Mia Costa's "How Politicians Polarize"

Mia Costa is a social scientist studying political representation, political behavior, and the politics of race and gender in the United States.

She works at Dartmouth College as an Assistant Professor in the Department of Government and Program in Quantitative Social Science.

For Spring 2025, she is a visiting professor of Government at Harvard.

Costa applied the “Page 99 Test” to her new book, How Politicians Polarize: Political Representation in an Age of Negative Partisanship, and reported the following:

Page 99 of my book introduces the method of one of the tests I use in the book –a conjoint experiment– and explains how this design allows researchers to estimate the causal effect of multiple things at once, such as a representative's race, party, gender, and rhetoric. The page explains that a basic survey experiment might lead respondents to make assumptions because they are only given limited information. For example, respondents might assume that a candidate using negative partisanship is also ideologically extreme. The passage provides a technical overview of how the method works and why it is well suited to testing how people evaluate candidates based on specific features.

Would readers get a good idea of the whole book just from this page? Not really! But they would get a peek. The page reflects the book's empirical approach and interest in identifying causal effects. And it hints at, but doesn't quite spell out, the main question for this particular test, which is about how political elites view negative partisan rhetoric as an electoral strategy.

This is only one part of the larger argument and patterns shown in the book, and page 99 doesn't yet cover what I find in this particular experiment! I go on to show that candidates and politicians think that voters do not respond favorably to partisan attacks. This is important: a lot of politicians spend time talking about the other party, but they don't actually think it wins votes. So why do they do it? Turns out there are other incentives for engaging in such partisan attacks. Negative partisan rhetoric boosts a politician's national profile, goes more viral on social media, and rakes in out-of-state fundraising. It might not get more votes, but it gets more attention. This type of language by politicians has consequences too. I show that politicians who make partisan attacks are not as effective at actual lawmaking. The good news: politicians actually talk about policy more than they talk about the other party. But it's just not what we hear about.

Visit Mia Costa's website.

--Marshal Zeringue