https://hdl.handle.net/1721.1/81436 2026-04-05T01:28:54Z 2026-04-05T01:28:54Z Caputi, Theodore https://hdl.handle.net/1721.1/139841 2022-02-04T03:21:07Z 2022-02-03T00:00:00Z

Recreational Marijuana Dispensaries and Fatal Car Crashes Caputi, Theodore Fatal car crashes are a leading cause of death among younger Americans and have become a central concern in the US marijuana policy debate. I provide new evidence on the effect of marijuana on traffic fatalities by exploiting zip code level data on the opening of recreational marijuana dispensaries in five US states. My intra-state differences-in-differences approach both increases power relative to past analyses and eliminates the potential of time-varying state-level confounding. I find that recreational marijuana dispensary openings increased the rate of fatal car crashes by approximately 6%. These effects are not observed in a placebo analysis of retail pharmacy openings, and the increase in fatal crashes is concentrated at nighttime, after most dispensaries have closed. Collectively, these findings suggest that the effect on fatal car crashes is driven by marijuana impairment rather than increased traffic.

2022-02-03T00:00:00Z Andrews, Isaiah https://hdl.handle.net/1721.1/89188 2019-04-11T14:16:27Z 2014-09-04T00:00:00Z

Robust Two-Step Confidence Sets, and the Trouble with the First Stage F-Statistic Andrews, Isaiah When weak identification is a concern researchers frequently calculate confidence sets in two steps, first assessing the strength of identification and then, on the basis of this initial assessment, deciding whether to use an identification-robust confidence set. Unfortunately, two-step procedures of this sort can generate highly misleading confidence sets, and we demonstrate that two-step confidence sets based on the first stage F-statistic can have extremely poor coverage in linear instrumental variables models with heteroskedastic errors. To remedy this issue, we introduce a simple approach to detecting weak identification and constructing two-step confidence sets which we show controls coverage distortions under weak identification in general nonlinear GMM models, while also indicating strong identification with probability tending to one if the model is well-identified. Applying our approach to linear IV we show that it is competitive with approaches based on the first-stage F-statistic under homoscedasticity but performs far better under heteroskedasticity.

2014-09-04T00:00:00Z Auclert, Adrien Rognlie, Matthew https://hdl.handle.net/1721.1/89187 2019-04-11T14:16:27Z 2014-09-04T00:00:00Z

Unique Equilibrium in the Eaton-Gersovitz Model of Sovereign Debt Auclert, Adrien; Rognlie, Matthew We provide a proof that Markov Perfect equilibrium is unique in the standard infinitehorizon incomplete-market model with a default option which, following Eaton and Gersovitz (1981), has become a benchmark for quantitative analyses of sovereign debt (Arellano (2008), Aguiar and Gopinath (2006), Aguiar and Amador (2014)).

2014-09-04T00:00:00Z