We’re approaching the four year anniversary of the economic recovery, and it still doesn’t feel like much of one, what with the unemployment rate at 7.7 percent and wages stagnant over the last half a decade. Various economists have gone to extensive efforts to ask whether this weak labor market is driven by not by an overall shortage of demand, but rather by something peculiar to the job market at this point in time.

It is an appealing story: Deeper, structural problems in the U.S. labor market that were masked by the housing bubble of the mid-2000s have come to the fore. It’s appealing to a lot of people, particularly those who wish to argue that monetary and fiscal stimulus won’t help thing.

One variation of this is the “skills mismatch” story, which holds that the problem ailing the U.S. economy isn’t that there are enough job opportunities, but that American workers don’t have skills to do the jobs where there are opportunities. It’s seemingly plausible, but if that were a widespread problem, you would expect to see wages in those in-demand sectors skyrocketing. Except for a few narrow specialties, that isn’t happening. A second variation is the “geographical mismatch.” Some parts of the country are doing OK, goes this theory, but people can’t move to take them because they are underwater on their houses. That also hasn’t really stood up to much scrutiny, as it would imply worker shortages in the stronger job markets, which outside of a handful of places in North Dakota and other energy-rich locales, isn’t really to be found in the data.

Economists at the Federal Reserve Bank of New York have a new piece up at their Liberty Street Economics blog that takes another whack at the idea that structural problems are behind the nation’s labor market woes. Rather than specific skills, could the unemployment problem be traced to some American workers ability to adapt to a workplace that requires creativity and flexibility?

The New York Fed economists,  build on work by David Autor and others on “labor market polarization,” the divide between different types of work. Rather than divide jobs by industry, they divide them by “routine” versus “nonroutine.” Routine jobs are those where the worker has explicit instructions and follows well-defined rules. Nonroutine jobs are those that require more flexibility, problem-solving, and creativity. While in general these track with the jobs that require more versus less education, it’s a little more subtle than that. Some jobs that don’t require advanced education are nonroutine, such as waiters and police officers. Many sales and other office jobs count as routine.

The long-term trend is unmistakable: More and more of the nation’s jobs are nonroutine. With machines increasingly able to do routine tasks like assembling things (in manufacturing) or handling routine sales transactions (such as through online orders), a rising proportion of the nation’s jobs require workers who have the flexibility and adeptness to adjust to circumstances.

Is Job Polarization Holding Back the Labor Market- - Liberty Street Economics


So there’s no question of the long-term trend. But could the trend have accelerated in the wake of the crisis? Could part of the weak jobs story be the fact that only people who have the nimbleness to leap from task to task and improvise on the job have opportunities? And given that not everyone has that ability, could this be the thing that is holding back the job market?

The New York Fed economists, Stefania Albanesi, Victoria Gregory, Christina Patterson, and Ayşegül Şahin, posit that if that were the case, you would expect a few things to be true. The unemployment rate should have risen more among nonroutine workers than those who do routine work. Workers who have been in nonroutine jobs should be finding new work faster than those who can only handle the repetitive tasks of routine work.

But it doesn’t show up in unemployment (“If job polarization were responsible for the sluggish recovery, we’d expect the unemployment rate among routine jobs to fall more slowly than the rate among nonroutine jobs,”write Albanesi et al. “The chart shows that this isn’t the case.”)


It doesn’t show up in data on which group stays unemployed for longer. (“If job polarization is causing long-term unemployment, we should expect long jobless spells to be more common among workers with routine occupations,” they write. “We see from the next chart that this isn’t supported by the data”)


The same can be said for the rate at which people in each group find jobs, and the ratio of people moving from routine to nonroutine jobs compared to the reverse. Neither of those seem to have changed much in the last few years, either.

So add it all up and the result is this: The long-term trend toward a workforce that requires people to be more flexible remains intact. But it doesn’t seem to have accelerated in the last few years in a way that can explain the deep recession and slow jobs recovery.

There is no question the American job market is changing, and it is particularly harsh for those who don’t have advanced skills and a certain entrepreneurial willingness to embrace change. But that doesn’t change the fact that what fundamentally ails the labor market is that there isn’t enough economic activity happening. It’s demand, stupid.