Robust Confidence Intervals for Average Treatment Effects Under Limited Overlap

Rothe, Christoph

doi:10.2139/ssrn.2554886

Cited by 11 publications

(33 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This idea could also be used in other settings where the finite sample accuracy of inference faces a similar bias-vs-normality trade-off, such as inference on average treatment effects under unconfoundedness with limited overlap (e.g. Rothe, 2017).…”

Section: Extensions and Remarksmentioning

confidence: 99%

Bias-Aware Inference in Fuzzy Regression Discontinuity Designs

Noack,

Rothe

2019

Preprint

Self Cite

View full text Add to dashboard Cite

Fuzzy regression discontinuity (FRD) designs occur frequently in many areas of applied economics. We argue that the confidence intervals based on nonparametric local linear regression that are commonly reported in empirical FRD studies can have poor finite sample coverage properties for reasons related to their general construction based on the delta method, and to how they account for smoothing bias. We therefore propose new confidence sets, which are based on an Anderson-Rubin-type construction.These confidence sets are bias-aware, in the sense that they explicitly take into account the exact smoothing bias of the local linear estimators on which they are based. They are simple to compute, highly efficient, have excellent coverage properties in finite samples. They are also valid under weak identification (that is, if the jump in treatment probabilities at the threshold is small) and irrespective of whether the distribution of the running variable is continuous, discrete, or of some intermediate form.This Version: June 12, 2019. We thank Tim Armstrong and Michal Kolesár and numerous seminar participants for helpful comments and suggestions. Financial support of the ERC through grant SH1-77202 is gratefully acknowledged.

show abstract

Section: Extensions and Remarksmentioning

confidence: 99%

Bias-Aware Inference in Fuzzy Regression Discontinuity Designs

Noack,

Rothe

2019

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Small cells are prevalent in specifications that include many two-way interactions and are saturated in at least a subset of controls. 5 When the controls are discrete, limited overlap -a major concern in research designs relying on unconfoundedness-type identification assumptions -can be viewed as a small cell problem (e.g., Rothe, 2017). Moreover, categorical controls, when incorporated through a set of indicator variables, give rise to small cells if some of the categories are sparsely populated.…”

Section: Introductionmentioning

confidence: 99%

Omitted Variable Bias of Lasso-Based Inference Methods Under Limited Variability: A Finite Sample Analysis

Wüthrich

Zhu

2019

SSRN Journal

View full text Add to dashboard Cite

“…Several statistical techniques have been proposed for robustly estimating the causal effect with respect to a single response variable-see, for example, Ding et al (2016), Rothe (2017), and the references therein. However, to the best of our knowledge, robust causal inference for a nonlinear function of multiple causal effects such as the iROAS defined by (1.2) has received little attention in the literature so far.…”

Section: Introductionmentioning

confidence: 99%

Robust Causal Inference for Incremental Return on Ad Spend with Randomized Paired Geo Experiments

Chen¹,

Au²

2019

Preprint

View full text Add to dashboard Cite

Evaluating the incremental return on ad spend (iROAS) of a prospective online marketing strategy-that is, the ratio of the strategy's causal effect on some response metric of interest relative to its causal effect on the ad spend-has become progressively more important as advertisers increasingly seek to better understand the impact of their marketing decisions. Although randomized "geo experiments" are frequently employed for this evaluation, obtaining reliable estimates of the iROAS can be challenging as oftentimes only a small number of highly heterogeneous units are used. In this paper, we formulate a novel causal framework for inferring the iROAS of online advertising in a randomized geo experiment design, and we develop a robust model-free estimator "Trimmed Match" which adaptively trims poorly matched pairs. Using simulations and case studies, we show that Trimmed Match can be more efficient than some alternatives, and we investigate the sensitivity of the estimator to some violations of its assumptions. Consistency and asymptotic normality are also established for a fixed trim rate.

show abstract

Robust Confidence Intervals for Average Treatment Effects Under Limited Overlap

Cited by 11 publications

References 27 publications

Bias-Aware Inference in Fuzzy Regression Discontinuity Designs

Bias-Aware Inference in Fuzzy Regression Discontinuity Designs

Omitted Variable Bias of Lasso-Based Inference Methods Under Limited Variability: A Finite Sample Analysis

Robust Causal Inference for Incremental Return on Ad Spend with Randomized Paired Geo Experiments

Contact Info

Product

Resources

About