Cross-Border Mental Health Effects of Cannabis Legalization: Spillovers from New Jersey to Pennsylvania

• In Event: Mental Health

Abstract

Legalization of recreational cannabis has increased over time, with 24 states now permitting recreational use. Several additional states, including Pennsylvania, are actively considering cannabis legalization. In some cases, proof of residency is not required to make purchases, raising the possibility that legalization in one state could directly increase recreational use in neighboring states. We consider the consequences of New Jersey's (NJ) 2021 recreational cannabis legalization for the mental health of Pennsylvania (PA) residents, a setting that offers a sharp policy contrast across a long shared border, as PA continues to operate a comparatively more restrictive medical-only program. Using zip-code–quarter inpatient discharge data from 2018–2022 and information on driving time from each PA zip code to the nearest NJ dispensary, we estimate difference-in-difference models with Mahalanobis distance matching of exposure of NJ dispensaries on PA psychiatric admissions, using both a binary treatment indicator and a continuous treatment that lets the effect vary with this travel time. We find that hospitalization rates for mental health disorders in PA zip codes close to NJ rose by approximately 1.06 per 1,000 per quarter (about 8%) following the legalization of recreational cannabis in NJ. Hospitalizations for depression and alcohol/tobacco disorders increase the most, rising by 14.14% and 7.2%, respectively. After allowing the treatment effect to vary with travel time to NJ dispensaries, we find that hospitalizations rise as travel time falls, particularly for dementia and other organic mental disorders. These results are robust to alternative definitions of the treatment areas, pandemic sample exclusions, and pass multiple placebo tests. Our estimates indicate that recreational cannabis legalization generates interstate mental health externalities primarily through geographic access, suggesting that the standard practice of evaluating cannabis policy at the state level alone may systematically understate the full social costs of reform, and that policy-relevant cost–benefit analyses should more carefully account for the spillover burden borne by residents of nearby non-legalizing neighboring states.

Author

• Haiming Li, Lehigh University
  Presenting Author