One question goes to the heart of shaping expectations for that matchup: Why does everyone think the economy stinks? The answer is critical, given that this election is probably going to be close and that a variety of research suggests that the incumbent party’s chances are better when the economy is going well. President Biden, trailing Donald Trump in early polls, will need all the economic tailwind he can get.
Many commentators on the left have focused on a purported gap between what they see as objective data signaling a strong economy (in particular, persistently low unemployment) and middling to poor consumer sentiment, as in the University of Michigan’s monthly survey. This gap is sometimes attributed to partisanship — Republican voters being unwilling to give any credit to Mr. Biden — and at other times to media bias or misinformation driven by social media.
But a more careful look at the numbers reveals a different answer, and it requires no great mystery to solve, no inexplicable gap in the data.
Consumers don’t think the economy stinks. Rather, they quite rationally have mixed feelings about this economy — and they’ll reveal different things depending on exactly what you ask them.
They are pessimistic about the future, but that’s a matter of prediction, not misinterpreting the current economic situation. And here’s the good news for Mr. Biden: They’ve noticed that the data has been improving.
The terms “consumer sentiment” and “consumer confidence” are sometimes used interchangeably, but in fact, they reflect two distinct, longstanding monthly surveys often cited by economists. First, there’s the University of Michigan’s Index of Consumer Sentiment and, second, the Conference Board’s Consumer Confidence Survey.
One is not inherently better than the other. The best approach, as I usually recommend with polls, is to average them. They actually show rather different things: The Michigan numbers are bearish (although growing less so), and the Conference Board’s are bullish. That’s because they focus on different parts of the economy.
The Michigan survey puts a lot of weight on voter assessment of pocketbook conditions, like whether it’s a good time to buy major household items. The Conference Board, meanwhile, asks consumers for their appraisal of the employment and business outlook but nothing that really gets directly at things like consumer prices.
Also, and this is often overlooked: In both surveys, the majority of the questions are about voters’ predictions about future economic conditions and not how they think the economy is doing at the moment. For example, the Michigan survey asks about the possibility of a severe economic downturn over the next five years — a question that is notoriously hard even for professional economists to answer.
Fortunately, instead of one measure of consumer confidence, Michigan and the Conference Board publish separate subindexes, one focused on consumers’ perceptions of current conditions and the other about their outlook for the future. So we actually have four measures: two major surveys each asking two varieties of big-picture questions.
In these surveys, from January 1978 to January 2021, consumers’ assessments of current conditions usually tracked each other well. But in summer 2021, they began to diverge — and not just a little but hugely.
For the graphic below, I’m normalizing these four data series such that they’re all on the same scale, with a mean of 100 and a standard deviation of 20. This just means we can make an apples-to-apples comparison. A score of 100 represents the average consumer outlook between 1978 and 2024.
Why the divergence? The Michigan survey’s questions are highly sensitive to inflation, whereas the Conference Board’s are not. And spring 2021 is when inflation really began to ramp up, as a white-hot-recovery summer ran headlong into supply chain disruptions, the Delta variant and an injection of stimulus cash that led people to splurge on everything from revenge travel to meme stocks. It was a deeply strange economy — good for businesses and good for job seekers but sometimes awful for consumers.
So while the Conference Board numbers have consistently been above average, at roughly a score of 120 on my normalized scale, the Michigan ones took longer to recover. However, they have rebounded recently, reflecting a deceleration of inflation since roughly mid-2023, perking up to 82 on my adjusted scale in the January 2024 reading after having bottomed out at 41 in June 2022.
If you’re wondering why a rebound took so long — or why the numbers are still below average — there are a lot of good explanations. First, although inflation numbers when reported in the news typically focus on the year-over-year change, that’s not necessarily how consumers see them. Prices in December 2023 were only 3 percent higher than they were a year earlier, but they were 10 percent higher than they were two years earlier and about 18 percent higher than three years ago.
It takes some time for consumers to adjust to the new normal. Historically, Michigan consumer sentiment is more closely correlated with the two-year change in inflation than the one-year shift. If so, the timing could work out well for Mr. Biden, since the period of peak inflation will be farther in the rearview mirror by the time people vote this November.
But it’s a mistake to assume that consumers have just been reacting to news accounts of high gasoline or fast-food prices instead of actually observing the impact on their bottom lines. People’s pocketbooks really aren’t in great shape — income growth has struggled to keep up with inflation.
Per capita disposable income is historically one of the variables that most accurately predicts election outcomes. Although heavily affected by the timing of Covid stimulus payments, nothing about this data suggests that consumers have had a smooth economic ride under Mr. Biden. While corporate profits have soared to record levels, Americans quickly spent down the savings they built up during the pandemic.
It’s not just that goods have cost more; people have also been spending more on an inflation-adjusted basis. Often, that’s a sign of healthy economic demand. But consumers may be getting the short end of the stick as companies use algorithm-driven price discrimination to induce them to spend more on things they don’t necessarily want or need.
In short, consumers’ assessment of the current economic situation has been rational. They accurately report in the Conference Board survey that the business and labor outlook has been good. And they accurately report in the Michigan data that their pocketbooks were in bad shape because of inflation but are now recovering. But what about their future outlooks?
The Michigan and Conference Board surveys closely overlap and tell the same story. Consumers were in an optimistic mood for roughly the first six months of Mr. Biden’s term, with both surveys usually showing above-average forward-looking numbers. Then the Delta variant and the period of extremely high inflation hit in midsummer 2021 and knocked the wind out of Mr. Biden’s promise of a rapid return to normalcy. Inflation was more persistent than economists were initially expecting, and the S&P 500 lost around a fifth of its value on an inflation-adjusted basis in 2022.
Combined with the profound disruptions of the pandemic itself, there has been a lot of anxiety-inducing economic news for consumers. Although optimism is up in recent surveys, it’s not surprising that it’s taken some time to process everything.
There are other long-term factors pointing toward greater pessimism. For almost a quarter-century, a majority of voters have consistently thought the country is on the wrong track. There are many indications of a rise in poor mental health (and equally many hypotheses for why that’s happened). Many Americans have existential concerns about the long-term future for reasons ranging from environmental degradation to runaway artificial intelligence.
Fundamentally, Mr. Biden’s challenge is that it’s hard to persuade voters who are used to constant doomscrolling that it’s Morning in America again. The incumbency advantage seems to be declining; it’s been 40 years since a president won re-election by a double-digit margin.
But there is good news for Mr. Biden: Voter perceptions about the economy are not just vibes — in fact, consumer sentiment has tracked the objective data well. That data, particularly the pocketbook numbers that were the weak point before, has begun to improve, and that leaves the door open for a potential second Biden term.
It will be a close call. His numbers against Mr. Trump haven’t improved yet — in fact, they’ve gotten slightly worse lately — even as consumers’ mood has become more buoyant. His age is still a big concern for voters (yes, Mr. Trump is old, too), and the Democratic coalition is bitterly divided over the Israel-Hamas war and other issues.
Polls show that Mr. Biden has lost the most ground with lower-income voters — even as the robust labor market has helped the working class. His campaign, however, has said it will replay its 2020 strategy, with a heavy emphasis on Mr. Trump and a lesser one on the economy. It’s plausible that this is a mistake. Mr. Trump is no longer the incumbent president. And working-class Democrats don’t necessarily have the instinctual dislike for Mr. Trump that college-educated progressives do.
Still, we ought not to take an overly deterministic view of the relationship between the economy and elections. With any sort of presidential election forecast, we’re limited in making reliable inferences because of small sample sizes. This is only the 12th presidential election, for instance, since Michigan began regularly publishing its consumer numbers. We’re in dangerous territory where models sometimes fail. No previous presidential incumbent has been as old as Mr. Biden — and no major-party challenger has been as old as Mr. Trump.
If Mr. Biden loses, it may be because the relationship between the economy and perceptions of the president has weakened — not because voters are mistaking a good economy for a bad one.
Graphics by Sara Chodosh.