Tort Reform, physician behaviors and health outcomes -- Experiences of Joint and Several Liability in Pennsylvania and New Jersey

Presenter: Shinyi Chou, Lehigh University

Abstract

This study focuses on the effect of change in Joint and Several Liability (JSL) in 2002 on the cesarean rates in Pennsylvania. We also look at the effect of JSL reform on the cesarean section rates based on mothers' socioeconomic status, especially women with Medicaid insurance.
Before 2002, JSL followed the so called “deep pocket” rule which allowed plaintiff to collect the entire payment from whoever had the highest ability to pay for the damages. Therefore malpractice risks faced by physicians were relatively lower considering the fact that hospitals were always the ones with deep pockets. However, as part of tort reform in Pennsylvania in 2002, the rule was altered to require full recovery from the defendant whose liability is 60% or higher. As a result, physicians faced a higher level of malpractice risk, which increased the probability they practiced defensive medicine. In such context we investigate physicians’ behavior change via certain procedure use and patient health outcomes using hospital birth data to identify the effect of tort reform on social welfare.
We apply a Difference-in-Difference model to inspect the practice of defensive medicine as physicians' response to the tort reform in JSL using Pennsylvania birth records from PHC4 inpatient data and New Jersey birth records from HCUP inpatient data from 1994 to 2006. Considering the fact that the JSL reforms occurred in each state under different time frames, we develop a series of tests to identify the average treatment effect of JSL reform on cesarean section rates for either treatment group or control group. We consider county fixed-effect and year fixed-effect in this model, and control for patient characteristics such as admission types, age, race, insurance type and previous cesarean section records. Next we specifically evaluate the effect of JSL reform on cesarean section rates on mothers who are enrolled in the Medicaid program within a Difference-in-Difference-in-Difference framework.
Preliminary empirical results show the JSL reform has a negative effect on cesarean section rates, which is consistent with the results provided by Currie & MacLeod (2006); we observe the same impact of JSL reform on cesarean section rates in New Jersey in 1995. We also find evidence on the effect of JSL reform on the cesarean section rates with mothers enrolled in the Medicaid program. In the end, we perform several falsification checks and find out the results are quite robust.