I confess that I admire the post-Katrina idea of trying to spend money on infrastructure only where critical masses of people settle and resettle. It is somewhat circular, to be sure, because more people will move to a neighborhood if they know they will be enjoy government money, but it is one reasonable way of directing resources to places where they will do the most good. Most alternatives for allocating post-disaster aid involve more central planning, and that is likely to be more error prone, politically difficult, and even corrupt. But if the basic question is how to spend money (and how much to spend) on a population much smaller than originally found and dislocated (which is not the case to the east of New Orleans), some difficult decisions need to be made - unless we are prepared to undo property titles (perhaps in return for just compensation, and perhaps for chits (or more compensation) for those who resettle in the area) and let people settle in as if it were a new frontier. Only an academic or an overly imaginative politician would pursue that idea.
I confess also that there is something about the political maneuvering that is pleasing. Here is an optimistic story: Politicians take the temperature of the country's taxpayers and send more money to a desperate locality the more the people are sympathetic, and the more they think the money will not be wasted in local hands. New Orleans is doing poorly because its (past) corrupt habits, failed schools, and spotty public services do not inspire confidence, and so the "voters" choose not to send too much more money that way. In the normal course of events, it is difficult to treat different localities in differential fashion, and voters do not pay much attention. Annual expenditures by the federal government will sometimes penalize but sometimes reward the most inefficient states and cities. But when forced to go back to the "market" for new capital on a large scale, as New Orleans must post-Katrina destruction, there is more information and more of an opportunity to focus attention on whether this is a good investment and whether the funds will be in good managerial hands. It is easy to say that Mississippi (which is receiving more money per capita or per destroyed house) has tighter political links to the White House and Capitol than does Louisiana (and that there is no shrinking population problem, as suggested above), but we might add to this the idea that even Mississippi (hardly a national favorite) seems more trusted by the taxpayers at large. Disaster has forced these jurisdictions to go to the market (of public opinion and of politics, if only because the the politicians do not want to look bad after the fact) much as corporations might make taxable distributions in order to be "forced" to go to the market for new capital. This is the essence of the cash-flow theory of why corporations pay dividends even when tax disfavored; in the post-disaster political world the claim is not that governments encourage disasters, but rather that the threat of disasters (and then the opportunity for large scale federal relief) can discipline local governments. In this view. Giuliani's value was in offering the image of someone who could be trusted to use relief funds wisely post-9/11. Had the Lower Manhattan architectural bickering begun earlier, New York might not have done so well.
I think we are accustomed to thinking this way about post-war relief on an international scale, though the international example is of limited use. It is true that we were more likely to fund the Marshall Plan the more our voters thought Europe would be well run and not wasteful. And we are less accustomed to thinking of domestic relief in this manner. But war on a large scale usually brings about changes in government, often dictated or influenced by the victorious powers, and domestic disasters do not do so, and may even serve to keep local leaders in place. Post-war relief is thus less likely to discipline pre-war politicians than is post-disaster domestic relief likely to discipline pre-disaster voters and politicians.