Yes, Trade Deficits Do Indeed Matter for Jobs

The issue of currency management by U.S. trading partners that increases U.S. trade deficits has become a front-burner issue in debates over the proposed Trans-Pacific Partnership (TPP). The discussion about whether or not trade deficits can really affect U.S. employment, however, occasionally gets very muddled. Here’s a quick attempt to un-muddle a couple different issues.

Trade deficits and overall employment

Trade deficits occurring when the U.S. economy is stuck below full employment and at the zero lower bound (ZLB) on short-term interest rates are a drag on economic growth and overall employment, period. And this describes the U.S. economy today, so a reduction in the trade deficit in the next couple of years spurred by a reversal of trading partners’ currency management would boost growth and jobs.

The logic is simple—exports boost demand for U.S. output while imports reduce demand for U.S. output. When net exports (exports minus imports) fall, then aggregate demand is reduced. Trade deficits are the mirror image of capital inflows into the U.S. economy, and there are times when these capital inflows can reduce domestic interest rates and boost economic activity, providing an offset to the demand-drag caused by trade flows. Today is not one of those times—further downward pressure on already rock-bottom interest rates (particularly since most of these inflows go into U.S. Treasuries) do very little to boost domestic investment to counteract the demand drag from trade flows.

Yes, during times when the economy is much closer to full-employment, or when the Fed has room to counteract the demand-drag from trade deficits, one can see trade deficits that do not harm overall job-growth (thought these would still absolutely change the composition of jobs in the U.S. economy—more on that below). And yes, there is two-way causality between trade deficits and overall GDP growth—rising trade deficits can drag on growth and faster GDP growth can lead to higher trade deficits (as faster domestic growth means faster import growth, which all else equal will increase trade deficits). And yet none of this changes the basic fact that if the trade deficit was reduced in coming years by ending widespread currency management by our trading partners, the United States would see a pick-up in output and employment growth.

Trade deficits and manufacturing employment

The bulk of the U.S. trade deficit is concentrated in manufactured goods. This means that even when the U.S. economy is at full-employment that U.S. manufacturing employment is lower than it would be absent trade deficits (with jobs rising in non-manufacturing sectors like construction).

Again, the logic here is straightforward: changes in manufacturing employment are equal to the change in domestic demand for manufactured goods plus the change in net manufacturing exports (measured as a share of manufacturing output—this is essentially a measure of foreign demand for U.S. manufactured goods) and minus productivity growth. So, if U.S. consumers demand 4 percent more manufacturing output in a given year and net exports are flat and productivity growth is 4 percent, this means manufacturing employment will not change, as productivity growth allows the current manufacturing workforce to satisfy the rise in demand for manufactured goods.

However, if U.S. consumers demand 4 percent more manufacturing output, but net exports (again, measured as a share of manufacturing output) fall by 2 percent while productivity rises by 4 percent, then employment in U.S. manufacturing will fall by 2 percent. Essentially, overall demand for manufactured goods is depressed by falling foreign demand (falling net exports) and this keeps domestic production lagging behind productivity growth.

Sometimes this discussion about trade and manufacturing employment gets a bit confused when people discuss manufacturing employment as a share of overall employment, and trends over long time-periods. This manufacturing share has indeed fallen steadily (not uniformly, but steadily) over time. And over a long period, this decline is indeed mostly because productivity growth is fast in manufacturing relative to other economic sectors.

But trade has had a co-lead role (at least) in manufacturing employment decline since 2000. And the easier way to see the trade deficit’s impact is to look at the level of manufacturing employment instead of its share of the total. This manufacturing employment level has not fallen steadily over time, as rapid productivity growth in manufacturing was balanced by rapidly increasing demand for manufacturing output.

In fact, manufacturing employment was roughly stable between 17 and 19.5 million for 35 years between 1965 and 2000. Since 2000 however, mammoth losses led it to fall under 12 million at its trough. The initial fall after 2001 was due to the recession in that year. But employment levels never recovered in the following expansion because of rising trade deficits. Then the 2008/09 recession further crushed down employment levels, and the still-high trade deficit in manufacturing is a key reason for non-recovery so far.

One can see this in the figure below, which plots the level of manufacturing employment against the goods trade deficit excluding oil and agricultural goods (manufacturing is mostly what’s left there), measured as a share of GDP. The relationship here is strongrising manufacturing trade deficits are associated with smaller manufacturing employment. How much exactly has trade contributed since 2000 to manufacturing’s decline? We’re updating some previous work at EPI to assess this and will release results soon, but the answer is surely a not-trivial amount.

trade deficit manu

Yes, the relationship between trade deficits and jobs can be nuanced, but it’s really not that hard. In today’s U.S. economy, trade deficit reductions engineered by ending currency management would boost U.S. output and employment, and trade deficit reductions will (all else equal) always and everywhere boost manufacturing employment.

  • I have doubts about the 4% productivity rise postulated in this article.

    The rising US mfg productivity stats may be mostly a chimera. In her research paper, “The Debate Over the State of US Manufacturing,”

    published July 2012, economist Susan Houseman argues that computer industry rises in value-added and labor productivity mask a very serious stagnation of productivity and precipitous drop in employment in the rest of American manufacturing. And to make it worse, those computer industry rises in productivity and value-added are themselves mostly a chimera, mostly for 2 reasons: 1) they reflect statistical adjustments for increased computing power rather than any drop in the labor required to produce a computer, and 2) the price decreases per computing power also reflect increased offshoring and outsourcing of component inputs from low-cost countries. To cap it off, our merchandise trade deficits increase as as more and more computer manufacturing moves to Asia.

  • Where does the article’s 4% rate of productivity increase come from?

    • lambchowder

      its hypothetical but not unreasonable

  • lambchowder

    I’m confused. How do capital inflows bring down interest rates? I mean, I guess more jobs = more savings but that seems to be missing that they would require a lot more lending.