Cross-country income mobility comparisons under panel attrition: the relevance of weighting schemes

Abstract

This article aims to present an assessment of the effects of panel attrition on income mobility comparisons for some EU countries by using the European Community Household Panel (ECHP). There are different possibilities of correcting the attrition problem by means of alternative longitudinal weighting schemes. The sensitivity of mobility estimates to these attrition correction procedures is tested in this article. Our results show that ECHP attrition is characterized by a certain degree of selectivity but only affecting some variables and countries. Different probability models corroborate the existence of a certain nonrandom attrition. The model chosen to construct the longitudinal weights to correct attrition offers up rather different results than those obtained when Eurostat's longitudinal weights are used. Although attrition does not seem to have a great effect on aggregated mobility indicators, it does have a decisive effect on decomposition exercises. Our tests reveal certain sensitivity of income mobility measures to the weighting system used.