by Michael Dorf
In my Verdict column for this week, I discuss Monday's SCOTUS oral argument in United States v. Texas. After discussing standing and the merits, I argue that, in the event that the Court does not simply divide 4-4 and affirm without opinion, the case provides an opportunity for some well-chosen dicta affirming that there are constitutional limits on the president's exercise of prosecutorial discretion. I explain why both progressives and conservatives should be able to get behind such a proposition.
Here I want to address a different question about unilateral executive power, concerning the president's authority with respect to foreign affairs.
Let's begin by recalling that a little over a year ago, 47 Republican Senators, led by Oklahoma's Tom Cotton, issued an open letter to the leaders of Iran warning that the nuclear deal that was then being negotiated was a mere executive agreement, and thus subject to revocation by a future president with a mere penstroke. The Cotton Letter (as it came to be known) was greeted with mostly negative attention. In a new article published in Political Science Quarterly, I try to mine the Cotton Letter for broader lessons about the relation between international law and domestic U.S. law.
The link above takes you to the PSQ website, but the article is behind a paywall. Accordingly, I will very briefly summarize the article for those of you whose .edu internet access does not make all knowledge freely available to you, before moving on to one further issue. Here's the summary:
1) On its face, the Cotton Letter makes a fundamental error, confusing the mechanisms by which international agreements become effective as between sovereigns with the mechanisms by which they are implemented in domestic law.
2) Whether or not the author and signers of the Cotton Letter are confused about the relation between inter-sovereign effectiveness and domestic implementation, foreign sovereigns have reasons to care about the latter, not just the former.
3) In particular, foreign sovereigns appropriately should focus on two issues: a) whether the mechanism by which the obligation is made renders it automatically effective domestically; and b) how "sticky" the obligation is.
4) I then categorize various forms of international agreements and obligations by automaticity and stickiness. I discuss self-executing and non-self-executing treaties, congressional-executive agreements, sole executive agreements, customary international law, and non-binding agreements. (The Iran deal is probably best understood as a non-binding agreement, but its explicit references to UN Security Council implementation make the categorization complicated.)
If you have access to PSQ, I urge you to check out the article. If not, sorry.
Now, onto a related question about stickiness. Suppose that Bernie Sanders becomes president and renounces NAFTA (or as he pronounces it, NAFTER). Could a president do that? As a matter of international law, withdrawal is typically governed by the agreement instrument itself, and NAFTA authorizes withdrawal by any party six months after giving written notice.
The relevant provision doesn't specify how a party withdraws, but I agree with Julian Ku, who suggested in 2008 that because the international dimensions of NAFTA are an executive agreement, the president can withdraw without seeking congressional approval. Indeed, the general rule is that a president can withdraw from a treaty without any sort of congressional approval, even though Senate approval was required to make it operative in the first place. It would seem to follow a fortiori that a president can withdraw from an executive agreement without congressional permission.
If President Sanders, President Trump, or some other president were to withdraw from NAFTA, would that be the end of NAFTA? Surprisingly, maybe not.
NAFTA became domestically effective via a statute, the NAFTA Implementation Act. I am not a trade law expert, and so what follows could be wrong, but my reading of the Implementation Act suggests important gaps. In various places, the Implementation Act specifies that various otherwise-operative provisions will no longer be in effect in the event that a NAFTA country withdraws from NAFTA. Presumably these provisions were written with Canada in Mexico in mind, but it is possible that they have the U.S. in mind as well. The interesting thing--and this is the point that I could be completely wrong about--is that the Implementation Act appears to indicate that even if the president were to withdraw from NAFTA, some provisions of the Implementation Act would remain in effect. Put differently, it appears (at least to my untrained eye) as though the NAFTA Implementation Act doesn't simply implement NAFTA. It has provisions that are just freestanding legislation.
I'd be happy to be corrected if I'm wrong about that analysis, but for now, let's assume that I'm right. Certainly it is conceptually possible for a piece of legislation that is styled an implementation act to specify that in the event that the whole international agreement blows up, some provisions of the implementation act would nonetheless remain valid. Under these circumstances, "implementation act" would be a misnomer, but there would be no larger difficulty.
What should the default rule be? Suppose the president signs an executive agreement with the heads of state of one or more other countries. To get around any errors I may have made regarding NAFTA, let's call this imaginary executive agreement SHMAFTA. Congress then passes the SHMAFTA Implementation Act, and for over two decades, both SHMAFTA and the SHMAFTA Implementation Act are in effect. Then the president withdraws from SHMAFTA, after giving the required notice. Suppose that the SHMAFTA Implementation Act is silent on the question whether it or any part of it will or will not remain in effect following presidential withdrawal from SHMAFTA. Should that silence give rise to a presumption that Congress wants the Act to cease having effect following withdrawal from SHMAFTA? To a presumption that Congress wants the Implementation Act to continue in effect? To no presumption at all but simply to the use of ordinary (if inadequate) tools of statutory construction?
I don't think there is any SCOTUS case law on this question, and I can see arguments pointing in each direction. Commonsensically, it seems that the point of the SHMAFTA Implementation Act is to implement SHMAFTA. Once the president withdraws, there's nothing to implement. Thus, the Implementation Act is tacitly conditional on the continuation in force of SHMAFTA.
Yet it's at least somewhat significant that the president and Congress chose to proceed via a congressional-executive agreement rather than via a treaty. If the president had entered into a non-self-executing treaty with the advice and consent of the president, and Congress enacted implementing legislation of a non-self-executing treaty, it would be apparent that Congress indeed meant the legislation to be dependent for its effect on the continued force of the treaty. This would be especially clear if, per Missouri v. Holland (which is still good law, at least just barely), Congress only got the power to enact the implementing legislation in the first place by operation of the treaty. But by proceeding via a congressional-executive agreement in reliance on Congress's Article I, Sec. 8 power to regulate foreign commerce, there is at least a plausible argument that Congress intended the legislation to stand on its own.
Or one might say that courts ought to examine an Implementation Act provision by provision. Some provisions will make little sense in the absence of the international agreement. Others will be more clearly freestanding. A no-presumption approach could well lead to the conclusion that some but not all of the provisions of the SHMAFTA (or NAFTA) Implementation Act actually implement the agreement.
How these questions are resolved could matter if a President Sanders withdraws from NAFTA, but Congress (with Republicans still in control of the House) does not repeal the NAFTA Implementation Act.