Apologies for the delay, the flu bug set me back enough to cancel class and to be unable to re-engage this important dialogue sooner. I hope a couple more posts will be enough to leave this discussion in the hands of readers for their own judgment.
To re-state my understanding of the Constitution’s design: Treaties were to be hard to enter into (hence the 2/3 Senate ratification requirement), but easier to enforce than under the Articles of Confederation, where compliance depended on the willingness of state legislatures. If a treaty is a valid treaty, Congress’ power to implement the treaty is not constrained by any “reserved” legislative powers of the states; the Constitution ensures that the legislative powers to implement treaties lie with the national government. This is a structural inference from the treaty-making power in Art. II and also a result of the necessary and proper (NP) clause. There are limits on what treaties can do, but those limits are to be found in various other provisions of the Constitution (Eugene is correct that those limits are likely not exhausted just by the Bill of Rights) and in the requirement that treaties must be valid exercises of the treaty power.
The Constitution was specifically designed to overcome “The Treaty of Peace” problem: peace treaties often require a nation to honor the claims of foreign creditors, eg, and Congress was giving the power to override state contract/debt laws in order to enforce the terms under which the Revolutionary War was ended. So far, I don’t think any of the responses from Nick, Ilya, and Eugene have yet explained how their views would enable Congress successfully to enforce the Treaty of Peace. In my view, it’s a serious strike against any interpretation of the Constitution if it cannot explain how the Constitution solves one of the fundamental problems to which the Constitution was specifically designed to be a solution.
Nick’s approach is particularly odd to me because it generates the conclusion that the national government can trump state legislative powers if it makes a treaty self-executing, but not if the treaty requires domestic legislation to be implemented. Nick gets to this view, in part, by claiming that Congress’ exercise of one enumerated power cannot give Congress additional legislative powers it does not have already. I want to say more about that claim of Nick’s, in addition to my earlier argument that the national government’s war powers have always stood against Nick’s view.
Nearly every exercise of power by Congress under the NP clause also seems to be inconsistent with Nick’s claim, unless I misunderstand that claim. Congress traditionally had no power to regulate intrastate railroad rates, for example, but if it regulates interstate rates through its commerce clause powers, then it can regulate intrastate rates as a necessary means of making the interstate regulatory regime effective. Or, Congress has no enumerated power to create national corporations or to create a Bank of the United States; yet once Congress is create currency, paying soldiers and sailors, purchasing property, and the like, it has the power to charter the Bank as a means of making effective the exercise of these other powers.
Here is Nick’s apparent answer to this problem, from his article at n.91:
Similarly, cases like Houston, East & West Texas Railway Co. v. United States (Shreveport Rate Cases), 234 U.S. 342 (1914), are not to the contrary. That case upheld an order of the Interstate Commerce Commission regulating intrastate railroad rates, because the order was necessary to maintain its regime of interstate rates. But to say that Congress can regulate intrastate railroad rates only when and because it is also regulating interstate railroad rates is not quite the same as saying that regulating interstate railroad rates expands the power of Congress to reach intrastate rates. The case is probably best read to hold that a single act of Congress (the Interstate Commerce Act of 1887) regulating both interstate and intrastate rates is necessary and proper to carry into execution the power to regulate interstate commerce. It does not follow, however, that an act of Congress regulating only intrastate rates would be constitutional – even if there were already another act of Congress on the books regulating interstate rates.
In other words, assume that (1) X alone is within Congress’s power; (2) Y alone is not; and (3) Y is necessary to carry X into execution. It may be that a single act of Congress X + Y is constitutional, because X + Y may fairly be described as a law regulating interstate commerce. It does not follow, however, that Y could ever be enacted alone, even after the enactment of X, because Y alone could never be described as a law regulating interstate commerce. Evaluation of the Article I power to enact a statute may rightly depend on the content of the whole statute, but probably should not depend on the existence of other statutes already enacted. The question in each case should be whether any given statute – all of it, in itself – may be said to be an exercise of an enumerated power (citations omitted).
Thus, Nick’s view is that it would be unconstitutional for Congress to regulate intrastate commerce in a statute passed after Congress had regulated interstate commerce, but constitutional if Congress regulates both interstate and intrastate commerce at the same time in one statute. Needless to say, no Supreme Court case has come close to endorsing that position, as far as I know, and I will let readers decide how persuasive they find it. In addition, laws like the one creating the Bank of the US — and many laws enacted under the NP clause — are not enacted at the same moment as exercises of the enumerated powers to which those later laws are necessary and proper. The Bank of the US law was a freestanding law enacted after the national government was engaged in other activities to which the Bank was viewed as necessary. But Nick is driven to his claim about how congressional powers purportedly work by his view that self-executing treaties can displace state legislative power (the equivalent to a comprehensive federal law that regulates both interstate and intrastate commerce in one moment) but not non-self executing treaties.
On the historical record, Nick takes issue with my statement that long before Missouri v. Holland it was “close to universally accepted” that Congress’ power to enforce treaties was not limited by any “reserved” legislative powers of the state. Ironically, one of the strongest pieces of evidence I can offer (in a blog post) for that statement is: Nick’s own article. Before making that statement, I re-read Nick’s articles with a specific eye out for every piece of historical evidence it offers to support Nick’s view, since I assume Nick would have marshaled all the supportive evidence. Yet I was surprised how thin that evidence turns out to be; Nick reprises virtually all of it his short blog post.
This evidence consists of (1) one newspaper article from 1879; (2) the position of one Senator, Wilson Cary Nicholas of Virginia, during debates over the Louisiana Purchase — but from my recollection of those debates, this statement was isolated and it was not an issue that anyone else engaged, agreed with, or took issue with it, because it stood askew to any of the issues actually being debated. But leaving that aside, if one Senator once made such a statement, that’s not much of a basis for concluding that there has long been a significant understanding, even if a minority position, within the political branches, of the anti-Missouri v. Holland view; (3) a statement in one Supreme Court case in 1836 (Nick’s post says “cases,” but he cites only this one majority opinion) and in St. George Tucker’s treatise on the treaty power. Yet neither this Court case nor the treatise, as I understand them, supports Nick’s particular view: neither takes the view that self-executing treaties can override state legislative power but non-self-executing ones cannot. These two statements, on their face (I haven’t gone back to the sources to read them in context), support a different view, closer to Ilya’s, which is that no kind of treaty can expand the legislative powers of Congress. And they remain two statements, in one treatise and one 1836 Court decision.
Having read Nick’s article, I said the Missouri v. Holland view had been “close to universally accepted” throughout U.S. constitutional history — not universally accepted. I know enough constitutional history to know that there is always at least a few bits of support that one can find for most views on almost any difficult issue in constitutional history. But based on the evidence offered so far, I remain surprised by how little evidence there appears to be for Nick’s view throughout American constitutional history. For the evidence on the other side, showing how central it was to the Constitution’s design and structure that the U.S. be able to honor its treaty commitments and for the historical understanding of the treaty power, see the articles referred to in my earlier posts by Dan Hulsebosch and David Golove. I stand willing to be corrected on that point and now that the Supreme Court will be hearing the Bond case, perhaps we will learn much more about what the full historical record shows on these issues.