Success of a tax amnesty: At the polls, for the fisc?

This paper provides a model for analyzing an individual's voting decision on tax amnesty and the change of his reporting behavior given amnesty is granted/rejected. Success at the polls is predicted to increase with higher penalties meted out to detected evaders while success in terms of reporting would be jeopardized. These and other implications of the model are empirically tested for two Swiss tax amnesty proposals, one rejected, the other approved by a majority of citizens/taxpayers.