Does Relationship-Specific Investment Depend on Asset Ownership? Evidence from a Natural Experiment in the Housing

Real world contracts are simple- but are they incomplete? Are con-tracts restricted by transactions costs to rely on residual control rights al-though a complex mechanism would implement a more efficient allocation? Only if the answer is yes, the incomplete contracting approach is justified. I study empirically the classic application of the incomplete contracting approach: the hold-up problem. I compare the frequency of relationship specific investments, such as kitchen or bath renovations, in owner occu-pied and rental housing. Using the variation in home ownership induced by natural experiment of the German reunification, I show that the allocation of ownership matters: Relationship specific investments are significantly less frequent in rental housing. Although relatively straight forward con-tracts could implement the first best and could render apartment ownership irrelevant, they are not used, and a large equilibrium inefficiency persists