Support for Anti-democratic Norms Is Conditional, Not Consistent

• In Event: Elite Incentives to (Not) Adhere to Democratic Norms

Abstract

Recent work debates how much public support for democracy helps prevent democratic backsliding. We argue that one of the reasons for this continued uncertainty is that typical measures of support for democracy assume that the construct is unidimensional—people either stand completely for or completely against democracy. We conduct a psychometric analysis of two different measures of support for democratic norms: one administered by Polarization Research Lab and one administered by Bright Line Watch. On both measures, we find that support for democratic norms is conditional, not consistent. That is, attitudes about democracy vary not only across individuals but also across norms. We have known for decades that measures of diffuse support for democracy may be limited as indicators of actual willingness to support democracy in practice. Our contribution is to show that it may not be possible to meaningfully combine even questions about specific norms into a single latent dimension. We further test whether individual norm questions predict substantively important political behavior (e.g., election denialism) better than either diffuse support for democracy or aggregated support for specific norms. In particular, we consider how these individual items predict willingness to accept elite behavior that violates some—but not all—of those norms.

Authors

• Gavin Ploger, University of Michigan

• Sean Westwood, Dartmouth College

• Yphtach Lelkes, University of Pennsylvania