This recent post by John Quiggin laments the lack of diversity in state governmental structure in the United States. Despite the fact that they are almost entirely free to experiment, states tend to favor tripartite systems of government with separate judicial, legislative, and executive branches, and in every state but one even have our bicameral house.

Of course, despite the basic similarities, state governments do have quite important differences in structure. Some have line-item vetoes, some have powerful near-unitary executives, others divide power and privilege the legislature. Some entrust basic rights and liberties to a massive state constitution and then delegate power to the judiciary to guard against tyranny. Others prefer popular sovereignty provisions that make referendum the ultimate check. Unfortunately, Yale, like almost any other law school, offers no class on state administrative law, or state constitutional law, or . . . anything very similar.

At any rate, I mention the original Crooked Timber post because I wonder exactly how much freedom the U.S. Constitution does give to states in fashioning their own form of governance. There is, of course, the Republican Guarantee clause, which Taney (in dictum) dismissed as non justiciable in Luther v Borden, but there are other considerations too. Article 1, Section 4 presupposes that states will have legislatures, which are constitutionally charged with regulating elections in the default instance. Article 1, Section 2 presupposes that states will have an "executive authority" (which it imagines in the singular). Article IV, Section 4 makes similar suppositions. Article VI seems to envision that there will be "judges in every state", and they are probably required by the Due Process Clause at any rate.

So one question is what work these constitutional presuppositions do. Although the Constitution clearly never says "Every state shall have a legislature, and an executive authority, and judicial officers", there are a number of clauses of the constitution that would simply make no sense if the state had none. One possibility is to give states ultimate freedom (within the bounds of the elusive Republican Guarantee clause). Another would be to allow them to designate their Prime Minister, Minister of Elections, Speaker of the House, or whoever else as the "executive" within the meaning of the federal constitution-- but this challenges the too-accepted notion that it is the federal government and not the states that determines what the federal constitution means. A third option would be to suppose that there is a brooding-requirement-of-tripartite-government that states simply must adhere to, as a sort of structural inference from the Constitution. Of course, even within the general restrictions of having a separated or unseparated three-branch structure (and being required to adhere roughly to the so-called one-man-one-vote rules) states have a great deal of leeway to experiment with things like instant runoff voting, parliamentary systems, and so on. Still, given the large number of structure tools they can play with within the three-branch system, it is not so clear why they would want to reinvent the wheel.

[At any rate, the question of presuppositions of the U.S. Constitution is not unique to the electoral context. As my classmate Steve Sachs once pointed out to me, the Seventh Amendment requires jury trials when the amount in controversy exceeds twenty dollars. Does this require the U.S. to use dollars to value things? If we wished to adopt the Euro would we be constitutionally required to maintain a fixed exchange rate so as to render clear the right to jury trial? What if we were to stop printing dollars and they were to slowly disintegrate and deteriorate until there was no longer a liquid exchange market?

Similarly, the Twentieth Amendment, among other provisions, makes explicit reference to dates on the Gregorian Calendar. Does this forbid Congress from adopting a new method of keeping time? Does this require us to keep track of the old Gregorian days for Constitutional purposes even if we switch calendars for all other federal business? There were some odd folks in the early Congresses who thought that Congress's power to "fix the standard of weight and measures" did not permit it to establish the Metric system (because that would imposing a new standard of measure rather than "fix"ing the one in common use) but that seems silly by modern lights-- even to an originalist, I would think.

None of this is meant as a general critique of originalism, textualism, written constitutions, or whatever else. But for nearly any theory of constitutional interpretation, there is a question about what to do with background assumptions that would turn sentences of the text into actual nonsense if they were rendered false.

Not very long ago, a surprising crowd of libertarian bloggers-- including me-- evinced their reluctant support for Section 8 housing vouchers-- which are effectively transferable rent vouchers, and a far more effective form of welfare spending than government housing projects/boondoggles. Matthew Yglesias-- I had thought-- agreed with us.

In his latest post defending the Davis-Beacon act, Yglesias seems to argue that it will be good for progressive housing politics to strengthen unions because then you'll have a constituent that will lobby hard for building lots of public housing projects. Of course, this will have the side effect of making Section 8 even less politically salient than otherwise, and as Yglesias has noted, Section 8 does not exactly have a lot of natural special-interest support, since it actually succeeds at moving most of the money from rich and middle-class taxpayers to the targetted recipients.

And furthermore, while Yglesias is right that giving union workers extra money with every government construction project will encourage them to lobby for the government to build more things, is there any reason to suppose the government will build the kinds of things Yglesias wants them to build? Governments build not just public housing projects but sports stadiums, massive corporate complexes, military bases, and a lot of other things that are not-- so far as I know-- an integral part of the progressive vision, but are often popular with other relevant interest groups. Why help Unions and the Poor when you can help Unions and Boeing?

It's as if all Yglesias cares about is putting more tax-financed bricks-and-mortar on the ground, and if that's what leftist coalitional politics have come to, they are in a sad state.

UPDATE: Yglesias emails to make clear that he is for section 8 housing vouchers now as previously, and that he is also for government construction projects generally, be they pork or not. The trouble I was trying, artlessly, to make clear was that to some extent these things compete against each other. Unions will not just funnel money from taxes to construction projects (which Yglesias thinks is good) but also from efficient welfare programs to very wasteful ones, or to construction projects that are not welfare programs at all. As he makes clear in his original post, he is basically fine with this, but I do confess that I am confused about why anybody else is supposed to be.

UPDATE TWO: A reader adds:

I don't think you fully appreciate that there is in fact quite a constituency for Section 8, consisting of builders and owners of low-income housing, investors in the same (check out IRC Section 42), lenders to owners and builders of the same (who get Community Reinvestment Act credit), lawyers ... who work on affordable housing deals, investment bankers who do bond financing for such housing (check out IRC Section 142(d)), etc. Admittedly, the benefits of giving vouchers to Louisiana refugees may be too diffuse to mobilize this constituency, but in general this group will certainly support more funding.