Unemployment benefits and parental resources: what helps the young unemployed with labour market integration?

Abstract

This article deals with the question of how different resources affect the labour market integration of the young unemployed. Previous research has often focused on the effects of unemployment compensation benefits on labour market outcomes. However, for young unemployed people additional parental resources may be even more important. The article examines the effects both of unemployment benefits and family-related resources. The main hypothesis it advances is that individuals who have access to different resources will enter more stable jobs after unemployment than those lacking such different kinds of support. Using data from a longitudinal study on youth unemployment in Germany, I find that subsequent job duration increases with the receipt of unemployment benefits but decreases with parental support. Accordingly, benefits actually improve post-unemployment job tenure, whereas parental support functions rather as a ‘safety net’ enabling young people to quit a job.

Keywords:

• job stability
• parental support
• unemployment
• unemployment benefits
• youth unemployment

Notes

1. Theories on intergenerational mobility and social reproduction of status also provide strong arguments for positive parental influence on young people's working careers by assuming different aspirations and strategies of different social classes for maintaining parental status (cf. Erikson and Goldthorpe 1992, Breen and Goldthorpe 1997). Even if aspirations do not differ between social classes, the resources are still unequally distributed (Hillmert and Jacob 2003).

2. It may even be the case that notably parents from higher social classes are unwilling to support their child in the case of unemployment or that high-status parents prefer to support re-entering education and gaining further qualifications.

3. Strictly speaking, the analyses are confronted with two sequential durations (previous unemployment duration and subsequent job tenure) that can both be censored. If unemployment duration and subsequent job tenure correlate, the results may be biased (e.g. if job tenure after censored long-term unemployment is not observed). The problems of sequential life times, dependent censoring and potential bias are discussed in more detail by Jacob (2005).

4. I only look at jobs embarked upon directly after unemployment. For further empirical analyses on leaving unemployment for different destinations using this data-set, see Dietrich and Kleinert (2005); on exits into education and training, see Jacob (2004); for other current research on the employment prospects of young unemployed see, for example, Blanchflower and Freeman (2000) and Bradley (2004).

5. I have also estimated models on subsequent employment stability (i.e. on the duration of uninterrupted employment) using the same covariates. Due to the high degree of correspondence between first jobs and employment careers owing to the short observation period, the results did not differ greatly.

6. This variable was only applied to persons who reported that they received unemployment compensation and whose formal eligibility for unemployment benefits could be deduced from their previous job histories.

7. The items used are for: (a) financial support – Did your parents give you advice in money matters (on a scale from one (very often) to five (never))? Is your main source of income money given by parents?; (b) emotional support – Did you talk with your parents about personal matters? Did you receive warmth and affection from your parents?; and (c) practical support – Did your parents help you in some practical way? Did your parents do you a favour? Did your parents advise you with regard to training, study or work?

8. The survivor function shows the probability distribution of job tenure T. It is defined as G(t) = Pr(T > t), which is the probability that the job's duration is at least t and that job termination occurs later than t. If N is the size of the population, N·G(t) is the number of individuals who have not yet left their first job after unemployment up to t.

9. I also ran a model on subsequent employment stability (i.e. on the duration of uninterrupted employment) using the same covariates. Owing to the high degree of correspondence between first jobs and employment careers in the sample, the results did not differ greatly.

10. To explore these regional differences in more detail, I also ran a model including local youth unemployment rates as a proxy for competition and the availability of jobs for young job seekers. However, the effect was very small and insignificant (results available on request from the author).