Supporters of gun control measures often say their ideas are common sense, suggesting that nearly every American supports them. If you spend five minutes discussing the topic with people at Everytown, Giffords or any other gun control group, they will inevitably point out that polls show as many as 90 percent of Americans support universal background checks, a policy that would require everyone to pass a check to obtain a firearm.
Do they, really?
You can certainly find surveys, including from reputable pollsters, in which 80 percent or 90 percent of respondents say they support universal background checks. But when the idea has gone to a vote, it has performed worse than the polls suggest. In Nevada, a 2016 ballot measure to strengthen background checks got a little more than 50 percent support — barely enough to pass. In Maine, another measure that year lost with 48 percent of the vote. Hillary Clinton carried both of those states in the 2016 election. I would love to believe that nearly every American supports stricter gun laws. It doesn’t seem to be true.
It’s not just guns. Policy proposals very often overperform in issue polls, according to a recent study that looked at available polling and ballot measure data across 11 topics from 1958 to 2020. The findings apply to both liberal and conservative causes. The more popular a policy is in polls, data showed, the more likely it is to underperform on Election Day.
These polls distort our democracy in important ways. Political parties shape their agendas and priorities based on polls that appear to overestimate support for these ideas in the real world. This can make politicians more extreme; if they believe their causes have public support, they will be less likely to moderate. So politicians and their parties become less representative of their voters, and the public becomes disenchanted with leaders they now see as out of touch. Meanwhile, Americans don’t get the policies they actually want.
The upshot is not that we should ignore issue polls. The recent study, from researchers at Catalist, George Washington University and Vanderbilt University, found that surveys often predict overall public support for a policy, even if they can’t predict the exact share of Americans who will vote for it. Still, the findings indicate that we should read polls more cautiously and ease up on the belief that any debated idea really has near-unanimous support in our polarized country.
Let’s look at other examples of issue polling misses. Here are some other issues that tend to get more liberal support. Despite promising polls, ballot measures dedicated to them have underperformed or even lost, sometimes repeatedly, on Election Day:
In some ways, the charts above understate the polling misses. Given that Washington is a liberal-leaning state and carbon taxes are a liberal cause, you would expect the idea to get higher support there than it would in a poll of all U.S. voters. Yet the opposite happened — multiple times.
The trend is not exclusive to liberal issues. Some conservative measures have seriously underperformed and lost in recent years, too:
Again, the charts understate the polling misses. You’d expect a conservative cause like school choice to do better in a conservative state like Kentucky than it does in a national survey. But it did much worse — and even lost.
What’s going on here? Several things at once. Americans are generally moderate, incremental and have a status quo bias, priming them to vote no on just about any change. Once an idea goes from a survey to the voting booth, political polarization takes hold and support ends up breaking down closer to what partisan ties would predict. And voters probably perceive national polling questions differently from state or local ballot issues.
Pollsters’ questions rarely match ballot measures’ language, and even small wording changes can have big effects. Similarly, voters might be enticed by an idea that’s cleanly framed by a pollster but grow skeptical once they see how it would work in reality. Lastly, the people taking the surveys may not give much thought to their answers or have an acquiescence bias — a tendency to say yes.
Put another way, issue polling measures public opinion “before it gets pulled into the political tornado,” said Jocelyn Kiley, a director of research at Pew Research Center. People might at first like the idea of requiring a background check for every gun purchase. But what happens when they find out their favorite politician opposes the policy? Or that the requirement might apply not just to transactions at stores and gun shows but also between a father and his son? Or that alternative ideas could combat gun violence without changing so much so quickly?
Exceptions exist. Abortion rights outperformed the polls in some recent elections, and ballot measures on issues that are less polarized by party, such as marijuana legalization, land closer to the polls. Across most topics, though, polls tend to find higher support for policy changes than they have in the real world.
So how can you gauge public opinion? Don’t rely on just one data point. Look at a variety of surveys that ask about the topic in different ways. Look at the outcomes of previous ballot initiatives. Keep in mind thermostatic backlash: Americans’ tendency to recoil at major policy changes. Ask yourself if you are reading poll results more hopefully than empirically.
These findings should influence how America’s leaders govern — for their own sake, if not the country’s. In recent years, we have seen the damage that public leaders can do to their own causes when they misread public opinion. By largely neglecting anger over inflation and immigration, Joe Biden hurt his party’s chances in last year’s election, and President Trump has subsequently dismantled much of what Mr. Biden did in office. Politicians who neglect the public mood risk losing not just their elections but also their legacies.
America’s leaders often say that we have more common ground than our political conversations let on. But we’re also a deeply divided country on many important issues. The nation’s governance and policies should reflect those divisions.
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: [email protected].
Follow the New York Times Opinion section on Facebook, Instagram, TikTok, Bluesky, WhatsApp and Threads.
German Lopez is a writer for the editorial board. @germanrlopez
The post Single-Issue Opinion Polls Are Bunk appeared first on New York Times.
