Duke University economist Ed Tower (who, I shamelessly boast, is a regular reader of Cafe Hayek) read my earlier post on Scott Sumner’s reaction to a paper by Olivier Blanchard and Jason Furman. Ed then, in response, sent to me the following e-mail, which I share here with his kind permission.
A better analogy might be a Canadian who commutes across the boundary. In November he drives his camper from Vancouver to Miami and in March he drives it back. The US has a trade deficit in November and a trade surplus in March.
Now for your example. Your initial transaction is recorded as a current account surplus for the US and since the American holds Canadian currency now as a capital account deficit. But when Smith builds a house on land he owns in Pennsylvania and sells the house to Canadian Justin for Canadian dollars that is accounted for as a surplus on direct investment and a deficit on short term investment. It doesn’t enter into the current account.
Now if in both cases the house had been paid for by Canadian maple syrup. In the first case, there would have been no net impact on the current account. In the second case the US would have recorded a current account deficit.
I tell my students that y=c+i+g+x-m, and one can use this identity in either a national or a domestic sense. Domestic means what is going on on American soil. National means what is going on with Americans. In that domestic Pensylvania house case there is no x. In the national Pennsylvania house case there is US current account surplus.
In the mobile home domestic case there is a current account surplus. In the moble home national case there is a US current account surplus.
None of this refutes your important point that the current account deficit doesn’t tell us anything useful without knowing the exact story. And of course even if we do know the exact story there is nothing wrong with the current account deficit.
Keep up the good work. Ed
I very much like Ed’s point. I did, however, send to Ed this follow-up e-mail (to which he responded “Perfect”).
If Canadian Justin sells hockey sticks to Americans for U.S. dollars and then immediately spends those dollars on a house in Pennsylvania, these transactions together raise the U.S. current-account deficit, do they not?
Separating the transactions, it is true that the U.S. current-account deficit rose the moment the Canadian received the U.S. dollars. But just as someone would correctly note that the U.S. current-account is kept balanced if the Canadian immediately spends the U.S. dollars on (say) U.S.-grown oranges, it seems correct to say that if instead the Canadian spends the U.S. dollars on a PA house the U.S. current-account deficit rises.
I don’t have time now to develop the thought that I’m about to share, but I suspect that it’s relevant to the above discussion (although further thought might prove me to be mistaken). The preamble to the thought is this: accounting, while very useful not only for conducting business but also for doing economics, is itself not economics. Accounting is record-keeping. When done correctly and wisely it helps to prevent business people (doing business) and economists (doing economics) from slipping inadvertently into double-counting or, on the other side, from failing to notice and, hence, account for some relevant flow of funds or resources. (Accounting, of course, serves other functions as well.)
And the thought itself is this: economic transactions are typically part of plans. An economic actor earns income not as an end in itself but as part of a plan of how to conduct his or her life. To summarize in perhaps a manner too simplistic, Joe goes to work as part of his plan to earn income, which itself is part of a larger plan to spend money both today (“currently”) and tomorrow (following his saving and investing). And all of these activities are part of Joe’s plan to improve his and his family’s life. The plan need not be specific. It is sufficient to realize that Joe’s waking up at 5:30am and trudging off to work cannot be understood in isolation. Joe does that for a reason – and the reason occurs later in time (namely, his receipt of a paycheck). But neither can Joe’s receipt of a paycheck be adequately understood in isolation. Instead, it is part of Joe’s plan, however vague around the edges and subject to change, to improve his material standard of living.
Economic transactions seldom (never?) occur in isolation. Each transaction is sparked by some previous transaction and – usually as per the transactors’ plans – gives rise to future transactions. Change the plans and the future transactions change. Therefore, accounting for each transaction in isolation – as if that transaction speaks for itself – is economically misleading.
For example, to properly judge if an international transaction raises the U.S. current-account deficit requires more than knowledge of that transaction in isolation. Every time non-Americans located outside of America sell goods or services to denizens of America, the immediate effect is an increase in America’s current-account deficit (or, if America had no current-account deficit, a decrease in America’s current-account surplus). But what do foreigners do with the U.S. dollars that they receive as payment for their exports to America? If they spend all of these dollars on U.S. exports during the same time period in which they earned these dollars – where “time period” is defined as a time period over which the sums of international transactions conventionally are recorded and reported (say, the month of March, 2017), the U.S. current-account deficit is neither recorded nor reported as rising during this time period – even if foreigners made all of their sales to Americans on the first day of the time period and didn’t buy anything from Americans until the final day of the time period.
Time matters. Plans matter. And because plans necessarily play out over time, usually the longer the time horizon that is accounted for, the more economically meaningful and revealing are the accounting figures.
Coda: Good accountants understand all of anything that might be correct in my above ramblings. My ramblings are addressed not to accountants but to economists, to students of economics, and to people who use (and misuse) national-income-accounting figures to assess and formulate economic policies.