Research
Peer-Reviewed Publications
Does financial aid affect voting? Using California's Cal Grant and a GPA cutoff, we estimate causal effects on registration and turnout—with implications for student aid design.
Do partisan election officials tilt outcomes? Using regression discontinuity around close clerk elections, we find they don't—evidence that U.S. election administration is more robust than critics suggest.
Who runs U.S. elections? First comprehensive analysis of local election officials—their authority, demographics, and variation across states.
Working Papers
For a century, scholars have debated whether ballot measures educate or overwhelm voters. Using 1,547 measures merged with 480K survey responses (2006–2020), I show that traditional two-way fixed effects initially support Matsusaka's "educative hypothesis"—but these findings evaporate under modern DiD estimators for staggered treatment. Even high-salience morality measures fail to produce lasting civic engagement effects.
Does the president's party shape who receives FEMA relief? Using an RD design around close House elections (2001–2020), I find evidence against standard favoritism predictions. Mechanism tests point to compositional differences in disaster vulnerability rather than strategic targeting: effects concentrate in white, affluent districts and don't respond to electoral cycles.
Does minority representation on school boards reduce racial discipline gaps? Using CVRA-induced transitions to single-member district elections across 230 California school districts (2011–2024) and heterogeneity-robust DiD estimators (Sun-Abraham, Callaway-Sant'Anna, BJS imputation), I find that the Black-White suspension gap increases by 2–3 percentage points. I validate the finding with randomization inference, triple-difference, and HonestDiD sensitivity analysis. BISG-predicted board race composition reveals the mechanism: SMD elections primarily elect Hispanic, not Black, members, and only same-race representation reduces the discipline gap.
Projects
Built CalBallot, a searchable database of 12,000+ California ballot measures across all 58 counties (1998–2026), from local school bonds to statewide propositions. The data pipeline aggregates records from multiple sources (Ballotpedia, CA Secretary of State, CEDA, NCSL, ICPSR), deduplicates via fingerprinting, and enriches with AI-generated summaries. Includes regional browsing, advanced filtering, data exports (CSV/JSON), and a BYOLLM AI chat interface. Explore CalBallot →
Policy Research
Health Workforce (UCSF Healthforce Center)
- California's Primary Care Workforce: Supply Characteristics and Pipeline 2017
- Forecasted Supply, Demand, and Pipeline of Trainees, 2016–2030 2017
- California's Current and Future Behavioral Health Workforce 2018
D.C. Health Workforce (D.C. Policy Center)
First publicly available analysis of D.C.'s health workforce. Combined ACS data, FOIA records, and original geocoding.
- COVID-Era Health Care Workforce Capacity in Washington, D.C.
- Inequalities in Health Care Need and Demand Across the District
- Who Is Providing COVID-19 Care in the Washington Metropolitan Area?
- Licensed Health Care Clinicians Increased During the Early Pandemic
Government Technology (GovExec.com)
Analysis & Opinion
Clips available upon request.