title

Document published in Torres Adrian, Mario. 1993, Research on Social Policy: Proposals for a Future Agenda. Edited by Mario Torres. Ottawa: International Development Research Centre. http://www.idrc.ca/socdev/pub/documents/torres.txt

The recent history of Third World development has been stimulating in some cases and has recompensed collective and individual efforts. Unfortunately, in many other cases it has been an experience in collective frustration. The case of Latin America is illustrative; it is said that the region's efforts during the decade of the 1980s were lost. The economic crisis, the impoverishment of the lower- and middle-classes, the dismantling of major social programs, environmental deterioration and violence have shadowed achievements in the democratization of political regimes and in relative social mobility and cultural integration (ECLAC, 1990a; World Bank, 1990). Worst of all, the path to recovering lost ground seems to have vanished.

Some proposals have begun to emerge, and many measures to be implemented. These include such programs and strategies as: "changing production patterns with social equity" (ECLAC, 1990b), "education for all" (WCEFA, 1990a), and "sustainable development" (ECLAC, 1991). While clarity exists with respect to the new economic models to be followed, proposals in the social arena have yet to be defined. This is not difficult to explain, for at base is the crux of historic change: human development. The response to the changes that Latin American societies require for advancement must not only be scientific and political, but also profoundly ethical.

The present document seeks to discuss some aspects of this complex topic, taking social policies as an entry point. After all, it is social policy that, through action or omission, will forge new generations, the only real resource that Latin America and the Third World have to achieve a better future. In this respect, we do not intend our discussion to focus on human development models or social policy content since, this being our first assumption, there is basic agreement about their final goals. Instead we are aiming at examining obstacles to the implementation and evaluation of new policies, since, this being our second assumption, the most critical issues concern the actual mechanisms for social interventions. This position does not deny the importance of conceptualizing human development. Rather it points to the fact that present circumstances demand such conceptualization to be developed along with practice. Thus, we will deliberately leave this issue pending, in the hope of stirring up reflections based on experience rather than on theory. Our discussion opens with an operational definition of what is to be understood by "social policies," and is followed by a theoretical proposal to be analyzed. We will conclude with the review of a set of problems related to these issues.

The conceptualization and practice of social policies cannot take place outside a wider reflection of the many aspects implied in Third World development, otherwise, the context for their design and evaluation would be lost. It is questionable, for example, to consider social policies from a strictly sectoral perspective, or as a complementary resource for other policies oriented towards economic adjustment or productive restructuring. More appropriately, they should be considered as a strategic part of a set of development policies. There are ethical reasons for this viewpoint, in addition to very pragmatic reasons: only human development can guarantee in the long term that actions taken in the economic, technological and natural resource management spheres can reach their objectives.

The conceptualization of social policies is in itself a difficult exercise, for not everyone understands them in the same way. In this document, social policies are understood as actions designed and implemented by the state, or with state support, to achieve personal human development. Typical social policies involve activities that seek to promote human survival, early child development, nutrition, integral education, learning capacities, family development, the prevention of health problems, social security and social participation (United Nations, 1989). These activities are not the equivalent to social services, which are the instruments of social policies. Nor do they refer to all social aspects of development. They do not, for example, include the management and reproduction of resources whose value is based on scarcity, nor the reproduction and conservation of environmental resources. Specifically, they relate to strategic actions at crucial moments and aspects in the life of a human being, which may be attended to within a framework of activities carried out by social institutions.

There is no definition of human development that is universally accepted, but there does seem to be consensus that personal development cannot exist where there is poverty, unemployment and inequality (MacPherson, 1982; UNDP, 1990). The persistence of large sectors of the population who are unable to satisfy their basic needs in health, education, employment, housing or recreation, who suffer from a lack of material and economic opportunities for survival, and who do not enjoy the same living conditions and opportunities as other sectors, has made it evident that economic growth, wherever it has occurred in Latin America, has not been sufficient to bring about personal development. Human development, in all its essential aspects, is still a goal to be reached by large population groups in the region.

It is important that this concept of human development, understood here as "essential development", be considered again in Latin America. In the face of the failure of economic policy actions taken in the eighties, the need to reorient the development process towards productivity, international competition and the incorporation of modern technology has been proposed (ECLAC, 1990b). Although the search for a development model based on these principles may be correct, social policies must not be placed in a complementary or remedial role. If this were to occur, the economic proposal would fail for -as the history of the post?war development in the Third World has shown- there is no economic development if there is no human and social development. The search for Latin America's new path must begin by centering discussions about development policies on the individual and collective aspects of human development. This, however, will not be an easy task in the 1990s.

The decade of the nineties was marked at birth by a deep crisis in the socialist utopia, which had apparently chosen a more humanist development model than that of Western capitalist societies. However, the crisis of socialism in Europe indicates that the policy instruments used did not lead to the levels of social and economic development necessary for success. The influence of this fact on the conceptualization and implementation of new social policies in Latin America promises to be remarkable. Although attempts may be made to limit discussions to technical issues, it will be impossible to avoid the ideological debate and the impact of political party repositioning in the wake of the decline of European socialism.

The new social policies will also have to consider the persistence of mechanisms of dependence and domination. The economic, industrial and technological development thus far attained in Latin America has not eliminated the conditions of dependence so widely discussed in the region during the fifties and sixties. The impact that dependency has had on the collapse of Latin America's social programs -as deterioration of commercial exchange terms and foreign debt payment made them even more unfeasible or precipitated their dismantling- cannot be denied. Discussions of this situation appear to have lost their timeliness and dependence has apparently disappeared beneath a cloak of greater economic integration and the emergence of sub-regional markets. Yet dependence has not disappeared. Moreover, dependence and domination will surely continue to affect the design and implementation of new social policies. The social implications of the search for pragmatic relations with the developed world will be important to follow. What will the social impact be of the proposed regional model for an open economy?

To all the above concerns must be added the fact that class structure, although somewhat transformed, continues to feed large social and regional inequalities. The domination of capital cities over interior regions has not disappeared. The concentration of wealth has not notably diminished, and in some cases has increased (ECLAC, 1990a). In addition, the state has not improved its efficiency or effectiveness. To the contrary, the state has become a heavy weight on societies who are already exhausted by populist, party-tied and even irresponsible management. Underdevelopment processes have not weakened. Although transformed, they continue to exist and have become stronger. Current and future social policies will undoubtedly be affected by all these circumstances.

There is hope that the new economic policies will create a more favorable institutional framework for redistributing the benefits of development. This hypothesis is not really new. Development planning, which began after World War II, was basically an exercise of economic planning in which social aspects were treated within a residual model. It was expected that the benefits of development would reach the entire population as a byproduct of industrialization, greater employment, urbanization, and increased per capita income. With higher incomes, families would be able to meet and solve their basic needs. The state need only intervene when that was not possible, such as in the case of abandoned children, the elderly, or the socially marginalized. In the sixties, however, the social aspects of development began to receive growing consideration. Education, health, housing, family welfare and other issues were paid more and more attention because the residual effects of economic development were proving insufficient. They were also paid more attention because, for ideological and populist reasons, governments became involved in promoting rapid social modernization (Hardiman and Midgley, 1989). In many cases this position led to the disorganized growth of social services, chronic fiscal deficits and a populism, which, in the long run, also contributed to the social services crisis in the eighties. In addition, the implementation of social policies was so centralized that they were identified as programs of whatever government was in power, not as state policies, which rendered them even more vulnerable to change or dissolution.

It is now argued that the introduction of competition and privatization will allow new social policies to be more efficient and effective (Irarrázabal, 1990). However, this may not necessarily equate to human development. It will therefore be necessary to investigate how these principles can operate within the context described above in order to achieve essential development.

It is possible to distinguish several roles for social policies: remedial, preventive and developmental (MacPherson, 1982; Dubey, 1980). Emphasis placed on the need to minimize the negative social impacts of new economic policies may lead to a markedly remedial concept. It is necessary to insist on the fact that the new social policies cannot be reduced to assistance programs for abandoned children, the elderly, the marginalized population, or to emergency programs for the hungry and unemployed. Social policies have to go further: they should be instruments for the development of human beings, so that they may reap by themselves the benefits expected from the new economic model.

An important feature of human development problems is that their solutions are not exclusively "social" (Boeninger, 1982). Technological, economic and environmental ingredients are all required. For example, early childhood stimulation requires the adequate training of mothers to handle mother-child interactions. It also requires a minimum of technology, income or employment conditions that allow mothers time to interact with their children, as well as favorable social and natural environments. Social treatments require psychological, social, technical, economic and environmental resources. As a result, solutions are difficult to design and implement. From a scientific perspective, it means that the treatment of human development problems is an extraordinary challenge that can only be confronted through an approach amenable to multidisciplinary, multidimensional and inter institutional research and social intervention.

A systemic approach could be useful for this purpose. Already applied in several fields for the analysis of phenomena from different angles and perspectives, the systemic approach permits the capture of an array of inter-relations and interactions. In the field of social sciences, this theoretical proposal has been intrinsically linked to the thoughts of the structural-functionalists in the sixties and to the ideological and theoretical controversy the proposal generated. Nevertheless, it would be worthwhile to reexamine the systemic view as a tool for the implementation and evaluation of social policies. It may prove to be fruitful at a moment when the dominant paradigms of social sciences are in crisis and a large number of partial generalizations abound. To this end, the systemic approach could be provisionally understood as a theoretical modality that allows the conceptualization of a set of behaviors, values, technologies, and environmental resources that affect human development from different angles and at different moments.

The need for a systemic analysis framework has become evident with the multisectoral character of classic social problems. Education, health, social security, employment, recreation and community development cannot continue to be sectorally treated. The most recently accepted positions in world forums and endorsed by governments, mention for example, that efforts to reach the goals of universal health and education should begin with actions by diverse social groups and by different sectors of the state (WCEFA, 1990a, 1990b, PROMEDLAC, 1991). The newest academic and research approaches follow the same lines. Theory, research and practice suggest the need for multidisciplinary, intersectoral and inter institutional approaches. However, the social sciences do not seem sufficiently prepared to meet this challenge.

During the past decades, the region's predominant social science paradigms were not oriented towards social engineering, but towards the diagnosis and critique of current social structures and development models. However, recent historical circumstances are leading towards greater pragmatism. The differences between capitalist and socialist development models are no longer as clear as they once were. Discussions now tend more towards the "how" of doing things to reach greater efficiency and effectiveness. These new attitudes and requirements demand clear social intervention models. The problem is that these types of resources are not available. The proposal here is to complement the systemic approach with the social engineering concept, not with the idea of reducing social policies to technocratism but of presenting the need to tackle social problems both in their immediate and long-term implications. A case that illustrates this point is that of family planning. Here, not only do the daily emotional and sexual aspects of a couple have to be considered, but also the long-term impact of their decisions on family organization and upon the size and structure of the population. Although the proposal may generate some disagreement, there is nothing in principle opposing a reconceptualization of the idea of social engineering within the context of present needs and changes.

The task is methodologically complex because there is not always consensus on what human development needs are (Irarrázabal, 1989). Societies, particularly the underdeveloped ones, are characterized by the existence of heterogeneous interests. This is particularly so in Latin America, where inequality has increased. Research based on neo-positivist assumptions has thus far been unable to capture this diversity of interests. Neo-Marxist orientations, on the other hand, are too general and have tended to oversimplify reality. Apparently, research on social policies will have to be epistemologically "conciliatory", if something like that is indeed possible.

Cook (1985) proposes that scientific research on this area be oriented towards "multiplism", taken to mean the execution of studies converging towards the same theoretical objective, despite each study's differing data sources. Multiplism can be considered as a call to research with different actors, diverse data analyses, competing interpretations, multiple definitions, varying methods, different research tasks, multiple causal models, competing hypotheses, and analysis of the generalization of results over diverse populations, time periods and contexts. It is obvious that one of the major problems of this proposal will be the integration of different valuative perspectives, hypotheses and results. Multiplism should not be seen as a panacea, but as a working perspective in tune with the systemic approach required in the development of new social policies.

Independently of whether the epistemological and methodological proposal of "multiplism" is accepted or not, it is clear that research on new social policies will have to innovate upon current valuative, conceptual, methodological and institutional frameworks. The remainder of this document discusses the diverse problems to be sorted out in the areas of data collection, the elaboration of diagnoses and evaluations, the implementation of interventions, and the human resource training required to face this formidable challenge.

The implementation and evaluation of the new social policies will require information, which responds to new multidisciplinary and intersectoral perspectives. Creativity will be necessary to assure that data collection and analysis is timely, reliable and useful. Although this may sound like a simple task, these requirements present a great challenge given the current situation of scarcity of all sorts of resources in the institutions concerned. It will be difficult to rely on traditional methods such as national censuses and surveys, and complex multivariate analysis. This is not only because they are insufficient, but also because they are costly and difficult to process. In most countries, administrative decentralization and regionalization will require instruments that can be managed by individuals working in interior regions. In addition, democratization processes have created an important opportunity for participation by local organizations and communities. Within this context, the centralization and authoritarian control of information will increasingly be less viable, despite the fact that for the time being, decentralization will not necessarily mean greater autonomy in decision-making. Although decentralization is being actively sought, as in Chile and Colombia, it is necessary to consider that central governments will still retain a great deal of control, and that the habits and inertia imposed by daily work routines will persist for some time.

The multisectoral execution of social policies should be more viable at the local level given the closeness of this level to beneficiaries. Information generation and exchange may possibly be facilitated by mutual support between local government officials, the private sector, and community leaders. Nonetheless, this road will have many obstacles.

Much of the information, which is periodically collected in Latin America, is lost because it is not readily accessible. In some cases it is not processed; in others, administrative obstacles makes it inaccessible. In several countries laws forbid access to primary sources of census data. Although this may seem surprising, there is not always a clear objective for gathering information. Much data continues to be collected because it is legally required, but once collected, it is simply piled up and stored away. One example of this concerns the case of data on migratory border movements. Resources to make use of what already exists are insufficient; when new resources appear, they are usually earmarked for items that cannot be extracted from the previously collected material. Another serious problem relates to governmental administrative periods, including those of ministers and directors, which affect data collection and processing. Depending on the length of the period, it is common to see programs paralyzed every four to five years. As the state's research capacity constantly decreases (and the same happens with action-oriented NGOs), evaluations cannot be performed. Furthermore, there are no channels to disseminate what work is being done. Even internally, within a single government agency, differing departments have no knowledge of each other's work. In planning agencies, all work is duplicated with every administration or director. And, in addition to this waste of resources, information does not reach users. Teachers, social workers, physicians, judges, attorneys, etc., continue to work blindly; the only concrete data they receive is related to new budget cuts.

The problems are legal, political or administrative; they are also technical. Information on small towns (populations under 100,000) is not collected during the periodic household surveys that are made in several countries because the sampling frameworks do not produce data, which can be generalized at this level. This situation occurs not only in rural areas, but also in poor marginal areas of urban centers. One of the challenges will be to create data collection systems that will allow the strategic organizations in these areas -schools, health centers, municipal authorities, development corporations, professional organizations, etc.- to have a minimal flow of information. The involvement of the private sector, especially of private enterprises and NGOs, in the task of supporting research and generating information needed for regional and local planning should also be considered, for they tend to be much closer to the problems than are central government agencies.

Whatever means are found to link the few or many existing human and institutional resources at the local level, it will also be crucial to have social indicators that monitor efficiency and effectiveness. A major difficulty is that currently available methodologies produce indicators which are too general and which are unable to identify the sources of problems relative to, for example, persistent school drop-out rates, adolescent pregnancies, child abandonment or unemployed youths. It will be necessary to examine how to identify target populations, how to obtain information in a rapid and valid way, and how to reach beneficiaries. For these purposes, it is not only necessary to conduct research from a substantive perspective, it is also necessary to carry out methodological research. Unfortunately, this type of research, closer to social engineering, is not yet popular.

Another great challenge will be to bring the benefits of the technological revolution in the field of informatics to local communities, organizations, and governments. Access to the most modern technology in this field is no longer an unreachable dream. Probably this is the moment to reexamine basic assumptions with respect to its use, and to recognize that the instruments needed to resolve problems must be placed in the hands of beneficiaries, since the state is increasingly unable to respond. Advanced scientific research must not only provide viable methods for collecting necessary data, but also user-processing systems. Union leaders, youths, mayors, health workers, school teachers and others must be able to have at least a minimum level of understanding of the significance of such information for improving employment, health and/or educational quality in their communities.

In several countries, municipalities have been granted the right to make social policy. This makes it necessary to generate data and processing programs to support municipal activities. A serious difficulty at the local level arises from the fact that macro-level sectoral policy objectives differ. It is not uncommon to observe that high?risk populations are differently defined by the health, justice or educational sectors. In practical terms, this means that although the health promoter, teacher, judge and agricultural extension worker sit at the same table and although they understand their daily local problems, their capacity for action will be limited because they work within different administrative frameworks, with different policy definitions and with data bases that are neither complementary nor cumulative. Under these circumstances, how can programs be monitored and evaluated? How can experiences be complemented?

Some of the ideas currently being explored include: municipal information systems that can integrate vital statistics, census information and continuous statistics; the elaboration of programs that will allow the desegregation and manipulation of census data at the level of small geographic units (Conning, Silva and Finnegan, 1988); data banks on local and regional investment projects for the operation of development corporations; the installation of health alert systems based on methods that collect and process data on infant mortality in a rapid and reliable manner (Guzmán, 1988; Arretx, 1990); programs to produce national and sub-national population projections (ECLAC, 1988); and the development of indicators on educational quality and relevance. All these ideas are oriented less towards data collection than they are towards the development and implementation of information systems that will allow data to be collected and analyzed in a manageable and timely manner. The current circumstances require accessible, user-friendly and reliable procedures. Although it may seem surprising, this will be a very difficult test for the methodological imagination of social scientists.

The identification of beneficiaries is not a simple task because social policy objectives are usually formulated in a very general manner. Social crisis and acute social inequity in the region have led to the proposal of focusing social policy efforts on groups in extreme poverty and need. This has been the philosophy of the Chilean regime, and one, which has pragmatically tended to become generalized throughout the region. It should be noted that the proposal of organizing social policies around intervention targets should be thoroughly discussed, since many social problems tend to have their roots, not only in individual characteristics or local circumstances but also in the way in which the distribution of development benefits takes place. Some typical examples of the important structural determinants of poverty are income distribution, land-holding structures, and property systems of urban lands. Historically, experience has shown that policies, which seek better income distribution, tend to benefit society as a whole, and lead to development. One problem that Latin America has yet been unable to solve is extremely unequal income distribution. Discussion of the role of these factors in social development should not be set aside, but included in the desire for effective social engineering.

The multidimensional character of social policy impacts makes it difficult to identify all but the most direct and immediate of all possible beneficiaries, as well as any possible negative impact. This last aspect includes the identification of those who might be prejudiced, since social policies will no doubt affect the interests of some or many social groups. Social research based on a systemic approach would provide more possibilities to examine this diversity of positive, neutral and negative effects (Miller and Fredericks, 1987). Similarly, it is crucial to understand that the beneficiaries of some social policies may not be those of others, and that while the interests of a single individual may be well met by some actions, they may be badly affected by others.

With respect to the above, some interesting examples should be mentioned. Multiple school shifts, which favored greater school attendance in Colombia and in other countries, also contributed to problems of vagrancy as students were left with a lack of activities during off-school hours. In the end this had a negative impact on community life. In several countries, pregnant adolescents are expelled from school (in order to keep the "bad example" from spreading). This has meant psychological and social ruin for the young mothers and the worst possible family environment for their newborn. Requirements that some educational costs be absorbed by families, in order to improve the quality of local education, will surely affect portions of the family budget intended for other things in the short-run. The policy of creating child welfare homes will benefit children in terms of opportunities for early interaction, but at the cost of crowding and poor learning practices. However, in the long run, these programs may be more beneficial than direct food subsidies if they allow mothers freedom to seek employment outside the home. The expansion of health and social security services may be beneficial for groups who have not had access to them, but will perhaps result in the deterioration of services offered to other beneficiaries.

These examples show that the concept of "beneficiary" is actually just another variable in the development of social policies. Focusing social policies is thus a questionable operating criterion, from both philosophical and social engineering perspectives. However, thus far it has been the only criterion proposed and implemented. The Chilean experience is interesting. In this case, the identification of beneficiaries was based on individual data, since the policy was to provide individual benefits. A preliminary evaluation of results has drawn important conclusions. The methodology does not allow families or other groups to which the individual in question belongs to be identified and worked with. Therefore, it is not possible to operate over the proximate determinants of poverty, which in many situations condition poverty at the level of the individual (Vergara, 1990). This lack of attention to proximate determinants indicates the presence and influence of implicit values in social policies. While social research had already shown the importance of contextual factors, ideological orientations prevailed nonetheless.

Within this context, it is not only important to consider the characteristics of the immediate family nuclei and community, but also the quality of goods and services at the local level (Vergara, 1990). Social policies have centered on problems of coverage rather than quality, or on the fact that quality varies geographically. Services provided to a poor marginal zone differ in quality from those provided in an integrated urban zone. Thus, even when the right of access to a benefit is granted, social differentiation in supply can distort the effects sought. It is not unusual to find that most beneficiaries are in the least marginal areas, given their better position in the social structure and their better access to the sources of power. This situation clarifies the need for a social ecological approach towards social policy, in order to adequately identify beneficiaries and to analyze the benefits offered. Social policies have a geographical expression that must be considered.

Given the large volume of the poor population and the magnitude of their unsatisfied needs, it is illusory to think that poverty will be eradicated by the more precise focusing of available resources (Vergara, 1990). No doubt it is necessary but it is not sufficient. It is crucial to assign sufficient funds to achieve this objective and, furthermore, to follow a policy which includes the identification of individuals and their contexts in its methodology. A systemic approach is indispensable, not only to facilitate an effective individual impact, but also to act upon the proximate determinants of poverty. If this is not done, social policies can only follow an inefficient and remedial approach.

Some examples can illustrate the complexity of this type of problem. It is not enough to provide mothers with food supplements for their children, if the internal family rules of food distribution favoring men over women, and the head of the household above all, are not considered. It has been observed that the price differences between subsidized food and that sold at the market generally cause mothers to resell what they receive. It is not enough to reeducate mothers on better nutritional habits, if food conservation and cooking methods are not also considered. A family's size and structure, which might include non-nuclear members helping with food purchases and preparation, are factors which may weigh more in the choice of foodstuffs than all possible technical arguments about the better nutritional quality of certain types of food. It is not enough to design training courses for teachers to improve instructional quality without also considering school resources, local culture and the incentives used to motivate teachers to apply what they have learned. It is not enough to provide the right to attend a health center if the appointment system is inadequate and there are long waiting periods. In the long run, only non-working mothers, with the fewest health problems and the best economic conditions, will be able to attend.

In all the above cases, it has been recognized that community participation is important in the process of qualifying demand and identifying action nuclei. This has been postulated many times, and perhaps with excessive enthusiasm, since it is a rather difficult task. On the one hand, a community tends to be made up of very heterogeneous, socially and economically stratified populations, where solidarity is more the wish of the researcher or social promoter than a reality. On the other hand, it is difficult to find a situation in which a community would reject an offer of services. Under these circumstances, the identification of intermediate agents is very important. In this respect NGOs can play a new role in adapting government activities to local needs, since under the new decentralization schemes they are being encouraged to provide their experience in identifying local productive and social projects (Vergara, 1989). It must again be noted that only by using a systemic approach will it be possible to analyze and act within this great social complexity.

The past implementation of social policies has left several lessons: currently existing alternatives dominate the type of program results; implementation is a process occurring in a series of stages that are not always successive and integrated; policies tend to pursue multiple objectives that tend in practice to compete among themselves; and decisions made at the level closest to implementation are those that tend to have greatest impact (MacLaughlin, 1985).

At the local level, the primary determinants of how and how well programs are run include factors such as: the background and training of personnel; the degree of commitment among managers; the local system of demands, which may be complementary to, in competition, or in conflict with the objectives sought; and support for local groups, including the potential beneficiaries (O'Toole, 1989). Because these factors vary from one place to another, the services provided will also vary. Simply having managers committed to a project, finding that the program objectives are not in conflict with political or economic interests, and that there is an interested community will lead to total success in one place, while in another, unmotivated managers, conflicts of interest with local "caciques" and an indifferent population will translate to complete failure.

In the long term, the manner in which a program is designed weighs less than the manner in which it is implemented. Procedure is most important for success. A change of school curriculum implemented by a school director through discussions with teachers may lead to very different results from those achieved by the posting of a simple administrative bulletin stating the new contents for the next term. The first case may result in the active and informed participation of teachers; the second may encounter passiveness, resistance, and even confusion. The procedures that are chosen will make the same program very different when applied in diverse locations.

The implementation of policy involves a complex process of successive and sometimes simultaneous stages, especially when working within organizations with a minimal degree of institutionalization. Where underdevelopment is greater so is complexity, since policy implementation generally parallels the institutionalization of organizational frameworks. This makes implementation a learning and adjustment process more than a simple act of installing a new program. It is necessary to acquire new capacities, learn new rules of behavior and be converted to new beliefs. It is necessary to learn to fulfill new tasks, provide services in accordance with new objectives, and feel and think in a different way (MacLaughlin, 1985).

The multiple dimensions and categories of social policies foster the existence of diverse implementation philosophies. One of them is to seek compliance with new dispositions, invoking norms and sanctions. Another is induction in which reinforcement and incentives are used to motivate the population. Yet another is beneficence, which is based on technical assistance and donations (Bardach, 1980). These philosophies are not mutually exclusive. One or the other may be used as the implementation process moves from one level to another and progresses through time. Those who participate in the implementation process emphasize different aspects and philosophies. At the end, policy becomes whatever the participants recognize as central, in accordance with their own interests. In this way unforeseen objective can come to light during implementation such that the final results may not be the ones expected. Thus, the implementation of social policies becomes part of a complex social intervention process. The original social policy is clarified, specified and modified as intervention in the social body takes place. The decision, which often has greatest impact, is the one made by the health promoter, teacher or local bureaucrat at the moment when service is provided (MacLaughlin, 1985). The face of the public service is that of these individuals, who are anonymous to the system, but finally invest policies with their vital content. Their actions are crucial, as can be shown through any number of examples.

A revealing case is that of a study that detected that many women who tested positive in a Papanicolaou test (a routine introduced by several health programs in the region) never understood the results communicated to them by the physician and never showed up for treatment despite the fact that they may have been mortally ill (Ramos and Pantelides, l990). Examples can also be found in schools, vocational training centers, family orientation agencies, children's welfare homes and nutritional centers. In some instances, these were cases, which did not greatly affect the lives of individuals, but in other cases, they have had fundamental consequences in terms of individual and collective development.

A major risk in the social intervention process, implicit in the development and implementation of social policies, is that the system lacks the capacity to control and evaluate final products. Perhaps the worst consequence of the state crisis in Latin America is the present impossibility of making anyone accountable for his or her bad -or good- work. The loss to society, the community and the individual of the capacity to "audit" the system is one of the current crisis' most negative consequences for human development. Research is urgently needed to generate knowledge with respect to the recovery of public and social responsibility, and to make effective control mechanisms operational. One of the greatest costs of social policies, while not exclusive to them, either, is inefficient bureaucracy and the corruption of public agencies. A typical case is that of losses due to corruption in social security systems (Urrutia, 1990). There is nothing more demoralizing, nor more ethically corrosive, than a corrupt and unpunished state. Wherever such a state exists, any attempt at social ethics for human development crumbles.

Social control or the "auditing" of social policies is a complex area requiring research from the perspectives of both supply and demand. Social policies surge in response to a group of needs for products, techniques and services. Both the state, through its programs, and the private sector seek to satisfy that demand. But as is well known, this demand does not appear in an integrated or coordinated fashion. Ideally, social policies require the coordination of three basic systems: 1) an information system providing a data base on needs, goals and achievements; 2) a planning system leading to adequate resource allocations; and 3) an efficient budget execution system which can place resources where they are required when they are required. Each of these systems is an area that should be "audited" by society. Social research in this area must answer the following questions:

What type of institutional arrangements will be needed so that the identification of needs and goals, the measurement of achievements, the allocation of resources and the execution of expenditures can be controlled in some way by the state and by the communities involved?

How can control be exercised over NGOs, donors and development agencies with respect to social responsibility?

How should control of the social impact of economic, technological and other kind of policies be exercised?

How should the impact of these policies on social policies be monitored?

The modernization of the state structure is an issue that must be implicitly or explicitly considered in order to answer these questions. The bureaucratic power structure divides into sectors actions that should be carried out in an integrated fashion, thereby affecting the state's capacity to implement redistributional social policies as well. It is not unusual to find that actions attempted from social sectors such as education or health are undermined by actions undertaken by the economic, finance or industry sectors. Sectoral specialization sharpens the conflict of interests between diverse policies and makes communication between sectors handling different concepts and data difficult. If sectoral rules of conceptualization and implementation differ, the integrated use of knowledge is impossible.

Regulations covering the state's sectoral actions are of great importance. Each government sector has operative rules. This may be an advantage or a disadvantage in responding to local needs in a flexible and timely manner, and represent an obstacle to the multidisciplinary use of knowledge. For professionals whose careers take place in the public sector, policy and program norms are practically their only source of knowledge. This is frequently observed in the health and education sectors, where innovations occur almost exclusively when service provision regulations change. Sectoral norms also define the space available for action by promoters, teachers, physicians and others. One of the problems of implementation at the local level will be the negotiation of space permitted by each sector's norms and practices. It is hoped that with decentralization and regionalization, local governments will enjoy greater working space to compromise between local needs and the state's vertical structure. This is most crucial in social policies, for which the cluster of local factors is highly important at the moment of implementation.

Certainly, one should not ingenuously believe that the implementation of new policies could take place outside the framework of existing institutions in the region. It is true that the implementation of selective policies will require specialized institutions. In the early stages, however, there is no alternative but to base activities on what already exists and make use of available institutional, material, technical and human resources (Vergara, 1990). That is why the experience and stability of institutional resources is so important for policy success. Unfortunately, the state and economic crisis has led to discontinuity. The underdevelopment in which both public and private institutions operate has in many cases led to the dismantling of some structures in order to create others. Instable financial resources, turnover in personnel, and normative changes in administration prevent the successful and effective application of social policies. Successions of educational reforms and health plans provide well-known examples of these situations. In any case, whatever amount of social policy that can be produced in the region must be implemented by institutions in crisis.

All these problems indicate that social policy implementation will be a difficult process. It is one thing to discuss the need to face new social policy design, implementation and evaluation challenges in a multidisciplinary and multisectoral manner, as part of a systemic approach. It is quite another to achieve integrated activities in practice. Attention must be paid to the institutional frameworks that must be used. Research on this last issue is urgently required as a prerequisite to reaching a minimal rationality in and establishing priorities for the allocation of public and private funds. The integration of policies is a basic requirement since poverty problems respond to sets of individual, family and collective determinants. That effective social interventions should adjust themselves to a systemic approach is a principle being accepted in public forums. Nevertheless, it will be difficult to apply, given the complexity of the implementation process, the weight of underdevelopment, the anachronisms of state structures, and the lack of appropriate social engineering systemic models.

Research on effective social intervention within the current framework of crisis and all types of scarcities, presents perhaps, the greatest challenge, and requires the most urgent answers. The major question is how to do things. Decentralization and regionalization, the new roles of local governments, and the democratization process are new and very favorable conditions to be taken into account. But not much time is available. In most cases, there is immediate need for quickly conducted studies designed to support professionals and technicians working at the local level with very few resources, pressured by the seriousness and urgency of unsolved needs, and without concrete intervention tools or adequate diagnoses of the situations they hope to act upon.

Evaluation is implicit in any discussion of social policies. Once evaluation exercises are begun, a distinction must be made between criteria on values that orient policies and criteria related to more technical aspects (Klein, 1982). Academic discussion of social policies tends to refer to the values of policy content; in this case, it is a meta-activity that goes beyond the apparent content, oriented towards the underlying social philosophy. The technical discussion is quite different, it does not question the underlying values; rather, it accepts them and operates from them. Both types of discussion are complementary and required; however, another current challenge is to refocus these discussions within a framework consistent with changes in the role of the Latin American state and recent developments in the fields of capitalism and socialism.

What values will orient the new social policies? Current history will redefine the content of values such as social equality, equity, individual achievement or freedom of choice. The discussion of social goals and their translation into specific social policy objectives is basic to redefining the limits of state, society and individual liability with respect to human development. However, this necessary academic discussion will not be easy, since the value structures corresponding to models prevailing until recently will persist, despite ongoing historical changes. Scholars will need to display courage and imagination to creatively discuss fundamental values, in accordance with historical changes, and experiences of social policy success or failures.

Another even more difficult challenge is to link evaluation as a meta-activity to evaluation as a technical activity. Ideally, to make technical evaluations viable, the objectives must be clearly defined, and cause-effect relations understood between social intervention activities (for example, the kind of educational curriculum or family planning service) and the expected effects (such as greater educational achievement or fewer pregnancies). However, social policies are often ambitious and very often have general objectives. Thus, the cause?effect relationships to be considered are not only multiple, but become increasingly blurred as a result of complex interactions between organizational, political, social, cultural and legal factors at different levels. The solution, starting from a clear identification of social indicators, is not easy. Any idea of a social welfare dimension (education, or ideal family size) is in itself multidimensional, valuative, and resists being encapsulated in a simple series of social indicators (Klein, 1982). Known social theory and available experience do not allow us to be certain as to what social indicators actually mean. But efforts must be pursued, there is no other way to carry out technical evaluations, without which it will not be possible to learn whether underlying goals are being historically achieved or not.

To the problem of uncertainty as to the meaning of social indicators, another may be added: what time frame is necessary to achieve a significant impact? To the uncertainty as to the causal mechanisms of our as yet elemental social engineering, we must add the uncertainty about the time horizon required for social change to become institutionalized. Under these circumstances, what will be requested of the social researcher and planner?

Experience with positivist evaluation models has not been satisfactory. The classic evaluation system based on the hypothetical-deductive model supposes a rationality that is non-existent in current social policy, and requires clear and operationally defined objectives, stable program parameters, and criteria for measuring success. Inputs such as policy and program results are considered in a static fashion as within an experimental design, and the results are treated as if they were the effects of a controlled treatment. Inputs and results are defined in constant and unidimensional terms, and rarely incorporate elements such as context and the implementation process. An environment with simple structures is presupposed, as is a hierarchy with singular authority, few complex interactions, and a rational rewarding system (MacLaughlin, 1985). Experience with Latin American social policies and programs proves it impossible to make these types of assumptions. Objectives are not only general but also changing, implementation mechanisms vary through time, and there are no definite authority or reward structures.

Research on social policy will have to explain, at a technical level, not so much what should be done, but how to do it, what will result, and why. Research must focus more on analysis of the processes involved, rather than simply on identifying whether objectives are realized. However, although social interventions will have to be discussed from a technical point of view in order to answer basic questions, their meta-political discussion will be unavoidable. Research must, therefore, generate objective technical results on which to base this discussion, and must then be able to elaborate a new social ethic to which all those providing social services - that final but crucial point of all social policy systems - must adjust themselves.

The many difficulties previously noted will surely mean that the distance between academic and technical discussions will never disappear entirely (Klein, 1982). However much this must be accepted, the situation that must be avoided is to allow financial and political pressures for immediate and clear results lead to the abandonment of academic discussion, or to reduce reflection on social development to technical evaluations. The separation between these types of discussion is one of the greatest intellectual and practical risks the region currently faces.

Conceptually and practically, the systemic approach to social policies requires a new profile for the researcher and the professional of the social sciences. The traditional disciplinary focus, based on formation in sociology, anthropology, economics, psychology, education or law, is no longer sufficient. New social policies must be oriented towards a multidisciplinary and intersectoral focus, which will lead to innovations in social planning procedures.

Social planning has been carried out from a sectoral perspective, which is consistent with traditional disciplinary and institutional specializations. Thus, services are offered by sector: education, health, employment, justice or social security. This, in practice, has kept social planning separate from economic planning. Since the time in which a working group presided by Gunnar Myrdal met in 1969 under United Nations auspices to revise the concept of social planning and recommended integrated planning, little progress has been made towards a concept that is only now recovering special interest (United Nations, 1971). At that time, it was proposed that social planning should be understood as socially committed economic planning: the planner should be an economist with social awareness (Hardiman and Midgely, 1989). However, little importance was attached to implementing these concepts in terms of training needs, organizational responsibilities and professional roles. The proposal assumed that common professional training, especially in economics, would be sufficient for social planning. The risk of applying strict economic efficiency criteria to social plans and policies was not foreseen. Recent Latin American academic practice has produced a divorce between different disciplinary traditions, to the point that economic criteria predominate in practical planning and development leadership (as shown by current government practices). With few exceptions, final decisions on social programs are made by the Ministries of Economy and the central banks throughout the region.

Many factors have contributed to this situation. Among the most important is the fact that social engineering models for human development have advanced less rapidly than economic models for the analysis of economic growth. There is, of course, a large potential that should be taken into account. But to do so, human resource training must be directed to producing innovative professional profiles.

Multidisciplinary formation based on the handling of quantitative and qualitative techniques is required, along with the mastering of new information technologies, and a clear systemic and practical orientation. As opposed to economic planners, social planners must have a different methodology, appropriate for considering social needs and social problems (Hardiman and Midgely, 1989). But, in light of what has been discussed thus far, this professional "must be" will be difficult to achieve. Here is, surely, Latin America's main challenge in implementing the social policies it needs so much. Because, in addition to all the above, the new social policies will have to be designed, implemented, and evaluated by the few or many professionals remaining in government, NGOs or universities after these years of crisis.

The formation of necessary human resources should be addressed to creating the technical and managerial capacity needed to work in both the public and private sectors. The capacity to win credibility from potential beneficiaries and mobilize community resources will be crucial. If the new social polices are to be based on the principles of privatization and the free operation of competition mechanisms in institutionally open markets, the training of new professionals will require drastic changes. The concept of a "social manager" will have to replace that of a social administrator.

The idea of social management is not popular among social scientists. However, institutional changes and the processes of economic and technological transformation will require greater capacities in developing solutions and evaluating results. Social engineers with a systemic, non-fragmented view of social problems will have to emerge, hopefully as a result of action rather than through omission by professional training and social research centers.

We have consciously left aside discussions about what human development is and how to define social policy content, but expect to have shown that despite clarity and agreement on these definitions, there is still a long and difficult road ahead before effective results are achieved. When dealing with an issue as complex as this, it may prove more rewarding to search simultaneously for conceptual clarity and solutions to the problems faced in practice.

Human development is basically an ethical enterprise. In the public sector, social policies are the major institutional instruments for its promotion and achievement. But for achievements to indeed occur, several conditions must be met. Policies should be based on a systemic perspective of knowledge, including consideration of the social, technical and environmental aspects affecting social problems. Data and collection methods should be accessible to those who make decisions and provide services. Knowledge of problems must be socialized and even democratized: both the collective and the individual should and may take part. The state, the society and the individual must be able to control levels of social liability among those involved in the implementation process. The ability to evaluate the achievement of goals and values must exist.

Human development as an ethical activity is not, then, merely a question of academic axiology. It also involves science and technology. Therefore, during the region's current historical moment, social research must target new methodological and practical directions. It is not enough to propose goals; the path towards them must also be indicated, as must the manner in which the results can be measured. All this will demand the ability to summon ideological pragmatism and managerial capacity from social science professionals. Human development is no longer a matter of good intentions. Meeting this challenge may perhaps be the best way of rescuing the social utopia that has always illuminated Latin America's most dramatic historical moments.

Arretx, Carmen
1990 Informe del Seminario-taller sobre la Experiencia de Aplicación del Método del Hijo Previo en América Latina. Santiago: Naciones Unidas, Centro Latinoamericano de Demografía (CELADE). January 29-31.

Bardach, E.
1980 "Implementation Studies and the Study of Implements." Paper presented at the Annual Meeting of the American Political Science Association, Washington, D.C.

Boeninger, Edgardo
1982 "Application of the Social Sciences to Public Policies: Producers, Consumers, and Mechanisms of Mediation." Stifel, Laurence D., Ralph K. Davidson & James S. Coleman. Social Sciences & Public Policy in the Developing World. Lexington, Mass.: Lexington Books, 251-276.

Cook, Thomas D.
1985 "Postpositivist Critical Multiplism." Shotland, R. Lance & Melvin M. Mark. Social Science and Social Policy. Beverley Hills: SAGE Publications, Inc., 21-62.

CEPAL
1989 Paquete para Proyecciones Demográficas Nacionales y Subnacionales por Microcomputador (PRODEM). Santiago: Comisión Económica para América Latina y el Caribe, Centro Latinoamericano de Demografía (CELADE).

1990a. Magnitud de la Pobreza en América Latina en los Años Ochenta. Santiago: Comisión Económica para América Latina y el Caribe.

1990b. Transformación Productiva con Equidad. Santiago: Comisión Económica para América Latina y el Caribe.

1991 "El Desarrollo Sustentable: Transformación Productiva, Equidad y Medio Ambiente." Notas sobre la Economía y el Desarrollo. No. 502/503.

Conning, Arthur, Ari Silva and Lawrence Finnegan
1988 "Small-Area Census Data Services by Microcomputer: Applications of the REDATAM System in Latin America and the Caribbean." Paper presented at the IASSIST 14th Annual Conference, Washington. Santiago: CEPAL-CELADE.

Dubey, Sumati
1980 "Conceptual Framework for Comparative Analysis of social policies." Social Development Issues. Vol. 4, no. 3: 63-78.

Hardiman, Margaret and James Midley
1989 The Social Dimensions of Development. Social Policy and Planning in the Third World. Aldershot: Gower Publishing Company Limited.

Guzmán, J. Miguel
1988 "El Procedimiento del Hijo Previo: La Experiencia Latinoamericana." Notas de Población. No. 46-47: 41-63.

Irarrázaval, I.
1989 Algunas Notas sobre el Concepto de Igualdad de Oportunidades y otras Estrategias del Estado en el Area Social. Documento de Trabajo No. 118. Santiago: Centro de Estudios Públicos.

Irarrázabal, I.
1982 "Promoción del Desarrollo Social Privado a Nivel Local: Una Propuesta." Estudios Públicos. No. 38: 149-192.

Klein, Rudolf
1982 "Evaluation and Social Policy. Some Reflections on Ideas and Institutions." Evaluation and Program Planning. Vol. 5: 133-140.

McLaughlin, M. W.
1985 "Implementation realities and evaluation design." Shotland, R. Lance & Melvin M. Mark. Social Science and Social Policy. Beverly Hills: SAGE Publications, Inc., 96-120.

MacPherson, Stewart
1982 Social Policy in the Third World. The Social Dilemmas of Underdevelopment. Sussex: Wheatsheaf Books Ltd.

Miller, Steven I. and Marcel Fredericks
1987 "The Unintended Consequences of Social Policymaking: Some Sociological Considerations." Sociologia Internationalis. 25 (1): 68-84.

Naciones Unidas
1989 Principios Normativos para las Políticas y Programas de Bienestar Social para el Desarrollo en un Futuro Próximo. New York: Departamento de Información Pública.

O'Toole, James K.
1989 "The Economics of the Implementation of Social Policy: Searching for the Missing Link." The Social Science Journal. Vol. 26, no. 2: 115-130.

PROMEDLAC
1991 Proyecto Principal en la Esfera de la Educación en América Latina y el Caribe. Declaración de Quito. Cuarta Reunión del Comité Regional Intergubernamental. April 22-25.

Ramos, Silvina and Edith A. Pantelides
1990 Dificultades en la Prevención Secundaria del Cáncer de Cuello de Utero: Las Mujeres y los Médicos frente a una Citología Cervical de Resultado Positivo. Documento CEDES/43. Buenos Aires: Centro de Estudios de Estado y Sociedad.

UNDP-United Nations Development Programme
1990 Human Development Report 1990. New York-Oxford: Oxford University Press.

United Nations
1971 "Social Policy and Planning in National Development." International Social Development Review. No. 3: 4-15.

Urrutia M., Miguel
1990 "Población y Gasto Social." Paper presented at the Seminar on Population and Development, Paipa. Bogota: Centro de Estudios sobre el Desarrollo Económico - Corporación Centro Regional de Población.

Vergara, Pilar
1990 Políticas hacia la Extrema Pobreza en Chile. 1979-1988. Santiago: Facultad Latinoamericana de Ciencias Sociales (FLACSO).

Vergara, Carlos
1989 "The New Context of Social Policy in Chile and the Space for Non-governmental Organizations." Charles Downs et al. Social Policy from the Grassroots. Non-governmental Organizations in Chile. San Francisco & London: Westview Press Inc. Boulder.

WCEFA-Inter-Agency Commission
1990b. Meeting Basic Learning Needs: A Vision for the 1990s. New York: UNICEF House.

World Bank
1990 Poverty. World Development Report 1990. Oxford: Oxford University Press.