This week, Canada released its own economic impact assessment of the new North American Free Trade Agreement (NAFTA), the United States-Mexico-Canada Agreement (USMCA), or as it’s referred to in Canada, CUSMA. Canada has yet to approve the deal in its parliament, while the United States and Mexico have already completed the ratification process. Canada is likely to approve the deal, and this assessment of its benefits will not make or break the process, but nevertheless warrants a close look, as it has broader implications for how we should think about evaluating the impact of trade agreements in the future.

A Numbers Game

Let’s start with the numbers. The economic impact assessment released by Global Affairs Canada (GAC) estimates that implementing USMCA “would secure GDP gains of $6.8 billion (US$5.1 billion), or 0.249%” over a 5 year period (p. 59). However, the report also notes that motor vehicle exports to the United States would decline by US$1.5 billion relative to the existing NAFTA. Why? The report cites the more stringent rules of origin (RoO) requirements of the USMCA as likely to lead to an increase in production costs.

Compare this to the U.S. International Trade Commission’s (USITC) report, which estimated gains for the U.S. economy at $68.2 billion, or a 0.35% increase in real GDP, over six years. Now, these figures are not directly comparable since they’re looking at the impact on their own respective countries, but the significant gap in reported gains is notable. It is worth pointing out that the USITC report also found a negative impact from the auto RoO.

A recently updated study by Dan Ciuriak, Ali Dadkhah, and Jingliang Xiao for the C.D. Howe Institute, a Canadian think tank, conducted an assessment of the USMCA, looking at the impact on all three economies. They found a negative impact on real GDP across all three NAFTA countries— a decline of ‑0.396% for Canada, ‑0.791% for Mexico, and ‑0.097% for the United States.

What gives? In making these assessments, the reports all rely on certain assumptions that impact the outcomes of their models. But two major issues stand out—what are the factors the studies rely on for their estimates, and what are they comparing the changes in the USMCA to?

As Jeffrey J. Schott from the Peterson Institute for International Economics notes, the USITC’s very positive growth figure “is entirely due to USITC estimates that the USMCA will induce more US investment by reducing uncertainty in policies on data, ecommerce, and intellectual property rights.” But questions have been raised about what this reduced policy uncertainty will actually produce. It’s the first time the USITC has tried to quantify the impact of binding commitments on data flows and data localization in terms of reducing uncertainty for markets.

But if this reduced policy uncertainty is excluded from the model, the USITC estimates that the net impact of USMCA on the economy is actually negative, at ‑0.12%, or approximately a loss of US$22.6 billion. This is due to the new provisions that will actually reduce trade in North America, which my colleague Simon Lester and I have written about earlier. The good thing about this figure is that it actually gives a comparison of NAFTA vs. USMCA, so we have a clear sense of the purported benefits of the “new” deal.

But what does Canada do? The GAC report does not model reduced uncertainty, saying that:

Some provisions under CUSMA could also help reduce policy uncertainty in certain areas such as services, investment and digital trade, and result in a positive impact on businesses; however, modelling such gains is challenging and relies heavily on the assumptions made. Therefore, these types of benefits were not evaluated in this study. Furthermore, many of these obligations have already been implemented by Canada under CETA, and by Canada and Mexico under the CPTPP (p. 47).

This is why the economic gains estimated by the GAC report are more modest, on the whole. Interestingly, however, the reason the topline figure is positive and not negative in the report is that the baseline used is not NAFTA. Instead, the GAC report compares USMCA to no NAFTA at all, plus the continuation of steel and aluminum tariffs on Canada due to U.S. Section 232 measures. The reasoning for this choice is justified as follows: “Canada was presented with two choices: risk the consequences of a U.S. withdrawal from NAFTA or pursue a modernized Agreement.” USMCA or bust!

Aside from the fact that there has been an intense debate on the legal hurdles to a unilateral withdrawal from NAFTA by the United States, the economic impact would be significant, complicating U.S. threats to withdraw. Would President Trump really risk such a move? This has always been unclear, and for Canada to base its assessment as if this was a certain outcome should raise more than a few eyebrows.

Furthermore, the right comparison is whether the USMCA improves upon NAFTA, which the GAC report did not calculate. A figure based on this comparison was reported by Ciuriak et al. (noted earlier). The lead author of that study, Dan Ciuriak, has noted that taking the new GAC report’s scenario into account, he finds a comparable estimate of gains from his study’s model. 

Therefore, if one were to assume reasonably similar estimates for NAFTA withdrawal and the impact of Section 232 tariffs on steel and aluminum in the counterfactual put forward in the GAC analysis, and if GAC were to make its own estimate of the marginal impact of the USMCA relative to the status quo (NAFTA in place, plus no Section 232 tariffs), these estimates would be in the general ballpark of the results published by Ciuriak et al. That is to say, if GAC did take the USMCA vs. NAFTA comparison, it would find a negative impact.

What Does this Mean?

The big issue raised with the discussion above is that we need to be sure we are evaluating agreements based on the differences between the new and old rules, or even with other rules that may be better. For example, for Canada and Mexico, it would be more useful to also compare the outcomes of USMCA with the Trans Pacific Partnership (TPP) agreement, from which the United States withdrew. As the excerpt from the GAC report above notes, many of the reductions in uncertainty come not from new provisions in USMCA, but things Canada has already agreed to under its agreement with the European Union, and the TPP. So did the United States get more out of USMCA than it did from TPP? The answer is probably not. In fact, the USITC report admits as much:

the Commission’s baseline does not take into account the various market liberalizations and binding commitments that Mexico and Canada have undertaken as signatories of the CPTPP, including commitments applying to data localization and data transfer. Since these are key policies that drive Commission model results in estimates that provide higher weights to the value of policy uncertainty, it is unclear whether estimates with higher weights for policy uncertainty would apply to the current (post-CPTPP signing) policy context (p.58, emphasis added).

In addition, these economic assessments raise an even more fundamental question, not least because Canada was not even planning to release a report until opposition parties called for it. The evaluations are supposed to be objective and offer a clear picture of the costs and benefits. However, the USITC and GAC report both find positive gains, whereas many trade policy experts have consistently urged caution and suggested that the USMCA is actually more of a NAFTA-minus deal than NAFTA-plus. If the reports simply provide support for government policy, and have to contort themselves to do it, are they really useful assessments?

Another study by Mary E. Burfisher, Frederic Lambert, and Troy D. Matheson from the International Monetary Fund (IMF) found a negligible impact on GDP (0.02% for Canada, ‑0.01% for Mexico, and 0.00% for the United States). They also note that “key provisions in USMCA would lead to diminished economic integration in North America, reducing trade among the three North American partners by more than US$4 billion (0.4 percent) while offering members a combined welfare gain of US$538 million” (p. 23).

This is precisely why it is important to look at independent analysis from scholars and experts in order to build a critical picture of what these government reports contend. It may also be worth asking whether other independent international bodies should have the ability to offer these assessments upon request (and to have their results considered alongside official government reports), such as the World Trade Organization (WTO) or IMF, in order to ensure political neutrality of the findings.

What we do know from existing studies is that the USMCA provides minor improvements to NAFTA, but that these gains could have been achieved had the United States stayed in the TPP. For Canada and Mexico, the benefits just don’t add up. Other than avoiding political confrontation with the United States, it is not clear that ratifying USMCA does anything for Canada or Mexico. In fact, it does not even reduce uncertainty, or the possibility for conflict with the United States.

Just this week, the United States Trade Representative signaled his support for U.S. withdrawal from the WTO’s Government Procurement Agreement (GPA), which Canada is also part of. In USMCA, Canada agreed to exclude provisions that would guarantee access to the U.S. procurement market, as it believed it could rely on the GPA rules. If the U.S. withdrawal from GPA does occur, this would bring Canada back to the negotiating table to sort out yet another trade conflict with the United States. Add to this the ongoing concerns over the Trump administration’s use and abuse of unilateral trade actions, such as Section 232, and all that reduced uncertainty seems like it was a mirage.