The worsening crisis in Europe and the slow recovery in the U.S. are distracting attention from a longer-term problem.

Across the developed world, well-paid jobs in the middle of the labor market are in decline. The issue isn’t simple, nor are the remedies, but governments must start paying attention.

The name for what’s going on is job polarization. Wages at the top are still rising, and at the very top they’re rising fast. Wages in a lot of menial and low-skill occupations are at least holding their own. But as research by David Autor, an MIT economics professor, and others has demonstrated, earnings in the middle of the labor market have been falling for years.

This issue is distinct from the heated debate over structural versus cyclical unemployment. To the extent that the U.S.’s 8.1 percent unemployment is structural -- because people have the wrong skills or are living in the wrong place and can’t move -- increasing demand through fiscal or other stimulus won’t work. To the extent it’s the result of cyclical changes in the economy, adding to demand is the answer.

Structural unemployment has always been lower in the U.S. than in other advanced economies, thanks to its famously flexible labor market. In this recovery, however, long-term unemployment has surged, and many workers have dropped out of the workforce altogether. If structural unemployment is rising, that’s grounds for concern.

Hollowing Out

But job polarization -- which began long before the recent recession -- is a worry in its own right. The issues are connected: Both involve matching workers’ skills and employers’ needs. But job polarization is more about the structure of employment and less about the characteristics of the unemployed. Look at it this way: At full employment, declining earnings in the middle of the labor market will still be a problem.

The main cause of this hollowing out is technology. Tasks such as cleaning a house or mowing a lawn aren’t easy to automate -- and because that kind of manpower is cheap to begin with, there’s little incentive. At the other end of the scale, advanced human skills are still needed to design software, play professional basketball or negotiate a corporate merger. Those tasks can’t be automated either. But in the middle is a swath of tasks where automation is feasible and wages are getting squeezed.

Computers are very good at skilled-but-routine tasks, and getting better all the time. Cheaper, too. Add the surge in high-bandwidth telecommunications, and you have what amounts to a new industrial revolution. But this time, good middle-income jobs, not muscle-power on farms or in factories, are under assault.

There are only two ways to respond. You could try to slow down the revolution, which would be a formula for economic decline, or help people adapt, which would spread the revolution’s benefits more equitably. The second is the right way to go. The catch is that the policy prescriptions aren’t simple.

A finding emphasized by Autor and others is that the labor market still rewards going to college, and in fact the so-called college premium has lately increased. Since the mid-1980s, demand for college graduates has been running ahead of supply. Getting more people through college will bridge the gap, and equip more workers for tasks that require human talent.

College Completion

Making sure people who start college finish with a degree is at least as important as getting more to start college in the first place. Completion rates in the U.S. have fallen, especially for men -- a main cause of the lagging supply of graduates. One of the worst things you can do for your standard of living is go to college, borrowing heavily to do so, then drop out.

To improve completion rates, high schools must do a better job of preparing students for higher education. Good work habits and proficiency in reading and basic math are the sine qua non of a successful college experience. That shouldn’t need saying, but colleges complain that high schools are failing to deliver. And here’s an issue that’s been especially neglected: better advice so that students can match themselves more skillfully to college courses and jobs.

Pushing more people through college will help, but it has limits. A surge of college graduates in the 1970s caused relative wages to fall. And there are signs that the college premium is itself flattening in the middle, while growing steeper at the top. College graduates do better in the labor market than those with less than a four-year degree, but not nearly as well as those with the right kind of degree. This underlines the importance of matching, both of aptitudes to disciplines and of disciplines to careers.

Maybe the biggest change demanded by job polarization is a rethinking of the idea of careers for life. In the U.S. and other countries, this is still very much the cultural norm: You train for one thing and do it for 40 years, gaining seniority through experience and on-the-job training. The pace of change, together with the fact that working lives (through choice or necessity) are lengthening, is making this model obsolete.

In the future, people may need to plan and prepare for multiple careers. Adaptability must become a frame of mind -- and schools and colleges must themselves adapt to this need. The economy will continue to demand labor, and an increasingly productive economy will be able to pay better wages. But matching people to jobs will need to be done more quickly and with greater flexibility. Education in the U.S. and other advanced economies faces a new and far-reaching challenge. The worker of the future needs to be trained to be trainable.

To contact the senior editor responsible for Bloomberg View’s editorials: David Shipley at