"Political Preferences and Migration Decisions of College-Educated Workers" (with Jinci Liu) [R&R: AEJ Applied, Latest version: August 2024]
In brief: We study the consequences of political polarization along educational lines in the United States. We show this type of polarization has increased dramatically in the last 15 years. We show that as a result, the election of a Republican governor substantially reduces new migration of college graduates into the state. We extend a spatial model of sorting to quantify the consequences for state economic activity and the spillovers onto other states.
"Top-Down Political Control of Prosecutorial Discretion" (with Ben Grunwald) [Reject and resubmit: AEJ Policy, Latest version: November 2024]
In brief: In 2003, Attorney General John Ashcroft issued a memo dramatically reducing the discretion of federal prosecutors for deciding what charges to file. We use an obscure rule in the Senate confirmation process to generate exogenous variation in compliance across the 90 federal prosecutorial districts. We find that district directors with more partisan loyalty implemented the memo more faithfully, and this led to a large increase in quits among prosecutors in response.
"Rules vs. Discretion in Allocating Labor Union Financial Investigations" (with Chloe Nibourel) [Latest version: December 2024]
In brief: There are widespread calls to replace bureaucratic discretion with rule-based decision making based on observable data. We study the agency that investigates US labor unions, where these calls have been made. We propose a method to infer in which district offices these strategies are used, and use the turnover of district directors. Most rule-based, data-driven allocations perform worse than bureaucratic discretion, but a minority of approaches perform better.
"The Freezing Point of History: Extreme Temperatures Hinder Elite Emergence" (with Arash Nekeoi, Fabian Sinn, and Jeffrey Shrader) [Latest version: January 2025]
In brief: We study how harsh environmental conditions -- the type shown to affect infant health and common in the developing world -- affect the emergence of the types of history-making elites whose extraordinary contributions shape long-run development. Harsh winters raise mortality, reduce education and earnings (especially the probability of top outcomes) and reduce the number of people from affected birth cohorts who leave a mark on history (as measured by having a Wikipedia page).
"Policing Misperceptions: Experimental Evidence on Reform Openness" (with John Kuk and Matthew Nanes) [Latest version: August 2024, Pre-analysis plan]
In brief: We test whether providing basic information about policing practices -- without overt attempts at persuasion -- can facilitate compromise on support for police reform. Voters generally underestimate the prevalence of reform-oriented practices. But while providing accurate information makes Republicans more open to reform, the same information makes Democrats more extreme and skeptical towards reform.
"When Do Super-Majority Rules Moderate?" (with Anton Arbman Hansing and Daniel Spiro) [Latest version: August 2023]
In brief: Standard theoretical approaches to analyzing supermajority rules have an implicit assumption that mechanically implies that these rules must moderate legislation. We point this out, show empirically this assumption is untrue, and analyze a new model in which supermajority rules might moderate legislation, or might polarize it by forcing compromise with party extremists. We show novel trends of a rise in extremism in the Senate, and that this has eroded the moderating value of the filibuster.
"Unit-specific trends lead to biased estimates of average treatment effects in difference-in-difference settings" [Latest version: November 2024, Key simulation (very simple Stata program)]
In brief: Never run this regression: Y_{it} = a_{i} + c_{i}t + d_{t} + B(Treat_{i} x Post_{t}) + e_{it} (i.e., one with unit-specific trends and a single post-treatment dummy)
"(Not) feelin' the (cross) pressure: Experimental evidence on issue misalignment and polarization" (with John Kuk and Matthew Nanes) [Latest version: December 2024, Pre-analysis plan]
In brief: Despite high levels of partisan polarization, even for the most controversial issues, a significant minority of voters hold the views of the opposite party. We study this in the context of police accountability. We show that most voters who hold mainstream views of the opposite party over-estimate how many within their own party agree with them, and under-estimate how many in the opposite party do. When informed of actual rates of agreement, these voters become more moderate and open to compromise.
"Congressional Elections and Union Officer Prosecutions" [Review of Economics and Statistics (2023): 105 (5), Policy Brief, Longer never-to-be-published version, Replication package, Original 2016-2017 JMP]
In brief: I show that members of Congress influence which labor unions have officers investigated and indicted. I use novel data on indictments, campaign contributions to link unions to Congressional candidates, and a regression discontinuity to estimate the effects of the union-supported vs. union-opposed candidate taking office. I find that both pro-union candidates protect supporters and anti-union candidates target opponents. Both formal and informal mechanisms of influence seem important.
"The Effects of Import Competition on Unionization" (with John Ahlquist) [American Economic Journal: Economic Policy (2023): 15 (4), Policy Brief, Online Appendix, Replication package]
In brief: We study how Chinese import affected US unionization, drawing on others' identification strategies. Within manufacturing, imports cause modest declines in union density. Across place, however, imports drive large increases outside manufacturing. This is not from would-be manufacturing workers (who shift to low-wage jobs), but their spouses and children (who shift to high paying jobs in healthcare and education). Overall Chinese imports prevented 25% of the decline in US unionization.
"Adjusting to Rain Before it Falls" (with Nels Lind and Jeffrey Shrader) [Management Science (2023): 69 (12)]
In brief: We use geographic variation in how responsive weather patterns are to changes in Pacific Ocean temperatures (ENSO) to estimate the effects of rainfall on construction employment, depending on how far in advance the increased rainfall could be predicted. We show much greater responsiveness to more predictable rainfall (which cannot be reconciled with frictionless perfectly competitive labor markets) and trace out implications of climate change induced rainfall volatility for workers and firms.
"Agency Breadth and Political Influence" (with Zachary Breig) [Journal of Economic Behavior and Organizations 188 (2021): 253-268, Online Appendix, Policy Brief]
In brief: We study a legislature's attempt to influence a bureaucratic agency. We present a model where legislative influence has spillovers onto other tasks the agency performs. As agencies become broad, these spillovers become more consequential and equilibrium influence decreases. We find robust support for this prediction among 70 US federal agencies, and rule out several alternative explanations.
"Did the War on Terror Deter Ungoverned Spaces? Not in Africa" [Journal of Development Economics 151 (2021): 102648, Online Appendix, Policy Brief]
In brief: Following the 9/11 terrorist attacks, the Bush Administration's foreign policy demanded countries take active steps to control ungoverned spaces or risk a US invasion. I develop a difference-in-difference strategy to determine whether African governments attempted to reclaim their territory in response. I find robust, conclusive evidence that they did not. I can rule out small effects. I argue that broad-based deterrence is unlikely to reduce ungoverned space.
"Partial Automation and the Technology-Enabled Deskilling of Routine Jobs" [Labour Economics 69 (2021): 101973, Online Appendix, Policy Brief, Oren Danieli's similar but better paper]
In brief: I argue that standard theories of automation ignore the fact that technology rarely replaces a routine workers entirely. Rather, it typically simplifies tasks, allowing less-skilled workers with machines to perform them. As a result, a high minimum wage makes automation strategies more expensive by raising the cost of low-skilled "machine users," thereby slowing the replacement of routine jobs. Using a triple-difference strategy, I find support for this claim among US states.
"Public-Private Competition for Workers: Spillovers from Government Pay Adjustments" (with Ziho Park and Gustavo de Souza)
In brief: We study the implementation of federal locality pay adjustments, which raised federal workers' wages in some cities not others. We study the spillovers onto private sector employers in affected cities and industries. We find pay increases drove up private sector pay, but didn't affect employment or the number of firms.
"Forward-Looking Labor Adjustment" (with Nels Lind and Jeffrey Shrader)
In brief: We propose an approach to use variation in the timing of forecasts to infer adjustment frictions. We implement this using rainfall forecasts and construction employment. Adjustment frictions are large.
"Democratic Accountability and Municipal Policing in France" (with Chloe Nibourel)
In brief (project funded, data collection ongoing): We explore the effects of the rise of municipal police departments in France on low-income and immigrant youth there.
"Substantially Equivalent? Enforcement of Housing Discrimination Laws and School Segregation"
In brief: The 1968 Fair Housing Act specified that states with their own anti-discrimination laws would receive significant funding and replace the federal government in enforcing the laws. I study the effects of this transfers of control. State enforcement improves desegregation, but only under Democratic administrations.
"Electronic Voting and Political Patronage" (with Mauricio Romero) [latest version: May 2014?]
"Community Monitors vs. Leakage: Experimental Evidence from Afghanistan" (with Eli Berman, Michael Callen, Luke N. Condra, Tarek Ghani, and Mohammed Isaqzadeh) [latest version: July 2017]
"Expanding Governance as Development: Evidence on Child Nutrition in the Philippines" (with Eli Berman and Joseph Felter) [Earlier version: NBER Working Paper 21849, Coverage: VoxEU, Chris Blattman, latest version: December 2016]
Note: Making public-facing replication data packages is really hard. Sharing data is easy. If I have something that you want (data or code) then I'm happy to share it. Just email me.