The Difference Between Polls and Public Opinion
4 min read
Issue polling can make you think that voters are oblivious. A recent poll by the Los Angeles Business Council finds that although 61 percent of respondents in the nation’s second-largest city think L.A. should substantially increase the number of new housing units, just 40 percent believe that doing so will make housing more affordable. Nearly half think that doing so will drive up the cost of housing and push residents out.
But here’s a more generous interpretation of that poll: Angelenos want the housing crisis to be solved. And when they hear a pollster offer them potential solutions to the problem, they express their agreement. They’re not policy experts; they’re transmitting their values and priorities. They want cheaper housing, they want more options for where to live, and they don’t want people to be forced from their current neighborhood. They just expect—quite reasonably—that working out the details is up to somebody else. That is, after all, the point of representative democracy.
Polls are incredibly useful in trying to demystify public opinion. They’re much better than heuristics such as “I ran into a guy at the grocery store” and “I saw a lot of lawn signs on my morning commute.” But issue polls are still merely tools that help uncover public opinion. They’re not public opinion itself.
Voters do have intuitions about what kinds of policies sound better to them. On housing policy, for instance, poll respondents are typically most excited about demand-side policies that have a clearly identifiable beneficiary. In this spirit, President Joe Biden’s new plan to lower housing costs calls upon Congress to expand rent vouchers and give first-generation homeowners up to $25,000 in down-payment assistance, among other measures.
Many people are cooler on supply-side policies that ease the construction of more housing, even though this is the only systematic answer to the housing crisis. Promises to streamline regulations sound amorphous and vague, and they directly benefit developers, who are routinely cast as villains. Sometimes, giving voters what they most want—in this case, an answer to an affordability crisis—can come only through seemingly unpopular means: building a lot of dense new housing.
By pushing initiatives more directly tailored to poll results, policy makers can appear to be addressing a crisis. But a popular course of action may not be a solution to the core problem. In March, Arizona Governor Katie Hobbs, a Democrat, vetoed an ambitious supply-side housing reform designed to produce more starter homes; the following month, she pledged $13 million in financial assistance for “up to 500” first-time homebuyers. How can helping a few hundred people buy a house seriously address a cost-of-living crisis that affects millions?
When pundits and political operatives take issue polling too literally, they may come to see the public’s views as hopelessly contradictory. You want fewer taxes and more social services? How could some of the same voters who said, in May 2020, that the government hadn’t gone far enough in shutting down businesses during the early pandemic also believe that all businesses that establish social-distancing protocols should be allowed to reopen?
A lot of polling is intentionally misleading, though. As ABC News’s G. Elliott Morris explains on his Substack, “The landscape of issue polling is particularly fraught with partisan advocacy organizations and biased surveys.” We should also be wary of polls that ask respondents to render instant judgment on oddly specific questions. What does it even mean to say that 54.6 percent of Americans think that the funding structure for the Consumer Financial Protection Bureau is unconstitutional? Or that 81 percent of Arizonans support conserving 30 percent of America’s land and waters by the year 2030?
How a pollster frames any question affects the response. More respondents might consider the 30 percent goal excessive if you told them that it would require setting aside more than 600,000 square miles of additional land—an area more than twice the size of Texas. Political context matters too. If you told people in Arizona, a polarized purple state, that the conservation goal is a Biden-administration priority, would 81 percent of voters still support it?
It’s reasonable to view a complicated policy proposal more or less favorably based on who is promoting it. If Biden doesn’t represent respondents’ values in policy areas that they understand well, they might rationally view an unfamiliar conservation proposal with more suspicion.
Political professionals tend to look down on the mental yardsticks that real people use when answering poll questions. In 2012, one Democratic pollster, frustrated with voters who embraced individual elements of Obamacare while rejecting the overall law, declared, “The first lesson you learn as a pollster is that people are stupid.” But policy makers miss something important when they decide that polls—about health care, housing, or anything else—reveal the shortcomings of the respondents rather than the complexity of the issues.
