Thanks for noticing my post. Note, however, I'm not saying that the system of subsidy by locality should be totally dismantled. But there are some pretty obvious gains that could be had if the practice was scaled back. For example, I wonder how many communities have used TIF districts to attract a big box retailer? If a lot of communities are doing so, and if the distribution of the retailers would be the same if TIFs didn't exist, then it is a socially wasteful practice.

But as I also pointed out, it is not in the individual best interest of any one locality to cut back on the process by itself because they would be putting themselves at a disadvantage vis-a-vis the others. If all the communities except one compete for the retailers then they will be the last place the retailers will go.

Though I can't recall chapter and verse off the top of my head, I've heard of some pretty questionable TIFs over the years. It is a practice that is based on a reasonable idea, but the actual usage has gone beyond its original intent. That was my point.

I read your post on quality of place. When you talk about politicians being responsible for "quality of place" and using that to justify immigration over outsourcing and the like, it does sound a little bit like mercantilism, albeit in a different way.

I agree that economic models don't include "quality of place". I'm trying to imagine how to include it without it becoming an awful mess. But I've always maintained that an economic model is a starting point for a policy discussion, and not necessarily the ending point.

Furthermore, I'm not sure I like the idea of letting politicians be entirely responsible for "quality of place"!

William — We don't disagree at all, then. I don't mean to overstate my point. My suspicion, actually, is that very many of the subsidies offered by localities, turn out to be precisely the kind of rent-transfer mechanisms that public-choice types accuse them of being. Many of the rest are only welfare-enhancing to local residents because the competitive landscape is (or is perceived to be) distorted by other localities' offer of subsidy, as when New York pays a Wall Street firm to stay on Wall Street, to match an incentive package dangled by New Jersey. If neither NY or NJ had offered, nothing would be different, except that taxpayers would have kept some of their money. It is certainly not my claim that subsidies are always, or even usually, good, whether in a broad sense, or from the perspective of subsidizing localities.

It is my strong claim, though, that occasionally these subsidies and incentives are profoundly successful, both in a local and global sense. Even if the normal scenario is error or corruption, there are cases where some combination of competence and luck literally saves places and dramatically improves the lives of the people bound to them. Anecdotally, I'd place the city I was born in (Baltimore) and the nation called China in that category. I am quite cynical about the general competence and and public-spiritedness of politicians. But, at present, we have very few institutions capable of looking after "quality of place". Local government development efforts do exist, and do sometimes succeed. I'd like better political institutions, better approaches to measuring and improving quality of place, and a pony too. But I won't issue a blanket condemnation of using the tools we have now. On a case-by-case basis, in full cognizance of the dangers, I enthusiastically support certain kinds of local subsidy.

Mercantilism is a good word to bring to this discussion. Whatever its technical and historical meaning, mercantilism has come to signify patterns of trade or trade distortion that turn what ought to be positive sum games into zero or negative sum games. I don't think that all competition between places qualifies as mercantilism. Competition to attract people and investment may well make all places better, as different localities work to develop their advantages and keep up with escalating standards for a good place to live. I think that has certainly been true within the United States. Ironically, I think we are more likely to create mercantilism by following an orthodox economic line, treating people and firms as atoms and ignoring place. By opting out of what could be healthy competition, places that might have competed fail relative to those acting strategically to protect or enhance their competitive position. It may well be that there should be rules governing competition between places, to prevent destructive practices like NY/NJ bidding wars. But those have to be developed cooperatively.

Suppose New Jersey insists on trying to woo Goldman Sachs and offers enough subsidy to make a move plausible. The worst thing New York could do would be to refuse on principle to bid. (Let's presume for the sake of argument that GS' presence does enough good to be worth a matching bid.) If it does refuse on principle, it has manufactured mercantilism. It has lost and New Jersey has won in a zero-sum game, a game that New Jersey will have every incentive to play again in the future. On the other hand, if NY matches the bid, and aggressively tries to poach NJ firms, that ends up creating costs for both parties, and an incentive to come to a cooperative arrangement that discourages the practice. The states might still compete for new firms by offering great infrastructure and reducing red-tape, or try to woo talented people with great parks, or whatever. That's "good competition". But states only have an incentive to refrain from destructive practices if it's clear that competitors will do unto them if they do not. The best way to support positive-sum free trade is to credibly threaten to abandon it if others do.

Unfortunately, standard economic models argue that victims of mercantilistic subsidy should unilaterally opt out, and profit from others' expensive distortions. If states were mere collections of people, that would be a good strategy. But since states are places, profiting from, say, cheaper financial services offered by Goldman-NJ also means a much longer commute for New Yorkers who want to work there, fewer and worse restaurants in the neighborhood of the old GS HQ, an of talented employees, contractors, and supporting firms, a smaller NYC tax base, etc. When states are places and not collections of people, the standard model doesn't work.

As you say, "an economic model is a starting point for a policy discussion, and not necessarily the ending point." I wish more economists brought that point of view into discussions of trade.