Effects of Cross-Cutting Political Talk in Non-Political Online Spaces

The final paper in this ICA 2024 conference session is by Talia Stroud, who begins by noting that cross-cutting exposure is seen as normatively good – but exposure to cross-cutting views has also been found to potentially increase polarisation. Where such cross-cutting exposure takes place matters, then; cross-cutting exposure in inherently non-political spaces might be more productive here than it is in explicitly political spaces.

This links to intergroup contact theory (and, I assume, Ray Oldenburg’s concept of the third place), but in deeply bipolar nations such as the United States even the establishment of such non-political spaces might now be increasingly difficult. Tania’s team explored this through an experimental study that asked participants to engage in online parenting groups on Reddit, and variously introduced no political debates into such groups; introduced them in the second half of the process only; or introduced them right from the start. Participants assigned to these groups were always half Democrats, and half Republicans, and they were required to post around three times per week over several months.

Rules of each subreddit were set by moderators, and two confederates posted discussion prompts in each subreddit, drawing on posts found in actual subreddits outside the study. Partisan animosity was unaffected by these experimental treatments; however, the political subreddits were seen as more negative than the non-political subreddits. Toxicity was very rare, though, and predominantly in the political subreddits.

Any little dose of politics in these groups is making people engage, then, but still turning people off the forum. There appear to be no bias-reducing effects from cross-cutting exposure.