Job Matching on non-separated Occupational Labour Markets

This work refers to analyses of matching processes on occupational labour markets in Germany. Up to now, all studies in this field are based on the crucial assumption of separate occupational labour markets. I outlined some theoretical considerations that occupational markets are probably not completely separated. By using information about similarities of occupational groups I constructed an "occupational topology" and finally tested my hypothesis of non-separated occupational labour markets with Spatial Econometric Estimators, particularly with a restricted version of a Spatial Durbin Model that includes “spatial” lags for regressors. The results show considerable dependencies between similar occupational groups in the matching process. Particularly, the results indicate occupational specific set-up and set-down processes in similar occupational groups. This has important implications for estimating the matching efficiencies of unemployed and vacancies, because the matching process is not only determined by the unemployed and vacancies in the same occupational group but also by those in other occupational groups.