Article
Statistics & Probability
Matias D. Cattaneo, Luke Keele, Rocio Titiunik, Gonzalo Vazquez-Bare
Summary: The regression discontinuity (RD) design is a credible identification strategy for program evaluation and causal inference in nonexperimental settings. However, RD treatment effect estimands are local, so developing statistical methods for extrapolation of these effects is key. A new design-based method relying on the presence of multiple cutoffs is introduced, with an easy-to-interpret identifying assumption mimicking common trends. Illustration with data on a subsidized loan program in Colombia offers new evidence on program effects for students away from the eligibility cutoff.
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
(2021)
Article
Mathematical & Computational Biology
Matias D. Cattaneo, Luke Keele, Rocio Titiunik
Summary: We provide a practical guide for the analysis of regression discontinuity (RD) designs in biomedical research. The key concepts, assumptions, and estimands within both the continuity-based and local randomization frameworks are introduced. Modern estimation and inference methods within both frameworks are discussed, along with empirical falsification tests for supporting key assumptions. The importance of considering fuzzy RD designs and RD designs with discrete scores in biomedical research is also emphasized. Three empirical applications are presented to illustrate the discussion, and replication materials are provided for researchers to conduct RD analysis using publicly available data and statistical software in Python, R, and Stata.
STATISTICS IN MEDICINE
(2023)
Article
Health Care Sciences & Services
Federico Ricciardi, Silvia Liverani, Gianluca Baio
Summary: The regression discontinuity design is a quasi-experimental design that estimates the causal effect of a treatment when its assignment is defined by a threshold for a continuous variable. Bandwidth selection is an important decision in this design analysis, and the proposed methodology considers units' exchangeability as the main criteria for selecting subjects. The validity of the methodology is demonstrated through simulated experiments and an example on the effect of statins on cholesterol levels.
STATISTICAL METHODS IN MEDICAL RESEARCH
(2023)
Article
Multidisciplinary Sciences
Abhiroop Mukherjee, George Panayotov, Rik Sen, Harsha Dutta, Pulak Ghosh
Summary: Despite the urgent need to track COVID vaccine effectiveness, many countries struggle to calculate standard VE measures from their public health data. In this study, researchers used regression discontinuity design (RDD) to estimate VE based on health records from West Bengal, India. The results showed an VE of 55.2% against symptomatic disease, 80.1% against hospitalizations, and 85.5% against intensive care/critical care/high dependency admissions or deaths. These measures can also be used by other data-deficient countries with age-based eligibility for any vaccine to inform their immunization policies.
Article
Surgery
Megan Lane, Nicholas L. Berlin, Kevin C. Chung, Jennifer F. Waljee
Summary: Understanding causal association and inference is critical to study health risks, treatment effectiveness, and the impact of health care interventions. This review highlights several methodologic options to deduce causality from observational data.
PLASTIC AND RECONSTRUCTIVE SURGERY
(2023)
Article
Multidisciplinary Sciences
Jakob Runge
Summary: Detecting and quantifying causal relations in ecosystem functioning is challenging and involves reasoning about underlying assumptions. A global study on grasslands highlights the importance of considering confounding, nonlinearity, and determinism in modern causal inference approaches in ecology.
NATURE COMMUNICATIONS
(2023)
Article
Statistics & Probability
Juan D. Diaz, Jose R. Zubizarreta
Summary: Regression discontinuity designs are commonly used for causal inference, but are limited to simple settings. We propose a framework for complex discontinuity designs that encompass multiple treatment rules. Covariates play a central role in identification, estimation, and generalization of causal effects. We discuss estimation approaches based on matching and weighting, and find that grade retention in Chile has a negative impact on future grade retention but is not associated with dropping out of school or committing a juvenile crime.
ANNALS OF APPLIED STATISTICS
(2023)
Article
Ecology
Suchinta Arif, M. Aaron MacNeil
Summary: Recent developments in computer science have advanced the use of causal inference and causal diagrams in observational studies, providing a unified approach to variable selection across different methodologies. This paper demonstrates how causal diagrams can be extended to ensure proper study design under quasi-experimental settings and highlights the importance of routinely applying causal diagrams in ecology research.
Article
Statistics & Probability
Maxime Rischard, Zach Branson, Luke Miratrix, Luke Bornn
Summary: This study examines the premium on house price for a particular school district in New York City using a novel implementation of a geographic regression discontinuity design (GeoRDD). By modeling spatial structure with Gaussian processes regression, the research identifies significant price differences along borders, with one border showing a statistically significant 20% higher price for houses on the more desirable side.
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
(2021)
Article
Economics
Matias D. Cattaneo, Rocio Titiunik
Summary: This article provides a curated review of the methodological literature on regression discontinuity design, focusing on the continuity framework and the local randomization framework. It discusses designs and parameters, estimation and inference methods, and validation and falsification approaches.
ANNUAL REVIEW OF ECONOMICS
(2022)
Article
Agricultural Economics & Policy
David Wuepper, Robert Finger
Summary: This article showcases the increasing use of regression discontinuity designs (RDD) in agricultural and environmental economics to identify causal effects. It discusses recent applications, best practices, identifying assumptions and testing methods. The article also covers basic empirical issues, advanced topics such as panel data utilization, modeling heterogeneous treatment effects and extrapolation of local estimates. It further demonstrates how agricultural economists can combine RDD with remote sensing and environmental modeling. The article concludes by highlighting emerging opportunities and providing implications for research and policy.
EUROPEAN REVIEW OF AGRICULTURAL ECONOMICS
(2023)
Article
Environmental Sciences
Ke Li, Weihong Yuan, Jianglong Li
Summary: This study finds that metro transit in China has a positive impact on reducing air pollution, with varying effects during different time periods.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2022)
Article
Economics
Zhuan Pei, David S. Lee, David Card, Andrea Weber
Summary: In this paper, we propose a polynomial order selection procedure based on the asymptotic mean squared error of the local regression RD estimator, which performs well in large sample sizes typically found in empirical RD applications. This procedure can be easily extended to fuzzy regression discontinuity and regression kink designs.
JOURNAL OF BUSINESS & ECONOMIC STATISTICS
(2022)
Article
Economics
Andrii Babii, Rohit Kumar
Summary: This paper examines the estimation and inference of isotonic regression at the boundary point, which is crucial for analyzing monotone regression discontinuity designs. The paper demonstrates that isotonic regression is inconsistent in this context and provides the asymptotic distributions of boundary-corrected estimators. Interestingly, the boundary-corrected estimators can be bootstrapped without subsampling or additional nonparametric smoothing unlike the interior point. Monte Carlo experiments indicate that shape restrictions can significantly enhance the finite-sample performance of unrestricted estimators. Finally, the paper estimates the causal effect of incumbency in U.S. House elections using isotonic regression discontinuity design.
JOURNAL OF ECONOMETRICS
(2023)
Review
Public, Environmental & Occupational Health
Michele Hilton Boon, Peter Craig, Hilary Thomson, Mhairi Campbell, Laurence Moore
Summary: Regression discontinuity designs have been widely applied in health research and could be used more widely still. However, shortcomings in study quality and reporting suggest that the potential benefits of this method have not yet been fully realized.