Date of Degree


Document Type


Degree Name





David A. Jaeger

Committee Members

Michael Grossman

Theodore Joyce

Subject Categories

Health Economics


drunk driving, ride-hailing, driver training, taxis


This dissertation comprises three essays on the economics of driving and health. The first examines the effect of the ride-hailing service Uber, its launch in New York City in May 2011, and the resulting drunk-driving outcomes in the counties that received service compared with those that did not. The difference-in-differences estimation of this effect implies a 25 to 35 percent decrease in the alcohol-related collision rate for the affected New York City boroughs, or about 40 collisions per month. With differentiated treatment effects for each affected county, the difference-in-differences effect is higher for Manhattan, middling for the Bronx and Brooklyn, and lower for Queens. A synthetic control analysis shows pronounced effects over time in the Bronx and Brooklyn, and a permutation test confirms the effect is not commonly reproducible using untreated counties.

In the second essay, I survey and evaluate the literature on the drunk-driving effects of ride-hailing services like Uber. All but one of these studies finds a measurable negative effect of ride-hailing presence on drunk-driving measures, and I replicate this outlier study with checks for robustness to empirical choices. The robustness checks do not overturn the result. Special emphasis is placed on the differences in interpretation between the five studies’ estimated results that distinguish their external validity despite their focus on the same essential research question.

In the third essay, I investigate the effect of revisions to Australian state graduated driver licensing laws on road fatalities of Australian youths between 16 and 21 years of age. While all of the Australian states and territories implemented a three-tier graduated licensing scheme for the protection and education of young drivers in the 1990s, many states also implemented revisions to those laws that increased the restrictions and hurdles for young drivers navigating the graduated licensing tiers. The difference-in-difference-in-differences estimation controls for age, state, and time effects as well as all interaction effects, but finds no measurable effect of increased driver-training restrictions on youth road fatalities. An aggregated specification comparing ever-treated youths to older cohorts finds one marginally significant, positive effect that is not robust to changes in specification.

This work is embargoed and will be available for download on Saturday, February 01, 2020

Graduate Center users:
To read this work, log in to your GC ILL account and place a thesis request.

Non-GC Users:
See the GC’s lending policies to learn more.