Employment in the United States has shifted away from goods production to services. During that shift, the U.S. balance of trade has been negative, leading many to blame international trade for the decimation of U.S. manufacturing jobs.

In Washington, some politicians react by proposing protectionist policies. Their explicit assumption is that imposing trade taxes and creating other barriers will rejuvenate industrial employment. That’s a very tenuous proposition, as is explained below.

But it’s also important to understand that debates over protectionism have major implications for global capital flows. That’s because a nation with a trade deficit, by definition, has a capital surplus. As a result, if politicians succeed in reducing the amount of tradeable goods entering a nation, they also will reduce the amount of capital flowing into a nation.

Keeping in mind the link between trade and capital flows, let’s consider the debate over international trade, which is often based on concerns that the growing U.S. trade deficit has a negative impact on jobs. When imported goods are substitutes for domestically produced goods, it is easy to see why a trade deficit can be detrimental to goods sector employment.

However, in the United States, declining goods sector employment was accompanied by an increase in employment in other sectors. To what extent are trade deficits actually responsible for the decline of U.S. manufacturing employment? Will employment return to manufacturing and to other goods producing industries when the U.S. trade deficits become trade surpluses?

President Trump’s promise to Make America Great Again included protectionist trade policies meant to bring back U.S. manufacturing jobs. Trump’s whole view on trade is that other people and countries are taking advantage of us. China and Germany were branded currency manipulators to gain an unfair advantage in trade with the U.S. Germany’s response was that the size of its trade surplus is largely a consequence of factors beyond its control, such as the price of oil and the value of the euro, as well as the ability of its companies to compete on the world stage. By penalizing imports that compete with U.S. goods, the Trump administration hopes to reverse the secular shift of employment from goods-producing industries to services-producing industries. The commonly held view among Trumpians and also in policy circles is that reversing U.S. trade deficits generated by U.S. households and government borrowing should cause labor to flow back to goods-producing industries (see, Bivens, 2006 and Scott, Jorgensen, and Hall, 2013). See figure 1

As the United States trades bonds for foreign goods, labor shifts away from domestically-produced goods and is reallocated to the services and construction sectors, which are less substitutable for foreign goods. The implication of this idea is that labor should eventually flow back into the U.S. goods sector to produce the extra goods needed to repay the debt.

Caliendo, Dvorkin and Parro (2015) find that the China trade shock in the early 2000’s resulted in a loss of 0.8 million U.S. manufacturing jobs, about 25 percent of the observed decline in manufacturing employment from 2000 to 2007. Service industries benefited from cheaper intermediate inputs imported from China leading to a rise in service sectors job creation. The China trade shock had heterogeneous effects across different labor markets and while some labor markets were adversely affected, most regions and sectors gained from the increased commerce with China, and overall welfare gains were sizable.

Other examples of sectoral reallocations associated with a trade imbalance include Spain after joining the European Community in 1986 when output in Spain’s traded sectors fell by more than 10 percent (Cordoba and Kehoe, 1999). In the 1990’s, access to international capital markets had a similar effect on the Baltic countries (Bems and Hartelius, 2006).

Lastly, a sudden retreat was Mexico’s experience which was accompanied by a shift away from services back towards goods producing sectors (Kehoe and Ruhl, 2009). See figure 2
Another factor affecting manufacturing jobs is technological innovation. Technological advancement boosts productivity and output growth in both the short and long term (see, Brynjolfsson and Hitt, 2003). In fact, productivity and output growth of firms were up to five times greater over the long-term, due to technological innovations. However, these leaps in productivity also result in fewer jobs, especially for blue collar workers. Economists agree that the biggest threat to manufacturing jobs has been automation, not trade, offshoring and immigration. Although imports have had a negative effect on jobs in parts of the country (see, Autor, Dorn and Hanson, 2016), automation has had a bigger effect than globalization as fewer workers are needed to do the same amount of work.

Was it trade or automation that led to the decline of manufacturing employment? Using evidence on sector-specific labor productivity, Kehoe, Ruhl and Steinberg (2017) estimate the impact of differential productivity growth on the decline in goods-sector employment in the U.S. KRS (2017) highlights three important facts: first, between 1992 and 2012, labor productivity in the goods sector grew at an average of 4.2 percent per year, compared to only 1.3 percent per year in services. Second, the goods trade balance generates most of the fluctuations in the aggregate trade balance since the United States has consistently run a trade surplus in services. Lastly, while reallocation away from goods and into services has been consistent, reallocation into construction was temporary, demonstrating that a shift into one sector from another sector can be reversed.

In addition, KRS (2017) assesses quantitatively the relative contributions of asymmetric labor productivity growth and the saving glut – the increased demand for saving in the rest of the world that made foreigners more willing to trade their goods for U.S. bonds. KRS (2017) finds that U.S. trade deficits only accounts for 15.1 percent of the observed decline in the employment share of goods producing sectors from 1992 to 2012. The bulk of the remainder is attributed to faster productivity growth in goods producing sectors compared to other sectors of the U.S. economy. The U.S. sectoral reallocation is driven by differential productivity growth and not trade, implying that as the United States repays its debt, its trade balance will reverse, but goods-sector employment will continue to fall.

Lastly, empirical evidence suggests that, all else being equal, openness leads to more competition that lowers firm costs. Trade liberalization encourages knowledge transfer and technological advances that raise productivity and the standard of living. Bussiere et al. (2011) shows that protectionist measures harm real GDP growth and the competitiveness of those implementing the measures, as well as the countries being targeted.

The shift in American jobs away from manufacturing and into the service sector that can be attributed to foreign trade is small. However, the rewards from international trade have been reaped, directly or indirectly, by all Americans. The primary driver of this shift was technological innovation that made labor more productive, raised living standards and caused employment in the service sectors to grow by more than the overall decline of manufacturing jobs. See figure 2

Economists widely agree that trade liberalization has permanent positive effects on economic growth. On the other hand, protectionist policies that restrict foreign trade hinder economic growth by raising production costs, slowing the accumulation of productive resources and the propagation of new technology. Protectionism is unlikely to override automation and to reverse the declining trend in manufacturing employment. To enact protectionist policies in an attempt to bring back long-gone manufacturing jobs would be another economic faux pas by the Trump administration since it would reduce U.S. competitiveness and slow economic growth.

And if protectionist policies are enacted, and if they are “successful” in reducing the amount of foreign goods purchased by Americans, that automatically means foreigners will be investing less in the U.S. economy. That distortion in capital flows would be an underappreciated and underrecognized source of risk to the American economy. Moreover, if U.S.-instigated protection triggers similar moves by other nations, the result could be 1930s-style tit-for-tax protectionism that threatens the global economy.