Skip to main content
ABC News
Economists Are Bad At Predicting Recessions

Every president’s election-year nightmare — a recession — is suddenly looming over the 2020 race. In a survey released earlier this week by the National Association of Business Economics, 38 percent of economists predicted that the country will slip into an economic downturn next year, and another recent poll of economists put the chances of a recession in the next 12 months at 1 in 3. Those predictions are getting a lot of attention, and it’s not hard to see why — an economic slowdown in the middle of the presidential election cycle could reshape the race, potentially changing the calculus of Democratic primary voters and undermining President Trump, who has made the strong economy a central selling point of his presidency.

But take a deep breath before you spend a lot of time trying to figure out how a recession would change Trump’s reelection chances: Although the economy does have a big effect on an incumbent president’s odds of winning a second term, economists have a terrible track record when it comes to predicting recessions.

“Very, very few recessions have been predicted nine months or a year in advance,” Prakash Loungani, an economist at the International Monetary Fund, told me.

This doesn’t mean a recession won’t strike in the near future. Over the past few weeks and months, there have been some worrisome signals about the country’s economic health, fueling broader concerns about an impending recession. But exactly when the next economic downturn will come — and specifically whether it will interrupt the 2020 election cycle — is extremely uncertain. “There’s no economic data or research or analysis that suggests we can look 12 months into the future and predict recessions with any confidence,” said Tara Sinclair, a professor of economics at George Washington University.

Instead, and despite the recent rash of stories about economists’ predictions, economic downturns usually come as a surprise. A 2018 study conducted by Loungani and others looked at 153 recessions in 63 countries between 1992 and 2014 and found that the vast majority were missed by economists in both the public and private sector. This was painfully true in the case of the global financial crisis in 2008, which wasn’t officially declared a recession until it had been going for almost a year. Other studies have found that in general, forecasters are too sunny about economic growth.

Part of the problem, according to Loungani, was that in the past, economists were unwilling to risk their reputations by predicting an imminent recession that never came to pass. So it might actually be a good thing, he said, if more economists were now willing to sound the alarm. But Sinclair noted that even now, relatively few are pointing to an immediate crisis. In one of the polls, for instance, the share of economists who said they were expecting a recession this year fell from 10 percent in February to 2 percent in July.

And even if economists are more willing to be wrong these days than they were a decade ago, the task of predicting recessions itself hasn’t become easier. We have plenty of clues about how the economy is doing, but a system that’s so big, complex and deeply intertwined with human psychology and actions will always be difficult to predict.

Recently, for instance, the financial world flew into a tizzy over the inverted yield curve, which is generally seen as a reliable harbinger of an economic downturn. An inverted yield curve appears when short-term investments pay more than long-term ones, and it generally reflects a pessimistic mood among investors about the economy’s future performance. When the yield curve stays inverted for three months — as it did earlier this year — that’s a clear sign that a recession could be coming, according to research by Campbell Harvey, a finance professor at Duke University. But it’s not a guarantee, since an inverted yield curve doesn’t itself cause a recession. Instead, it’s a reflection of how investors feel about the economy’s future — and those feelings could be off-base. Other economists, like Sinclair, also said they’re not sure yet what the inverted yield curve means — and Harvey added that although it has a good predictive track record, it’s just one signal in a complex economic landscape.

Even if the inverted yield curve proves prescient and a downturn does come, we don’t have a good way to pinpoint when it will hit. According to Harvey, recessions have followed inverted yield curves by anywhere between six and 22 months. That’s not a small range, especially in political terms — it’s the difference between an economic slowdown that begins just before the Iowa caucuses and a recession that starts five months after the next presidential inauguration.

And in the meantime, consumers, investors and policymakers will all keep doing things that affect the economy. It’s possible that the anxious headlines about an impending recession could become self-fulfilling if everyday people respond by saving their money instead of spending it. Or maybe the opposite will happen, and smart policy responses to early warning signals could ward off a recession or make it less damaging.

Either way, the unpredictability of human behavior will frustrate anyone trying to pin down exactly when a recession will arrive. That doesn’t mean economists should stop making forecasts or that signals like the inverted yield curve aren’t useful. But anyone looking at predictions about when the next recession will land should take those forecasts with a big grain of salt. “Given historical patterns, a recession is likely to come again, so we need to be talking about what we’re going to do when it hits,” Sinclair said. “But we have to be open about the fact that we don’t really know when that will be.”

Amelia Thomson-DeVeaux is a senior editor and senior reporter for FiveThirtyEight.