Abstract

This thesis consists of essays investigating job mobility and measurement error. Job mobility, captured here as a change of employer, is a striking feature of the labour market. In empirical work on job mobility, researchers often depend on self-reported tenure data to identify job changes. There may be measurement error in these responses and consequently observations may be misclassified as job changes when truly no change has taken place and vice versa. These observations serve as a starting point for this thesis. Chapter 3 explores the level and determinants of job mobility in Ireland, using the Living in Ireland Survey, the Irish component of the European Community Household Panel. One of the findings is that the rate of voluntary (i.e. employee initiated) job mobility in Ireland trebled over the period 1995 to 2000. A decomposition technique indicates that composition changes only explain around one-fifth of the increase, while the remainder reflects changes in operation of the labour market. Chapter 4 uses Monte Carlo simulation techniques to investigate the performance of a modified probit estimator developed by Hausman, Abrevaya and Scott-Morton (1998) that controls for misclassification in a dependent variable. An analysis of the data indicates that there is the possibility of substantial measurement error which may make it difficult to capture job changes. The Hausman et al. (1998) estimator is used to formally control for measurement error in models of job change for Ireland (Chapter 5) and other European countries (Chapter 6). The main findings are that the true rates of job change are being severely undercounted in several countries and also that similar factors are important in determining job changes across countries. Finally, Chapter 7 contributes to the existing literature that examines the impact of job mobility on wage growth. It finds that by controlling for measurement error in job changes, the effect of job mobility on wage growth is larger than prevailing estimates suggest.