And if you believe this, I’ve got a great deal to sell you: The economic impacts of the revised NAFTA (USMCA) Agreement

The North American Free Trade Agreement resulted in growing trade deficits with Mexico and steep U.S. job losses after it was implemented in 1994, increasing the bilateral trade gap by at least $97.2 billion and costing at least 682,900 jobs through 2010.

Can NAFTA 2.0 do any better?

The U.S. International Trade Commission’s (ITC) new report on the economic impact of the U.S.—Mexico–Canada Trade Agreement, released last week, projects that the revised NAFTA (USMCA) will have tiny impacts on the economy. The ITC estimates the deal will increase GDP by 0.35% when it is fully implemented (six years after it takes effect), or roughly 10 weeks of growth. Similarly, it projects that 175,000 jobs will be added in the domestic economy, a 0.1 percent increase in total employment (based on CBO projections for the economy in 2025), or roughly as many jobs as the economy adds in a normal month, over the next six years. And it claims real wages will rise about one-quarter of a percentage point (0.27 percent), roughly 4 percent of what workers are expected to gain, in real terms, over the next six years, if promised gains in output and employment are realized.

But there are strong reasons to doubt that these gains will be achieved. The ITC results show that the deal will yield remarkably small gains, and those gains rest on questionable assumptions about how the deal will help workers and the economy. Perhaps the most problematic finding in the ITC study (p. 25) was that labor provisions in the USMCA “would increase Mexico union wages by 17.2 percent, assuming that these provisions are enforced.” Given that unionization rates in the durable goods sectors of Mexican manufacturing are reported to be 20.2 percent (Table F.4), these would be massive impacts, indeed. Yet Mexican workers will not benefit unless there are mechanisms to ensure that labor rights enforcement does improve, but those provisions do not yet exist in the agreement.

Thus, it is not surprising to find that the AFL-CIO, other labor unions, and many members of Congress are demanding that “swift [and] certain enforcement tools” are included in the deal before it is submitted to Congress. These concerns also apply to segments of the agreement that pertain to the environment, access to medicines. Furthermore, the assumption that Mexican union wages will increase 17.2 percent seems especially heroic, within the 6-year adjustment period in the ITC model (p 23.), in light of the struggles that will be required to unionize such a large share of the labor force.

The ITC’s economic impact projections are built on a series of heroic assumptions that are built into its “computable general equilibrium (CGE) model.” The ITC model assumes the economy is always at full employment; that trade deals do not cause trade imbalances, job losses, growing income inequality, or downward pressure on the wages of most workers, and environmental damages;, and that the more expensive drugs, movies and software don’t otherwise harm consumers or the economy.

In particular, the ITC study assumes that the overall U.S. trade balance is unchanged by the deal (despite its significant changes in the rules of governing, trade, labor and the environment). Peter Dorman pointed out the problems with CGE models nearly two decades ago, as have many others, and yet the ITC staff, and other economists keep using them anyway.

ITC studies of the impacts of trade and investment deals have a particularly poor track record of projecting the actual pattern of trade following recent trade deals. In addition to being wrong about the overall size and direction of trade flows, the ITC also failed to “correctly identify the winning and losing industries in trade with Mexico and Korea.” The ITC claimed (Table 2.2) the US-Korea KORUS deal would improve that trade balance, but the trade deficit increased, costing more than 95,000 American jobs. And the most infamous, 1999, ITC estimate (Tbl. ES-4) was that China’s entry into the WTO would increase the bilateral trade deficit by about $2 billion after the deal is fully implemented. At last count, that deficit is up $272 billion, costing 3.4 million U.S. jobs.

The ITC’s claims about supposed (net) benefits of the NAFTA-2 (USMCA) deal stand in contrast to those of several other studies. Researchers from the International Monetary Fund recently used their own CGE model to estimate the economic impacts of the trade deal. They found that the USMCA, alone, would result in slight increases in welfare for Mexico and Canada and a small net welfare loss of $794 million for the United States.

Some of the most important, and controversial, changes to the NAFTA agreement, embodied in the USMCA, involve new rules of origin (ROO) that raise the level of regional content for motor vehicles, parts and some other products (including steel and aluminum in vehicles) which must be achieved in order to qualify for special, duty-free treatment under the agreement. New standards were also developed for labor value content (LVC) that require certain minimum shares (40 to 45 percent) of cars and trucks must be produced with labor earning at least US$16 per hour. [1]

The ITC report estimates that the USMCA’s ROO and LVC requirements would increase net employment in U.S. auto and parts production by more than 28,000 jobs, and that U.S. investment will increase by “$683 million per year to meet new demand for U.S. produced engines and transmissions. (ITC report 19).” But these findings are certainly controversial.

In their recent IMF working paper, Burfisher, Lambert and Matheson, found that the USMCA agreement (alone) would result in a small ($275 million) increase in the U.S. trade deficit, and that the negative sectoral impacts would be more significant. In particular, they found the agreement would reduce output of motor vehicles and parts in all three countries by an average of 0.8 percent ($7.6 billion), and by 0.2 percent in the United States alone. Higher vehicle costs (due in part to tariffs) will reduce vehicle demand, and some auto and parts production is likely to shift to lower wage locations such as China, Vietnam or Malaysia. This finding stands in contrast to the ITC assumption that the USMCA will increase motor vehicle production in the United States..

It is time for a new approach to analyzing trade deals. The CGE models are built to analyze the impacts of reducing tariffs on trade. But deals such as the USMCA are about much more than trade in goods and increased efficiency in the economy. They are trade and investment deals, and their most important, negative effect on U.S. workers is that they make it attractive for multinational companies to offshore production to low wage countries such as Mexico. The CGE models make gross, simplifying assumptions about how non-trade issues affect workers and the economy, and yet provide an unearned degree of perceived rigor about these effects, when, in fact, they are assuming answers rather than carefully exploring their impacts.

The USMCA nibbles at the edges of the USITC trade model, and the basic approach to negotiating trade deals, but doesn’t appear to change its basic tenets.. The USMCA will continue to encourage firms to offshore production. The locations may change, from Mexico and Canada to China, Vietnam or Malaysia, but the results will remain the same.

These negotiations, and the models used to assess their impacts, aren’t getting the job done for working Americans. It’s time to go back to the drawing board, for all of us.

[1] ROOs for textiles and apparel were also tightened, and some other trade restrictions involving agricultural trade (dairy, eggs, poultry, cotton, peanuts & related products) between the U.S. and Canada, and some trade in small packages were relaxed.