Mitchell’s Musings 4-2-2018: Identity vs. Behavior

28 Mar 2018 9:57 PM | Daniel Mitchell (Administrator)

Mitchell’s Musings 4-2-2018: Identity vs. Behavior

Daniel J.B. Mitchell

As we resume these musings after the end of the winter quarter at UCLA (when I am teaching and too busy to write them), it’s worth looking back at events of that period. And it’s worth looking back at some confusions that arose in that period, too. One of those events was the announced imposition of steel and aluminum tariffs. Let’s put aside the question of whether those tariffs were a good idea, either politically or economically. There has been much confusion in the popular news media as experts – who generally opposed the tariffs - were interviewed. From Forbes:

…Our deficit will not be reduced by enforcing trade rules, renegotiating existing trade pacts, or forming new ones. While these policies might shuffle the deck and alter the bilateral trade balances the U.S. has with other countries, they will not alter the overall U.S. trade balance. Indeed, policies aimed at eliminating so-called “unfair trade practices” — while attractive to many businessmen, trade unionists, most progressive activists and all who harbor mercantilist sentiments — are wrongheaded and futile.

Why? The simple analytics of the trade deficit prove the utter futility of the Trump administration’s trade policies. In economics, identities play an important role. These identities are obtained by equating two different breakdowns of a single aggregate. Identities are interesting, and usually important, by definition. In national income accounting, the following identity can be derived. It is the key to understanding the trade deficit:

(Imports - Exports) ≡ (Private Investment - Private Savings) + (Government Spending - Taxes)

Given this identity, which must hold, the trade deficit is equal to the excess of private sector investment over savings, plus the excess of government spending over tax revenue. So the counterpart of the trade deficit is the sum of the private sector deficit and the government deficit (federal + state and local). Therefore, the U.S. trade deficit is just the mirror image of what is happening in the U.S. domestic economy. If expenditures in the U.S. exceed the incomes produced in the U.S., which they do, the excess expenditures will be met by an excess of imports over exports (read: a trade deficit)…[1]

The implication of this observation seems to be that if the ultimate objective of the Trump administration was to improve the U.S. trade balance (make it less negative than it is), raising tariffs is a fool’s errand. If you don’t change saving and investment behavior, the trade deficit will remain the same. Similar objections have been raised when the question of doing something with exchange rates is the issue.

It is true that as a matter of national income accounting, the balance of saving vs. investment is the reverse of the trade balance. One is the mirror image of the other. So it is true that if you raise tariffs while saving and investment remain the same, the trade balance will be unaltered. If the tariffs you raise cut down on imports, but the trade balance is unaltered, then it must be the case that exports will have (somehow) dropped by the same amount. Again, this observation is pure accounting. The observation is true by construction.

You can say the same thing about exchange rates. If the dollar is devalued and imports fall as a result – and if saving and investment are unaltered – then it must be the case that exports will fall by the same amount.

But there is a fundamental problem here. Who says that changing tariffs in a big way, or changing exchange rates in a big way, will leave saving behavior and investment behavior unchanged? Perhaps a thought experiment will be helpful. Imagine we raise tariffs on everything and keep raising them higher and higher. Imported goods and their domestic substitutes are thus becoming more and more expensive, both relative to non-traded goods, e.g., haircuts, and relative to exported goods. And imagine that saving and investment are unaltered. Surely, at some point – maybe a 500% tariff, maybe a 1000% tariff – imports become prohibitive. If we started with a negative trade balance and now imports are reduced to zero, in order to preserve the negative trade balance, exports would have to be negative! But there is no such thing as a negative export. Something is clearly wrong here. And what is wrong is the idea that tariffs (or exchange rates) don’t affect saving and investment.

Here’s a different scenario, one more realistic than the extreme case above, but also more realistic than the assumption that saving and investment are immutable. Suppose there is a big increase in tariffs (or a big dollar devaluation). Prices tend to rise a result. Particularly if the economy is at or near full employment (as it currently is), the Federal Reserve will raise interest rates to resist the price effect. Investment will likely fall. It’s likely that saving will rise.

In past musings and elsewhere, I have noted that if the Trump administration really wants a zero-trade balance, it could achieve it with the Warren Buffett plan – a kind of cap-and-trade approach that relies on a market mechanism to bring about balanced trade.[2] If that were done – and there is no sign that my advice is being taken – since exports would be equal to imports – there would also be behavioral responses in saving behavior and investment behavior such that in the end, saving would = investment.

To repeat, this musing is not focused on whether Trump policies with regard to tariffs on steel and aluminum are wise. However, I will say that those tariffs appear to be based more on political symbolism than anything else. And I am not saying that pointing to national accounting identities is useless. For example, you can certainly argue, based on the identities, that the Trump tax cuts – which create governmental dissaving – will tend to worsen the trade balance. For them not to do so, they would have to raise private saving (by some magic) or reduce investment (which Trump doesn’t want). But that is another story. The simple point here is that in evaluating any major policy change, you can’t assume that key macro variables such as the rates of saving and investment are going to be unaffected. Nor can you ignore likely responses from policy actors such as the Fed.





Employment Policy Research Network (A member-driven project of the Labor and Employment Relations Association)

121 Labor and Employment Relations Bldg.


121 LER Building

504 East Armory Ave.

Champaign, IL 61820


The EPRN began with generous grants from the Rockefeller, Russell Sage, and Ewing Marion Kauffman Foundations


Powered by Wild Apricot. Try our all-in-one platform for easy membership management