We evaluate rationales for income-based speeding fines, as famously applied in Finland, where a $250 offense for a low-income speeder can cost $100,000 for the rich. We consider four potential policy goals—externality mitigation, redistribution, equal compliance, and proportional punishment—using linked Finnish administrative data and an original survey. First, using a "job-loss" design, we find that marginal speeding costs decrease with offender income, leading us to reject externality mitigation as a rationale. Next, we assess the redistributive rationale. In standard models, indirect taxes aid redistribution only if the taxed action serves as a signal for earnings ability. We recover this signal by differencing speeding behavior with respect to causal effects of income on speeding, estimated from within-individual earnings variation and inheritance shocks. The estimated signal is negatively correlated with earnings, implying that Mirrleesian redistribution rationalizes a lower fine on the rich. Finally, we use our survey to assess whether preferences for equal compliance or proportional punishment are motivating rationales. Respondents trade off fixed and income-based speeding fine policies, where the latter vary in the induced across-income distribution of behavior and schedule "steepness." Net of redistribution, respondents are on-average willing to forgo €144 million in government revenue to implement income-based policies. However, these valuations are insensitive to induced behavior or schedule-steepness, ruling out our candidate rationales. Given our findings, speeding fines should incorporate some income-dependence but need not feature the extremes in Finland. Residual support for income-dependence suggests that current policy reflects either unmodeled instrumental considerations or direct, non-instrumental preferences.

We identify occupancy fraud—borrowers who misrepresent their occupancy status as owner-occupants rather than investors—in residential mortgage originations. Unlike previous work, we show that fraud was prevalent in originations not just during the housing bubble but also persists through more recent times. We also demonstrate that fraud is broad-based and appears in government-sponsored enterprise and bank portfolio loans, not just in private securitization; these fraudulent borrowers make up one third of the effective investor population. Occupancy frauds obtain credit at lower interest rates, suggesting a motivation for undertaking fraud. These fraudulent borrowers perform substantially worse than similar declared investors, defaulting at a 75% higher rate. We also provide evidence consistent with fraudulent borrowers' defaults being more "strategic," suggesting that this population poses a risk in the face of declining house prices.

About 1 in 5 eligible households fail to claim the Earned Income Tax Credit (EITC), which constitutes the largest means-tested cash-transfer program in the US. According to IRS estimates (Plueger, 2009), most of these non-claimants do not file a tax return (a requirement to claim the EITC) and have incomes below the income filing threshold (the threshold above which a household is "required" to file taxes). Drawing on IRS tax-return data, I investigate the contribution of the filing threshold to incomplete EITC take-up by exploiting a 2003 reform that increased the threshold level, theoretically inducing missing mass in the filing distribution. A conceptual framework clarifies how missing mass estimates can be used to measure a lower bound on filing costs for individuals who are marginal to the reform.