Abstract

In a classic study, Korpi and Palme argue that large, universal welfare systems have more of a redistributive impact than systems which target resources primarily to the needy. They call this phenomenon the 'paradox of redistribution'. This article re-examines the validity of the 'paradox of redistribution' using two different methods. Particularly, we compare 'classic' OLS multiple regression with the 'newer' fuzzy-set analysis. Due to iets focus on causal heterogeneity, fuzzy-set analysis is said to allow a closer fit between theory and data. We argue here that this is not necessarily the case. We do acknowledge, however, that in fuzzy-set analyses, causal heterogeneity comes to the fore as a prime characteristic of the analysis, whereas other methods mostly require several and complex models designed by the researcher to come to the same conclusion. This suyggests that fuzzy-set analysis, more than regression analysis, is suitable as an explorative research method.

How to Cite:

Rottiers, S., Marx, I. & Van den Bosch, K., (2006) “De oorzaken van een gelijke inkomensverdeling: een vergelijking van een multiple OLS regressie en een fuzzy-set analyse”, Tijdschrift voor Sociologie 27(4), 376–404. doi: https://doi.org/10.21825/sociologos.86667