Why the U.S. Has a Monopoly on Jobless Recoveries

Most rich countries hire back workers after a recession. The U.S. replaces them with machines.

There's no return.

pHOTOGRAPHER: Guillermo legaria/afp/getty images

In a series of earlier columns, I suggested that the U.S. has systematically underestimated the intangible value of jobs, while focusing too exclusively on economic efficiency. The real question is how can the U.S. turn that idea into real effective policy, without severely damaging the economy in the process. When looking for policy ideas, one place to start is other rich countries. And when we look at other developed nations, what we find is that in recent decades, they seem to do a better job putting people back to work after recessions.

In the U.S., we’ve been seeing a phenomenon sometimes called the jobless recovery. Look at a graph of the prime-age employment-to-population ratio since 1948:

Bounce-Back 'V's, Jobless-Recovery 'U's

Employment population ratio, 25-54 year olds

Source: Federal Reserve Bank of St. Louis

There’s a clear pattern. Before the 1990s, every downward dip except for the early '70s looks sharp and angular, like a letter V. But after 1990, they all look rounded, like the letter U. Shifting from V-shaped to U-shaped recoveries means that the economy is taking longer and longer to put people back to work after bad times hit, leaving them to languish for years on the couch playing video games.

Economists have recently discovered that it’s middle-skill routine jobs -- think of cashiers, telemarketers, or cooks -- that tend to get eliminated in jobless recoveries. In a landmark paper titled “The Trend is the Cycle: Job Polarization and Jobless Recoveries,” Nir Jaimovich and Henry Siu found that it’s these workers who aren’t being hired back in the U.S. after recessions hit. In fact, the much-feared phenomenon of job polarization -- the separation of the labor market into low-paid grunt work and high-paid knowledge work -- happens entirely during these U-shaped recoveries.

But does this happen in other countries? If not, there might be policy steps the U.S. could take to prevent this from happening. In a new paper called “Is Modern Technology Responsible for Jobless Recoveries?,” economists George Graetz and Guy Michaels looked at 17 different developed countries, from 1970 through 2011. The title refers to the hypothesis that companies replace routine workers with machines. Graetz and Michaels basically find that the modern jobless recovery is a phenomenon unique to the U.S., and that other nations manage to quickly re-employ their middle-skilled workers once bad times end.

First, the authors demonstrate that other rich countries have had snappier recoveries than the U.S., at least when it comes to employment. While employment started to lag behind gross domestic product in U.S. recoveries after 1985, in other developed nations there was no change. That implies that U.S. businesses recovered their footing by replacing people with technology, while companies overseas simply hired people back.

The authors then look specifically at industries that use more routine, middle-skilled workers. These industries weren’t particularly slow to recover in other countries. And when Graetz and Michaels look at routine jobs themselves, these tended to recover at about the same speed as other jobs -- except in the U.S., where job polarization occurred during recessions.

The conclusion here is that the U.S. is different. When bad times hit, American companies try to replace people with machines, while foreign companies hire back workers. Why this difference exists is a mystery. Unions might have something to do with it. I suspect that differences in financial systems might play a role -- U.S. companies tend to be funded by markets, while European and Asian firms usually depend on banks. A third possibility is that overseas companies engage in so-called labor hoarding -- essentially, maintaining long-term relationships with workers in which workers get temporarily laid off in bad times, but with the understanding that they’ll get hired back as soon as the company can afford it.

Whatever the reason, the question is whether the U.S. can be more like other countries, and what that might cost. If replacing humans with machines improves productivity, then there’s a natural tradeoff here -- protecting routine jobs might hurt long-term  growth. Countries such as Japan and Germany have generally had slower productivity growth than the U.S. since the late 1980s, and this difference might be a big reason.

On the other hand, it’s possible that trying to reduce headcount in recoveries wasn’t what made U.S. companies more productive. It could be that long-term relationships between workers and companies are actually beneficial, and the U.S. is wasting human talent and energy by letting these relationships fall apart every time the economy slows down. In that case, policy should encourage companies to maintain long-term contact with laid-off workers, and hire them back as soon as possible.

So more research is needed. But it does look like other countries have managed to avoid the curse of the jobless recovery, and U.S. policy makers should think about the possibility of emulating them.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

    To contact the author of this story:
    Noah Smith at

    To contact the editor responsible for this story:
    James Greiff at

    Before it's here, it's on the Bloomberg Terminal.