Association for Public Policy Analysis & Management

In 2014 almost 19,000 deaths were attributable to prescription opioid overdose in the United States, with another 10,500 deaths associated with heroin overdose. The American opiate epidemic has spread rapidly since the early 2000s and continues to worsen, with devastating consequences for families and communities. While there is a growing literature on the effects of prescription opioid policies at the state level, less is known about the link between legitimate opioid prescribing in physician offices and opioid abuse and death in aggregate. This paper will combine data on all Medicare Part D prescribing for opioids at the individual physician level in the U.S. with county-level accidental drug death data for the years 2010-2014 to assess the strength of this link. In particular, the paper presents evidence about whether linkages are observed between measures of office-based physician competition (HHI and concentration ratios for overall and opioid Medicare prescribing), Gini coefficients for the distribution of prescribing in a county, and the presence of high-volume prescribers and the level of county death rates to assess the nature of any spillovers from licit prescribing to mortality outcomes. Further, state opioid policies (e.g. prohibitions on doctor shopping, active prescription drug monitoring programs, limits on opioid refills, and the like) are included in the models to determine the degree to which such policies can mitigate, or interact with, the contribution of physician prescribing to opioid-related overdose deaths.