Mixed signals: to what extent does wage scarring vary with the characteristics of the local labour market? -preliminary draft, do not quote-

Abstract

Continuous work-life histories are constructed using matched BHPS-LFS data, in order to test for regional variation in the impact of unemployment experience on future wage growth. The main hypothesis under test is whether unemployment spells experienced in high unemployment regions are seem by future employers as more a characteristic of the region than a negative productivity signal. Consistent with previous studies, empirical results highlight significant and persistent average wage penalties due to interruption that depends on previous labour market status. Strong regional differences are found in the impact of redundancy on wage growth. This is contingent on the labour market tightness and urbanity of the regional in which this unemployment was experienced. However, no evidence is found supporting the main hypothesis in the UK. It is likely that human capital explanations still play a substantial role, given that average unemployment durations are likely to be lower in tight labour markets with better re-employment prospects.