Gautam Gowrisankaran, Associate Fellow

Estimating Hospital Quality Using a Bayesian Selection Model

In this project, Professor Gowrisankaran estimated the quality of care for pneumonia patients being treated by hospitals in Southern California, using Bayesian methods to control for patient selection based on severity of illness. The use of Bayesian methods has a number of advantages but is very computationally intensive. Because of the interactive nature of the Bayesian estimation algorithms, Professor Gowrisankaran was able to parallelize the computation.

This project sought to estimate the effects of competition for both Medicare and health maintenance organization (HMO) patients on the quality decisions of hospitals in Southern California. It was found that increases in the degree of competition for HMO patients decrease risk-adjusted hospital mortality rates. Conversely, increases in competition for Medicare enrollees are associated with increases in risk-adjusted mortality rates for hospitals. In conjunction with previous research, estimates indicated that increasing competition for HMO patients appears to reduce price and save lives, and hence appears to be welfare improving. However, increases in competition for Medicare appear to reduce quality and perhaps reduce welfare. The net effect of a given merger on hospital quality depends on the geographic distribution of different payer groups.

Another project concerned the causes and magnitudes of network externalities for the automated clearinghouse (ACH) electronic payments system, using a panel data set on individual bank usage of ACH. This researcher constructed a model of ACH usage to identify network externalities from correlations of increases in usage levels for banks within a network, and from increases in usage following exogenous increases in market concentration of size of competitors. He structurally estimated the parameters of the model by matching equilibrium behavior to the data, using simulated maximum likelihood and a data set of localized networks. Estimation requires repeated computation of the equilibrium for different parameter values and values of the structural unobservables. Preliminary estimates reveal moderate network externalities that are principally due to lack of bank usage. Fixed costs of adoption are very variable across banks. Moderate subsidies to banks that adopt ACH are welfare improving. This study could potentially help shed light on the extent of any underuse of electronic payments technologies and other high-technology products.