This article studies the period preceding the current opioid epidemic by leveraging the 1974-1982 RAND Health Insurance Experiment. We document novel facts about the widespread use of prescribed opioids, prescriber characteristics, and diagnoses linked to opioid prescriptions in what we refer to as Wave Zero. Additionally, we exploit random assignment to health insurance plans to estimate how opioid use adjusts to differences in insurance generosity. We find that more generous plans increase use by raising outpatient visits and providers' likelihood of prescribing opioids, without changes in filling rates.

With Cecilia Diaz-Campo