WORKING PAPERS
Licensing Policy Stringency and Crash Outcomes: Triple-Difference Evidence from U.S. States
Abstract: This paper examines how driver licensing stringency affects road safety outcomes using state-level policies and fatal accident data from 1994-2022. I develop a comprehensive stringency measure using structural equation modeling and employ a triple-difference approach exploiting variation across states, time, and crash types. Identification is validated through reverse causality and parallel trends tests. Stricter licensing requirements significantly reduce young driver fatal accident rates, with young drivers experiencing net reductions of 4.5-5.5 fatal crashes per 100,000 licensed drivers but facing 7.7 percentage point licensing rate reductions. The pattern of results suggests policies operate through both screening out high-risk drivers and improving behavior among the licensed. Insurance markets show limited incorporation of policy-driven risk reductions into pricing.
Full paper available on request (PDF; slides also available).
WORKS IN PROGRESS
Accident Affinity Model: Multiplicative Analysis of Age-Specific Crash Interactions
Abstract: This research develops a novel multiplicative model to analyze age-specific crash propensities and interaction effects. Using comprehensive crash data from 1994-2022, the model employs stacked regression with age-group-by-year fixed effects to capture temporal changes in crash behavior across demographic groups. Findings reveal that teen crash propensity declined by 28% over the study period while maintaining stable interaction patterns with other age groups. The multiplicative structure demonstrates that targeted interventions for specific age groups generate systemic safety benefits across the entire driving population through reduced interaction effects.