Moreover, we show that 'seeing the other side', casual exposure to a politically diverse set of acquaintances, contributes to correcting the misperceptions that partly fuel animosity. We find both cross-sectional and experimental evidence in favor of this complementary mechanism

Traditionally, close relationships are thought to buffer partisan animosity by exposing citizens to cross-cutting political discussion. We find that heterogeneous discussions still take place frequently, even during a contentious election, and still reduce partisan animosity

Both out-partisan close relationships and out-partisan acquaintances are associated with lower partisan animosity. In the paper, we complement traditional measures of affective polarization with a behavioral-incentived experiment where expressing animosity costs respondents actual money.

Nuancing worries about partisan echo chambers, we first show that both close relationships and acquaintances remain politically heterogeneous, even in today's polarized political climate. Especially acquaintance networks expose US citizens to more diverse political contexts

New publication in @sociologicalsci.bsky.social with Delia Baldassarri. We show that in the US, social relationships remain politically heterogeneous, and that heterogeneous close ties and acquaintances lower partisan animosity through distinct mechanisms. A thread on the main findings:

Moreover, using difference-in-differences models, we show that an increase in network education levels is predictive of an increase in progressive immigration attitudes and feeling close to progressive parties, and a decrease in feeling close to far right parties.

This homogeneity matters for attitudes and voting: citizens with more tertiary-educated ties hold more progressive attitudes on immigration and the EU, are more likely to vote for progressive parties, and are less likely to vote for the far right

Close relationships in the Netherlands show significant homogeneity by education levels. 40-50% of citizens do not report a single close relationship with a different education level, and more than 75% reports a mostly similarly-educated social network

In the latest issue of @cpsjournal.bsky.social, my article with Jonne Kamphorst in which we propose that educationally homogeneous social networks contribute to the stabilization of political conflict on the cultural dimension. A short thread on the main findings:

Link: doi.org/10.1177/0010...

Moreover, using difference-in-differences models, we show that an increase in network education levels is predictive of an increase in progressive immigration attitudes and feeling close to progressive parties, and a decrease in feeling close to far right parties.

This homogeneity matters for attitudes and voting: citizens with more tertiary-educated ties hold more progressive attitudes on immigration and the EU, are more likely to vote for progressive parties, and are less likely to vote for the far right

Close relationships in the Netherlands show significant homogeneity by education levels. 40-50% of citizens do not report a single close relationship with a different education level, and more than 75% reports a mostly similarly-educated social network

Finally, reported experiences of perspective-getting, feeling included in the discussion, commonality and reduced misperceptions are both widely reported, and are related to increased sympathy, but not to increased willingness to have discussions. (6/7)

I find no evidence that the discussion changed opinions.

Remarkably, Conservative voters are found to soften their views just before the discussion, when asked to provide them publicly to their discussion partner. (5/7)

But: there's a limit to the love. As people's opinions were further apart, the depolarizing effect of the discussion decreased.

Indeed, I suggest in the paper, unexpected agreement may be partly driving the salutary effects of these discussions in this article and in others. (4/7)

The discussion immediately increased sympathy for the other side, as well as willingness to have close out-partisan friends and willingness to have discussions. The latter effect was still visible 2-3 weeks later. (3/7)

I partnered up with an NGO to pair UK Labour and Conservative voters for a brief chat-based discussion on immigration or redistribution.

We designed a setting where people could freely discuss and potentially disagree, refraining from providing moderation or discussion tips.👇 (2/7)

Can cross-partisan discussions reduce political polarization? Even when people disagree?

In a new publication, just out in Nature Communications Psychology, I study these questions with an experiment in the UK.

https://www.nature.com/articles/s44271-023-00051-8. (1/7)