جستجو، استخدام و شرایط بازار کار

This paper attempts to investigate how employers' recruitment strategies (in terms of the recruitment method used and the applicants' characteristics) change in response to different conditions on the relevant regional labour market. Our empirical results show that the hiring of unemployed candidates and the use of the public employment service are events more likely to happen in a slack regional labour market. Moreover, the use of advertisements and the hiring of already-employed job seekers are more likely to occur in the presence of excess demand on the relevant regional labour market. This supports the view that shifts in recruitment strategies may be driven by cyclical fluctuations in expected variations (in both the size and composition) of the pool of potential applicants.

The present paper aims to investigate employers' hiring behaviour over the business cycle. Job search and hiring are in fact continuous activities in dynamic economies (Holt and David, 1966). However, search and hiring intensity vary in response to different macroeconomic conditions (for a supply side perspective, see Osberg, 1993). A common result of studies on gross job creation and destruction Leonard, 1987, Davis and Haltiwanger, 1992 and Cramer and Koller, 1988 is that both take place simultaneously at any time of the business cycle, i.e. in shrinking as well as in expanding industries. Furthermore, activities on one side of the market have external effects on the costs of the other side (thin and thick markets effects, Diamond, 1982). Along these lines, Pissarides (1994) developed a model of on-the-job search and two categories of jobs: “good” (more expensive to set up, with high productivity and high wages), and “bad” (less expensive to set up, low productivity and low wages). In this model, employers have a rather passive search behaviour: they do not signal whether they are posting a bad or good job, and limit themselves to offering the vacant job to the first applicant who is ready to accept it. Unemployed job seekers accept both good and bad jobs, while employed job seekers accept good jobs only. On-the-job search increases employers' expected benefits from creating good jobs. Therefore, during periods of expansion, when more employed job seekers become available in the labour market, it should be possible to observe the creation of more vacancies and of relatively more good jobs. This model predicts that a larger share of already-employed workers are hired during booms when a relatively large proportion of good jobs are created. Conversely, in a recession, hiring from the pool of unemployed would become more important, and consequently, fewer vacancies with a relatively high proportion of bad jobs are observed. The predictions from Pissarides' model are consistent with observed gross worker flows in the labour market. Burda and Wyplosz (1994) show, for different countries, that the outflow from unemployment becomes more important during recessions, but it becomes much less important during periods of economic recovery. In contrast, the importance of job-to-job transitions is less evident during recessions, but they become much more important during periods of economic expansion (see also Burgess, 1993). Micro-econometric studies on employers' recruitment behaviour, however, suggest that, unlike what is assumed in the model, employers' search behaviour is in fact not passive. For example, Barron et al. (1985) show that employers do search for applicants actively and that they spend a considerable amount of time screening applicants. Barron et al. (1997a) show that employers tend to screen more applicants for more important positions (with high educational requirements and involving training). Barron et al. (1997b) endogenize the information about the quality of the match. The authors found that both applicants and vacancy characteristics strongly influence firms' search both at the intensive (signal offered by the hiring standards) and at the extensive (number of applicants screened before a job offer) margin. In general, employers can activate different recruitment methods to reach certain groups of job seekers. The public employment service, for example, best reaches unemployed workers, whereas advertisements appear to best reach employed workers (see Gorter et al., 1996, Van Ours, 1994 and Lindeboom et al., 1994). The evidence from all these studies is built on single cross-sections of (filled) vacancies, and therefore changes in recruitment strategies over the business cycle cannot be captured. In their recent article, Burdett and Cunningham (1998) advocate that the analysis of employers' search would be greatly improved if the market conditions the firm faced at the time of search could be taken into account. In this respect, Russo et al. (2000) show that employers' search at the extensive margin changes in response to different conditions in the labour market. When faced with a tight labour market, employers tend to switch to the use of the recruitment methods that trigger the highest number of applications in order to make up for the scarcity of job seekers.2 This paper attempts to add to this literature by analysing changes in recruitment behaviour at the individual (firm) level at different points of the business cycle. By doing so, we attempt to uncover the role of employers' preferences for hiring among different categories of applicants (with particular attention to the unemployed) The structure of the paper is as follows: Section 2 discusses the effect of different labour market conditions on recruitment and postulates how employers will adapt. Section 3 contains a description of the data set, and Section 4 introduces the statistical model (to test the postulated adaptive behaviour). Section 5 presents the results of the empirical analysis. Finally, Section 6 concludes.

6. Conclusions The aim of the present paper is to analyse empirically how employers adjust their recruitment strategies (in terms of recruitment method used and type of applicant sought) in response to changes in the labour market conditions. In particular, we find that the use of the public employment service and the hiring of unemployed candidates are more likely to happen when there is excess supply in the labour market. The use of advertisements and the hiring of already-employed job seekers are more likely to occur in presence of excess demand. The empirical analysis also confirms the importance of hiring standards. In particular, when higher educational standards are required, employers are more (less) likely to: (i) use advertisements (the public employment service); and (ii) hire employed (unemployed) job seekers. Moreover, we found a significantly positive association between, on the one hand, the use of temporary help agencies and the public employment service instead of advertisements and, on the other hand, the hiring of unemployed instead of employed applicants. Our results suggest that employers do change their recruitment strategies in response to cyclical fluctuations in the expected size (and composition) of the pool of potential applicants.