… because they effectively destroy endogeneity!
This paper recently arrived on the NBER website, and I was sucked in by the title alone. It’s a relevant question – are pricey hospitals really delivering on what you’re paying them? The endogeneity problem is obvious: the rich people who go to these high cost hospitals are inherently different from their poor-hospital counterparts. So how do we circumvent the problem?
Ambulances! Ambulances work in zones, so people who live very near each other and are demographically similar could go to different hospitals in emergencies. This creates a factor for hospital choice – geographic proximity to a hospital – which is unrelated to the other characteristics which would normally confound a hospital efficiency analysis. How cool is that?
The results are quite interesting. High-cost hospitals do better in terms of their one-year mortality rates, but teaching hospitals that adapt to the latest technologies and trials do the best, regardless of cost. The authors conclude that the prices do not necessarily reflect treatment quality, but rather treatment intensity and attention to procedure.
I had come across papers before that exploited people on either sides of borders (for school zoning laws, different tax rates, etc.) but never ambulance zones and hospital choices. I loved the extension of that method to answer some pressing health economics questions.