We study how people’s ties to one another and to political life hold, loosen, and change.

Project spotlight

Whither the Progressive Revolution? The Structure of Attitudinal Change in the United States across Cohorts, 1900–2000

We compare Americans born from 1900 to 2000 on more than 100 General Social Survey questions, from civil liberties and race to abortion and gender roles. Trends across survey years understate how much cohorts differ, and change across cohorts slowed among those born after World War II before picking up again among those born from 1970 to 1990.

American attitudes are still liberalizing across cohorts. Where change has stalled or reversed, it has done so among later cohorts rather than through a backlash among older or less educated Americans, and the strongest divide has moved from race to party.

Project spotlight

Framing Effects in Informal Communication Networks and the Problem of Embeddedness

Fedor A. Dokshin and Sébastien Parker · Social Psychology Quarterly, 2026

Framing is a common recommendation for building consensus on issues such as climate change, but most of the evidence comes from survey experiments that leave out who is talking and to whom. In an agent-based model of informal discussion networks, framing moves opinion far less once people talk mostly within their own party or can push back against the other side, and under partisan homophily it can widen the gap between parties as one party’s members persuade one another.

In the model, fostering discussion across party lines, or making party identities less salient, shifts opinion at least as much as having more people use a frame, and does far more to reduce polarization. Campaigns that succeed with framing may owe it less to the frame than to reaching people beyond these networks.

Project spotlight

Included on Whose Terms? Citizen Voice in France’s Democratic Consultation

France responded to the Yellow Vest movement with the Grand Débat National, whose online platform let anyone who registered answer open questions, and the polling firm OpinionWay summarized the answers at the government’s request. I compare the categories of that official summary with the millions of answers it summarizes. About a third of the answers lie in groups of similar answers that its categories do not cover, and the official summary is closer than the average answer to a vocabulary of rules, implementation, and the allocation of resources. Where it differs from what contributors wrote, the difference points toward the answers from more advantaged départements, beyond each département’s share of the answers.

Who gets heard in a consultation depends not only on who takes part but also on how a summary prepared for those who govern differs from what contributors wrote. When a participatory or deliberative procedure ends in a summary, comparing that summary with what people said is a way to ask who gets heard after they have had their say.