Estimating the Gains and Losses from Free Trade

Posted on by

Yesterday, our post “Costly Trade with China,” cited a study by the same name by the Economic Policy Institute which estimated the jobs lost to trade with China between 1997 and 2006 at 2.1 million. I wondered whether anyone had done the calculus as to what the offsets to these job losses might be, and Mark Thoma was kind enough to send me a reply in the form of a quote from a recent Bernanke speech:

After all, the United States is a big country, and we can certainly achieve many of the benefits of specialization by trading within our own borders. How important is it for the health of our economy to trade actively with other countries? As best we can measure, it is critically important. According to one recent study that used four approaches to measuring the gains from trade, the increase in trade since World War II has boosted U.S. annual incomes on the order of $10,000 per household (Bradford, Grieco, and Hufbauer, 2006).2 The same study found that removing all remaining barriers to trade would raise U.S. incomes anywhere from $4,000 to $12,000 per household. Other research has found similar results. Our willingness to trade freely with the world is indeed an essential source of our prosperity–and I think it is safe to say that the importance of trade for us will continue to grow.

Now we didn’t look at the study cited; we merely took $8,000 per household (just to be conservative) and 100 million households (92 million is actually a better number) and concluded that if you merely looked at the jobs lost to China versus the income gains, each job would have to be worth $380,000 for American to come out worse.

The problem is that isn’t the right calculus. Comparing gains from World War II onward veruss losses from trade with a single country from 1997 to 2006 is clearly apples and oranges. Doing the right computation is pretty difficult, and I have a sneaking suspicion no one has attempted it. The Bradford, Grieco, and Hufbauer study estimates income gains from World War II. You would therefore need to estimate total job losses since then attributable to more liberal trade, and come up with wage estimates (actually, wage plus benefits, most importantly health care) for those jobs. You would also need to ascertain whether there has been any loss of bargaining power due to trade (likely) and also factor that impact into the wage levels of jobs that remained.

An aside: one would likely find that the jobs lost exceeded the number of unemployed. There is substantial underemployment (determined by taking the number of males of employable age and deducting those in prison and the disabled. The math is more complicated for women, given the need to allow for childrearing), and common estimates are 20%.

It is unarguable that labor, both blue collar and white collar wages have been under pressure for a considerable time (ex CEOs and hedge fund and private equity mavens). Paul Krugman and others frequently point to the fact that average worker wages have been stagnant since the 1970s. The officialdom believes that growing income disparity is due more to technology than to globalization, but if so, why does the stall in wage gains date to the 1970s? The implementation of PCs and related technology didn’t take hold in a major way until the mid-1980s. Similarly, the corporate workers I know are all afraid of losing their jobs (not a good position to be in if you want to bargain) and working much harder than people in similar roles were a decade ago. Is this attributable to technology, globalization, or some synergistic mix of this plus other factors? I raise the question simply because I think it is very difficult to tease out, and therefore any assumption or analysis on this front is legitimately subject to question.

The biggest gap in the analysis, however, is the asymmetry between economic losses and gains. The free trade proponents act as if a dollar gained in trade can be offset on a one to one basis against income lost due to job losses. But behavioral finance tells us that ain’t so.

Even in games involving small amounts of money, researchers have found that most participants exhibit loss aversion: they don’t like to participate in games where they might lose everything they have, or worse, wind up with a loss, even if the odds are stacked in their favor. While that might be irrational when small amounts are involved, it isn’t crazy at all if you are talking about sums that are significant relative to one’s financial health.

So in other words, the average American would not take a bet that would give him a $10,000 gain if he might lose his job. Indeed, the lab examples tell us he’d need an expected gain of several multiples of his wages and benefits to willingly take that bet.

But per above, that is the tradeoff that economists tell us that more open trade is creating. That’s why people who are exposed to the real possibility of job loss are so skeptical of free trade. And as outsourcing makes people who were heretofore protected, like lawyers and bankers, also vulnerable, expect the resistance to become more broad based.

There’s a tendency to characterize the anti-free trade contingent as economically illiterate. However, the behavioral finance perspective says they are completely rational.

Print Friendly, PDF & Email

One comment

  1. Anonymous

    There’s a tendency to characterize the anti-free trade contingent as economically illiterate. However, the behavioral finance perspective says they are completely rational.


    Analyzing free trade only in terms of aggregate statistics is simplistic (to put it kindly) and doesn’t explain an individual’s assessment. Always beware the fallacy of composition.

Comments are closed.