The Dangers of Extreme Counterfactuals

By Gary King, Langche Zeng. 2006

573 citations recorded by the source corpus.

Subfields: Economics and Econometrics, Statistics and Probability

How researchers see this paper

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

In Economics and Econometrics

In Statistics and Probability

Citations in the corpus

Synthetic Control Methods for Comparative Case Studies: Estimating the Effect of California’s Tobacco Control Program (2010, methodology): “As a result, the synthetic control method safeguards against estimation of “extreme counterfactuals,” that is, those counterfactuals that fall far outside the convex hull of the data (King and Zheng 2006).”
Generalized Optimal Matching Methods for Causal Inference (2016, background): “Moreover, whereas linear regression is subject to dangerous extrapolation [52, 4], matching with weights in the simplex (corresponding to convex combinations), including subsets, inherently prohibits extrapolation.”
Asymptotic inference of causal effects with observational studies trimmed by the estimated propensity scores (2018, background): “Unfortunately, it is common to have limited overlap in covariates between the treatment and control groups, which affects the credibility of all methods attempting to estimate causal effects for the population (King & Zeng, 2005; Imbens, 2015).”
Causal Inference without Balance Checking: Coarsened Exact Matching (2012, background): “Section 2.6 discusses what to do when the analyst chooses to prune treated units when no reasonable match exists among the pool of available controls.”
Improving the Interpretation of Fixed Effects Regression Results (2018, background): “…being unlikely to ever occur in the real world, such “ extreme counterfactuals, ” which extrapolate to regions of sparse (or nonexistent) data rest on strong modeling assumptions, and can produce severely biased estimates of a variable ’ s effect if those assumptions fail (King and Zeng 2006).”
An Introduction to the Augmented Inverse Propensity Weighted Estimator (2010, background): “Advance Access publication December 14, 2009 Political Analysis (2010) 18:36–56 doi:10.1093/pan/mpp036 An Introduction to the Augmented Inverse Propensity”
MatchIt: Nonparametric Preprocessing for Parametric Causal Inference (2011, methodology): “See King and Zeng (2006, 2007) for information about the convex hull in this context and as a measure of model dependence.”
Matching as Nonparametric Preprocessing for Reducing Model Dependence in Parametric Causal Inference (2007, methodology): “King and Zeng (2006) offer methods to detect the problem of model dependence; we show how to ameliorate it, when possible.”
Cem: Coarsened Exact Matching in Stata (2009, methodology): “…type of inﬁnite regress, researchers use entirely diﬀerent technologies for the ﬁrst step, such as kernel density estimation (Heckman, Ichimura and Todd, 1997) or dropping control units outside the hyper-rectangle (Iacus and Porro, 2009) or convex hull (King and Zeng, 2006) of the treated units.”
How Conditioning on Posttreatment Variables Can Ruin Your Experiment and What to Do about It (2018, background): “After all, published research in political science identified post-treatment bias (in passing) as problematic a decade ago (King and Zeng 2006, 147–148).”
A Primer on Inverse Probability of Treatment Weighting and Marginal Structural Models (2016, methodology): “In other words, there are no treated participants who are similar or comparable to any of the untreated participants on the totality of the observed covariates, and any comparison between the two groups might be due to extrapolation (King & Zeng, 2006).”
How Much Should We Trust Estimates from Multiplicative Interaction Models? Simple Tools to Improve Empirical Practice (2016, unknown): “If either of these two conditions fails, the conditional marginal-effect estimates are based on extrapolation or interpolation of the functional form to an area where there is no or only sparse data and therefore the effect estimates are fragile and model dependent (King and Zeng 2006).”
Comparative Politics and the Synthetic Control Method (2014, unknown): “We illustrate the main ideas behind the synthetic control method by estimating the economic impact of the 1990 German reunification on West Germany.”