Fiscal Reflation in One Country

President-elect Trump wants to cut taxes and increase investment in U.S. infrastructure (or at least provide a tax break for existing infrastructure investment) and doesn’t seem especially worried if the result is a larger fiscal deficit.

The call for larger fiscal deficits has some parallels to the agenda I think makes sense for balance of payments surplus countries like Germany, or Korea—though I have always advocated for more progressive tax cuts than those proposed by President-elect Trump, and wanted East Asia to use its fiscal space to finance an expansion of social insurance.

But just as fiscal expansion should reduce the external surpluses of those countries that now run sizable balance of payments surpluses, fiscal expansion in a country with a sizable balance of payments deficit, in any conventional macroeconomic model, implies a bigger balance of payments deficit.

The numbers on the fiscal side could be significant. Michael Feroli of J.P. Morgan estimated that full implementation of the proposed corporate and individual tax cuts, together with higher spending on infrastructure and defense, would raise the fiscal deficit by about 3 percent of GDP if they were adopted in full (see Gavyn Davies blog); Ryan Avent of The Economist has a similar estimate.

A 3 percent of GDP fiscal expansion, according to the IMF’s coefficients, raises the external deficit by about 1.4 percent of GDP (the coefficient on fiscal deficits was raised to 0.47 in the IMF’s latest external balance assessment; see table 3).* A lot of the rise in U.S. demand from a fiscal stimulus would be shared by the rest of the world. With the U.S. economy operating at close to potential, the fiscal stimulus would lead to the Fed to raise rates -- and that in turn would push up the dollar.

There is no immaculate adjustment. A 1.4 percent of GDP increase in the current account deficit is consistent with a 10 percent or more rise in the dollar (see Joe Gagnon, quoted in Neil Irwin’s well-argued analysis of the tensions between President-elect Trump’s various goals).

Of course, many keen political and economic observers (including the previously cited Michael Feroli) expect that the actual fiscal package will be far smaller—maybe one percentage point of GDP a year over several years.

That would scale down the impact, not eliminate it.

And I can think of one reason why the rise in the U.S. external deficit could be smaller than implied by the simple application of the standard global coefficient for estimating the balance of payments impact from a fiscal stimulus, and three why it could be bigger.

It could be smaller because the proposed fiscal expansion includes a lot of low-multiplier tax cuts. Tax cuts to upper-income groups (and firms) are likely to be saved, not spent. Higher private savings would offset the fall in public savings; the fiscal deficit would rise—but import demand would not go up by as much (see Jared Bernstein and the FT’s Martin Sandbu, among others).

But the increase in the U.S. external deficit could also could be bigger than expected for a few reasons.

(1) The Federal Reserve has noted that in the current global context—with lots of savings globally and low (or negative) policy rates in many surplus countries—relatively small changes in U.S. interest rates can have a relatively big impact on the dollar. That should amplify the trade impact of any increase in U.S. interest rates—either a change in the expected path of the Federal Reserve or a rise in the term premium -- from a large fiscal stimulus.

Federal Reserve Governor Lael Brainard in September:

"... recent research suggests that changes in expectations regarding the path of policy in the United States relative to other major economies lead to exchange rate movements that appear to be several times bigger than they were several years ago."

Bigger exchange rate moves, of course, imply more pressure on the parts of the U.S. economy that specialize in the production of traded goods.

(2) The IMF’s standard model also may not capture the dynamics of the U.S. income balance (the current account is the balance on interest and dividend payments to/from the rest of the world as well as the balance on trade).

The U.S. income balance in the current account is distorted by the large (currently tax-deferred) profits earned by U.S. firms overseas. Those profits drive the surplus in the income balance, as foreign firms operating in the U.S. report surprisingly small profits relative to the tax-deferred profits U.S. firms report offshore.** But set that aside for a moment. The rest of the income balance is fairly straight-forward. The United States essentially makes the interest payments to the world that you would expect from a country that has run an external deficit for a long time, and now has a substantial external debt stock.

Net debt to the world is about 50 percent of U.S. GDP. The average rate now is about two percent, leading to—in ballpark terms—interest payments on the U.S. external debt of about one percent of U.S. GDP.

Over the past few years the fall in the nominal rate has pulled down interest payments even as the stock of external debt has increased. Should both short-term and long-term rates start to rise the dynamics will shift. And the math is pretty simple: every percentage point increase in the nominal rate raises U.S. interest payments to the world by about 50 basis points of U.S. GDP.

(3) Oil, and its own unique supply dynamics. Modelling the oil balance separately from the overall balance of payments always yields better results.

Right now, parts of the U.S. oil industry are in effect a global swing producer (if estimates of the global cost curve are right, the U.S. will produce a lot more oil if oil is at $60 than if oil is at $40). Right now, U.S. oil demand is rising and U.S. production is falling. 2016 crude oil production will be smaller than 2015 production by about 0.6 million barrels a day, according to the Energy Information Administration (natural gas liquids production is still rising). In the third quarter, ever so imperceptibly, the U.S. real petrol deficit started to inch up (data here, see exhibit 11). This dynamic should be more obvious in the next few quarters of data. While U.S. production has not fallen by as much as might have been expected 18 months ago thanks to the fall in the U.S. cost curve, U.S. production growth is unlikely to keep up with U.S. consumption growth—so a big driver of the U.S. improvement in the U.S. current account since 2010 should go (modestly) into reverse.

Add these factors up, and it seems reasonably likely the U.S. current account deficit will be back above four percent of U.S. GDP within the next couple of years if there is a significant fiscal loosening. And the U.S. (goods) trade deficit could easily approach six percent of U.S. GDP. Concretely, if the adjustment comes from a fall in exports, that is well over a million lost jobs in the traded sector (the Commerce Department’s estimates imply a percentage point of GDP in exports supported about a million jobs in 2015). The domestic part of the U.S. economy would do well, but not the outward facing bits.

In order to get both reflation and a rebalancing that favors the U.S. traded goods sector (and the price sensitive parts of services), the tax cuts and new infrastructure spending really needs to come from countries that are now running external surpluses: Germany, the Netherlands, Sweden, Japan, Korea, Taiwan, and the like.

A stable rather than depreciating yuan and a somewhat stronger Mexican peso also would be a big help. The yuan and the peso together account for about a third of the broad dollar index.

Note: edited after posting for clarity; I used "adjustment" without clarifying that the adjustment I had in mind was a bigger external deficit.

* The IMF’s estimated coefficient is a bit on the high side in my view, but it is hard to get estimates of less than 0.3.

** Technically, tax-deferred profits show up as reinvested earnings in the balance of payments. The gap in actual cash dividends in the income balance is small. But profits earned abroad and held abroad are reported as an income payment to the U.S. parent (a positive in the current account) that is reinvested abroad (in the subsidiary).