Why Unions Still Matter: The Effects of Unionization on the Distribution of Employee Earnings (Job Market paper)
Abstract: This paper estimates the causal effect of unionization on the distribution of employee earnings using a regression discontinuity design that links administrative records on individual earnings to union certification election results. The results suggest unions raise the lower end of the distribution by around 30 log points, with a much smaller effect on the upper tail, and a modest effect on average earnings. Estimates of average effects by baseline earnings quantile suggest the distributional effects correspond to individual-level earnings effects that vary by skill. Unionization also appears to reduce employment of the lowest skilled workers. These results are consistent with a model of union wage setting in which unions set wages so as to maximize the probability of certification, subject to a minimum profit constraint for the employer. The optimal union wage schedule pays low-skilled union members above marginal product but reduces the return to skill. The estimates suggest that about one quarter of the increase in the variance of log earnings from 1979 to 2009 can be accounted for by falling U.S. private sector unionization rates, a larger fraction than earlier studies have found.
Exact inference for a weak instrument, a small sample, or extreme quantiles
Abstract: This paper describes a randomization-based inference procedure for the distribution or quantiles of potential outcomes for a binary treatment and instrument. The method imposes no parametric model for the treatment effect, and remains valid for small n, a weak instrument, or inference on tail quantiles, when conventional large-sample methods break down. The method is illustrated using simulations and data from a randomized trial of college student incentives and services.
Quantile Treatment Effects in the Regression Discontinuity Design (forthcoming at Journal of Econometrics, with Markus Froelich and Blaise Melly)
Abstract: We introduce a nonparametric estimator for local quantile treatment effects in the regression discontinuity (RD) design. The procedure uses local distribution regression to estimate the marginal distributions of the potential outcomes. The estimator attains the optimal rate of convergence for one-dimensional nonparametric regression and is asymptotically normal. We illustrate the procedure through Monte Carlo simulations and an application on the distributional effects of a universal pre-K program in Oklahoma. We find that participation in a pre-K program significantly raises the lower end and middle of the distribution of test scores.
The effects of public sector collective bargaining rights
Abstract: Widespread public sector unionism emerged only in the 1960s, as individual states opened the door to collective bargaining for state and municipal workers. I exploit differences in timing across states to construct estimates of the causal effects of public sector collective bargaining rights on pay, benefits, and employment for teachers, fire fighters, and police. Perhaps surprisingly, estimates that allow for state fixed effects and state-specific trends show little effect on teachers' pay, hours, benefits, or employment, despite significantly increasing union presence among teachers. For police and fire fighters, however, the results show modest positive effects on pay and benefits.
Did Vietnam Veterans Get Sicker in the 1990s? The Complicated Effects of Military Service on Self-Reported Health (Journal of Public Economics 2010, With Joshua Angrist and Stacey Chen)
Abstract: The veterans disability compensation (VDC) program, which provides a monthly stipend to disabled veterans, is the third largest American disability insurance program. Since the late 1990s, VDC growth has been driven primarily by an increase in claims from Vietnam veterans, raising concerns about costs as well as health. We use the draft lottery to study the long-term effects of Vietnam-era military service on health and work in the 2000 Census. These estimates show no significant overall effects on employment or work-related disability status, with a small effect on non-work-related disability for whites. On the other hand, estimates for white men with low earnings potential show a large negative impact on employment and a marked increase in non-work-related disability rates. The differential impact of Vietnam-era service on low-skill men cannot be explained by more combat or war-theatre exposure for the least educated, leaving the relative attractiveness of VDC for less skilled men and the work disincentives embedded in the VDC system as a likely explanation.
Long term effects of military service on the distribution of earnings
Abstract: I estimate the long term effect of military service on quantiles of earnings and education using the Vietnam draft lottery eligibility status as an instrument. I compare the local quantile treatment effect estimator studied by Abadie, Angrist, and Imbens (2002) to the instrumental variables quantile regression technique developed by Chernozhukov and Hansen (2008). Ordinary quantile regression shows a large negative association between service in Vietnam and earnings of white men, with the effect increasing in magnitude for the upper quantiles. Quantile treatment effects estimates show the opposite pattern, although much smaller in magnitude, with a small negative effect at the lower end of the distribution, and a small positive effect at the upper end. This suggests the ordinary quantile result is due to heterogeneous selection effects. The two methods of quantile treatment effects estimation give similar results.