Comment deleted
Expand full comment

Personally I have viewed the largest cause of increasing American political polarization as a direct result of the algorithmic changes that our society has undergone, primarily (but not only!) with respect to the Internet.

The short version of this is that the Internet proliferates content that is emotionally outraging at significantly higher virality factors than non-polarizing content. You can pick who you want to blame for this, but the truth is that it is not a single person, party, website, or company, but rather the system as a whole and the incentives that construct it.

Facebook profits off of polarization in the same way Twitter, Youtube, CNN, Fox, and, well, everyone else does. Even if these entities decide "Profit is nice, but we would rather have less polarization", they will *still* have a difficult time reducing it, because they are going against the incentive structures of the entire system: content that outrages is still significantly more viral and memetic, and will thus spread more rapidly and become more commonplace, in the exact same was covid and its various strains will continue to spread even as you attempt half-baked countermeasures.

Similar to the author of this novel, however, it's difficult to come up with a solution that is actually feasible to implement. I'm hoping that society, at many levels, will start to collectively realize what is occurring and (very slowly, but eventually) take steps to remedy this with improved algorithms that do a better job at suggesting constructive rather than destructive content.

And of course I should add, that what I have mentioned here is still only one factor. There's many factors at play given how complex a society of our size is, but I predominantly think this is currently the leading factor, and that it has more causal factor than many other contributing factors.

Expand full comment

I think the "negative polarization" bit works the same mechanics as that old unhappy customer adage: "your average unhappy customer tells 10 people about their experience, whereas the happy customer tells only three" (or whatever the numbers were...)

Expand full comment

Did Klein discuss the hypothesis that, well, maybe the US is just too big and too diverse to continue to exist as a single entity in the absence of a totalitarian regime?

Expand full comment

> And he suggests granting statehood to Puerto Rico and DC, because if we guarantee that the Democrats always win, then the Republicans will have to change their strategy

I think people overestimate the effect that adding guaranteed-Democrat Senate seats would be. "Tentin Quarantino" on Twitter (https://twitter.com/agraybee/status/1276634862548647936) said in June:

> But if you were to add 4 Democratic senators tomorrow, you'd have 51 Democrats, still two short of the 53 needed for a majority in a senate with 104 people.

David Shor (https://twitter.com/davidshor/status/1277781856721743880) replied:

> I lot of the thinking here is driven by the inability to think through the derivative of the (50+x)/(100+x) function

Expand full comment

how quaint and how surprising it is to hear somebody discussing polarization as an issue between Democrats and Republicans.The polarization is between those who know we are ruled by a corrupt oligarchy intent on bankrupting a majority of the country, and those who like to pretend we live in a democracy.

Expand full comment

> Klein links this polarization to "vetocracy", the idea that it's impossible to do anything nowadays because somebody will prevent you. Sometimes this is a literal presidential veto. Other times, it can be something as stupid and venal as the party out of power using filibusters and every other dirty trick to make sure nothing improves, because if something did improve the party in power would get the credit

This seems a little scrambled to me. More polarization actually could make things move! "Nothing" gets done now because in a closely-divided Senate, there are a few holdouts on both sides - Collins, Manchin, and the like. If they gave into polarization, the majority would be able to ram through its agenda.

Expand full comment

I'm struggling to understand the 'loss of geography' section. In particular:

> As cable TV, Internet, and fast travel made national news more of a thing, and people switched from local-politics-as-part-of-daily-life to national-politics-as-entertainment, it turned out that almost everyone in Kansas was more similar to everyone else in Kansas than any of them were to people in New York, and so Kansas went solid Republican and New York went solid Democrat. But a national Democratic Party that has to include Kansas Democrats under its tent is a Democratic Party that's going to naturally be a bit conservative, a bit sensitive to the interests of religious people, a bit sensitive to the interests of farmers, etc. Once you move from within-state sorting to national-level sorting, things get a lot more sorted very quickly.

Wouldn't this *decrease* polarisation at the national level? I think I get the first bit: when people are focusing on local or state-level politics, the Kansas political parties find their way to the Kansas center, so that ~half of Kansas voters see themselves as Democrats; but when people start focusing on national-level politics, most of the Kansas Democrats realise that they lean right relative to the rest of the country, and so they become Republicans. But I can't figure out the last two quoted sentences.

Expand full comment

I read this book earlier this year and I found it very engaging (read it in 2 days), but was overall unsatisfied with his proposed solutions; particularly it did read to me that he was saying that the cure for polarization is for the Democratic Party to achieve complete and total victory.

Expand full comment

In case you are interested, here's my review from last year of Ezra's book:


Expand full comment

Here's my issue with the Zach Goldberg chart: If you made one of those 60 years ago, most of the issues would have been stuff like interracial marriage, Jim Crow laws, damming the Grand Canyon (which Goldwater proposed). Go back another 50 years and you have child labor laws, women's suffrage, workplace safety laws. Go back further when public libraries, national parks, belief the Chinese will never ever assimilate, and the question of should humans be allowed to own other humans - were the hot button issues of the day. Stuff we basically identify now as progress.

In 20 years gay marriage will be the same - duh, of course gay people should be allowed to get married. The only people opposing it will be the hardcore wingnut types who still oppose inter-racial marriage now. Same with healthcare - of course the US shouldn't be the only wealthy nation on earth to deny people healthcare based on pre-existing conditions - that's just common sense. Of course [X group] will assimilate, and of course immigrants aren't a net drain on society.

History leans progressive. Many of the hot button issues of today become common sense of tomorrow. You can always make a chart that makes it look like progressives are going off the rails.

Expand full comment

I hope someday you sincerely try to tackle the "why cant we go to the moon and why is our government objectively awful?" question.

It seems to be the most pressing current question. Unfortunately, It does not appear to me that the left takes it seriously. The left answer appears to be "more of what we have been doing, and do it with new people."

On the right, there are some who write directly about the question and try to sincerely answer it, though very few people of stature bother to address the arguments.

I would love to see you take a serious shot at answering that question from the ground floor.

Expand full comment

Something that bothered me about the book is that it doesn't link the people's affective polarisation with the political parties. There's definitely an increase in political polarisation across the world - incl Marine le Pen in France, Brexit, German far right parties rising fast, India, Hungary and more. I'm not sure how to square that seemingly benign affective polarisation chart with the political outcomes. And while Dixiecrats move would've had a role in creating a better party mix, I'm not sure why that would've led to more polarisation relative to all the other party mixes that have happened in the past.

One option, which I wish the book had discussed better, is the rise of mass media. Media creates a transparent information economy for the whole country. It ensures the speed of information transmission is rapid or instantaneous. So any political action you take has to have an immediate and positive effect. If that's the case, then naturally compromise gets harder, because you compromise to get something you want later. But the immediacy of feedback makes us all hyperbolic discounters. Makes us do things like McConnell not giving everyone $2k even though that would've won him all 3 branches of the govt.

Ultimately it makes us all greedy algorithms. And that's not a US specific topic, it's true around the world.

Expand full comment

I think both the comparison of whether Democrats or Republicans are radicalizing more, and of the relative success of their "rebel flanks," might be improved by separating economic and cultural issues.

QAnon and "it's racist for white people to eat sushi" have something in common: elites think it's dumb, but it doesn't really affect their bottom line, so they're willing to play to it when the polling says they can pander to those activists without weirding out normies.

Social democracy, much less a more vigorous model of socialism, *would* threaten the people who run the Democratic party. So Democratic leadership is willing to move left on race and gender, at least on the representational aspects of those issues, while making sure to keep the economic and foreign policy left out of control of the party.

Expand full comment

I thought the claim in the "Republicans suck" section was similar to what's discussed in "Neutral vs. Conservative: The Eternal Struggle"


The "neutral" mainstream institution leans left, which leads to a conservative offshoot which is more extremist. Fox News is not just the mirror image of CNN - it is more extreme, more actively aiming for partisan ends, more detached from reality & echo-chambery. And something similar is true of US polarization writ broadly.

The "Neutral vs. Conservative" post was more about the dynamic, and how the lack of neutrality of the mainstream institution plays into things. Klein's argument (I imagine) is more about the fact that this pattern exists. (Klein maybe also puts more of the responsibility on the rightward offshoots rather than the left-neutral institutions which inspire them, but that seems like a secondary question).

Expand full comment

It's not a huge issue with the article, but the impression I got on the first read through was that at the start of part 3 you were only linking Ezra Klein's article as an offhand reference, as you often do. So I did not expect the next paragraph to be outlining what Klein said in that article instead of his book stance, which you had been almost solely quoting him on up to that point. It took me until most of the way through the paragraph to realize my misreading, thinking at first that you were going to detail Klein's argument for something to the effect of "Polarization causes problems but the alternative is so much worse."

Expand full comment

I like Ezra Klein, but this sounds a bit like "The problem is that the outgroup is too polarized."

I guess there's no such thing as a completely objective and neutral analysis of this stuff though, so I'll probably still read and enjoy the book, and just price in the fact that he's a Democrat to my own biased interpretation of it.

Expand full comment

"One point kind of in support of this - ask Democrats their favorite news source, and you get a long tail of stuff (most popular is CNN at 15%, then NPR at 13%, and so on). But ask conservatives and it's dominated by FOX (47%). Does this lack of news-source diversity reflect a lack of ideological diversity? Could be. "

If anything, I think causality might flow the other way, if there's any relationship at all. IIRC, Fox is by far the newest of the bunch; all the others have been around for many decades, and Fox filled in a gap starting in the 90s by being the only one that wasn't extremely liberal. (Perhaps part of the polarization story should include why all the major news networks were liberal by that point in time! Were they always that way? Did they follow Congress, but precede the public? Has media trust decreased over time? My impression is yes, but I don't know, but it might be relevant.) I think the most likely explanation is really "there aren't enough conservatives who want to go into journalism to support more major networks."

(If you're interested in questions of media bias, I recommend the book Left Turn by Time Groseclose).

But also, I would want to see data indicating Republicans actually are less ideologically diverse before accepting that there is anything to explain. I don't trust Klein's impression, since Republicans are his outgroup, and https://en.wikipedia.org/wiki/Out-group_homogeneity exists. Speaking of, you discussed party identity, but perceptions about the other party might be more important than the reality: http://gsood.com/research/papers/partisanComposition.pdf

That paper suggests perhaps the easiest intervention for decreasing polarization, namely literally just tell the truth about both parties: "When provided information about the out-party’s actual composition, partisans come to see its supporters as less extreme and feel less socially distant from them."

Moreover, if journalistic bias contributed to popular polarization, could academic bias have done the same? Academia is probably even more liberal than journalism (e.g. https://www.researchgate.net/publication/264003803_Political_Diversity_Will_Improve_Social_Psychological_Science) and it looks largely like liberals and leftists pushing conservatives out rather than conservatives choosing to leave. Much is made of the "decline of trust in expertise" but if major institutions are increasingly biased, isn't distrust a logical reaction?

Expand full comment

Perhaps the polarization was mostly kicked off by the Democrats finally stepping on a couple of middle Americas sacred cows and middle America deciding to dig in. Specifically in regards to Christian moral values being the guiding factor in our cultural/political norms and American patriotism/nationalism/exceptionalism no longer being a widely accepted truth. That split seems to more or less have solidified around 2005 with the Iraq war and lgbt issues being a couple of the main splitting points.

Expand full comment

I'm honestly surprised that you missed two related, enormous, obvious-in-retrospect things:

1. Polarization among people tracks with the development of a right wing ecosystem, starting with talk radio in the 90s, then Fox, up through today. This ecosystem, in particular Fox, were explicitly built to cultivate a monolithic partisan voting block. It worked and continues to work. The other party is just "everyone who opposed this", with a plurality of that coalition backing/sorting into a vaguely liberal policy platform.

2. The GOP didn't get more *ideologically* extreme, but more *institutionally* extreme - or rather, anti-institutionally. The GOP of 20 years ago would have balked at sending out checks, but Trump backed it. Third is because the GOP is ideologically rudderless, and not really more conservative than it used to be. The old GOP also wouldn't have encouraged an angry mob to attack the capitol by perpetuating bald-faced lies about election fraud. This is the dimension on which the GOP has gotten much more extreme - dangerously so, as Klein correctly points out.

The institutional degradation under Trump was genuinely new and concerning, along a dimension completely divorced from liberalism/conservatism. If that's the only lens you look at things through, you're missing most of the story.

Expand full comment

You didn't mention it in this review, but does Klein discuss the transition from delegates picking candidates in smoke-filled rooms to open-primary systems where every step is determined by the popular vote? To my eye, one of the primary drivers of polarization has been politicians becoming more afraid of primary elections than general elections.

Expand full comment

hi everyone. i'm a "it's because of single round, first-past-the-post elections, resulting in a two-party equilibrium" guy. so i just wanted to chime in to say: i think it's because of single round, first-past-the-post elections, resulting in a two-party equilibrium.

Expand full comment

I'm not American, but the poetical story I always told myself is that Americans are fighting themselves after they lost the things they fought for together. To me, it seems that over the past 50 years Americans cared about lots of cool things like spreading democracy and capitalism, and now they only care about dealing with local problems. And who causes all of the local problems? The other party.

And this isn't a chicken and egg thing. Everyone seems to say Americans care about local issues now because of Polarization. I think it is very clearly the opposite, Americans are polarized because they don't clearly perceive any external threats to themselves (which bdw, America looks like it has a ton of external threats it should be concerning itself with).

Expand full comment

I think you're missing a crucial point with the "Republicans moving further right vs Democrats moving further left" issue, which is the social versus economic split. I don't have the fancy data to support this, but if you look at social issues (such as gay marriage or race equality), I don't think many people would say that Democrats haven't moved further left. However, if you look at issues like reducing poverty, increasing labor laws, etc, you'll see a very different view. In the into the 1970s, both parties often supported increasing social security (Truman created Social Security), labor rights (Nixon introduced OSHA), nationalized healthcare (Johnson created Medicare and Medicaid), the minimum wage, and more. George H.W. Bush signed a tax hike while every Republican president since has been in favor of tax cuts. Right now, Democrats support continuing those changes while Republicans generally prefer to go back on those changes.

Expand full comment

Some events that could have contributed to polarization of the general US population around 2000:

* Newt Gingrich's strategy of partisan obstruction.

* Founding of Fox news

Expand full comment

>I much prefer the Ezra Klein who writes things like Why We Can't Build: America's Inability To Act Is Killing People. Here he makes all of the impassioned and convincing arguments he avoided in his book

Regarding how Ezra has different styles in the book and elsewhere, he said in a podcast 'I only put things in the book that I think I could prove...this is my first book and I wanted it to be very grounded. So there are things that I think are true but cannot really prove it, so they did not go in the book'. It is at 19:40 in episode 'Your Questions Answered'. https://podcasts.apple.com/us/podcast/your-questions-answered/id1081584611?i=1000479525981

Expand full comment

There is a lengthy, deeply grounded, and deadly serious argument that the Republican party has actually gone totally off the rails. It's a shame to see it mostly dodged here. But I point you to the events of January 6th as a good starting point, and you can work your way backwards. Or start with https://en.wikipedia.org/wiki/It%27s_Even_Worse_Than_It_Looks

Expand full comment

Robert Putnam taught us that social capital (i.e. the network of relationships among people, both formal and informal) has been dropping in the United States for 75 years across all incomes, education levels, and demographics. This has led to decreases in social trust, cooperation norms, and a sense of shared identity. It is not hard to believe that this alienation of the American people from each other is leading to higher degrees of polarization. The specifics that Klein talks about certainly are part of the question, but it all sits in a much broader context.

Expand full comment

I want to chime in on the idea of polarisation being some kind of lynchpin of problems, that if we fix everything would become a lot easier. I live in New Zealand, which is less polarised than the USA (you can see that in the graphs above). But despite this, our government (whether red or blue) has been strikingly unable to do anything to solve a number of the big problems the country faces. NZ has the worst housing costs in the developed world, one of the lowest productivities in the developed world and has so far done almost nothing to reduce our carbon emissions despite having a substantial proportion of our capital and people deployed in areas threatened by flooding and sea level rise. We're going in backwards in education levels, too. This seems inconsistent with the idea that if polarisation were reduced, the USA would find a way to solve its problems (but then again, maybe aside from the polarisation the USA otherwise has very good institutions? I'm not sure).

Expand full comment

Enjoyed the post, minor nitpick, you refer to both Bill and Hilary Clinton as Clinton in the same paragraph, and it's a little confusing.


I think liberal, conservative, and polarized are terms that are all used in unclear ways here. For example the graph linked to in the twitter post reads to me as Democrats have become more Democraty than Republicans have become Republicany. The issues in the graph, for the most part, don't have an accepted coherent philosophy unifying them, other than Republicans take one side of the issue, and Democrats take the other. Reading the graph this way fits with the Democratic narrative that most Americans agree with the Democratic position on most issues:


So it's not surprising that Democrats are more Democraty, because Democratic positions are more popular. In this respect the Republican party is more extreme because it differs more from the median American position.

The other graphs also make me wonder. Does it make sense to measure historic liberal and conservative values of various parties, if the hot issues of the time were different, or the opposite positions on said issues were considered progressive? As an example I challenge you to consistently map US foreign policy to liberal or conservative over US history.

Along a similar set of lines I wonder what is being measured by the polarization over history graphs. Is it measuring difference in opinion or willingness to cooperate? Difference in opinion probably reduces the likely hood of cooperation, but it is not sufficient. Both parties may want to pass infrastructure reform, but are unwilling to make minor concessions lest their opponents get a policy win. Regardless of how polarization is measured, I think it's fair to say that we are not at a polarization high point considering we have not had any duels or canings on congress recently.

I'm surprised Klein doesn't raise Gerrymandering as a possible cause of polarization since this is a common talking point in Vox among other places.

Expand full comment

"Trump holds basically the same positions that Americans in the mainstream of either party would have held in a less polarized time (eg 1995); Clinton holds positions that everyone in 1995 (including her husband) would have thought insane, radical, and ultra-far-left."

Yes, except if we rewind to 1970, Clinton's positions (at least the economic ones) look moderate and Trump's... well, 2016-Trump didn't have any coherent positions, but the Republican Party would look insane and ultra-far-right. Both parties shifted to the right a lot in the 1975-2000 period, and the Democrats starting to swing back left is part of the more recent growth in polarization.

My belief remains that having political differences between parties is a good thing, though I'd distinguish between 'ideological diversity' and 'polarization', because the latter case implies having precisely two poles (/tribes), which *is* a problem but also may be decreasing a bit now (and in any case is the fault of the two-party system).

Expand full comment

> If you had some limited number of resources, and you wanted to improve (US) politics as much as possible so that the government made better decisions and better served its populace, what would you do?

Lawrence Lessig makes a compelling argument in Republic, Lost (https://www.goodreads.com/book/show/11814478-republic-lost) that the best way to improve US politics would be reversing Citizens United and getting money out of politics. Because both parties are in such intense competition, they spend all of their time fundraising instead of governing. If we made fundraising illegal, they'd have nothing better to do than good government, and they'd stop trying to get you to hate the other party.

Expand full comment

The article about polarization data from 9 countries article doesn't seem to totally back up the idea that the internet/social media isn't to blame. Looking at their raw numbers, instead of their 2nd derivatives, Britain and Australia are roughly equal to the modern day US (which matched my understanding of their political climate), but are shown as having about an unchanging level of polarization over the past 20-30 years.

It could be that one conclusion is that the internet has nothing to do with polarization, but another potential one is that the internet raises polarization up to a certain cap, and other countries were already at that cap and have been even before the internet. Looking at the graphs, only 3 countries show a trend downwards from their current level of polarization, and they are all still (excluding Germany) currently at a level of polarization that roughly equals the US (45).

While this is a more complex explanation and thus goes against Occam's razer, it also seems really unlikely that the internet has nothing to do with polarization. After all, the Loss of geography section seems like it perfectly fits with the internet being to blame.

Expand full comment

"That means the Republicans are more ideologically uniform - Christians are genuinely similar to other Christians, but Jews are only superficially similar to Muslims by virtue of their non-Christianness."

Republicans are less ideologically uniform in Congress. Probably less ideologically uniform among the general population, too. In both the Democratic and Republican parties, the primary opposition is from the right.

All evidence suggests the American party system (based on the electoral college) naturally tends toward polarization. America depolarized beginning in the 1920s largely because old debates were becoming less relevant, and the 1930s created the left-right ideological axis as the new primary difference between the parties, resulting in depolarization over the next forty years as the old pattern of polarization dissipated. In the 1970s and later, party constituencies became increasingly coherent to match the left-right ideological axis on which issues were debated. Even as late as 1976 you could see a presidential election that wasn't TOO different from that of 1876. No longer.

Expand full comment

I'm a little surprised to see so little discussion of religion. A big part of the changing dynamics of the 80s into the early 90s was the complete embrace of the Republican part by evangelicals, along with their re-entry into the political sphere after generations of trying to separate from society. Much of the socially cultural issues that caused deep divisions from 1980-2010 were in that area - abortion, gay marriage, school prayer, flag burning. The weird thing is that seemed to have a delayed effect up through the mid 2000's, when you people started abandoning organized religion, and the Democrats really started to be the party of the youth. This then naturally merged into the racial divisions that were exacerbated by right-wing opposition to Obama as a legitimate president.

I would have liked to see much more discussion in the book about the dimensions and correlations of religious/political polarization in the US. That could explain why we are so different from other WEIRD countries.

Expand full comment

This is a good essay that gets to a good place and raises lots of good points, but I'm going to quibble with you on one point and then just point something out, but let's just understand I'm only quibbling on the point, not the overall piece.

OK? Cool.

Observation: I just don't think it's very useful to compare the United States of America to Germany (as Klein did in the post you cited), because America is just so big. Let's see a polarization graph for the whole European Union and then we can talk.

More size, more types of people, more extreme people, more differences in culture that make it harder for people to communicate.

People in small countries have a great deal more culture common ground which makes communication easier.

A small town Kansan (like I grew up) does not have a lot of context in common with a Brooklyn-native (the people of where I live now). The mutual suspicion there is acute. I don't think it's quite so acute for a rural vs. Urban German.

That's more an observation.

To quibble: I don't agree with you that the GOP is reacting to the craziness of the Dems. I agree the Dems are crazy! But the GOP has had its own crazy that presaged the current crazy for a long time. I feel like people have forgotten the late 90s era of the Christian Coalition not-quite-dominated-but-heavily-influenced Republican party.

That was quite crazy too. Not as crazy as now but it was crazy. And it sewed a lot of seeds for today's identity politics, particularly along the sexual faultlines.

In sum, each has brewed their own crazy and they both go way back. Each has also positioned themselves in contrast to the other, as well, though. Both things have happened. History is a mess. Anyway: Moderatism doesn't sell tickets.

And as an aside I've always had a hard time with moderatism on some issues. For example: I want to see a pretty hard ass approach to climate change. The time for moderation has passed. And even if it hasn't passed, if it turns out we go to far, we'll still come out with a nicer, cleaner, more healthy world on the other side of going hardass on it, so it's fine regardless. The utility value of going green is quite high to all even if the threat is overstated.

But I don't think it is.

This feels a little like a digression but I don't think it is. A mushy middle isn't really the answer. Restoring a consensus of nationhood and we all have more of a responsibility to that consensus than we do to political points seems more important.

To your point about an EA of politics: I've long thought crushing gerrymandering would be the best effective altruist solution here. If districts were more politically muddled and primaries were more of a contest we would see a different politics.

Gerrymandering is more important than the money. Usually people get mad when I say that but I suspect this crew won't.

Expand full comment

The thing that always strikes me about US political polarization is how clean of a rural/urban split it is, where urban areas all over are strongly Democratic and rural areas all over are strongly Republican. When did it start being this way? Do other countries see this same split?

Expand full comment

"I'm not sure I fully understood Klein's explanation of exactly how this happened. After all, since identities are not 100% correlated (ie not all farmers are Protestant and not all city-dwellers are poor), you can't actually do this for the whole population at once. I think Klein would say that these correlations went from kind of random, to the ones that capture the biggest slice of the population."

Two things here:

1) People underestimate how very NOT diverse America was until recently. We were an 85%-90% white country from 1860-1960, and over 90% Christian, so those were not identities that you even could polarize around. There was some slicing and dicing of the various Christian sects (e.g. Catholics overwhelmingly supporting Democrats), but I think that's how this has shifted--the "big" identities have become much more polarized because they've become smaller.

2) Klein sort of misses something in this analysis, which is that polarization can be explained as much by the *erosion* of identity as by the *realignment* of identity. The question of the book is of course "Why are we polarized?" but as you do a good job explaining, the real question is more, "Why did we go from a system where you had conservatives and liberals supporting each party, to one in which conservatives overwhelmingly support Republicans and liberals overwhelmingly support Democrats?" But you could also pose the question as "What kept a bunch of conservative white people in the Democratic Party for so long?" The answer in the South was that the Democratic Party was completely tied up in white Southern identity. In the North it seems to be labor unions, or more broadly that they saw the Democrats as the "working man's party," that kept them in the party. I think the breakdown of the labor movement--and the class consciousness that came with it, not just the material benefits--kept a lot of conservatives in the Democratic Party.

As far as what polarized the parties, I see it as (in chronological order):

1) Civil Rights Act: As you said, this led to the decline of the Democratic Party in the South, essentially pushing conservative white Southerners closer to the GOP)

2) Reagan: He actively campaigned against moderate Republicans before becoming president and strongly pushed the idea that the Republicans should become a conservative party, basically making the conservative wing of the party ascendant

3) Iraq War

4) Obama: I don't think he tried to polarize the country, but he was basically the liberal dream candidate--black, son of an immigrant, intellectual, more dovish than Kerry, Gore, and Clinton--and the conservative nightmare. The 2010 election was basically the end of the conservative Democrat in Congress, and I don't think you can disentangle that from conservatives who had historically voted for Democrats being turned off by Obama.

5) Trump: Basically the mirror image of Obama

Expand full comment

RE international comparisons: which other developed countries have two-party systems? From a cursory Wikipedia, it seems like the answer is Malta and that's basically it.

Other contenders:

South Korea often only has two big parties, but which two seems to change every election.

Australia has two big groups, but one is the Coalition of two parties, and the senate is elected via PR and has 20% minority representation.

I am more familiar with the UK and don't think it counts, we had a coalition government pretty recently and the Scottish National Party are pretty big.

Any others?

Expand full comment

Inequality began its explosive rise starting in the late 70s due to globalization, the loss of labor bargaining power, the rise of the managerial class, etc. As the producer of the world’s reserve currency, the US has a unique role in globalization which magnifies the effects (both good and bad). The massively divergent outcomes among members of society, plus what I would argue resembles stagnation at lower rungs, created fertile ground for a loss of social trust and a rise in angst. In the absence of deep, productive debates on the true underlying issues, we simply default to tribal partisanship and culture wars.

Expand full comment

Surprised I haven't heard anyone saying this - what about wealth and income inequality? Given the correspondence, seems as if an explanation would have more power if it related to them both. https://images.app.goo.gl/biMLpT78K9rK1QJr7

High levels of blatant inequality in the 1880s-1910s, pushes polarization until enough solutions get passed, things go down, people are happy, welfare state starts stagnating and retreating, inequality goes up, pushes polarization.

Why would wealth inequality push Polarization? People are unhappy, don't know why, blame the outgroup.

Alternatively, only way to suppress class consciousness is to give people some other identity to care about.

Expand full comment

If there's any discontinuity around that time, it'd be 9/11 and Iraq. I think 9/11 was the original cause, but it took a couple years to show up, because in the short term it was a unifying factor. It wasn't until Iraq happened, and started to go differently to how Bush suggested it would, that it exploded into polarization.

This is a guess, obviously, but it seems to fit well enough.

Expand full comment

Maybe this is a consequence of coming of age in the 2000s, but I'm always going to associate political agreement and ability to get stuff done with the wars in Afghanistan and Iraq, as well as the creation of DHS, domestic spying and other civil liberty issues, etc., that came after 9/11. Democrats and Republicans agreed about a whole lot of that! Yes, I see that there are many problems with polarization. But you don't have to reach back very far to find examples where "inability to get stuff done" would have looked a lot better than the alternative.

Expand full comment

Does the chart of polarization show the UK becoming less polarised? Because that does not match my impressions. I would say less until the mid 90s and then flat for a bit (maybe not quite - Blair was dominant so the Tories going down weird routes for a bit did not affect much) and then the left reaction against Blair started with Iraq/Bush which started moving bits of Labour apart just as the Tories went centre enough to have a chance to grab back power - the 2008 crash made this worse and then the Tories won and while socially liberal they were harsh cutters of government services, which made polarisation more urgent. The Lib dem implosion as part of the coalition destroyed the middle for a bit, then Brexit/Corbyn/Boris sharpened the cultural aspect until there was very strong polarisation.

So I would expect a U shape with the minimum late 90s early 00s

Expand full comment

I don't think anyone has mentioned Bush/Gore and 9/11? Those are two events that were huge in the popular consciousness that fit the timeline. Of course you can't draw a line in the sand, can you - just why was the 2000 Presidential race so evenly split - in my memory both Bush and Gore were pretty unremarkable candidates. All I know is that you still can find people bitter over the outcome, which naturally ought to contribute to a "win-at-all-costs" mentality. And then you had 9/11 and everything that followed. Attack on American soil! It was quite a shock to the system after a decade of relative peace and harmony. People were digging in on jingoistic patriotism versus the other side advocating a more measured response and saying, No the US is the true aggressor here. And throughout the Iraq War the justifications became more murky and Bush expanded his power in unpopular ways and you had things like the Patriot Act and Guantanamo. Bush was so hated that a faux-documentary was made about his assassination and what would be the aftermath. I don't have a point to all this, perhaps it's too much of a just-so story. But I definitely see some of the origins of the current culture war in those times.

Expand full comment

I think an interesting subject for further inquiry is why the Democrats have been more successful at fighting off insurgencies that Republicans.

My moderately informed opinion is that institutional Democrats have a lot more institutional support than institutional Republicans.

If someone primaries Mitch McConnell, they will get a lot of favorable coverage (in the primary campaign) even if that person's views were much more abhorrent. If AOC were to attempt to primary Chuck Schumer, it would be seen as a threat to the republic.

When Trump was launching his campaign, everyone thought it was hi-larious how he was taking down the Jeb Bushes and Marco Rubios of the world. I recall a lot of tweets about how Trump maybe really nuts, but what's really nuts is the numbers in Rubio's tax plan!

Meanwhile, the leasing institutionally conservative publication published an "against Trump" cover. But it didn't matter.

Bernie Sanders was always a threat. He was pressured to quite. People were really concerned about "Bernie-bros.", etc. Try talking about Jill Stein in certain quarters.

Expand full comment

I'm more than a little surprised that population density didn't come up once.

The simplest way to describe the difference between republican and democratic coalitions is population density. Democrats are predominantly an urban coalition, and republicans are more rural / suburban.

When you factor in that most economic growth has taken place in cities, and cities are always where almost all culture is produced, you see a pretty obvious pattern:

- cities are where almost all economic power is concentrated

- cities are where almost all cultural power is concentrated

- cities are dominated by democratic politics

He's got a point that AOC/Bernie have less influence inside party politics - but there are plenty of SJW extremists writing in the New York Times. Sure, QAnon is totally off the rails and there isn't anything like that on the left, but the closest left wing equivalents are tenured sociology professors.

Whenever I leave California to go visit my parents in Ohio, I feel like I understand the sense of despair and anger and frustration that people here feel, towards the place that took a lot of their kids, and responds by calling them racist, bigoted people for not going along with this increasingly radical agenda, which seems coupled to economic policies that are AWESOME if you leave your hometown and move to a big city and learn to do really insanely complicated things for a giant corporation, but otherwise are kinda meh.

It's been difficult for me to escape the conclusion that you can't possibly have a stable political coalition that covers "city dwellers" as more and more of the country movies into cities - unless those city dwellers are united against a common enemy. All around San Francisco in 2014, there was a big tech backlash, with people spraypainting '#DieTechieScum' on the sidewalk. It all went away around the time Michael Brown was killed. I watched how quickly all the big tech companies joined into BLM, and remembered thinking at the time "now we have an enemy that we can all feel good about hating."

If this hypothesis is true - that increasing urbanization drives political polarization - the solution would be... some kind of ... thing.. .which caused lots of people to start leaving big cities all at once. Maybe spreading tech wealth all over the country, and increasing the frequency of interaction between highly trained technological professionals and blue collar workers. Who knows what would cause such a thing?

Expand full comment
Feb 10, 2021Liked by Scott Alexander

One theory of the cause of increased partisanship since the 70's that has been growing on me is transparency/sunshine laws. Apparently, pre 1970's congressional/comittee meetings and votes were all done closed door to the point where we don't even know how specific legislators voted on specific legislation. But beginning in the 1970's a wave of legislation calling for increased "accountability" opened up these proceedings to the general public. But, the theory goes, special interest groups, corporations, media interests etc are a LOT better at holding their representatives accountable than the average joe voter. Where previously they were still motivated to enact regulatory capture, but ultimately didn't know if it was working and couldn't easily organize against someone who defected, now they could. And the legislators now knew that they could.

And where previously they could give a good partisan speech but then go behind closed doors and compromise or acknowledge when someone from the other side made a good point, now there was no longer any backstage. The theatre element of politics is on and you can never be seen to not be treating the other side like they're the bad guys and we're the good guys.

Also with the transparency laws, the media gains power because they are the largest influencers of the only group that can possibly counter the special interest groups, the average voters. So the value of influencing the average voter goes up which, of course, increases competition for that influence. And the more the battle for that influence heats up the more you need to band together to win(media consolidation), the more you gain from demonizing the other side, and the more incentivized you are to do it.

The link I posted above has some really compelling research and arguments that link transparency laws to polarization, regulatory capture, government waste, pork, bill size inflation/obfuscation, wealth inequality, increased campaign spending, decreases in crossing the aisle voting, increases in "bang for your buck" for lobbying spending, wage stagnation/inequality and a whole host of other problems. It's a surprisingly robust theory.

It also offers an easy solution. Closed doors and secret ballots.



Expand full comment

Keeping people polarized seems to be the way political parties gain votes, all over the world. It is just a numbers game. In India, this means a politician running for election in a constituency being of the right caste/religion and promising just the right amount of favors to the right groups. Not too much, because that is a waste of resources.

The only time voters don't care about this is if the politician has an extraordinary personality. Then they vote for him or her even without being promised favors. So it is identity/personality.

It seems like the same thing happens in America these days. Is "Why we are polarized" really any more interesting than this?

Expand full comment

I think he's wrong that "one party called itself "Democrat", but had few similiarites to the Democrats in the North". The primary axis of politics was still a right vs left one focusing on fiscal/economic issues, with the secondary/regional one that southerners differed from northerners on being a smaller axis. Klein's co-founder Matt Yglesias has a more realistic view of how important Dixiecrats were to things central to Democratic politics.

Expand full comment

I feel like this talk of "polarization", and which party has gotten more "extreme", conflates a few things, and I think the talk of how extreme the two parties are on various popular issues obscures a few things.

This is a point I've made several times before -- the word "extremism" really conflates two different things. One of these things is, well, just object-level positions that are extreme, which is fine and in fact you tend to end up there if you try to be at all coherent. The other is "meta-level extremism" -- resorting to or endorsing violence, threats, etc., and which rather than "extremism" would be better called "illiberalism".

This discussion of polarization is similar -- is there a necessary connection between the parties moving further apart on object-level issues, and them becoming more illiberal in their attempts to get their way? I don't see that there is. It's what people expect, but the two are logically separate, as I see it.

But it's the illiberalism that's really dangerous, not the having of different positions. This is why I think looking at the positions of Democratic and Republican candidates and saying "well, the positions of the Republican candidates would have been fairly ordinary in the 1990s" is misleading. Sure, maybe if you restrict to object-level issues, but if you look at the meta-level issues, about how our democracy should function, not at all! I mean, we recently had the then-president encouraging the storming of the Capitol building, a bit hard to say that would have happened in the 1990s. I think when you look at the meta-level it's clear the Republican party has gone off the rails much more than the Democratic party. (And yes obviously there is a substantial illiberal wing there too, what with the SJers and all, but as mentioned above, they're not in control of the party, nor do they seem to rise to the same level.)

And you could say that well that wouldn't have happened in the 1990s because things were less polarized then, but my point is that these are two separate things; I think it's a mistake to automatically collapse together polarization and illiberalism. Even if they typically go together, that link should be made explicit rather than assumed.

Expand full comment

I think the US is more polarised because it has a hard two party system, third parties exist but they have close to zero relevance on the national stage. Other countries like UK, Ireland, Canada, Australia, New Zealand, etc, don't – they have what you'd call a soft two party system, in which there are two major parties, but third parties are represented in the national legislature, and sometimes the governing party has to rely on the support of a minor party to govern.

I personally think having a Presidential rather than Parliamentary system is part of what gives the US a hard two party system. The Democrats and Republicans have always been quite big tents with weak party discipline. Other English-speaking countries, the major political parties tend to be somewhat smaller tents with stronger party discipline, which I think promotes the major parties being narrower coalitions which leaves more space for third parties to exist. You need stronger discipline when the executive depends on a majority in the legislature to continue to govern. The independence which the executive has in the American model reduces the need for strong party discipline in the legislature branch.

First-past-the-post voting is another contributing factor, although it can't bear all the blame because Canada and the UK have it too. I think the parliamentary system dilutes the negative impact of first-past-the-post. Plus both Canada and the UK have far greater internal cultural diversity than the US does. Many people in Scotland and Wales view themselves as historically distinct nations, and they have aspirations for independence and parties which cater to that aspiration (SNP and Plaid Cymru). Northern Ireland politics still revolves around the conflict between the Unionist and Nationalist communities, which gives Northern Ireland its own completely distinct party system from the rest of the UK. In Canada, you have two distinct language communities (Anglophone and Francophone), and a significant minority of Quebecois still want independence. People say there is a big cultural difference between New York and Alabama, and no doubt there is, but it really pales in comparison to the internal cultural differences within Canada and the UK – despite their differences, people in New York and Alabama both view themselves as members of a common nation with a shared past and a shared future, whereas significant numbers of people in the UK and Canada don't.

Expand full comment

"He says that the Republican Party represents the modal American on various characteristics" What does "modal" mean in this sentence?

Expand full comment

"Breaking the Two Party Doom Loop" also argues against the "vetocracy" and alludes to assymetrical issues in our political alignment, but with a more keen eye towards the Senate (which leads to dramatic changes in the power of your vote based on state population) as a key driver.

Expand full comment

If the Internet as an explanation for increasing polarization is lacking because it raises the question "why only here?" and racial resentment is lacking because it raises the question "why now?", it seems like these issues might answer each other's questions? Is there another country which was primed for race-aligned polarization before the Internet that did not become more polarized?

Expand full comment

My intuition is that the change in the media landscape (the Martin Gurri thesis) is a significant factor. Specifically, the change from a small set of gatekeeper voices limiting the overton to an infinite number of voices, reflecting the opinion of all people and creating "common knowledge". We went from a world where in 2008, Barack Obama couldn't publicly support gay marriage to a where where today, if a prominent person did not support gay marriage, they would be ex-communicated. I think it's important to think through the causal mechanism behind how this happened.

One possible theory as to why the changing media landscape did not increase polarization in the rest of the first world - is that the rest of the world follows the news/entertainment complex of the US rather than their own domestic one. Jon Stewart and Stephen Colbert were watched all over the developed world; I suspect this polarized people 'against' American republicans without having a significant impact on how they view their own country.

Another plausible theory is that the changing media landscape only would have increased polarization if the country was ripe for an increase in polarization. Factors like Newt Gingrich and Gerrymandering could have sufficiently changed the situation to allow polarization to take off.

Expand full comment

I am not sure if someone else mentioned this, but isn't the obvious response to your point on technology that the internet is itself not neutral? The experience of many Europeans of politics on the internet tends to be with specifically American politics, whereas Americans will not have the same experience.

Imagine if Americans browsed a Chinese internet with political discussions dominated by discussions about e.g. Xi Jinping's new anti-corruption campaign or whatever instead of about domestic political affairs.

Expand full comment

Does Klein provide detailed evidence about the Dixiecrat/racist southern Democrat shift to the Republican Party after the GOP in Congress passed for the Civil Rights Act, or does he just repeat the myth and not provide any foundation?

Because while the myth of the "Southern Strategy" is popular among the left who want to believe Republicans are racists, it doesn't actually fit the electoral facts. See https://www.nytimes.com/2006/12/10/magazine/10Section2b.t-4.html, for:

" In their book “The End of Southern Exceptionalism,” Richard Johnston of the University of Pennsylvania and Byron Shafer of the University of Wisconsin argue that the shift in the South from Democratic to Republican was overwhelmingly a question not of race but of economic growth. In the postwar era, they note, the South transformed itself from a backward region to an engine of the national economy, giving rise to a sizable new wealthy suburban class. This class, not surprisingly, began to vote for the party that best represented its economic interests: the G.O.P. Working-class whites, however — and here’s the surprise — even those in areas with large black populations, stayed loyal to the Democrats. (This was true until the 90s, when the nation as a whole turned rightward in Congressional voting.)"


"Wealthy Southerners shifted rightward in droves but poorer ones didn’t." Is the model of the Southern Strategy theory that poorer southerners weren't racist?

The facts are that the least racist southern states turned Republican first, and the most racist states last. The opposite of what you'd expect if the switch was based on racism. It makes no sense that racists in the south would switch to the less racist party. Remember, the southern Republicans voted in favor of the civil rights act while southern Democrats voted against it. So the theory is that southern racists decided to.... join their ideological enemy in their state and start voting for them instead??? It's not a very coherent story.

Rather, the middle class increased. Small business owners increased. People moved into the south from northern and western Republican strongholds. That's what the data in the book shows how the south turned Republican.

Klein has a nice just-so story, but I'd bet he doesn't have the detailed voter-level data to back it up, because the myth of racists switching wholesale to the GOP post-civil rights act has already been disproven by Johnston and Shafer's work. Anyone still promoting it needs to address their data.

Expand full comment

I recall watching a C-SPAN lecture a few years ago about Eisenhower. In 1952, both the Democrats and the Republicans wanted Eisenhower to be their candidate. When the Republicans won that tug-of-war, the Dems enlisted Adlai Stevenson. Stevenson ran a half-hearted campaign because he, like apparently every single other American, thought Ike would make a great president. In 1956 Stevenson had gotten a lot grumpier about Republican policy, and ran a more forceful campaign, but of course still lost. I wonder vaguely if the nearness or distance of a devastating war with lots of middle-class and upper-middle-class loss of American life plays a role in our divisiveness.

Expand full comment

Scott - you missed the part of the book that implicates social media and explains Klein’s change in attitudes.

I’d have to look it up (listened to the audiobook) but there’s a chapter that details media strategies at Vox where they deliberately try to “activate identities” and find that’s the easiest way to drive clicks. That could mean (examples form memory) benign things like 90s kids and left handed people, or any part of the intersectional collation.

The feedback loop here is- the more you profit off of activating identities, the more incentive you have to taller all of your behaviors to that paradigm—everything you do becomes part of your profit-driven identity activation until you forget how to see things any other way.

If you listen to the audio version of the book (like I did) you get to hear the giddiness in his tone change as he discusses the dashboard software Vox uses to monitor traffic on articles. It’s hard to hear it and not think of him and something of an addict.

Expand full comment

My hunch is that gay rights are underrated in the polarization surge of the past two decades.

When I showed up at my elite US college in 2004, the student population definitely leaned Democratic, but there was a live split - I think 4 of the 7 guys in my freshman suite voted for Bush, as did a substantial minority of the other folks in my dorm. For the most part, though, those people were uncomfortable with the Republican party's position on gay rights; I recall a conversation between my roommate (an ambitious and extremely outspoken Republican) and one of our dorm neighbors, the daughter of a well-known Republican official, in which they talked about that as an issue where they disagreed with the party's position but hoped it would change soon. These were two extremely central members of the Republican coalition as it existed in 2004, but the social milieu of college students had moved fast enough that they were dissenting towards the nationwide minority of liberals on one of the central issues of that year's campaign.

The salience of gay rights and the Iraq war during the years I was in college was absolutely devastating to the Republican party's standing within that cohort. If that class was 60-40 for Kerry (I'm just guessing) it was probably 80-20 for Obama, and those two issues were doing almost all of the work. Shit, a few years after that an old friend of mine went to Harvard Business School; I asked him what the political breakdown of the class was and he said it was basically monolithic. At a business school! That would have been unthinkable 10 years earlier.

Think, then, what that meant for institutions that hire college educated workers; if their workforces were moderately Democratic-leaning in the 90s, all of a sudden in the 2010s they'd be massively Democratic-leaning without any real change in the hiring mix. Already liberal institutions, meanwhile, all of a sudden have an audience that's richer (for fundraising purposes) and more culturally similar to their own college educated workforces, and easier to talk to, and that facilitated a lot of them getting lazy about communicating outside their most natural idiom.

There's a whole Fox News centric story to tell on the other side of this, but I don't really understand it very well.

Expand full comment

Always worth remembering that although "my side is just so much closer to being right than the other side, which is compromised of intellectual and moral barbarism" sounds very jejune *that doesn't mean it's not right*.

You're right that the US hasn't been able to accomplish much during a period of hyper-polarization, but this is more a symptom of the US's very poorly designed constitution. Americans are taught that their constitution reflects a special genius in its design, but actually it's engineered to produce gridlock unless people can come to a consensus. As you note, through sheer fluke, consensus was possible for most of the first few hundred years (except for that civil war...) but now those good times are coming to an end. Hyper-polarization isn't going away, so if I were looking to save America I'd suggest learning to live with it by moving towards a unicameral legislative branch on a parliamentary model (no senate, no presidential veto, reduced supreme court oversight). At present the US might be going the way of the Polish parliament under the liberum veto: http://www.demos.org/sites/default/files/publications/Whose%20Voice%20Whose%20Choice_2.pdf

Unicameral parliamentarianism has two advantages:

1. Things can actually be done.

2. Voters know who to blame when things go wrong- the governing party.

It also has zero chance of happening right now. It's probably more likely than hyper-polarization going away in less than 30 years without a civil war or similar decisive event, though.

Expand full comment

Can't help but think the simple solution here is to cut the Gordian knot & do a secession. As Scott has observed, the blue tribe is way less attached to America memes than the red tribe, so let red America stay as the United States & both coasts can split off. Florida can be it's own country, with Donald Trump as king! Then you can go back to Kansas having real contests between Kansas dems and Kansas repubs who actually represent their state - beating back Moloch with good fences.

Or not. That ship may have sailed, and with everything being so interconnected these days it might not be so easy to escape the sufficiently intense competition. Brexit hasn't seemed to spur a wave of localist English politics, the most major development recently (imo) has been the internationalist neoliberal Starmer purging his Labour party of leftists. Maybe going small just means placing yourself under the thumb of a hegemonic China.

I still think some kind of secession movement is likely in coming days. Red tribe anger is spiking & can only simmer for so long before it finds an effective outlet. The blue tribe meanwhile seems to be content to fight fire with oil in the form of a domestic war on terror. There's lots of speculation about a civil war brewing but I view this last year as instructive: despite everyone having guns & being incandescently furious, hardly any combat fatalities happened. My take is that the vast majority have too much to lose to throw away their lives on the barricades. Still mad as hell though, and in this case ignoring your enemy might just make them go away.

Expand full comment

From the Discord:

> Really the most interesting thing is the very end of the post, why is the US facing polarisation in a way that say Australia or Germany is not

> Yeah, that's the only thing in the review that actually made me think. One of my leading hypotheses was that social media has been driving US polarization, and this pretty much disproves that unless you add a ton of epicycles

> On the flipside, in that David Shor interview he linked a study looking at broadband rollout that did find it had significant effects boosting polarization. ( https://www.jstor.org/stable/26379489?seq=1 ) Someone want to post this in Scott's substack comments btw?

Expand full comment

The timing of congressional polarization seems more consistent with Roe v. Wade being a driving factor than the CRA, and also with the story conservatives tell about themselves (sincerely, I think). Moreover, during the 50 years after the CRA, white people consistently became less racist while black people got better off and more access to power. 'White identity politics' then became more pronounced during a period when black incomes were actually pretty stagnant compared to the previous half century, so I don't really buy racial resentment being a driving force.

As others have said, religion seems under-discussed as a factor. Maybe secularism finally hitting the US in a big way in recent decades has made conservatism increasingly desperate and pessimistic; other trends have had the same effect. The right these days seems to be driven by deep-seated pessimism and is constantly anticipating its own extinction, or at least that's my impression, and it seems at odds with Klein's.

Expand full comment

re: "After Lincoln's death, his successor Andrew Johnson decided this sounded hard and gave up. Within a few decades, the South was back to being a racist, paramilitary-violence-prone one-party dictatorship."

While Andrew Johnson was that bad, this implies that he was responsible for events but this was bitterly contested the whole time. (You might say the country was polarized.) Republicans in Congress kept Reconstruction going (overriding Johnson's vetos), and then Grant was elected. It really fell apart in Grant's second term and ended when he left office.

Also, this seems to imply that there was ever a time after the Civil War when the South was not a racist and violence-prone place? But the Reconstruction-era governments survived because the U.S. army was still occupying the South, and even so there was plenty of racist violence the whole time.

I recommend Ron Chernow's biography of Grant for a better view of this. (It's been a while so I might not have remembered it all correctly.)

Expand full comment

Is the point that the USA is only now coming into the same levels of polarisation as other developed western nations, but is struggling to cope with it an argument that the governmental system (congressional rather than parliamentary) is no longer fit for purpose?

Expand full comment

Are you familiar with the Will Wilkinsons thesis that the parties have been shifting along urban/rural party lines? This is an ongoing shift that has been further exasperated by globalization. These two ingredients alone, along with the downstream cultural implications, seem to me to explain the polarization change much more than any race narrative does.

Expand full comment

One argument that I've found to be somewhat compelling is that as the federal government plays a larger role in local politics the conversation around national politics grows more polarized. That polarization isn't necessarily new, but that it would have been more focused on local issues in the past. The reason news and political discussion transitioned to national is not necessarily a technological one, but that as the federal government began to have a direct and tangible impact on everyone's daily life the only reasonable response is to care more about what it's up to. I'm not totally sure if this lines up with the polarization graphs, or even how you'd exactly measure the the level of impact the federal government has on daily life, though.

Expand full comment

>This does not seem obvious to me. Trump holds basically the same positions that Americans in the mainstream of either party would have held in a less polarized time (eg 1995); Clinton holds positions that everyone in 1995 (including her husband) would have thought insane, radical, and ultra-far-left.

When you say "Trump's positions," Are we talking "What Trump stated his positions were on the campaign trail" or "What Trump actually succeeded at implementing as President"? Because the latter might be described as mainstream, but the former definitely can't. Not unless "Ban all Muslims from entering the country," "Make Mexico pay for the construction of a border fence," and "the Democrats can't win an election without massive voter fraud" are mainstream Republican positions.

That last one in particular should carry a *lot* of weight, since you're approaching this from a "how to avoid civil war" perspective. "Candidate is looking into extralegal ways to retain power" is much more worrying in that regard than "Candidate holds extremely left-wing economic stances."

Expand full comment

Interesting that World War 2 as a great antipolarization event isn't really mentioned either. Obviously the R or D doesn't seem some bad when you're fighting alongside them against Nazis...

And you'd expect the afterglow to match quite some time, especially in a time when all the "non-partisan" experts who won the war where widely trusted to order the economy, govern, etc... Between aging and the shine coming off in the Vietnam era, you would expect that to fade around the 1970-ish inflection point you note.

I also would hypothesise that as faith in the "Best and Brightest" continues slipping away, it's easier to assume those in charge are incompetent. Given human nature, this makes it a lot easier it to look at the other side and blame them for wrecking everything, as opposed to a grudging "they are on the other side but I still respect their competence, expertise, etc" which would work against polarization.

I wonder if there are any good "faith in elites" or "faith in experts and institutions" measures to check this against? Anyone know?

Expand full comment

The nationalization was accelerated under Newt. From post war to early 90s, Republicans did great in national (presidential) elections, and poorly in local congressional elections. Newt's plan to win congress was to nationalize all politics, and institute the Hastert Rule (https://en.wikipedia.org/wiki/Hastert_Rule), which accelerated polarization.

Expand full comment

I felt confused about the cross-country polarization graph, and went and read the paper it's from (https://www.brown.edu/Research/Shapiro/pdfs/cross-polar.pdf). Here are some things I learned, in case they're of interest to others:

The authors don't assume a two-party system; they "extract each respondent’s party identification, and exclude 'leaners' who only choose a party identification in response to a second survey prompt" and then look at the average difference in affect (normalized to a 0-100 scale) between someone's own party and all the other parties they reported feelings about. This is a better method than what I imagined and makes the graph more interesting!

For example, Canada's Conservatives (the right-wing party) had a schism right at the beginning of the trend (1987) and reformed in the middle of it (2000). I would have expected the reformation to increase polarization (since now right-wing people have no alternative parties) but the trend shows a decrease. Their appendix figure showing how many people declare a party affiliation doesn't jump up in 2000, either (which is what I'd expect to see if many pre-2000 right-wing voters were excluded as 'leaners' between parties). So the Reform-PC merger didn't really impact the polarization trend. Neat!

I am still not 100% clear on how they measured "party affiliation", but I now think it's something sensible. I was initially unsure how they could do this; in Canada, about 1/3 of people vote for neither the Liberals or Conservatives (leading, as others have noted above, to a "soft" two-party system) and very few people are officially registered party members. The paper reports the exact surveys they extract affliliation from, and the Canadian Election Study data is online (here's 2015 for example: https://search1.odesi.ca/#/details?uri=%2Fodesi%2FCES-E-2015.xml). I'm still not 100% percent sure which question they used for party affiliation, but my guess is that it's "Which party do you think you will vote for?", which seems pretty sensible to me.

Expand full comment

Polarization wouldn't be as big of a problem if it didn't lead to unrepresentative governance and political paralysis. The big culprits here are stuff like partisan gerrymandering, the filibuster, and the Senate - all undemocratic institutions that work to limit partisan accountability at the ballot box for their failings in governance.

Expand full comment

Polarization still feels pretty squishy to me. The bias of declinism will mess with our instincts here.

But there's a whole book on it from a really smart guy, so playing along... I might go with internal migration patterns.

Urbanization took off in the last half of the century, it self-selects younger people with certain skills who, most importantly, are high on "openness to new experiences" measures, which anti-correlate with certain conservative positions. It clusters them in certain regions. They leave low population areas that swing faster in response to small demographic changes.

(OTOH... are internal migrants really less conservative? And what about Virginia as a case study? Is it any "less polarized" as a state since migrants turned it blue? Probably not. Some gaps and further research needed for this one...)

For a structural cause, poli sci friends of mine have talked a lot about the anti-pork thesis. Not sure the original source, but the claim is that increasing restrictions to prevent "pork" - handouts to congressional districts tacked on to other legislation - actually reduce the ability of parties to deal. It prevents you from peeling off a few moderates who could go home and claim, well, I voted with the other side, but brought home real benefits for constituents. Everybody hates pork in theory, but it actually a useful currency, it provided liquidity to the political system. Without it you just get gridlock and party line voting, because the tangible incentives to break have been cut down tremendously.

Or maybe we're just perceiving more "polarization" because of the pronounced gap between the median voter (50, no college degree), and the median social media user (20s, college)? So both overreact to the thing they can't control, the media / political environment? (I worry this comically understates the diversity of opinions in both politics and social media, so it's not really my favorite explanation.)

If you're still looking for answers, it might be worth contrasting Yglesias with "The Right Nation" from ca. 2000, from writers with different priors than Matt but also I think *trying* to be objective (and likewise failing in parts). The gaps between the two books might say more than either volume alone.

Expand full comment

I think there's an important missing piece here, which of course has happened in part in the last year, has been the polarization inside the Republican party. I know two people who used to hang out and complain about the Democrats. Now they hang out and the talk about anything other than politics, since one of them is a big fan of Trump and the other is a Never-Trumper Republican. Liz Cheney, the No.3 House Republican and daughter of former VP Dick Cheney was censured by her state's Republican Party for voting against Trump. It's of course still very much in the air what will happen with the party, but if it doesn't change, then something is going to have to budge.

Expand full comment

Just out of curiosity, what on earth does "Percentage of Overlapping Members First Dimension" mean?

Expand full comment

I feel inclined to make two notes, which aren't really related to each other:

1) The framing about Kansas Republicans and Kansas Democrats seems to miss that prior to a certain point, states tended to be really safe for one party or another -- there was an extremely stable system between the end of Reconstruction and the 1896 realignment, another really stable system (in which there was a Republican majority) until the New Deal, nearly uniform Democratic rule for about half a decade, and then another stable alignment from 1938-1952. When I read about Taft-Hartley and the rancor of the 1946/1948 elections, I don't see this taking place in a period of low polarization; looking at Senate and presidential races from this era, states tended to behave quite predictably (though the set of swing states was larger than it is now). Then Eisenhower comes in, banishes anti-New-Deal sentiment from the party, and you get a period of really intense depolarization -- by 1960 basically every state apart from the Deep South and northern New England is just 50/50 Nixon/Kennedy, like the nation -- and then polarization doesn't pick up again until the 1990s (defined as states preferring specific parties in a consistent way). The 1992 election -- well into your period of repolarization -- was the last one on record in which polling showed most Americans might've voted for more than one candidate (though this was because of Perot's presence), and Senate elections in the 1980s seemed to have patterns that verged on random, as the incumbents elected during the polarizing times of the 1940s had nearly all left by then. By 2004, there was a correlation again between the Senate races and the presidential one.

2) *very culture war content* Is the social justice movement really an identity-politics movement in the same way there are identity politics movements in countries with party list systems? In Israel, there is a party for Russian immigrants, parties for different kinds of Haredim, and so forth, which are strongly supported by those sectors, have negligible support outside of those sectors, and rather than having consistent positions just track public opinion from within those sectors. In the Netherlands, there is an identity politics party for Turkish immigrants (which gets votes almost entirely from Turkish immigrants) and another for hardcore Calvinists (which gets votes almost entirely from hardcore Calvinists).

Social justice doesn't seem to be like this; the minorities it claims to represent don't support referenda proposed by SJWs (like the affirmative action referendum in California), and often prefer candidates who are not very doctrinaire SJWs (like Biden). I'm not sure comparing it to more straightforward foreign identity politics movements is useful.

Expand full comment

There were a few significant events that happened between 2000 and 2005...

1. The Republican president and congress decided to start 2 wars, neither of which went well. That mortally wounded the "both parties are the same" lie in a very public and ongoing way.

2. Fox News came into its own as a "conservative" news outlet. On the other hand, RW AM radio (eg Rush) had been a fixture since the eighties, and I don't know how much extra scope Fox provided.

3. The Supreme Court had a number of decisions (eg Bush v Gore) that pushed Democratic activists further away from the Republican party

4. This was the start of the Gay Marriage issue ramping up, which took ~10% of the population and made one of their defining personal attributes polarized.

Expand full comment

>Why We're Polarized and the rest of Klein's oeuvre make a strong case that you would try to do something about polarization. Solve that, and a lot of the political pathologies of the past few decades disappear, and the country gets back on track.

But hold off on proposing solutions until you really understand the problem. It also possible that polarization is an effect of other things, and that treating those things would be better than trying to treat polarization directly.

Expand full comment

>my own model here is that the social justice movement pioneered a much angrier, more radical, more in-your-face style of identity politics

I would suggest that these "identity politics" weren't very radical at all and that this is a misrepresentation of fairly standard debates. Can you find any evidence that they were "angry" or "in-your-face"?

Expand full comment

There was an interesting FiveThirtyEight post a few days ago that was relevant to this: https://fivethirtyeight.com/features/why-a-biden-blowout-didnt-happen-and-why-a-2024-blowout-is-unlikely-too/

Basically, they pointed out that the elections from 1988-2020 bear a lot in common with the elections from 1876 to 1900, while the elections in between are quite different. 1876 to 1900 and 1988 to 2020 are two periods that are highly polarized, and have extremely close presidential elections (not one margin of 10 points or more), with the popular vote loser winning the electoral college twice each. 1904 to 1984 is a period that is not very polarized, and the majority of presidential elections have a margin of over 10 points (often with landslides for opposite parties not far apart).

I think that these periods can be refined a bit further if we think about the history of US party systems. It's traditional to say that we are now in the Sixth Party System (https://en.wikipedia.org/wiki/Sixth_Party_System, and links to the earlier ones). The first party system was 1796 to 1820 or so, Federalist vs Anti-Federalist. The second party system was 1828 to 1852 or so, Democrats vs Whigs. The third party system was 1860 to 1892 or so, with Reconstruction Republicans against Redeemer Democrats. The fourth party system was about 1896 to 1928, with Progressives in both parties, and a Populist third party. The fifth party system was about 1932 to 1968, with the New Deal system. The sixth party system was about 1972 to 2012, with the Republican Southern Strategy.

It seems that the third system and the sixth system were highly polarized, with very close elections, while the second, fourth, and fifth party systems were very anti-polarized - the two parties chose an obscure point of tariffs as their main differentiation, but otherwise existed only as meaningless electoral coalitions, that intentionally kept race issues out of Congress as long as they could.

Expand full comment

Introducing national propositions on issues, like some states now have, and like Switzerland, might reduce polarization by, for example, letting people oppose abortion but vote for Democrats, or raise the minimum wage but vote for Republicans.

Expand full comment

How does a discussion about the fall and rise of polarization over the court of the middle of the 20th century not mention the phrase, "New Deal Coalition", even once? It even shows up in your graphs: They all have spikes in the late 20's and they drop sharply in the late 70's. The coalition of ethnic whites(/organized labor/Catholics), liberals, blacks, and Southern whites created an electoral hegemony that, when it could stick together, was able to absolutely dominate the government from 1932 to 1976 (with fading but still present power until the 90's). This meant that the Republicans were thrust into an adapt or die situation, so they adapted by moderating towards acceptance of the New Deal vision of government (and, when they didn't, they died really hard, in 1964). There's a reason the Republicans did not control a unified Congress between 1956 and 1996 (and the short control in the 50's was super fluky and had to do with a super-popular Republican President's coat-tails in Eisenhower -- 1932 is really the correct year).

The New Deal Coalition collapsed for a lot of reasons. The cultural unification of the mass industrialization era declined over the course of the 1950's and 1960's, so by the 1970's it became possible for one party to be consistently tarred as the 'liberal' party. Nixon's 'moderate Republican' image was tied up much more with his stances on the New Deal government and much less with his cultural liberalism, because he pretty much had none. The Republicans could start biting into the ethnic white/Southern white portion of the Coalition which was becoming increasingly upset about the social and cultural change happening in the 1960's and 1970's -- yes, including the Civil Rights Act, busing, and other racial issues, but also including things like views about ostensible Democratic failures on the international stage (while anti-communism and foreign policy hawkishness was something both parties had strong wings in favor of in the post-war era, and Nixon was the one who pulled us out of Vietnam, the Democrats got associated with the anti-war movement in a lot of people's minds) and an increasingly libertine cultural approach (drugs, sex, and rock and roll). Jimmy Carter was the last Democratic President to be able to pull the New Deal Coalition together and he embodied these things so much to so many people that the Coalition collapsed around him in 1980.

There's a lot more to the story (the beginnings of the conservative movement go further back than the late 60's -- the rise of the religious right, too, which WAS a separate thing at first), but the broad picture is there: American politics was very underpolarized in the middle of the century because it had a hegemonic party. When the Conservative movement reached its apotheosis with Reagan, it actually failed to establish a similar hegemony, only barely managing to pull the culture towards themselves during the Clinton years (while Clinton triangulated well after the drubbing liberal Democrats had gotten in 1980, 1984, and 1988, he was not himself a conservative Democrat, although he sometimes played one on TV). What cultural pull the conservatives had managed to establish declined over the whole period and, I think, collapsed in the Bush years. They never managed to create the kind of cultural consensus created by the Great Depression, the war, and the New Deal and they never managed to create the electoral hegemony the Democrats had enjoyed in those years. So the country just kept slipping towards polarization.

Yuval Levin takes on this question in what at least seems to me from your description of Klein's work to be a better way in 'The Fractured Republic', focusing on the sense of cultural consensus and unity that was built up in the early 20th century and the way that it declined in the later 20th century. He doesn't dwell too hard on the details of the New Deal Coalition, either, but I think he does a better job of examining the underlying mechanics. Klein's focus on racism as the motive force of US history isn't insightful, it's just the standard party line for him and his now, cooked up so they can have a free hand to recreate American society in their image. The Dixiecrats and their motion through American politics were an important force, but they were one among many. Levin focuses more on the thesis that consensus, unity, and depolarization are not actually a natural thing, they were an aberration created by the relatively unique conditions of the early 20th century (mass industrialization, the rise of radio and television as highly centralized mediums, the mass impoverishment of the Great Depression and the mass mobilization of the Second World War, etc) and that the breakdown of this unity and consensus over the course of the later 20th century is a natural fragmentation of society. He, like Klein, doesn't necessarily think that polarization can or should be dealt with, but he does think we need to learn how to live and how to work with it. Build institutions that are capable of functioning in a social environment where polarization is common, such as in the late 19th century.

Post: I also have some quibbles about the timing of the Dixiecrat's inception and its role in depolarization. White Southerners were Democrats from the start, as soon as they started regaining the right to vote in the later 1860's and throughout what was left of Reconstruction. These 'Dixiecrats' survived through a century of drastically changing levels of polarization. Even the level of black suffrage changed mostly independently of polarization in this period: black suffrage in the South was substantially gone by 1900 but this is 20-30 years before any of your synthetic polarization indices show decreases. White supremacy had been re-established and became an article of political faith long before it became a stabilizing force in national politics.

Overall, I'm dis-satisfied with the level of actual historical discussion in this discussion of history. Polarization is a complex phenomenon, it takes actually digging into the details to understand it. A few synthetic indices aren't really going to tell you much on their own.

Expand full comment

Plenty of democracies are equally polarised, but (i) elections are fair and seen as such, since they are run by an independent election committee, and (ii) the winner can govern, since the prime minister is elected by parliament, so "winning" means winning both the legislature and the executive. The U.S. has (i) a throughly corruptible electoral process, and (ii) a byzantine governing system, which makes it likely that one party controls the executive and the other controls at least half the legislature. As long as politics was local, this peculiar system worked OK. Politicians who brought home the bacon could do whatever they wanted in Washington, and voters didn't know or care. But now, voters are better informed, so that U.S. politics has become nationally polarised - just as it has always been in smaller countries. But unlike these other countries, the U.S. system isn't built for these conditions.

Expand full comment

Does Ezra discuss that the parties united against the Cold War commies, but then came apart when they had no common enemy?

Expand full comment

> "Trump holds basically the same positions that Americans in the mainstream of either party would have held in a less polarized time (eg 1995); Clinton holds positions that everyone in 1995 (including her husband) would have thought insane, radical, and ultra-far-left."

I recently realized that many people wouldn't frame the issue this way. I was listening to, of all things, a Malcolm Gladwell podcast about civil rights lawyers in the 1950's, in which he offhandedly summarizes some of their early experiences as "they were losing". *Losing?* I thought. That didn't seem right; they were building support, raising awareness, and getting better results with each successive case. In hindsight they were already well on their way to near-total victory. Then I realized that Gladwell and I had different mental baselines. My baseline was a near-zero rate of change; anything above that baseline was a civil rights "victory" because it shifted the long-term status quo. Gladwell's baseline was a near-infinite rate of change; anything below that was a civil rights "defeat" because it allowed injustice to persist for that much longer.

The "who's changed more" analysis of political polarization depends on a similar discrepancy. In the absence of more sophisticated statistics (e.g. measuring skew in the distribution of political positions), claims about which side "has become more extreme" should be taken as claims about what the correct rate of change is. And such claims can't be separated from object-level political positions.

So for whatever it's worth, I'm not very convinced by this argument against Klein's framing. I disagree substantially with Klein both on how fast political consensus ought to change (slower!) and on whether its change since the 90's has been for the better (no!). Even so, I think it's clear that the correct rate of political change for societies like ours isn't zero, or even very close to zero (say, <5% of the rate that the median Democrat's position has changed). If nothing else, technological changes necessitate some political changes; also the status quo leaves, shall we say, plenty of room for improvement.

Expand full comment

My own view is that local-politics-as-part-of-daily-life vs national-politics-as-entertainment is the crux of this problem, not polarization. Politics has become a spectacle, where what's at stake is less about policy outcomes and more about symbolic gestures and abstract point-scoring in a reality TV game show. That's why both sides are less willing to compromise -- unlike policy outcomes, symbolic victories are zero-sum. Also, if the fight is about abstractions neither side can ever really win, so the same fights continue indefinitely.

This gets obscured because activists will often organize around policy issues, but often these are just totems in the symbolic fight. How effective has the Woke Left been at actually improving the lives of marginalized people? How effective has Trumpism been at restraining the Woke Left? If the people in either camp really cared about effectiveness, wouldn't they consider changing tactics? I see no signs of it; I see both sides doubling down on symbolic gestures, as if it's an end in itself, as if that's what politics is really all about.

Yes, of course empty symbolism has always been a part of politics. I'm saying it's come to occupy a much larger mind share in the last few decades, and I think this is 100% driven by changing media technologies & the incentives they create: Decontextualizing everything; forcing complex issues into simple, flattering narratives; pushing emotional buttons, etc. etc.

If this is happening at the same time as rising polarization it can look like polarization is driving the dysfunction & gridlock, but I think that's a red herring. The problem isn't that people disagree -- democratic republics are designed to handle that -- the problem is that activists on both sides have lost sight of instrumental goals & grown disconnected from reality.

Expand full comment

It is very late and I should not be awake, but I have ADHD and a crazy idea. Please excuse me for a moment as I start sounding like a conspiracy theorist and don't want to stay up even later tracking down sources.

I live in Texas. I grew up here even though I wasn't born here. I am also a Democrat. Yes we exist. The reason I'm a Democrat is also what I believe to be the source of why we got so polarized: Cracked.com.

Hear me out: there was a time that I was rooting for Rand Paul to win an election (can't remember what he was running for but I wanted him to win). This was around the time the Tea Party was picking up steam. I was fresh out of high school, I had burned a bunch of bridges, and was staying with relatives because my dad had kicked me out as soon as I graduated. I had nothing better to do than surf the internet all day everyday for about 6 months straight. I was about as republican as an uninformed Texas-dwelling high school graduate could get. But that all changed when (the fire nation attacked) I found Cracked.com.

At the time, Cracked was at its popularity peak. Millions of clicks on every article. The ones I want to draw particular attention to here though are the regular "true life experience" articles they ran. These were listicles "written" by people actually living through different extreme/unlikely/ horrific experiences: the guy who survived over a year in AlQueda custody, the lady who escaped from a cult, working prostitutes and other sex workers, homeless people, I think they got that guy who cut off his own arm that inspired the movie 127 hours, people who survived natural disasters and military insurrections, members of "weird" religions, people who worked behind the scenes on everything from television shows to politicians to sewage waste treatment plants etc.

Around the same time I was getting exposed to these people with heart wrenching tales of survival and oppression, Cracked's most popular writers were pulling further and further left. David Wong and John Cheese are the main writers I'm thinking of here: Wong was more abstractly political, but Cheese was the real converter.

Read his articles and you'll find stories of growing up in squalor with unhinged people, with a subtle, almost subliminal implication that these horrific people are only like this because they've been born, raised, and inculcated with the values and virtues necessary to survive while living in such a poor (in more ways than one) environment.

All of these taken together paint a picture that is violently against the conservative rhetoric of "individual responsibility", "disregulation", and "American exceptionalism".

Remember, Cracked at this time had readers in the millions everyday, and they were leaning more and more left as time went on. The only reason I don't still go to the site anymore is that they fired a bunch of really good writers which lead to two of my favorite video series getting canceled (After Hours and Some News if you're curious). They were able to get a young Texas republican to change parties in less than a year. Imagine what they did to people who were already on the fence.

Even though I don't still go there though the damage had been done: unless the Democrats do something to totally ruin the progress we've made, I'm a member for life.

Spider Jerusalem once said "Journalism is a gun with a single bullet, but if you aim it just right, you can blow a kneecap off the world". My (conspiracy) theory is that Cracked was in just the right place, at just the right time, to provide exposure therapy on a massive scale to the younger population

Expand full comment

Why Ireland is not as polarized as the US? Our elections use single transferable vote in multi-member constituencies. The argument that you have to vote for the Xs if you don't want the Ys to win doesn't work here - you can vote for the Zs, the Qs, the Ns... and finally, okay, last option on the ballot goes to the Xs, since they are at least better than the Ys.

The UK has a first-past-the-post system, where there are multiple parties but only two realistic options in most constituencies, and any government will be lead by either the Conservatives or the Labour party, and it is increasingly polarised.

Expand full comment

I think you cannot talk about polarisation in different countries without looking at voting systems, for that radically changes incentives.

Here in Germany, we do not have the Winner Takes All system of the US, so the incentives for political parties are totally different. If you have more than two parties representing an non-miniscule part of the population and are not in the situation whetre one party has enough votes to govern alone, demonization of the Others will not work, because you cannot really demonize someone and then form a coalition with them.

This is what happend in Germany, in my impression: We were pretty clearly polarized in the 70s and 80s, with the CDU representing the more affluent, more conservative, more christian, more nativist parts and the SPD representing the poorer, more liberal, more foreigner-friendly parts. The FDP were the enocomically liberals and made a coalition with one of them, contributing the 8-10 % needed to top the 50%. Between 61 and 98, the FDP was part of every government, and the other part was either SPD or CDU.

But then, times changed. The Greens entered the picture in the 80s, Die Linke in the 90s. Votes spilt between 5 parties meant smaller chunks for the big parties, and a need for more diverse coalitions. We had SPD/Green coalitions from 98 to 05, than a coalition of SPD and CDU, the big old enemies.

Why? Because if the CDU had stayed as conservative as it was in the Kohl years, it would have trouble to start a coalition with either SPD, Greens, Die Linke, and their own numbers have been decreasing since 83, when the Greens entered the picture. Merkel moved the party towards the middle, while Gerhard Schröder did the same with the SPD, so much so that it is hard for most to tell their politics apart.

But it is not a uniform trend. The rise of the radical ADF in the 2010s had the CSU (something like part of the CDU, it's complicated) try to move back to the old days of ultra-conservativism, only to then do a 180 and try to be like the Greens now. It's hard to predict a trend now.

That a radical party like the AFD could get 10+ % is a total novum for us, so the radicalization of the internet may well be a real factor driving parts of the electorate towards polarisation. It just isn't the only one.

Genereally, my impression is that the modern times offer more diverse identities, which in a voting system other than Winner Takes All will lead to the creation of more parties. These force the parties to be open to coalitions, which tends to move them towards the mainstream. At least, that is what I thing happened in Germany. In Israel for example, the story is a totally different one, from what I can tell.

Expand full comment

The graph about polarization in sweden feels like it ends right when it would have gotten interesting. In 2010, the sweden democrats (SD, conservative nationalists) got into parliament, and stayed there. Throughout basically their entire stay they have been met with exclusions (invite every party into the discussion, except SD; open to cooperation with every party, except SD and anyone open to cooperating with them), protests and various accusations and scandals about being nazis. And just last months I've seen news indicating them getting called pigs, misogynists, etc by prominent politicians.

Perhaps other countries aren't decreasing their polarization *now*, it's just we're a decade or two behind the US in terms of the rise of polarization?

Facebook was for sure a thing even in 2010, but I don't recall twitter being mainstream at that point.

Expand full comment

Scott is there not a certain tension between on the one hand the fact that you vote 'almost straight Democrat' by your own account, and your strong record of opposition to anyone else ever claiming that the Dems are better than the Repubs. Like, if the Dems are really no better why vote for them?

Expand full comment

That graph for polarization in Britain as a function of year is the clearest sine curve I've ever seen and it hurts my heart to see a straight line fitted to it.

Expand full comment

That graph of polarization in Britain as a function of year is the clearest sine curve I've ever seen and it hurts my heart that they fitted a straight line to it.

Expand full comment

Excellent post that I would not have known I needed to read until I read it. Thank you.

”A Republican writing a similar book might argue that although the Republicans have the advantage of being able to say "We have beliefs/characteristics X, Y, and Z", the Democrats have the advantage of being able to say "People with beliefs/characteristics X, Y, and Z are the enemy". And Klein has already shown that negative partisanship is more powerful than positive partisanship! Having a clearly defined set of people to be against can be more unifying than being anyone in particular yourself.“

Nicely summarizes this current political climate. Hate fighting hate? 🤔

Expand full comment

American polarization is not exactly the same as regular polarization as America is not ideologically polarized but socially so. The average establishment Dem and the average establishment Republican don't have very much ideological divergence especially when compared to the rest of the Western world where the overton window is considerably larger. American polarization really rests on historical animosity and America's obsession with relatively superficial things like identity politics.

Expand full comment

If you think polarization is a bad thing, which seems implied, then, surely, the only thing you can possibly do to reduce it, is to take a step towards the other pole.

Expand full comment

I think Roe vs Wade is a big part of the story.

There are huge numbers of single-issue voters who won't vote for a pro-choice (or pro-life for the other side) candidate. Back before RvW, the parties were less polarized on abortion but it's almost at the point where it's impossible to get chosen as a candidate for the Democrats if you are pro-life (and vice versa).

For large swathes of social conservatives, the only reason to vote for a Republican president is because they will nominate another conservative SCOTUS justice who will overturn Roe v Wade.

If your frame for evaluating a candidate is always "do they support abortion rights?" (I am moderating my language here) then all other considerations become rounding error. Other cultural issues (guns, prayer in school, SSM, trans rights) come along for the ride but abortion is the ur-issue that started to drive the parties apart.

It's worth considering this in the light of the constitution's guidance that "The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people". Many people on right see the abortion issue as not just a moral issue but as an attack on the Constitution.

If the power to regulate abortion were reserved to the states then the people of Kansas could make different judgments than the people of Hawaii and the politicians of Kansas (and Hawaii) could be held to account for those decisions. As it stands, the nationalization of this decision by RvW gives cover to politicians in Kansas as they make ever more extreme rules that they never have to account for locally because they know they will be struck down by SCOTUS.

I predict that, if RvW is ever overturned, the urgent need to vote for Republican presidents will be taken away and some of the polarization will fade.

Expand full comment

There are two reasons I can think of that the Tea Party has been more successful than the AOC wing: money and votes.

My mental model (which may be wrong) of the Tea Party is that they tend to be older and thus tend to vote more and tend to have more disposable income. My model of the AOC wing is that they are younger and thus less likely to vote and less likely to have disposable income.

Politicians want votes to get re-elected. And politicians want campaign contributions because they have expenses to run a campaign, and you can use the rest creatively for your own amusement.

Expand full comment

"(W)here nobody could tell you the difference between the two major parties (or, rather, they would wax rhapsodic about the parties' differing positions on a 1920s treaty, and then when you interrupted and said "no, I mean today", they would say "oh, no difference"). This was baffling to me. I would ask Irish people how they chose which party to vote for, and it would usually be something along the lines of "ah, we're a Fianna Fail family, always have been, always will be, someone from Fine Gael killed my great-grandpa during the war. Never bothered to look at either side's position on the issues, but still hope to get around to it one day."

Well, yes 😀 Though both parties which were centre-right (Fine Gael being slightly more to the right, see the Blueshirts, https://en.wikipedia.org/wiki/Blueshirts , our homegrown version of the Blackshirts and Brownshirts and Fianna Fáil having some Old IRA connections being slightly more economically left) moved even more to the centre to compete for the (growing) middle-class vote, where Labour joined them (and left the traditional working-class vote to be scooped up by Sinn Féin) and both went more for the 'business friendly, light-touch regulation' economically while moving socially more liberal.

Before that, the major differences were the economic/class ones - Fine Gael being seen as the party of the "strong farmer" and the incipient professional classes, aspiring to more respectability. Fianna Fáil being slightly more the worker's party, not as left-wing or urban working-class as Labour but the party of, say, the farm labourer as distinct from the farmer (my grandmother, who worshipped Dev, said "he made the farmers pay their workers a decent wage" and my grandfather, being a farm labourer, she always voted straight FF, reared her kids to do likewise, and at least my mother raised us to do the same).

In the 1941 poem by Donagh MacDonagh below, the "teacher, solicitor and bank clerk" would be the archetypal Fine Gael party members and voters:

Dublin made me and no little town

Donagh MacDonagh

Dublin made me and no little town

With the country closing in on its streets

The cattle walking proudly on its pavements

The jobbers, the gombeenmen and the cheats

Devouring the fair-day between them

A public-house to half a hundred men

And the teacher, the solicitor and the bank-clerk

In the hotel bar drinking for ten.

Dublin made me, not the secret poteen still

The raw and hungry hills of the West

The lean road flung over profitless bog

Where only a snipe could nest

Where the sea takes its tithe of every boat.

Bawneen and currach have no allegiance of mine,

Nor the cute self-deceiving talkers of the South

Who look to the East for a sign.

The soft and dreary midlands with their tame canals

Wallow between sea and sea, remote from adventure

And Northward a far and fortified province

Crouches under the lash of arid censure.

I disclaim all fertile meadows, all tilled land

The evil that grows from it and the good,

But the Dublin of old statutes, this arrogant city

Stirs proudly and secretly in my blood.

Expand full comment

I think Klein misses something very important, and to conservatives, very obvious about "vetocracy." It ties in so well with his overall point that I can only assume he misses it because at heart he's a Democrat who cares as much about the Democrats winning (getting their, and his, preferred policy) as he does making polarization better.

If the Democrats (or Republicans) want to pass a bad law, and do not want to pass a different law that is a good law, then the Republicans (or Democrats) are better off making sure nothing at all passes instead of what their opponents want. Mitch McConnell is a sometimes hero of the Republicans because he has been able to sit on the Senate and stop the Democrats from putting any of their stupid (to the Republicans) policies into place. For instance, look at the $15 minimum wage discussion now. Democrats are saying $15. Republicans are against $15. If the Democrats asked for ~$9, I bet the Republicans would vote that in, and quickly. But they're not asking for $9 ("good law," compromise), they're asking for $15 ("bad law," no compromise). The Republicans have made it abundantly clear that $15 is too high. And recall, Republicans are more often from rural areas where $15 is a LOT higher than it is in a big city. A few years ago San Francisco's equivalent of $17/hour was Alabama's $11. A $15 minimum wouldn't affect SF very much, but would destroy Alabama's economy. If the Democrats decide to push forward with $15 instead of a compromise, what possible reason would there be for Republicans to pass the law? They should refuse to pass the $15 minimum wage, and defeat it in any way possible.

This isn't purely a Republicans Refuse To Pass Good Laws issue, as there are times Republicans want to pass a law and it's defeated by the Democrats. That said, it is much more likely that Republicans want to keep things as is, rather than changing. It's probably part of the core reason we call them conservatives. That shouldn't be surprising. Democrats more often want to change things, try new things, etc., than Republicans.

Expand full comment

As a democrat, I feel like one of the big sources of polarization is gun freedom/control. At least it's used as a wedge issue and my impression is that there is a sizeable chunk of the populace that will always vote for the person most in favor of a broad view of the second amendment. Even though I support basic gun control for the same reason I support drivers licenses, I'd really encourage my party to just drop the issue entirely. For two reasons:

1) Though gun violence is problematic, it's exaggerated and any gun control that had the slightest chance of passing would have minimal impact on such a thing. We've made 0 progress in my lifetime, maybe it's time to give up?

2) Can't we trade this for support in another realm? I would gladly take single-payer healthcare in exchange for completely dropping gun control as an issue. Single-payer healthcare is actually fairly popular so this isn't totally an implausible trade if we thought dropping gun control nets a 5pp gain in the polls.

Yet I feel like my party will continue aggressively supporting gun control precisely because it's the opposite of the other party. If someone were looking to deescalate political polarization, gun control would be where I'd start. I'd also be curious if anyone with more contact with pro-gun rights people could comment on how likely any would be to change their voting if gun rights weren't an partisan issue.

Expand full comment

Some additional factors to consider:

First is the effects of a series of crises - the Great Depression and WWII, followed by the Cold War. This produced a generation where it was the norm for partisanship to be subsumed by more pressing concerns. I don't think it's a coincidence that the growth in partisanship among the public really took off in the 1990's when the Silent and greatest generation were dying out coinciding with the end of the Cold War. I think younger people (I'm 52) like Klein don't fully understand how the Cold War and the ideological battle against Communism generally influenced American politics more generally by forcing a consensus that bled over into other areas and was led by two generations of people that thought in those terms.

The second factor is the growth in power and authority (perceived and real) of the federal government as compared to state and local government. When issues become nationalized and are adjudicated at the federal level, then that raises the stakes. This also applies to the courts. A good example of this is Roe v. Wade which unintentionally created the binary division over abortion we see today and prevented the healthier kinds of political debates on the topic seen in other countries. Even Justice Ginsberg, in her earlier years, had reservations about Roe:

"Suppose the Court had stopped there, rightly declaring unconstitutional the most extreme brand of law in the nation, and had not gone on, as the Court did in Roe, to fashion a regime blanketing the subject, a set of rules that displaced virtually every state law then in force. Would there have been the twenty-year controversy we have witnessed, reflected most recently in the Supreme Court’s splintered decision in Planned Parenthood v. Casey? A less encompassing Roe, one that merely struck down the extreme Texas law and went no further on that day, I believe and will summarize why, might have served to reduce rather than to fuel controversy."

And abortion has been one of the major catalysts for ideological sorting today.

Expand full comment

>And the Dem presidential nomination went to Joe Biden, a moderate who wouldn't look out of place running for president in 1988 (in fact...).

1988-edition Joe Biden wouldn't recognize 2021-edition Joe Biden, and not just due to some extra wrinkles and less hair. 2021-Biden would not even approach being a 1988-moderate.

Your cheap joke devalues your analysis and glosses over the differences that mark 33 years of changes in the party as reflected by that blank canvas of a politician.

This should not be taken as saying people can't "evolve their views," either; just that doing so requires actual change, and if he has actually changed your joke is moot.

Expand full comment

No story about American polarization can really be started without discussing the urban / rural divide. People like Klein always think that big ideas shape history, but they don't. Trends and forces do, and most trends and forces come down to expressions of human behavior and their access to technology. Our least politically polarized time was shortly after the founding, when a town had to literally meet in a church or a barn and hash out their politics, most political decisions didn't extend beyond the decision makers, and most meetings were done in groups under Dunbar's number. The counter argument to this usually starts with federal/anti-federal stuff, but that was the vast minority of political decisions being made. Most where on the town/county level and never made it into the history books because it just was an agreement to use 4 rails on fences instead of 3 because Goodman Yoder's cow kept getting into Mrs. Hancock's flowers.

Now, we extract above-average academic achievers from their home communities (while making academics more reliant on social capital), place them in an environment with a ~4 year churn of the population, move them to be in charge of lots of stuff they don't have skin in the game on, and wonder why we don't have the cultural institutions that keep bad things from happening.

In the mean time, we concentrate the academic underachievers (aka low human capitol) out in the hinterlands (rural or urban) and wonder why depression, crime, and drug use is rampant and why social institutions like churches and mutual-aid groups like unions fail.

Does anyone here know anyone from high school who a smart go getter that turned down college to chose to start a small blue collar business in their rural or ghetto home town and contribute to the human capitol of the community?

Expand full comment

I think this is missing a structural explanation: we're polarized because the American system is structurally set up to favor two parties and no more. This is an accidental by-product of the constitutional requirement that the president be the majority choice of the electoral college (if no one wins a majority, the vote is thrown to the house of representatives, which no one wants). This gives ordinary voters strong incentives to choose one of only the two most prominent choices. Parties can therefore only define themselves on one principal axis, and other possible axes are after thoughts, and only two parties can at any given time emerge as a source of plausible nominees for president. Third parties are doomed to irrelevance for presidential elections.

Now, this need not determine which parties seat members of congress, and indeed sometimes third parties do win some seats, but the gravitational pull of the big contest for president drags the rest of the federal contests into alignment with this same two-party system, creating effective federal duopoly.

At the state and local level it's possible to move somewhat further away from this two-party trap, which is why you see many more parties fielding plausible candidates for local positions. But the whole system is still tugged towards duopoly as what seems like an unintended consequence of the electoral majority requirement in the constitution.

Now, the two parties square off and define themselves in opposition to each other. There are no third parties to join in coalitions to help elect candidates, so the kinds of compromise that coalitions can build are impossible. This need not cause polarization, but it seems very likely to be a factor intensifying it.

Expand full comment

Increasing globalization in the 1970s and the rise of the internet/IT in the late 1990s and early 2000s have slowed income growth among people without some level of post-high school education. I think a more useful dichotomy is between people who are doing well relative to the median vs people who aren't. This also seems to correlate better in time. Your review makes it seem like Klein focuses a lot on racial dichotomy (white vs not white), and while I think this is a factor, I don't think it's the primary one.

Expand full comment

The colors of the links on substack are lighter than the standard link color. For me, that causes them to blend into the background and makes reading harder. Any chance we could get the color to be a bit darker/more saturated? (There are some web accessibility standards which provide a good guide on which colors should be used on a white background)

Expand full comment

Regarding the reason why the coalitional makeup of the Democrats would make them more structurally resistant to ideological polarization than the Republicans - I haven't read the book, but I have listened to Ezra Klein talk about this on podcasts, and the point he made there made a lot of sense to me. It goes basically as follows -

The Democratic party cannot become as ideologically polarized as the Republican party, because the racial coalition aspect also translates into having a lot of conservatives. There are a very large number of black, hispanic, and other non-white voters who are conservatives by inclination, but stick to the Democratic party because they (rightly or wrongly) see the Republicans as the Racism Party.

If you're a black conservative, your choice is between the Democrats, who may be more left-wing than you like, and the Republicans, who are basically still seen as the party of the Dixiecrats. Again, whether or not the reader agrees with this idea, it is clearly one accepted by the great majority of black voters in the US. And we see similar patterns across many other groups who make up the Democratic coalition.

But this means that as a result, the Democratic voter base is not merely more racially diverse, but also ideologically more diverse. You need to make enough room in the party not to completely drive out voters like this, and thus the reason the Democratic party has remained more centrist overall than the Republican party, which doesn't have any equivalent group of liberal voters who are "forced" into the party.

Expand full comment

Could there be a tradeoff between polarization and corruption?

In Paraguay when you ask someone how they choose which person to vote for they often answer that they'll vote for so-and-so because they promised some inane, selfish, short-term material benefit.

People here vote for the person who promises them a public sector job (a position that will open up after the candidate wins and fires all the supporters of the other party), or the one that will help them cut the line to get free healthcare for their parents in the overloaded public healthcare system, or even just the one who throws the best parties.

This has many disadvantages that the United States doesn't have and cannot afford to have. For example it leads to a public sector that is full of lazy incompetent bootlickers. But it does have an advantage: when the population is poor and concrete material conditions are at stake in every election, there is little room for abstract ideological considerations.

Some concrete examples of what I mean:

1) I know some pro-choice feminists who voted for a literal Catholic priest for president, just because his concrete proposal to spend more money on hospitals in rural areas was more important for most women than an abstract agreement on the moral question of abortion.

2) Said Catholic priest won an election by allying himself with free market liberals and literal communists. Nobody here thought this was weird at all. After all, the point of elections is to win and appoint your supporters to well-paid positions, not to advance some meaningless ideological agenda.

3) I know someone who has, in different elections, voted for a corrupt fascist party, a corrupt technocratic neoliberal party, a social democrat party, a Trump-style populist party and a corrupt agrarian socialist party. I know this because they voted for the same party every time, and that party changes ideology depending on who is their most charismatic member at the time.

Expand full comment

"Every so often, people ask what an effective altruism of politics would look like".

That actually sounds really interesting! Can anyone point to any ideas or discussions around this?

Expand full comment

Given the recent themes in your blog about about journalism, I'm surprised you didn't discuss that material at all. What did you think of it?

Expand full comment

I tends towards these explanations that aren't mine, but I feel are heavily under-discussed:

1. Two party system: not like the UK with 2 big parties, but really the US seems unique in only having 2 real options for leadership and president. This may have worked temporarily, but it seems obvious to me that human nature will turn this situation in two warring sides each considering the other to be the ultimate enemy. Most other democracies have smaller parties with enough power to keep the big ones from going too extreme.

2. End of the cold war: the out group used to be a scary foreign country that threatened to overtake and destroy America. Such an external threat simply doesn't exist anymore. It seems natural to me that everyone needs an outgroup and now that wars between superpowers is a thing of the past, people will find their outgroup in local politics.

Really eager to hear criticism of these explanations!

Expand full comment

Well, this really misses the point.

The very notion of representative democracy and a free country relies on a system that is widely trusted to be structurally fair. Things like the peaceful transfer of power rely on extremely large majorities believing that it is better to be a loser under a fair system, than to be a winner in an unfair system, where you have only won because you successfully deposed the other side in a coup. In the US at the moment:

* Nobody believes that the US system is structurally fair

* Nobody apparently seems to believe that a structurally fair system is even possible

* Nobody is even trying to construct or promote a system on the basis of its structural fairness

In the absence of faith in the system, what you end up with is everybody voting and acting for personal gain. This includes trying to reconstruct the system in ways that favor you. This manifests as the kind of extremely political polarization that we're currently living through.

It was not always thus: the American Dream was the shining vision of the systemic economic fairness of the United States. MLK's vision of a colorblind society was a vision of a structurally fair social system.

The notion that this is about the Republicans swinging right while the Democrats stay where they are is frankly laughable. If anything, the Republican leadership clings hopelessly to the notion that the American system is structurally sound, even as the economic unfairness of globalization has totally discredited that notion with their voting base, hence Trump. The Democrats meanwhile have turned social and legal unfairness into their calling cards, spent four years discrediting the 2016 election result and whipped up a mob who want to burn the whole thing down and replace it with ... what, exactly? This is not particularly clear, but my guess would be some kind of system where "fairness" is given out as political patronage. This is normally the case when a system is substantially designed by angry left-wing protest groups who want it to favor minority interests. It is hard to find historical examples where this has been attempted without significant downside!

I can't really see any of this getting better any time soon. Redesigning political systems is like trying to change the wings of an airplane while it is flying. Doing it in the current environment is like trying to change the wings while flying through a storm.

Expand full comment
AnonymousFeb 10, 2021

I am pretty convinced that the recent, perceptually sharp, increase in polarization is due in large part to the abolition of congressional earmarks.

There are lots of articles about this — here's one to start:


"While pork-barrel spending, in many cases, can undoubtedly undermine voter confidence in government by making them associate government with corruption, its role as a possible curative for hyperpolarization seems to make it a net positive practice for preserving American democracy.

Since many scholars argue that the number one goal of all politicians is to win re-election, removing incentives that assist that goal would consequently reduce the chances of politicians being willing to compromise.

Therefore, when pork-barrel/earmark spending, a type of spending designed specifically for incumbents to boast of their legislative achievements on behalf of their constituents, is removed, politicians must turn to other methods to win and maintain the favor of their constituents to protect them from both primary and general election challenges.

Since politicians can no longer appeal to their constituents’ material interests, they are driven to appeal to their social and ideological interests."

Expand full comment

"Ezra Klein is great." This opening salvo felt like a thesis. I was hoping in the course of reading to adjust my opinion of Ezra upwards. But instead I feel most deflated... Ezra Klein's "Why we can't Build Anymore" a quick article of his outperforms his full length book. The concluding sentence felt like praise, but was it?

Expand full comment

My folk understanding of this was that it was an outgrowth of the McGovern-Fraser commission making party nominations much more small-d democratic. This lead to candidates more directly chosen by primary voters, leading to greater ideological sorting of politicians, leading to greater ideological sorting of voters, leading to more extreme candidates winning primaries, etc. This seems to dovetail with stuff like Cory Gardner losing his reelection campaign by 10 points after completely failing to hew to the center: he was terrified that doing so would get him primaried. The timing doesn't seem like a perfect fit for this story, but it's not terrible, and because this story is one about a snowballing feedback effect it has the virtue (vice, really) of not committing to a tightly constrained timeline.

Expand full comment

Have there been any studies on comparing how effective are different ways of lowering polarization? Or how cost-effective?

Expand full comment

Thankfully we Irish have figured out both parties are identical, and replaced the main competion with a random batch of murderers.

Expand full comment

An interesting question in polarization: how many times does the average punter play a game of Three Card Monte before deciding that the game is rigged against them and checking out.

Similarly, how many times can a person be on the losing side of an election before deciding that democracy is rigged against them and becoming an insurrectionist/revolutionary?

It is entirely possible that any form of stable centrism is invariably unstable, since it typically leads to a stable group being excluded from decision making and therefore steady attrition in democratic participation. Once you get to the point where you have a similar number of people who have lost faith in the system as are voting for the winner, you're almost invariably going to have something happening like the Capitol insurrection.

Maybe a lack of political swings with most candidates looking basically the same should be a major red flag of a system about to go off the rails.

Expand full comment

> they don't seem to become much more polarized until 2000-something

The chart showing the the snapshots of public's ideological distribution in 1994, 2004, and 2015 leaves out such an interesting chunk of time. There's an animated version on this page, https://www.pewresearch.org/fact-tank/2014/06/12/7-things-to-know-about-polarization-in-america/ . Watching that, I see a significant shift right among Republicans from 2004-2012 (while Democrats basically stay put), and then both both parties shooting apart from 2012 forward.

I wonder what was happening in each of those periods?

Expand full comment

My observation, and Scott hints at it, is that this short essay from the NYT identifies the most important cause of polarization https://www.nytimes.com/2016/11/20/opinion/sunday/the-end-of-identity-liberalism.html?_r=0 The reason seems pretty obvious: the current national obsession with identify politics within US large corporations and cultural institutions naturally forces people to follow one of their most basic instincts, which is to take sides with their side.

Expand full comment

> Along with the nationalization of politics came a nationalization of news. In the old days, you either read the city paper or watched your one local news channel. Since these papers/channels had both Democratic and Republican readers, they would try pretty hard to avoid offending either group, and had some incentive towards objective journalism.

As far as I can tell, in many other countries, most of the media have always been national, with much of it having various partisan alignments. The American situation where, for a long time, the mainstream media at least tried to be neutral, was somewhat unusual.

This could explain that other countries have long been as polarized as the US is now: their media have always been national and partisan, like the US media are now. I'm not claiming that this is actually the explanation: I have no idea if it is, or if the political polarization charts are even meaningful (especially when compared between different countries).

Expand full comment

Always felt that a lot of polarization could be traced back to efforts to get more people involved in politics. Back when I was growing up, a sizeable proportion of the population were ambivalent. A constant refrain (which I was dubious about) was "how do we get more people involved in politics?".

Of course the way you do this is simple. You make it entertaining, and you make it angry. This approach has been so successful that it's hard to read about sport without mostly consuming politics. You could argue that the politics is now secondary, as America's politics serves as the globe's entertainment.

Expand full comment

There's a name for the effect where the title followed by the author's name spells out something funny. It's called 'The Man Who Melted Jack Dann'.


Expand full comment

Regarding vetocracy and infrastructure: "American litigiousness developed specifically in the 1970s – it’s exactly how what the authors of the paper call citizen voice is enforced. In contrast, on this side of the Channel, and to some extent even generally on this side of the Pond, laws are enforced by regulators, tripartite labor-business-government meetings, ombudsmen, or street protests. French riotousness is legendary, but its ability to systematically change infrastructure is limited, since rioting imposes a real cost on the activist, namely the risk of arrest and backlash; in contrast, it is impossible to retaliate against people who launch frivolous lawsuits."

From https://pedestrianobservations.com/2019/07/15/costs-are-rising-us-highway-edition/

The "citizen voice" part is later expanded in an article of its own ("surplus extraction").

Expand full comment

I'm confused about the claim that low levels of polarization are abnormal. Here's a toy model which explains my confusion. Let's say we have a two-party system which differs only on one dimension (e.g. liberal-conservative). Naively, you'd expect the party platforms to be as close as possible, so that they might capture as much of the electoral middle as possible. They'd try to differ just enough so that voters could distinguish them, and not more than that.

This doesn't seem to be the case with Republicans and Democrats. So where does the model go wrong? Well, one way it can break down is if the electorate is polarized and people too far from the center become disaffected and stop voting for "their" parties. My sense was that something like this happened in the 1960s through 1970s - voters were polarized on civil rights before the parties tackled the issue, and 60s counterculture was another (related) source of grassroots divisions. (I got this picture mostly from Bill Schneider's "Standoff: How America Became Ungovernable." Schneider also claims that it's the fact that the US is a particularly religious country that has caused polarization to keep increasing past the 60s - the conservative backlash to the 60s counterculture has been much stronger.)

I don't really see how party polarization could ever happen *before* mass polarization. I mean, if we tweak the toy model to reflect the fact that parties aren't one-dimensional, there's another route for polarization: party platforms might move to extremes on issues that most people don't care about either way, and some of the base cares about a lot. Then the public will take its cue from the party and polarize as a result. Something like this probably happened on some issues (gay rights maybe?), but given that the US parties have been getting increasingly unidimensional [https://legacy.voteview.com/political_polarization_2015.htm "a single dimension accounts for about 93 percent of roll call voting choices in the 114th House and Senate "], this just doesn't seem to be most of the story.

So I guess given the lack of data on mass polarization before the 90s, I would expect that normal people *were* issue-polarized before 2000... though maybe not as affectively polarized?

Does my reasoning check out? I'd love to see holes poked in the toy model.

Expand full comment

Curent polarization is no mystery. Starting in 2000 Democrats explicitly embraced the strategy set out in "The Emerging Democratic Majority," which was basically to exploit immigration to create a dominant non-white majority.

Then, starting around 2012 (as documented by Jonathan Haight), American Universities basically lost their minds and starting pumping the "woke" ideology that everything that ails society is "whiteness" and "white privilege" and the invisible force of "systemic racism."

Because the MSM is downstream from the university monoculture, they became lockstep propaganda outfits dedicated to propagating the new religion. The essence of wokism is not debate or compromise but morally demonizing your opponents as immoral "racists," "Misogynists," "homophobes," etc. The more you do this, the more morally superior you are to your fellow un-woke Americans, whom you should look down on and despise. The NYT, CNN, NPR, are now dumpster fires of hate directed at Trump voters, Republicans and White people generally.

In short, politics has been transformed from concrete debates about policy -- taxes, spending, healthcare -- to a never ending moral outrage. It is this aspect of moral condemnation that makes the subjects "resentful." For what it's worth, the Democrats initiated this, Republicans and Trumpists are just reacting.

Oh yeah, and Klein is one of the bad people at the front leading this process with the execrable Vox. So it's no surprise that he can't get it. He's a victim of his own propaganda. He can't see anything but virtuous leftists and their bigoted opposition.

In other words, Klein can't diagnose the problem because he is the problem.

Expand full comment

Democrats and Republicans weren't exactly the same in the early 20th century. Traditionally Democrats were the party of unions, and labor rights whereas Republicans were the party of business interests. That has been fairly constant, however, as of 2016 Republicans seem to be making inroads with union workers, even as they continue to oppose and dismantle labor laws.

Expand full comment

My comment is re: republicans representing "modal" American. I believe that is entirely wrong, as the Trump coalition brought together arm-in-arm Orthodox Jews with Holocaust fans, hispanics with people who thought they are rapists, women with incels, asian immigrants with those who hate immigrants, etc, etc. It's almost beautiful, kinda. The glue that held them all together was a mutual sense of "hey man, on the maga train whatever you want to believe, we won't judge". So whether it's QAnon, falun gong, religious devotion, anti religious devotion, hoping to finish the Holocaust and eliminate the zionists, hoping to empower the zionists, refugees from totalitarian states wanting to make a totalitarian state because it's anti totalitarian (the kind they don't like)... None of this matters, it's just a big tent of broadly anti-liberal individuals and groups.

A final point: the craziest thing on the 2020 campaign was this: Trump ads in Spanish with people waving Mexican and cuban flags. And broadly winning them.

Expand full comment

It makes sense for US and Canada to be more affected by trends towards national media, since these are much bigger countries. I'm not sure when Sweden started having national television, but I'm guessing it happened at the same time we got television.

Expand full comment

I feel like I've spent most of my adult life surrounded by people claiming we're experiencing increasing polarization and not believing it. I guess that's in part because I tend to take a very long view of history and we live in a country that has historically experienced tremendous internal violence, whereas my own lifetime has been steadily increasing internal peace, marked largely by the decrease in crime since the 90s. That has nothing to do with political polarization, but I remember going to school in a place where we weren't allowed to wear red or blue because kids were getting shot in the streets by whoever was wearing the other color, and whatever hysteria comes from the fact that every now and again kids in the suburbs shoot up their schools, the actual danger of being a person in America is so historically low that whining about government gridlock just makes me sigh. If we can't get optimal tax or housing policy and healthcare costs too much, I guess that sucks, but it's a heck of a lot better than my memories of the news every single night talking about some little girl getting shot in gang crossfire trying to buy ice cream from the ice cream man. That seems like cognitive bias on first glance, but actual statistics bear out that life has gotten much better in the past 30 years by any reasonable measure and that is even more the case globally. Much of Africa and South Asia was very nearly an unlivable hellhole in the lifetimes of people still alive and that is largely not the case any more.

But I've probably been convinced at this point. Even your own comment section has a good 5% or so of people who seem like they wouldn't be out of place on any of the boards I frequent for sports fans, insisting that the other team is everything wrong with the sport, seemingly forgetting that free agency is a thing and the guys they love today they hated just a few years ago. Things have gone tribal, in spite of the fact that I don't buy that Americans really have distinct enough social groups to warrant calling them tribes. I've lived in seven states, born in California, currently in Texas, lived in rural Kentucky and absolute bum fuck Appalachia in North Carolina, and I just don't see it. Talk to people for five minutes and they're the same damn people. But when all they know of each other is the news and the Internet, they seem so easily convinced they're on opposite sides of a grandiose struggle for the soul of the future that will determine whether our great nation stands or falls. That hasn't been true since 1870.

The reasons seem obvious as just-so stories. The news, politicians, addiction-based ad tech companies posing as public squares all have tremendous incentive to incite outrage. But why now and why here more than comparable countries? Do we simply interact with each other less? Is it car culture and unwalkable suburbs? We just aren't physically near enough to actually talk to people as might be more encouraged by the basic literal structure of European cities? Weaker national identities thanks to not having nationality inherently tied in with ethnicity and religion as in many European countries? No near-peer external existential threat after the fall of the Soviet Union forcing us to look internally for the enemy? Or is it really just regression to the mean after a near century of cooperation created by the historical accidents of Dixiecrats and the New Deal Coalition that are now gone? Is comparing the entire United States to any single European country just a flawed comparison? What does the comparison look like if we consider polarization between different parts of Europe? England doesn't seem thrilled with the continent. Germany doesn't seem to be getting along swimmingly with Greece. Russia just invaded Ukraine a few years ago. Is polarization specifically along political party lines really different or are we just turning to the only distinction we can plausibly conceive of with people who can make enough of a local impact for us to care?

Expand full comment

I think there's a potential reason why the internet and social media might have played a larger role in polarizing the US, and it's that the US is just more important than, say, Germany. US politics has become a sort of global focal point for political debate. Nobody cares which German party you support, but if you, even as a non American, support the Republicans you can be ostracized by your peers, kicked from discord servers, cancelled on Twitter, etc.

I'm not sure which direction the causality flows, but I'm pretty confident that the internet has been uniquely accelerating of polarization in the US.

Expand full comment

No mention of 1972 Roe v Wade??? Semi-socialist sympathizing Christians which were long time family Democrats, but believe abortion is wrong - such folk have been essentially exiled from the Dems. And now make up a huge part of the Reps, along with their desires for more (good!) gov't against the older smaller gov't Reps.

Dem control of colleges has also created an elite that is NOT allowed to denigrate others, but it's OK, even encouraged, to denigrate and demonize Republicans.

And Democrats have always loved to Hate the evil ones - so they falsely claim Reps are evil, whenever there is mere policy disagreement.

Racism is now the focus because accusations, including false accusations, can be so damaging to any career; and Dems have little problem in lying about Reps (look at Dan Rather's lies about Bush).

Reps being hated by Dems, especially elitist Dems acting "superior", cause a hate-based reaction.

The "Political Correctness" that has been gaining strength is now more like PC-nazis.

Expand full comment

Viewing it as a Canadian, I’d say it went something like this. Bush v Gore was widely perceived then and since as being a partisan decision. 9/11 was kind of mixed, but featured lots of infringements of civil rights and things like that. Then there was the Iraq War, which featured lots of “with us or against us” rhetoric, then turned out to have been based on misinformation.

Getting all flustered about Bush was, IMO, legitimate. But then nobody calmed the hell down, in part because it'd been eight years of normalizing people being angry and extreme about these things. And it's just kept building from there.

Expand full comment

"He refers to data 'showing Congressional Republicans have moved further left than Democrats have moved right', which I think is a typo (isn't the usual argument that Republicans have moved further right than Democrats have moved left?)"

I listened to the audiobook not long ago, and I believe there this section says "showing Congressional Democrats have moved further left than Republicans have moved right," acknowledging a point you make later in your post. It's a bit challenging to verify that in an audiobook--can anyone else corroborate?

Expand full comment

Long time lurker, almost never commentator here. I"m surprised you were confused by why polarization started to get worse among the general public in the US around 2005-ish. The answer is social media.

Expand full comment

My working theory for what changed around 2000-2004 in US politics is that, prior to that time, we did not identify things as "red states" or "blue states":

> In the days following the 2000 election, whose outcome was unclear for some time after election day, major media outlets began conforming to the same color scheme because the electoral map was continually in view, and conformity made for easy and instant viewer comprehension. On election night that year, there was no coordinated effort to code Democratic states blue and Republican states red; the association gradually emerged. Partly as a result of this eventual and near-universal color-coding, the terms "red states" and "blue states" entered popular use in the weeks following the 2000 presidential election.


This identification gave rise to a bunch of effects, including feeling like a member of the outgroup if you were "blue" aligned in a "red" state and vice versa (where previously those identities may not have mattered as much in day-to-day life) and also color-coded maps of the US that lent themselves to mistaking land area for political representation (which also activated identities of aggrievement or assailment, depending).

Expand full comment

There is an interesting question here about Ireland, because it has a much more polarised politics now. The conventional mainstream media story is "sex scandals in the Roman Catholic Church and the Global Financial Crisis". I think this misses at least one other big factor (peace in Northern Ireland that has enabled Sinn Féin to legitimise themselves).

The two traditional parties (Fine Gael and Fianna Fáil) have both declined in numbers and eventually got forced into forming a coalition with each other - while Sinn Féin now operates as a somewhat left-wing opposition party to the merged neoliberal government.

This poses an alternative way that American politics could have polarised, which is the rise of a third party that would have pushed two neoliberal parties together.

But, well, honestly, I would say that Klein is answering the wrong question. The right question is: given that the US is polarised, how do you design a political system that doesn't completely fall apart. The unique thing about the USA is that it's a political system that enforces a strong form of Duverger's Law (ie pushes to two parties very hard) and also requires cross-party working for effective governance. Other countries are either two-party (Australia, UK) but have a political system that puts one party in power and means they never need votes from the other, or require cross-party co-operation in building a coalition, but are multi-party, and therefore the co-operating parties are still not all parties (Ireland where there's a FF/FG government and an SF opposition, Germany, Netherlands, etc).

[Note that Sinn Féin originates from taking a third position on the 1920s treaty, but that position was so unpopular (it was "restart the war with Britain") that they essentially ceased to exist until 1969 when they were revived by, uh, restarting the war with Brtain (in Northern Ireland) - this means that they are not associated with the treaty/civil war by most Irish people, unlike FF and FG]

Expand full comment

I think it's likely that American polarization has structural roots. The U.S. is an unusual country in its combination of first past the post voting and strong presidentialism. Most of the world's most successful democracies (going by ratings from Freedom House or from the EIU Democracy Index) have proportional representation and parliamentary systems. In those systems, the head of state is usually a monarch or an indirectly elected figurehead president, and the head of government (prime minister) is also elected by the parliament. The upper chamber, if it exists, is also usually elected indirectly, and usually is much less powerful than the lower chamber. All this indirect election may seem less democratic superficially, but it arguably has much more democratic outcomes.

In the U.S., very little can be accomplished legislatively unless one party holds the presidency, the House, and 60% of the Senate seats OR there is significant bipartisan cooperation. Once enough polarization has happened that bipartisan cooperation is unworkable, the incentive structure encourages hyper-partisanship, because the only viable path to achieving any political goal in such a system is total domination.

Many structural reforms of varying levels of viability would alter the incentive structure, and would probably make polarization decrease: abolishing the filibuster in the Senate (more parliamentarism), any form of proportional representation in the House such as Lee Drutman's plan to emulate the Irish system that Ezra Klein mentions in the book (less FPTP), reducing the power of the president and/or of the Senate (more parliamentarism), alternative voting systems like approval voting or range voting or ranked choice voting (less FPTP), or alternative executive structures such as delegating some executive powers to the House or replacing the Presidency with a Swiss-style executive council (more parliamentarism).

Why is Canada not polarized? I think it has to do with its level of parliamentarism. Canada has FPTP voting, but, like the Finns in their proportional democracy, Canadians vote only for their district's parliamentary representative rather than directly electing the Most Powerful Person in the Land Who Represents the Soul of the Nation. That's less polarizing. After they vote, a governing majority or coalition forms, which has the power to legislate and get things done instead of just screaming at each other all the time. That's also less polarizing. Seemingly as a result, Canada's main political parties are moderate, political discourse is more about ideas than in the U.S., and voters are less sorted ideologically. Some voters who identity as liberals vote for the Conservative party and some voters who identify as conservatives vote for the Green party. See this paper, which shows this in surprising charts:


I think reducing polarization via structural reforms is an important cause area for rationalists and effective altruists to focus on. Polarization, or at least its U.S. manifestation, is extremely wasteful of resources (think of the hundreds of millions spent on Senate campaigns this year). It makes political discourse become about demographics and geography and whether people like you feel like they're winning or losing, instead of being about policy and ideas and how to actually make the world a better place. It makes it really hard to get anything done even if it's really important (like stopping climate change). And – I can't back this up with data – but it seems like it's got to be damaging to public health by causing everyone to feel constantly anxious (which, in turn, distracts people from thinking about things that really matter – things like Kelsey Piper's latest Vox article on the history of smallpox eradication and its implications for today). Oh yeah, and it might destroy liberal democracy, leading to autocratic government (which is what usually happened when Latin American countries tried to copy the U.S. system).

As to why polarization is happening so much NOW, I don't know the answer, but I suspect it might have something to do with rising income inequality. See this article (which also discusses some of Peter Turchin's ideas):


Expand full comment

It feels important to say that whatever the case was for the Republican party being different from the Democratic party before the 2020 election, the case now is overwhelming.

Heaps of top Republicans parroted Trump's baseless lies about election fraud. Trump and his faithful encouraged and organized a mob to interrupt the election's certification. Perhaps most concerning to me, 147 members of congress voted not to certify the election results. It's hard to keep track of all the other attempts to obstruct the electoral process along the way, but suffice it to say this wasn't nearly the half of it.

This is all a million miles from the Bernie or AOC contingents, and it's not the least bit fringe: just about the entire Republican party is involved, including almost every one of its biggest names. Let's not let this one go unremembered. Whatever the republican party was in the recent past, a faction of anti-democratic opportunists have since taken the reins, and their influence looks to have fully metastasized.

(In case this sounds like the sort of thing only a leftist partisan could believe, here's a conservative-born publication with hundreds of articles making the same point: https://thebulwark.com/)

Expand full comment

So, you recognize that "polarization" can be based on past conflicts or death. You know we had a Civil War in the US over slavery. You know the South defected to the GOP right after the 1964 Civil Rights Act, having supported FDR for an unprecedented four terms as President when he ran on repealing Prohibition, expanding the role of government in helping people, and allying with Communists if need be to protect American ideals. Yet you badmouth the idea that they're just racist or following race-based tribalism.

Plainly, the Fox News Channel, which is a propaganda organization created by Roger Ailes when he left "America's Talking", deliberately made race-based tribalism respectable again. Everyone on the Left who had elderly relatives not on the hard Left reports them going far-right when the cancer known as Fox News got up and running. My late grandfather told me we should put gays in concentration camps, and that was one of the milder changes people have reported.

I also recall your own graph showing that support for immigrants was growing, among Democrats and Republicans, before the pestilent rot called the FNC began. At that point support tanked among Democrats, and stagnated for Republicans. Then, when Democrats realized the FNC narrative was a pile of poisonous lies, support went up to where it might have been naturally. Republicans stayed in the gutter.

I see people still pushing falsehoods about how this happened, so let's be very clear. Roger Ailes was an operative in the Nixon administration who proposed they create GOP TV, in order to counter the media's bias against treasonous criminals. Later, he ran the network that became MSNBC. It was called America's Talking, and it had a talk show hosted by Roger Ailes, as well as a segment purporting to showcase failures of government. That wasn't enough propaganda for Ailes, which is why he left to realize his long-standing dream of founding GOP TV. None of this is secret.

You also rather oddly say that nothing was happening re:race from 2000-2005. Ignoring the whole Iraq War, and everything on Fox News, you also either fail to grasp why Michael Bloomberg had zero chance of winning the Democratic nomination for President, or you choose to ignore the way he <a href="https://www.youtube.com/watch?v=sFivw-dB4gA&ab_channel=TheDailyShowwithTrevorNoah">outright lied</a> about facts he later confessed to, on the record.

Expand full comment

I think it's unfair to not mention the (IMO accurate) view that the senate is biased in terms of political representation in the context of the DC and PR statehood arguments. The democratic party picked Biden over Bernie et al in part because it needs to win more than 50% of the vote to win elections. The Republican party has become a lot Trumpier over the past 4 years due to not suffering a penalty for nominating a candidate that lost by 2 million votes. (I know I'm equivocation between the EC and senate here but I think similar arguments hold). I think "reduce electoral bias in favor of Republicans" is a highly plausible antipolarization mechanism and you don't give it a fair reading here.

Expand full comment

The primitive brain Is wired to place more importance on avoiding poisonous berries... primitive people shared danger knowledge in order to survive... The nightly news taps into this primitive fear response.. captivating audiences = ratings = add revenue. Political parties tap into fear and hate captivating, polarizing = growing your base.... the attention economy exploits the primitive brain...whaddyagonnadooo?!

Expand full comment

I wrote about Ezra's failure on this and nailed it 7 years ago. He promptly blocked me:



Game theory + US system (game rules) = the more power DC gains, the more polarized we become.

The US political system requires "MOVE!" Choose one of 50 sovereign flavors. Americans are the genetic elite of moving. This is the game the founders invented. States compete in a political free market for common currency and labor.

The only hack to the supremacy of the US system people who hate their homeplace so much, they do not MOVE! someplace else, they move to DC and make all states the same, and try to manage the polarization.

Ezra knows this theory and refuses to confront it directly.


Expand full comment