One size does not fit all: Quantile regression estimates of cross-country risk of poverty in Europe

Using a macro panel of 31 European countries, this study shows that the application of quantile regression (QR)procedures to the estimation of poverty risk reveals that poverty determinants and cross-country differences in poverty levels are more reliable than those emerging from conditional mean estimations. The extent and significance of interquartile differences of estimated coefficients suggest that economic growth, income distribution, public expenditure, investment, education, and the labor's share of income – aproxy for functional income distribution or social antagonism – have strong but differentiated effects on poverty reduction. Low-institutional quality exemplified by the high public-sector-corruption perception has a significant concomitant adverse effect on poverty and interacts with economic cofactors in determining interquartile differences of estimated coefficients. Results show that the implementation of a common European Union policy against poverty should consider cross-country interquartile differences and avoid a one-size-fits-all uniform philosophy