Does partisanship promote anti-democratic impulses? Evidence from a survey experiment

ABSTRACT

Emerging research argues that liberal democracy, including American democracy, is showing evidence of erosion and is in retreat around the world. Our research provides a counterbalance to concerns for American democratic decline by exploring the willingness of Americans to grant the president more authority for partisan reasons. We draw from a survey experiment administered to 1000 Americans embedded within the 2018 Cooperative Congressional Elections Survey (CCES). This experiment evaluates the influence of co-partisanship on support for undemocratic actions to empower presidents. We find that on average, respondents oppose a series of intensifying actions, across all partisan groupings and treatment conditions. We also find that co-partisanship with the president affects Democratic responses, but not those of Republicans. Democrats are less opposed to empowering President Obama relative to President Trump. But, overall, Republicans are less opposed to undemocratic acts than Democrats across all conditions. We interpret these results as relatively good news for American democracy. Partisan polarization may drive relative differences in support for expanding presidential power, but overall support for democracy, separation of powers, and institutional checks and balances remain high.

Acknowledgments

We are grateful for support from the University of Miami Department of Political Science.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 YouGov maintains panels of individuals who volunteer to complete surveys over the Internet. Participants included in the data set were selected using YouGov's matched sample methodology entailing two steps: First, a representative target sampling frame of U.S. Citizens was created using demographic data from many sources, and second, for each member of the target sample at least one member from the pool of opt-in participants was selected for inclusion. The matching process was based on sex, age, race, years of education, interest in politics, employment status, Evangelical or born-again Christian status, marital status, partisanship, and ideology. The result is a data set comprised of participants with the same measured characteristics as the target population. Use of human subjects in research was approved by [University of Miami] Human Subjects Research Office on 09/13/2016 (Protocol #MOD00013692).