flows, social control, face-to-face interaction, rational self-serving individuals, efficiency maximizing, and shared value-enhancing. This abundance of paradigms or quasi paradigms - which sometimes compete or overlap with each other - has fuelled efforts to develop superior languages, or metalanguages and general theories. However, many of these have a comparable degree of apparent elegance and coherence, and the problem is slightly biased toward the sphere of metaparadigms.

Hence, state-of-the-art theory can be described as a set of different-order paradigms and subparadigms, each of which selects an independent or major explanatory variable, develops a typology of institutions according to the main categories of the chosen variable, and explains or predicts a range of dependent variables or structural and behavioural features of the identified institutional types. Table 1 illustrates some of the most commonly selected variables.

Table 1. Some commonly selected variables.




Main categories




Specialized diffuse








Individualistic- collectivistic

Formal-informal networks



Placid-competitive systems



Bureaucracy is too important a part of everyday life to be merely a topic of theoretical controversy. A literature has evolved that focuses on complex sets of partially overlapping, partially excluding paradigms of strategic intervention in a wide variety of organizational contexts. These approaches are broadly framed within one or another of the theoretical paradigms. They also tend to select as the independent variable the easiest one to manipulate.

Some of these variables include administrative relationships (such as organizational charts, fluxes, and processes); legal frameworks (especially for public institutions); incentives, time, and motion (the industrial engineering perspective); planning and monitoring systems (such as quality-control techniques and strategic decision-making); leadership styles (such as management by objectives or participatory management); the human factor (T groups, Z theory); the environment (interinstitutional awareness, strategic linkages); and the role of donor agencies (the blueprint model, logframe, second and third generation, direct support, and learning process approaches).

Building Research Institutions in Latin America

General theories on institutions can be used to identify some of the factors that might affect the success of research institutions in Latin America. The best way to identify helpful questions is to focus on those aspects of general theory that are most relevant to the subject under consideration. This discussion will now turn to the topics of institution building and research institutions in Latin America.

What constitutes a "built" institution depends to a large extent on what is meant by "building." Is it an institution's ability to survive, grow, and undertake new programs, or its ability to contribute to overall social well-being? Is it the capacity to satisfy relevant actors (owners, politicians, employees) or to maximize rationality and achieve stated goals? Is it the capacity to change the environment or to adapt to it? If institution building is recognized as a multidimensional concept, as it should be, how do we measure components such as effectiveness, efficiency, morale, and adaptability?

Institution building should be recognized as an extremely relative concept. Max Weber ( 1947) theorized that the bureaucratization of social life resulted from the rationalization of Western society. According to his theory, if the formal or bureaucratic organization of human activity is viewed as a mechanism for achieving collective goals, then an institution can be considered more or less "built" when it achieves those goals. It is also known that some institutions achieve their goals more efficiently than others. Price (1978, pp. 203-204) provides an excellent summary of factors contributing to the success of institutions:

Organizations which have the following mechanisms are much more likely to have a high degree of effectiveness than organizations which do not have the mechanisms:


The organization's economic system should be characterized by: i) high degrees of division of labour, ii) specialized departmentalization (except where there is a high degree of complexity), iii) mechanization (except where there is a high degree of professionalization), and iv) continuous systems of assembling output.


The organization's internal political system should be characterized by: i) high degrees of legitimate decision making, ii) rational-legal decision making, iii) centralization with respect to tactical decisions (except where there is a high degree of complexity), and iv) maximum degree of centralization with respect to strategic decisions.


The organization's external political system should be characterized by: i) a high degree of autonomy; ii) an ideology with high degrees of congruence, priority, and conformity; iii) cooptation; iv) major elite cooptation; v) a high degree of representation; vi) major elite representation; and vii) a major elite constituency.


the organization's control system should be characterized by: i) a high degree of sanctions; ii) a norm enforcer-norm conformer relationship which is basically secondary; iii) a sanction system with a high degree of grade; iv) a collectivistic sanction system; v) high degrees of vertical communication and horizontal communication; and vi) a communication system which primarily is instrumental, personal, and formal.

This organizational profile may be understood as a general version of the specificity that Israel (1987) identified as one of two basic incentives to performance in a recent World Bank survey of 159 developing countries. The other incentive - competition - refers to an institution's relationship with its environment and the pressures brought to bear upon its personnel (Table 2). In short, there are good reasons to believe that complex organizations are more likely to attain their goals when they meet the criteria of specificity and competitiveness as set out here.

Table 2. The second incentive: competition.


Predominant category of pressure





Economic competition





Competition surrogates






















"Building" is a purposeful, voluntaristic activity in which deliberate effort is made to strengthen an institution. Thus, although specific, competitive organizations are easier to build than their nonspecific, noncompetitive counterparts, there are still certain features that can be manipulated to increase the probability of success. These more controllable variables are the subject of the specialized literature on institution building. Milton J. Esman's Institution Building Concepts (1967) remains the classic in this field (Table 3).

Table 3 Esman's classic model.
















Several sophisticated measuring techniques have emerged in the literature on institution building based on Esman's concepts (see Bjur 1983). So have sets of bivaried hypotheses (Duncan 1975) and detailed action guidelines (see, for example, the 38 strategies recommended by Derge 1968). This specialized literature is extremely pertinent to our analysis of scientific and technological research institutions in Latin America.

At the heart of bureaucracy is a dilemma that has been discussed in many classic texts. Two of the key determinants of institutional success may in fact work at cross purposes. An institution's success in achieving collective goals depends on its ability to reduce uncertainty by dividing complex processes into routine components. At the same time, an institution's competitiveness depends on its ability to create an environment that motivates and stimulates its personnel.

It is clear that scientific research is one of the most uncertain of endeavours. There is no such a thing as an "art of invention" (or ars inveniendi), although natural philosophers pursued it until the 18th century. This means that formal organization, although still necessary, is less likely to be crucial to modern science than to other fields such as industry or the military. Furthermore, the "soft" organizational style of most research institutions is less conducive to success. (We might conclude that bureaucracy does kill creativity.)

Hence, it would appear that it is the control and external political systems that make or break research institutions. The evidence does suggest, although not conclusively, that both motivation and environment are critical to the success or failure of scientific organizations.



The creation of knowledge cannot be monitored, hired by the piece, or externally controlled like many other bureaucratic activities. Rather, it is a unique and absorbing process that demands the intense, voluntary dedication of the researcher.

Historically, values and motives have played a vital role in scientific development. Here are some examples from different contexts:

· The social value placed on scientific knowledge (see Bernal 1974);

· The impact of specific sets of cultural and religious values on scientific attitudes (see Merton 1957);

· The importance to scientists of peer recognition (see Stein 1982);

· The finding that scientific creativity increases in an atmosphere that "avoids isolation and domination and provides frequent stimulation combined with autonomy of action" (Pelz 1976); and

· The psychoanalytical hypothesis that scientific vocation is a "neurotic search of substitutes to aggressive and sexual instincts" (Kubie 1954, p. 109).

For our purposes, the most important motivational issue surrounding scientific activity relates to institutional incentives. This refers to the way an organization's system of rewards and punishments encourages or discourages certain behaviours. Many research institutions in Latin America appear to reward behaviours other than scientific creativity. Examples include teaching ability (or is it popularity among students?), conformity (or is it skill at playing in-house politics that can exert pressure not to produce, lest colleagues lose face), and the popularization (or is it "ideologization"?) of knowledge, rather than its creation. It is critical that an institution's reward structure be well understood before it can be changed to better support scientific creativity.



With respect to institutional environment, Price's emphasis on cooptation (see his description of the external political system) seems to conflict with Israel's emphasis on competition. However, what appears to be a contradiction also suggests a general hypothesis that could be used to guide research on scientific institutions. The hypothesis is that the success of a research organization requires both a high degree of social legitimization (Price's emphasis) and a high degree of competition among organizations and research staff, which serves to provide momentum (Israel's emphasis).

This hypothesis can be demonstrated with historical and cross-sectional evidence. Here are several important contributions. Israel (1987, p. 132) has described political commitment (social legitimacy) as "an essential ingredient in the success of an institutional development program" and the critical influence of "macropolicies" on institutional development. Ben David (1960) has provided a classic analysis of how decentralization and competition advantaged German and American medicine over the British and French models. Schultz's (1975) work has shown how institutions rise and fall in response to changing social demands. Powelson (1972) has explained how leading elites create new institutions to "model" development.

Therefore, it is suggested that the following environmental issues be placed on the research agenda:

· Political commitment to scientific development;

· Macropolicies and national strategies affecting science;

· Interinstitutional competition for research resources (recognition and prestige, human, and financial resources);

· The social demand for science and its products; and

· Roles and perceptions of national elites concerning science.

These should be accompanied by an assessment of the institutional effectiveness of external systems relating to the environment (such as information gathering, planning, marketing, feedback collecting, and extending linkages)

It could be argued that institutionalization is just another name for development. However, institutions in developing countries have only begun to be studied. Little attention has been paid to research institutions and even less to the factors identified here as critical to their success. Because the guidelines suggested in this chapter have been derived from the experience of developed countries, they require careful adaptation to be meaningful to the Latin American context.

Of course, science is universal (although scientists are not, as Pasteur has reminded us), and scientific production is subject to an international division of labour. This brings us to a tricky question, which is both empirical and normative. What kind of scientific research do we want for Latin America? Is it realistic, or indeed wise, to aim for high quality, pioneering scientific research in our countries? Or, should we concentrate on borrowing and adapting from others, taking advantage of our position as late-comers to the field? I lean toward the second approach with several qualifications. First, there should be a large enough pool of highly trained scientists to ensure a scientifically minded society, proper teaching of the basic sciences, and the efficient adaptation of world knowledge. Second, the arts, humanities, and social sciences should remain unique to our countries because culture and nationality do make a difference. Third, another exception is science or technologies capable of playing a strategic role in national development in areas such as agriculture, tropical medicine, and the new industries of knowledge for the world market.

Another key question is whether an institution's success should be judged on the inherent intellectual value of its products or on the potential of these products to contribute to the transformation of the natural or social universe.

This leads to two more variables that affect institutional success. The first is the type of research design normally employed by an institution. Van de Vall and Bolas (1978) have shown that some research designs tend to yield more applicable results and others tend to advance substantive knowledge. The former are broad in scope, employ ideographic concepts, strive for less logical elegance, and have a wider implemental span. The latter have a narrow scope, employ nomothetic concepts, tight logical deduction, and shorter implemental spans.

The second concerns institutional networks. Formal organizations interact with other social actors in complex and unique ways. Relationships may be complementary, competitive, or regulatory. Institutions dedicated to pure and applied science establish linkages with other institutions that can contribute to their success. (To confirm the importance of this variable, we have only to observe the differences among research units that are attached to universities, attached to private enterprise, or are independent.)

The question of pure versus applied research keeps bumping up against the larger issue of Latin America's proper position within the international order of science. Science is a circuit of many steps and institutions that endlessly (if partially) feed upon one other. These include basic research, technological research, the training of new scientists, the development of academic and professional communities, and the diffusion and production of goods and services.

The circuitous, multistage nature of science has two major implications for the agenda proposed here: when evaluating and explaining the success of a given research organization, consideration should go beyond national institutions to international networks of associated agencies. For example, it could be that the cost-benefit ratio for a host country turns out to be quite different from that of a third country and the wider international community.

A nation's capacity to make truly significant contributions to science depends largely on the critical mass of cultural, human, and financial resources that it can devote to the effort. Science is a long-term process that is nourished by patient trial and error, by diverse and heterogeneous approaches, and by accumulation, feedback, and acceleration. Without such a critical mass, a country is in no position to fully use the fruits of knowledge.



The existing literature on complex organizations, particularly that on the institutional development of Latin American scientific and technological research institutions, suggests that priority should be given to the following institutional components:

· Specificity (including types, degrees, determinants, and implications);

· Competitiveness;

· Values and motivational structures of the scientific community (with an emphasis on institutional incentives);

· The relationships between an institution and its environment (including legitimacy, competition, and external systems);

· The basic orientation of research (including research designs and institutional networks);

· The position of an institution in the international scientific order (including relevant networks of world institutions and its stage in achieving a critical mass of resources).