We analyze the welfare impacts of a reform that expanded an eligibility notch in one of the world's largest cash transfer programs, Bolsa Famíilia. We develop a novel framework to bound the welfare impacts of reforms to transfer programs featuring notches using two sufficient statistics: (1) the number of households bunching at the old notch who move toward the new notch, and (2) the number of households who "jump" down to the new notch. We estimate these two statistics using longitudinal administrative data and a difference-in-difference strategy. Despite strong evidence of behavioral responses to this reform, we find that the corresponding efficiency costs are small relative to the equity benefits: the reform's MVPF is between 0.90 and 1.12. Because the Bolsa Família eligibility threshold is based on self-reported income, our findings suggest that the efficiency costs of targeting based on self-reported income can be relatively small even in high-informality settings.

This paper derives necessary and sufficient conditions for allocations to be incentive compatible in multidimensional screening problems that satisfy a generalized single crossing property. We then devise a numerical method based on these results to solve multidimensional screening problems. Importantly, our numerical method can be applied to multidimensional screening problems for which existing approaches cannot be applied. We apply this method to several numerical examples in the context of multidimensional optimal taxation. In addition to illustrating how to apply our theoretical results and implement our numerical method, our simulations highlight the importance of bunching in optimal multidimensional taxation. Finally, we prove that bunching must occur in multidimensional optimal taxation problems when the social planner has sufficiently redistributive preferences.

This paper uses schooling decisions to estimate curvature in utility of consumption in developing countries. We show that the elasticity of marginal utility of consumption (EMUC) can be recovered from three sufficient statistics: (1) the price effect of schooling; (2) the income effect of schooling; and (3) the consumption cost of sending a child to school. We estimate the consumption cost of schooling using experimental variation from the roll-out of Progresa, a conditional cash transfer program in rural Mexico. Using estimates of income and price effects in the Progresa setting from Bergstrom and Dodds (2021), we estimate the EMUC at 1.62.

This paper develops a general theory of optimal taxation with multiple dimensions of heterogeneity. The main technical hurdle in developing this theory is the possibility that individuals have multiple optimal incomes. Using a variational technique, we derive optimal tax formulas that account for the possibility that individuals have multiple optima and, hence, account for the possibility that individuals jump between their optimal income levels when we perturb the tax schedule. We quantify the magnitude of these effects, thereby augmenting the optimal tax formulas from Saez (2001) with additional  "jumping effect" terms. We provide a partial characterization of when individuals with multiple optimal incomes may exist under the optimal tax schedule. Finally, we derive a new methodology to simulate optimal income tax schedules with multidimensional heterogeneity. We implement this method, showing that individuals with multiple optimal income levels can exist under the optimal tax schedule.

Using a general labor supply model in which individuals choose how much to work conditional on productivities and preferences for consumption relative to leisure, we show that the mapping from earnings and hours worked to productivities and preferences can be expressed entirely in terms of reduced-form labor supply elasticities. We investigate the role productivities vs. preferences play in driving income inequality in the United States. Benchmark labor supply elasticity estimates from the literature imply productivities drive most of income inequality. Preferences become increasingly important, relative to benchmark, with larger income effects or larger differences between earnings and hours worked elasticities.

Conditional cash transfers (CCTs) are a popular type of social welfare program that make payments to households conditional on human capital investments in children. From a targeting perspective, compared to unconditional cash transfers (UCTs), CCTs are costly because they exclude some low-income households as access is tied to normal investments in children. However, we argue that conditionalities on children's school enrollment offer an unexplored targeting benefit over UCTs: CCTs target money to households who forgo a discrete amount of child income. We show that the size of the targeting benefit relative to the targeting cost of CCTs is directly related to the consumption differences between schooling and non-schooling households and two elasticities already popular in the literature: the income effect of a UCT and the price effect of a CCT. We estimate these elasticities for a large CCT program in rural Mexico, Progresa, using variation in transfers to younger siblings to identify income effects. In this setting, we find that the targeting benefit is a similar magnitude to the cost of excluding some low-income households; this implies that 33% of the Progresa budget should go to a CCT over a UCT based on targeting grounds alone.

This paper develops a general framework to construct optimal policy reforms. We show that if a policymaker can control how fiscal externalities are spent, the welfare-weighted marginal value of public funds (WMVPF) is the relevant sufficient statistic for determining optimal reforms. If a policymaker cannot control how fiscal externalities are spent, the welfare-weighted net social benefit (WNSB) is the sufficient statistic. If a policymaker can control how a fraction of fiscal externalities are spent, the relevant sufficient statistic is an "augmented internal WMVPF". We provide several stylized examples illustrating how constraints on fiscal externalities can have meaningful impacts on the optimal reform direction.

This paper develops a general theory to recover the inverse welfare function that rationalizes a given tax schedule as optimal. Our theory allows for complex environments including the presence of multidimensional tax schedules, bunching/jumping behavior, optimization frictions, general equilibrium effects, and externalities. We show how inverse welfare functions can be used to characterize the set of welfare-improving tax reforms and to test for Pareto efficiency. We numerically construct several inverse welfare functions for piecewise-linear income taxation, taxation with optimization frictions, joint income and property taxation, income taxation with labor demand and endogenous wages, and taxation with inequality aversion.

This paper quantifies how much people would be willing to pay to remove frictions that impede them from working their ideal number of hours using two sufficient statistics: (1) the percentage difference between ideal (i.e., frictionless) and actual hours, and (2) the Hicksian elasticity of ideal hours with respect to the after-tax wage rate. We implement this method to construct estimates of the willingness-to-pay to remove frictions in the United States and Germany. There are three core findings: (1) the cost of adjustment frictions (an omnibus measure encompassing, for example, fixed costs of adjustment, discrete choice sets, and search costs) is large for any reasonable value of the Hicksian ideal hours elasticity, even when accounting for endogenous wages, multiple labor supply decisions, and dynamic decisions; (2) the cumulative cost of adjustment frictions and tax misperceptions is even larger: individuals would be willing to pay at least 10% of their income on average to remove these two frictions in hours worked; and (3) adjustment frictions appear to be much more costly than tax misperceptions.