PS: Political Science & Politics


Thermostatic Voting: Presidential Elections in Light of New Policy Data

Jørgen Bølstada1

a1 Center for Comparative and International Studies, ETH Zurich


Existing studies imply a model of “thermostatic voting”—a phenomenon characterized by negative feedback from government policy to election outcomes, suggesting that a party's success in setting policy diminishes its electoral prospects. This phenomenon could give politicians an incentive to constrain the fulfillment of public demands, which would conflict with the notion of electoral accountability, which also forms part of the theoretical framework in question. This article addresses this paradox and provides new data that expand an existing time series of American policy liberalism. Employing the new data, the article identifies thermostatic voting in American presidential elections, but in light of the analysis, certain empirical features are also identified that reduce the possible incentive to withhold promised policy changes.

Jørgen Bølstad is a post-doctoral researcher in European politics at the Center for Comparative and International Studies (CIS), ETH Zurich, Switzerland. He was a Fulbright Fellow at Harvard University, 2010–2011, and obtained his PhD from the European University Institute in Florence, Italy. He can be reached at

The author wishes to thank Mark Franklin, Carolien van Ham, Till Weber, and anonymous reviewers for valuable feedback. He is also grateful for support from the European University Institute, the Fulbright Foundation, and the Norwegian Research Council (project no. 184566).