The Role of Physical Activity in the Unemployment-Health Relationship— Evidence Based on German Panel Data

Studies have documented adverse effects of unemployment on various health outcomes, yet it remains an open question whether these effects are causally related to changes in health-related behavior. We focus on physical activity as a potential mediator and whether changes in physical activity can explain how unemployment impairs health. Competing theoretical perspectives suggest that unemployment may either restrict or increase physical activity. We use thirteen waves of the German panel study "Labor Market and Social Security" (N_it = 57,124; N_i = 13,123) as well as fixed-effects models in combination with mediation analysis techniques. Based on time-distributed fixed effects models, we assess temporal dynamics. We find a negative total effect of unemployment on self-rated health (β = -0.063; p < 0.01). At the same time, unemployment increases physical activity (β = 0.173; p < 0.001) and physical activity positively affects health (β = 0.046; p < 0.001). We also find a positive indirect effect (IE= 0.008; p < 0.001) that partly counterbalances the negative unemployment effect—implying that the effect is 13 percent stronger net of physical activity. The indirect effects hold for all phases of unemployment. Physical activity suppresses part of the health deterioration associated with unemployment. The temporal effect patterns suggest that increased available time, rather than stress relief, explains the rise in physical activity.