Association for Public Policy Analysis & Management

U.S. local governments are on the front lines in the battle against the national opioid epidemic. Yet, there is a dearth of systematic understanding regarding how and to what extent local policymakers are responding to this complex problem. Recent research suggests some opioid policies such as administering naloxone, promoting needle exchange, and expanding treatment programs, reduce opioid-related overdose deaths and prolong life, while other policies targeting opioid supply can promote transition to heroin use (Pitt, Humphreys, & Brandeau 2018). Researchers contend that a comprehensive, portfolio approach may therefore be more effective, but there is currently little investigation into the fuller gamut of policy alternatives for addressing opioid-related problems, and what factors shape such utilization at the local level. To fill this gap, we are surveying (2018-2019) county governments across five geographically dispersed and politically “purple” states (Colorado, Florida, Ohio, Pennsylvania, and Washington) with both rural areas and large population centers, to learn about their opioid policies and what environmental, organizational, and institutional factors facilitate or hinder their utilization. We first employ item response theory (IRT) to determine which opioid policies (19 in total) may require a greater latent commitment for utilization. We then test a linear regression model that compares the extent to which problem severity (2016 overdose deaths rate per 100,000), political preference (2016 presidential election vote share), governmental capacity (population, median income, dedicated budget and staffing levels), and other local characteristics predict latent commitment to opioid policymaking. Preliminary findings from Colorado and Pennsylvania only (N = 64, response rate = 49%) suggest governmental capacity is generally a key driver, and political preference and problem severity also play important roles. Findings also show some capacity and political measures displaying nonlinear relationships with policy commitment. Implications for researchers and policymakers will be discussed.