Why Economists’ Predictions Are Usually Wrong

They almost always fail to foresee a recession before it happens. But there are ways they can improve their insights.
by | November 2015
As global market forces have become more volatile, revenue forecasting has gotten harder for U.S. economists. (AP)

When the Chinese stock market took a plunge in late August, it triggered a 1,000-point drop over just a few hours in the U.S. stock market. It also triggered worries about whether the sell-off was a sign of a slowing economy. The response from many economists was that the market was simply due for a “correction” and that the figures that really matter to the economy -- job growth and unemployment -- should remained unscathed.

So far, that’s held true. But the fact is, economic forecasting is an extremely inexact science. Most of the time, economists tend to predict fiscal growth well. But when it comes to crucial pivots in the economy -- the major downturns and upswings -- they’re dead wrong almost every time.

In the last recession, it took the Survey of Professional Forecasters nearly a year to start adjusting their jobs forecasts downward after the economy began to turn in 2008. The discrepancy was so big, in fact, that the group’s year-ahead employment forecast predicted there would be 7 million more people employed in mid-2009 than was actually the case.

Budget forecasting has never been easy, but these days it’s getting even harder. Revenues have become increasingly volatile, and global commerce is ever more intertwined -- as China’s market tumble laid bare. “For someone to be a good recession forecaster, they need to predict it more than once,” says Tara Sinclair, chief economist at the jobs site Indeed and a professor at George Washington University. “That’s where we’ve pretty much seen this epic failure, where nobody’s been able to consistently forecast.”

The lag time in accurate data can be a big culprit. State forecasters tend to rely heavily on the jobs growth and employment data released monthly by the U.S. Department of Labor, because job growth is connected to increases in income and sales taxes -- two of the biggest revenue raisers for states.

However, the monthly figures are typically revised two or even three months after they are first published. So forecasters don’t really have an accurate picture of March’s economy, for example, until it’s already June. “That makes it really much harder to project those turning points, because you’re always looking in that rearview mirror,” says Juliette Tennert, director of economic and public policy research the University of Utah. Tennert was a forecaster for the state of Utah during the recession and, more recently, the state’s budget director.

This process has real consequences for governments. The foundation of all budgets is the forecast of how much the state will earn that year in revenues. When forecasters miss a downturn, lawmakers have to cut spending mid-year to balance the budget.

So how can economists better predict recessions? Sinclair thinks that more economists should focus solely on predicting major turns in the economy. But there just isn’t much incentive for economists to work only on recessions, which tend to happen about once every six years.

Tennert says forecasters should also track unemployment claims in their state because they provide a better real-time indication of the jobs economy. A big help, she adds, is getting out in the field and talking with a region’s business leaders about what they are seeing in the economy.

State forecasters usually build a small cushion into their revenue forecasts, but budget offices and policymakers also need to develop other cushions -- rainy day funds, for example -- in case of a downturn. “It would be nice if you got to under-forecast by 10 percent so when you get to a downturn everyone’s OK,” Tennert says. “But when you have these really critical needs at play, you just can’t be in that situation."