• How much do mandatory minimums matter? - under review

Abstract: I estimate the causal effect of mandatory minimum (MM) eligibility on federal drug sentencing using a regression discontinuity design with extrapolation to disentangle statutory impacts from prosecutorial selection. I find that MM eligibility increases sentence length uniformly across case types by about 10 months (14%). This includes defendants with low criminal history, indicating limited protection for low-level offenders. To assess which types of cases are affected by selection, I compare extrapolated counterfactual sentence lengths against observed sentences. I find evidence that charging manipulation is localized among minority defendants. These results indicate that racial disparities in MM sentencing are driven by prosecutor charging decisions rather than by features of the MM statute.

PDF Link, Online Appendix 

• Prosecutor tradeoffs and race: evidence from a circuit split

Abstract: Prosecutors face a tradeoff between gathering evidence and pursuing charges. I examine how differences in evidentiary costs affect racial disparities in charging decisions by exploiting a circuit split in federal drug sentencing rules. Comparing high- and low-cost regimes, I find that mandatory minimum policies reduce preexisting race gaps by lowering penalties for minorities. However, when no gap exists initially, benefits accrue mainly to White defendants, creating disparities. High-cost regimes also show higher pre-policy punishment levels. The findings suggest raising evidence costs can reduce existing disparities but may introduce new ones when race gaps are absent.

PDF Link 

• Prosecutorial discretion and outcome disparities (With Andy Yuan) - R&R at American Law & Economics Review (ALER)

Abstract: We examine the causal effects of prosecutors' incentives on racial and gender disparities. Blakely v. Washington 542 U.S. 296 (2004) significantly disincentivized state prosecutors from pursuing sentence enhancements by raising their burdens of proof from "preponderance of evidence'' to "beyond a reasonable doubt.'' Through a regression discontinuity design, we find that a higher burden of proof reduces defendants' likelihood of receiving a sentence enhancement by 47%. Furthermore, we find striking evidence that Blakely eliminated the entire preexisting gender gap of men being 24% more likely to receive sentence enhancements than women. However, we find no evidence suggesting a racial gap of sentence enhancements both pre and post Blakely.

PDF Link

• Do legal actors compensate? - measuring prosecutor and judicial responses to an exogenous change in charging costs (with Andy Yuan)

• The efficacy and equity of reducing traffic stops (with Jacob Meyer)

• Physical environment and crime: evidence from streetlight modernization (resting)