A Revealed Preference Approach for Ranking City Quality of Life

This paper presents a new method for ranking city quality of life that requires weaker assumptions than the Rosen and Roback hedonic methodology. A simple revealed preference approach is used. If a person can raise his wages and lower his rentals by moving from city a to city b, then city a must be nicer than city b. This approach differs from the hedonic approach because a separate fixed effect is estimated for each city. The fixed effect represents the dollar value of the entire bundle of a city's local public good vector. The revealed preference approach relaxes the assumptions that: 1) All relevant city specific attributes are observed, and 2) the functional form of how attributes affect wages and rentals is known. I also show how this estimator can be used to study city quality of life dynamics from a single cross-section of rentals and home prices across cities