A few minutes ago I sat down to write a short explanation of why Olivier Blanchard and Jason Furman are deeply mistaken about the nature of trade deficits (and, hence, mistaken also in the reasoning they use about the balance of trade to analyze a border-adjustment tax [BAT]). But as I began to formulate my blog post, I noticed that Scott Sumner, over at EconLog, criticizes Blanchard’s and Furman’s analysis. I endorse all that Scott says. (By the way, Scott’s criticisms apply also to Martin Feldstein’s flawed defense of the BAT.)
Here’s the most egregiously mistaken sentence in Blanchard’s and Furman’s paper:
But if foreign debt is not to explode, trade deficits must eventually be offset by trade surpluses in the future.
As Scott writes in response:
This is flat out wrong, and it’s not even debatable. The official trade deficit does not measure the increase in net indebtedness; as a result the measured U.S. trade deficit can (and likely will) go on indefinitely.
What Scott says.
It continues to baffle me – and, I suspect, it baffles also Scott – that so many economists persist in asserting that every $1 increase in a country’s trade (or current-account) deficit necessarily means a $1 increase in the indebtedness of the people of that country. Scott’s example of why this persistent assertion is mistaken is clever and very useful: according to the conventions of national-income accounting, a large class of what are in reality exports and imports are recorded, not on the current account, but on the capital account. The large class of exports and imports that Scott identifies as being treated this way is real estate.
Here I steal the gist of an example of Scott’s. When American Jones builds eight mobile homes at his Pennsylvania factory, ships these homes to Canada, and sells them, for cash, for a total of $750,000, that transaction is recorded on the U.S. current account as $750,000 of U.S. exports (and on Canada’s current account as $750,000 of Canadian imports). (“Yippee!” shout the Trumpians! “Exports are good! Take that, Canada!”) But when American Smith builds a house on a parcel of land that he owns in Pennsylvania and then sells the house and land, for cash, to a Canadian buyer for $750,000, this transaction is recorded, not on the U.S. or Canadian current accounts, but instead on the U.S. and Canadian capital accounts. This $750,000 shows up in the U.S. trade statistics, not as $750,000 of U.S. exports, but as $750,000 of foreign investment in America. And so it swells the U.S. capital-account surplus by $750,000 – which means that it also (as a matter of accounting) necessarily swells the U.S. current-account (“trade”) deficit by $750,000. (“Omigosh!” scream the Trumpians! “The trade deficit is bad! Do something! Canada is killing us!”)
Yet even even quick reflection on these two different examples reveals that (1) they are economically identical in all essential respects, and (2) in the example in which the U.S. trade deficit rises, Americans’ indebtedness is no more increased than it is in the example in which the U.S. trade deficit did not rise. American indebtedness rises in neither example.
Repeat after me: A TRADE DEFICIT IS NOT SYNONYMOUS WITH INCREASING INDEBTEDNESS. It’s just not – yet there seems to be, even among many competent economists, a powerful will to overlook this reality.