To my macroeconomist friends: if you are going to do urban economics, please read the urban economics literature first.

I saw a paper from a famous macroeconomist a week or so ago that proposed that cities with high incomes relative to house prices produce more utility than those cities with low incomes realtive to house prices. Using this metric, he concluded that Flint was among the five highest utility cities in the US. This might have been a clue that there was something wrong with his utility measure.

The systems of cities literature (see Jan Brueckner's chapter in the Handbook of Urban Economics) and the quality of life literature (see Stuart Gabriel Joe Mattey & William Wascher's RSUE paper) shows that in a country with mobility, utility tends to get equalized across cities, and so that places with high house prices relative to income have more non-housing amenities than places with lower house prices. I can testify to the reasonableness of of this, as while Los Angeles is expensive (as well as congested), I do not find myself tempted to move anywhere else, suggesting that I, at least, derive a great deal of utility from living here (I recognize that lots of people are not so enamored of LA, but enough of us are to keep the price of housing high relative to other places). Other places like LA include New York, London, Paris, Singapore, Tokyo, Hong Kong, Sydney, etc.

This produces an efficient outcome, for if Los Angeles were less expensive, it would be even more congested. On the other hand, St. Louis' cheap house prices should eventually attract people back to it. But it also produces an unfair outcome, because it is very difficult for low income people in Los Angeles to find reasonably priced housing in reasonable locations. Perhaps the goal of housing policy should be to allow everyone to be able to choose the city in which to live, while at the same time distorting the relative prices of cities as little as possible. I am not sure how one does both.