Think Again. Free Trade is Not the Enemy Here.

After 25 years mired in a backwater of wonky acronyms — FTAs and NTBs, TPP and TTIP — trade policy is today front and center. In his first week as president, Donald Trump started by removing the U.S. from the Trans-Pacific Partnership championed by Barack Obama, and then signaled he is seriously considering a 20 percent border tax on imports from countries like China and Mexico.

I don’t want to get into a detailed accounting of the pluses and minuses of free trade agreements — they tend to be small net economic positives; often their geopolitical implications are more important. Nor do I want to go into the complex consequences of implementing a border tax — if you tax an imported component that goes into a “Made in America” car, that will hurt, not protect, American autoworkers (not to mention car buyers).

Instead, I want to make three broader points, in increasing order of importance:

The first point is the core homily of international economics. But it bears repeating, because it is true.

1.    Free trade improves living standards primarily because it lowers the cost of goods and services we buy.

Clothes and appliances are cheaper in Walmart stores in America because many of those products are sourced overseas. iPhones would be more expensive if they were made in America, rather than assembled in China from parts made in Germany, Japan and Korea.   

We have tended to pay less attention to the second point about the distributed impact of trade, as opposed to its aggregate consequences.

2.    Free trade generates (concentrated) losers who need compensation for their losses.

Political economists have always understood the asymmetric impact of trade. While the winners are dispersed (all consumers), the losers are concentrated (people whose jobs go offshore). This creates political pressures for protection. But because the aggregate gains from trade outweigh the concentrated losses, governments can maintain policies of free trade by using some of the gains to compensate the losers.

The social democratic bastion of Sweden, for example, has long been a free trader because it is good for the economy as a whole. But Swedish governments then use some of the gains from trade to compensate the losers from trade — with a generous welfare state and extensive retraining programs for dislocated workers. So free trade and the welfare state go together, not only in Sweden but also throughout northern Europe.

One could argue that the U.S. should be more like Sweden, staying the course on free trade, cushioning the costs of job losses in the short term, and giving dislocated workers the skills they need to compete.

Needless to say — and unfortunately, from my perspective — telling Americans their country should become more Swedish is a non-starter. Instead, the American-led free trade ethos is under more threat today than at any time since 1945 because regular citizens blame their economic woes on trade.

I believe this view is largely mistaken, and that is my third and most important point.

3.    Rapid technological change coupled with economic stagnation have caused more of the pain felt in Middle America than free trade.

That means the path forward for the U.S. is to stimulate more economic growth and to help more people benefit from technological change, not to close the door on trade.

Let me make this point by comparing the decade of the 1990s (when trade really took off in the U.S.) with the last decade (2006-2015). In the 1990s, inequality in the U.S. increased considerably. But people at all parts of the income distribution did better in absolute terms.

According to the U.S. Census Bureau, the real income of American households in the top 20 percent (quintile) of the income distribution increased by over 25 percent during the 1990s. This increase was more than twice as fast as for the bottom and middle quintiles, but people in the lower 60 percent of households still experienced more than 10 percent growth in real incomes.

Contrast this with the last decade. Income inequality has increased, but not as much as in the 1990s (measured by the gap between the changes in income at the bottom and top of the U.S. income distribution). Real incomes (that is, correcting for inflation to reflect purchasing power) in the top quintile only increased by 2.4 percent from 2006 to 2015. But they declined by 6.6 percent for the bottom quintile, and were completely flat in the bottom half of the income distribution.

In sum, income inequality increased more in the 1990s than it has in the past decade. The difference is that everyone in America has done less well in the past decade than they did in the 1990s. And the people at the bottom of the income distribution have become absolutely worse off since 2006.

The biggest change in the American labor market is the impact of technological change, not rising trade.

What explains these outcomes? Listening to the Trump administration, your one-word answer would be “trade.” But that answer is at best incomplete. Indeed, I think it is mostly wrong. Most economists think “skill-biased technological change” (i.e. the Internet) has more to do with the problems for people on lower incomes (and with less skills). Moreover, the biggest story in the U.S. is how poor the U.S.’s growth record has been in recent years, certainly when compared with the 1990s.

The data in the above table is from the World Bank’s World Development Indicators (my go-to source on global economic statistics). Here is what it tells me.

In the 1990s, trade (exports + imports) represented 21.5 percent of U.S. GDP — very small by the standards of developed economies, but understandable given the massive size of the American domestic market. The Internet was born in the 1990s, but only one in eight people had access to it. Annual economic growth was 3.2 percent, about a percentage point less than the 20 years after World War II, but very strong compared with what followed.

Compare this with the last decade. Trade was still less than 30 percent of GDP, three-quarters of Americans had access to the Internet, but the average growth rate dropped to 1.3 percent.

In terms of changes from the 1990s to the last decade, trade grew by about one-third — a big change, no doubt. But Internet penetration increased by 600 percent. And growth shrank by 60 percent. Both changes dwarf that for trade.

These very simple comparisons make a point that I believe is consistent with more rigorous analysis. The biggest change in the American labor market is the impact of technological change, not rising trade. And the biggest story for the standards of living of American households is the dramatic decrease in the growth rate (if anything, rising trade has increased growth).

What that means to me is that the two biggest challenges facing the Trump administration are: 1. to increase the growth rate, and 2. to help more Americans benefit from the IT revolution. Trump’s proposed tax cuts and infrastructure spending should help with growth. Things like online education should help more Americans gain the basic IT skills needed for so many jobs today.

Trade is an obvious political whipping boy. But it isn’t justified by the economics. We can only hope the protectionist impulses of the Trump administration stay more in the realm of rhetoric than reality, symbols rather than tariffs.

Geoffrey Garrett is Dean, Reliance Professor of Management and Private Enterprise, and Professor of Management at the Wharton School of the University of Pennsylvania. Follow Geoff on Twitter.