In 1996, statutory sick pay was reduced for private sector workers in Germany. Using the empirical observation that trade union members are dismissed less often than non-members, we construct a model to predict how absence behaviour will respond to the sick pay reform. We show that union members may have stronger incentives to be absent and to react to the cut in sick pay. In the empirical investigation, we find a positive relationship between trade union membership and absence due to sickness and observe more pronounced reactions to the cut in sick pay among union members than among non-members. These findings suggest that more flexibility in the use of paid absence due to sickness constitutes a private gain from trade union membership.