NAFTA Termination: What Would FDR Do?

October 16, 2017

As negotiators meet in Arlington this week to discuss the future of the North American Free Trade Agreement (NAFTA), the Trump administration is again rattling nerves. While no formal proposal is publicly available, the U.S. Trade Representative is reportedly calling for stricter Buy America rules in government procurement, a five-year sunset clause (i.e. the agreement would not be renewed unless countries agreed), and an opt-in/opt-out on the controversial investor-state dispute settlement process. Oh, and Trump himself again suggested he might just pull the U.S. out of the pact.


The U.S. Chamber of Commerce is not happy. Mexican CEOs are not happy. The American Farm Bureau is not happy. Even congressional Democratic critics of NAFTA are urging Trump to slow down.

My bet is that these business entreaties will be successful. As President George W. Bush trade official Phil Levy told The New York Times, “Find me the last trade agreement that U.S. passed with the chamber in opposition,” Mr. Levy said. “You don’t have a chance. It’s hard enough with the U.S. Chamber in favor.”

Thus, I predict business will get what it wants—either in the form of a NAFTA 2.0 that looks like the Trans-Pacific Partnership, a return to the NAFTA 1.0 status quo ante, or a NAFTA 1.5 with changes at the margin that are symbolically significant but don’t affect U.S. business interests too much.

However, academics have been debating the most extreme outcome suggested by Trump’s tweets: Does the president even have the authority to unilaterally exit without a say so by Congress?

What Would FDR Do?

Tufts University’s Joel Trachtman is a leading exponent in the “no” column and makes a strong case in this paper that Congress has a role to play. His contributions are both doctrinal and empirical. On the first front, he makes the argument (following major cases like U.S. v. Curtis Wright (1936)) that presidential unilateral authority waxes and wanes. When it comes to trade, however, Congress’ lead constitutional role in regulating commerce argues for legislative say-so on termination. On the second, Trachtman waded through over a hundred instances of treaty termination and finds that in most instances (over 95% of the time), termination was pursuant to a congressional action of some type. If the probability of how a judge will rule based on past decisions is the law, Trachtman has good arguments.

However, I was interested in the exceptions to the rule, where executives appeared to act unilaterally, in that gray area without unambiguous congressional support. Trachtman discusses some of them, while others are discussed in more detail in papers by Curtis BradleyJean GalbraithJonathan York Thomas, and the Congressional Research Service. Moreover, one can access the State Department’s internal correspondence of some of the key cases on the Foreign Relations of the United States online archive. These include:

  1. An 1899 decision by the McKinley administration to suspend certain provisions of an 1850 Swiss trade pact.
  2. The 1933 decision by the Franklin Delano Roosevelt (FDR) administration to withdraw from the multilateral 1927 Convention for the Abolition of Import and Export Prohibitions and Restrictions.
  3. The 1936 decision by FDR to terminate an 1871 trade pact with Italy, after Mussolini started ramping up protectionist pressures.
  4. The 1939 decision by FDR to terminate a 1911 trade pact with Japan (subsequently endorsed by Congress).
  5. The 1944 decision by FDR to terminate the 1929 Inter American Convention for Trademark and Commercial Protection.
  6. The 1979 decision by the Carter administration to terminate a Taiwan mutual defense pact after establishing trade and diplomatic relations with China (which became a major Supreme Court case when members of Congress challenged it before the Supreme Court in Goldwater v. Carter).

These cases give a bit more support to executive power on NAFTA than Trachtman and others seem to think.

First, they are all directly or indirectly related to trade.

Second, two-thirds of the cases came from FDR—an administration that prided itself on a mix of economic nationalism and internationalism; an administration committed to rethinking the nation’s role in the world (much as Trump has, albeit in a much darker, rhetorical direction and with no real achievements to show). By this reading, it’s to be expected that Trachtman’s stats would go the way they did: Most presidents do not try to remake U.S. global strategy.

Third, a major justification for these actions was an executive unilateral determination that the world had changed since the treaties had initially gone into place. Under the international law doctrine of changed conditions (rebus sic stantibus), governments can claim to opt out of their obligations. For instance: The State Department justified the 1944 decision because the treaty was seen as ineffective and not conducive to U.S. interests. They justified the 1933 decision because FDR had just signed the National Industrial Recovery Act, which allowed unions, companies, and the president to penalize imports if they endangered or made ineffective the Act’s industrial policy goals. FDR and his free trading Secretary of State Cordell Hull claimed that they couldn’t achieve those goals—so important in the wake of the Great Depression—while meeting the international obligations. Moreover, other countries had been opting out of the 1927 pact to look out for their domestic recovery, meaning the U.S. was justified in doing likewise.

In correspondence with his Italian ambassador (to be relayed to Mussolini), the Secretary of State justified the 1936 decision using rhetoric that is similar to the types of things you hear from the present administration about NAFTA:

The treaty of 1871 is clearly obsolete. The economic life and conditions of the respective countries and of the world at large have changed fundamentally. Not only is the existing treaty obviously inadequate to govern the economic relations between the United States and Italy today, but actual instances of its inadequacy have arisen and tend to multiply. There would seem to be no necessity, after a lapse of sixty-five years, to argue the question whether a new treaty is needed by two advanced commercial and industrial states… notwithstanding the provisions of that treaty for most-favored-nation treatment, the commerce of the United States is now being discriminated against in ways that are definitely harmful to Americans who desire to trade with Italians.

In the memos reproduced by the State Department’s legal advisor G.H. Hackworth from around that time, he wrote that the law on termination “is by no means settled.” Citing a 1909 legal memo from State, he wrote:

[Beyond congressional authorizations] A third method of terminating a treaty is by notice given by the President upon his own initiative without either a resolution of the Senate or the joint resolution of the Congress … It is thus seen that there are three methods of terminating a treaty, so far as the United States is concerned, and that the choice of method would seem to depend either upon the importance of the international question or upon the preference of the Executive.

Then citing his own 1936 memo on the Italy question (and the first two examples above), he wrote that:

Recognizing this action taken in 1898 as a precedent, which is confirmed by the procedure followed on the occasion of the withdrawal of the United States from the convention for the abolition of import and export prohibitions and restrictions in 1933, I have no doubt that you may authorize the giving of a notice to Italy of the intention to terminate the treaty of 1871 without seeking the advice and consent of the Senate or the approval of Congress to such action.

Takeaway: Trump Isn’t FDR, 2017 Isn’t 1933

To be clear, I don’t think denunciation is politically likely, for the reasons discussed above. But the historical-legal record is more ambiguous on denunciation’s legality than I would have thought.

And even if unilateral termination were possible, its practical implications would be limited. Trachtman correctly argues that the administration can’t unilaterally undo the domestic implementing legislation that accompanied NAFTA. (In FDR’s time, trade policy was in some ways more of an executive affair.) So the issue is whether Trump could denounce the pact while the U.S. still follows the accompanying statute. In that case, we’d have a kind of Zombie NAFTA where the U.S. extended privileges to Canada and Mexico while not being technically a member of any trade pact. As I’ve suggested on Twitter, this path would likely bring WTO litigation.

The world has changed, in several ways.

First, international legal complexity a la 2017 would suggest the path taken by FDR would be less likely today.

Second, contemporary presidents have also been far less ambitious in their economic interventionism. Not unrelatedly, the 1933 NIRA was struck down by the Supreme Court. Trump has not even attempted to get such a serious industrial policy mandate from Congress, and he would probably be rejected if he tried.

Finally, a few months ago, I would have been willing to believe that the Supreme Court might have followed its 1979 precedent by conservative (then) Associate Justice William H. Rehnquist that treaty termination is a non-justiciable political question. By October, I’m not so sure. Trump has burned bridges with the law enforcement and legal community over the Muslim ban, Russia investigation, and environmental policy rollbacks. Congressional Republicans and the business community are beginning to distance themselves from him. None of this is dispositive for how the courts would rule on a unilateral NAFTA termination, but it does suggest an atmosphere of distrust of his motives and competence. Indeed, if Trump attempted termination, he might find a Court willing to put him in adult day care supervision. That might be the right call in this case, but it would be an unambiguous rollback of executive power that would hamstring presidents of both parties for years to come.