Advertisement

SKIP ADVERTISEMENT

What Do Swing Voters Think? Meet @American__Voter

Only 23 percent of adult citizens are either all liberal or all conservative across seven major issues.

A Trump rally last month in Battle Creek, Mich. Voter behavior is more nuanced than it may appear.Credit...Pete Marovich for The New York Times

Alexander Agadjanian and

Mr. Agadjanian is a researcher at M.I.T. and Mr. Morris is a writer for The Economist.

Imagine you’ve scooted into a red booth in an unfussy local diner somewhere in Michigan, not unlike those portrayed in the numerous articles reporters have dispatched from the Midwest since the 2016 election. One booth over, you’re overhearing a middle-aged white man talk about his politics with a buddy of his.

You find out over the course of your meal that he’s a moderate Democrat who wants to keep Obamacare protections in place and opposes concealed-carry, but who also supports mandatory minimum sentencing and favors deporting illegal immigrants. He also happens to mention that he voted for Donald Trump.

This sort of conflicted, “cross-pressured” voter often appears in vigorous debates over swing voters in quasi-hypothetical terms. However, we know from the Cooperative Congressional Election Study (C.C.E.S.), a nationwide scientific survey, that this aforementioned voter in Michigan is a very real, living, breathing man, who was among the roughly 65,000 Americans asked about their identities, policy preferences and voting behavior by the study.

As researchers, we saw back in 2017, near the start of the Trump presidency, how the voices in online discourse belied the ideological idiosyncrasies of the actual electorate. So we decided to shoehorn that reality right into the belly of the discourse beast by creating a Twitter account called the American Voter Bot, which turns the raw C.C.E.S. data into neat, tweet-length profiles of real-life individual voters’ stances on issues. (We posted the code for this on GitHub for anyone to check out.)

We looked at just seven salient questions the C.C.E.S. asked adults in 2016: Whether they supported concealed-carry permits, the deportation of illegal immigrants, the circumstances under which they support abortion, the strengthening of the Clean Air Act, mandatory minimum sentences for nonviolent drug offenders, the Affordable Care Act and raising the minimum wage.

Those With Mixed Views Outnumber the Purists

More than three-quarters of American adults have a mix of conservative and liberal views on seven key issues. Only 23 percent of adult citizens are either all liberal or all conservative by this measure. The seven issues are: concealed-carry permits, deportation of illegal immigrants, abortion, the Clean Air Act, mandatory minimum sentences for non-violent drug offenders, the Affordable Care Act, and raising the minimum wage.

Of 7 policies, citizen

has liberal views on:

Percentage of all

American adults:

17%

14

14

13

13

12

11

6

All 7

6 of 7

5 of 7

4 of 7

3 of 7

2 of 7

1 of 7

None

Of 7 policies, citizen

has liberal views on:

Percentage of all

American adults:

All of them

6 of 7

5 of 7

4 of 7

3 of 7

2 of 7

1 of 7

None of them

17%

14

14

13

13

12

11

6

By The New York Times | Source: analysis of 2016 Cooperative Congressional Election Study by Alexander Agadjanian and G. Elliott Morris

Only 17 percent of those who responded to the survey held liberal positions across all seven issues and only 6 percent were conservative across the board, meaning that just 1 in 4 Americans had the perfectly consistent policy views held by each party’s most visible politicians and some of the biggest voices online.

Our automated bot works by selecting four policy preferences at random for each anonymized voter, who are identified by their age, education, race, gender, home state and 2016 vote choice. Then the bot tweets one profile out with this information every hour, allowing the Democratic electorate that dominates Twitter the chance to e-meet the voters who may be outside of their online bubble. (Among the bot’s followers are many “Twitter famous” journalists based in elite coastal institutions.)

The bot introduces you to some interesting characters. Take the Floridian white man in his 40s who recently sprouted onto our feed. He says he’s conservative, but doesn’t feel like he fits with either party. He thinks women should always have the right to choose when to have an abortion, is pro-Obamacare, and wants the federal minimum wage to be $12 an hour. But curiously enough, he’s a Trump voter. Why? Well, he opposes the Clean Air Act — might he have a job in a fossil fuel industry and feel threatened by Democrats’ quick embrace of radical environmental policy in light of climate change?

Of course, it’s not possible to know through a survey. But just by getting a peek at a voter beyond his one ballot can inspire a much more humble reading of the electorate’s behavior.

You don’t need to look much further than the latest far-left-leaning Democratic debate to see how uniform policy preferences have become in our two-party system. The candidates’ reluctance to reject purity tests, as well as their attempts to avoid “getting ratioed” by their loud online base, often leads them to well-intentioned yet hard-line rhetoric and stances that are unrepresentative of even many Democratic voters.

As that party’s primary rages on this year, we may need reminders of people like the 59-year-old white woman from Massachusetts we came across who voted for Hillary Clinton three years ago but is a self-described conservative who doesn’t think women should always have a right to an abortion. Could a zinger at a town hall that puts down pro-life arguments earn a candidate plaudits online and on MSNBC, but make others like her feel ignored or vilified? Even for the unwaveringly pro-choice, it’s a politically handy question to keep in mind.

Over the course of a day, our bot introduced followers to a white moderate Mississippian who opposes Obamacare and supports the Clean Air Act, a 20-year-old non-college-educated Hispanic Trump voter from California, a white woman Clinton voter from Michigan who supports deporting undocumented immigrants, and a pro-choice Republican from North Dakota who didn’t vote in 2016.

These are certainly all unique combinations. But the ubiquity of this general pattern shows that even as America geographically continues what the journalist Bill Bishop calls “the Big Sort” — where people increasingly live siloed with others who think, act and vote like them — American voters, and nonvoters, remain complex as a whole.

Naturally, being met face to face with those who you feel bear responsibility for a personally threatening president is an unpleasant reality for many of our followers, who reply to some profiles with anger. In response to one 33-year-old moderate Democrat who didn’t vote in 2016, several users commented with variations of “thanks for nothing.”

One self-described liberal who supports raising the minimum wage and opposes deporting undocumented immigrants but somehow ended up voting for Gary Johnson in 2016 led another follower to exclaim in a comment, “I want to smack him on the side of his head.”

Mere hours after news broke that Justice Anthony Kennedy (a supporter of Roe v. Wade) would retire, the bot tweeted the profile of a college-educated, pro-choice, moderate Republican white woman who voted for Donald Trump in 2016. The profile quickly went viral, for all the wrong reasons. It’s true that, on paper, this voter’s behavior doesn’t quite add up. To that, there is a solid body of research that questions whether much of the polarization that exists today is less informed by policy differences than by “exposure to messages attacking the out-group,” a more salient, visceral force at times.

Still, regardless of whether it should inform any electoral strategy, these randomized profiles aren’t just a quirk of a single survey. Political science research has long established that the American public’s beliefs are a mixed bag. But those findings have failed to manifest in public discourse, which is increasingly held online.

We all really do contain multitudes. No matter what you make of anyone’s politics, hopefully our Twitter bot can be a tiny reminder of that.

Alexander Agadjanian (@A_agadjanian) is a research associate at the M.I.T. Election Lab. G. Elliott Morris (@gelliottmorris) is a data journalist for The Economist.

The Times is committed to publishing a diversity of letters to the editor. We’d like to hear what you think about this or any of our articles. Here are some tips. And here’s our email: letters@nytimes.com.

Follow The New York Times Opinion section on Facebook, Twitter (@NYTopinion) and Instagram.

Advertisement

SKIP ADVERTISEMENT