≡ Menu

Do Trade Deficits Have to Eventually be Repaid?

Scott Sumner thinks that trade deficits have eventually to be repaid.  I think he’s mistaken – at least on any meaningful definition of the word “repaid.”

In practice some portion of a trade – or, current-account – deficit must be repaid.  For example, the $1,000,000 that Mr. Nonamerican lends to Uncle Sam when Mr. Nonamerican buys U.S. Treasury notes instead of buying American exports must be repaid – and when it is repaid Mr. Nonamerican will use the proceeds to buy American exports (or he’ll push the transaction forward one more ‘period’ by reinvesting the proceeds in dollar-denominated assets).

But suppose instead that Mr. Nonamerican uses the $1,000,000 that he earned on sales of goods to Americans, not to buy U.S. Treasury notes, but to open a restaurant in Fairfax, Virginia.  What is there here to repay?  Repay to whom?  And repay for what?

It’s true that Mr. Nonamerican opened this restaurant in Fairfax with the hope and expectation of earning a profit on it.  If his hope and expectation turn out to be true, then the net present value of his investment rises to some figure greater than $1,000,000.  Let’s assume that it rises after one year to $1,500,000.  And let’s further assume that Mr. Nonamerican then sells his Fairfax restaurant to Ms. American for $1,500,000 and that Mr. Nonamerican then immediately spends all of his sales proceeds on American exports.  I suspect that Scott might call this sale (and purchase of American exports) a repayment of that part of trade deficit that was created one year earlier when Mr. Nonamerican chose not to buy American exports with his $1M but instead to use all of these funds to open a restaurant in America.  (Note that I do not deny, in this example, that Mr. Nonamerican’s investment in the U.S. in year one leads to an increase in American exports in year two.)

I believe, though, that to label as “repayment” either the dollars paid to Mr. Nonamerican for his restaurant, or the American exports that he buys with these proceeds, is misleading.  First, to call it “repayment” implies that it satisfies some debt obligation.  But in this example – which is highly realistic – there is no debt involved.  No one loaned any funds to, or borrowed any funds from, anyone.

Second, the $1.5M that Mr. Nonamerican receives when he sells his Fairfax restaurant is value that he – Mr. Nonamerican – created.  He created the original $1,000,000 of value by successfully offering to American buyers the goods that he produced, exported to America, and sold here for that sum.  He created the remaining $500,000 of value by launching a restaurant in Fairfax and operating it so successfully that its value increased.  Had he operated the restaurant in a way that did not create value, the Fairfax restaurant would be worth one year later no more than, and quite possibly less than, $1,000,000 (in real terms).

Suppose that Mr. Nonamerican operated the restaurant so poorly that its value one year after he opened it is only $600,000.  He sells the restaurant to Ms. American for that sum, and Ms. American then transforms that restaurant into some other business – perhaps another restaurant or perhaps some other, entirely different business.  As in the case of Mr. Nonamerican’s successful operation and then sale of the restaurant, in the case of his unsuccessful operation of the restaurant there is no transfer of funds either made or destined to be made that is appropriately classified as a “repayment” of the original set of transactions that one year earlier raised the U.S. trade deficit by $1M.

Let’s look at third possibility: Mr. Nonamerican retains ownership indefinitely into the future of his Fairfax USA restaurant.  Presumably he keeps this restaurant open as long as he earns profits from doing so.  Again, it was the prospect of earning these profits that prompted Mr. Nonamerican to open the restaurant to begin with.  Some people – perhaps Scott – might call this stream of profits (or the American-made goods purchased with them) “repayment” of that $1M portion of the U.S. trade deficit that was created when Mr. Nonamerican chose to open the restaurant in Fairfax.  But again, in what way are these profits repayments?  They are the monetized value that Mr. Nonamerican himself created.  These profits would not exist were it not for his entrepreneurship, risk-taking, and efforts.  So anyone who insists on classifying these profits as American (re)payments to Mr. Nonamerican should recognize that Americans were able to make these “(re)payments” only because Americans were thereby enriched, at least to the tune of the value of these repayments, by Mr. Nonamerican’s operation of his Fairfax restaurant.  (Furthermore, there’s something more than passing strange about calling “repayment” that portion that becomes profit out of the voluntarily payments that American consumers pay for the food and drinks that they voluntarily order in that restaurant.)

Of course it’s true that all commercial transactions are made in the expectation of some gain, be it in money or purely in form of subjective utility.  In this sense everyone expects “repayment” for his or her participation in any commercial transaction.  The American who opens a restaurant in Fairfax expects such “repayment” no less and no more than does the non-American who opens a restaurant in Fairfax.  Likewise, the American who lends $1M to Uncle Sam expects repayment of her loan no less and no more than does the non-American who lends $1M to Uncle Sam.  The fact that accounting conventions cause some such transactions to be classified as ones that cause “deficits” or “surpluses” in artifactual accounts (here, the current account), while other identical transactions cause no such changes in any measured accounts, is economically irrelevant.


I fully agree with and endorse what I take to be Scott’s essential point – namely, that foreigners, like domestic citizens, typically do not go about with the intention of bestowing economic gifts on strangers.  Every commercial transaction is done in the expectation of some gain, where the gain ultimately comes in, or is expected to come in, the form of the receipt of real goods and services.  What I object to is labeling as “repayments” all of the expected receipts from one subset of commercial transactions, when the receipts from economically identical other commercial transactions are not labeled “repayments.”

Language matters, and in no area of economics or economic policy is the language so geared to create false impressions as it is in the area of trade and trade policy.  Foreigners want to gain from their trades with us.  We want to gain from our trades with them.  But because the term “repayment” suggests – or is taken to imply – a debt, it is simply too misleading to call the expected gains from trade “repayments.”