Book Summary: Why We’re Polarized by Ezra Klein

If, like me, you’ve been watching US politics recently and wondering how it got to be the way it is, you might find this summary of Why We’re Polarized by Ezra Klein an enlightening read.

Buy Why We’re Polarized at: Amazon (affiliate link)

Estimated time: 25 mins

Key Takeaways from Why We’re Polarized

• America has polarized significantly over the past few decades:
  • Political polarization has increased.
  • Swing voters who are genuinely persuadable are going extinct.
  • Polarization by itself isn’t necessarily bad. But what is concerning is the rise in negative partisanship or hatred of the other side.
• How did this happen?
  • Political polarization was so low in the past in large part because of voter suppression of African Americans. This changed after the Civil Rights Act 1964.
  • America’s demographic shifts and Obama’s presidency have made certain groups (older, white, Christian voters) feel threatened.
• Polarization leads to more polarization:
  • People polarize more, and faster, as they become more politically engaged. The most politically engaged in turn polarize the rest of the population.
  • The Internet has increased polarization because outrage is more viral and motivating than policy discussions. The primary system also incentivizes candidates to appeal to the most polarized voters.
• The US political system was not designed to deal with national-level polarization:
  • When the Founding Fathers set up the system, they assumed people identified more with their state than as an “American” or other broad identity.
  • The US system has many undemocratic elements that allow—and incentivize—a minority to frustrate the majority’s will.
  • The Republican Party has been particularly vulnerable to forces of polarization because its base is more homogenous and its geographical advantage allows it to win elections without popular majorities.
• Klein offers several half-hearted suggestions:
  • At a structural level, we could remove tools that allow minorities to hold majorities hostage (e.g. filibuster, debt ceiling), incentivize parties to broaden their bases, and lower the stakes so that even losing parties still have a voice.
  • At a personal level, we should focus more on local and state politics, and be mindful of the way that politicians and others try to trigger our identities.

Detailed Summary of Why We’re Polarized

America has polarized significantly

Political polarization has increased

Though it might seem laughable today, people in the 1950s considered the lack of polarization in US politics to be a problem. In those days, whether a politician was a Democrat or a Republican didn’t tell you much about whether he was a liberal or a conservative. Voters struggled to know who to vote for because there was so much overlap between the parties and diversity of opinion within each party.

Political polarization has grown since the 1990s. Between 1994 and 2017, the average partisan gap more than doubled. So if you’re a Democrat, the Republican Party today poses a bigger threat to your vision of a good society than the Republican Party several decades ago did. Of course, the same is true for Republicans looking at the Democratic Party, too.

Swing voters are going extinct

Persuadable swing voters are going extinct. As the parties polarized, it became harder to remain undecided. One analysis found that the number of people who were actually undecided between the two parties had plummeted from 22% in the 1980s to just 7% in 2000.

From 2004, the Republican Party started to turn its attention to mobilizing its base instead of trying to persuade undecided voters. Since voting is optional and often difficult in the US, base mobilization can be a very effective strategy. For example, after Mitt Romney lost the 2012 election, Republicans initially focused on strategies to win non-white voters (at the time 43% of Democrats were non-white but only 9% of Republicans). But they ended up choosing a quicker path to victory in 2016—mobilizing almost 7 million white people who had voted in 2008 but not in 2012.

Polarization isn’t necessarily bad

People often use the words “polarized” or “partisan” as a synonym for “extreme. However, extremism involves a value judgment, while polarization does not.

Political scientists have long debated whether America is polarizing or merely sorting:

Divisions have always existed. Even when political polarization was at its lowest in America in the mid-20th century, there were still deeply divisive issues. The difference is where the division lies. Divisions inside parties get dealt with through suppression or compromise, while divisions between parties get addressed through conflict.

In a way, polarized political parties make sense. When parties are divided on an issue, the issues can get aired and sometimes even resolved. Political parties can also act as “shortcuts”. None of us have the time or energy to fully consider every political issue, so we can just defer to a party we trust.

Negative partisanship has also increased

The problem isn’t a rise in polarization, but a rise in negative partisanship. That is partisanship driven not by positive feelings towards one party but negative feelings towards the other. A 2016 Pew Report found that over 50% of independents voted because they disliked one party more than another.

Politicians understand that feelings often matter more than thoughts. Negative partisanship motivates people to take more costly political actions like turning up in meetings, planting yard signs, volunteering and donating. The most effective way to fire up your base is to focus on how bad the other side is.

Negative partisanship also explains why so many conservative Republicans who disliked Trump nevertheless voted for him in 2016. They hated Hillary Clinton and her coalition more.

How did polarization happen?

Lack of polarization in the past was because of voter suppression

Before the Civil Rights Act 1964, the main priority for whites in the South had been to enforce white supremacy. Race was simply not “on the table” as a political issue for much of the 20th century. The southern Democratic Party (the “Dixiecrats”) had a coalition with the national Democratic Party to enforce segregation in the South. In return, the Dixiecrats gave the national Democrats the votes to control Congress.

The Dixiecrats faced very little competition in the South because of voter suppression. In 1944, in the old Confederacy states, only 5% of age-eligible African Americans were registered to vote. And when the eligible voters tried to exercise their rights, some were beaten or even killed.

Things changed after the Civil Rights Act. Many southern Democrats found they no longer agreed with the national Democratic Party’s ideology of redistribution, particularly when that included redistribution to African Americans. Some also realized that ideals of “small government” could stop the federal government interfering with states’ white supremacist policies. The parties began to sort around the issue of civil rights. Southern conservatives joined the Republican Party and northern liberals joined the Democrats.

Demographic changes have made older, white, Christian voters feel threatened

America has undergone major demographic shifts since the 1950s, particularly in race. (There have also been changes in gender and religion, but Klein skims over those.) On average, white Americans are older than other groups, with the most common age being 58. By contrast, the most common ages for Asian-Americans, African Americans and Hispanics are 29, 27, and 11, respectively. For the first time in 2013, a majority of US infants under the age of one were non-white.

Amy Chua suggests that whites could afford to be generous when they represented a clear, dominant majority. They could afford to be universalist, inclusive and enlightened without risking anything. Today, however, no group in America feels comfortably dominant. America’s demographic changes combined with the Republicans’ aging base gives them a sense of urgency.

Michael Tesler finds evidence that Obama’s presidency itself made America significantly more divided by race. That was despite the fact that, in Obama’s first term, he discussed race less than any other Democratic president since Franklin Roosevelt.

Everyone engages in identity politics

The term “identity politics” has been weaponized to describe politics practiced by members of marginalized groups and make them appear self-interested. By contrast, issues raised by dominant groups are framed as more rational topics for debate.

However, identity influences all politics. Identity is far more powerful than policy in people’s voting decisions. Research shows only a weak correlation (around 0.25) between how much a person identifies as a conservative or liberal and how conservative or liberal their views actually are. We naturally have strong group identities because humans evolved to live in groups.

Both Republicans and Democrats appeal to identities. But some identities are so pervasive they’re almost invisible. “American” and “Christian” are both identities, yet people don’t usually think of them as such. “White” is also an identity, which tends to emerge in times of threat.

Political power lagging cultural power

Perceptions of how quickly America’s racial demographics have changed are even faster than the actual shifts. A 2013 survey found that, on average, people believed 49% of America was non-white, when the actual answer was just 37%.

Part of this is because cultural power runs about a decade ahead of demographics as brands and media chase younger, urban, and more diverse consumers. In contrast, political power generally runs a decade or more behind demographics. That’s because older people vote at higher rates and tend to live in rural areas that hold disproportionate political power.

What about the economy?

Many people feel uncomfortable talking about race and prefer to explain Trump’s rise by referring to the economy. While evidence does suggest that economic anxiety and racial resentment are related, there’s a chicken-and-egg question about which one came first. Most people seem to think economic anxiety after the 2007-2009 recession drove racial resentment. Klein believes it’s the other way around.

One reason is that, before Obama’s presidency, American’s feelings about black people had little impact on their perceptions of the economy. After Obama, this changed. Voters with high levels of racial resentment became more likely to believe the economy was doing poorly.

Another reason is that the populist Left has also offered economic solutions to the financial crisis, but the populist Right seems to have been far more successful overall. Right-wing populism has also grown in other countries regardless of how their economy was doing, or how generous or stingy their social safety net was. Eric Kaufmann finds that migration crises tend to give sharp boosts to populist-right numbers but economic crises alone do not.

Polarization leads to more polarization

Stacked identities increases polarization

Our political and non-political identities reinforce each other. How someone votes can now predict their religion, race, ethnicity, gender, neighborhood, and even their grocery store. On Big Five personality tests, “openness to experience” predicts liberalism while “conscientiousness” predicts conservatism.

Even though many of these preferences don’t start out political, they can still have political impacts. For example, you might decide where to live for non-political reasons, but your decision might still increase political polarization. A 2017 Pew survey found that 65% of Republicans would rather live in a community with larger houses that were further apart, whereas 61% of Democrats preferred smaller houses within walking distance of schools and shopping. Such preferences may explain, or contribute to, the growing urban-rural divide in politics. No dense city consistently votes Republican, and few rural areas vote Democratic. The dividing line sits at around 900 people per square mile.

Shared identities can act as a bridge across partisan divisions. But identities that are stacked on top of each other increase polarization. If one conflict threatens multiple identities at the same time, compromise becomes more difficult. Ideological differences are much easier to resolve when the stakes are lower.

Political engagement increases polarization

Several studies show that when people become more informed, they become more vulnerable to motivated reasoning on politicized issues (but not unpoliticized issues).

The same applies to misperceptions of facts about the other party. A 2018 study found that as consumption of political media increased, Democrats were more likely to overestimate the percentage of Republicans who were Southern or evangelical, and Republicans were more likely to overestimate the percentage of Democrats who were gay, lesbian or bisexual.

The most politically engaged people and institutions polarize everyone else

There’s a reinforcing feedback loop at play whereby those who are the most politically engaged (e.g. political elites and institutions) become more polarized than the general public, and those elites and institutions then cause the public to polarize. The elites and institutions then polarize even further in order to appeal to a more polarized public. [Klein uses the term “polarized” here, but I feel he’s talking more about tribalness or negative partisanship.]

Media

Perhaps the best example of this is the media. During the early days of the Internet, people thought it would be great for democracy because it would make information so much more accessible. But surveys show that average levels of political engagement and awareness have not increased.

Klein offers several explanations for this:

• The attention economy was far less competitive. TV networks wouldn’t schedule news and entertainment at the same time, so most people watched some news even if they weren’t all that politically engaged. Today, news competes with literally everything else for your attention. No one has to follow politics, and it has become much easier for people to ignore the news entirely.
• The incentives today are to provoke outrage. In the past, newspapers would try to earn higher profits by monopolizing a local market. They would try to be relatively balanced in order to appeal to a broad audience. Today, media outlets’ business models are very different. Instead of trying to appeal to everyone, their incentive is just to get enough people to care deeply enough to subscribe and to share their articles. The best way to do this is to stoke outrage and engage people’s sense of identity.

The media is biased. But it’s not so much a bias consistently towards the Left or Right. It’s a bias towards outrage.

Primary elections

Before the 1970s, party elites were gatekeepers of the presidential nomination process and would block “outsiders” like Donald Trump and Bernie Sanders. Since 1972, the nominations are determined by party primaries, and only the most fervent party supporters vote in primaries (less than 30% of eligible voters participate in primaries). [See How Democracies Die for more on this.]

Small donor financing

There are broadly two types of donors in the political system:

• Individual donors give as a form of identity expression. To appeal to these donors, candidates have to get their attention and move their emotions.
• Institutional donors give as a form of investment. These donations drive the obscure laws and processes that the media tends to ignore.

The Internet has made it much easier for campaigns to ask individuals for donations. While small-donor democracy sounds great in theory, it also has a dark side. When the parties control the money, candidates are incentivized to be moderate and appeal to a broad base. But when candidates raise money themselves, the incentives change. Their incentive is to appeal to the most polarized and engaged people, since fervent supporters are more likely to donate.

[I’m not sure about this reasoning. Klein says elsewhere that the Republican Party had already begun to switch from persuading swing voters to mobilizing its base as early as 2004, which is before (or near the start of) the rise of small donors. Besides, why would we expect party officials to be more moderate than individual donors? After all, party officials are likely more politically engaged.]

The combination of weak parties and stacked identities

We’re left with a situation in which parties are weaker, in that they’ve lost control over who they end up nominating. But parties are also strong in that, once a candidate is nominated, they’re virtually assured to get the support of the party’s broader base.

If you look at the numbers and demographics behind the 2016 election, Trump’s win looked less like a weird glitch and more like a continuation of existing trends. Because party polarization was so strong, many Republicans who didn’t like Trump nevertheless felt he was the only option. In fact, the Republicans might have enjoyed a more decisive victory had they chosen a more conventional candidate. Hypothetical polling matchups consistently show that someone like Marco Rubio would’ve beaten Hillary Clinton with a convincing popular vote majority.

The US political system was not designed for parties

To understand American politics, we need to look beyond individual narratives to the broader system and the incentives within it.

The system was designed for state identities, not national identities

The American political system is built on a deep sense of place. That’s why the system privileges rural areas over urban ones. The Founding Fathers deliberately designed it to protect (geographical) minorities from the rule of majorities.

The Founding Fathers expected people to identify more with their particular city or state than with the United States. Members of Congress were meant to represent their districts, which may not line up with the national parties’ interests. The system required constant compromise to work. If a bill benefited most people but disadvantaged a particular district, giving that district some sweetener (e.g. a road or hospital) could help pass it.

When the Constitution was drawn up, the Founders’ assumption was correct. A textual analysis of books suggests that writers referred to their state identity more often than their national identity. It wasn’t until 1968 that national identity took a clear lead. Ticket-splitting also used to be common, with people voting for one party at the state level and another at the national level. One study found that between 1972 and 1980, the correlation between the Democratic share of the House vote and the Democratic share of the presidential vote was only 0.54. By 2018, this correlation had risen to 0.97.

Today, almost all Americans will list their nationality as one of their top 3 identities. Even though local politics has a bigger impact on people’s daily lives than national politics, the former gets far less attention in the media—again, because people follow politics because of identity, not because of the issues. Legislators largely vote along party lines.

Minorities have various ways to frustrate the majority

America’s political system was not designed to give full power to popular majorities but to frustrate them.

In a parliamentary system, whichever party holds a majority in the legislature also holds executive power. By contrast, in the US’s presidential system, the legislature (Congress) and executive (president) are elected separately. This can result in divided government, where two opposing parties each have some claim to democratic legitimacy. In a 1990 paper called “The Perils of Presidentialism”, Juan Linz noted that the vast majority of stable democracies were parliamentary systems, and that presidential systems tended to fail. America appeared to be an exception.

The reason this presidential system nevertheless worked in the past was because, up until 40 years ago, political competition used to be pretty weak and majorities were often lopsided. When competition is lopsided, the subordinate party can only wield influence by cultivating good relationships with the other side. When competition is close, however, the subordinate party is incentivized to frustrate the majority. That’s because people are more likely to vote for a change when government feels broken.

Why the Republican Party has polarized more

Both parties have experienced the forces of polarization but the Democratic Party has weathered them better. Even when top Republicans expressed unease with some hardball tactics, they seemed to be helpless to resist their base.

There are two reasons for this:

• Lack of diversity. The Democratic Party has many competing factions, so it has to make a lot of compromises amongst them. The Republican Party on the other hand has a very homogenous base in terms of race, ideology, and media diet, and their interests are more aligned.
• Geography. Geography does not favor the Democrats. To win elections, they have to appeal to voters well to the right of the median voter. Meanwhile, Republicans can win by appealing to voters much further to the right of the median American.

In other words, Republicans can win elections even if they use more extreme and controversial tactics that turn off the median voter. Democrats can’t. If the system were structured so that Republicans had to compete for more kinds of voters, it could certainly reform itself to do so. After all, there are moderate Republicans leading blue states who are very popular.

Possible solutions

Klein tentatively suggests some reforms, with the disclaimer that he is much more confident in his diagnosis of the problem than in how to solve them. He also notes that there isn’t an “end state” to American politics, and there is no one best way for the system to work, so he offers only suggestions for how to improve from the current situation.

Klein’s suggestions fall roughly into two categories: structural and personal solutions.

Structural solutions

Polarization alone is not a problem, but it’s incompatible with how the political system was designed to function. We should therefore reform the political system so it can function despite polarization.

Possible reforms include:

• Bombproofing. Get rid of tools that allow a minority to hold a majority hostage, like the debt ceiling and filibuster.
• Broadening. Ensure both parties are incentivized to build broad coalitions. This might include abolishing the electoral college, allowing proportional representation, and making voting easier.
• Lowering the stakes. Find ways to guarantee each party a certain degree of political power even if they lose. For example, make it easier for minority parties to bring full bills to the floor in Congress, or depoliticize the appointment of Supreme Court judges.

Personal solutions

In terms of what you can do personally, Klein suggests:

• Focusing on state and local politics. You can do much more to impact state and local politics than you can for national politics. Your local official would probably be glad to meet for a coffee. And even if your heart lies in national politics, experience with local politics will make you more effective in the long run.
• Identities. While our political identities might be polarized, we all have many other identities (e.g. parent, sister, coach), and some of these identities will be more motivating than others. Be mindful of the ways that others try to trigger your identities and intentional about how you respond. Remember that identities cannot resolve policy debates.

Other Interesting Points

• For much of American history, most newspapers were explicitly partisan and often had “Democrat” or “Republican” in the name.
• When Klein was launching Vox, he saw other news sites less as competitors and more as collaborators because if another site converted someone into a politics junkie, that person would become more likely to read Vox.
• Lots of stuff on groupish and tribal behaviours covered in books like Political Tribes and The Righteous Mind.
• Also lots of stuff on motivated reasoning and how more information, math ability, scientific literacy, and engagement with media from the other side doesn’t help people reason better when an issue is politicised. See The Scout Mindset, How Minds Change and, again, The Righteous Mind on this.

My Review of Why We’re Polarized

I listen to Ezra Klein’s podcast and generally find him to be very insightful and balanced. Despite being a liberal, he made the strongest steelman I’ve ever heard for why Senate Republicans refused to hold hearings for Merrick Garland’s nomination in 2016, effectively “stealing” a Supreme Court seat.

As an outsider to US politics, I was more interested in the stuff unique to America. Humans everywhere are groupish and tribal, but negative partisanship and political dysfunction in the US seem unusually high. So I found the parts about the history of voter suppression, the downsides of a presidential system, and how the US system was set up based on state identities to be more engaging. I was also shocked to learn about the extent of Russia’s interference in social media, which only got a passing mention. This made me wonder whether part of the polarization in the US is simply because there’s been more foreign interference in the US than in other democracies.

Why We’re Polarized is very well-researched, but I found Klein’s writing quite messy. There are many overly long quotes from other books, as well as names of relatively unimportant politicians that felt distracting (at least for me). Klein’s message also gets a bit muddled in parts. Early on, he clearly explains the difference between polarization on the one hand and extremism or negative partisanship on the other, and argues that polarization itself is not bad. But in the second half of the book, there were several times where he seems to use “polarization” in the negative sense. Overall, a bit more editing could have helped present a clearer and more streamlined message.

Some reviews of Why We’re Polarized have criticized the book for its relatively weak solutions. I think this is unfair. It’s common for books diagnosing enormous social problems to conclude with some solutions so that readers can finish on a note of optimism instead of despair (e.g. Why We Fight, Evicted and Destined for War). I get the sense that Klein’s publisher strongly encouraged him to offer some solutions as he admits that even writing his suggestions made him “queasy”. I can respect that. We can’t expect any easy fixes to the question of “how to fix American politics” and it would be disingenuous to pretend otherwise.

Let me know what you think of my summary of Why We’re Polarized in the comments below!

Buy Why We’re Polarized at: Amazon <– This is an affiliate link, which means I may earn a small commission if you make a purchase through it. Thanks for supporting the site! 🙂

If you enjoyed this summary of Why We’re Polarized, you may also like:

• Book Summary: How Democracies Die by Steven Levitsky and Daniel Ziblatt
• Book Summary: Thinking in Systems by Donella Meadows
• Book Summary: The Righteous Mind by Jonathan Haidt