The Effect of Cost-sharing on Opioid Use

Over the last two decades, drug overdoses and number of deaths related to opioid abuse have been persistently increasing. To combat this problem, both the federal and state governments have introduced a range of supply side policies---including the reformulation of oxycontin and state prescription monitoring programs (PDMP). Use of demand side policies to address this epidemic, however, is not common. Recently some states, including California and Kentucky, introduced bills to tax opioid consumption. The empirical evidence on whether consumers of opioid drugs would respond to the introduction of tax on opioid, however, is almost non-existent. In this paper, I fill this gap by presenting a new evidence on the effect of cost-sharing on opioid consumption. To identify the causal effect, I use the closure of the Medicare Part D coverage gap (also known as donut-hole) under Affordable Care Act and the variation in Medicare eligibility month (which is determined by birth month) among age 65 Medicare beneficiaries as exogenous sources of variation. The difference in the difference estimate shows that closing the Medicare Part D coverage gap increases number of opioid prescriptions of individuals who turn age 65 in the first quarter of the year in a range of 8.3-11 percentage points. The increase in opioid consumption is statistically significant both at the extensive and intensive margin. The effect is stronger on the consumption of pure narcotic analgesic drugs than others. It does not, however, affect the consumption of narcotic combination drugs and the consumption of those individuals that turn age 65 in the second quarter of a given year.