Winning

A couple weeks ago I was at a bar in London with a rather literary set. I think I had the second-fewest Substack subscribers of anyone there. That’s how you know you’re hanging out with the cool kids.

“Well, Yarvin’s just not a very impressive speaker,” One of my new friends was telling me about his perspective on various New Right thinkers. “He was early on that whole anti-democracy aesthetic, but there’s really nothing there. Not much going on with him.”

I agreed. I don’t think so highly of Yarvin.

“Steve Sailer, too, I mean he tries to make it all scientific but he’s still an edgelord at heart. Nothing going on there either. ”

Again, I agreed. I told a brief story about Steve Sailer being half-Jewish, but this wasn’t news to my interlocutor.

“So, who do you think really gets it? Who’s writing the most important stuff?”

I kind of froze. I didn’t have a hip, esoteric answer. Eventually I offered:

“Well, this certainly isn’t the cutting edge or anything, but I like what they’re doing at The Argument.”

The political pundits I read usually aren’t the most radical. The Argument is a new publication started over the summer by Jerusalem Demsas that makes a positive argument for abundance liberalism. They have a polling arm and are unabashed in their attempts to help steer the Democratic party towards victory in the next election. I believe that this is important work.

Now, I don’t blame anyone else for wanting something more exciting. There’s a natural emotional resonance with big ideas, and even from a purely utilitarian perspective there are times when radically different conceptions of our society and our politics are the most productive thing to discuss. The right new framework can be immensely valuable.

I just think that right now, when considering electoral politics it’s more important to get the current systems right than to conceive of new systems. And what matters most in politics in 2025 is that Democrats win elections. The Argument is aligned with this goal, but upon further reflection, the name I should have given is that of David Shor. Shor champions something that’s been called “popularism,” which is the idea that in electoral politics we should consciously optimize on winning elections by taking popular positions.

Shor1 is a data scientist who has worked to try to elect Democrats for his entire professional career. When he was just 20 years old, he was overseeing political forecasting for Obama’s 2012 campaign. Next he joined Civis Analytics, a data science firm where he continued to work on political causes as their Head of Political Data Science. In the summer of 2020 he posted a link to research showing that violent protests hurt Democrats electorally while nonviolent ones helped; this drew fierce backlash on social media, and Shor lost his job at Civis.

He quickly bounced back, founding his own company. Shor’s Blue Rose Research is now the leading progressive data analytics firm, and they played a prominent role in Kamala Harris’ campaign.

There’s a tension between his personal politics and his prescribed policies for the campaign trail. Shor himself is an avowed socialist with a red rose in his X bio, and yet he pushes campaigns to take moderate positions on issues where the electorate is to the right of the Democratic party.

Typically, cultural issues are where the Democratic party is far to the left of the median voter. Left-economic populism, on the other hand, sometimes polls very well; for instance, wealth taxes are popular, as are expansions to social security and Medicare. Shor believes that when the public disagrees with progressives, it’s the critical for progressives to be disciplined in decreasing the salience of those issues so that they play less of a role in a voter’s decision at the ballot box.

The other side of this argument is that we can win by staking out positions that are currently unpopular because they can generate new energy and shift the opinions of the electorate. There’s an excellent Politico article from 2022 that breaks down these dynamics by presenting two different cocktail parties, one hosted by David Shor and the popularists and the other representing the Democratic party’s progressive, activist flank.

The other approach, termed “inclusive populism” in the article, can sometimes work. Gay marriage is a notable example of an issue where the party’s left flank championed an unpopular position and the electorate eventually came around. But that was a period when leading Democratic politicians often publicly took moderate positions on cultural issues—Obama notably did not support gay marriage in his 2008 campaign—and these days we have fairly robust evidence that, ceteris paribus, moderates perform better in national politics. The median voter theorem might just work.

Popularism doesn’t mean that we can never take positions that the electorate disagrees with, but it does mean that we conceive of such things in terms of trade-offs. Sometimes the trade-off is worth it! But often you should place a very large premium on winning.

This works in practice because voters don’t closely follow the details of governance on most issues. They care about cultural alignment and economic vibes, but they’re not tracking industrial policy or AI regulation. This means Democrats can win by sounding less weird on cultural issues while maintaining the latitude to govern competently on the complex questions that will actually determine America’s future.

Let me start with where I’m coming from. I want an America that embraces its melting pot heritage by welcoming immigrants who want to work hard from all over the world. I want an America that provides a robust social safety net to people who would otherwise struggle during a difficult period in their life. I want an America where people are free to build and structure their families as they choose.

Frankly, for these reasons I would have wanted Democrats to beat Republicans, in both 1984 and 2024. But even if you don’t agree with these values—if you think that immigration is damaging our social fabric, or that redistribution is too distortionary, or that abortion is unethical—you should probably still want Democrats to win.

The two parties just face very different incentive structures. Our politics has evolved such that pretty much all of the crazy voters are now Republicans. It wasn’t always this way; it used to be that, for instance, anti-vaxxers showed up at low levels across both parties. But over the past decade, educated professionals have gravitated toward Democrats while generic suspicion of institutions has become much more common among Republicans. This created a self-reinforcing cycle: as educated people moved left, the GOP became more welcoming to institutional skeptics, which pushed even more educated people away.

This has been termed the “crank realignment” by Matt Yglesias. In a post-“crank realignment” world, Republicans face terrible incentive structures. Because their voters often believe crazy things, they end up appointing people who either believe crazy ideas or feel pressured to support them. The nomination of RFK to head HHS was an olive branch to the crank contingent, and now we have top government officials talking about how autism rates are higher in circumcised children because their mom was taking Tylenol.

Republican voters also venerate Trump, so he doesn’t get held accountable.2 The Republican party today is more or less the party of Trump, and so if Trump gets a bad idea in his head we all have to deal with the consequences. The intra-party system of checks and balances in policy generation has gone away; now when Trump decides he wants tariffs we get tariffs, even if there’s no electoral base for it and every economic analysis says it’s a terrible idea. He also gets to do things like demand the Justice Department pay him $230 million without any electoral consequences or real pushback from within the GOP.

The Republican party is also open to influence from deep-pocketed contingents; Peter Thiel and his allies have taken over the party’s tech policy, and now even Marjorie Taylor Greene is saying that Trump is prioritizing tech industry insiders over regular people. Crypto billionaire Justin Sun was charged by the SEC with fraud in 2023; he proceeded to invest $75 million in Trump’s crypto project, with most of that going to Trump’s family directly, and when Trump took office the SEC halted its prosecution. We also see this pattern with executives like Paul Walczak and Trevor Milton and Changpeng Zhao, and SEC enforcement has gone down across the board, especially in crypto. Again, there’s no pushback here from the Republican base when the country is sold out like this.

The Democratic base, by contrast, holds its leaders accountable. Thanks to that increased education polarization I mentioned, Democrats are much more likely to be college educated and much more likely to be highly engaged with politics. Engaged voters are more likely to respond to poor behavior from elected officials, creating natural accountability mechanisms: Democratic politicians who engage in serious misconduct face consequences. After he was accused of many different instances of sexual harassment, Andrew Cuomo was pushed out from his position as Governor of New York, and he won’t be elected Mayor of New York City either.

The Democratic party does face a different challenge. In the wake of the 2024 election, some prominent liberal pundits (Matt Yglesias, Ezra Klein, Adam Jentleson) wrote explicitly about “the groups,” a nebulous set of progressive non-profits that have powerful voices in Washington and push Democratic lawmakers to take unpopular positions. Because they are well-organized and face no counter-pressure within the coalition, politicians acquiesce. This is what prevents Democrats from taking positions more popular with the electorate: a fear of rocking the boat with this class of activists.

But a crucial difference is that unlike with the Republican cranks, the voters that are aligned with “the groups” make up a small minority of the electorate, although they are unusually visible on the internet: especially platforms like BlueSky. Thanks to this, Democratic leaders don’t actually need to listen to “the groups.” As Yglesias has argued, it’s incumbent on leaders like Chuck Schumer and Hakeem Jeffries to have the discipline to turn down “the groups” and campaign on policy that aligns more closely with the electorate.

Encouragingly, some Democrats are recognizing this, such as the aforementioned Adam Jentleson. Jentleson was actually featured in that Politico article on popularism—but he was one of the organizers of the other cocktail party, the one standing against popularism. Yet last month he launched the Searchlight Institute, which we could call an “anti-group.” It’s a think tank that has the explicit goal of curbing the influence of the other “groups” in order to get the party to take positions aligned with the electorate. He’s had a change of heart. I can only hope others come around.

So even with the influence of “the groups,” Democrats have far better epistemics than Republicans and are far less corruptible. Democrats’ problems are fixable; they’re about political courage rather than deep-rooted structural issues.

And why does it matter that our government makes rational choices instead of anti-rational ones? Well, I believe that these are very important times. The next decade will require state coordination on a scale we haven’t seen since World War II.

We need to ensure we don’t create misaligned superintelligent AI. We need to protect Americans from what Xi Jinping has called “spiritual opium,” the addictive digital products designed to capture attention that degrade both cognition and social trust. And we need America to remain a credible global leader standing against the rising tide of nationalist, zero-sum populism worldwide.3

But let me focus on one challenge that illustrates why governance quality matters: the need for America to maintain its position as a superpower in the face of China’s rapid technological development.

We’re in a race with China on AI development, where competent regulation and industrial policy could be the difference between American hegemony and a Chinese century. We’re dangerously dependent on Chinese exports of the rare earth metals that are essential for our military and economy. China has already shown willingness to restrict these materials as leverage in negotiations, and their ability to flex that muscle in a prolonged conflict should scare us.

Thanks to NVIDIA, we still have a significant lead over China in AI chips. Depending on the pace and magnitude of AI development, this may be all that matters. We also have a steadfast ally in Taiwan, which produces most of the world’s advanced semiconductors. But Taiwan’s position is precarious. We need to develop domestic semiconductor production to ensure we’re not vulnerable if China moves on the island and America isn’t able or willing to stop them.

Free trade has been enormously beneficial globally, but our geopolitical environment now requires targeted industrial policy, strategic investments in manufacturing capacity, and regulation that serves national interests.

This requires exactly the kind of governance Democrats are (relatively) better positioned to deliver: evidence-based policymaking, resistance to regulatory capture and corporate interests, and the ability to make long-term investments in unglamorous but critical infrastructure. Republicans, with their crank base and veneration of Trump, simply cannot deliver this. Their incentive structures make competent governance nearly impossible.

This is why popularism matters. Not because voters are always right or because we should slavishly follow every poll. But because in 2025, winning elections is how we ensure competent governance at a moment when incompetence could be catastrophic. For now, I care most about putting the right people in charge so that our country can handle the challenges ahead. That means Democrats need to win, and that means rolling up our sleeves and doing some popularism.