Social protection in Latin America: one region two systems

Social protection systems in Latin America have experienced deep structural change in the last two decades. Policies, programmes and institutions have emerged in all countries in the region directed at population groups characterised by low incomes and informal employment. Social assistance – budget-financed and rules based programmes providing transfers and services to disadvantaged groups – reaches around one third of the population in the region. It plugs a longstanding gap in the region's social protection institutions, historically rooted in Bismarckian social insurance. Structural change, the chapter argues, has resulted in dual social protection institutions: social insurance covering workers in formal employment and social assistance supporting low-income groups in informal employment. The chapter tracks the expansion of social assistance, documents this dualism, assesses explanations for how we got here, and speculates on the future of social protection institutions in the region.

Since the return to democracy in the 1990s, Latin American countries have embraced remarkable policy activism in social policy (Levitsky and Roberts 2012) associated with a newly acquired responsibility of governments to secure social rights. The most visible change is associated with the emergence of large-scale social assistance programmes and policies targeting low income and informal employment groups. They include an expansion of budget-financed transfers to older people (Rofman, Apella, and Vezza 2013; Bosch, Melguizo, and Pages 2013). And the implementation, in all countries in the region, of large-scale conditional income transfer programmes providing regular and reliable transfers to families in poverty conditional on children attending school and accessing primary health care (Cecchini and Madariaga 2011). In addition most countries have strengthened the provision of social services to disadvantaged families (Rossel 2013).

The emergence of social assistance runs counter the historically dominant role of social insurance institutions in the region. It has resulted in a growing dualism in social protection institutions in the region reflected in the type and generosity of support as well as in dedicated institutions. Today, most governments in the region have a Ministry of Labour and Social Security alongside a Ministry/Agency charged with Social Development (Cecchini and Martínez 2011). The expansion of social protection in the region since the return to democracy can best be described as vertical rather than horizontal. Low income and informal groups have not been incorporated into existing social insurance institutions, but into parallel social assistance institutions.

Growing dualism in social protection raises important questions about the future of welfare institutions in the region. The structural limitations of social insurance institutions are well understood (Mesa-Lago 2007). And the emergence of social assistance has received a great deal of attention (Barrientos and Santibañez 2009). Yet, research on the implications of dual institutions is scarce.1 Levy (2008; 2013) has raised strong concerns on the likely impact of social protection dualism on economic incentives and productivity, especially in the context of Mexico. He has advocated a radical path to eradicating dualism (Anton, Hernandez, and Levy 2012). At a more programmatic and institutional level, Ferreira and Robalino (2010) argue for the need to consolidate dual institutions through social protection reforms. The ambitious and influential 'Cepalista' approach has been modified to account for the expansion of social assistance and the stagnations of social insurance. Filguera et al. (2005) advocate for a 'basic universalism' with a focus on scaled down protection for all. Huber and Stephens (2012) support this 'basic universalism'. MartínezFranzoni and Sánchez-Ancochea (2017) advocate for a form of universalism with differentiated protection thus retaining dualism. Teichman (2007) and many others see social assistance as a 'low cost' version of social insurance and argue for a return to the former.

With few exceptions, this literature is primarily normative. It tells us what 'ought' to happen without fully engaging with what explains the current dualism. Huber and Stephens (2012) go furthest in this direction. They provide a detailed analysis of the political and economic factors explaining the stagnation of social insurance, but their support for 'basic universalism' is not grounded on equivalent causal processes.

The main objective of this chapter is to highlight the urgent need for research on the causal processes that led to the growing dualism in social protection institutions in the region. It adopts a regional perspective to address this question. Inevitably, there is significant variation across countries in the region in terms of the reach and scope of social insurance and social assistance. However, there is sufficient similarity to support a regional perspective. It focuses on the two main components of social protection: social insurance understood as programmes addressing work and life course contingencies normally financed through payroll contributions by workers and employers; and social assistance understood as budget-financed rulesbased programmes providing regular and reliable transfers to low income households.

The chapter is organised around four main sections. The next section offers a region-wide perspective on the growth of social assistance. The following section characterises the emerging dualism in social protection institutions. The section after that identifies potential explanations for the emergence of dualism. A final section speculates on the future of dualism in the region.