On the Specification of Propensity Scores, With Applications to the Analysis of Trade Policies

Millimet, Daniel L.; Tchernis, Rusty

doi:10.1198/jbes.2009.06045

Cited by 87 publications

(69 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The topic has been addressed in the literature with focus on both variable selection and functional form specification. [4][5][6][7][8][9][10] Through simulations, Millimet and Tchernis [8] study the case of overfitting in terms of including both irrelevant variables and higher order terms as well as underfitting in terms of excluding relevant covariates and higher order terms. Their conclusions were in general positive towards overfitting and they found that the cost in terms of efficiency was small in the simulations under study.…”

Section: Introductionmentioning

confidence: 99%

Effects of correlated covariates on the asymptotic efficiency of matching and inverse probability weighting estimators for causal inference

Pingel

Waernbaum

2014

Statistics

View full text Add to dashboard Cite

In observational studies, the overall aim when fitting a model for the propensity score is to reduce bias for an estimator of the causal effect. To make the assumption of an unconfounded treatment plausible researchers might include many, possibly correlated, covariates in the propensity score model. In this paper, we study how the asymptotic efficiency of matching and inverse probability weighting estimators for average causal effects change when the covariates are correlated. We investigate the case with multivariate normal covariates, a logistic model for the propensity score and linear models for the potential outcomes and show results under different model assumptions. We show that the correlation can both increase and decrease the large sample variances of the estimators, and that the correlation affects the asymptotic efficiency of the estimators differently, both with regard to direction and magnitude. Moreover, the strength of the confounding towards the outcome and the treatment plays an important role.

show abstract

Section: Introductionmentioning

confidence: 99%

Effects of correlated covariates on the asymptotic efficiency of matching and inverse probability weighting estimators for causal inference

Pingel

Waernbaum

2014

Statistics

View full text Add to dashboard Cite

show abstract

“…Zhao (2008) studies sensitivity of propensity score methods to their specifications though Monte Carlo experiments and finds that, under exogeneity, treatment effect on the treated are not sensitive to the specifications. Millimet and Tchernis (2009) find that over-specifying the propensity score estimator does not impart much of a penalty in terms of inconsistency and inefficiency.…”

Section: Introductionmentioning

confidence: 85%

Estimating treatment effects on healthcare costs under exogeneity: is there a ‘magic bullet’?

Basu

Polsky

Manning

2011

Health Serv Outcomes Res Method

View full text Add to dashboard Cite

Methods for estimating average treatment effects, under the assumption of no unmeasured confounders, include regression models; propensity score adjustments using stratification, weighting, or matching; and doubly robust estimators (a combination of both). Researchers continue to debate about the best estimator for outcomes such as health care cost data, as they are usually characterized by an asymmetric distribution and heterogeneous treatment effects,. Challenges in finding the right specifications for regression models are well documented in the literature. Propensity score estimators are proposed as alternatives to overcoming these challenges. Using simulations, we find that in moderate size samples (n= 5000), balancing on propensity scores that are estimated from saturated specifications can balance the covariate means across treatment arms but fails to balance higher-order moments and covariances amongst covariates. Therefore, unlike regression model, even if a formal model for outcomes is not required, propensity score estimators can be inefficient at best and biased at worst for health care cost data. Our simulation study, designed to take a ‘proof by contradiction’ approach, proves that no one estimator can be considered the best under all data generating processes for outcomes such as costs. The inverse-propensity weighted estimator is most likely to be unbiased under alternate data generating processes but is prone to bias under misspecification of the propensity score model and is inefficient compared to an unbiased regression estimator. Our results show that there are no ‘magic bullets’ when it comes to estimating treatment effects in health care costs. Care should be taken before naively applying any one estimator to estimate average treatment effects in these data. We illustrate the performance of alternative methods in a cost dataset on breast cancer treatment.

show abstract

“…Specification (1) is under‐specified ; Specifications (3) and (4) are over‐specified . While the under‐specified case is arguably the most important in practice, we also examine the over‐specified cases since Millimet and Tchernis () find that over‐specifying the propensity score model when the CIA holds is warranted. We explore whether this conclusion extends to the current situation.…”

Section: Resultsmentioning

confidence: 99%

“…Millimet and Tchernis () and Busso et al . () provide evidence of the superiority of the normalized estimator in practical settings.…”

Section: Estimationmentioning

confidence: 98%

Estimation of Treatment Effects Without an Exclusion Restriction: With an Application to the Analysis of the School Breakfast Program

Millimet

Tchernis

2012

J of Applied Econometrics

Self Cite

View full text Add to dashboard Cite

SUMMARY The increase in childhood obesity has garnered the attention of many in policymaking circles. Consequently, school nutrition programs such as the School Breakfast Program (SBP) have come under scrutiny. The identification of the causal effects of such programs, however, is difficult owing to non‐random selection into the program and the lack of exclusion restrictions. Here, we propose two new estimators aimed at addressing this situation. We compare our new estimators to existing approaches using simulated data. We show that while correlations might suggest that SBP causes childhood obesity, SBP is likely to reduce childhood obesity once selection is addressed. Copyright © 2012 John Wiley & Sons, Ltd.

show abstract

On the Specification of Propensity Scores, With Applications to the Analysis of Trade Policies

Cited by 87 publications

References 33 publications

Effects of correlated covariates on the asymptotic efficiency of matching and inverse probability weighting estimators for causal inference

Effects of correlated covariates on the asymptotic efficiency of matching and inverse probability weighting estimators for causal inference

Estimating treatment effects on healthcare costs under exogeneity: is there a ‘magic bullet’?

Estimation of Treatment Effects Without an Exclusion Restriction: With an Application to the Analysis of the School Breakfast Program

Contact Info

Product

Resources

About