Back in 2011, Michael Spence and Sandile Hlatshwayo described how employment levels in the tradable sector had remained stagnant from 1990 to 2008 while employment levels in the nontradable sector increased considerably. Spence and Hlatshwayo warned that this trend was problematic for a number of reasons. Productivity growth tends to be much higher in the tradable sector than in the nontradable sector, and though compensation and productivity don’t march in lockstep, compensation gains are more sustainable in a sectors experiencing productivity growth than in those that do not. And because much of the expansion of employment in the nontradable sector was financed by the public sector, fiscal constraints will tend to limit further expansion.
Moreover, broad swathes of sectors that have traditionally been understood as nontradable, like education and health care, are becoming tradable. Recently, Aaron Chatterji considered the implications of the accelerating transformation of “eds and meds,” sectors that have become a bulwark of local labor markets in much of the country, including Rust Belt labor markets that have been particularly hard hit by steep declines in manufacturing employment:
Education and health care jobs are so attractive because unlike manufacturing jobs, which have steadily declined over the last 25 years, they are largely shielded from global competition. As a society we continue to spend large sums of money, both in the public and private sector, on educating our students and caring for the health of our citizens. Since good jobs will increasingly require more education and our population is aging, the long-term outlook for these sectors looks positive. Education and health care also create jobs across income distribution, providing work for home health aides as well as college professors.
However, while the total number of jobs in these sectors could grow, it is not likely that all regions would benefit equally. For example, one might take for granted that there will be growing demand for orthopedic surgeons in Toledo, Ohio, and educational administrators in Iowa City. But the same forces that led other industries to cluster in specific regions (think technology in Silicon Valley or banking in New York) are now sweeping through education and health care.
As edX and other new educational institutions build new platforms for distributing course materials and (just as importantly) devise new, more reliable tools for assessing educational outcomes, it seems likely that we’ll see polarization across incumbent educational institutions as some flourish and expand their reach in the new environment while others flail, and perhaps go out of business. The medical sector, similarly, will change as a growing number of consumers embrace “travel surgery,” in which people travel to the most efficient providers for a given procedure. (A similar dynamic might shape the construction sector in the years to come.)
And so Chatterji ends his short piece on an ominous note:
[W]hile the number of education and health care jobs could indeed grow significantly in the coming years, that does not directly imply job growth in small and midsize cities that depend on these sectors.
In fact, the opposite situation could unfold for places that are not world leaders in providing education and health services. This category includes most places outside major metro areas. Toledo and Iowa City are quite typical in their heavy reliance on education and health care sectors. Without job growth in these industries, there are few remaining employers in most places left to make up the difference.
Instead, we might see the same dynamic of winners and losers observed in other industrial sectors, as top universities and hospitals become larger and absorb most of the increase in students and patients from across the nation. While these shifts might increase economic efficiency and gross domestic product over all, they will leave even fewer opportunities for good jobs in the places that need them the most.
Chatterji could be underestimating the benefits of regional specialization. As regions decline, we could encourage more domestic migration. But even the most well-designed relocation voucher program won’t turn depressed cities into ghost towns. So Chatterji’s concern could flow from the fact that the prospects for the residual residents of declining cities look set to deteriorate in the years to come.
What is more interesting still is that these dynamics aren’t limited to the United States. Dani Rodrik has observed that many developing economies are experiencing “premature deindustrialization,” i.e., manufacturing as a share of total employment is declining in Brazil, India, and China before ever reaching the peak levels that had been reached in the U.S., Britain, Germany, and Sweden. The diffusion of new business models like blended learning might mean that developing economies won’t be able to mimic the post-industrial transitions of the affluent market democracies — they will have to identify new empowering innovations to increase employment levels. If they succeed, the world will be a much better place. If they don’t, we will see the rise of new Rust Belts in countries that can ill afford to sustain vast numbers of underemployed people.