The Conceptions of Social Policy: Universalism versus Focalization *
* This paperwork was originally published by the magazine Nueva Sociedad No 215, May-June 2008, ISSN: 0251-3552. It was developed on the basis of a paper prepared for the Andean Corporation of Promotion (Corporación Andina de Fomento - CAF) and it was presented in the Workshop on Social Policy held in Caracas on January 18, 2008. It has been included in the current edition of the Public Budget International Journal upon permission granted by the institution. The author thanks Horst Grabe, Julio Bolvinik, Alfredo Sfeir and other participants in the workshop. He also thanks Martin Hopenhayn and Juan Carlos Ramirez for their comments. The multiple interactions with Luis Miguel Castilla, CAF chief economist, about the report of the CAF 2007, and Rebecca Grynspan, regional director for Latin America and the Caribbean of the United Nations Development Programme (UNDP), over several years, have served to refine some of the concepts expressed in the essay. These conceptions are, of course, exclusive responsibility of the author. ** Former Minister of Finance and Agriculture of Colombia, was Executive Secretary of the Economic Commission for Latin America and the Caribbean (ECLAC) and Deputy Secretary General of the United Nations Economic and Social Affairs. He currently teaches at Columbia University.
The struggle between two conceptions of the social policy has come back to life in recent years. While the first of them, which has served as the essential grounds for the social policy reform in the last decades, emphasizes the focalization of State grants towards the poorest sectors and the design of public-private schemes, the second defends the need of a firm foundation for social policy on different principles, among which universality and solidarity may be highlighted. The latter conception recovers the roots on which social policy was developed during the post War period –and in some cases, before that–, a policy that, however, had limited progress in Latin America, mainly due to the restricted scope of formal employment to which its development was associated.
This paper revises this debate and inclines in favor of the universalist and solidarity conception on the basis of a social citizenship view. It argues that the best focalization is a universal social policy, and even more, that focalization should be viewed not as a substitute, but as a complement –and, in fact, as an instrument– of universalization. In addition, it presents some thoughts about the growing segmentation problems in the social policy systems in the region. However, it points out that full application of these schemes generates a high demand of fiscal resources, which most of the countries with an endemic problem face: the weakness of their tax structures. Therefore, progress towards universal social policy schemes, in accordance with a view of citizen’s rights, shall demand greater efforts to increase and improve the structure of tax revenue.
The essay is divided in four parts. In the first one, a conceptual debate is revised and the essential arguments in favor of a universal social policy are introduced. In the second one, the available evidence about focalization of social public spending is analyzed. Then, we briefly develop the subject of taxation. Finally, the main conclusions and implications are introduced. Although wide in its approach, we should point out that the paper addresses social policy itself and sets aside the wide field of relations between economic policy and social policy, on which I have focused my attention in other papers.
The struggle between two conceptions of social policy
The social policy schemes of the past and the impact of market reforms
Modern conceptions of social policy as the State’s responsibility originate in the acknowledgment by liberal States, of the need to offer basic public education, both secular and universal, in addition to some basic health services, such as social services inherent to the progress of modern societies. From the late 19th century, it added the Bismarckian view of social security and the appearance of a labor movement that, both directly and through its political expressions, demanded not only the creation of a legislation that encourages and protects employment, but also an integral system of social policy. These ideas gradually spread in the industrialized countries throughout the 19th century and, above all, the 20th century, with the development of the Welfare State. The correlation of this process was the unprecedented expansion of the size of the State and the consequent demand of resources to fund it.
In Latin America, these trends were expressed in a more limited manner. The initial progress was achieved in a group of countries, especially in the Southern Cone and Costa Rica. It was the reflection of an equally early economic progress, but also, during some stages of the history of these countries, of a strong social development commitment. The reforms encouraged by José Batlle in Uruguay, in the middle of the 1910’s, were the most outstanding development.
However, in most Latin American countries, progress towards universal public education and public health systems was only definitely made after the Second World War. On the other hand, the scope of social security both in the field of pensions, as well as health and professional risks, has always been limited to the formal wage earning employment, due to the influence of the Bismarckian principle of associating social security to formal employment in economies where the scope of the latter has always been limited. Besides, the varied strength of wage earning worker groups translated into great differences in the benefits they received. The combination of the limited scope with these corporate elements1 shows how the system developed in the region, as a segmented and incomplete Welfare State (Ocampo).
In this manner, although inspired by the same universal principles that had fostered the development of the Welfare States in industrialized countries, the social policy of the region only partially progressed. One of the corollaries of this was that, at the end of the industrialization stage addressed by the State, the social policy only reached, in most of the cases, those sectors of the population with average income (including workers from the modern factory sector, which in general, belong to these sectors) and thus, continued excluding the poorest sectors, especially in the field of secondary school education or the access to the most advanced social protection systems. In addition, the poorest sectors were concentrated in rural areas, where the scope of the social policy was by far more limited.
The economic reforms in the eighties and nineties –and in some countries like Chile, even beforehand– acknowledged this problem. Such reforms, developed under the influence of the World Bank (WB) in the framework of the fiscal hardship that resulted from the debt crisis, led to an alternative view based on three basic instruments: focalization, private participation supported by demand subsidies and decentralization. The first of these instruments, focalization, looked for consistency between the imperative of rationalizing public spending and the need for social policy to effectively reach the poorest sectors. One of the newest elements was the design of beneficiaries’ identification systems. The other two instruments tried to rationalize the State apparatus, either by means of the development of public-private schemes, or through decentralization to introduce the beneficiaries to the state authorities in charge of the provision (or assignment) of grants. To these instruments, specific programs for vulnerable populations or to face the costs of the structural adjustment were added; among which, the social emergency funds in their different variants can be highlighted.
The reforms that were initiated combined these instruments with the old schemes of social policy in a variable manner. In the case of basic education, public education remained dominant. Public-private participation schemes based on demand grants (scholarships), such as those implemented in Chile, progressed in this way, in a limited manner. On the contrary, in health, and especially in pensions, public-private participation systems –and even the privatization of pension systems– were imposed in different countries, although, in others, public schemes remained prevalent, and were subject to a rationalization tending to make them financially feasible. In any case, even in those countries where they were privatized, the State kept the regulatory responsibility over the pension and health systems, as well as over sustainability and other financial obligations (minimum pensions and health grants). Besides, focalization achieved its best development by means of conditioned grant programs, which were originally designed as social emergency mechanisms (such as the Progresa program in Mexico) or as supplementary instrument of universalization of educational services (such as Bolsa Escola in Brazil), but they gradually evolved towards wider support coverage programs, through the income transfer to the poorest sectors (Oportunidades in México and Bolsa Familia in Brazil) and then applied in many other countries in the region. On the other hand, the decentralization progressed more in the two countries with a strong federalist tradition (Argentina and Brazil), as well as in two formally centralist but with “federal vocation” (Bolivia and Colombia), but also, with different intensity, in many others.
The result of the above is that today, three types of social policy schemes coexist, matching, in a certain manner, the three Welfare State models of Gøsta Esping-Andersen (1990), but which combine in one country in many cases – and that had tended to variably lose the attribute so much highlighted in this author’s work about «decommercialization»–. The first one is a system with strictly universal vocation and predominantly public organization, although with different degrees of decentralization, which is found in primary and secondary school education and, in competition with private institutions and in university education. The second one is a segmented scheme and with different corporative elements, prevalent in the social security in a wide sense (pensions, health and professional insurance). The third one is a strictly focused scheme, whose best example is the conditioned grant programs, but which has expressions in many other policies addressed to certain social groups. In fact, the proliferation of specific programs, which, in many cases, have a structure of «geological layers » (new plans introduced as innovations by new governments, but, which often are partially overlapped with old programs that do not disappear), seems to be an outstanding characteristic of the social policy in recent decades2.
In spite of innumerable reforms –and in some cases, useful innovations–, the social policy seems to have lost its unity and strategic nature, characteristic of the welfare states of industrialized countries and of the best developments from the past in Latin America. Fernando Filgueira et al. (2006, p. 37) have characterized the social policy in the region as a «persistent corporatism combined with a liberal reform » lacking a pillar of defined non-contributive social services. A way of expressing this loss is that the excessive emphasis on the three instruments mentioned above –focalization, public-private schemes and decentralization–, instead of on the principles that should guide it, ended up defeating the strategic view of social policy.
The return of a universalist view
Not in vain, therefore, the return to a strategic view of the social policy subordinates the instruments of said policy to the principles over which the latter must be founded. They, in turn, are conceived as derivations of the formulations about rights and social citizenship and, as such, place the social policy in the center of the agreement and social cohesion.
This view has a long tradition. We must remember that, parallel to the development of the Welfare State, a new body of human rights was formulated (most of them, second generation rights), and was finally honored in articles 22 to 27 of the Universal Declaration of Human Rights and, subsequently, in the United Nations International Pact on Economic, Social and Cultural Rights. In modern conceptions, this body of rights, which expresses equality, solidarity and non-discrimination values, is considered indivisible from civil and political rights, clearly formulated from the end of the 18th century, especially in the Declaration of the Man’s and Citizen’s Rights of the French Revolution, which ensure the rights of the persons above the State’s power and to the participation in public decisions (Cepal 2000a).
It is interesting to remember that this body of economic, social and cultural rights is considered in the preamble of the Universal Declaration of Human Rights as an extension of the principle of freedom. Said preamble points out, in fact, that the United Nations search «to promote social progress, and (…) raise the life quality within a wider concept of freedom», reproducing a concept that had already been adopted in the United Nations Charter. This conception has direct roots in Franklin D. Roosevelt’s political views which inspired the Charter, and is associated, in turn, to the concept of social citizenship by T.H. Marshall (1992), and in more recent times, to the «development as freedom» by Amartya Sen (1999). One of the correlates is the view of democracy as an extension of citizenship, in its triple dimension of civil, political and social citizenship, divulgated in Latin America by the UNDP’s report (2004). It is worthwhile observing that the view of social citizenship –or of «Social State of right»– is present in the constitutions of several countries in the region, among them, Brazil, Colombia and Venezuela.
One of the most precise formulations of these ideas is the chapter about social policy principles of the document Equidad, desarrollo y ciudadanía by the Cepal (2000a), where the principles of universality, solidarity, efficiency and integrality are set forth. The first of them is associated to the concept of basic social services provided by the State as the citizen’s rights. According to this conception, therefore, education, health and social protection are more than services or goods. The second principle is related to what is obvious, particularly in highly unequal societies: that the guarantee of access of the more disfavored sectors to said rights demands the application of the principle of solidarity, which also expresses a basic social objective: the need to build more integrated societies. The third one –efficiency– points out the need to use public resources better, in order to achieve more benefits; while the last, –integrality – relates to the strong interrelations existing between the different dimensions of social development (or, from the negative side, the multiple dimensions of poverty).
This formulation refers us to two dimensions certainly correlated: the relation between economic and social rights and the level of economic development of a country, on the one hand, and the demand of said rights, on the other hand. About the first one, it is important to point out that the only political statement that «Every person, along with his/her family, has the right to a suitable life quality, health and welfare, and especially, food, clothes, housing, medical care and the necessary social services »3, or the right to education and social security4, does not allow the creation of wealth or the distribution of the inexistent. In this sense, as was commented in a prior paper:
Its implementation must be compatible with the level of development achieved and (…) with the ‘fiscal pact’ prevailing in each society, and thus avoiding it translates into unsatisfied expectations or into macroeconomic imbalances which affect, in other ways, the social sectors which need protection. Equity, in this sense, must be understood as setting goals that the society is effectively able to reach, given its level of development. That is to say, its point of reference is the achievable, as a minimum, and therefore, as the debates about economic and social rights have interpreted it, the maximum among the achievable. (2004, p. 159.)
This implies that the demand of economic and social rights is subject to the limits of what is possible, at a certain moment and given the economic development of a country. In fact, the attempt to conceptually demand rights, without taking into account what a society may give to all citizens may result in the distribution of the already scarce resources to only a few. Therefore, although certain levels of juridical requisition are inherent to any formulation of rights, in the case of economic and social ones, what is equally or perhaps even more important is political requisition. In other words, the political authority (the Constituent Assembly or the Congress) should specify which achievements within the field of economic and social rights the society expects to reach, within the restrictions imposed by the level of development, but also under the principle that the «maximum among the achievable» must be reached . Around this, of course, the political debate is structured on different democratic options, as well as on the generation of the necessary resources to ensure the supply of the basic services the society, through the political authorities, defines as essential at a specific moment.
This view is associated to the concept of «basic universalism » by Carlos Molina (2006)5, who mentions a series of basic social services and essential risk coverage that must be extended to the entire population, with homogeneous quality standards and supplied on the basis of the principle of citizenship. This also refers us to the concept of merit goods from the welfare theory that the Cepal (2000a) translated –alternatively and correctly– as «social value goods»; that is, those goods and services the society considers must be received by all its members as citizens. These goods and services may be understood, therefore, as a tangible expression of the social and economic rights and evidence the society’s authentic preferences that go beyond individual preferences.
The scope of these «social value goods » must reach not only the level of development, but also the demands of a specific economic organization model on the citizens. In this sense, contemporary economies generate two specific demands, besides those that the social policy has acknowledged for many decades. The first one derives from the extension of society, knowledge and the higher education and training requirement demanded by today’s economy and society. The second one is the greater economic insecurity generated by a society that faces accelerated changes in a context of greater competence. For this reason, some authors have pointed out that there must be a positive association between economic openness and social spending, as occurring in the countries of the Organization for Economic Cooperation and Development (OECD) (Rodrik). In other words, this is consistent with different views of the Scandinavian Welfare State, in the sense that a more elaborated social protection system is essential in more open economies, as said services are an alternative to the labor protection based on the barriers to foreign trade (Thalen).
Finally, it is worth mentioning that this formulation also seeks to correct one of the greatest dangers faced by social policy: segmentation. This problem tends to get worse in the focalized policy schemes, but it is also present in systems where multiple service suppliers exist, some of which tend to «skim» the market or to discriminate between the different demanders, according to their level of income or their site location. The educational system is a special source of concern. It is usually said that education is a social equality mechanism – a principle that, as I have already pointed out, has deep liberal roots – and that the different educational achievements are one of the fundamental determinants of the inequality in income distribution. It is important to remember that, especially in highly unequal societies, the educational apparatus is also a powerful mechanism of social segmentation, which tends to reproduce the existing inequalities and, in certain cases, to spread ideas and practices that separate the elites from the rest of the citizens. Just as this problem shows up in systems that generate education for the rich and education for the poor, there is also a tendency to generate health care systems for the rich and health care systems or the poor, as well as room segmentation in the cities and many other mechanisms which reproduce or widen these differences.
The universalist paradigm has been the object of two main criticisms: that it demands many fiscal resources and that, given scarce public budgets, this may generate social policy systems in which human resources end up badly focalized. As we will see hereinafter, the previous criticism is valid, the latter is not. In fact, the exactly opposite statement may be formulated: the best focalization is a universal social policy.
This second criticism –that the universal policies generate a poor focalization of resources– is associated, in some formulations, to the view that middle classes count on the capacity to get the public resources distributed in their favor, thus excluding the most disfavored sectors. This would be done, for example, from the trade union associations (which, as I have already pointed out, in countries having limited formal employment systems correspond, in general, to middle income sectors) or from their influence on political parties.
This view is deficient, due to at least three different reasons. First, because even though the extension of social benefits may be mostly seen as «conquering» the middle sectors, the democratic systems generate a pressure so that these benefits get generalized to the entire population later. This is, in fact, the history of the Welfare States in industrialized countries. In that context, as we will see, marginal spending oriented to extend the coverage of social services is highly progressive. A second way of thinking about this subject is that the capacity of the social policy to attract the middle class may be seen as evidence and a guaranty of a policy that provides homogeneous quality services (Grynspan). Besides, this is essential so that the middle class supports the necessary taxation levels to fund the high social public spending necessary to support these systems, as mentioned in studies referring to industrialized countries (Korpi/Palme). In addition, with a few exceptions, the proposal of some defendants of focalization to redistribute the middle sector’s public resources towards the poorest sectors lacks political realism.
Secondly, in Latin America many of the homes belonging to the «middle » sectors earn relatively low incomes. This becomes evident when you observe quintiles 3 and 4 of income distribution. In fact, according to the estimations of the Cepal6, in countries with a lower per capita income (such as Bolivia, Honduras or Nicaragua), some of the homes in quintile 3 belong to the poor population. In these countries, most of the families in both quintiles receive incomes lower than the equivalent of two lines of poverty, which makes them very vulnerable to fall into that situation. In the countries with a per capita income similar to the regional average, such as Colombia or the Dominican Republic, almost all the homes in said quintiles, earn less than the equivalent to three lines of poverty. Only in four countries (Argentina, Chile, Costa Rica and Uruguay) it may be said the bulk of those homes in quintile 4 obtains incomes higher than the equivalent to three lines of poverty.
This, in turn, reflects a widely known fact: the poor distribution of income, typical in Latin America, is associated to the concentration in the richest decile (see, for example, Cepal 2006b, chap. I). It is even possible that said middle sectors are amongst those who experimented some of the strongest tensions in recent decades, associated, for example, to the decreasing returns of secondary school in terms of higher incomes, or the reduction in formal employment until the beginning of this century.
Finally, a social policy consistent with the promotion of a greater social cohesion should offer a proposal to the entire society. In fact, one of the greatest problems of the «focalizing State» is that its capacity to convene citizens has been limited, among other things, because it does not offer the minimum the society expects from the State: a proposal for the entire society and not only for parts of it.
Universality and focalization
The redistributive effects of social public spending
Different papers by the Cepal (2000a, 2000b, 2006c and 2007) provide a comparative summary of the studies on the redistributive effects of social public spending (in the case of those published by the Cepal, 2006c, based on the research made by the IDB). Graphics 1 and 2 summarize the two most recent studies, which mostly correspond to estimations based on public spending from the late nineties and the beginning of the current decade.
These studies indicate that the distributive effects differ significantly among different types of spending that may be grouped in three large categories. The first one includes the most redistributed spending: social work programs and those areas of the social policy with universal or quasi-universal coverage, such as primary education and some health services. The second category comprises services with intermediate coverage, such as secondary education and spending on housing and sanitation. In this case, the distribution is not generally progressive (except in some countries), but it is not far from being equally distributed and, therefore, it is much better than the distribution of primary income. Spending in health is between the two categories above mentioned: it is slightly progressive, particularly in the most recent study.
Finally, the third category includes social security (pensions) and university education, where the benefits tend to be concentrated in the highest deciles. But even in these cases, said distribution is, on average, slightly better than that of the primary income (although in some cases it is worse).
Although it is true that the assistance programs confirm the interpretations of those who defend focalization, in the sense that greater selectiveness implies a better redistributive effect, the fact is the impact is limited because the amounts allocated to them are relatively modest. The conditioned grant programs (such as Bolsa Familia and Oportunidades) are the most important innovation in this field, but there are other programs which also generate highly distributive effects, especially the nutrition programs addressed to the infant population. However, focalized programs that generate a great impact, are those precisely characterized by the extension of its coverage; that is, by the trend to «universalize» its benefits within a target population. In addition, one of the greatest virtues is that they use transfers as leverage to make the beneficiary sectors use the education and health universal social programs.
As these programs absorb small proportions of social public spending, their contribution is less than one fifth of the redistributive effect of said spending (Cepal 2007, chart II.20). Thus, the highest redistributive impact comes from the education and health programs with wide coverage. In fact, according to all the existing studies, these are the programs with a greater impact on the poorest sectors’ effective income. Moreover, the extension of these programs, from a limited coverage to a wider or universal coverage, confirms a conclusion that was shown in the pioneer study about this subject in Colombia (Selowsky): social spending, in these cases, is highly redistributive in the margin.
Although the studies about this subject are not strictly comparable, they are useful to illustrate this issue. Chart 1 shows a comparison between implicit incomes received by Colombian homes for the concept of different types of public spending in education at two moments in time, 1974 and 1992 (Selowsky; Vélez)7. As observed, spending in primary education was already highly progressive in 1974, but it became even more progressive in the following years. As far as the benefits of secondary education, in 1974, they concentrated in the middle sectors, but with the extension of the coverage, they moved towards the low income sectors. And the benefits of university education, which was highly regressive in 1974, were already less concentrated than primary income in 1992, thanks to the extension of the coverage.
This implies that spending that was considered regressive before, became progressive as the coverage was extended. This is, in particular, the case of secondary education, but also the case of other services, such as water piping and sewage. In both cases, marginal spending was, therefore, highly redistributive.
The conclusion of this analysis is that the policies included in the second category above mentioned, secondary school and housing and sanitation spending – that is, the social services under universalization process–, will acquire greater progressiveness as greater coverage levels are achieved. In all of them, the spending expansion associated to the coverage extension –that is, marginal spending– is strongly progressive, perhaps not less progressive than the spending included in the first group (primary education and basic health care services).
The third category includes two types of social services, university education and social security, of a very different nature. The first one is perhaps the clearest case where the redistributive criteria should prevail more than in the past, by means of the design of a scholarship system associated to the income of the students’ families. This financing system is different from free education schemes, which must prevail clearly in primary and secondary education. This statement should be analyzed, however, in two different senses. On the one hand, the university is also a source of knowledge production and, in consequence, it should receive appropriate resources for that. On the other hand, public university may be a strong mechanism of «social desegmentation» in countries where the spaces for meeting among citizens are very limited. In industrialized countries – and, in particular, in European countries– the public university has played, in fact, said role. Therefore, a scholarship system runs the risk of deleting this function of the public university; it is possible, in fact, that this function has already been strongly eroded. A manner of handling this problem, which is applicable to other sectors of social policy, would be to keep a highly competitive system, with public and private agents, but awarding the public ones general grants to increase their attraction. In fact, this is the United States university design. In this case, quality control over the public university offer could be achieved through competition.
In addition to university education, the second area included in this category is social security (pensions). At this point, it must be pointed out that the estimations exaggerate their regressive impact on distribution. The first reason for that is that, in general, the contributions (present and past) from those who benefit from the systems are not deducted. To explain it simply: if the spending is regressive because access is limited, the contributions funding them are, for this same reason, progressive, because they fall on the higher income sectors. Under these conditions, the net regressive effect is much lower than the one regularly estimated and it is possible that the grants incorporated to these systems are progressive, even highly progressive8. In addition, an important part of those appearing as budget contributions are, really, the employer’s contributions the State makes to its workers. This contribution appears as regressive due to the high proportion of workers with university education in the public sector. The distributive effect of social security should be estimated, therefore, based on the grants awarded by the State charged on the general budget, net of its employer’s contributions. Even in this case, as we will see in the next section, the central hypothesis of this paper is still valid: spending progressiveness–or, perhaps better, less regressiveness– is associated to a wider coverage. The highest progressiveness will be reached through a policy addressed to universalizing the access to social security. As the Cepal (2006a) already analyzed deeply, in economies where an important proportion of employment is informal, this will demand to combine the existing contribution system with a non-contributive support. In a scheme of this nature, the State’s net grants will have, by definition, a highly progressive effect.
The general conclusion derived from this analysis is clear: the higher progressiveness levels of public spending are closely related to the extension of coverage. Therefore, the best focalization is universalization. Focalization,–or perhaps better, selectivity, to eliminate the strong conceptual and even ideological content acquired by the term – must play a subsidiary role in this scheme. However, it may fulfill three particular functions. First of all, in countries where the levels of poverty are still very high, social assistance programs may have important redistributive effects and must be, therefore, a pillar of the system. A requirement that must be fulfilled by these programs, within the framework of the principles of the social policy we have formulated, is its universal vocation towards the social groups that are considered objectives (nutrition programs, conditioned grants and pensions for poor elderly people, among others). But said programs should be considered as subsidiary of a basic social policy with universal character and, for this reason, they should be integrated, as far as possible, to said policy. The second function of focalized programs is to allow the design of special plans adapted to certain populations (such as the indigenous peoples) or specific groups of the population (for example, pension systems that take into account women’s reproductive activity). The third function is to be a bridge to ensure that sectors with difficulties in accessing basic universal social services may obtain said access. In all these cases, focalization (or selectivity) must be seen as an instrument of universalization, never as its substitute9.
The progressiveness of spending in different countries
The estimations about the total redistributive effect of social public spending in Latin America corroborate these appreciations. Graphic 3 shows the relation between the redistributive effect of social spending in the different countries (estimated as points of the Gini coefficient) and an aggregate indicator of the social policy development, the Human Development index from the UNDP, although taking into account only education and health components of said index. The first of these variables includes not only the impact of focalization in the strict sense, but also that of the social spending dimension. Both factors are reinforced, because, since a better focalization –and, in consequence, more redistributive impact – is associated to more universal systems, spending dimension should also be further, as explained below.
The data indicates, in fact, that the countries where public spending has a more redistributive effect are those that developed the State’s social services earlier and the ones that have achieved the best coverage extension in their education and health care systems: Argentina, Chile, Costa Rica and Uruguay (to them, we must add Cuba, which is not usually included in the analysis of these issues).
Note that in some estimations oriented to the exclusive analysis of focalization, the redistributive effect of public spending in Chile stands out as the most redistributive10, something that does not occur, however, when the joint impact of focalization is combined with the dimension of spending11.
The graphic shows that the countries with an intermediate degree of development in their social policies, such as Brazil, Colombia and Panama, also generate redistributive impacts on the intermediate social spending, whereas, less redistributive impact is achieved in nations with less relative development, such as Bolivia, El Salvador, Honduras and Guatemala. Other countries, such as Ecuador, Mexico and Peru, stand out because they have a limited redistributive impact on their social policies with medium human development indicators 12.
A view of the specific data corresponding to social security (pensions) reveals more complexities. In this case, the evident relation is between the redistributive impact, measured by the quasi-Gini of the corresponding spending, and the degree of labor formality (chart 2). In general, data corroborate that the spending is more redistributive as long as there is more access, measured through the degree of labor formality. However, there are notorious exceptions, positive as well as negative. Among the positive ones, Brazil and Bolivia are outstanding, as they have developed universal basic pensions with a non-contributive character. In the negative aspect, some systems have achieved a very limited coverage of the pension regime, given its degree of labor formality. Honduras and Colombia are the most outstanding cases (also Guatemala, but the reason for that may be the low labor formality). The data about the redistributive impact corresponding to Argentina in 2002 only refers to non-contributive pensions and, in consequence, it is not strictly comparable to the rest.
The conclusions of this analysis, as well as from the prior section, coincide with the conclusions of Korpi and Palme (1998), who, upon analyzing the social protection systems of industrialized countries, have called «the redistribution paradox»: that the total redistributive effect of social public spending is higher in countries with universal systems than in those countries which apply focalization criteria (particularly Anglo-Saxon countries). According to these authors, one of the main reasons for that is the strong political support of social public spending by the middle class in those countries with more universal systems, which is essential to obtain the support of said sectors at the high levels of taxation needed to fund them.
It is worthwhile mentioning, in addition, that the estimated redistributive effect mostly refers to indirect transfers to homes equivalent to the cost of provision of the social services they received from the State. Only in a few cases, they associate with direct income transfers (pensions and some social assistance programs). Some recent studies have highlighted that Latin America’s delay in the development of said direct transfers is one of the most important explanations for the high levels of inequality in income distribution in the region, compared to that of the OECD countries (World Bank, chap. 5; OECD, chap. 1).
Finally, it should be noted that the data in graphic 3 refer to the social policy impact on the secondary distribution of income. A further argument that should be highlighted refers to the effect of social policy on primary distribution of income. The best illustration of this issue is provided by the analyses corresponding to the industrialized countries. Said analyses clearly corroborate that the countries in Continental Europe, which count on more universal welfare systems, have better income distribution than those countries where the focalization principle is used more actively (Alesina/Glaeser). Undoubtedly, the causes are two-way: more equalitarian societies claim universal social policy systems, and the latter, in turn, contribute to generate more equalitarian societies. If the social policy is going to contribute to reduce the excessive inequality in income distribution which characterizes Latin America, it must therefore go with a social policy of universal nature –including, as the World Bank and OECD studies point out, elaborated systems of income transfers.
This association between the development of social policy systems and the primary distribution of income is not uncommon in Latin America, but it has tended to weaken with the distributive deterioration experienced in the last decades by almost all the countries of the region with an advanced degree of human development, as part of the tendency that the Cepal (2006b) defined as the «convergence towards more inequity ». Of the four countries with higher distributive impact of social policy, the cases of Costa Rica and Uruguay still tend to corroborate the association between social policy and better income distribution (and, it may be added, support to democratic systems), although the first has experienced distributive deterioration in recent decades. Argentina, instead, experienced the deepest tendency towards distributive deterioration between the seventies and the beginning of the current decade and, in spite of its recent improvement, still has inequality levels not very different from the regional average. Chile also experienced a distributive deterioration in the seventies and since then it can be found amongst the countries with the highest levels of inequality.
The fundamental restriction for a universal social policy is the demand for public resources. One of the advantages of focalization – and perhaps one of its greatest political attractions– is that it requires fewer resources. Thus, its spreading was associated to the efforts to rationalize public spending. This fact reveals one of the fundamental weaknesses of Latin American’s development: the fragility of its fiscal structure and the weight of the indirect systems. Progress towards a universal social policy will not be possible, in consequence, without a new fiscal pact, to use the term coined by the Cepal (1998), which significantly raises public income and turns fiscal policy in a progressive instrument.
The works by the Cepal (1998), the WB (2006, chap. 5) and the OECD (2007, chap. 1) confirm these appreciations. The study by the OECD shows that the average fiscal burden in Latin America is only half of that of the OECD (17% against 36% of the GDP). Big disparities appear in direct taxes (5% in Latin America against 15% in the OECD) and in the contributions to social security (3% in Latin America against 9% in the OECD). The WB, in turn, estimates that, given the international patterns of relation between taxation and income levels, Latin America should collect four additional points of GDP, above all, in the concept of income tax, especially from individuals. As it was pointed out in these studies, indirect taxation does not show, on the contrary, similar remnants. This confirms that the countries in the region have more regressive structures than the countries of the OECD and that, progress in the field of direct taxation may contribute to reduce the excessive distributive disparities that characterize the region.
In this framework, a controversial issue is the contributions to the social security and taxes to fund other types of public programs (such as labor training programs) which are collected from the payroll. The fundamental criticism to these contributions is that they constitute a tax to labor which reduces, therefore, the generation of formal work and deepens labor informality. This observation is undoubtedly correct and has led to proposals to partially fund these benefits with general taxes. However, it does not seem possible or convenient, that the entire social policy system rests on general taxes. A brief look to the OECD’s data shows, in fact, that industrialized countries could not fund the high benefits of social security without the new points of the GDP collected through this via. But, in addition to not being financially feasible, a system only based on general taxes would eliminate an essential element of political economy: the contributive systems generate a feeling that the worker belongs to the welfare state, which is irreplaceable and which implies not only rights (services), but also duties (contributions).
For this reason, the idea of separating the financing of social policy from the contributions to the payroll becomes stronger in those cases where public goods are generated, or important externalities (programs to prevent transmissible diseases or labor training, for instance) and there is no direct relation between the tax payers (or what the company contributes in the name of the worker) and those who receive the benefits. But when the benefit principle is applied, as occurs with social security in a wide sense, the contributions bound to the payroll are irreplaceable. The key for the progress of social protection in a context where formal employment remains limited should be to consider, in consequence, a careful mixture of the contributive and non-contributive pillars of the system (Cepal 2006a), but not the elimination of the former.
The importance of taxation is confirmed, finally, in the tendencies of social public spending verified in Latin America. In fact, one of the most positive tendencies in the nineties was the significant increase of social public spending, which rose, in average, from 12.8% to 15.5% of the GDP, thanks, partly, to the substitution of other kinds of spending. The increase, although generalized, was more notorious in the countries which were lagging behind. However, this tendency was followed in recent years by a more paused progress, as in 2004-2005 when said spending was 15.9% of the GDP (Cepal, 2007, chart II.6).
In summary, progress in the field of fiscal policy in Latin America resides inevitably, in a change of taxation and, especially, in direct taxation and contributions to social security. This is one of the aspects where it may be verified whether the return of equity to the agenda of the regional development has deep roots or is just planted in a rhetoric field.
This paper argues that the approaches of social policy that have prevailed in recent decades, with the emphasis on focalization, must leave room for schemes based on the concept of social citizenship and, therefore, on the principles of universality and solidarity. The result of this will be the development of social services that may be universalized effectively, and in turn, said universalization must be the main instrument to achieve a greater focalization of public spending on the low income sectors.
In a scheme as the one proposed, focalization (or better yet, selectivity) should be seen as an instrument of the universalization, rather than its substitute. It may perform three basic functions: develop social work programs with universal vocation within certain social groups, adapt the programs to the specific characteristics of some sectors and to act as a bridge towards the universal programs, as occurring with grants.
A further issue is the importance of fighting against segmenting the access of different social groups to different services. As pointed out above, this problem tends to be acute in focalized policies, but it is also the result of systems in which there are multiple service providers, some of which tend to discriminate between the different petitioners. A special source of concern in that sense, are the high levels of segmentation that characterize educational systems, which implies that the systems where multiple actors take part require stricter standards oriented to the correction of these problems.
Finally, a fundamental requirement in the turn towards a universalist conception of social policy is to strengthen tax systems and to make them more progressive. In consequence, the progress towards a social policy of a universalist nature will not be possible without a new fiscal pact that reinforces, in particular, direct taxation. Moreover, given the much slower pace of the increase of social public spending during the current decade, the efforts to deepen the progresses of social policy will be much more demanding in the future, in fiscal terms, compared to the past.
Alesina, Alberto and Edward L. Glaeser: Fighting Poverty in the US and Europe: A World of Difference, Oxford University Press, New York, 2004.
Banco Mundial: Reducción de la pobreza y crecimiento: círculos virtuosos y círculos viciosos, World Bank / Mayol Ediciones, Washington, D.C.-Bogotá, 2006.
Corporación Andina de Fomento (CAF): Reporte de economía y desarrollo 2007/2008. Oportunidades en América Latina, CAF, Caracas, 2007.
Comisión Económica para América Latina y el Caribe (Cepal): El pacto fiscal. Fortalezas, debilidades, desafíos, Libros de la Cepal Nº 47, Santiago de Chile, 1998.
Comisión Económica para América Latina y el Caribe (Cepal): Equidad, desarrollo y ciudadanía, Cepal / Alfaomega, Bogotá, 2000a.
Comisión Económica para América Latina y el Caribe (Cepal): La brecha de la equidad: una segunda evaluación, Cepal, Santiago de Chile, 2000b.
Comisión Económica para América Latina y el Caribe (Cepal): La protección social de cara al futuro. Acceso, financiamiento y solidaridad, Cepal, Santiago de Chile, 2006a.
Comisión Económica para América Latina y el Caribe (Cepal): Panorama social de América Latina 2004, Cepal, Santiago de Chile, 2006b.
Comisión Económica para América Latina y el Caribe (Cepal): Panorama social de América Latina 2005, Cepal, Santiago de Chile, 2006c.
Comisión Económica para América Latina y el Caribe (Cepal): Panorama social de América Latina 2007, Cepal, Santiago de Chile: versión preliminar, November 2007.
Cornia, Giovanni Andrea and Frances Stewart: «Two Errors of Targeting» in Dominique van de Walle and Kimberly Nead (eds.): Public Spending and the Poor: Theory and Evidence, Johns Hopkins University Press, Baltimore-London, 1995, pp. 350-386.
Esping-Andersen, Gøsta: The Three Worlds of Welfare Capitalism, Polity Press, Cambridge, 1990. [There is a Spanish edition: Los tres mundos del Estado de bienestar, Alfons el Magnànim, Valencia, 1993.]
Filgueira, Fernando, Carlos Gerardo Molina, Jorge Papadópulos and Federico Tobar: «Universalismo básico: una alternativa posible y necesaria para mejorar las condiciones de vida» in Carlos G. Molina: Universalismo básico. Una nueva política social para América Latina, Banco Interamericano de Desarrollo / Planeta, Washington, D.C., 2006, pp. 19-55.
Grynspan, Rebeca: «Universalismo básico y Estado: principios y desafíos» in Carlos G. Molina: Universalismo básico. Una nueva política social para América Latina, Banco Interamericano de Desarrollo / Planeta, Washington, D.C., 2006, pp. 75-81.
Korpi, Walter and Joakim Palme: «The Paradox of Redistribution and Strategies of Equality: Welfare State Institutions, Inequality and Poverty in the Western Countries» in American Sociological Review vol. 63 No.5, 10/1998, pp. 661-687.
Marshall, Thomas Humphrey: «Citizenship and Social Class»  in T.H. Marshall and Tom Bottomore: Citizenship and Social Class, Pluto Press, London, 1992.
Mkandawire, Thandika: «Targeting and Universalism in Poverty Reduction» in José Antonio Ocampo, K.S. Jomo y Sarbuland Khan (eds.): Policy Matters: Economic and Social Policies to Sustain Equitable Development, Orient Longman, Himayatnagar; Zed Books, London and Third World Network, Penang, 2007, pp. 305-333.
Molina, Carlos Gerardo (ed.): Universalismo básico. Una nueva política social para América Latina, Banco Interamericano de Desarrollo / Planeta, Washington, D.C., 2006.
Ocampo, José Antonio: Reconstruir el futuro. Globalización, desarrollo y democracia en América Latina, Grupo Editorial Norma / Cepal, Bogotá, 2004.
Organización para la Cooperación y el Desarrollo Económico (OCDE), Centro de Desarrollo: Perspectivas económicas de América Latina 2008, OCDE, Paris, 2007.
Programa de las Naciones Unidas para el Desarrollo (PNUD): La democracia en América Latina. Hacia una democracia de ciudadanas y ciudadanos, Aguilar / Altea / Taurus / Alfaguara, Buenos Aires, 2004.
Programa de las Naciones Unidas para el Desarrollo (PNUD): Informe sobre Desarrollo Humano 2007/2008, PNUD / Grupo Mundi-Prensa, New York-Madrid, 2007.
Rodrik, Dani: Has Globalization Gone Too Far?, Institute of International Economics, Washington, D.C., 1997.
Selowsky, Marcelo: Who Benefits from Government Expenditure? A Case Study of Colombia, World Bank, New York, 1979.
Sen, Amartya: «The Political Economy of Targeting» in Dominique van de Walle and Kimberly Nead (eds.): Public Spending and the Poor: Theory and Evidence, Johns Hopkins University Press, Baltimore-London, 1995, pp. 11-24.
Sen, Amartya: Development as Freedom, Alfred A. Knopf, New York, 1999. [There is a Spanish edition: Desarrollo y libertad, Planeta, Barcelona, 2000.]
Thalen, Ingela: «Equidad y economía global: un comentario desde Suecia» in Heraldo Muñoz (ed.): Globalización XXI. América Latina y los desafíos del nuevo milenio, Aguilar Chilena de Ediciones, Santiago de Chile, 2000.
Vélez, Carlos Eduardo: Gasto social y desigualdad: logros y extravíos, Departamento Nacional de Planeación, Misión Social, Bogotá, 1996.
1 Here I use the «corporate» concept as do other authors in literature about social welfare systems, to refer to the tendency of these systems to offer different services for different groups of workers. It is worthwhile highlighting that this use differs from the one in political science. Therefore, it does not refer to the existence or non-existence of corporate political systems.
2 This classification is not necessarily exhaustive, but I use it here because it allows direct comparison with Esping-Andersen’s already known conceptual scheme.
3 Article 25.1 of the Universal Declaration of Human Rights.
4 Articles 26.1 and 22 of the Universal Declaration of Human Rights.
5 This contribution, also essential for the debate, came from a project promoted by the Inter American Institute for the Social Development of the Inter American Development Bank (IDB). See, in particular, the «Introduction» to said collective work, written by the editor, as well as Chapter 1 (Filgueira et al.).
6 In this regard, see Cepal (2007), chart annex 6.
7 The estimated data for Colombia according to Cepal (2007, chart II.16), corresponding to 2003, do not differ significantly from those in chart 1 for the year 1992.
8 Although referred to social security in health, the study by Vélez (1996) illustrates this issue very well. Thus, even though the coverage of said system was slightly regressive in Colombia in 1992, net grants were highly progressive, with a quasi-Gini of -0.345; in fact, not very different from the one on primary education.
9 Please, note that these arguments, as well as the ones introduced in the following section, have a positive meaning: trying to show that the universalization criteria are clearly superior in terms of redistributive impact. There is also critical literature about focalization that points out the information errors, distortions in incentives and the stigmatization problems which characterize said systems. See, among others, Giovanni Cornia and Frances Stewart (1995), Thandika Mkandawire (2007) and Amartya Sen (1995).
10 See, for example, Cepal’s estimations (2006c).
11 In stricter terms, the data corresponding to that country are not comparable to the others, because they exclude health contributive systems and pensions managed by the private sector and which, in other countries, remain the State’s responsibility. A corollary of this is that the calculations about the spending redistributive effect should consider, in the future, resources and services managed by private or charity entities participating in the social policy systems, or, what is equivalent, they should be net estimates of all the contributions from those who access the corresponding systems.
12 We have excluded data referring to Nicaragua, which show the opposite effect, because the amount of social spending incorporated in the corresponding calculations widely exceeds the dimension of social spending the Cepal is currently estimating for said country.