*****
2. Greg Mankiw links to an evaluation of ARRA (the stimulus) that finds a negative impact on jobs. What doest his have to do with the states and interstate commerce? Well, the study is done with state-level data and more crucially the model specification is unable to incorporate the impact of interstate commerce which is - to put it mildly - an enormous omission. One nice thing about the study is that it does attempt a sophisticated identification strategy. It uses two instruments - state sales tax intensity (to instrument the decline in state budgets - which I've always noted has been an important contractionary force that people ignore), and state highway funding formulas (which should be exogenous to the recession and are an important determinant of ARRA funding levels). I'm especially glad to see them instrument for state budget losses. But here's the problem - the only impact they can estimate is between states that received relatively more or relatively less stimulus. So if Virginia received $900,000 in stimulus and Maryland received $1,000,000 in stimulus they're going to try and identify the impact of that extra $100,000 of stimulus. Theoretically that could be fine - it'll provide a marginal effect that can be attributed to the whole package. My first (more minor) concern is that when you infuse a lot of money into an economy like this you're ultimately going to run into bottlenecks and - yes - crowding out at least on the margin and in certain areas. Is that marginal effect of the last $100,000 the same as the marginal effect of the first $700,000? Likely it's not.
That's a relatively minor concern, actually. The bigger concern is the interstate commerce point. A lot of commerce is done across state lines and that can't be accounted for with this sort of estimation strategy because the ARRA funds going to Virginia not only are not used to estimate job levels in Maryland - the job levels in Maryland are actually counted against the Virginia impact estimates. To simplify things, you can think of the model as asking "what is the effect of (VA stimulus-MD stimulus) on (VA jobs-MD jobs)". If MD stimulus positively impacts VA jobs or if VA stimulus positively impacts MD jobs, you're going to actually reduce the estimated marginal effect. I don't know if this sort of thing completely eliminates the prospect of state-level studies of fiscal multipliers, but it's certainly something that needs to be taken into account.
That's a relatively minor concern, actually. The bigger concern is the interstate commerce point. A lot of commerce is done across state lines and that can't be accounted for with this sort of estimation strategy because the ARRA funds going to Virginia not only are not used to estimate job levels in Maryland - the job levels in Maryland are actually counted against the Virginia impact estimates. To simplify things, you can think of the model as asking "what is the effect of (VA stimulus-MD stimulus) on (VA jobs-MD jobs)". If MD stimulus positively impacts VA jobs or if VA stimulus positively impacts MD jobs, you're going to actually reduce the estimated marginal effect. I don't know if this sort of thing completely eliminates the prospect of state-level studies of fiscal multipliers, but it's certainly something that needs to be taken into account.
*****
3. Matt Yglesias has a good treatment here of arguments against the clear English of the commerce clause that amount to "well if you can do that, what can't you do?". It's never been an especially impressive argument. Yglesias points out some obvious things you can't do, including violating other Constitutional provisions like the first amendment and rights to due process. I've noted before that you also can't violate the general welfare clause - special priveleges for the sake of private welfare skirt Constitutionality in a real way that belies these claims about "if you can do that what can't you do?". But more importantly, these arguments demonstrate a real lack of commitment to the very idea of republican virtue. The Constitution is an important document because it limits the state, thereby protecting incursions on liberty. But since when has it been the only thing standing between the state and liberty? It has always been recognized that for a republic to succeed you need a virtuous populace. You need a populace that won't pursue inappropriate uses of power (or rectify the situation when such powers are pursued).
People act as if we can't have a meaningful and successful republic if the Constitution doesn't provide the ultimate and final demarcations of power. This is an excessively myopic critique, in my mind. We want to be able to achieve public ends with the republican institutions we have set up. Half the petitions raised by Jefferson in the Declaration were complaints about George III not letting the colonial legislatures pass laws that were for the public good. Since the beginning of the republic, it has been understood that we want a government that allows us to govern ourselves - that gives us the ability to make important public investments and decisions. A nit-picking Constitution threatens that, so instead we have a Constitution with real restrictions on the state - but restrictions that are open to interpretation, and yes - deliberation. It amazes me that this very idea that things are left open to deliberation and interpretation is viewed as a threat to liberty, rather than a source of real liberty.
In a free society, we deliberate within a framework of broad restrictions on the state, and if we want to keep that free society we have to preserve the republican virtues that are required for the preservation of liberty. The Constitution is a tool, not a master. It helps us preserve liberty - it's not a free pass on deliberation or a guarantee of success or an ultimate bulwark in defense of liberty. It's like a marriage. The contract itself doesn't guarantee anything. You have to work at your goal.



