China: Be Careful of Bilateral Trade Numbers

A report just came out from the US-China Business Council that seems to be getting a lot of play in the press. Among other things the report repeats a widely accepted claim that since the US no longer produces the kinds of goods that it is importing from China (the evidence for this is always anecdotal, so I have no idea if it is true), any attempt to contract the US trade deficit by getting the Chinese trade surplus to contract could not possibly succeed. Here is how the report starts:

An updated study released yesterday blaming widespread US job losses on trade with China is again based on flawed analysis and distracts from the real challenges facing the US economy and the trade relationship with China, the US-China Business Council (USCBC) said today. “The Economic Policy Institute’s latest study, ‘Unfair China Trade Costs Local Jobs’ is once again built on the faulty assumption that every product imported from China would have been made in the US otherwise. As I said two years ago, this assumption is clearly wrong–several decades wrong, in fact,” said John Frisbie, USCBC’s president.

“Think about the television in your home. The label on the back probably says ‘Made in China.’ Fifteen years ago the label likely would have said ‘Made in Japan’–but it was still an import.” Frisbie continued, “Much of what we import from China replaces imports from other countries, not products we make in the US today. A jobs impact study that ignores the facts undermines its own credibility.”

As of the end of 2008 (the latest data available), the United States was the world’s largest manufacturer–and likely remains so today. US manufacturing jobs, on the other hand, have been in a long decline over the past four decades, long before China came on the scene, and now constitute about 9 percent of total US employment. “The main reason for the decline in manufacturing jobs is productivity, not China. The US makes more with fewer people, primarily because of productivity and technology advances,” continued Frisbie.

Although I agree with much of what the report recommends, and some of what it asserts, I have to say I am more inclined to blame the SCBC for “flawed analysis” than the EPI. Mr. Frisbie claims that the EPI’s estimate of job losses is “built on the faulty assumption that every product imported from China would have been made in the US otherwise.”

I think he is making two mistakes. First, the contraction in China’s trade surplus need not be the result only of a contraction in Chinese exports. It could come about just as easily as an expansion of Chinese imports, in which case it wouldn’t matter whether Americans can produce what the Chinese produce, but rather whether the Americans can produce what the Chinese want. This line of reasoning suggests, by the way, that an RMB revaluation is probably better for the US than import tariffs on Chinese goods, since a stronger RMB should stimulate Chinese demand for imports.

Second, and much more importantly, I haven’t read the EPI report but even without reading it I am pretty sure that the EPI did not make that assumption. In fact I am reasonably sure that no one in the world has ever made that assumption. It is just too silly to be plausible.

For that reason it is also pretty easy to refute. I think this this what they call a straw dog — an argument you set up largely because it is so easy to knock down. The EPI’s real assumption, I suspect, is that adjustments in the global balance of payments occur on a multilateral basis. Rather than assume a one-for-one replacement by US manufacturers of Chinese imports, they actually assume a global adjustment in which China produces less in the aggregate and the US produces more.

If that is the case, then they are right — although before I get accused of defending the EPI report remember that this doesn’t necessarily mean that the aggregate increase in US production will match the aggregate decrease in Chinese production. China’s trade surplus can very well shrink without a corresponding shrinking of the US trade deficit, but just not for the reasons the USCBC says.

The failure to recognize that adjustments are more likely to occur via shifts in multilateral trade than shifts in bilateral trade is, I think, the basic problem with the USCBC’s report. The real flawed analysis, in other words, is likely to be the USCBC’s focus on bilateral trade between the US and China as the only thing we need to consider when looking at global balance of payment adjustments. (As an aside, I remember one of my professors in graduate school begging his students never to look at bilateral trade numbers because they generally obscured far more than they revealed.)

So, according to the USCBC, if China produces something which the US cannot or will not produce, then contraction in China’s trade surplus will have no effect on the US trade deficit. End of the story, right?

No, not at all. A shift in the terms of the trade between the two countries, whether caused by tariffs or caused by revaluation, can also tilt the balance in both countries between consumers and producers, and this will affect both countries’ trade accounts. Basically, if the cost of US imports rises, Americans will reduce overall consumption and shift their consumption behavior, and this can directly affect American producers, even though neither Mr Frisbie nor I will be able easily to predict which ones.

This can also happen internationally. Even if Chinese exports to the US are fully replaced by exports from another country, let’s call it Mexico, Mexico’s overall trade balance is determined by a number of policies and conditions, including domestic wages, domestic interest rates, and the value of the currency. These are likely to be affected by the new surge in Mexican exports to the US — Mexican wages will rise, or unemployment drop (either of which raises Mexican consumption), the Mexican peso will strengthen, and Mexican interest rates will rise, both of which affect the balance between producers and consumers. The result of some or all of these changes could easily be a surge in Mexican imports or a reduction in other Mexican exports, either of which may be good for US manufacturers.

I don’t want to overstate this argument — just to clear up some wide-spread and very muddled thinking. In my last post I pointed out that I think Beijing may be right to argue that a contraction in the Chinese trade deficit will not automatically lead to an equivalent contraction in the US trade surplus. But the reason has absolutely nothing to do with the USCBC criticism of the EPI study. There are too many other things that determine how a shit in China’s balance of payments position will affect the US balance of payments position, including whether or not a Chinese revaluation will permit other countries, especially in Asia, also to revalue.

But too many people make the same mistaken argument — why focus on the Chinese trade surplus, they say, when the US makes none of the things China makes. Even if this were true, and it probably isn’t, it is irrelevant. So even though I might agree with the USCBC’s conclusion, I wouldn’t agree with their reasoning.