Introduction

Abstract

This Introduction discusses the main features of the approach adopted in the book. It offers a brief review of the main literature, highlighting key influences on the approach in the book, and new departures. It provides a map to the structure and contents of the book.

This book provides an account of social protection institutions in Latin America. It aims to develop a systematic understanding of the contribution of social protection institutions to shaping economic and social cooperation in the region. It is motivated by an acknowledgement that we lack a settled theory of social protection institutions in Latin America. Comparative study of social protection in the region has accelerated in recent decades. The availability of harmonised data on social protection programmes, their features, implementation, and outcomes has made this possible. Regular household survey data, attitudinal data, and programme evaluation data offer a range of tools to support the study social protection institutions. A substantive comparative research literature built on improving comparative data is making a significant contribution to our understanding of the role and scope of social protection in the region. This book contributes to advance research further towards the development of a theory of social protection in Latin America. By theory it is meant a systematic and comprehensive analysis of the contribution of social protection institutions to shaping social and economic cooperation.

The development of a theory of social protection in Latin America has been delayed by two approaches common in the literature.¹ Implicit in much of the literature on social protection in the region is the assumption that our institutions are on a slow march towards fully fledged European welfare states, taken as the inescapable point of destination. The implication flowing from this assumption is that social protection institutions in the region are best understood as underdeveloped welfare states. Researchers interested in social protection in Latin America have much to learn from theories and methods developed to study welfare states. They provide a wealth of knowledge and practice that constitute an essential toolbox for social protection researchers worldwide. However, a global perspective on social protection will confirm that welfare states are a special case of institutional development, far from an inescapable destination. A general theory of welfare institutions, global in its scope, will be delayed in its advance if its parameters are set by a special case.²

A theory of social protection in Latin America has also been delayed by ‘excess normativity’. It is apparent to any student of social protection institutions in the region that they fall short of securing minimum living standards for the population. There is consensus among researchers and policy makers that access to social protection is inadequate for the population groups that need it most. Some social protection institutions may even contribute to raise income inequality. Social assistance does not reach all those in poverty and might be insufficient to lift those it reaches out of poverty. Social protection in Latin America is pro-old, gendered, and compensatory. There is a large, and important, literature comparing these outcomes to normative standards and proposing reforms to achieve them.³ Recent basic income proposals are perhaps an extreme example of this approach. However, an understanding of the contribution of social protection institutions to social and economic cooperation is a necessary condition for a successful reform of these institutions. On its own, a normative approach is unlikely to succeed.⁴

1.1 Why Focus on Social Protection?

Social protection is a component of social policy. Social policy is understood as the collective provision of transfers and services to the population of a country in response to agreed social goals and financed from government revenues. It includes transfers in kind or public services and transfers in cash or income transfers. The provision of public services includes inter alia education, health care, housing and sanitation, and national security. They are transfers in kind.

Social protection consists of publicly supported programmes directed at ensuring minimum standards of living for workers and their dependants. It describes institutions and policies protecting citizens from life course and employment income risks, ensuring minimum living standards, supporting investment in their productive capacity, and facilitating political participation (Barrientos, 2013). Social protection programmes are mainly associated with income transfers and with regulations ensuring minimum income levels, such as minimum wage or job termination regulations.

Societies provide a range of income transfers in response to the needs of population groups. They include, inter alia, child and family transfers, maternity benefits, subsidies to low paid workers, and pension benefits for older groups and people with disabilities. For our purposes, they can be arranged in two main groups: insurance and assistance. Insurance transfers address the impact of a range of contingencies on income or employment. They are in the main financed by contributions from workers and employers and compensate affiliates when affected by pre-specified contingencies. Assistance transfers provide income subsidies to low-income groups to supplement their consumption and support human capital investment. Assistance transfers are budget financed. Insurance transfers can be financed from a range of sources: payroll contributions to occupational funds, payroll taxes, or budget subsidies.

Analytically, social protection institutions can be arrayed on a continuum.⁵ Social insurance is financed from mandated payroll taxes and covers the entire population. Occupational insurance is financed from payroll contributions and is restricted to contributing members only. Social assistance is financed from budget revenues and entitlements are based on need and residence. Insurance and assistance embed forms of social stratification specific to countries or regions. This is why the qualifying term ‘institution’ is essential. Institutions are best described as “rules and procedures (both formal and informal) that structure social interaction by constraining and enabling actors’ behaviour” (Helmke & Levitsky, 2004, p. 727). Understanding social protection institutions necessitates some understanding of the stratification they embed.

As the review of the main literature below will underline, most studies focus on the broader field of social policy and examine social protection as a component. In this respect, research in the Latin American region is heavily influenced by the welfare state literature in which welfare states combined transfers in cash and in kind. However, this book focuses exclusively on social protection and leaves aside any detailed examination of public services. There are some gains and some losses from adopting this approach.

Basic services such as education and health care provide transfers in kind. At the micro-level they are essential to secure well-being, productivity, and full participation in society, same as social protection. Basic services have many complementarities with social protection. Take for example the influence of education on skills or the contribution of health care to labour productivity. Some income transfers support access to education and health, as in conditional income transfers. Some active labour market policies combine services and transfers to maximise the employability of people with disabilities, or the long term unemployed.

There is a macro-level complementarity which can best studied when examining services and transfers together. Studies of social spending are often deployed to characterise government social policy effort and changes over time. Comparative studies of the evolution of welfare states among early industrialisers have been able to identify a specialisation on services, especially education, in the United States and a specialisation on income transfers in European countries (Lindert, 2004). Cross-welfare states comparisons resting on one component would fail to get the whole picture (Alesina et al., 2001).

In the context of Latin America, however, a focus on social protection is justified. Latin American countries have not constructed comprehensive welfare states and, aside from health insurance, they developed social protection institutions largely independently from public services. Education provision has been universalised across the region, health provision has more recently began to be extended to most of the population, but social protection has remained selective. Research on social spending has demonstrated that trends in service provision often diverge from trends in social protection. Researchers focused on the impact of democratisation or globalisation on social spending, for example, have concluded that service provision and social protection appear to have a different political logic (Avelino et al., 2005; Kaufman & Segura-Ubiergo, 2001). It makes sense therefore to tackle social protection on its own. The main argument developed in the book would suggest that the distinctiveness of social protection institutions is associated with their role as a stratification mechanism. This provides a strong justification for looking at social protection separately.

Furthermore, combining social protection with basic service provision in the Latin American context can be misleading. The unique place given to Costa Rica and Uruguay as exemplars of successful social policy provision in Latin America rely to an important extent in their provision of basic services. If the focus is relocated to social protection instead, these two countries would remain on the top half of the rankings but not at the very top.

1.2 Theories of Social Protection in Latin America

The analysis below takes existing accounts of social protection in the region as its point of departure. This section provides a brief introduction to the key literature.

Mesa-Lago (1978) offered an early account of emerging social security institutions in the region, including country developments. It also provided an early classification of countries according to the development of their institutions. It is a canonical text and will be referred to often in the discussion that follows. Two contributions to this book need underlining. Mesa-Lago makes it clear from the outset that the main feature of social security institutions in the region is the stratification they generate. As he puts it, “ … is my contention that … a variety of pressure groups with divergent powers constitute the predominant factor in the historical inception of a ‘stratified’ social security system in Latin America and that the system generates significant inequalities in the distribution of its services” (Mesa-Lago, 1978, p. 4).⁶ He identifies three main models in the emergence of social security institutions, one in which pressure groups influence state recognition of occupational funds; a second in which the state co-opts pressure groups by granting them social security protection; and a third in which a combination of pressure groups and political parties influence state provision of social security. The relevant point to underline here is the identification of causal path, indicated by arrows, capturing alternative stratification effects.

An important strand of literature studies the emergence and development of social protection and social policy in Latin America by reconstructing countries development paths and the conditions that enabled them.

Segura-Ubiergo (2007) researches the emergence and development of welfare institutions, welfare states in his designation, in Latin American countries. He focuses on three main theoretical approaches drawn from the welfare state literature: (i) a modernity approach that associates the development of welfare institutions with economic development and especially industrialisation; (ii) an approach that associates welfare institutions with globalisation and economic openness; and (iii) the power resources theory that associates welfare states with the demands of the working class and left parties against a context of deepening democratisation. These three theoretical approaches are tested against institutional development paths of social policy in countries in Latin America.

He focuses on two periods: 1930–1970 and 1970–2003. Relying in Qualitative Comparative Analysis to interrogate data for the earlier period, Honduras, Panama, and Cuba are excluded due to missing data, he finds that the paths shown by welfare institutions can be used to group countries into two clusters: one cluster exhibiting strong economic development, low openness, and democracy or left parties; and another cluster exhibiting weak economic development, high openness, and limited democratisation. The first cluster includes Southern cone countries and Costa Rica. The second cluster includes Andean and Central American countries. His findings suggest that Latin American countries welfare development paths run counter to theories of welfare institutional development emphasising responses to the uncertainties associated with increasing openness. They provide some support for the industrialisation explanation and for the power resources theories. Countries in the first cluster “had a longer history of continuous democracy and stronger labor movements than the group of non-welfare state countries” (Segura-Ubiergo, 2007, p. 71).

Analysis of the second period 1970–2003 interrogates correlates of social spending. The main findings indicate that all three theories of welfare state development could apply in Latin America, but not in the direction predicted. According to Segura-Ubiergo, late and capital intensive industrialisation did not generate a ‘critical mass of workers’ in Latin American countries. The absence of democratic institutions meant that “demands for compensation were either not voiced by groups that may have suffered the consequences of trade openness or not heard because governments were not accountable to citizens for their actions in the public realm through regular elections”(Segura-Ubiergo, 2007, pp. 263–264). Whereas trade openness encouraged demand for social protection in early industrialisers, in Latin American countries welfare institutions developed against the context of protectionist policies.

Huber and Stephens (2012) extend power resources theory to include two dimensions with particular influence in Latin American countries: the role of veto players and that of international actors. Compared to early industrialisers, Latin American countries have a more diverse class structure explaining limited redistribution and a sui generis development path for welfare institutions. To cite, “…the class structure of Latin America is inhospitable to class organization and class political mobilization compared to that of Western Europe…the scope for union organization is much more limited in Latin America …These same features of the class structure have not been favourable to democracy, …weaken the density of civil society and thus the potential for autonomous (i.e. non-clientelistic) political mobilization of lower class groups” (Huber & Stephens, 2012, pp. 25–26). A more diversified class structure and the influence of international organisation explains the presence of weaker and segmented welfare institutions in Latin American compared with European welfare states.

Haggard and Kaufman’s (2008) book is motivated by the ambition to develop a general theory of welfare institutions, with a global scope. It provides as analysis of the development path of welfare institutions across world regions. In their view, welfare institutions development paths can be studied by focusing on ‘critical realignments’ that is a “focus on discontinuities in patterns of political domination… the emergence of new ruling coalitions and political incorporation or exclusion of working class and peasant organizations” (Haggard & Kaufman, 2008, p. 7). Critical realignments help set in place welfare institutions in response to interest coalitions. The incorporation or co-optation of workers organisations to the political system facilitates the development of social protection institutions. “In Latin America, this process of accommodation resulted in the extension of social insurance coverage to some portion of the organized urban working class, programs that became the template for further expansion of entitlements…[ but they were not AB] associated with new political opportunities for the rural poor” (Haggard & Kaufman, 2008, p. 346). They pay close attention to conditioning factors, particularly country differences in terms of democracy, economic growth, fiscal constraints, and development models, but these conditioning factors are mediated through the quality of the social contract in place for specific countries or sets of countries. “Latin America’s stratified welfare system reflected the urban biases of the political coalitions that challenged oligarchic rule. These biases were reinforced by the turn toward import substitution” (Haggard & Kaufman, 2008, p. 79).

Cross-country comparative studies identifying similarities and differences in social policy and social protection institutions offer a complementary perspective to the study of development paths. Following the approach developed in Esping-Andersen’s three worlds of welfare capitalism, this set of studies aims to cluster countries in Latin America according to key features of their social policy (Antía, 2018; Filgueira, 2005; Martínez Franzoni, 2008; Sátyro et al., 2021). This is a fertile area of research because the grouping of countries around their social policy and social protection institutions can help identify structural explanations of welfare institutions and provide evidence for the relevance of alternative theories of social protection.⁷

This brief review of some key contributions to the theory of social protection institutions in Latin America help introduce some of the tools and concepts to be applied in the analysis in this book. Mesa-Lago’s early emphasis on the stratification effects of social protection as its key feature and the hint towards causal inference will be important building blocks for the argumentation in this book. Haggard and Kaufman’s notion of critical realignments as an essential tool to separate qualitatively distinct periods in the evolution of social protection, and the substantive role of worker’s political incorporation or exclusion, will provide an important dimension of the analysis. Huber and Stephens’s emphasis on the role of class differentiation and democratisation in making sense of social protection in the more recent period will guide the discussion below. Comparative analysis identifying country clusters will be taken up and expanded. These insights are central to the analysis in this book.

1.3 Some New Departures…

The analysis in the book takes up many of the themes and findings from this literature. It also makes some new departures. This section will introduce them.

1.3.1 Wage Earners Are the Object of Social Protection Institutions

The analysis in this book takes wage earners as the main object of social protection institutions and their theory. By wage earners, it is meant the population dependent on labour earnings. They are that section of the population whose wellbeing is largely dependent on the income they can generate from selling their labour. They are by far most of the population in the region, especially when their dependents are included. Wage earners as a population category include salaried workers, own account workers who are not employers, unpaid family workers, and their dependents. The justification for focusing on wage earners, a broad category, is that they are the object of social protection. No wage earners, no social protection.

Research on social protection is often focused on a subset of wage earners. The emergence of occupational insurance funds in Latin American countries is predicated on a growing urban industrial working class. The expansion of social assistance in the 2000s aimed to reach population groups on low and irregular incomes and often dependent on informal and precarious employment. Attempting a more comprehensive analysis of social protection requires incorporating these population groups within a broader category, wage earners.

There are some important implications from a broader focus. The more significant implication is to shift the conventional focus of the analysis away from inter-class differentiation and coalitions and replacing it with a focus on intra-class differentiation and coalitions.

In the existing literature on social protection in Latin America, especially the literature influenced by welfare state theory, it is conventional to segment wage earners into two ‘classes’ the ‘working’ class and the ‘middle’ class. In welfare state theory, the working class includes industrial blue-collar workers, while the ‘middle class’ includes white-collar office and services workers. This distinction is central to welfare state theory, in as much as the three ideal types of welfare states—social democratic, liberal, and conservative -, are explained by alternative forms of working- and middle-class coalitions (van Kersbergen & Vis, 2014). Classes have a central role of in this approach because they enable researchers to read social protection preferences directly from occupational data.

It is questionable whether this approach is appropriate in the Latin American social protection context.⁸ As noted above, Huber and Stephens (2012) argue that class⁹ is significantly less homogeneous in the Latin America context, and that in any case reading social protection preferences from occupational class membership is far from straightforward. Taking income as opposed to occupational affiliation as the metric of class membership indicates that ‘working’ and ‘middle’ classes are insufficiently distinct. Figure 1.1. below shows the centile distribution of income in Brazil. It is hard to eyeball the middle class in the Figure.¹⁰

Fig. 1.1

Percentile distribution of income in Brazil. Data Source: WIID (UNU-WIDER, 2022)

The Figure shows the mean income of each percentile of the population in Brazil reported in WIID 2022 for 1999 and 2009 in 2017 PPP USD, estimated from household survey data.

This is not to suggest that the wage earner category is a homogeneous group. In fact, a central argument of the book contends that social protection institutions are a mechanism to stratify wage earners, to be discussed further below. The approach adopted here acknowledges contestation across occupational groups within wage earners. Intra-class differentiation is very real and have a longstanding influence on social protection institutions in the region. Intra-class differentiation and coalition building will be explored in some detail later in the book. The approach does not presume there is a homogenous set of distributional preferences within wage earners. Stark differences in distributional preferences across groups of workers are all too apparent in a region where welfare institutions are so markedly segmented. In the analysis that follows no presumption will be made that distributional preferences could be reliable read from the occupational or income position of groups of wage earners. The focus on wage earners shifts attention the focus away from inter-class differentiation and coalition and towards intra-class differentiation and coalition.

In the context of social protection in Latin America a further advantage of this approach is to include groups with low income and dependent on informal employment. They are a significant section of wage earners, around one half of workers in the region when measured as the proportion of workers not contributing to social protection schemes. Until the turn of this century, they only figured in social protection debates as an amorphous group devoid of agency and excluded from participation, lacking the capacity to mobilise and organise around social protection demands. Segura-Ubiergo claims that “the fragmentation and lack of collective identity of the poor themselves … leaves them without interests to be represented” (Segura-Ubiergo, 2007, p. 111). Haggard and Kaufman argue that the “… fate of the poor and vulnerable is therefore never in their hands alone but will depend on the self-interest of other social groups and the formation of cross-class coalitions with an interest in equity and social justice” (Haggard & Kaufman, 2008, p. 362). Huber and Stephens note a more recent shift: “…the poor were not organised and did not constitute an effective pressure group, least so under authoritarian regimes…Since democracy had been (re)installed, however, politicians began to pay more attention to the poor as potential voters” (Huber & Stephens, 2012, p. 171). Recent analyses of the politics of social protection and social policy reforms in the region emphasise the substantive role of low income workers and excluded groups (Garay, 2016; Kapiszewski et al., 2021; Silva & Rossi, 2017). This issue will be taken up in more detail below in connection with the expansion of social assistance.

In the study of social protection institutions, intra-class conflict and contestation constitutes a more reliable, flexible, and productive approach. It can accommodate a hard distinction between manual and intellectual work made by researchers studying industrialisation processes in the first half of the twentieth century. It can also accommodate the sectoral divisions among wage earners associated with fordism, and its weakness, in Latin America. It can also accommodate a blurring of this distinction associated with tertiarisation, global production networks, and job polarisation in the knowledge economy.

To repeat, the main justification for focusing on a broader category of wage earners is empirical, they are the main object of social protection institutions in Latin America.¹¹ Change and innovation in social protection institutions over time has direct effects on wage earner stratification, and on intra-class differentiation and coalition building.

1.3.2 Social Protection Institutions Are Stratification Mechanisms

In the literature on social protection institutions in Latin America, and elsewhere, it is common to find that they are justified in terms of specific social goals. For example, to co-opt and control potential challenges to elites; or to protect labour power from globalization or industrialisation-caused social risk; or as a means to ensure productivity and economic growth. In this literature, social protection institutions are normatively assessed in terms of their adequacy and/or efficiency in securing these goals. However, establishing whether social protection meets normatively defined standards is not an explanation why particular institutions are in place or are missing. There are numerous strategies that policy makers could follow to address these goals, aside from social protection. Provision of a basic income, say equivalent to median labour earnings, to all residents in a country would help achieve many of the social goals listed above, without the need for social protection.

Social protection institutions are not engines of equality or poverty reduction, although they may secure these goals, but they are primarily engines of wage earner stratification. Capitalist production requires wage earners, that is independent workers able and willing to work for a wage.¹² Production in advanced forms of capitalism require workers with specific skills and motivation. Social protection institutions help ensure that wage earners are available in the quantity and quality needed to ensure capital accumulation. Social protection institutions support the stratification of wage earners under capitalism. Alternative types of social protection institutions generate different forms of stratification, as the analysis in the book will demonstrate. A comparative examination of social protection institutions will throw light on different forms of stratification.

The view that the primary contribution of social protection is wage earner stratification is well established in the literature. The analysis and evaluation of social protection systems focuses on their effects on the wellbeing of wage earners and on wage earners’ incentives to work. Effective social protection systems support adequate living standards and maximise work incentives for distinct categories of workers. A central issue in the economics literature on social protection is precisely the ‘incentive compatibility’ features of social protection programmes and policies (Aizer et al., 2020; Barth & Greenberg, 1971; Lazear, 2018; Ravallion & Chen, 2015). The dominant research focus is on whether social protection interventions maximise labour supply and productivity incentives for categories of wage earners, skilled and unskilled workers for example. Interventions that mollify work incentives are perceived as problematic.

Research and policy perspectives on social protection underline the fact that wage earners are its object. Risk perspectives suggest that the role of social protection is to mitigate the effect of social risks on workers (prudential) (Rehm, 2011). Productivity perspectives suggest that the function of social protection is to improve the productivity of workers by facilitating adaptation to economic change (mobility) (Ippolito, 1997). Human capital accumulation perspectives focus on the role of social protection in encouraging skills acquisition and firm attachment (social investment) (Moffitt, 2002). Redistribution perspectives emphasise the need for social protection to ensure social trust and cohesion (equity) (Attanasio et al., 2009). Poverty and inclusion perspectives underline social protection effects on redressing disadvantage and exclusion (citizenship) (Barrientos, 2007; Garay, 2021). Acknowledgement that social protection is fundamentally directed at wage earners runs through all perspectives on the role of social protection.

An understanding that social protection is a mechanism for wage earner stratification is at the core of welfare state theory. For Esping Andersen and Myles, a “… quick historical glance at social reform will dispel of any notion that the welfare state was pursued for egalitarian reasons. Its foundations were typically laid by conservative reformers who, like Bismarck, sought primarily to reproduce, rather than to alter, prevailing social hierarchies. When the opportunity arose, socialists pushed for social policies that would better the conditions of workers, eliminate poverty, and equalize opportunities” (Esping-Andersen & Myles, 2012, p. 639). Some researchers claim that the evolution of welfare institutions is best explained as status preserving social mechanisms (Garcia-Fuente, 2021).

These examples demonstrate that the view that social protection is primarily a stratification device is well understood in the literature.

1.3.3 Social Protection Institutions and Dominant Forms of Stratification

The emergence and dynamics of the three main social protection institutions in place in Latin America—occupational insurance funds, individual retirement savings plans, and social assistance—reflect dominant forms of stratification. In the ‘long century’ (1900–2020) it is possible to distinguish four periods associated with significant social change with implications for social protection institutions

The first period includes the first wave of industrialisation (1880s–1930s) (Bénétrix et al., 2012; Williamson, 2011). This period is dominated by factor stratification leading to the emergence of an urban working class as an economic and political actor. Factor stratification describes the development of social groupings based on ownership of factors of production: capital, labour, land. This form of stratification precedes the emergence of large scale social protection institutions, but it is a necessary precondition. During this period early industrialising countries in the region saw the emergence of autonomous working class organizations with incipient mutual aid associations (Illanes, 2003). The antagonistic nature of state institutions precluded the development of publicly supported social protection institutions. Research has considered the extent to which factor stratification sets parameters for the development of social protection institutions that follow (Wibbels & Ahlquist, 2011).

The suspension of world trade associated with the two world wars put to an end the first wave of industrialisation in Latin America and encouraged a shift to import-substitution-industrialisation. In the more industrialised countries, the state took an active role in supporting industrialisation, described as State-led development (Astorga & Herraz-Loncán, 2020; Bértola & Ocampo, 2012). Sectoral stratification became central to the new political economy. Development strategies based on import-substitution-industrialisation were stratified by sector. State-led development was supported politically by a coalition of industrial and financial capital and the nascent urban industrial working class. The process leading to the growing participation of working class organisations in the political process has been described as a first incorporation (Collier & Collier, 1991). State support for selected economic sectors was replicated and reinforced by Bismarckian occupational insurance. This process is absent in Central American countries with a dominant agro-exporting economy, and Andean countries where industrialisation was incipient.

The exhaustion of state-led development import substitution policies in the 1980s signalled another critical realignment with implications for social protection. Export-led-growth policies rose to dominance, leading to extensive economic liberalisation. The opening of the relatively protected economies of the region involved far reaching structural adjustment, impacting more strongly on previously protected sectors, and placed considerable pressure on worker organizations and occupational insurance funds in these sectors. Skill stratification replaced sectoral stratification as the dominant form of stratification. Pension reforms and the introduction of individual retirement plans in some countries sought to remove the influence of worker organisations on social protection policies and institutions.

The commodity boom in the 2000s reinforced the dominance of skill stratification. It encapsulated a more gradual change in the domestic and global economy. Global production networks involved new forms of integration in the growing global economy. Technological change placed rising demands on the expansion of the pool of skilled workers, encapsulated in the ‘knowledge economy’ (Chorev & Ball, 2022; Gasparini et al., 2021; Iversen & Soskice, 2015). The emergence of conditional income transfers and old age transfers (social pensions) signalled a significant shift in social protection provision in the region. For the first time, a programmatic and rules-based set of social protection institutions reached low wage and informal population groups in all countries in the region (Barrientos, 2013; Cecchini & Madariaga, 2011; Villatoro & Cecchini, 2018). It embedded new political strategies for connecting these groups to social policy, described as a (second) incorporation (Collier, 2021; Roberts, 2021; Silva & Rossi, 2017).

The three dominant forms of social protection in Latin America: occupational insurance funds, individual retirement savings plans, and social assistance can be associated with critical realignments and with associated shifts in the dominant forms of stratification. The confluence of diachronic, synchronic, and analytical perspectives in the study of social protection institutions in Latin America will anchor and frame the discussion in the book. This will be examined in greater detail in Chap. 3 of the book.

Four main periods in Latin America’s social protection (Table 1.1):

Table 1.1 Social protection and dominant forms of stratification

1.3.4 Methodologically, from Correlation to Causality

Researchers on social protection in Latin America have employed a variety of methodologies. Qualitative methods have been deployed to construct case studies and comparative cross-country studies. Data reduction methods have been employed to assess similarities and dissimilarities in social protection provision and institutions across countries (Martínez Franzoni, 2008; Pribble, 2011). QCA has also been employed to research institutional configurations and their path over time (Cruz-Martínez, 2014; Segura-Ubiergo, 2007). There has been growing interest in establishing the impact of social protection institutions, especially social assistance by interrogating quasi-experimental evaluation data and household survey dataset (Molina Millán et al., 2019; Saavedra & Garcia, 2017). Attitudinal data has been used to throw light upon the political effects of social assistance programmes (Araújo, 2021). Quantitative studies using time-series cross-section and panel data have been favoured by researchers investigating the correlates of social protection development (Haggard & Kaufman, 2008; Segura-Ubiergo, 2007) and the impact of globalisation and democracy in social spending (Avelino et al., 2005; Huber et al., 2008; Kaufman & Segura-Ubiergo, 2001). These studies estimate linear structural models. Research in social protection shows healthy methodological diversity but with the growing availability of harmonised cross-country data regression methods are becoming dominant.

The analysis in the book will draw from these studies, but it will pay especial attention to causal inference. Causal inference has been at the forefront of the recent spread of impact evaluation studies of social protection interventions. The analysis in the book will seek to extend a focus on causal inference to the study of social protection institutions in the medium term. Of course, it is not possible to experiment with social protection systems, nor is it possible to control for the complexity of social change or the heterogeneity of behavioural responses. And the sample of countries is fixed at eighteen. But it is possible to pay attention to the effects of social protection institutions, identify counterfactuals and data generating processes, and account for the temporal dimension of institutions and effects. The analysis in this book will therefore make a concerted effort to identify and implement methods appropriate to capturing causal factors.

Gelman and Imbens (2013) draw a distinction between two types of enquiry. ‘Causes of effects’ inquiries seek to understand what factor or factors are responsible for phenomena. For example, what explains the expansion of fragmented occupational insurance funds in Latin America? Or, what explains the emergence of conditional income transfers in Mexico and Brazil? An alternative type of enquiry focuses on establishing the ‘effects of causes’. Do conditional income transfers reduce labour supply among recipients? Do occupational pension benefits increase income inequality? The ‘causes of effects’ type of enquiry is backward looking, it attempts to find an explanation for the presence of an effect. By contrast, the ‘effects of causes’ enquiry is forward looking, it follows from some intervention or manipulation. Gelman and Imbens suggest that ‘effects of causes’ research, if properly designed and implemented, is capable of yielding (probabilistic) causal explanations, but ‘causes of effects’ research can only at best identify non-causal factors and suggest a list of potentially valid causal factors. Its role is to clear the ground for the formulation of ‘effects of causes’ type enquiries.

This is helpful in forming expectations regarding the contribution of alternative methodologies. In the Latin American context, research into the effects of social assistance interventions have been successful in delivering reliable assessments of their impact on recipients. Their approach is in the process of being replicated in other areas of social protection. Research into the factors behind the emergence and evolution of social protection in the region has been less successful in generating definite findings but has been effective in discounting the strength of some hypotheses and in the process contesting the predictions of some theories.

Paying attention to causal inference involves, as an essential first step, a concern to spell out the theoretical framework underpinning causal relations between variables. Glynn and Quinn (2013) underline the point that “a theoretical understanding of the substantive problem, particularly regarding the nature of any temporal dependencies, is an important part of a successful estimation strategy” (p. 3). It also involves a concern to spell out causal associations. There is growing interest in the use of graphical methods to identify causal relations between variables. The structural causal model of Pearl (2009) will help to establish the quality of these relationships as a first step to establishing whether identification is feasible (Elwert, 2013; Morgan & Winship, 2015). The application of graphical methods will also help to guide research towards appropriate empirical counterparts and estimation techniques.

1.4 The Approach in This Book

1.4.1 The Need for Theory

Data on the outcomes of social protection institutions in Latin America has improved exponentially in the last two decades. Information on country programmes, their reach, and financing has never been as comprehensive and reliable as the present time. The expansion of conditional income transfers has encouraged the application of quasi-experimental techniques to identify reliably their effects. This is now in the process of being extended to other type of programmes. Harmonisation of household survey data as well as attitudinal data has greatly expanded the range of research questions that can be addressed. Data gaps remain, especially in terms of reconciling survey and administrative data for occupational insurance and social assistance, but these are no impediment to detailed research on social protection in the region. Our knowledge of the design and outcomes of social protection institutions has improved significantly.

The main challenge remains to develop a theory of social protection institutions capable of explaining why Latin America has stratified social protection institutions with generous provision for skilled workers in large firms and limited or non-existent provision for the rest. How best to explain the persistence of selective and ineffective institutions? What factors can explain the compensatory, pro-old, and compensatory bias embedded in social protection institutions? How best to explain the strength of barriers to successful reform? Providing reliable answers to these questions is urgent but requires constructing a working theory of social protection institutions in the region.

1.4.2 A Late Industrialisers Perspective

Research into the development of welfare institutions in late industrialisers¹³ has been heavily influenced by theories developed to study European welfare states. It is undeniable that there is much to be learned from the study of the politics of welfare states. It presents us with an advanced and sophisticated set of theoretical perspectives, empirical observations, and methodological tools (Castles et al., 2012). But it is important to avoid researching welfare institutional developments in late industrialisers through the prism of welfare states. Latin American countries show significant differences in their economic and social structures with early industrialisers, easily demonstrated by comparison of the size of their economies, level of development, and tax revenues and spending (Gough & Therborn, 2010). But this is not the primary reason for seeking to construct a distinct theory of social protection institutions in Latin America. European welfare states constitute a notable achievement, securing a set of social rights capable of protecting their citizens from the adverse effects of capitalism on workers and their dependents. In the view of some researchers, European welfare states made possible the accommodation of capitalism and democracy (Beramendi et al., 2015). However, taking a global perspective it is hard not to see welfare states as institutional developments restricted in space and time. To the extent that welfare states are a special case of institutional development, theories developed to study them are unlikely to work as general theories of social protection institutions. Whilst making use of the knowledge base developed for the study of welfare states, it is important that theories of social protection institutions in Latin America, and elsewhere among late industrialisers, are constructed with a strong measure of independence.

Recent developments in the study of welfare states have emphasised the presence of distinct models of welfare state institutions, as the welfare regime literature confirms (Esping-Andersen, 1990; Farragina & Seeleib-Kaiser, 2011). Research into the evolution of welfare states over time does not provide strong support for convergence in European welfare states (Powell & Barrientos, 2004). Perhaps more relevant to the argument in this book, comparative research on welfare provision in European and Latin American countries does not provide support for a global convergence in welfare institutions.¹⁴ The proposition that Latin American countries welfare institutions are on a development path to become welfare states in the European mould lacks sufficient supporting evidence.¹⁵

The challenge is to develop a theory of social protection institutions in the Latin American region that best captures the evolution, design, and outcomes of the institutions in place in these countries.

1.4.3 A Dynamic and Comparative Perspective Is Essential…

An important step forward to developing a theory of existing social protection institutions in the region is to develop an understanding of their origins, evolution, change and permanence. A dynamic perspective on these institutions is an essential building block for theoretical development.

From their origins in workers’ organizations in the first industrialisation wave at the onset of the twentieth century to the rapid expansion of conditional income transfers at the start of the twenty-first century, social protection institutions in the region show considerable change, but also resilience. In fact, the three main sets of institutions in place today emerged as a policy response to social change. State supported occupational insurance emerged alongside the first incorporation of worker organisations into mass politics in the second wave of industrialisation. They were stronger in countries which adopted explicit import-substitution-industrialisation (ISI) policies. Individual retirement savings plans replaced were introduced at the turn of the twenty-first century as part of the economic liberalisation imposed by export-led strategies. Occupational insurance funds undermined by labour market liberalisation and deindustrialisation. Whereas occupational insurance funds came alongside worker incorporation, individual retirement savings plans sought to reverse this process. The extent to which individual retirement savings plans replaced, competed with, or complemented, social insurance funds reflected the strength of incorporation reversals. Social assistance expansion in the twenty-first century originated in a second wave of incorporation. This time around, informal, and low paid workers were the main focus of incorporation, associated with the conflictive demands of the commodity boom, tertiarisation, and the knowledge economy. Accounting for these three sets of institutions is essential to developing a working theory of the structure of social protection in Latin American countries today.

Developing a theory of social protection in Latin America requires that we study current institutions, their evolution, and change.

1.4.4 Comparative Social Protection

The book takes a comparative approach as Latin America is a large region with significant variation in the size and composition of social protection across the countries within it. The comparative approach is multidisciplinary, with strong contribution from political economy, social policy, politics, and public finance.

The focus on Latin America as a region using a comparative perspective is justified by the distinctive features of their social protection institutions, their dynamics and resilience. Latin American countries provide an interesting set, sharing sufficient common ground to support comparative research while at the same time exhibiting sub-regional clusters in terms of economic development, state capacity, politics, and demography. Southern cone countries and Mexico industrialised earlier and more deeply, with the implication that occupational pension funds and individual retirement savings plans are more developed there. Central American countries, integrated early on with Northern markets via agricultural exports and foreign multinational investment (Sánchez-Ancochea & Martínez Franzoni, 2015; Segovia, 2022). Their industrialisation was weaker and came much later that in the larger economies of the region. Their agro-exporting model has proved resilient. Andean countries have experienced periods of intense economic growth, driven by global demand for the products of their extractive industry. A comparative perspective is essential if the aim is to construct a regional perspective that can accommodate country differences. The social protection institutions in place in the region show a sufficient degree of similarity to support the development of a common theoretical proposition. At the same time their differences help to identify the causal factors behind their evolution and outcomes.

1.4.5 Shortcuts: Pensions and Social Assistance

Social protection in Latin America includes a variety of programmes, policies, and institutions. They provide pensions, maternity benefits, unemployment support, compensation for workplace accidents and chronic illness, conditional income transfers, to name a few. Reference will be made to these programmes in the book, where relevant, but the primary focus will be on occupational pensions, individual retirement savings plans, and social assistance transfers. They are associated with critical realignments and societal change in employment and political participation. They are the principal gateways for access to social protection by the population. They account for the bulk of spending on social protection. Analytically, they are the most significant models of social protection in the region.

Applying Occam’s razor, a focus on these three sets of programmes will be sufficient to reach into the core of social protection in the region, including its variation across countries. Occupational pensions and individual retirement accounts can be accurately described as instruments for horizontal redistribution. They transfer resources for the same people across their life course. Their transfers are based on group membership in the former and on personal saving in the latter. Social assistance transfers, on the other hand, are an instrument for vertical redistribution. They transfer resources from the better off to disadvantaged groups facing poverty and vulnerability. Pensions and social assistance represent the two types of redistribution at the core of social protection.

1.5 The Structure of the Book

1.5.1 Framework

Chapter 2 discusses methods appropriate to the study of social protection in Latin America. It reviews and assesses the main methodological approaches employed by researchers in the past and, building on these, it aims to identify methods appropriate to developing a theoretical perspective capable of explaining the evolution and current configuration of social protection institutions in the region. Recent trends in social research methods reveal a renewed emphasis on causal inference. The application of quasi-experimental methods in the study of the outcomes of social protection interventions, initially in the context of conditional income transfers, has focused attention on the causal effects of social protection interventions. The chapter assesses the challenges associated with the application of causal inference models in the context of research on institutions relying on observational data. It argues the potential outcomes approach offers a systematic framework for incorporating attention to counterfactuals. It makes a case for the use of graphical casual models to help discriminate causal versus non-causal association between variables, thus refining researchers’ hypotheses and linking the model to potential empirical counterparts.

Chapter 3 presents the structure of the argument in the book. It introduces, develops, and applies a framework capable of explaining the emergence, structure, and dynamics of social protection institutions understood in Latin America. The central argument is that social protection institutions are an important mechanism of stratification in the capitalist economies of Latin America. The emergence of social protection institutions is explained by industrialisation. Industrialisation creates a distinct class of dependent wage earners and in doing so it sets the conditions for the emergence of social protection institutions. Once in place, the shape and dynamics of social protection institutions is best explained by their stratification effects. There are three dimensions of stratification effects that are especially relevant to understanding social protection institutions: their effects on employment, their protection effects, and their effects on the political incorporation of wage earners. The confluence of diachronic (critical realignments and the emergence of social protection institutions), synchronic (the combination of core institutions in the present), and analytical perspectives facilitates a theoretical understanding of social protection institutions in the region.

Chapter 4 is concerned with the design, implementation, and outcomes of current social protection institutions in the region. It examines in turn the three core social protection institutions: occupational pensions, individual retirement savings plans, and social assistance. Occupational pensions in Latin America are restricted to selected groups of workers, mainly skilled workers in large firms. They are also highly fragmented and dependent on public subsidies, although the level of subsidies is increasingly contested. Recent trends in labour force participation in occupational pensions largely suggest stagnation during this century. Individual retirement savings plans are grounded in individual workers saving capacity. They show three distinct institutional settings. In Colombia and Peru, individual retirement savings plans compete with occupational insurance. In Costa Rica and Uruguay, they complement occupational insurance for advantaged workers. In the rest of the countries that have adopted them, a transition process will make them the sole mandated retirement income institution. Individual retirement savings plans have not proved wildly successful in the region, and it is fair to say their institutional structures are in transition. Social assistance greatly expanded in the new century. Most countries in the region have implemented large scale old age transfers and conditional income transfer programmes. By 2015, social assistance transfers reached around one third of the population in the region.

Chapter 5 examines the stratification effects of social protection institutions on employment. The conventional approach, supported by current theories of welfare institutional development, is to assume that the structure of employment shapes social protection institutions. This makes sense when researching the origins of welfare institutions. However, once social protection institutions are in place, they exercise considerable influence on employment. The direction of causality is now from social protection to employment. The discussion in the chapter explores the stratification effects of social protection institutions on the structure of employment. The chapter discusses social protection contribution to three changes in employment in the region: tertiarisation, job polarisation and relocation, and participation in global production networks. Approaching the association between social protection and employment as a causal link brings new insights on social protection institutions in the region.

Chapter 6 studies the protection effects of social protection institutions. It involves examining the main outcomes of the core social protection institutions and of tax and transfer systems. The most significant outcome of social protection provision in Latin America is its stratification. Occupational pension schemes and individual retirement plans reach better off sections of the population while social assistance supports low-income groups dependent on informal employment. A detailed analysis of the incidence of pension benefits, from occupational pensions and individual retirement savings plans, confirmed they reach less than half the population aged 65 and over. Tracking income support for the population aged 65 and over reveals a ‘hockey stick’ shape, with most older people lacking income support or receiving at best a minimum level, and with a small proportion of pensioners receiving very generous pension benefits. Assessing the protection effects of social assistance has been greatly facilitated by the spread of quasi-experimental evaluation studies. Conditional income transfer programmes have measurable effects in improving the consumption of participant households. They reduce poverty in the short term and facilitate investment in schooling and health with implications for the productive capacity of participants in the longer term. Evaluation studies largely confirm the short term positive effects of conditional income transfers. Studies evaluating the protective effects of old age transfers find reductions in poverty and vulnerability among recipients and their households. Tax-transfer systems in the countries in the region have limited effects of poverty and inequality.

Chapter 7 offers an examination of the incorporation effects of social protection in Latin America. The analysis focuses on two intertwined elements in the analysis of the incorporation effects of social protection institutions and change: (i) the features of the ‘critical realignments’ that is the incorporation effects of the emergent institutions; and (ii) the theories that can be brought to bear to explain change. The main findings from the chapter can be summarised as follows. First, developments in social protection in the region are primarily explained by processes of incorporation and disincorporation of wage earners. Second, in Latin America the incorporation and disincorporation effects of social protection institutions confirm that intra-class (wage earners) fragmentation and aggregation is far more important than inter-class coalitions for the politics of welfare institutions. Third, identifying and studying critical realignments is essential to explaining changes in social protection institutions and the resulting processes of incorporation and disincorporation. Fourth, theoretical frameworks developed to explain the emergence of welfare states enhance our conceptual and methodological toolbox, but there are few gains in extrapolating them to the Latin American context. It makes sense to seek to develop theoretical frameworks from an understanding of the Latin American context. Fifth, the literature on the politics of social protection appears to have an in-built bias towards granting a primary role to the incentives of rulers/politicians. Focusing on the incorporation effects of social protection institutions shifts attention to wage earners, the primary actors of social protection. Research on the likelihood of broader wage earner coalitions is the gateway to more inclusive and egalitarian social protection in Latin America.

Chapter 8 examines differences and similarities in social protection institutions across countries in the region with the aim of identifying country groupings or clusters. From a comparative perspective, studying institutional differences and similarities is an important step towards theory development. The methodological approach implemented was borrowed from the literature in social policy regimes seeking to identify country clusters in the region. This literature groups and ranks countries in the region based on selected social protection indicators. A review of research on social policy regimes raised some conceptual and methodological issues. They help define an approach to grouping countries based on the causal outcomes of social protection institutions—employment, protection, and incorporation - and repeated for the periods associated with institutional development. The results from this analysis reported in the chapter largely confirm the findings in the social policy regime literature. There are three main clusters of countries: Southern cone countries plus Costa Rica, Mexico, and Panama; Andean countries including Paraguay; and Central American countries. This result is in line with the findings from the social policy regime literature and more generally the country groupings identified by early social protection research like those in Mesa-Lago (1978). Repeating the cluster analysis for different periods provides additional insights. The temporal analysis suggests a large measure of continuity in the country groupings over time, but it also shows disruption in the membership of the clusters during the neoliberal period. Cluster memberships loosen up during this period, especially for countries like Chile and Mexico. Paying attention to measures of the distribution of transfers introduces fluidity in the membership sets of Andean and Central American countries.

Chapter 9 brings together the main strands of the argumentation in the book. It will perform two main functions. It will draw out some pointers from the discussion in the book for the construction of a theory of social protection institutions in Latin America and will speculate on implications for policy.