Political Determinants of Universal Social Policy in the Global South, 1990–2020

• In Event: Determinants and Consequences of Social Policies

Abstract

Over the last two decades there has been a large expansion of public social spending, particularly as low- and middle-income countries introduced broader and more generous social programs (Dorlach, 2020, E. Yörük et al, 2022). What are the political determinants that have enabled certain countries from the Global South to engage in more universalist social policy? In this paper I investigate 3 political determinants of the universalist provision of social policy: social mobilization, party institutionalization, and electoral competition, controlling for economic factors and features of institutional design. I rely on a panel data analysis of 109 countries in the Global South and 3,100 observations (country-years) ranging from 1990 to 2020. Data comes from the most recent Varieties of Democracy 2022 database and the 2020 Database of Political Institutions.
Panel data analysis offers the unique opportunity to prevent endogeneity by controlling the part of the unobserved error that is fixed at the unit level and invariant in time. In consequence, I firstly estimate a fixed effects model at the country level. To take advantage of the opportunity that the time dimension inherent in panel data offers to deal with reverse causality, I estimate a second model which controls for fixed effects but lags several of the variables of interest.
Despite its advantages, cross-sectional time-series also features challenges. Particularly, it can produce temporally autoregressive and cross-sectionally correlated error terms and biased and parameter estimates. To address this, I also estimate a Prais-Winsten rho transformation model. Lastly, to approach the joint effect of a dynamic panel data model with a lagged dependent variable, as well as the control at the unit level, the final model relies on the Arellano–Bond strategy of using all available lags as instruments. I find that there is consistent support for the hypothesis that electoral competition is causally related to increasing universalism in social policy, regardless of the orientation of the ruling party. There is mixed support for the positive effect of party institutionalization over universalism, and there is support for a relationship of mass mobilization and the outcome variable as well which, however, goes in the opposite direction of the one predicted by the literature. These results are robust to several specifications.
This paper addresses three important gaps in the literature about recent social policy provision in the Global South in general, and about expansion and the turn towards universalism. First, as a critical systematic review of 3,263 pieces of research of social policy expansion in the Global South reveal (E. Yörük et. al, 2022), the literature is largely descriptive and impact-oriented, whereas analytical studies on the determinants of social assistance programs remain under-examined: only 8.2% this sample deals with them. Second, the analysis of the political determinants of social policy expansion remains less examined than structural (economic) institutional (state capacity and revenue collection, veto points in the political system) and ideational approaches. This paper contributes to cover these gaps.
Finally, this analysis provides quantitative testing for several hypotheses that emerged in qualitative historical comparative research, such as the effect of the institutionalization of the party system (Pribble 2013) or electoral competitiveness and social mobilization (Garay 2016). Although the seminal analysis of the effect of the ideological leaning of the government on social policy expansion of Huber and Stephens (2012) was developed through quantitative evaluation, their analysis relied on cross-sectional linear regression, which has restricted purchase to evaluate causal links. Recent testing of these three hypotheses has been pursued mostly through qualitative techniques and case studies. The link for government ideology and social policy expansion has been tested trough case study analysis for Argentina, Botswana, Chile, Colombia, Mauritius, and South Korea (Niedzwiecki and Pribble 2017; Ulriksen 2012; Yang 2013; Holland & Schneider 2017). Testing of the social mobilization hypothesis has been performed trough case study analysis for Bolivia, Brazil, India, Thailand (Anria & Niedzwiecki 2016, Harris 2017, Gibson 2019, Agarwala 2013, Harris 2017). Finally, quantitative tests for the link of party institutionalization and social policy expansion are scarce, and have been done by Rasmussen and Knutsen (2021), who rely on country and year-fixed effects models, but only control for electoral characteristics. This paper thus addresses these limitations and offers a rigorous contribution to the ongoing debate on the political determinants of social policy expansion in the Global South, a debate relevant to the well-being of citizens of the regions most characterized by poverty and inequality in the world.

Author

• Oswaldo A Mena, Graduate Center at the City University of New York