External Validity in Fuzzy Regression Discontinuity Designs

Bertanha, Marinho; Imbens, Guido W.

doi:10.3386/w20773

Cited by 78 publications

(104 citation statements)

References 23 publications

(37 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One is to estimate our treatment effects separately for sub-samples of the data defined by pre-treatment characteristics (e.g., boys versus girls). Second, using a straightfoward technique recently introduced by Bertanha and Imbens (2014) we examine whether our complier population is distinctive.…”

Section: Validity Of the Rd Designmentioning

confidence: 99%

“…To examine the empirical relevance of this concern, we follow the suggestion recently introduced by Bertanha and Imbens (2014). They recommend examining the continuity of outcomes, separately for children who took up the "treatment"…”

Section: Treatment Heterogeneitymentioning

confidence: 99%

“…However, we also find evidence that these effects are heterogeneous. Using an approach recently introduced by Bertanha and Imbens (2014), we present evidence on the heterogeneity that distinguishes the "compliers" from the "never takers" and "always takers" in our "intent to treat" (ITT)…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

The Gift of Time? School Starting Age and Mental Health

Dee

Sievertsen

2015

View full text Add to dashboard Cite

In many developed countries, children now begin their formal schooling at an older age. However, a growing body of empirical studies provides little evidence that such schooling delays improve educational and economic outcomes. This study presents new evidence on whether school starting age influences student outcomes by relying on linked Danish survey and register data that include several distinct, widely used, and validated measures of mental health that are reported out-of-school among similarly aged children. We estimate the causal effects of delayed school enrollment using a "fuzzy" regression-discontinuity design based on exact dates of birth and the fact that, in Denmark, children typically enroll in school during the calendar year in which they turn six. We find that a one-year delay in the start of school dramatically reduces inattention/hyperactivity at age 7 (effect size = -0.7), a measure of self regulation with strong negative links to student achievement. We also find that this large and targeted effect persists at age 11. However, the estimated effects of school starting age on other mental-health constructs, which have weaker links to subsequent student achievement, are smaller and less persistent.

show abstract

Section: Validity Of the Rd Designmentioning

confidence: 99%

Section: Treatment Heterogeneitymentioning

confidence: 99%

See 1 more Smart Citation

The Gift of Time? School Starting Age and Mental Health

Dee

Sievertsen

2015

View full text Add to dashboard Cite

show abstract

“…Here we discuss some supplementary analyses that can be done to assess the generalizability of the local average treatment effect. This section builds on the discussions in Angrist (2004), Hirano, Imbens, Rubin, and Zhou (2000), Imbens and Rubin (1997ab) and Bertanha and Imbens (2014). The Bertanha and Imbens (2015) discussion is primarily in the context of fuzzy regression discontinuity designs, but their results apply directly to other instrumental variables settings.…”

Section: Generalizing the Local Average Treatment Effectmentioning

confidence: 99%

The Econometrics of Randomized Experiments

Athey

Imbens

2017

Handbook of Economic Field Experiments

479

297

View full text Add to dashboard Cite

In this chapter, we present econometric and statistical methods for analyzing randomized experiments. For basic experiments we stress randomization-based inference as opposed to sampling-based inference. In randomization-based inference, uncertainty in estimates arises naturally from the random assignment of the treatments, rather than from hypothesized sampling from a large population. We show how this perspective relates to regression analyses for randomized experiments. We discuss the analyses of stratified, paired, and clustered randomized experiments, and we stress the general efficiency gains from stratification. We also discuss complications in randomized experiments such as noncompliance. In the presence of non-compliance we contrast intention-to-treat analyses with instrumental variables analyses allowing for general treatment effect heterogeneity. We consider in detail estimation and inference for heterogenous treatment effects in settings with (possibly many) covariates. These methods allow researchers to explore heterogeneity by identifying subpopulations with different treatment effects while maintaining the ability to construct valid confidence intervals. We also discuss optimal assignment to treatment based on covariates in such settings. Finally, we discuss estimation and inference in experiments in settings with interactions between units, both in general network settings and in settings where the population is partitioned into groups with all interactions contained within these groups.JEL Classification: C01, C13, C18, C21, C52, C54

show abstract

“…RDD identifies causal effects local to a cutoff value; several authors develop conditions for extrapolating local effects farther away from the cutoff. These include estimation of derivatives of the treatment effect at the cutoff by Dong (2016) and Dong and Lewbel (2015); tests for homogeneity of treatment effects in fuzzy RDD by Bertanha and Imbens (2018); and estimation of average treatment effects in RDD with variation in cutoff values by Bertanha (2017). All these theoretical contributions rely on point identification and inference, and they are subject to both types of impossibility.…”

Section: Regression Discontinuity and Kink Designsmentioning

confidence: 99%

Impossible inference in econometrics: theory and applications

Bertanha¹,

Moreira²

2019

Self Cite

View full text Add to dashboard Cite

This paper studies models in which hypothesis tests have trivial power, that is, power smaller than size. This testing impossibility, or impossibility type A, arises when any alternative is not distinguishable from the null. We also study settings where it is impossible to have almost surely bounded confidence sets for a parameter of interest. This second type of impossibility (type B) occurs under a condition weaker than the condition for type A impossibility: the parameter of interest must be nearly unidentified. Our theoretical framework connects many existing publications on impossible inference that rely on different notions of topologies to show models are not distinguishable or nearly unidentified. We also derive both types of impossibility using the weak topology induced by convergence in distribution. Impossibility in the weak topology is often easier to prove, it is applicable for many widely-used tests, and it is useful for robust hypothesis testing. We conclude by demonstrating impossible inference in multiple economic applications of models with discontinuity and time-series models.

show abstract

External Validity in Fuzzy Regression Discontinuity Designs

Cited by 78 publications

References 23 publications

The Gift of Time? School Starting Age and Mental Health

The Gift of Time? School Starting Age and Mental Health

The Econometrics of Randomized Experiments

Impossible inference in econometrics: theory and applications

Contact Info

Product

Resources

About