Martin Henning (p), Rikard Eriksson, Zoltán Elekes

Laszlo Czaller

Already in the theories championed by Alfred Marshall, a quintessential argument was that established local industry specializations facilitated recruitment of skilled labour in particular fields, providing local clusters with new generations of specialized workers to constantly re-invent local specializations. On the other hand, in the spirit of Jane Jacobs, one could argue that a more diverse local recruitment behaviour signals openness to new local divisions of work and innovative combinations, and local diversification. However, the ambition of local actors to push local labour market in certain directions, is not the same as realizing these ambitions. Until now, it has been very difficult to study such local ambitions to change local labour market portfolios. However, as information about local job postings across time becomes increasingly available, it allows for the investigation of how actors on local labour markets aspire to expand their human capital portfolios in relation to existing specializations, which strategies and experiments are more successful, and where. The purpose of this investigation is therefore to analyse how such “local labour market push strategies”, as revealed by the structures of local job postings, are related to already established local job portfolios and to local economic outcomes. We derive information about this in a Swedish setting and for 290 Swedish local economies, by combining information from Swedish job postings published by the Swedish public employment service´s Jobtech initiative with information from the databases of Statistics Sweden. We first map the geographical structures of job ads and study which local economies pursue more diversified versus more specialized recruitment ambitions. We make use of a job skill-relatedness matrix, computed from text analyses from job postings. Second, we relate the structures of local job ads to already existing local portfolios of jobs. This allows us to identify the local economies that have the largest push towards new skill domains, as well as those that mainly push in line with existing specializations. Third, taking advantage of our time series, we measure which local economies that are more successful with their push, and if this matters for regional economic development. Very preliminary evidence suggests that given their level of diversity, local economies that pursuit more skill-focussed skill recruitment strategies are more successful than other local economies. This is an indication of the value of concerted efforts, or local agency, that collectively build knowledge resources on labour markets.

Anne Otto (p), Johannes Koenig

Martin Henning

We investigate the geographical patterns of labour market entry (at the level of German functional labour market regions (planning regions) for doctoral graduates starting their first position after receiving their doctorate. The mobility decisions of more than 23 thousand doctoral graduates in Germany from five graduation cohorts are examined using a unique micro level data set based on doctorate recipient’s social security records. Empirical findings suggest that factors influencing the choice of a working region after graduation differ depending on whether doctoral graduates remain in academia after graduation or switch to the private sector. The results indicate that especially researchers who remain in academia, tend to move to regions where they have stronger networks. For doctoral graduates who leave academia after graduation, on the other hand, a good fit between subject-specific qualifications und the structure of regional labor market seems to be relevant. At the same time, there are a number of regional factors that favour the choice of a region for both groups, such as low unemployment rate and regional amenites. Further, there is a high attachement to the graduation region. This result is particularly relevant for regional science, as it shows that doctoral graduates as group of most highly educated workers are less mobile than often expected.

Laszlo Czaller (p)

Anne Otto

Large labor markets have long been hypothesized to facilitate matching between employees and employers. Those looking for work in large markets have more job options, while employers can choose from a wider pool of applicants. Consequently, jobseekers in large urban labour markets are more likely to find jobs that match their skills and preferences. While little evidence has been found in favor of the ‘urban insurance effect’, that is, labour market size does not seem to shorten the duration of unemployment, the effect of labor market size on matching quality has been little researched. In this paper, we develop a simple economic model that establishes a relationship between matching quality and labour market size. Then, we analyze this relationship using administrative data on all registered job seekers in Hungary between 2012 and 2017. Matching quality is measured by the degree of skill similarity between occupations before and after unemployment using survey data on the skill requirements of Hungarian. Jobseekers also declare what kind of occupation they are looking for which enables us to analyse whether skill mismatches are already present in intentions.
Since skill similarity is bounded on both sides we use fractional response models instead of linear models to assess the extent to which labour market size contributes to matching process. We find that in large urban labour markets the distance between the preferred and filled (post-unemployment) job is smaller on average, which means that in such labour markets it is easier to find a job with similar skill requirements as the preferred job. Surprisingly, this result holds regardless of the duration of unemployment. Estimation issues arising from the endogeneity of labour market size are addressed using historical instruments while fixed-effects models are estimated to remove unobserved characteristics of workers. In the models where we control for unobserved individual heterogeneity, the effect of labour market size is identified for individuals who moved and were registered at least twice in the unemployment register in the period 2012-2017. This means that the identification of the effects is done on a highly selective sub-sample of jobseekers, nevertheless the results are very similar to the baseline models estimated on the full sample. These results are also robust to re-employment definitions, alternative measures of occupational similarity, and measurement errors of the preferred job title.