Abstract: Using a new dataset linking parents' bankruptcy filings to administrative records on children, and quasi-random variation in judge leniency, we document that recipients' children earn $2,145 more annually, have 5.7pp higher homeownership by age 30, and exhibit greater upward mobility. We collect case-level information on debt discharged and estimate a benefit of $2.92 in the present value of children's lifetime earnings for each dollar of relief, better than several child-focused interventions. Effects are largest for families in disadvantaged neighborhoods and with fewer resources — populations that receive less relief. Asset protection, increased educational investment, and geographic stability contribute to these results.

Abstract: We provide evidence that contemporaneous opportunistic trading by CEOs and independent directors indicates private information sharing and propose such trading as a new, observable measure of independent directors' informedness. We show that CEOs are more likely to transfer private information in complex firms where outsiders’ advice is likely more valuable, and when greater CEO power shelters them from closer monitoring scrutiny. We show that information sharing positively impacts the firm through enhanced future innovation, M&A, and investment outcomes, and improved firm performance, despite having a negative impact on monitoring outcomes. Our period-specific measure also enables a re-measurement of board independence.

Abstract: We document a political partisanship-based assessment gap that imposes disproportionate fiscal burdens on political minorities. Using property tax data matched with voter registration records across the United States, we find that political minorities—Republicans in Democratic-majority counties and Democrats in Republican-majority counties—face higher property tax assessments than the political majority within the same tax jurisdiction, despite being subject to identical tax administration and rates. This partisan assessment gap is economically significant, representing 40-50% of the previously documented racial assessment gap. In Republican counties, the gap is driven by within-neighborhood variation, while in Democratic counties, it is driven by variation across neighborhoods, enabled by higher levels of residential partisan segregation. Leveraging novel hand-collected historical data on assessor identities and their political affiliations, we show that assessor bias contributes to these gaps: disparities are weaker when the assessor shares the political affiliation of the minority group and larger otherwise. Moreover, the gap's magnitude varies with the partisan composition of county commissions---increasing with aligned political control and decreasing with minority party representation. These findings demonstrate that property tax assessments, though ostensibly neutral, become tools for redistributing fiscal burdens along partisan lines and generate inequities in homeownership costs.