This paper develops a novel bootstrap procedure to obtain robust bias-corrected confidence intervals in regression discontinuity (RD) designs using the uniform kernel. The procedure uses a residual bootstrap from a second order local polynomial to estimate the bias of the local linear RD estimator; the bias is then subtracted from the original estimator. The bias-corrected estimator is then bootstrapped itself to generate valid confidence intervals. The confidence intervals generated by this procedure are valid under conditions similar to Calonico, Cattaneo and Titiunik's (2014, Econometrica) analytical correction-i.e. when the bias of the naive regression discontinuity estimator would otherwise prevent valid inference. This paper also provides simulation evidence that our method is as accurate as the analytical corrections and we demonstrate its use through a reanalysis of Ludwig and Miller's (2008) Head Start dataset.
When the running variable in a regression discontinuity (RD) design is measured with error, identification of the local average treatment effect of interest will typically fail. While the form of this measurement error varies across applications, in many cases the measurement error structure is heterogeneous across different groups of observations. We develop a novel measurement error correction procedure capable of addressing heterogeneous mismeasurement structures by leveraging auxiliary information. We also provide adjusted asymptotic variance and standard errors that take into consideration the variability introduced by the estimation of nuisance parameters, and honest confidence intervals that account for potential misspecification. Simulations provide evidence that the proposed procedure corrects the bias introduced by heterogeneous measurement error and achieves empirical coverage closer to nominal test size than "naive" alternatives. Two empirical illustrations demonstrate that correcting for measurement error can either reinforce the results of a study or provide a new empirical perspective on the data.
Este artigo examina como o desempenho relativo no mercado de trabalho de cada profissão afeta a escolha profissional dos futuros universitários. O desempenho no mercado de trabalho é medido pela média e pelo desvio padrão dos salários recebidos por profissão e pela sua taxa de desemprego no censo demográfico nos anos próximos ao vestibular. O número de pleiteantes a ingresso na carreira é medido pelo número de inscritos no exame da Fuvest. Utiliza-se dados em painel para os anos de 1991 e 2000 para controlar pelo efeito específico de cada profissão. Os resultados apontam para um efeito positivo e robusto do salário médio da profissão sobre a escolha profissional, que persiste na análise de painel, e para efeitos negativos da dispersão salarial da renda e do desemprego que, no entanto, não se mostraram significantes. This paper examines the relative performance of individuals in each profession affects the demand for this profession by college entrants. The performance in the labor market is measured by the mean and standard deviation of wages earned by individuals in each profession and by the unemployment rate in the census in the years close to the entrance exam. The demand for the profession is measured by the number of students enrolled in the entrance exam in each career. We use panel data for the years of 1991 and 2000 to control for the specific effect of each career. The results show a positive and robust effect of the average income on the demand for the profession, which persists even in the panel data analysis, and negative effects of the wage dispersion and unemployment, which were not robust, however
This study documents two potential biases in recent analyses of UI benefit extensions using boundary-based identification: bias from using county-level aggregates and bias from across-border policy spillovers. To examine the first bias, the analysis uses a regression discontinuity approach that accounts for measurement error in county-level aggregates. These results suggest much smaller effects than previous studies, casting doubt on the applicability of border-based designs. The analysis then shows substantial spillover effects of UI benefit duration on across-border work patterns, consistent with increased tightness in high-benefit states and providing evidence against a dominant vacancy reduction response to UI extensions. (JEL E24, E32, J22, J64, J65)
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.