Job matching and unemployment : applications to the UK labour market and international comparisons
This thesis studies different aspects of job matching, mismatch, and their relationship with aggregate unemployment. The first part addresses the question of the structural rise in the unemployment rate in OECD countries by looking at the link between sectoral shocks and aggregate performance in an economy with heterogeneous labour. The type of sectoral shock considered in chapter 2 is the introduction of skill-biased technologies, that increase the relative demand for skilled labour at the expenses of the less-skilled. Unless the supply of skills adjusts accordingly to the increased demand, and/or relative wages are perfectly flexible, this shock has permanent effects on the aggregate unemployment rate, as shown in a non-competitive labour market model with skilled and unskilled workers. The calibration of this model predicts that a relevant proportion of the recent rise in British unemployment can be attributed to an unbalanced evolution in the demand and the supply of skills, while in continental Europe skill imbalances do not seem responsible for serious labour market problems. Finally, the impact of skill mismatch on US unemployment was limited in magnitude and almost completely offset by counteracting forces. The second part of the work uses a job-search approach to investigate the technical characteristics of the matching process between vacancies and unemployed job-seekers. Chapter 3 reviews the empirical search literature that has estimated hiring functions, concluding that recent work has successfully established the existence of a labour market matching function, in which both vacancies and unemployed workers contribute significantly to job formation. Chapter 4 considers a plausible alternative to a random meeting technology between employers and job-seekers, based on the existence of cheap information channels that save all traders the effort of locating matching partners. When combined with a proper handling of timing in the matching technology, this set-up provides novel results on the recent performance of the British labour market. In particular, it seems that the claimed deterioration of the search effectiveness of the unemployed cannot be explained by a lack of search effort per se, but by stronger competition that the registered unemployed face by other labour market segments. Chapter 5 provides an analysis of the matching process at the micro level, using individual duration data obtained from a British sample of unemployment entrants. The determinants of re-employment probabilities are here related to a search model in which transitions into employment depend on the probability of receiving a job offer and that of accepting a job offer. The analysis shows that the hypothesis of constant returns to scale in the matching technology, embodied in most bilateral search models, is not rejected by the data. Individual re-employment probabilities respond in fact to local labour market tightness, and are unaffected by its size.