Few books are as well-matched to the moment of their publication as Ezra Klein’s “Why We’re Polarized.” President Trump has just become the third president ever to be impeached — and the first one where the votes have fallen almost perfectly along party lines. Klein’s careful book explains how different groups of Americans can see politics through such different lenses, examining how various psychological mechanisms allow committed partisans to rationalize almost anything their party does.

Klein first came to my attention during the 2009-2010 health-care debate, when his Washington Post blog “Wonkblog” was the go-to site for understanding the political and policy dynamics surrounding that legislation. This book fully displays the attributes that have made Klein’s journalism so successful.

The book is undeniably wonky, in the best sense of the word. Klein is an astute reader of political science and social psychology, disciplines he takes seriously. I should disclose that I occasionally wrote for Wonkblog several years ago. But given how much Klein has done to elevate political and social science, it’s hard to find a political scientist not in his debt.

Klein’s book starts with the psychological underpinnings of polarization, and then looks at ways that today’s media landscape and political institutions generate feedback loops that amplify it. In this view, polarization is self-reinforcing. Political elites divide over a question, and then citizens, picking up on those divisions, follow the natural grooves of human psychology by dividing themselves into increasingly meaningful groups. Those emerging divisions, in turn, heighten politicians’ incentives to accentuate their divisions. Thick with insight, the book is especially compelling on how today's media environment fosters identity-infused content.

But Klein may have incorporated certain lessons from contemporary political science too well — picking up its blind spots and inheriting my discipline’s collective overemphasis on political psychology.

Of course, Klein’s book recognizes that polarization doesn't come from psychology alone, but from the interplay of political psychology and political context. Drawing on Liliana Mason’s book, Klein argues that contemporary polarization is especially pernicious because various social identities become aligned with partisan groupings.

“Today, the parties are sharply split across racial, religious, geographic, cultural, and psychological lines,” he writes. “[S]ince these mega-identities stretch across so many aspects of our society … they are constantly being reinforced.”

But demographics still don’t fully define what party citizens will identify with. While some social divisions have come to map onto party lines more closely — like whether someone lives in a rural or urban area — others have actually faded. Knowing whether someone was Catholic, a Southerner or a union member used to tell you more about their politics than it does today. And even for some of the most salient contemporary social identities, the connection between those identities and partisanship remains far from perfect.

Here in Philadelphia, if you walk west from the University of Pennsylvania, you pass first through an affluent neighborhood that’s about 30 percent black and 50 percent white before reaching a less-affluent neighborhood that’s more than 90 percent black. In the first neighborhood, more than 60 percent of adults have a college degree, while in the second, the figure is 12 percent. Yet those significant demographic differences are essentially invisible on political maps, since both neighborhoods back Democrats nationally by almost identical margins.

If millions of blacks and whites, of religious Christians and avowed atheists, all identify as strong Democrats, as Klein details, the power of that partisan identity can’t come straightforwardly from its fusion with racial, religious or class identities. In a two-party system, and in a highly diverse country of more than 325 million people, there are too many meaningful social identities for them to fully fuse with partisanship.

Klein’s title — “Why We’re Polarized” — prompts the question: Who is this “we” that is so polarized? Just how deeply does that polarization divide people who aren’t engaged in politics more or less professionally? The book details how the explosion of media options results in a nation where only those who are deeply engaged politically participate in the ongoing partisan battles. That more limited model of polarization is a compelling alternative. It’s also backed by research, including that of James Druckman and Matthew Levendusky and Samara Klar and co-authors. Both studies show that “affective polarization,” or polarization based on citizens’ feelings toward the other party, may be more confined to political activists than we sometimes think.

What’s more, there’s extensive evidence that voters aren’t especially polarized on many issues. For instance, on the hot-button issue of immigration, 67 percent of Americans backed increased border security in 2019. Voter polarization isn’t the main reason for congressional gridlock. And as David Broockman and Chris Skovron’s research suggests, public officials may have a skewed image of public opinion, based in part on what kinds of citizens actually contact them. Klein’s book handles these issues with subtlety. But if most citizens aren’t highly polarized on the issues, that’s qualitatively different from the way political elites are polarized.

Voters in general are only one of the influences on politicians, and only during elections do they make their influence felt. And elections are a very coarse signal about what voters want, since they might support or oppose a candidate for many reasons. That’s why political scientists and observers have argued so fiercely about who supported Trump in 2016, and more importantly, why.

Klein’s general characterization of polarization as a feedback loop is surely right. But elite and highly-involved political figures are likely to have much more influence on voters than the reverse.

Activists, donors and journalists are distinctive from voters at large. They can make their voices heard by politicians consistently in real time. These groups have gotten less attention from recent political science research, partly because they’re harder to study through the surveys and experiments we now emphasize. But they may well have fostered polarization more than we’ve yet acknowledged. They’re certainly overrepresented in the news media audiences Klein writes about so knowledgeably.

There is definitely a “we” that is highly polarized on issues and divided on a series of fundamental, identity-infused questions. But that “we” may be smaller than Klein’s book sometimes suggests.