By William Reinsch
Over the past several weeks we have witnessed an odd kerfluffle over how to calculate the trade deficit. It began with a report, subsequently denied, that the Trump administration was considering excluding reexports from the trade deficit calculation. Reexports are items that are imported into the United States and then reexported without significant change. Currently, they are counted as imports when they come in and exports when they leave. The result is not much change in our trade balance since they both enter and exit. For some years, however, some groups, notably Public Citizen, have argued that these items should not be counted as exports, presumably because they do not contain much American content (that, incidentally, is often wrong. Current estimates are that imports from Mexico average as much as 40% U.S. content). The practical effect of such a change, of course, would be to reduce the value of our exports and make our trade deficit appear larger. Cynics would suggest that that is the real reason behind the argument for this change.
There is no question that making such a change would be a mistake, and the administration should be commended for not falling for the proponents’ argument. Counting these items as imports but not as exports would introduce a fundamental asymmetry into our trade calculations that would skew the final result. Since they both enter and leave the country, the common sense approach is to count them both times, which is what we do. Second best would be not to count them at all. Counting them once — either as imports or exports — makes the least sense.
At the same time, the debate opens the door to the larger question of whether there is a better way to count imports and exports. The answer is that there is, although it is not an easy calculation and not one that does much good if only done unilaterally. The problem is exemplified in the work done by Kenneth Kraemer, Greg Linden, and Jason Dedrick at the University of California Irvine looking at the iPod and iPhone. In brief, these products are assembled in China and enter the U.S. as products of China even though their actual Chinese content is de minimis. If we instead counted all the products from China on the basis of where their actual value came from, it is estimated that would lower our bilateral deficit with China by as much as one-third. Of course, deficits with other countries would go up correspondingly, and the overal global deficit would stay the same. Even so, this would be a more accurate and more useful calculation than what we have now because it would allow us to see more precisely the amount and source of foreign content in our products. That, in turn, could lead to less finger pointing at the wrong targets and better policy aimed at the right targets.
There is, of course, no free lunch, and there are two problems with this proposal. One is the difficulty of calculating it. While many countries have some experience with value-added calculations because they have a VAT, we do not, and one lesson we can draw from the Apple product studies is that the necessary information can be exceedingly difficult to obtain.
The second is that the calculation may not be all that useful unless everyone does it the same way. That is why the WTO and the OECD have launched a project to try to develop an effective value-added methodology that hopefully all countries could subscribe to. Global adoption is some ways off, but it is definitely a move in the right direction. Of course, that won’t help those that simply want to cook the books to make our deficit appear larger, but I remain, for the time being, hopeful that reality and accuracy will prevail. The irony is that if it does it might be partly because those who prefer ‘alternative facts’ have given the debate some momentum.
William Reinsch is a Distinguished Fellow with the Stimson Center, where he works principally with the Center’s Trade21 initiative.