With Long Island Representative Peter King and New Jersey Governor Chris Christie blasting House Republicans today for refusing to pony up Sandy relief money, it's worth saying that the case for general skepticism about disaster relief is actually decent.
One problem is a threshhold issue. From the standpoint of any given person, having your house randomly destroyed isn't any better than having your house and a bunch of other peoples' houses destroyed. Why should people afflicted as a small part of a larger "disaster" get special help relative to people just hit by some other misfortunate? It's particularly striking when you see conservative politicians from disaster-struck areas like Christie and King out there talking about the government's need to step in and help those in need. Where are those guys on the day-to-day misfortunes of poverty and illness? Conversely it doesn't seem like a very good idea to systematically transfer money from low-disaster areas to high-disaster areas. You'd like to see people, buildings, and infrastructure move over time to areas where they're relatively unlikely to be destroyed by earthquakes or bad weather.
Now on the other hand, a natural disaster is clearly a situation that calls for a little deficit spending. And the federal government has a much greater ability than a state government to borrow money on a moment's notice. So in that sense, the case for federal disaster assistance is very strong.
Which suggests that the right approach to disaster relief money is to drop these ad hoc emergency bills and try to move to a bank model instead. Create a standing reserve fund that states and local governments can tap when the president declines an official disaster. The interest rate would be some small premium over the prevailing interest rate on federal debt. That way states suffering damage will be able to finance any necessary and useful repairs, but there won't be any net flow of resources over time to unusually high-disaster areas.