STATE (IN)ACTION AND INEQUALITY

In their World Bank study, de Ferranti et al. (Reference de Ferranti, Perry, Ferreira and Walton2004, 109–47) provide a very comprehensive overview of research on the historical social origins of inequality in Latin America, arguing that the factor endowments that the colonial powers encountered shaped the institutions that they imposed and that these institutions concentrated wealth and power in the hands of a small minority. Concentration of wealth and power, exploitation of indigenous and imported slave labor, and political exclusion of the majority survived into the independence period and entailed a neglect of the expansion of public education. High inequality generated clientelistic politics and thus prevented the emergence of broad-based reformist political coalitions after the advent of mass politics in the twentieth century. Elites left the state apparatus with a low capacity to provide for macroeconomic stability, property rights, and basic services, which stood in a relationship of mutual reinforcement with unequal social relations. The struggle over economic and political inclusion versus exclusion resulted in political instability and thus erratic economic policy and growth.

This diagnosis suggests a range of state actions that could potentially remedy the situation, but at least so far, most Latin American states have failed to take such action on a large and sustained basis. For instance, even in the second half of the twentieth century, Latin American states failed to implement effective land reform to correct for historic legacies, such that inequality and poverty remained widespread in the rural sector and contributed to massive rural-urban migration. Lack of assets, both land and capital, has also remained a problem in the urban areas. Huge numbers of migrants became squatters, building housing on public lands to which they never obtained title. Without collateral, they had no access to the credit system and thus no opportunities to build small enterprises. Public administrations were reluctant to issue land titles and they by and large failed to devise credit schemes for small entrepreneurs. Development banks typically catered to large and medium enterprises, leaving the promotion of small or micro enterprises mainly to civil society initiatives.

The import substitution industrialization model (ISI), followed by the more advanced Latin American countries from the 1930s on and by others later, generated economic growth and formal-sector jobs but came to depend on massive external borrowing in the 1970s and thus ended with the debt crisis in the 1980s. Even during its height, the formal-sector jobs it created were not nearly sufficient to absorb the increase in the labor force resulting from population growth and rural-urban migration. Thus, a dual economy emerged and reinforced income inequality. Government policy favored capital-intensive growth and foreign investment in industry and natural resources, not job creation in small enterprises and small farms.

The key to redressing inequality of market income is the system of taxes and transfers. However, the low levels of tax revenue, the structure of the tax system, and the patterns of expenditure have obstructed significant redistribution. Latin American countries as a whole have been under-taxing their populations, with an average tax burden of 14% of GDP in the first half of the 1990s, compared to 17% of GDP in a group of East and Southeast Asian countries (IDB 1998, 128). By 2002, this figure had risen slightly to 16%, compared to 36% for OECD countries (Centrángolo and Gómez Sabaini Reference Centrángolo and Gómez Sabaini2006, 53). Direct taxes, generally the most progressive element of taxation, amount to about 25% of tax revenue only, and of this amount some 60–80% typically come from corporate tax payments, while only 10–15% come from private individuals (ECLAC 1998, 72). Tax reform has been high on the agenda, but the dominant pattern of tax reforms over the past three decades in Latin America and the Caribbean has consisted of reductions of marginal tax rates on corporate and individual incomes and attempts to broaden the tax base (Tanzi Reference Tanzi2000). On the whole, the tax systems in most Latin American countries in the 1990s have been at best proportional and more commonly slightly regressive (Engel, Galetovic, and Raddatz Reference Engel, Galetovic and Raddatz1998; Gómez Sabaini et al. Reference Gómez Sabaini, Santieri and Rossignolo2002).

Even a distributionally neutral tax system could be a powerful tool for redistribution if it made a significant share of GDP available for social expenditures, and if these expenditures were allocated in a way that is less unequal than market income. ECLAC data for eight countries in the region show that the most progressive types of expenditures are spending on primary and secondary education, and that public spending on health care and nutrition is the second most progressive category (2005, 26). Within health and education, there are also regressive components, particularly if they involve public subsidies for privately provided services, but in general the progressive components tend to outweigh the regressive ones (de Ferranti et al. Reference de Ferranti, Perry, Ferreira and Walton2004, 263–64). Total spending for public education is progressive in Argentina, Chile, Mexico, and Uruguay, but regressive in Brazil (de Ferranti et al. Reference de Ferranti, Perry, Ferreira and Walton2004, 264).

Transfers as a whole are regressive, because the bulk of transfers goes to social security pensions, and these pensions are employment based and segmented and have earnings-related benefits. Typically, social security pension schemes were first established for privileged groups who then kept their privileged schemes even when pensions were later extended to other groups (Malloy Reference Malloy1979; Mesa-Lago Reference Mesa-Lago1978; Huber Reference Huber and Esping-Andersen1996). In the great majority of countries social security coverage remained confined to formal-sector employees, which means that often 20% to 60% of the economically active population remained excluded. In their review of a range of studies, de Ferranti et al. (Reference de Ferranti, Perry, Ferreira and Walton2004, 268–72) find that in most countries the regressive components of social security spending outweigh progressive components. ECLAC (2005) and Lindert, Skoufias, and Shapiro (Reference Lindert, Skoufias and Shapiro2006) also show that social security spending provides greater benefits to middle and upper strata.

Looking at recent trends in inequality in Latin America, we can see that it increased in most countries in the 1980s, remained largely constant in the 1990s, and decreased in a few countries in the 2000s. The economic crisis and subsequent opening of the economies was accompanied by a shrinking of the formal sector, an increase in unemployment, and an expansion of the informal sector. Official unemployment figures vastly underestimate the true dimensions of the problem, because people barely surviving with activities in the informal sector are not counted as unemployed. Within the informal sector, some people are making a good living, but the majority of people are barely making enough to help their families survive. Workers employed in small enterprises in the informal sector earn less than workers in the formal sector, even controlling for experience and years of schooling. The same is true for self-employed workers, the vast majority of whom are in the informal sector. Moreover, the difference between male and female earnings is larger among workers in the informal than in the formal sector and among the self-employed than formal sector workers (IDB 1998, 40). Thus, the size of the informal sector contributes to high inequality (de Ferranti et al. Reference de Ferranti, Perry, Ferreira and Walton2004, 159–61). In the 1980s and 1990s, roughly 8 out of every 10 new jobs created were in the informal sector (Tokman Reference Tokman and Huber2002, 161). It is important to keep in mind the magnitude of the phenomenon: Portes and Hoffman (Reference Portes and Hoffman2003) calculate a weighted average for eight Latin American countries of 46% of the labor force belonging to the informal proletariat in the 1990s, that is, non-contractual waged workers, casual vendors, and unpaid family workers.

There are a number of cash-transfer programs that are not employment based and earnings related and are progressive, such as non-contributory pensions and conditional cash transfers. Conditional transfers are targeted to poor households and make receipt of benefits conditional on participation in education and health programs. They generally are highly progressive. The comprehensive review of conditional cash-transfer programs by Fiszbein and Schady (Reference Fiszbein and Schady2009), however, shows that they are limited in coverage and financing, reaching a maximum of 0.5% of GDP. The largest coverage amounts to roughly 20% of the population in Brazil and Mexico and 40% in Ecuador.

Non-contributory, means-tested social-assistance pensions are still relatively scarce and poorly funded as well (Muller Reference Muller, Messner and Scholz2005). The most significant programs exist in Brazil for the rural sector and in Uruguay, accounting for 1% and 0.6% of GDP, respectively (Bertranou, van Ginneken, and Solorio Reference Bertranou, van Ginneken and Solorio2004, 7). However, these kinds of programs are highly effective in reducing poverty and destitution among their recipients, not only in Brazil and Uruguay, but also in Argentina, Chile, and Costa Rica (Bertranou, van Ginneken, and Solorio Reference Bertranou, van Ginneken and Solorio2004, 13). The recent Chilean pension reform will provide social-assistance pensions and supplementary pensions to the elderly with no or very low pensions who belong to the lowest three quintiles of income earners.

On the basis of a number of effort and outcome indicators, Argentina, Chile, Costa Rica, and Uruguay emerge as countries with the most effective social-policy regimes in the past 40 years (Huber and Stephens, Reference Huber, Stephens, Mainwaring and Scullyforthcoming). Chile slipped under Pinochet but has made great strides under the Concertación governments since 1990. Argentina was deeply affected by the crisis of 2001–02, but has restored health and education spending and continued transfers to the poor of working age through the emergency employment program. What distinguishes state policy specifically in Uruguay and Costa Rica from that of other countries is the consistently high investment in social capital, particularly in education and health services, and the effort to expand coverage of the pension system and integrate all pension schemes. Literacy in Uruguay reached 94% in 1976, the same as in Argentina; Costa Rica, at a much lower level of economic development, reached 91% in 1977. By 1980–86 the primary-school completion rate in Uruguay was 88 and in Costa Rica 75, the two highest levels in Latin America (Morley Reference Morley1995, 62). Coverage of the health system was expanded consistently. Infant mortality was comparatively low in 1970–75 already, with 53 per 1,000 live births in Costa Rica and 46 in Uruguay (Morley Reference Morley1995, 67), and it fell to 11 in Costa Rica and 14 in Uruguay by 2002 (UN 2003). Again, this is remarkable for Costa Rica, given the comparatively low GDP per capita. As of the early 1970s, Uruguay reached 84% of the population with health care through a combination of public services and mutual insurance schemes, the latter linked to social security financed through employer and employee contributions (Gaudiano Reference Gaudianon.d.). Costa Rica essentially built a unified national health system in the 1970s, comprising both the public and social security parts and marginalizing private provision. Coverage expanded from 38% in 1970 to 84% in 1979 and 85% by 1989 (Mesa-Lago Reference Mesa-Lago1994, 96). In the 1990s, the public network of primary and preventive care centers was greatly strengthened, at a time when participation of the private sector in health care was being expanded in other countries where access became more difficult for poorer sectors. The capacity to reform the system in Costa Rica in this direction was due to the fact that there were fewer powerful private stakeholders capable of blocking such reforms than elsewhere. In general, including advanced industrial countries, in systems where health insurance and provision are heavily private, total costs, both public and private combined, are high, and there is greater potential for inequality in access than in public systems.

CAUSES OF STATE (IN)ACTION

What explains the comparatively successful policies of Uruguay and Costa Rica, as well as the experiences of Argentina and Chile, and to what extent can these causal processes be generalized? Uruguay and Costa Rica have the strongest democratic records in Latin America. Uruguay has been democratic since 1911, with only two interruptions, and Costa Rica has been democratic since the early 1950s. This means that democratic processes are firmly institutionalized and political parties have had time to become major actors. In particular, in both countries parties with a social-democratic orientation were able to attract large support bases and leave their imprint on the social-policy regimes. Research on advanced industrial democracies has shown that incumbency of social-democratic parties is a key predictor of welfare-state generosity and of a redistributive welfare-state structure (Huber, Ragin, and Stephens Reference Huber, Ragin and Stephens1993; Bradley et al. Reference Bradley, Huber, Moller, Nielsen and Stephens2003). The PLN in Costa Rica and the Colorados in Uruguay, at least the factions identified with the Battlista legacy, have been promoters of investment in human capital and social safety nets. In addition, in Uruguay parties to the left of the Colorados have grown in strength and as competitors they have made it difficult for the traditional parties to steer away from a commitment to maintaining strong social policies. Since coming to power in 2005, the left-of-center government of the Frente Amplio has aggressively expanded progressive transfers and improvements in health and education. In Chile, Socialist presidents Lagos and Bachelet launched highly ambitious health care and pension reforms benefiting lower-income earners.

Results from our pooled cross section and time series analysis of data for 18 Latin American countries from 1970 to 2000 suggest that these explanations can be generalized to other Latin American countries. We find that strength of the democratic record is a significant determinant of public spending on health and education and on social security and welfare (Huber, Mustillo, and Stephens Reference Huber, Mustillo and Stephens2008). Kaufman and Segura-Ubiergo (Reference Kaufman and Segura-Ubiergo2001) find positive effects of democracy on health and education spending also, but a weak negative effect on social security spending.Footnote ¹ When we include the strength of different party blocs in the legislature or the political color of executives in the analysis, we find no effects on any of the social expenditures. This suggests that governments of all political orientations were more responsive to demands of their constituencies for greater state efforts in social policy.

In another pooled cross section and time series analysis of the determinants of inequality in Latin America we find a negative effect of the strength of the democratic record on inequality, as well as a negative effect of strength of left parties in the legislature (Huber et al. Reference Huber, Nielsen, Pribble and Stephens2006). This effect, along with the high correlation between democracy and left strength, suggests that extended periods of democracy make it possible for organizations representing the interests of the underprivileged to emerge, consolidate, and form a counterweight to the political influence of privileged groups. Where left parties achieve electoral strength, they use their parliamentary representation to shape policy in a progressive direction, via more progressive allocation of public expenditures, and presumably a host of other policies that affect distribution, such as minimum-wage laws.

If we now turn from differences within Latin America and the Caribbean to differences between these countries and advanced industrial democracies, it becomes clear that weakness of democratic institutions and weakness of the Left are major factors explaining the higher inequality in the former. Weakness of parties of the Left in turn is a result of the combination of weakness of democracy and weakness of organized labor. Weakness of organized labor is partly a function of economic structure, that is, a relatively small percentage of the labor force employed in industry, as a result of late industrialization based on imported technology. Partly, it is also a function of the weakness of democracy. Both parties of the Left and organized labor, rather than enjoying the democratic freedom to organize, suffered severe repression, forcing parties of the Left to abandon much of their reformism as the price of being allowed to participate in the political process (e.g., Apra in Peru 1932–62, AD in Venezuela 1948–58). Parallel to repression, state elites made efforts to encapsulate labor in institutions controlled from above. Whereas such efforts could in fact raise the level of labor organization, they thwarted the independent political articulation of labor and thus kept the base for parties of the Left weak.

The increase in electoral strength of left parties in the 1990s and early 2000s, precisely when unions were so debilitated, seems to challenge the argument about the connection between strength of organized labor and strength of the Left. However, it is worth remembering that the Labor Party (PT) in Brazil grew out of the labor movement there. Indeed, Madrid (Reference Madrid2004) found that the unionization rate was a consistently statistically significant and substantively important determinant of the Left's share of the vote in Latin American elections in the period 1978–2002. The recent rapid growth of left electoral support is largely a result of popular disenchantment with the economic performance of incumbent traditional parties, and the presence of a left alternative that has demonstrated governing capacity in subnational units. One should keep in mind that the same disenchantment generated support for the populist authoritarian Fujimori in Peru, where no credible left alternative was available. In other words, the recent expansion of electoral support for the most part does not rest on an ideologically committed and reliable social base, but one that is likely to diminish if incumbent left parties prove unable to modify public policy sufficiently such as to improve popular living standards.

Given the intrinsically political nature of social policy, redistribution of resources in the political playing field is a prerequisite for major progress in building redistributive social-policy regimes. Strengthening of democratic institutions is necessary but not sufficient. We have seen that strength of the democratic record does depress inequality, but so does political strength of the Left. Parties of the Left are at a disadvantage everywhere compared to parties of the Right in terms of the financial base of their supporters and supportive media coverage. Economically privileged groups support parties that protect these privileges, and privately owned print and electronic media do not tend to favor parties with redistributive programs that might affect the interests of their family owners or shareholders negatively. Thus, the increasing importance of television for political campaigns has worked to the disadvantage of left-of-center parties. Legislation on public financing of parties and of political campaigns, on limits on private contributions, and on provision of free and equal media access is in its infancy in Latin America (Posada-Carbó and Malamud Reference Posada-Carbó and Malamud2005). Accordingly, it has a low potential to counteract the translation of socioeconomic inequality into political inequality and thus into a barrier to state action designed to ameliorate the former.