The influence of previous labour market experiences on subsequent job tenure

• Autores: Carlos García Serrano, José María Arranz Muñoz
• Localización: Papeles de trabajo del Instituto de Estudios Fiscales. Serie economía, ISSN 1578-0252, Nº 17, 2003, págs. 1-37
• Idioma: inglés
• Enlaces
  • Texto completo (pdf)
• Resumen

  • The aim of this investigation is to analyse the influence of individual's previous labour market experiences on the duration of subsequent job matches in the Spanish labour market. In particular, we try to provide answers to four different issues. First, is there evidence that longer time spent in subsequent employment relationships is the result of longer previous unemployment spells (and longer search)? Second, how does the reason for previous job termination (end of contract or layoffs) affect the exit rate from jobs accepted by the unemployed? In other words, do the exit rates (or the duration) of subsequent employment relationships differ among the unemployed due to the influence of the way prior job terminated? Third, does the exhaustion of previous unemployment benefits influence the hazard rate from a future job match? And fourth, are differences in exit rates associated with different individual and job characteristics? This study draws on a sample of workers extracted from the Spanish administrative dataset HSIPRE for the period 1987-1997. Data come from longitudinal linkage of records from monthly payroll computer files for all registered unemployed workers who receive all types of unemployment benefits from the Spanish public agency (INEM, Instituto Nacional de Empleo) in charge of the administration of the payment of unemployment benefits. This dataset contains information not only on the duration of unemployment spells but also retrospective information on previous employment durations. It means that we are able to trace back and to date job separations and non-employment exits to address the effects of previous non-employment spells on subsequent job tenure. Moreover, information on the cause of previous job loss is available which allows analysing the effect of different sort of job separations on subsequent job tenure. We find evidence on the existence of a scarring effect: the duration of past non-employment damages the duration of future job matches. Stated differently, the longer the time spent in non-employment since previous job separation, the shorter the duration of subsequent re-employment relationships. We have also obtained that the reason for termination of prior employment relationships is relevant: workers whose previous job match terminated due to the ending of a temporary contract are very likely to come back to employment under another temporary job and have a higher probability of job termination (their new employment engagements last less time). The exhaustion of unemployment benefits also seems to exert a scarring effect on job duration: unemployed workers who exhausted their previous unemployment benefits find jobs whose duration is shorter in comparison with workers who exited the unemployment state before the time of exhaustion. These effects are also magnified for some categories of workers, as those aged over 45. These results point out that there are specific groups of unemployed workers for which the scarring effect of previous non-employment in terms of increasing the probability of re-entering unemployment is most important: those who come from temporary employment, those who wait until the exhaustion of unemployment benefits to accept a job, and those with longer non-employment spells. Previous studies for other countries have detected this scarring effect. These investigations point out that the cause of the scarring effect may be that past unemployment experiences may change preferences, prices and/or constraints that help determine current unemployment; that firms may hire workers taking into account their previous unemployment history; or that the unemployed may change their reservation wage as time in unemployment passes by and then accept short-term job matches or employment engagements characterised by having high destruction rates. In any case, the existence of scarring suggests that policies aimed at preventing unemployment would have a long-term impact on aggregate unemployment. As scarring seems to be particularly important in the case of the long-term unemployed and older workers, programmes aimed at the prevention of long-term unemployment and targeted to particular groups of workers should be in the public agenda. At the same time, active policies towards job finding would be efficient since they can reduce individual unemployment durations and speed up the return to employment. Moreover, appropriate on-the-job training and more stable jobs would avoid depreciation in acquired skills and undesirable effects linked to the employmentunemployment chain. Supply-side policies that make individuals more employable and increase work incentives should go hand in hand with demandside policies to generate more (stable) employment.