Training, wages and employment security: an empirical analysis on European data

Abstract

Data from the European Community Household Panel (ECHP) are used to assess the effects of employee training on the average wage and employment security of different labour market groups in EU countries. Significant training wage premia are found only in the case of young or highly educated employees. By contrast training appears to have a strong impact on employment security, measured through subjective measures, in the case of both older and low-educated workers. To reconcile this apparent contradiction, one needs to take into account that, as standard in the literature, wage premia are estimated on a truncated sample including only employed workers. Due to downward wage rigidity, those workers who are unable to maintain their productivity are more frequently laid-off – rather than experiencing a wage fall and be retained in employment – and thereby excluded from the sample.

Acknowledgements

This paper draws from unpublished background work made by the author for the 2004 OECD Employment Outlook. However, the views expressed here are those of the author and cannot be attributed to the OECD or its member countries. The Author thanks Sebastien Martin for excellent research assistance as well as Mark Taylor, the Editor, Francis Kramarz and participants to the 2004 NERO Meeting on Labour Market Issues in Paris for their comments. All errors are the Author's.

Notes

¹ One notable exception is Groot and Maassen van den Brink (2000), who study the impact of training on the employability for other tasks or jobs within the same firm.

² As explained above, match-specific effects on wage premia to training taken with current employers are eliminated by subtracting job-match-specific means from the stock of training taken with the current employer and using this difference as an instrument. A sensitivity analysis (not presented here) was undertaken by estimating wage equations with job-match fixed effects, and revealed that the two procedures give extremely close results as regards to training taken with the current employer.

³ Estimating separate equations by country and labour market group would have implied very small samples in many cases.

⁴ These findings are consistent with previous studies that typically find that the training premium increases in the aftermath of a job change (see Loewenstein and Spletzer, 1998, Parent, 1999, and Sicilian and Grossberg, 2001, for the United States, Fougère et al., 2001, for France, Blundell et al., 1999, and Booth and Bryan, 2005, for the United Kingdom, and Gerfin, 2004, for Switzerland). These papers usually interpret this fact as evidence of employers’ market power.