Item Nonresponse and Imputation of Annual Labor Income in Panel Surveys from a Cross-National Perspective (chapter 19)

Abstract

This chapter outlines the basic characteristics of the three panel surveys focusing on the incidence of item nonresponse (INR) with respect to labor income. It demonstrates the selectivity entailed by INR and investigates the time dependence of nonresponse behavior. The chapter describes the imputation methods applied in the three surveys. Based on rather typical empirical research questions using labor income, it demonstrates the effect of imputation on earnings inequality and mobility as well as on wage regressions. The chapter concludes with recommendations on how to deal with INR and imputation in cross-national comparative research. It provides an exhaustive description of imputation procedures used in Socio-Economic Panel (SOEP), British Household Panel Study (BHPS), and Household, Income and Labour Dynamics in Australia (HILDA).