Estimating Heterogeneity in the Benefits of Medical Treatment Intensity

Federal and state laws passed in the late 1990 increased considerably postpartum stays for newborns. Using all births in California over the 1995-2001 period, 2SLS estimates suggest that for the average newborn impacted by the law, increased treatment intensity had modest and statistically insignificant (p-value>0.05) impacts on readmission probabilities. Allowing the treatment effect to vary by pre-existing conditions or the pre-law propensity score of being discharged early, two objective measures of medical need, demonstrates that the law had large and statistically significant impacts for those with the greatest likelihood of a readmission. These results demonstrate heterogeneity in the returns to greater treatment intensity, and the returns to the average and marginal patient vary considerably.