with John L. Mikesell, and Justin M. Ross, 2019.
Can the federal budget be constrained using congressionally self-imposed rules? The Budgetary Enforcement Act of 1990 (BEA90), in effect from 1992 to 2002, is frequently held up as a political control that largely succeeded in doing so and its expiration is considered a watershed event. This paper finds evidence supportive evidence of BEA90 as a constraint on non-defense discretionary outlays from a synthetic control using Lasso regression methodology. The effect of BEA90 is estimated to be around 10 percent, but the sampling variation suggests that the probability these results arose from chance to be around 30 percent. Confronted with a common applied empirical problem of having a large number of potential predictors (over 2,000) and a relatively short time-series (12 years), we propose novel additional checks that successfully cross-validate the synthetic control’s counterfactual model.
With Philipp Doerrenberg and Danyang Li, 2022.
We estimate the impact of employers' sex and race on the labor supply decisions of workers using data generated in an online labor market. The labor task requires workers to transcribe information from gasoline receipts, and employer's race and sex are signaled via a photograph of a hand holding receipts. Our empirical analysis reveals several findings. First, workers are more likely to correctly identify the race and sex of white and male employers compared to female and black employers. Second, we find mixed evidence of discrimination on the extensive margin. On the one hand, those who correctly identify the race (sex) of the employer are more likely to work for black (female) employers compared to white (female) employers. On the other hand, those who incorrectly identify race (sex) are less likely to work for black (female) employers compared to white (male) employers. Third, white workers transcribe fewer pictures and do so less accurately for black employers relative to white employers. Fourth, both male and female workers transcribe more pictures and do so more accurately for female employers relative to male employers. Finally, results from a survey of mTurkers suggest that the gaps we identify are not driven by statistical discrimination.
with Philipp Doerrenberg, 2019
How do firms' avoidance and evasion opportunities affect market prices? We investigate the causal link between tax-evasion opportunities and prices in a situation where firms remit sales taxes and have access to tax-evasion possibilities. In light of difficult causal identification with archival data, we design a controlled experiment in which buyers and sellers trade a fictitious good in competitive markets. A per-unit tax is imposed on sellers, and sellers in the treatment group are provided the opportunity to evade the tax whereas sellers in the control group are not. We find that the equilibrium market price in the treatment group is lower than in the control group, and the number of traded units is higher in treatment markets. The results further show that the after-tax incomes of sellers in the evasion treatment increases despite trading at lower prices. Our findings have implications for tax incidence. In particular, sellers with access to evasion shift a smaller share of the nominal tax rate onto buyers relative to sellers without tax evasion opportunities. Additionally, we find that sellers with evasion opportunities shift the full amount of their effective tax rate onto buyers. Results from additional experimental treatments show that this full shifting of the effective tax burden is due to the evasion opportunity itself rather than the evasion-induced lower effective tax rate.
Income equality and labor effort: evidence from NHL salary caps.
with Philipp Doerrenberg, Andreas Peichl, and Daniel Waldenstrom
Budget Games: Impact of performance measures and priorities on budget allocation
with Whitney Afonso and Justin Ross
Tax policy changes and food donations
with Shellye Suttles
Tax Reform and Income Inequality: Evidence from the Jamaican kink
with Oronde Small