Does In-Party Affect Increase Factual Belief Polarization? A Rigorous Test of an Old Assumption

Recent studies argue that partisans have become increasingly polarized in their perceptions of objective facts, a trend commonly attributed to rising levels of in-party affect. The notion that partisans with stronger emotional attachments to their party are more likely to view even objective facts from a party lens goes back to the beginnings of modern political science. However, there is surprisingly little empirical evidence to support this claim. We test if higher levels of in-party affect increase the partisan gap in perceptions of objective facts using experimental and probability-based panel data from Sweden. We manipulate in-party affect by asking partisans to describe the things they like the most about their party. Although our intervention is successful, we do not find that the correlation between (block) partisanship and factual beliefs is stronger in the treatment group than it is in the control group. In other words, factual beliefs polarization is just as high in the group induced to feel higher levels of in-party attachment as it is in the group assigned to the placebo task. We replicate the null findings by utilizing a three-wave panel dataset fielded during a time when in-party affect is most likely to occur naturally, i.e. during an electoral campaign. Like previous research, we find a strong correlation between partisanship and factual beliefs, yet the strength of this relationship is not sensitive to changes in partisan attachment. These results cast doubt on the relevance of partisan identity and affect for explaining the partisan gap in perceptions of objective facts.