CEA Report Is Simply Not That Relevant to Current Trade Policy Debates

The White House Council of Economic Advisers (CEA) released a report last Friday touting the benefits of international trade for the American economy. The paper provides an interesting review of research on a range of trade’s economic effects, yet the report is largely irrelevant to current trade policy debates. Worse, when its findings are related to current trade policy debates, they are often reported in ways that could mislead readers.

The weaknesses of the report generally fall into one of three areas.

First, the overwhelming focus of the report touts the benefits of trade flows qua trade flows, and often even compares outcomes relative to a hypothetical scenario where trade barriers were raised so high that the U.S. economy became completely autarkic. Academics might find this interesting, but nobody in today’s economic debate argues for increasing U.S. trade barriers, let alone to historically never-seen levels. The CEA acknowledges this explicitly by noting that barriers to foreign imports coming into the U.S. economy are already extremely low and unlikely to be reduced significantly by treaties like the Trans-Pacific Partnership (TPP). Several times, the report alludes to potential benefits of the TPP and other treaties in pulling down barriers to U.S. exports abroad, but fails to mention what is by far the most important barrier to U.S. export success—several major trading partners (including some proposed TPP partners) managing the value of their own currencies for competitive gain vis-à-vis the United States.

Second, the report spends very little time on the most important non-currency issue regarding trade policy: the distribution of gains and losses. When the report does cite research on distribution, it is woefully incomplete, looking only at how the benefits from trade are distributed while ignoring the costs. The research on the comprehensive costs and benefits of this issue is pretty clear:  trade with labor abundant trading partners, like many of those in the proposed TPP, tends to lower wages for the majority of U.S. workers and provide gains only to the upper end of the income distribution.

Third, the report makes strong claims that “trade” has been valuable for improving living standards in many poor countries around the world. Yet it fails to acknowledge that it is largely countries that have been most resistant to simply following the prescriptions embedded in trade agreements like the TPP and in trade regimes like the World Trade Organization (WTO) that have seen the largest successes in leveraging global integration for the benefits of their citizens.

With these larger points in mind, we can see how several of the bulleted claims made in the report’s first page are largely irrelevant to current debates regarding trade policy.

CLAIM: U.S. businesses must overcome an average tariff hurdle of 6.8 percent, in addition to numerous non-tariff barriers to serve the roughly 95 percent of the world’s customers outside our borders.

Tariffs are a much smaller barrier to U.S. exports than currency management by our competitors. If the goal for U.S. policymakers in crafting trade agreements is to improve U.S. export competitiveness (and this is indeed a worthy goal), then it is odd not to address the single biggest barrier to U.S. exports—our trading partners’ practice of managing the value of their currency for competitive gain. A trade agreement really focused on the need to balance U.S. trade and help U.S. exporters would include a strong currency provision.

CLAIM: Exporters pay higher wages and the average industry’s export growth over the past twenty years translated into $1,300 higher annual earnings for the typical employee.

This is misleading. First, it ignores imports. While there is a wage premium for workers in export sectors, there is also a premium for workers in import-competing sectors relative to non-traded sectors. So, when the United States runs a trade deficit, this pushes more jobs into non-traded sectors, where wages are lower. Again, for trade policy the relevant issue is really whether or not we allow total traded-sector (i.e., exportable and importable industries) employment to be harmed by the current rules of the game governing trade. And since these current rules of the game do nothing to stop the currency management that leads to large U.S. trade deficits and lower traded sector employment in the United States, we do indeed harm wages this way.

Furthermore, the research that highlights correlations between exports and higher wages does not pin down causation that runs from exporting to higher wages. Instead, higher wages are often a by-product of the fact that more productive firms have greater export success.

Finally, the comprehensive impacts on wages of expanding trade with more labor-abundant trading partners are a well-studied issue in economics. The result for the United States is simply that such expanded trade leads to lower wages for the majority of U.S. workers. This is driven largely by what happens as workers (particularly those without 4-year college degrees, who constitute 70 percent of the workforce) are displaced from traded sectors by imports. This boosts the relative supply of non-college labor to non-traded sectors as these displaced workers compete to find employment there, and wages subsequently decline for all workers who resemble those displaced by imports in terms of formal credentials. In short, waitresses and landscapers don’t lose their jobs to imports, but their wages are hurt by having to compete with laid-off apparel workers.

CLAIM: Middle-class Americans gain more than a quarter of their purchasing power from trade.

This is deeply misleading and irrelevant to current debates. For one, remember, nobody is talking in current debates about increasing barriers to trade, so there is no threat to any gain Americans have received from imports. Further, remember that even the CEA argues that the TPP will do little to affect the price of imports (because barriers to imports coming into the United States are already low). The cited statistic claims to measure how much middle-class Americans gain relative to a counterfactual scenario where the United States imposed trade costs so high that imports fell to zero. This is a very extreme and irrelevant counterfactual.

For another, the study that generates this number is highly stylized. It does not measure the expenditure patterns of U.S. consumers directly, instead it infers them from aggregate data making a number of assumptions along the way. Further, the reported results are odd in that they seem to promise benefits for households up and down the income distribution that add up to a substantially greater benefit than is reported for the entire economy.

The study reports the result that the overall gains from trade stemming from cheaper imports for U.S. incomes equal 5.6 percent of total U.S. income. But the 10th percentile U.S. household sees 62 percent gains, the median household sees 29 percent gains, and even the 90th percentile sees 3 percent gains. The math on this is very odd. The bottom 50 percent of American households account for just under 30 percent of all consumption. If trade raised their incomes by an average of 29 percent (which is far too conservative an extrapolation of the study’s results—remember that it claims households at the 10th percentile saw 62 percent gains), then just the bottom 50 percent would contribute about 8.5 percent to overall income gains, which would more than exhaust the estimated overall benefits. This incongruence between overall benefits and benefits reported at various levels of the income distribution makes this a very hard number to properly interpret.

Finally, the study that generates this number only focuses on the benefits of trade—namely, lower prices for consumers. It makes no attempt to look at the costs of trade—lower incomes for producers and wage declines for the majority of workers. The majority of American workers have seen wage declines (not captured in the aforementioned study which only examined consumption patterns, not earnings patterns, of households) from increased trade with labor-abundant trading partners.

It is always true that if one only tallies the benefits of trade rather than both benefits and costs, one can generate large numbers showing that trade is good for everybody.

CLAIM: Over the past 20 years, the average industry’s increase in exports translated into 8 percent higher labor productivity, or almost a quarter of the total of productivity increases over that time… According to other estimates, the reduction in U.S. tariffs since World War II contributed an additional 7.3 percent to U.S. GDP.

On the first claim, these estimates again do not cleanly estimate the causal effect of policies that foster more trade. Instead, they could well simply be finding that industries with more productive firms tend to have greater export success. The second claim rests on a deeply flawed and exaggerated estimate of the gains from trade liberalization, which I have previously examined.

CLAIM: Trade raises labor standards and incomes abroad, helping developing countries lift people out of poverty and expanding markets for U.S. exports.

This is broadly true, but again largely irrelevant to today’s trade policy debates. The countries that have been the most successful in using global integration to lift their populations out of poverty have been largely those that have refused to restrict their own policy autonomy by signing trade agreements like the TPP, and which have felt free to flout rules governing trade embedded in institutions like the World Trade Organization. More importantly, the vast majority of the benefits from increased trade that could be obtained by poorer trading partners of the United States could be obtained unilaterally by them. That is, the benefits stem largely from cheaper import prices that these countries can achieve anytime they want by simply unilaterally reducing tariffs. They simply don’t need the TPP to realize these gains.

Crucially, the TPP actually contains measure to restrict trade and increase the prices of imports for many the members. This is through provisions that seek to increase the protection of intellectual property monopolies held by (mostly) U.S. corporations. The CEA report explicitly acknowledges this, noting that: “Branstetter, Fisman and Foley (2006) show that strengthening foreign intellectual property protection leads to more outward licensing from the United States.” By boosting the amount of money foreign consumers have to spend per unit of intellectual property, the TPP will likely restrict the demand for other U.S. exports in foreign markets.

To sum up, it seems inarguable that the United States is richer today than it would be in an alternative history where we had instead erected prohibitive barriers to trade and never imported a dollar of anything. But this is an uninteresting finding and one not relevant to today’s trade policy debates. What are much more contested are claims that changes to trade policy like those embedded in the TPP will make the majority of Americans and the majority of citizens in trading partner countries better-off. They likely will not.

  • SS50

    We’ve run about $10 trillion in trade deficits since NAFTA – roughly a $500 billion drain on our economy per year – 2-4 million jobs foregone. Workers have no bargaining power and employers have expanded their bargaining power.

    That is a boundary condition on any analysis of bad trade policy, and the starting point for any discussion of good trade policy.