Andrea Robbett
LightNews LightNews
banner
andrearobbett.bsky.social
Andrea Robbett
@andrearobbett.bsky.social
1.3K followers 940 following 15 posts
Experimental economist at Middlebury studying polarization, norms, social dilemmas. Co-author of Game Theory and Behavior and currently writing a behavioral economics textbook.
Posts Replies Media Videos
andrearobbett.bsky.social
Are you interested in spending a couple years in beautiful Middlebury, VT, teaching economics to engaged students, and hanging out in an research-active, supportive, and almost weirdly friendly department? @middecon.bsky.social is hiring a two-year VAP. Applications due Jan 20th and JOE link below.
November 13, 2025 at 2:15 PM
andrearobbett.bsky.social
💥Some news!💥 After four years of pondering, writing, and revising, I’ve sent a complete draft of my behavioral economics textbook off to MIT Press. More info about the book, including the full table of contents, is in the link below.
June 5, 2025 at 6:32 PM
andrearobbett.bsky.social
What about their punishment/reward behavior? Participants are less likely to punish/more likely to reward accurate and/or partisan answers. But when the two conflict, partisanship wins out: Wrong, congenial responses are rewarded more (punished less) than correct, uncongenial ones.
Figure shows the average financial adjustment and the likelihood of answers being punished or rewarded based on accuracy and partisan congeniality. For each dependent variable, inaccurate congenial responses are better received than accurate uncongenial ones.
December 18, 2024 at 7:27 PM
andrearobbett.bsky.social
Both treatments significantly increase the partisanship of answers and decrease their accuracy, and this effect is driven by the uncongenial questions. (Interestingly, there is a partisan difference: Democrats respond more to norm information while Republicans are influenced by punishment/rewards.)
Figure indicates that the proportion of correct answers decreases for uncongenial questions for both treatments, while there is little effect for congenial questions.
December 18, 2024 at 7:27 PM
andrearobbett.bsky.social
These gaps in normative evaluations are not purely attributable to biased beliefs about the correct answer, because respondents' assessments of their party’s norms were more partisan than the actual answers others (drawn from the same subject pool) gave when answering for a small financial reward.
Figure reports the proportion of respondents who answer a specific question correctly (when financially incentivized) versus those who identify a norm approving the correct answer.
December 18, 2024 at 7:27 PM
andrearobbett.bsky.social
We find clear partisan norms governing responses to factual questions: Both Ds and Rs view accurate -- but politically uncongenial -- responses as being socially inappropriate for members of their party and the norms differ by party for all but one question.
Image reports the social appropriateness of providing an accurate answer to 10 different true/false questions about politics used in the experiment. For all but one question, Democrats' and Republicans' responses differ significantly, with the mean rating falling on opposite sides of the origin according to the expected party valence
December 18, 2024 at 7:27 PM
andrearobbett.bsky.social
How do social norms influence partisans' factual statements about politics? Our new working paper (w/ @phminvt.bsky.social ) reports the results of an experiment conducted on election day. 🧵
Title page of paper: "Reality Bites: Partisan Beliefs as Enforced Norms" by Andrea Robbett and Peter Hans Matthews. Abstract: How do partisan social norms influence stated factual beliefs about politics? We report the results of two online experiments, conducted on the eve of the 2024 US election, in which participants evaluated factual political statements. In a motivating experiment, we first elicit social norms over reported beliefs and find that Democrats and Republicans identify divergent partisan norms over factual statements. In our main experiment, we find that these norms influence stated beliefs: Learning about which statements co-partisans believe are appropriate for the group and/or that co-partisans could punish or reward them both cause respondents' factual statements to be less accurate and more in line with partisan biases.
December 18, 2024 at 7:27 PM
andrearobbett.bsky.social
The MiddExLab Virtual Experimental Economics seminar series resumes next week! The talks are Thursdays at noon Eastern (NYC time) and all are welcome. Please sign up here to receive the Zoom links:

sites.google.com/view/middexl...

#econsky @ihaal.bsky.social @katymilkman.bsky.social
September 25, 2023 at 6:30 PM