Black/Hyman/Silver have a new draft paper, "Does Tort Reform Affect Physician Supply? Evidence from Texas," (via Robinette) that substantially undermines the empirical case for the conventional wisdom that Texas's 2003 reforms against medical malpractice lawsuits attracted more doctors to Texas. The result is highly counterintuitive: after all, even the authors acknowledge that the reforms dramatically decreased malpractice expenses for doctors. Are we to conclude that doctors do not respond to economic incentives?
Alas, the authors do not suggest any explanation for the phenomenon they describe. Possibilities:
- The supply of doctors is inelastic relative to after-expense income. This is a testable hypothesis, and would have dramatic implications for "bending the cost curve" of health-care expenditures if true.
- Employers of doctors offset the decrease in medical-malpractice expenditures by decreasing wages paid to doctors. This seems somewhat implausible, as many doctors are independent, and the ones that aren't probably aren't paying for their own malpractice insurance. But it is also a testable hypothesis. Too, if the health-care market in Texas responded to such a wage decrease by reducing costs to patients (or, at least, reducing costs to patients relative to the nationwide trend of rising costs to patients), that is also worth studying, and would be a benefit that may refute the overstated conclusion of the authors that "tort reform is a small idea, when it comes to the larger and linked questions of health care access and affordability."
- The quantity of doctors did not increase, but the doctors responded to the incentives by changing the mix and quality of services provided in any given year: more OB/GYNs willing to deliver babies rather than restricting themselves to less risky work; more doctors willing to work in emergency rooms; doctors spending more time seeing patients and less time in medical-malpractice-related activities like defending themselves in lawsuits, cover-your-ass documentation, and (for better or worse) defensive medicine. If the average practicing doctor is spending more hours with patients post-tort-reform than pre-tort-reform, doctor supply is increasing, even if the raw numbers aren't. I am not aware of any evidence for this, but economic theory would predict this result. It's not clear whether the data exists to test this hypothesis, but as in the parable of the drunk looking for his lost keys under the streetlamp, one should avoid drawing conclusions that contradict economic theory just because it is too difficult to test an alternative hypothesis consistent with economic theory. Too, if defensive medicine practices changed, as one predicts they would, have health outcomes changed for better or worse? (Professor Silver has argued elsewhere his concern that Texas doctors would take less care post-reform.) Again, this is difficult to test, especially since the adverse consequences of many defensive-medicine decisions, such as excessive CAT scans, won't be known until the additional cancers show up decades later. But it is both a potential benefit and a potential cost of tort reform, as we don't know to what extent doctors are properly weighing benefits and costs (including opportunity costs of more intensive treatment of a particular patient) at the margin. Kessler's study, backed to a lesser extent by the CBO, certainly suggests defensive medicine is wasted money at the margin in the state of the world without damages caps, but defensive medicine is surely different today than in the 1980s.
- For many doctors with low-risk practices, malpractice liability is not a large factor in their practice decision. But the malpractice liability crisis most heavily hit high-risk practices, like neurosurgery or OB/GYN or emergency-room care. Did Texas tort reform materially affect the supply of doctors in high-risk specialties, while the effect on low-risk specialties was overwhelmed by noise? This should be a testable hypothesis, but the data is poor because of a change in the way statistics were collected. The authors try to get around this by comparing 1997-2000 growth to 2008-2010 growth, but there's not necessarily a reason that one would predict a post-tort reform world to have a different post-equilibrium effect than a pre-tort reform world. One cannot rule out the hypothesis that doctors overreacted to the new incentive when tort reform was first imposed and that depressed new demand in later years. Of course, one cannot rule out the null hypothesis that a dramatic decrease in malpractice-insurance rates caused by tort reform did not increase the supply of high-risk doctors, though, again, one wishes for an alternative explanation for why doctors are not responding to economic incentives. (Note, too, that the authors' decision of excluding 2001-07 from the data has dramatic effects on the data. It's unclear to me why a reporting change in 2001 that would artificially increase the 2001-02 numbers relative to the 1999-2000 numbers should have an effect on the 2003-07 numbers, especially given the 2000-2003 declines that are being excluded.)
Can anyone think of other alternative hypotheses in the comments?
I remain skeptical that a wealth transfer from lawyers to doctors and patients didn't have positive externalities, but I, for one, am going to stop claiming that Texas tort reform increased doctor supply without better data demonstrating that. More study is needed to explain Black/Hyman/Silver's counterintuitive result, and partisans on both sides need to be more conservative with their policy claims. Earlier.