The purported benefits of the U.S.-Mexico Canada Agreement (USMCA, or NAFTA-2) for American workers are so tiny, one can hardly see them.

The U.S. International Trade Commission’s recent study of the economic impacts of the USMCA finds that it will have small, but positive, effects on U.S. output (GDP up 0.35 percent over six years), employment (176,000 jobs or 0.12 percent) and wages (up 0.27 percent). However, these projections are based on a number of questionable assumptions about the impacts of the trade deal, “assuming” for example that Mexico will adopt new labor legislation that will improve labor rights in that country, and “that these provisions are enforced” and Mexican union wages increase by 17.2 percent as a result. Furthermore, the ITC claims that U.S. wages will rise as a direct result of improved labor rights enforcement in Mexico, although that conclusion is not supported by the results of their own model.

These findings illustrate a much larger problem with the outdated modeling approach used by the ITC, which assumes that the purpose of trade and investment deals, such as the USMCA, is to reduce tariffs. However, the most important provisions of modern international economic agreements, such as the USMCA and the World Trade Organization, lay down rules governing matters such as foreign investment, services trade, government procurement, data transmission and storage, food and product safety standards, as well as labor rights and environmental standards. These rules govern how countries trade and businesses invest and how our economies are governed and regulated. At the end of the day, they determine who wins and loses, how income is distributed, the tradeoffs between corporate power and control, and whether the rights of workers, the public and the environment will be protected from transnational abuses from big business and big government.

Chapter 8 of the ITC report on the USMCA (p. 215) makes the following erroneous claim: The Commission estimates that the collective bargaining legislation will likely increase unionization rates and wages in Mexico and also increase Mexican output. This, in turn, would be expected to increase U.S. output and employment also, resulting in a small (0.27 percent) increase in U.S. real wages to attract the new workers.

This claim is not supported by the model results. Appendix F of the ITC report (Modeling the Labor Provisions, Table F.5 (p 327)) reports the results of a sensitivity analysis showing the impacts of various assumptions about the size of the Mexican union wage premium (17.5 percent, 32.7 percent, and 37.5 percent) on US macroeconomic variables, including GDP, total employment and wages. The first of these is the base case for the ITC’s overall estimates. These simulations resulted in no significant changes between the base case and alternatives (despite much higher assumed union wage premiums in Mexico) in the estimated impact of the USMCA on GDP (0.35 percent), wages (0.27 percent), or employment (176,000 jobs) in the United States, despite roughly doubling the assumed impact of collective bargaining on wages in Mexico (GDP and total U.S. employment increased very slightly in these simulations, by between 1/10 to 3/10 of 1 percent, as the Mexican wage premium was doubled).

Thus, it appears that in the ITC’s own model, dramatic increases in the assumed impact of Mexican labor standards, resulting in very large increases in Mexican wages, would have, at most, a de minimis impact on employment and output in the United States. This leads to a puzzle. If stronger labor rights enforcement in Mexico is not responsible for significantly higher output, employment or wages in the United States, what is?

It turns out, according to the ITC projections, that essentially all of the expected benefits from the agreement are expected to result from increased “certainty” for investors—almost entirely as the result of new regulations governing internet commerce and data transfer. These regulations are supposed to reduce uncertainty, leading to increased investment, output and employment in the United States. According to Table 2.6 (p. 56) of the USMCA report, U.S. output, total employment and wages would fall under the agreement (with declines of .12, .04 and .06 percent, respectively), but for the uncertainty provisions in the agreement (that is, in a scenario with no reduction in uncertainty, but all other elements of the agreement in place). With “moderate” certainty (the base case) these variables increase by .35, .12 and .27 percent, respectively, that is, the by net effect of the USMCA shown above.

Taken together, these two analyses of the ITC sensitivity analysis of labor rights, and the impacts of uncertainty reductions, provide convincing evidence that in the ITC models, Mexican labor rights have no significant impact on U.S. wages, and that all of the net estimated benefits of the USMCA are a product of the assumed benefits of “uncertainty reduction.” Specifically, the estimated U.S. wage premium is entirely explained by the uncertainty provision assumptions (as Dean Baker has noted). With no uncertainty benefits, the U.S. wage impacts of the USMCA are negative; with moderate (base case) uncertainty reduction, the wage impact is +0.27 percent. But these results also highlight deeper problems with the ITC models.

First, at what cost are the gains from “reduced uncertainty” achieved, and what form do they take? The model is a black box, and does not tell us. The deal’s proponets claim investment will increase. But where? In Mexico, the United States, or Canada? If it’s investment abroad, it is likely to result in more offshoring and job loss in the United States. These estimates provide a thin reed indeed on which to assume the deal’s benefits, particularly when the USMCA also contains a six-year sunset clause, and the president has frequently threatened to withdraw from NAFTA itself if he does not get what he wants from Congress on everything from approval of the USMCA itself to funds for his border wall with Mexico.

Finally, these results also illustrate a core problem with the ITC approach to modeling the impacts of trade agreements. The agency uses a computable general equilibrium or CGE model that is designed to estimate the effects of tariff cuts on trade flows. Thus, in order to evaluate provisions of trade deals, each provision must be translated into an “effective tariff cut” and the impact of the agreement as a whole is then evaluated via the estimated impacts of these “imputed” tariff cuts on trade flows. But many of the issues involved in modern trade agreements have much more to do with the distribution of income than with the level of trade. For example, improved labor rights in Mexico will redistribute income from capital to labor in that country. That may have an impact on the demand for labor in the United States. These interactions should be carefully evaluated as a system of interactions.

Similarly, the USMCA will require the U.S., Mexico and Canada to provide 10 years of exclusive patent protection for expensive biologic drugs, the most significant driver of increasing spending for prescription drugs. It would prohibit Congress from reducing this period of monopolistic patent protection for these drugs (as has been proposed), and limit competition with “biosimilar” products. This simply drives up profits for some drug companies, while increasing consumer costs in all three countries. This will cost lives, and reduce consumer income available to purchase other products, thus shrinking trade in and production of other products, along with national output and employment. Very few workers are involved in the production of biologic drugs (and in fact, many of the components are produced abroad and imported into the U.S. to take advantage of offshoring incentives in U.S. tax laws).

In both cases (labor law reform and patent protection) trade and investment deals likely have bigger impacts on the distribution of income than on GDP, employment or wages. Yet the CGE models used are designed to primarily evaluate all these components via their tariff-equivalent impacts on trade. The ITC needs to develop a new generation of models designed to evaluate the impacts of trade and investment policy on both the distribution of income, including the effects of labor rights and environmental standards enforcement, as well as the impacts of these deals on output, employment and wages here and abroad. These models should also transparently evaluate both the benefits and the costs, and the distributional impacts of domestic and foreign investment, including offshoring, as a result of these deals.

NAFTA and other trade and investment deals have been a disaster for American workers. We need a new approach that will rebalance trade and level the playing field for workers in the United States and other countries. And it is time for the ITC to develop the analytic tools capable of transparently exploring the complex distributional impacts of these economic arrangements.