SOEPpapers on Multidisciplinary Panel Data Research 577
Manski's partial identification allows less restrictive, therefore, more credible assumptions than the assumption of random treatment assignment to solve the evaluation problem. In this article the theory of partial identification is applied to the welfare effect of the euro cash changeover. When evaluating the impact of the euro cash changeover on individual welfare, Wunder et al. (2008) face the evaluation problem. Instead of arguing for a comparability of both treatment groups used (i.e. the British and the German Population), partial identification as a more robust technique is used for evaluating the effect of the euro cash changeover. Imposing less restrictive assumptions leaves out an answer about the direction of the welfare effect.