TABLE I.
Difference-in-Differences Estimates: Drug Overdose Death Rate
| Nontriplicate × | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
|
| ||||
| Panel A: All drug overdose deaths per 100,000 | ||||
| 1996–2000 | 1.173** [0.390, 2.374] | 1.290*** [0.421, 2.449] | 1.267** [0.062, 2.274] | 1.229** [0.017, 2.483] |
| 2001–2010 | 3.667** [1.521, 6.210] | 4.488*** [2.201, 6.395] | 3.561*** [1.321, 5.687] | 3.232** [1.011, 5.318] |
| 2011–2017 | 6.061** [2.812, 9.371] | 7.806*** [4.023, 10.439] | 5.240*** [3.213, 7.274] | 4.714*** [1.811, 7.253] |
| Joint p-value | .016 | .000 | .001 | .015 |
| Weighted | No | Yes | Yes | Yes |
| Covariates | No | No | Yes | Yes |
| Region-time dummies | No | No | No | Yes |
| Mean 1991–1995 | 3.890 | 4.436 | 4.436 | 4.436 |
| N | 1,377 | 1,377 | 1,377 | 1,377 |
| Panel B: Opioid overdose deaths per 100,000 | ||||
| 1996–2000 | 0.634** [0.083, 1.573] | 0.620** [0.112, 1.614] | 0.725 [−0.244, 1.621] | 0.821* [−0.189, 1.761] |
| 2001–2010 | 2.614** [1.115, 4.382] | 2.940*** [1.232, 4.249] | 2.081** [0.151, 4.192] | 2.271** [0.297, 4.402] |
| 2011–2017 | 5.002** [1.480, 8.292] | 5.899*** [1.764, 8.895] | 3.334*** [1.415, 5.613] | 3.284** [0.703, 6.012] |
| Joint p-value | .039 | .010 | .034 | .118 |
| Weighted | No | Yes | Yes | Yes |
| Covariates | No | No | Yes | Yes |
| Region-time dummies | No | No | No | Yes |
| Mean 1991–1995 | 1.189 | 1.476 | 1.476 | 1.476 |
| N | 1,377 | 1,377 | 1,377 | 1,377 |
Notes:
Significance 1%
significance 5%
significance 10%.
Outcome is all drug overdose deaths or opioid overdose deaths per 100,000. The reported coefficients refer to the interaction of the given time period and an indicator for whether the state did not have a triplicate program in 1996. Estimates are relative to preperiod 1991–1995. 95% confidence intervals reported in brackets are estimated by clustered (by state) wild bootstrap. All models include state and year fixed effects. Covariates include the fraction non-Hispanic white, fraction non-Hispanic Black, fraction Hispanic, log of population, fraction with college degree, fraction ages 25–44, fraction ages 45–64, and fraction ages 65+. “Joint p-value” refers to the p-value from ajoint hypothesis test that all three nontriplicate post effects are equal to zero and is also estimated using a restricted wild bootstrap.