This article is from the September/October 2010 issue of Dollars & Sense: Real World Economics, available at http://www.dollarsandsense.org


issue 290 cover

This article is from the September/October 2010 issue of Dollars & Sense magazine.

Subscribe Now

at a 30% discount.

dr. dollar logo

Dear Dr. Dollar:


Many times I have read (in writings by former Secretary of Labor Robert Reich, for example) or heard it said (by President Obama, for example) that technology is eliminating more manufacturing jobs in the United States than rising imports. Is that true?

—Kevin Rath, Oakland, Calif.



The former labor secretary and the president are right—at least if by “technology” we mean all sources of increases in output per worker, whether by greater use of machinery, work reorganization, or plain old speed-up.

Consider the years from 1990 to 2007. In this period, the supply of manufactured goods in the United States (production plus imports) roughly doubled (in inflation-adjusted, or “real,” terms). Had the share of imports stayed the same and had productivity (output per worker) stayed the same, employment too would have roughly doubled (assuming workers worked just as many hours, an issue I will come back to below). But that’s not what happened. In 2007, manufacturing employed about 76% as many people as in 1990, 13.431 million as compared to 17.695 million.

In fact, both productivity and the share of imports in total supply increased substantially between 1990 and 2007. Output per worker in manufacturing rose dramatically, by roughly 88% in real terms. Imports rose from 27% of the total supply of manufactured goods in the United States in 1990 to 48% in 2007. So the amount of domestic output, instead of doubling along with the doubling of supply, increased by only 43% (in real terms).

To figure out which of these factors—productivity or the share of imports in total supply—had the larger quantitative impact on employment, let’s look at two fictitious cases that allow us to examine each factor separately:

Case 1: Productivity increases by 88%, but the share of imports stays the same. If supply doubles, domestic output will also double. Employment will rise, but only by approximately 6% because with an 88% increase in productivity, each worker is yielding a larger output. Only 6% more workers are necessary to produce twice as much output.

Case 2: The import share increases from 27% to 48%, but productivity stays the same. If supply doubles, domestic output will rise by 43% (as it actually did). And if productivity stays the same, employment would therefore also increase by 43%.

It is clear that the increase in productivity had a quantitatively larger impact in limiting employment growth than did the increase in the share of imports. 

Job-loss in manufacturing presents serious problems. Some people lose their jobs and most are unlikely to get jobs that pay equivalent wages, even if overall employment picks up. These people suffer, and so do their families and communities. More broadly, the strong unions that developed in manufacturing are weakened, undercutting the economic and political strength of all working people.

But these negative impacts don’t have to take place. There is no good reason why we have to accept the choices of cutting foreign trade or cutting jobs, of reducing technological change or reducing jobs, or, for that matter, of destroying the environment (e.g., off-shore drilling and clear cutting of forests) or destroying jobs. Instead we can find ways to prevent the costs of economic changes from falling on workers, their families, and their communities.

One positive step would be to assure that productivity gains yield shorter work-weeks with no cut in pay—that is, no loss in the number of jobs. After all, this has happened before. Also, we need to establish the conditions for rebuilding strong unions in all segments of the economy—for example, by passing the Employee Free Choice Act and assuring that the National Labor Relations Board does its job. In addition, we need an array of good social programs such as, “Medicare for all” so people don’t lose their health care benefits when they lose their jobs. Likewise, better support for education and training programs for all workers—not just those affected by imports—are essential to facilitate their adjustment to change.

There may be good reasons to place some restrictions on technological changes or foreign trade—for example, when they generate environmental destruction, when they place workers in danger, or in conditions of virtual slavery (in the exporting countries). When we are forced to choose between the gains from trade or technology versus the well-being of workers directly affected, most often we should choose the well-being of the workers. The real solution to these problems, however, is to find ways to change the choices.

Whatever steps we take, we need to recognize that foreign trade is not the only cause—not even the quantitatively most important cause—of the decline in manufacturing employment.

Arthur MacEwan is a professor emeritus at the University of Massachusetts-Boston and a Dollars & Sense Associate.


Did you find this article useful? Please consider supporting our work by donating or subscribing.

« Back to Ask Dr. Dollar

end of article