Consider the last six recessions in the U.S. economy. In the earlier recessions of 1969–70, 1973–75, and 1981–82, aggregate employment began to recover within six months of the economy’s trough. However, in our three most recent recessions of 1990 – 91, 2001, and 2007 – 09, aggregate employment didn’t begin to recover meaningfully for 18 – 23 months after the trough. Although the current recession ended over five years ago, we still read about the slowly improving labor market. What explains these “jobless recoveries”?
Explanations for the slow job recoveries come in three primary flavors: demand side, supply side and structural. The demand-side story claims there is simply too little demand in the economy, and those who believe this is the source of the problem argue for more fiscal and/or monetary stimulus. The supply-siders typically argue that unhelpful incentives or high economic uncertainty result in slow business expansion, investment, and hiring, thus slowing the economy and job growth. The structural argument focuses on problems resulting from a mismatch of education and skills with the jobs that are available, including a related issue of “job polarization.”
While all three of these phenomena may contribute to slow job growth, I’ll focus on the job polarization story. For the last 20–30 years, economists have found that jobs have been increasing in non-routine cognitive labor categories that involve problem-solving skills, intuition and persuasion. These jobs often require a higher level of education and are typically high-income jobs. At the other end of the spectrum, non-routine manual labor jobs in such areas as food service, home health assistance, and security and janitorial services, often requiring less than a high school degree and offering low pay, are also increasing. However, the demand for middle-skilled (and middle wage) routine jobs has declined. These jobs involve procedural, rules-based activities that are easier to automate and outsource. The most significant declines in routine jobs have been concentrated in occupations such as clerks, tellers, cashiers, travel agents, mail carriers, equipment operators, fabricators, and drivers. It’s the low recovery rate of routine employment that distinguishes the three most recent recessions from earlier ones. In the recessions prior to 1990, routine jobs recovered right along with non-routine jobs. However, for the three most recent recessions, especially the Great Recession, the recovery of routine jobs was almost nonexistent.
With an economy that is experiencing job polarization—recovery of non-routine jobs at the high-skill and low-skill ends combined with nonexistent growth of routine jobs in the middle—which policies hold the greatest promise? To the extent that the jobless recovery results from a disappearance of middle-skill routine jobs, policies to stimulate overall job demand may have little effect. Since openings are shifting toward non-routine jobs that are not easily automated, it becomes ever more important that we emphasize education and training in non-routine cognitive skills. Improving education and strengthening workers’ ability to perform abstract tasks involving problem-solving, intuition and persuasion should be our highest priority.