Working Papers

Time-Limited Subsidies: Optimal Taxation with Implications for Renewable Energy Subsidies (with Owen Kay)

Accepted at the Journal of Political Economy

Pigouvian subsidies are efficient, but real-world subsidies are not Pigouvian when they have a limited duration. We show that such “time-limited subsidies” must be paired with investment subsidies to be efficient and that the change in production when a subsidy ends is the sufficient statistic for the optimal subsidy duration. We examine the US Renewable Energy Production Tax Credit for wind energy and find that firms reduce output by 5-10% when the ten-year subsidy ends, resulting in a loss of 6,000 GWh of energy annually—a figure that will increase as more turbines age out of subsidization.

Download Draft (Draft: March 2025)

2020: The Outstanding Third Year Paper Award

 
 

Monthly Capacity Factor (Seasonality Adjusted)

  • Population
  • Population

Why Choose Career Technical Education? Disentangling Student Preferences from Program Availability (with Brian Jacob)

Reject and Resubmit at the Journal of Human Resources

This paper presents the first evidence of how students make career technical education (CTE) course-taking decisions. Among the universe of Michigan high-schoolers we find large disparities in CTE access and participation by gender, race, and income. We decompose participation gaps between supply (access) and demand (preferences) with a simple discrete choice model. We find that student preferences for CTE content drive participation gaps by gender, inequities in access drive gaps by income, and school-level supply and demand factors combine to create the gaps by race. Policy simulations highlight the importance of accessible CTE delivery models within comprehensive high schools.

Download Draft (Draft: May 2023)

Media Coverage: Work Shift News

 
 

Number of CTE Programs in School

  • Black
  • White
  • Black
  • White

Self-Selection Around Policy Recommendations: The Case of Kindergarten Entry

Job Market Paper 2021-2022

This paper quantifies the efficiency and equity implications of influencing behavior with policy recommendations versus requirements in the context of kindergarten entry. Because recommendations allow opting in and opting out (“self-selection”), welfare comparisons depend on how gains accrue to those who self-select (“selection on gains”). Using variation from multiple birthday-based discontinuities, I estimate these heterogeneous effects among the universe of kindergartners in Michigan public schools and find that only higher-income families positively select on test-score gains. Consequently, whereas self-selection raises average test scores of higher-income families, it reduces those of lower-income families, creating an equity-efficiency tradeoff.

Download Paper (Draft: Dec 2024)

2021: John E. Parker Memorial Prize in Labor Economics

 
 

Money Well Spent? Partisan Reasoning and Polarized Support for Higher Education (with Reuben Hurst and Andrew Simon)

Submitted

We conduct three experiments to study the causes and consequences of polarized demand for public spending on higher education. In the first survey experiment we find that information reduces cross-party gaps in ideal policies by up to 32% and polarized communication with officials by 23%. We identify partisan differences in information processing as a key mechanism in the second experiment. Finally, by sending these letters to elected officials in a natural field experiment, we find that receiving constituent demand increases policymakers’ engagement on higher education issues. Our findings highlight how polarization may stem from diverging perceptions—not solely heterogeneous tastes.

Download Draft (Draft: March 2025)

 
 

Ideal Per-Student Spending

  • Control
  • Information Treatment
  • Control
  • Information Treatment

Loopholes and the Incidence of Public Services:
Evidence from Funding Career & Technical
Education (with Thomas Goldring, Brian Jacob, and Dan Kreisman)

Submitted

In 2015, Michigan increased its Career and Technical Education (CTE) funding and changed its funding formula to reimburse programs based on student progression through program curricula. Although this change nearly doubled program completion rates, student enrollment and persistence were unaffected; instead, administrators accelerated student progress by reorganizing course curricula around notches in the new funding formula. As a result of response heterogeneity, 30% of the funding increase is transferred away from high-poverty districts to more affluent ones, underscoring how supply-side responses to loopholes shape the incidence of public services.

Download Draft (Draft: March 2024)

 
 

Dollars of State CTE Funding Per Student

  • High-Poverty Schools
  • Other Schools
  • High-Poverty Schools
  • Other Schools

Capital Utilization in Production Function Estimation (with Michael Gmeiner)

A method to simultaneously estimate capital utilization rates and production functions is developed that exploits the fuel input choices of producers. Using Chilean data from 1979-1996 it is shown that the elasticity of output with respect to labor is systematically lower when accounting for capital utilization, and the elasticity of output with respect to capital is systematically higher. Total factor productivity is shown to be procyclical even when estimated independently of capital utilization. Output elasticities and capital utilization exhibit correlations with input quantities and investment which justify the validity of results.

Original SSRN Draft (Significant Revisions including Census RDC Proposal in Process)

 
 

Work in Progress

Effect Heterogeneity and Optimal Policy: Getting Welfare Added from Teacher Value Added (with Tanner Eastmond, Nathan Mather, and Julian Betts)

Though ubiquitous in research and practice, mean-based “value-added” measures may not fully inform policy or welfare considerations when policies have heterogeneous effects, impact multiple outcomes, or seek to advance distributional objectives. In this paper we formalize the importance of heterogeneity for calculating social welfare and quantify it in an enormous public service provision problem: the allocation of teachers to elementary school classes. Using data from the San Diego Unified School District we estimate heterogeneity in teacher value-added over the lagged student test score distribution. Because a majority of teachers have significant comparative advantage across student types, allocations that use a heterogeneous estimate of value-added can raise scores by 34-97% relative to those using only standard value-added estimates. These gains are even larger if the social planner has heterogeneous preferences over groups. Because reallocations benefit students on average at the expense of teachers' revealed preferences, we also consider a simple teacher compensation policy, finding that the marginal value of public funds would be infinite for bonuses of up to 14% of baseline pay. These results, while specific to the teacher assignment problem, suggest more broadly that using information about effect heterogeneity might improve a broad range of public programs—both on grounds of average impacts and distributional goals.

 
 
 

Dynamic Inefficiencies of Production Subsidies: Technological Progress, Investment, and the Race for Wind Resource

Firms often face simultaneous trade-offs when making entry and location decisions. For example, in the US wind industry, location decisions face a static tradeoff between wind resource and proximity to transmission, and entry decisions face a dynamic trade-off between early entry in preferred locations and waiting for technological progress. I show that market failures like inter-firm externalities from transmission investment can break Pigouvian subsidies calibrated only to pollution externalities, generating dynamic distortions by inducing entry at inefficient times and in inefficient places. The key parameters for this distortion are the effect of the market imperfection on entry timing and the degree of complementary between productivity (in this case wind resource) and technology (feasible turbine rotor diameter). I estimate the responsiveness to transmission spillovers using an event study design and find that entry increases by 10% after new infrastructure is built near a feasible site.

 
 

New Turbines Constructed After Grid Expansion

  • New Turbines
  • New Turbines

Broader Horizons: The Long-Run Impacts of Exposure to New Places (with Tanner Eastmond)

Data collection complete