This academic research changed my mind about China and trade policy

Years ago, I heard an argument from a guy who later went on to serve at a high level in the Obama administration that I thought was really smart and persuasive: American politics massively overstates the importance of trade policy in causing the big shock to the American economy that was induced by foreign manufacturing in the early 21st century.

The reason you can tell is that the big change in trade policy regarded Mexico (NAFTA, in other words), but the big change in terms of actual trade regarded China and the United States didn’t do anything to lower tariffs or other barriers to Chinese imports. The result is a kind of nonsensical deadlock in which people talk a lot about NAFTA, which was a real policy change, but are mostly angry about second-order consequences of trade with China, the country with which the United States runs far and away the biggest trade deficit.

The problem, on this account, is that Chinese exports surged primarily because China got better at making stuff, not because of anything that happened in American politics.

I found this convincing and have believed it for years. I’ve repeated it to other people, some of whom may also have found it convincing. And over the weekend I read a piece of research that convinced me I’ve been wrong this whole time. We didn’t lower tariffs on Chinese goods, but we did make a subtle policy change right at the end of the Clinton administration that’s made a big difference.

Here’s the subtle policy shift. Starting in 1980, Chinese-made goods were taxed at the relatively favorable rates applied to countries that enjoyed what was known in legal terms as normal trade relations with the United States. That status had to be affirmatively extended each year by Congress or it would expire, and it was extended each and every year for 20 years.

Starting in 2000 that changed, and China was granted permanent normal trade relations. That meant it would automatically keep the low tariff rates it had enjoyed since 1980 unless Congress affirmatively took them away.

Clearly, that’s a change. But since Congress really had routinely approved NTR for two decades, it never seemed to me like a particularly large change. Certainly not a big enough change to explain the wholesale transformation of the industrial landscape that’s taken place in the 15 years since it happened. Clearly the action must have been things that happened in China rather than this extremely modest shift in US trade policy.

But Justin R. Pierce from the Federal Reserve and Peter Schott from Yale’s School of Management have convinced me this is wrong.

In their paper “The Surprisingly Swift Decline of US Manufacturing Employment,” they deploy two statistical tests that reveal that the modest-sounding change was a really big deal in practice. They say the effect is plausibly large enough to explain the 18 percent decline in manufacturing employment that occurred between March 2001 and March 2007.

Pierce and Schott’s primary strategy is to exploit the fact that the gap between the NTR tariff rate and the non-NTR tariff rate varies from industry to industry. If the switch from NTR to PNTR were a big deal, we would expect to see bigger changes where the gap between the NTR and the non-NTR rates was biggest.

And, indeed, their regression analysis reveals “a negative relationship between the change in U.S. policy and subsequent employment in manufacturing that is both statistically and economically significant.” This change remains significant even when they control for changes in Chinese policy.

They also exploit the fact that the European Union, another large, high-wage economy, gave China its equivalent of PNTR many years earlier and had no significant policy change in 2000. They find “no relationship” between “the US NTR gap and EU manufacturing employment,” confirming that the correlation that exists in the American context isn’t spurious.

So why did American trade policy make such a big difference, even though tariffs didn’t fall? Pierce and Schott can’t say for sure, but they speculate that it has to do with patterns of investment.

Giving Chinese companies confidence that tariffs would stay low encouraged them to invest in production capacity aimed at supplying the American market. And giving American companies confidence that tariffs would stay low encouraged them to build supply chains around Chinese manufacturers.

Shifting production to Asia to save a few bucks on each item sold, in other words, might not make sense if there was a chance that you’d have to shift it back a year or two down the road. The old practice of keeping tariffs low but requiring the low-tariff status to be renewed on an annual basis gave US-based producers a significant amount of hidden protection. Despite the superficial appearance of openness to imports, the temporary nature of the low tariffs discouraged companies from making plans built around that openness.

Twenty years of repeated extensions of normal trade relations had only a relatively modest impact on the American economy, but making NTR status permanent led to a very rapidly displacement of American manufacturing work by Chinese imports in a way few PNTR proponents anticipated or have even acknowledged.