Abstract:
This study analyzes the relationship of individual risk attitudes and occupational sorting with respect to occupational earnings risk. By using the German Mikrozensus, a precise measure for earnings risk is computed as the occupation-wide standard deviation of wages. Following the procedure proposed by Bonin (2007), this earnings risk measure is used as dependent variable in cross-sectional and panel data estimations using the SOEP data of 2004 and 2006, including a measure of the individual willingness to take risks. The significant relationship in cross-sectional analyses vanishes when controlling for unobserved heterogeneity. Cross-sectional results seem to be driven by the correlation of unobserved ability and willingness to take risks, and are potentially biased by an attenuation bias due to unstable risk preferences. This study contributes to the existing literature by showing the importance of controlling for unobserved heterogeneity and instability of attitudes when examing the effects of personality traits in labor market decisions.
Abstract (Translated):
Diese Studie analysiert den Zusammenhang von individueller Risikoeinstellung und Berufswahl anhand des Lohnrisikos einzelner Berufe. Auf Basis des deutschen Mikrozensus und des Sozio-Ökonomischen Panels Deutschland wird gezeigt, inwieweit Panelregressionen helfen können, potenzielle Verzerrungen durch unbeobachtbare Heterogenität und Messfehler zu vermeiden. Die Ergebnisse unterscheiden sich stark von bisherigen Arbeiten auf Basis von Querschnittsregressionen.