How To Win An Election

Image Source

Eric Levitz has a thought provoking interview with David Shor up over at New York Magazine. Shor is a electoral whiz kid who seems to have been making waves in the world of liberal polling for some years, but only came to national prominence a few months ago when he was fired from his gig at Civis Analytics for tweeting that non-violent protests were more likely to shift American public opinion leftward than riots.

I don’t share Shor’s politics, but in this interview the whiz lives up to his reputation. His take on political behavior is fascinating. The general mode of analysis he lays out has applications that extend far beyond the topics covered in the interview. A good place to start is with his understanding of the non-partisan voter:

 The single biggest way that highly educated people who follow politics closely are different from everyone else is that we have much more ideological coherence in our views. If you decided to create a survey scorecard, where on every single issue — choice, guns, unions, health care, etc. — you gave people one point for choosing the more liberal of two policy options, and then had 1,000 Americans fill it out, you would find that Democratic elected officials are to the left of 90 to 95 percent of people. And the reason is that while voters may have more left-wing views than Joe Biden on a few issues, they don’t have the same consistency across their views. There are like tons of pro-life people who want higher taxes, etc. There’s a paper by the political scientist David Broockman that made this point really famous — that “moderate” voters don’t have moderate views, just ideologically inconsistent ones.

Some people responded to media coverage of that paper by saying, “Oh, people are just answering these surveys randomly, issues don’t matter.” But that’s not actually what the paper showed. In a separate section, they tested the relevance of issues by presenting voters with hypothetical candidate matchups — here’s a politician running on this position, and another politician running on the opposite — and they found that issue congruence was actually very important for predicting who people voted for. So this suggests there’s a big mass of voters who agree with us on some issues, and disagree with us on others. And whenever we talk about a given issue, that increases the extent to which voters will cast their ballots on the basis of that issue.

Mitt Romney and Donald Trump agreed on basically every issue, as did Barack Obama and Hillary Clinton. And yet, a bunch of people changed their votes. And the reason that happened was because the salience of various issues changed. Both sides talked a lot more about immigration, and because of that, correlation between preferences on immigration and which candidate people voted for went up. In 2012, both sides talked about health care. In 2016, they didn’t. And so the correlation between views on health care and which candidate people voted for went down.

 So this means that every time you open your mouth, you have this complex optimization problem where what you say gains you some voters and loses you other voters. But this is actually cool because campaigns have a lot of control over what issues they talk about.[1]

 But wait, readers of the Stage might ask, how does this square with the pattern of politics I have discussed in posts like “Public Opinion in Authoritarian States,” “Reason is For Stabbing,” and “Chinese Are Partisan Too?” Those essays describe public opinion as a creature of coalition instincts; where coalition leaders go, opinion soon follows. Shor instead talks about public opinion as a more immutable thing that American electioneers can appeal to but cannot change.

 Shor thinks the two views can be reconciled:

 It’s worth being precise about mechanisms. It’s true that political parties have enormous control over the views of their partisans. There’s like 20 percent of the electorate that trusts Democratic elites tremendously. And they will turn their views on a dime if the party tells them to. So this is how you can get Abolish ICE to go from a 10 percent issue to a 30 percent issue. If you’re an ideological activist, that’s a powerful force. If you convince strong partisans to adopt your view, then when the party comes to power, strong partisans will ultimately make up that administration and then you can make policy progress.

  The problem is that swing voters don’t trust either party. So if you get Democrats to embrace Abolish ICE, that won’t get moderate-ish, racist white people to support it; it will just turn them into Republicans. So that’s the trade-off. When you embrace unpopular things, you become more unpopular with marginal voters, but also get a fairly large segment of the public to change its views. And the latter can sometimes produce long-term change.

….I don’t think anyone ever says something like, “I think it was a good trade for us to lose the presidency because we raised the salience of this issue.” That’s not generally what people want. They don’t want to make an unpopular issue go from 7 percent to 30 percent support. They want something like what happened with gay marriage or marijuana legalization, where you take an issue that is 30 percent and then it goes to 70 percent. And if you look at the history of those things, it’s kind of clear that campaigns didn’t do that.

If you look at long-term trends in support for gay marriage, it began linearly increasing, year over year, starting in the late 1980s. But then, right when the issue increased in salience during the 2004 campaign, it suddenly became partisan, and support declined. After it stopped being a campaign issue, support returned to trend. Campaigns just can’t effect those kinds of long-term changes. They can direct information to partisans who trust them, and they can curry favor with marginal voters by signaling agreement with them on issues. But there isn’t much space for changing marginal voters’ minds. [2]

 The way to win an election then is to figure out what demographics are necessary for electoral victory, determine which issue sets are important to those demographics, further determine which of these issues meld most easily with your partisan program, and then do everything in your power to make the election a referendum on those broad topics. 

Important to Shor’s discussion is the idea that voters are not evaluating specific policy proposals so much as determining how much they trust parties to handle the general issue of concern:

One way to think about electoral salience and the effects of raising the salience of given issues, is to look at which party voters trust on a given issue, not just what their stated policy preference is. So if you do a poll on universal background checks for guns, you’ll find that they’re super-popular. But then, politicians who run on background checks often lose. In the same way, if you poll comprehensive immigration reform, it’s super-popular, even among Republicans. But then Republicans can run on anti-immigrant platforms and win. So how do you square that circle?

One way is to remember that these polls give us a very limited informational environment. You just throw people a sentence-length idea, which they’ve often never heard of before, and then people react to it. So it tells you how people will respond to a policy at first brush without any partisan context. But ultimately, when people hear from both sides, they’re gonna revert to some kind of partisan baseline. But there’s not a nihilism there; it’s not just that Democratic-leaning voters will adopt the Democratic position or Republican-leaning ones will automatically adopt the Republican one. Persuadable voters trust the parties on different issues.

And there’s a pretty basic pattern — both here and in other countries — in which voters view center-left parties as empathetic. Center-left parties care about the environment, lowering poverty, improving race relations. And then, you know, center-right parties are seen as more “serious,” or more like the stern dad figure or something. They do better on getting the economy going or lowering unemployment or taxes or crime or immigration.

If you look at how this breaks down in the U.S. — Gallup did something on this in 2017, and I’m sure the numbers haven’t changed that much since then — you see that same basic story. But there’s an interesting twist. One thing that Democrats consistently get rated highly on is improving race relations. And this points to the complexities of racial resentment. The way that racially charged issues generally get brought up in the U.S. is in the context of crime, which is a very Republican-loaded issue (in terms of which party the median voter trusts on it). Or it comes up in terms of immigration, which is itself a Republican-loaded issue. So even if voters acknowledge the massive systemic inequities that exist in the U.S., discussion of them normally happens in a context where conservatives can posit a trade-off with safety, or all these other things people trust Republicans on.[3]

You can clearly see this dynamic being played out in real time right now. Consider yesterday’s report in the New York Times, “How Chaos in Kenosha is Already Swaying Some Voters in Wisconsin:”

John Geraghty, a 41-year-old worker in a tractor factory, has barely paid attention to the presidential race or the conventions. Every day he focuses on survival: getting his son to sports practice, working at his job where he now wears a mask, and getting home to sleep, only to start over again the next day.

But when he woke up on Monday morning to images of his hometown, Kenosha, Wis., in flames, he could not stop watching. The unrest in faraway places like Portland, Ore., and Minneapolis had arrived at his doorstep, after a white police officer on Sunday shot a Black man in the back multiple times. And after feeling “100 percent on the fence” about which candidates he will vote for in November, he is increasingly nervous that Democratic state leaders seem unable to contain the spiraling crisis….

Don Biehn, 62, owner of a flooring company, was standing in line at a gun store on Tuesday afternoon. He said that he had never bought a pistol before, but that he had a business to protect. A former county board supervisor, Mr. Biehn said he had been calling county and state officials for days, trying to explain how grave the situation was.

Neither John Antaramian, the mayor of Kenosha, nor Jim Kreuser, the county executive, responded to requests for comment. (The positions are nonpartisan, but both men previously served in the State Assembly as Democrats.)

“There’s people running all over with guns — it’s like some Wild West town,” Mr. Biehn said. “We are just waiting here like sitting ducks waiting to get picked off.”

He added: “It’s chaos — everybody is afraid.”

Mr. Trump, he said, “was not my man,” but now he is grateful he is president.[4]

 What is happening here is exactly what Shor describes: for voters in Kenosha, the issue salience of “law and order” is spiking, and Republicans are the party they trust on that issue. Pointing this dynamic out a few months ago was what got Shor fired. He recognized that violent protests make the general issue of race-relations salient in a way unfavorable to the left. In contrast

What’s powerful about nonviolent protest — and particularly nonviolent protest that incurs a disproportionate response from the police — is that it can shift the conversation, in a really visceral way, into the part of this issue space that benefits Democrats and the center left. Which is the pursuit of equality, social justice, fairness — these Democratic-loaded concepts — without the trade-off of crime or public safety. [5]

This broad issue salience framework helps make sense of other electoral puzzles as well. Consider the problem of politicians who vote against measures that poll popular with their constituents:

 David Broockman showed in a recent paper — and I’ve seen this in internal data — that people who give money to Democrats are more economically left wing than Democrats overall. And the more money people give, the more economically left wing they are. These are obviously the non-transactional donors. But people underestimate the extent to which the non-transactional money is now all of the money. 

This wasn’t true ten years ago. So then you get to the question: Why do so many moderate Democrats vote for center-right policies that don’t even poll well [even when their donor base is more leftist than their constituents]? Why did Heidi Heitkamp vote to deregulate banks in 2018, when the median voter in North Dakota doesn’t want looser regulations on banks? But the thing is, while that median voter doesn’t want to deregulate banks, that voter doesn’t want a senator who is bad for business in North Dakota. And so if the North Dakota business community signals that it doesn’t like Heidi Heitkamp, that’s really bad for Heidi Heitkamp, because business has a lot of cultural power. I think that’s a very straightforward, almost Marxist view of power: Rich people have disproportionate cultural influence. So business does pull the party right. But it does so more through the mechanism of using its cultural power to influence public opinion, not through donations to campaigns. [6]

There are a few take-aways one can glean from all of this. The first is a tidy explanation of why voters rarely punish candidates for failing to live up to their campaign promises. If you understand statements like “build the wall” as a specific policy proposal, then Trump’s failure to do just that might seem like it would hurt his prospects with the anti-immigration crowd. But the reality is that voters do not make their votes on the basis of individual policies, but a general idea of whether the candidate can be trusted to handle the issue before him. Statements like “build the wall” are not so much policy commitments as electoral signposts designed to communicate the reliability of the candidate on the issue in question (in this case the message being, “this guy can be trusted to take immigration seriously”).

I encourage readers to take this framework and apply it outside of the American context. From this perspective, the recent elections in Taiwan were always the DPP’s to lose. Given what was happening in Hong Kong, the most salient issue was always going to be the cross-strait relations, and for a decade now this is an issue Taiwanese voters have trusted the DPP with more than they have trusted the KMT. Absent an economic crisis large enough to change this framing, the KMT was always going to struggle to compete.

An interesting question, which I can only speculate on, is whether this same framework applies to authoritarian states where competition between alternating partisan blocs is banned. My first take at the question is that it does. There are some issue sets that the Chinese populace does not have much confidence in the CPC’s ability to manage well (environmental issues and pollution being perhaps the most obvious). Over the last few months American leaders have been careful to draw a distinction between the Party and the people in an attempt to put additional pressure on Chairman Xi and his team. But if they have done this while simultaneously centering Sino-American conflict on issues whose issue salience is favorable for the Party, they may end up increasing Chinese confidence in the Chairman.

 [1] Eric Levitz, “David Shor’s Unified Theory of American Politics,” New York Magazine (17 July 2020)

[2] ibid.

[3] ibid.

[4] Sabrina Tavernise and Ellen Almer Durston,“How Chaos in Kenosha is Already Swaying Some Voters in Wisconsin,” New York Times (26 August 2020).

 [5] Levitz, “David Shor’s Unified Theory of American Politics,”

[6] ibid.