Paradoxes Lost and Found: The Dimensions of Social Welfare Transfers, Relative Poverty and Redistribution Preferences

Paradoxes Lost and Found: The Dimensions of Social Welfare Transfers, Relative Poverty and Redistribution Preferences

Korpi and Palme’s (1998) classic “The Paradox of Redistribution” contends that universalism reduces poverty much more than targeting because of the politics of welfare states. Though there have been fundamental changes to social policy, politics, and inequalities since the mid-1980s period Korpi and Palme analyzed, very few have reinvestigated their study. With data on the distribution of welfare transfers received, we develop measures of extensity (transfers as a percent of household income), low-income targeting (the low-income concentration of transfers) and universalism (the homogeneity of transfers across the population). We examine the relationships between these country-level dimensions and individual-level poverty and redistribution preferences. We also systematically compare the results within rich democracies and a much broader sample of developed and developing countries. While some results are consistent with Korpi and Palme, there are also key differences. Poverty is significantly negatively associated with extensity and universalism. Low-income targeting is not robustly associated with poverty, and is even occasionally negatively signed. Redistribution preferences are only significantly negatively associated with low-income targeting. We also show that while universalism is strongly associated with extensity, low-income targeting is also surprisingly positively associated with extensity and universalism in the broader sample. Therefore, we revise the paradox of redistribution into two new paradoxes. First, there is a clear mismatch between what reduces poverty and what matters to redistribution preferences. Second, in developing countries, the dimensions that best reduce poverty correlate with the one dimension that undermines support for redistribution. We conclude by discussing the implications for research on inequality, politics and social policy.