Earnings responses to social security contributions

- Earnings caps for social security contributions generate kinks in the budget set.
- An increase of the earnings cap of health insurance is used to estimate incidence.
- The bunching approach is used to estimate employment responses.
- I find equal burden sharing of social security contributions as well as negligible employment responses.

This paper utilises the discontinuities induced by earnings caps for social security contributions (SSC) in Germany to analyse the effect of SSC on gross labour earnings. Economic incidence is identified by exploiting an increase of a regional earnings cap of health and long-term care insurance as a natural experiment. Based on administrative data, difference-in-differences models are estimated. I find the burden of SSC is shared equally between employers and employees. An auxiliary analysis studies employment responses to SSC at the intensive margin by a modified bunching approach that is applied to the earnings caps of health and long-term care insurance. Finding employment responses to be negligible supports crucial identifying assumptions of the main analysis. Both results are robust and consistent with a standard partial-equilibrium labour market model.