No Evidence on Directional vs. Proximity Voting

The directional and proximity models offer dramatically differenttheories for how voters make decisions and fundamentally divergentviews of the supposed microfoundations on which vast bodiesof literature in theoretical rational choice and empirical politicalbehavior have been built. We demonstrate here that the empiricaltests in the large and growing body of literature on this subjectamount to theoretical debates about which statistical assumptionis right. The key statistical assumptions have not been empiricallytested and, indeed, turn out to be effectively untestable withexisting methods and data. Unfortunately, these assumptionsare also crucial since changing them leads to different conclusionsabout voter decision processes.