The case for a better way to poll – Grid News

Get the best of Grid in your inbox, every day.

The case for a better way to poll

In recent years, progressives have invested heavily in crafting a narrative which holds that all or almost all of their main policy ideas are overwhelmingly popular with the public.


Hear more from Matthew Yglesias on this story:


Planned Parenthood routinely claims that “79% of Americans don’t want to see Roe v. Wade overturned.” Sen. Kirsten Gillibrand, D-N.Y., said that “paid leave is overwhelmingly popular — even with Republicans.” And Sen. Bernie Sanders, I-Vt., said “the American people want action, not never-ending ‘negotiations’ and obstructionism.” This is naturally an enticing line of thought for progressives — all their ideas are popular, and they are held back in life only by the timidity of the Democratic Party’s elected leaders.

But is it really true? After all, if Americans are overwhelmingly in favor of liberal policies, how are Republicans winning elections roughly half the time?

ADVERTISEMENT

Explanations among progressives have varied over the years. The power of “dark money” has faded recently to the power of misinformation. Both forces play a role, certainly, in shaping American politics. But the main thing is that public opinion is simply much more conservative than progressives like to believe.

As one Democratic pollster who’s frustrated with some of the demands his clients put on him told me, “the most underrated idea in American politics is that lots of people vote Republican because they agree with them on policy.”

But to the extent that advocates and activists want to generate polls showing that the public agrees with them, it’s not that hard to generate favorable issue polls. In general, studies show that respondents have a bias toward answering yes to pretty much anything you ask them — a phenomenon known as acquiescence bias. Combine that with a favorable description of the proposal you have in mind, and it’s easy to create a poll showing your idea is popular.

In the real world, people don’t have detailed opinions on most subjects. And Americans don’t go into a voting booth and vote for or against issues. They vote for candidates. This makes a less-common approach to issue polling enlightening — rather than asking whether people favor this or that idea, pollsters sometimes ask respondents which party they prefer on a given issue. The results are more telling than a preference poll alone.

Grid wanted to explore the difference between the two types of polling and what they tell us about the electorate. We asked several detailed issue questions. Then we asked about party preferences. If you look across the issue space in a rigorous way, the closely divided nature of American voting patterns simply reflects public divisions about the issues. Some progressive ideas really do poll well. But so do some conservative ideas. And even in a polarized world, the public does have some clear preferences about which party they’d like to see put in charge of which issues — with both parties having clear areas of strength and weakness.

ADVERTISEMENT

Progressive preferences aren’t as strong as they seem

On seven of the questions we asked with our best effort at neutral wording, we found the public really does support the progressive stance.

But our polling did not find the kind of across-the-board support that progressives would like to believe exists among the general public.

When you ask specifically, as we did, 54 percent of respondents said they favor maintaining Roe while 31 percent say they want to overturn it. In terms of party preference, Democrats have an advantage — 44 percent to 32 percent, but it’s a weak one, with 23 percent saying they don’t have a clear preference. That’s a far less formidable bloc than Planned Parenthood’s 79 percent claim might have you believe. It’s also clear that most voters, in fact, support significant restrictions on abortion rights in the second trimester — restrictions that Democrats generally oppose and that are often forbidden under the Roe precedent.

Given the option to pick no preference, there is virtually no topic on which either party has majority support. There’s a clear Democratic advantage on climate, education and abortion rights offset by a Republican edge on the generic topic of government spending, but on all these issues key swing blocks say they have no clear preference. And the most basic explanation for why the GOP has a good amount of issue trust is that the public supports a range of conservative ideas.

The public backs conservatives sometimes

It’s less common to see popular right-wing proposals polled, but we found that there is strong support for requiring recipients of social assistance programs to pass a drug test in order to get their benefits. Most voters want the death penalty for murder and rape. And a majority of the public says we should reduce government spending in order to lower taxes.

ADVERTISEMENT


Or consider gun control. Background checks poll pretty well, but the Democratic agenda on guns is much broader than that, and there’s a considerable range of polling outcomes on the question of banning assault weapons. In a 2019 experiment, Gallup found that minor differences in wording generated huge swings in public opinion. Asked “are you for or against a law which would make it illegal to manufacture, sell or possess semi-automatic guns known as assault rifles,” more than half of respondents said no. But tweaking the question to “do you think there should or should not be a ban on the manufacture, possession and sale of semi-automatic guns known as assault rifles” created a 14-point swing in favor of the ban.

The muddled public

American public opinion is a conflicted jumble of progressive and conservative impulses with limited knowledge of the details of the issues, mixed feelings about the parties, and a preference for divided government and compromise. What’s interesting about polling on party trust is it reveals the enduring significance of this old conventional wisdom even in an era of polarization, presidential tweets, media fragmentation and whatever else has changed about the political system.

It’s common for Democrats to deliberately seek out coverage of their proposals as “sweeping” or “transformational” as if the mass public’s biggest concern about the party is that it’s not left-wing enough. But there’s no evidence that it’s true of the public at large.

In particular, the conservative results on abstract questions about levels of government spending are a reminder that the tactic of bundling many disparate policy items together into a single package is a risky move. The same public that might look favorably on the idea of higher spending on clean energy or education or paid leave does not favor the idea of a broad-based increase in the size of government spending or particularly want to see one party decisively seize control of tax and spending issues. According to our polling, 51.6 percent say we should reduce spending in order to cut taxes, and by a margin of 36-30 respondents trust the GOP more on the topic of government spending.

None of which is to say that it’s necessarily a mistake for partisans to press for assertive policymaking when they control the levers of power. Activists push for ideas they believe in, and an idea that’s good enough on the merits can be worth paying a political price to enact. But one of the most striking patterns in American politics is a tendency of all newly elected majorities to convince themselves that the public is demanding sweeping change only to be met with rapidly waning public standing and a backlash in the midterms. Underlying that cycle of overreach and backlash is the dynamic revealed in these polls — an electorate that leans left on some topics and right on others, and that does not like to see strong partisan control over any major issue area.

  • Matthew Yglesias
    Matthew Yglesias

    Editor-at-Large

    Matthew Yglesias serves as an editor-at-large at Grid, where he contributes original reporting on “Uniquely American problems.”