Sociological Methodology · 2004 · 744 citations · 29 references
Public HealthBiostatisticsCausal InferenceEconometricsExperimental DesignCausal Effect EstimationInstrumental VariablesMatching MethodsPropensity Score MethodsSelection BiasTreatment EffectsCausal EstimationCausal RelationshipsConfoundingInstrumental Variables EstimationMatchingMatching EstimatorsUnobserved ConfoundingUnobserved Heterogeneity
Propensity score matching estimates causal effects largely free of bias from observed covariates, but is vulnerable to hidden bias from unobserved confounders, while instrumental variable estimation offers an alternative that reduces precision and relies on untestable assumptions; together, these methods provide complementary insights into causal relationships. The study proposes using Rosenbaum bounds to quantify the strength of hidden confounding needed to invalidate matching results and introduces a method to assess the additional uncertainty in instrumental variable estimates so that the total uncertainty can be compared with that from Rosenbaum bounds. The authors employ Rosenbaum bounds to evaluate hidden bias in matching estimators and develop a procedure to quantify the uncertainty inherent in instrumental variable estimation, enabling a direct comparison of the two uncertainty assessments. Illustrating the approach, the authors analyze the impact of unemployment insurance on reemployment timing, post‑unemployment wages, and relocation probability using Survey of Income and Program Participation data.
Propensity score matching provides an estimate of the effect of a “treatment” variable on an outcome variable that is largely free of bias arising from an association between treatment status and observable variables. However, matching methods are not robust against “hidden bias” arising from unobserved variables that simultaneously affect assignment to treatment and the outcome variable. One strategy for addressing this problem is the Rosenbaum bounds approach, which allows the analyst to determine how strongly an unmeasured confounding variable must affect selection into treatment in order to undermine the conclusions about causal effects from a matching analysis. Instrumental variables (IV) estimation provides an alternative strategy for the estimation of causal effects, but the method typically reduces the precision of the estimate and has an additional source of uncertainty that derives from the untestable nature of the assumptions of the IV approach. A method of assessing this additional uncertainty is proposed so that the total uncertainty of the IV approach can be comparedwith the Rosenbaum bounds approach to uncertainty using matching methods. Because the approaches rely on different information and different assumptions, they provide complementary information about causal relationships. The approach is illustrated via an analysis of the impact of unemployment insurance on the timing of reemployment, the postunemployment wage, and the probability of relocation, using data from several panels of the Survey of Income and Program Participation (SIPP).
29
Journal of the American Statistical Association
Anne B. Koehler · International Journal of Forecasting · 1995 · 5.4K citations
Statistics and Causal Inference
Paul W. Holland · Journal of the American Statistical Association · 1986 · 5.1K citations