Shows, in an OLG model with matching, that neglecting labor market
frictions may seriously bias the analysis of pension reform.