Unemployment Insurance Claims After the LA Wildfire (with Peter Mannino, Roozbeh Moghadam, Jacob Morris, and Till von Wachter). California Policy Lab, 2025.
Administrative burdens and labor market outcomes: Evidence from California’s unemployment insurance modernization
Abstract: Using administrative UI data from California's Employment Development Department (EDD), I examine how administrative frictions in the early UI lifecycle affect spell duration and subsequent labor market outcomes. First, drawing on newly assembled data on UI applications, I characterize the distribution of delays between application filing and first benefit payment and document how these vary across claimant demographics, application type, and macroeconomic conditions. These descriptive findings establish that application-stage delays are widespread, systematically distributed, and concentrated in precisely the window prior evidence identifies as most consequential. Second, to estimate causal effects of early-spell liquidity on UI duration and reemployment, I exploit EDD's phased 2015 rollout of UI Online — an online certifications portal made available to approximately 400,000 active claimants over ten weeks. The rollout substantially reduced payment processing times, with most UI Online claims paid within one day. I use the plausibly exogenous timing of claimants' access to UI Online to estimate effects on UI spell duration, post-UI job match quality, and long-run earnings trajectories, with particular attention to heterogeneity by claimants' position in the unemployment spell at adoption.
Expedited justice? Assessing procedural reform in U.S. asylum court (using EOIR case data accessed as TRAC fellow)
Abstract: Can courts be sped up without sacrificing fairness? Existing evidence in criminal and civil contexts suggests that streamlining courts can cause net positive case outcomes, but this may not generalize to other legal settings. I study the equity-efficiency tradeoff of speed in immigration court, an especially complex legal system where defendants often face language barriers and lack institutional knowledge. Using publicly available administrative case data from the Executive Office for Immigration Review, I analyze the Dedicated Docket (DD), a 2021 policy that mandates judges decide cases within 300 days for certain families seeking asylum in the U.S. To estimate the policy’s effects, I exploit the sudden implementation of DD and instrument DD placement with families’ U.S. entry date relative to the policy’s start date. I find that DD placement significantly reduces time to case completion, confirming efficiency gains. Simultaneously, DD placement increases the probability of a removal order, indicating compromised equity. My findings show that procedural reform can generate unintended distributional consequences for vulnerable populations and underscore the need to account for institutional capacity in the design of legal reforms.
Estimating responses to changes in immigrant EITC eligibility
Abstract: The poverty alleviating effects and labor supply responses to the earned income tax credit (EITC) are well documented, but little is known about immigrant participation in the program. Since 1996 welfare reform, filing has required a social security number (SSN), thus excluding undocumented immigrants from claiming the federal tax credit. Since 2020, 11 states that offered a supplemental EITC expanded eligibility to tax filers using an individual taxpayer identification number (ITIN), effectively removing eligibility restrictions on undocumented immigrants. Using the Current Population Survey March supplement, I estimate the poverty and labor effects on likely undocumented immigrants with a difference-in-difference model. I find no evidence that changes in eligibility affect extensive margin labor supply among working age undocumented immigrants or that newly eligible EITC filers are less likely to have household incomes below the poverty line.