Do Hospital Stays Modify Locus of Control?

The assumption of stable non-cognitive skills is important in the economic literature. This paper proposes to test this assumption by investigating whether a specific non-cognitive skill, locus of control, is stable after the occurrence of a health-related event, namely a hospital stay. To do so, we use a representative and longitudinal dataset of individuals living in Germany (SOEP). Our results show that mean-level changes in locus of control are, on average, small. In addition, we report evidence that length and frequency of hospital stays have only little influence on locus of control. Young adult women (i.e., aged 18–35 years old) and old men (i.e., aged 66 years and more) experience, however, larger locus of control change. Empirical studies analyzing the impact of locus of control for these subgroups might, therefore, rely on a locus of control measurement prior to the desired outcome to limit reverse causality or simultaneity bias.