Inference in Differences-in-Differences with Few Treated Groups and Heteroskedasticity

By Bruno Ferman, Cristine Pinto. Review of Economics and Statistics, 2019

150 citations recorded by the source corpus.

Abstract We derive an inference method that works in differences-in-differences settings with few treated and many control groups in the presence of heteroskedasticity. As a leading example, we provide theoretical justification and empirical evidence that heteroskedasticity generated by variation in group sizes can invalidate existing inference methods, even in data sets with a large number of observations per group. In contrast, our inference method remains valid in this case. Our test can also be combined with feasible generalized least squares, providing a safeguard against misspecification of the serial correlation.

Subfields: Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/e11c3e495a1c2ee0624b4a14256ee3bf074494ef

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

Citations in the corpus

Placebo Tests for Synthetic Controls (2017, background): “For example, this condition is satisfied if we have different groups of units where time trends are different across groups but parallel within groups, as considered in Ferman et al. (2016) and Ferman and Pinto (2016b).”
Face masks considerably reduce COVID-19 cases in Germany (2020, unknown): “21).”
Doubly robust difference-in-differences estimators (2018, unknown): “…between years of schooling and real earnings in 1974; and (iii) an “augmented DW” specification, which adds to the “DW” specification the interactions between married and real earnings in 1974, and between married and zero earnings in 1974 - these two interaction terms were used in Firpo (2007).”
What’s trending in difference-in-differences? A synthesis of the recent econometrics literature (2022, unknown): “It is often assumed that ϵ ijt are iid mean-zero across i and j (and sometimes t ); see, e.g., Donald and Lang (2007), Conley and Taber (2011), and Ferman and Pinto (2019).”
ArCo: An artificial counterfactual approach for high-dimensional panel time-series data (2018, background): “4 The first difference can be attenuated in light of the recent results of Conley and Taber (2011) and Ferman and Pinto (2015), who proposed inferential procedures when the number of treated groups is small.”
Suffering in silence: How COVID-19 school closures inhibit the reporting of child maltreatment (2020, unknown): “…only for two months across all counties in the sample, and models with few treated units can lead to improper inference (Cameron et al., 2008; Ferman and Pinto, 2019; MacKinnon and Webb, 2017, 2018).10 To address these concerns, we follow Chetty et al. (2009) and Buchmueller et al. (2011)…”