Bill Fox sells cars, including all-American Chevrolets and Ram trucks, in an upstate New York town whose congressman lost re-election in 2014 by 20 points — the kind of place where, according to the usual media narrative, voters are angry about the loss of blue-collar jobs and worried about the economy. But like other car dealers, Fox is seeing near-record sales: Somehow, he said, consumers don’t seem as worried about the economy as the pundits say they are.
“We’re not seeing [anger] at all,” said Fox, a partner in Auburn-based Fox Dealerships. “The way I account for it is, the public sees economic indicators that are OK, their job’s not threatened, and they may be afraid of the future, but the monthly [car] payment is good.”
That kind of optimism might be surprising in a year when Donald Trump, at least according to one dominant media narrative, rode a wave of economic anxiety to the Republican presidential nomination and when voters routinely cite the economy as a top concern. In a recent NBC News/Wall Street Journal poll, for example, 48 percent of voters said the country is “losing ground” on economic issues, compared with 45 percent who said it is “making progress.”1 In a recent YouGov poll, 40 percent of respondents said the economy was getting worse, versus 21 percent who said it was getting better. A weekly Gallup survey that asked the same question found that 37 percent think the economy is getting better and 58 percent think it is getting worse. (Gallup’s survey, unlike YouGov’s, didn’t offer “about the same” as an option.) Seventy-two percent of voters say the U.S. is on the wrong track, according to a separate poll from Gallup, which hasn’t found a majority satisfied with the “way things are going in the United States” since George W. Bush’s first term.
Yet even as Americans tell political pollsters that they are worried about the economy, they tell a different story in a separate set of surveys that are used by economists and investors to forecast consumer spending behavior. On Tuesday, the Conference Board’s Consumer Confidence Index, hit a nine-year high. The index is still well below its all-time high, set in 2000, but is as good as during Ronald Reagan’s “Morning in America” era of the mid-1980s. A rival survey from the University of Michigan finds consumers a bit less cheery but is still consistent with a solid economy. (On the other hand, the less well-known Bloomberg Consumer Comfort Index has fallen sharply in September, although it is still up significantly over the past few years.) Even people with only a high-school education — whose economic woes are often cited in media reports explaining Trump’s rise — are about as confident today as they were before the recession began, according to the Michigan survey.
“After a little bit of softening in the spring, [confidence] picked up,” said Lynn Franco, the Conference Board’s director of economic indicators and surveys.
Which set of surveys should we believe? Political and economic surveys serve different purposes. Political polls seek to predict or explain voter behavior; confidence surveys are used by economists and investors looking for an early hint about how consumers will behave in the months ahead. But there is evidence that confidence surveys like the Conference Board’s and Michigan’s yield a more accurate reflection of the state of the economy.
A one-point increase in the Conference Board index, for example, is typically followed by an $8.3 billion increase in consumer spending over the next 16 months, according to a 2015 report by Moody’s Analytics. It found that consumer confidence swings predict about two-thirds of all movement in consumer spending, which is 70 percent of the economy. That pattern seems to have persisted this year, when consumer spending grew at a 4.4 percent annual pace in the second quarter as confidence rebounded from dips in late 2015 and early this year. Political surveys don’t have the same track record: For example, Pew Research Center’s version of the wrong-track question, which is similar to Gallup’s, does a poor job predicting major economic variables such as consumer spending or job growth.2
The confidence surveys, especially Michigan’s, have been honed over the years to improve their predictive accuracy. Michigan has been conducting its phone survey for 70 years, and its dozens of questions zero in on consumers’ specific financial circumstances as well as their broad view of the economy. Questions that researchers thought would have predictive power but did not have been weeded out, said Richard Curtin, the longtime director of the Michigan survey.
This year, confidence is probably high because wages are rising and inflation is low, partly a result of the nearly 40 percent drop in the price of gasoline since it neared $3.80 a gallon in 2014. “We measure economic factors by sticking close to what people’s actual experience is,” Curtin said.
The consumer confidence data is consistent with other measures that show the economy performing solidly but not spectacularly. (Consumer spending has held up better recently than other parts of the economy, such as manufacturing.) So why do voter-consumers pull the long faces when pollsters ask the right-track question? Former Federal Reserve Chairman Ben Bernanke recently argued that the election-oriented surveys may be skewed by political polarization and that the non-political surveys capture a truer picture of Americans’ economic outlook.
“When Americans are asked specifically about the economy, in an apolitical context, they are for the most part not nearly as pessimistic as the conventional wisdom would have it,” Bernanke wrote in a blog post for the Brookings Institution earlier this year. “Instead, they answer more or less as they have done in the past at a similar stage of the business cycle.”
It’s possible that Americans interpret similar questions differently in a nonpolitical and political context. Economic surveys more so than political ones are focused on short-term behavior. Someone with a good job but little saved for retirement might be positive about his immediate financial well-being and therefore report relatively high near-term confidence. But that same person might be nervous about his longer-term outlook and want political leaders to address these issues. That’s consistent with consumer confidence surveys, too: Both Michigan’s and the Conference Board’s show more confidence about current conditions than the future.
As far as markets are concerned, though, there’s no question which set of surveys is better. Investors get paid for being right, not for choosing the right ideological team; they long ago learned to pay attention to the confidence data, not political polls or accompanying noise, to understand consumer moods that affect behavior in the ways that count.