Volume 22, Issue 2

Growing evidence suggests that medical marijuana laws have harm reduction effects across a variety of outcomes related to risky health behaviors. This study investigates the impact of medical marijuana laws on self-reported health using data from the Behavioral Risk Factor Surveillance System from 1993 to 2013. In our analyses we separately identify the effect of a medical marijuana law and the impact of subsequent active and legally protected dispensaries. Our main results show surprisingly limited improvements in self-reported health after the legalization of medical marijuana and legally protected dispensaries. Subsample analyses reveal strong improvements in health among non-white individuals, those reporting chronic pain, and those with a high school degree, driven predominately by whether or not the state had active and legally protected dispensaries. We also complement the analysis by evaluating the impact on risky health behaviors and find that the aforementioned demographic groups experience large reductions in alcohol consumption after the implementation of a medical marijuana law.

The income gap between specialists and primary care physicians and among specialists is well established, but the drivers of this difference are not well delineated. Using the Community Tracking Study (CTS) Physician Survey, we sought to isolate and compare premiums paid to physicians for specialization and the proportion of time spent on offices visit rather than procedures. We divided medical subspecialties according the proportion of Medicare billing for Evaluation and Management (E&M) codes for the specialty as a whole. We report substantial differences in income across physician specialty, and over 70 percent of the difference in income remained controlling for factors that may confound the relationship between income and specialty including gender, location and type of practice, and hours. We note a large variation in premiums for specialization: 11.3–46.8 percent above family medicine after controlling for confounders. Classifying medical subspecialties by E&M billing as procedural versus non-procedural specialties revealed clear income differences. Controlling for confounders, procedural medical specialties earned 37.5 percent more than family medicine, as compared with 15.3 percent for non-procedural medical specialties. This analysis suggests that differences in physician income and resulting incentives are a direct consequence of the payment structure itself, rather than compensation for additional years of training or a reflection of different underlying demographics.