Green thumb on the scale: Nathaniel Stinnett quit his job at a law firm to focus on nudging more environmentalists into the electorate. Credit: Danielle Robertson

In a crowded field, no issue more spectacularly illustrates the failures of our political system than climate change. We are hurtling toward catastrophes that threaten the very existence of humankind, yet the matter is almost totally absent from political discourse. Donald Trump’s 2018 State of the Union address didn’t mention it. Neither did the official Democratic response. There are no fights being waged on climate policy in Congress; no government shutdowns based on it; no think pieces wondering whether Democrats should emphasize the environment over identity politics.

The reason for this is depressingly simple: polls show that the environment is a very low priority for most voters. And persuading them to care more is devilishly hard. Climate change is finely calibrated to thwart human psychology: the worst consequences—biblical flooding, widespread heat death, famine, unbreathable air—won’t be felt for decades, but avoiding them requires taking radical, society-altering actions right now. A stream of ever-more harrowing reports from scientists and journalists doesn’t seem to have jolted Americans out of their complacency, and we’re running out of time.

But what if persuasion is the wrong way of looking at it? What if the crux of the matter is not the people who don’t care enough, but the ones who already do?

In late 2013, a Democratic political operative named Nathaniel Stinnett noticed something odd about public opinion on the environment. Fresh off of managing a campaign to a painfully close loss in that year’s Boston mayoral election, Stinnett was looking over some polls that ranked political issues—the economy, national security, and so on—according to what percent of voters listed them as their highest priority. There were two rankings: one for all registered voters, and one for the subset who were likely to vote in the 2014 midterms.

As a seasoned campaign hand, Stinnett knew that politicians obsessively study opinion polls—but they care almost exclusively about likely voters. That’s why he was so struck by what he saw: the environment ranked much lower on the likely-voter poll, unlike other issues like immigration or abortion. That meant that people who listed it number one made up a smaller share of likely voters than of registered voters overall. They were, in other words, less likely than average to vote—up to 50 percent less likely, as Stinnett would later discover. He dug through other public opinion surveys and saw the same pattern. The stereotype of an environmentalist is someone deeply engaged in politics: the young Greenpeace volunteer gathering signatures on the sidewalk, the Birkenstock-wearing Baby Boomer who never misses a town hall meeting. But when it comes to voting, the stereotype is backward.

That may seem like yet another depressing data point. But to Stinnett, himself an environmentalist, it was actually great news, because it’s a lot easier to get someone to vote than to make them care about an Antarctic ice shelf. And it gave him an idea. If there were all these registered voters out there who already prioritized the environment, and simply weren’t voting, then the problem wasn’t really about persuasion; it was about turnout. The size of the gap between the two polls suggested that a potentially huge number of environmental voters, perhaps millions, were routinely sitting out elections. If he could find these people and get them to vote, Stinnett reasoned, they would start getting picked up in models of likely voters in future elections. The environment would climb higher in those likely-voter issue priority polls. Climb high enough, and politicians would start feeling that they can’t win without catering to environmentalists.

The stereotype of an environmentalist is someone deeply engaged in politics: the young Greenpeace volunteer gathering signatures on the sidewalk, the Birkenstock-wearing Baby Boomer who never misses a town hall meeting. But when it comes to voting, the stereotype is backward.

At the time, Stinnett and his wife were expecting their first child, and thinking hard about what kind of world she would grow up in. So Stinnett spent the next two years laying the groundwork for an organization that would take on the environmental turnout problem. In October 2015, he launched a nonprofit, the Environmental Voter Project (EVP). Traditional environmental activism includes turnout, but it centers around advocacy: coordinating rallies, lobbying elected officials, endorsing candidates, and the like. Stinnett wouldn’t bother with any of that. His organization would have exactly one objective: push environmentalists into the electorate, and trust politicians to respond in their own rational self-interest.

“There’s one thing you can always depend upon: politicians want to get reelected,” he says. “They will respond to the demand in the marketplace—just like if we drive 5,000 coffee drinkers to the door of Starbucks, they’re going to make more coffee.”

Electorally speaking, there are three main types of people in America: those who aren’t registered to vote; those who are registered and vote regularly; and those who are registered but don’t vote. The political process pays a good deal of attention to the first group—liberal and nonpartisan organizations spend millions of dollars every election cycle on voter registration drives and on lobbying to make registration easier. And both parties obsess over the second group, spending hundreds of millions of dollars on persuasion, messaging, and opinion research.

But the third group—the registered nonvoter—is largely ignored. Turnout is overwhelmingly the province of campaigns, and unless your campaign is especially well funded, contacting someone who is unlikely to vote is a risky use of limited resources.

“I’m not going talk to someone who occasionally votes,” says Dane Strother, a Democratic communication strategist, summarizing the attitude of a typical campaign. “I don’t have the time or the money. If you don’t vote, you don’t have a voice, and if you don’t have a voice, then we don’t care.”

Campaigns are built around winning the next election—not changing the electorate in the long term. But changing the long-term shape of the electorate is all Stinnett is focused on, and to do it, he’s bucking one of the most enduring features of U.S. politics. The system is set up to cater to people who voted in the last few elections. Unlikely voters are nearly invisible.

But, if Stinnett is right, they may be the key to saving the world.

What is an environmentalist?

I’m going to use Stinnett’s idiosyncratic definition: someone for whom the environment or climate change is a top political priority. This is important, because there’s a world of difference between preferences and priorities. Most Americans are already convinced that climate change is a problem. In a Reuters/Ipsos poll last summer, 72 percent said they want government to take aggressive action to stop global warming. But politically speaking, what Americans “think” doesn’t matter. The only opinions that count are the ones that drive voting decisions, and for most of that 72 percent, climate change simply doesn’t. “I would always work with great environmental candidates,” Stinnett says. “But it would have been malpractice if I told them to talk about an issue that voters didn’t care about.”

To change politicians’ incentives, Stinnett needed to flood the electorate not just with people who care about the environment, but with people who care about it more than just about anything else. (There’s some terminological fuzziness here, because not everyone who picks “the environment” on a poll means climate change—some people really care about pesticides or manatees. But there’s enough overlap that it makes sense to lump them together.) The first step was to come up with a way to identify every nonvoting environmentalist in a given state. The nonvoting part is standard practice among modern political practitioners. Your voting history is a matter of public record—not whom you vote for, but which elections you vote in. That history, plus other basic facts about you, has a lot of predictive power in future elections: if you voted in the 2014 midterms, you’ll probably vote in 2018 as well.

To figure out the environmentalist part, Stinnett hired Clarity Campaign Labs, a Democratic-leaning analytics firm, to build a predictive model, beginning in Massachusetts. They conducted thousands of phone surveys, asking people to rate their political priorities. Once they had a big enough sample of people who prioritized the environment, they matched it to a voter file—a list of names, addresses, and voting histories—that itself had been matched up with demographic and consumer data. That revealed correlations between personal information—like your age, what magazines you subscribe to, what stores you shop at—and political priorities. It’s akin to the way online advertisers use our browsing histories to figure out what we’re likely to buy.

Because the data sample is so big, the correlations have tremendous predictive value. The model, which is built anew for each state that the Environmental Voter Project expands into, assigns each registered voter a score from 0 to 100, representing the probability that they would pick environmental issues as a top priority. This allows Stinnett to isolate likely environmentalists, which he defines as people who score at least two and a half times higher than the average registered voter in their state. Then he zeroes in on the ones who are unlikely to vote in the next election. That’s his target group. (Technically, the EVP builds two models per state, using two different measurements of voter priorities, and combines the results to generate the target populations.)

Here’s what he has found. According to the model, environmentalists make up around 10 percent of registered voters nationwide—which tracks with recent public polling numbers. In any given election, however, their turnout rate lags between 10 and 20 points below the national average. Take November 2016, when there were 200 million registered voters nationwide and about 70 percent of them turned out. According to the model, there were 20.1 million registered environmentalists—but only half voted. If they had matched the overall turnout rate, they would have cast two million more ballots. The story was similar in the 2014 midterms. Turnout of registered voters was 44 percent, but among environmentalists it was only 21 percent. That meant more than fifteen million of them stayed home.

We can also zoom in on the swing states where the EVP has built state-level models from scratch. (The EVP expanded operations to five new states last year: Pennsylvania, Florida, Georgia, Colorado, and Nevada.) In Florida, about 547,000 registered environmentalists stayed home in 2016. Donald Trump’s margin of victory there was about 112,000 votes. In Pennsylvania, just under 300,000 environmentalists sat the election out; Trump won there by just over 44,000 votes. Both states have gubernatorial and senatorial elections this year, and because turnout is lower in midterms, the number of targets will be much higher than in 2016.

There’s a world of difference between preferences and priorities. In a poll last summer, 72 percent of Americans said they want government to take aggressive action to stop global warming. But what Americans “think” doesn’t matter. The only opinions that count are the ones that drive voting decisions.

The failure of our government to deal with climate change is, of course, driven by the lunacy and cupidity of the Republican Party, which caters to funders like the Kochs and the Mercers by denying, as a matter of party orthodoxy, that the issue even exists. But while Stinnett’s background is in progressive politics, the EVP is rigidly nonpartisan. The group doesn’t endorse or criticize any particular candidate; it doesn’t tell people whom to vote for. “We’re not changing who makes the policy,” Stinnett says. “We’re just changing who votes, and trusting that politicians are going to continue to want to get reelected.”

This gives him two ways to win. In the short run, success most likely means helping Democrats win more elections in close races. (Predictably, environmentalists are much more likely to be Democrats than Republicans.) In the long run, it means convincing more Republicans that their best chance to beat Democrats is by appealing to environmental voters—just as many Democrats have long felt pressure to cater to the small but highly engaged bloc of pro-gun voters. In that sense, the EVP is the NRA of saving the environment.

No straw, please.”

This comment, directed to a waiter at an upscale pub in downtown Boston, is one of very few outward hints that Stinnett is an environmental nut. (It prompts a long conversation about the wastefulness of disposable straws.) Stinnett, who cut short a career as a professional opera singer to go to law school and get involved in politics, speaks in a warm, crisply enunciated baritone, a vestige of his stage training. Before launching the Environmental Voter Project, he had a day job as a land use attorney in Boston. In his spare time, he worked as a high-level volunteer for Democratic campaigns, including as treasurer for Martha Coakley’s 2010 U.S. Senate run campaign. With a gleaming shaved head and well-fitting business-casual clothes, he looks more like a management consultant than a tree hugger.

In fact, environmentalists generally don’t fit the stereotypes. Stinnett has found, for instance, that black, Hispanic, and Asian people are all more likely than white people to prioritize the environment. “I would go so far as to say that a Latina grandmother in Phoenix or a young African American man in Miami are now more likely to care about environmental issues than white hipsters in Portland,” he says. Some of the predictive factors are things you’d expect, like having a college degree. But others are head-scratchers. Are you a pro-basketball fan living in Massachusetts with an AOL email address, a fourteen-year-old kid, and no landline? Congratulations: you’re probably an environmentalist.

What Stinnett has not figured out is why so many environmentalists don’t vote. There’s no obvious demographic overlap that explains it. Within any given population slice, environmentalists almost always have a below-average turnout rate. When Stinnett surveyed a few hundred of them on a battery of civic participation–related questions, the always-voters and the never-voters gave identical answers.

For now, Stinnett says, he has “embraced the black box.” Ultimately, all that matters is pushing these people into the system. And to do that, it’s more important to know what makes people vote than what makes them not vote.

American elections present two puzzles of human behavior. The first puzzle, given our low participation rates compared to other wealthy democracies, is why so few people vote. But the second puzzle is why so many vote. In all but the smallest elections, you are a statistical lock to never affect the outcome. So why bother?

Early academic attempts to answer this question, in the 1950s, applied an economic framework called rational choice theory: the benefit of voting equals the probability of deciding the outcome, multiplied by its importance, minus the cost of voting. (Or, in English, you vote only when the prospect of tipping the election outweighs the hassle.) The obvious problem with that model is that the expected benefit will always be negative—you will never vote—because you’re generally more likely to get hit by a bus on the way to the polls than to cast the deciding ballot. How could the possibility of being the pivotal vote ever outweigh the cost?

A more satisfying explanation comes from more recent research applying the insights of behavioral psychology, particularly by a pair of political scientists named Alan Gerber and Donald Green. The biggest driver of voting, they have found, may be the pressure that comes from social norms. It’s about identity, not outcomes. Call it the “everyone else does it” theory of civic participation. In a study published in 2008, Gerber and Green sent letters to people in Michigan warning that whether they voted or not would be publicized to their neighbors. That led to dramatically higher turnout than the control group, blowing any other direct mail–based get-out-the-vote techniques out of the water. Another group was told that the researchers would be sending them—but not their neighbors—an updated chart of their voting history after the election. They didn’t vote at quite the same rate as the public-shaming group, but still showed a huge increase. Perhaps this was because they wanted to reassure the researchers that they were good people. Or perhaps they wanted to reassure themselves.

To most people, the postcard from the Environmental Voter Project would have looked unremarkable—a generic reminder that an election was coming up. But Amy Bucher is a behavioral psychologist. Her day job is to nudge people to develop better habits. When the card showed up last summer in her mailbox in Boston, Bucher recognized the work of a fellow practitioner.

The mailer used “textbook” behavioral psychology tactics to encourage voting, Bucher wrote on her blog. One nice touch: the recipient’s name was handwritten onto the card. In this case, the recipient was her husband, who, unlike Bucher, doesn’t always vote in nonpresidential elections. The back of the card mentioned “critical issues facing our environment.” Here was a bit of luck, Bucher thought: the environment just so happened to be a top priority for her and for her husband, who works in clean energy.

Of course, luck had nothing to do with it. The EVP had targeted her husband because he was an infrequent voter and its model said he prioritized the environment. But otherwise, Bucher was right. Stinnett and his team harness behavioral psychology, especially voting-related research, to run their GOTV operations—the heart of what they do day to day.

When I visited their offices in downtown Boston last October—a few desks in a no-frills coworking space, plus one tiny windowless office for Stinnett—the three-person team crowded around Stinnett’s desk to discuss a study of two different direct mail scripts. (Direct mail is one of several techniques they use, along with phone banking, in-person canvassing, texting, and targeted online video ads.) Both mailers included language declaring that turnout was expected to be above average in the upcoming election.

“Isn’t it dishonest to say every election’s turnout will be above average?” I asked.

There was a pause. Stinnett looked me in the eye.

“Yes,” he said, as the room broke into laughter. But it’s probably an effective lie. In another experiment, Alan Gerber and a colleague told one group of voters that turnout in an upcoming election was expected to be low, and another group that turnout was expected to be high. Then they asked them whether they intended to vote. Rational choice theory would predict the first group to say yes more than the second: the fewer other people participate, the higher your chances of making a difference. But the researchers found the opposite: the voters who were told turnout would be high ended up being much more likely to say they intended to vote, and this effect was most pronounced among infrequent voters. It was social pressure at work. The lesson: if you want to get unlikely voters to vote, tell them everyone else is doing it.

I spent one morning tagging along with Kate Donaghue, one of the EVP’s volunteers in Boston, knocking on doors in the Charlestown neighborhood ahead of the November citywide vote. It was a quiet election, in which turnout was expected to be low and the incumbent mayor, Marty Walsh, was assured a lopsided victory. That made it a perfect environment for the EVP. In a low-turnout race, there are by definition more unlikely voters to target. And getting someone to participate in one is the best way to get them included in future likely voter models. From a campaign’s perspective, the people who vote when everyone else stays home are the surest bets to turn out in the next election.

It was a weekday, so at most addresses Donaghue just left a flyer like the one Bucher got in the mail. But some people came to the door, like Peter Nowak, a retired university administrator wearing a Boston Latin sweatshirt. Donaghue began by asking if he was planning to vote. “Yes, absolutely,” he said, just like the other dozen or so people we spoke to that day—even though we knew, from their voting histories, that this was almost surely not true.

But Donaghue maintained complete ingenuousness. “Oh, that’s wonderful,” she said, holding out the flyer. “Would you be willing to sign this pledge card? It just says what you just told me—that you’re going to vote.”

Here was the trap. If Nowak signed the pledge in front of Donaghue, he would be publicly committing himself to voting. And the EVP would mail him the pledge card back shortly before the election to remind him of his promise. The card itself included a reminder that one’s voting history is a matter of public record. It was all designed to leverage social pressure. Each step of commitment dramatically raised the likelihood of eventually voting. The EVP sets aside a control group in every election: nonvoting environmentalists who don’t get contacted. This limits the potential turnout bump—which is why a campaign would never do it—but is essential for figuring out what works and what doesn’t. So far, they’ve found that people who sign a pledge card are 21 percent likelier to vote than the control group, compared to 14 percent for people who say they will vote but don’t sign.

(The EVP doesn’t go so far as to threaten to out nonvoters to their neighbors, which risks a backlash. But most of its direct mail includes a not-so-subtle warning: “We may follow up with you after the election to ask about your experience at the polls.”)

Nowak, like almost everyone Donaghue talked to, signed the card. (I later learned that Donaghue is a superstar canvasser. Other volunteers had a lower hit rate.) “Thank you for being a good voter,” Donaghue said before turning to go—a deliberate appeal to a sense of identity. A few weeks later, I called Nowak. Curious whether the EVP’s model was working, and without mentioning the environment or climate change, I asked him to name the political issue most important to him.

Environmentalists generally don’t fit the stereotypes. Some of the predictive factors are things you’d expect, but others are head-scratchers. Are you a pro-basketball fan living in Massachusetts with an AOL email address, a fourteen-year-old kid, and no landline? Congratulations: you’re probably an environmentalist.

“I feel very strongly about the environment,” he replied. “I have adult children. I’m worried about, what kind of a world are they going to live in if the environment continues to get worse and worse?”

Then I asked about his voting habits. He said he “just about always” votes, but then admitted that that’s not so true in off-year elections—and that he might not have voted in the mayoral race if Donaghue hadn’t reminded him.

In January, Stinnett sent me results from the EVP’s work in last November’s elections. Unsurprisingly, the numbers were strongest in the Boston area, where the EVP combined remote outreach with in-person canvassing; nothing beats face-to-face interactions for boosting turnout. In Boston, turnout among the treatment group was 6.3 percent higher than the control. That may not sound like much, but in the world of get-out-the-vote it’s massive. It translates to 2,683 more environmental voters than would otherwise have voted, in a race in which total turnout was about 109,000—which means, if their numbers are correct, that the EVP was responsible for one out of every forty votes cast.

Elsewhere, where the EVP’s efforts didn’t include in-person canvassing, the results varied. In St. Petersburg, Florida, which had a contested mayoral race, the treatment group saw a 4.5 percent bump—a striking result for an operation done entirely remotely. (Alan Gerber, the political scientist, has written that a “typical non-partisan mailer increases turnout by less than 0.5 percentage points.”) In Atlanta’s hotly contested mayoral race, meanwhile, the EVP raised turnout by a more modest 2.8 percent. But even that meant adding 955 voters to the electorate. In December, Democrat Keisha Lance Bottoms would go on to win the runoff by only 821 votes.

What excites Stinnett more than the results of any one election is what may be happening to the treatment groups over time. The EVP tracked one group of targeted environmentalists in Massachusetts for four elections over the course of a year. By the end, the group was turning out a full 12 percent more than the control group. Stinnett hypothesizes that this cumulative effect is the result of campaigns spotting new likely voters and adding them to their GOTV efforts. Once the EVP nudges someone to vote in an election, especially a low-turnout one, that person goes from being invisible to campaigns to having the equivalent of a flashing sign next to their name in the voter file. Whatever the exact mechanism, the results align with another study by Gerber and a colleague that found that “voting in one election substantially increases the likelihood of voting in the future.”

Twelve percent of the environmentalists who didn’t turn out in 2016 would have been more than a million votes. But that’s a long way from where the EVP is now. In 2018, Stinnett says, the organization has identified 3.05 million environmentalists in six states who are very unlikely to vote, and will be targeting about 2.4 million of them. (The rest go into control groups.) Most of that will be through direct mail, online ads, and texting, because door-to-door canvassing, which requires boots on the ground, is much harder and more expensive to scale up. Extrapolating from his 2017 results, Stinnett expects to add “anywhere from 67,000 to 108,000 new environmental voters to the electorate,” depending on whether the results are more like Atlanta or more like St. Petersburg.

In the short run, success for the Environmental Voter Project most likely means helping Democrats win more elections in close races. In the long run, it means convincing Republicans that their best chance to beat Democrats is by appealing to environmental voters.

Eventually, Stinnett hopes to expand to all the states in which there are currently enough nonvoting environmentalists to make a real difference—probably just under half of states.

“The ultimate goal is to put ourselves out of business,” he says. “If our data shows that there’s no longer a significant environmental voter problem, well—the world doesn’t need another environmental lobbying organization or another endorsement group.”

Stinnett’s plan to save the world from climate change depends on a somewhat stylized theory of policymaking. Likely voters tell pollsters what their top issues are; pollsters tell politicians; and politicians respond to that revealed demand by supplying or emphasizing policies that cater to it.

This account leaves out other ways in which politicians take the temperature of their base. Republican legislators have to keep up with Fox News and talk radio. Officials from both parties pay attention to what they hear from constituents who show up to town hall meetings and protests.

More importantly, Stinnett’s plan seems to ignore two other drivers of politician behavior: money and ideology. Policy formation is far from a perfectly efficient market. Lawmakers cater to the desires of major donors and lobbyists, and they also enter office with ideological commitments—things they actually believe in. Both these factors were on high-profile display last year, as congressional Republicans voted for a deeply unpopular tax bill, and almost passed an even more unpopular Obamacare repeal.

Stinnett isn’t really so naive. In fact, his worldview could be described as idealistic cynicism: he believes politicians will so predictably try to maximize their chances of reelection that they will inevitably take action on climate change if they believe their seat depends on it. He’s not trying to fix the system; he’s trying to exploit it.

“Politicians go where the votes are,” he says. “We can always depend on politicians to poll likely voters, and nothing motivates a politician more than the prospect of winning an election.” The health care and tax cut votes were remarkable precisely because of how unusual it is to see lawmakers vote against their own apparent self-interest. “Yes, donors drive policy. But the only thing that money buys is votes.” The way to neutralize the power of big money is to convince politicians that there are more votes to be gained by defying donors than by obeying them.

And while ideology is real, there is no end to the examples of high-profile politicians reversing themselves on major issues when enough votes appear to be on the line. Recall Barack Obama on gay marriage, Hillary Clinton on the Trans-Pacific Partnership, and Mitt Romney on pretty much everything. Veteran politicians change long-held positions all the time to survive.

Those who don’t tend to end up like Bob Inglis. Inglis, a former Republican congressman from South Carolina, lost his seat to a primary challenge from Trey Gowdy during the Tea Party wave of 2010. His departures from right-wing orthodoxy had been several—supporting comprehensive immigration reform, voting for the bank bailout—but his “most enduring heresy,” as he puts it, “was saying climate change is real and let’s do something about it.” Today, he runs republicEn, a nonprofit that advocates for free-market solutions to climate change.

I recently asked Inglis to describe the role that issue polls play for members of Congress. “You go to a Republican conference meeting,” he said, “and there’s a good pollster there who’s telling you what the top issues are” among likely voters—especially likely primary voters. That helps the members decide what issues to emphasize.

“Campaign cash is helpful, but the most important thing is the level of the tide,” Inglis continued. In the 1970s, for instance, a candidate could have made Taiwan a central campaign issue, “because there was a high tide of fear about the red scare and Communists. But if you’re running in 2018, I’d tell you it is dead low tide on Taiwan. The Kochs and the Mercers could give you all the money you wanted; you couldn’t win on Taiwan.”

What Stinnett is trying to do, then, is raise the tide on the environment—at worst, to make it safe for politicians to talk about it, and at best, to make it dangerous for them to ignore it.

“If you flood the electorate with a bunch of people who really care about environmental issues, that will drive policy,” he says. “Or the politicians won’t get reelected. It is that simple.”

We understand much more clearly today why people vote than we did back in the days of rational choice theory. But the thing about behavioral psychology explanations is that they still don’t explain how voting can be rational. “Everyone else does it” may describe our unconscious motivation, but, as parents throughout history have reminded their teenage children, it’s not very logical.

There actually is a rational, self-interested reason to vote—it’s just not the one the rational choice theorists imagined. We tend to assume that voting is purely transactional: we cast a ballot to pick the winner. Stinnett’s insight is that voting is not merely, or even meaningfully, a transaction. It’s more like a dialogue. When you vote, you declare that you’re likely to vote again in the future, which increases the chances of future candidates paying attention to your opinions and priorities. That, not the infinitesimal chance of affecting the outcome of the election, is the rational, self-interested reason to vote.

And it applies to every election. If voting were just a transaction, then in our winner-take-all system, your vote would be wasted whether the final margin was fifty votes or 50,000. But from the perspective of joining the political conversation, your vote always has marginal value. No matter how small or uncompetitive an election is, you are adding to the aggregate demand for action on the things you care about.

The Environmental Voter Project tracked one group of targeted environmentalists in Massachusetts for four elections over the course of a year. By the end, the group was turning out a full 12 percent more than the control group.

“That person who never misses an election, they drive policy in this country,” Stinnett says. “Obviously, it’s important who wins. But regardless of who wins—even if you go into the polling booth and write your dog’s name—simply by announcing that you’re a regular voter, you become part of a very small and select group that policymakers care about.”

The failure of our national leaders to take climate change seriously may be both cause and consequence of environmentalists’ low voting rates: if neither candidate talks about the issue you care most about, you may see less reason to participate. But if you don’t participate, then candidates will see less reason to talk about your issue.

It’s a vicious cycle, but it’s up to voters, not politicians, to break it. Someone who sits out the democratic process because leaders have the wrong priorities is drawing the exact wrong conclusion.

“At a very deep level, many Americans obviously have a very cynical approach to politics,” Stinnett says. “But they never internalize it to a point where they realize that politicians follow more than they lead. Take your cynicism one step further.”

It’s fair to have contempt for politicians who take their cues from opinion polling rather than what’s best for the country. But since we know they do, there’s no excuse for not using it to our advantage. Our political life is besieged by anti-democratic distortions—big money, gerrymandering, nihilistically partisan right-wing media—that threaten to swallow the system whole. But they haven’t yet, quite. On climate change, as with so many other issues that bedevil us, votes are still the most powerful weapon we’ve got.

Gilad Edelman

Follow Gilad on Twitter @GiladEdelman . Gilad Edelman, a Washington Monthly contributing editor, is a politics writer for WIRED.