The paradox of redistribution in time: Social spending in 53 countries, 1967-2018

Korpi and Palme (1998) famously suggested the existence of a Paradox of Redistribution:although programs targeted to the poor may be more redistributive per unit of expendi-ture, universal programs reduce distributive conflicts, leading to bigger, more egalitarian wel-fare states. However, recent works question the existence of this trade-off. My paper addsa dynamic, long-term perspective to this literature: it analyzes the relationship between theprogressivity and the redistributive impact of social transfers in 53 rich and middle-incomecountries, using microdata from 479 household surveys harmonized by LIS. My results showthat the relationship between the redistribution obtained by social transfers and their pro-gressivity is non-monotonic and is contingent on initial policy positions: welfare states thatfocused on the poor have grown bigger and more egalitarian by moving up the income ladderto include richer constituencies, while welfare states that focus on the rich are unable to reachdown the income ladder and remain stuck at very low levels of redistribution.This reflects how social policies shape distributive conflicts: expanding upwards in theincome distribution narrows the gap between contributors and beneficiaries, easing distribu-tive conflict and allowing welfare state expansion. In contrast, expanding downwards draws aclear gap between contributors and beneficiaries, making welfare state politics zero-sum. Thisfits with evidence on the long-term evolution of universal welfare states - as they grew frommeans-tested cores to earnings-related universalism - and countries with elitist social policies,exemplified by Latin American countries, that remain captured by the middle classes andthe rich. In short, my results reinforce the idea that increases in redistribution are driven bystatus-preserving considerations - not by attempts at soaking the rich.