Generalized Synthetic Control Method: Causal Inference with Interactive Fixed Effects Models

Xu, Yiqing

doi:10.1017/pan.2016.2

Cited by 667 publications

(704 citation statements)

References 54 publications

Supporting

Mentioning

678

Contrasting

Unclassified

Order By: Relevance

“…In contrast, for Connecticut, there are negative effects for synthetic control and constrained Lasso and no significant effect for the factor model, while for other states such as Montana, EDR laws do not significantly impact turnout in any period. These findings are broadly consistent with the results in Xu (2017). As for the overall no-effects null hypothesis, the choice of the model for the counterfactual matters for the confidence intervals.…”

Section: Empirical Illustrationsupporting

confidence: 88%

“…To illustrate the practical usefulness of our methods, we revisit the analysis of the effect of Election Day Registration (EDR) laws on voter turnout in the United States by Xu (2017).…”

Section: (Ii) Approximate Validity Under Weak Assumptionsmentioning

confidence: 99%

“…Carvalho et al (2017) derive the asymptotic distribution ofᾱ in setups where the counterfactual is estimated based on Lasso and Li (2017) studies inference based on the constrained least squares estimator of Abadie et al (2010Abadie et al ( , 2015. Xu (2017) proposes an asymptotic bootstrap inference procedure based on factor models, but leaves the formal justification of this procedure for future research. By contrast, our approach will instead be based on permutation distributions, and will be shown to be formally valid exactly under strong exchangeability assumptions and approximately valid under stationarity and weak dependence of {u t } and very weak conditions on the estimator for P N t .…”

Section: Related Literaturementioning

confidence: 99%

“…In this section, we apply our inference procedure to analyze the effect of election day registration (EDR) laws on voter turnout in the United States as in Xu (2017). Voting in the United States is typically a two-step procedure since eligible voters must register prior to casting their ballots.…”

Section: Empirical Illustrationmentioning

confidence: 99%

“…Such setups frequently arise in applied economic research and there are various different approaches to estimate the policy effects of interest. A non-exhaustive list of methods includes difference-in-differences methods (e.g., Ashenfelter and Card, 1985;Card and Krueger, 1994;Bertrand et al, 2004;Athey and Imbens, 2006;Angrist and Pischke, 2008), synthetic control models (e.g., Abadie and Gardeazabal, 2003;Abadie et al, 2010Abadie et al, , 2015Li, 2017), penalized regression models for synthetic controls (e.g., Valero, 2015;Doudchenko and Imbens, 2016), factor, matrix completion and interactive fixed effects models for panel data (e.g., Bai, 2003;Pesaran, 2006;Bai, 2009;Kim and Oka, 2014;Gobillon and Magnac, 2016;Xu, 2017;Athey et al, 2017;Amjad et al, 2017), matching methods (e.g., Heckman et al, 1997Heckman et al, , 1998Dehejia and Wahba, 2002), as well as standard time series models. Doudchenko and Imbens (2016) and Gobillon and Magnac (2016) provide comparative overviews.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

An exact and robust conformal inference method for counterfactual and synthetic controls

Chernozhukov

Wüthrich

Zhu

2017

View full text Add to dashboard Cite

This paper introduces new inference methods for counterfactual and synthetic control methods for evaluating policy effects. Our inference methods work in conjunction with many modern and classical methods for estimating the counterfactual mean outcome in the absence of a policy intervention. Specifically, our methods work together with the difference-in-difference, canonical synthetic control, constrained and penalized regression methods for synthetic control, factor/matrix completion models for panel data, interactive fixed effects panel models, time series models, as well as fused time series panel data models. The proposed method has a double justification. (i) If the residuals from estimating the counterfactuals are exchangeable as implied, for example, by i.i.d. data, our procedure achieves exact finite sample size control without any assumption on the specific approach used to estimate the counterfactuals. (ii) If the data exhibit dynamics and serial dependence, our inference procedure achieves approximate uniform size control under weak and easy-to-verify conditions on the method used to estimate the counterfactual. We verify these condition for representative methods from each group listed above. Simulation experiments demonstrate the usefulness of our approach in finite samples. We apply our method to re-evaluate the causal effect of election day registration (EDR) laws on voter turnout in the United States.

show abstract

Section: Empirical Illustrationsupporting

confidence: 88%

“…To illustrate the practical usefulness of our methods, we revisit the analysis of the effect of Election Day Registration (EDR) laws on voter turnout in the United States by Xu (2017).…”

Section: (Ii) Approximate Validity Under Weak Assumptionsmentioning

confidence: 99%

Section: Related Literaturementioning

confidence: 99%

Section: Empirical Illustrationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

An exact and robust conformal inference method for counterfactual and synthetic controls

Chernozhukov

Wüthrich

Zhu

2017

View full text Add to dashboard Cite

show abstract

Removal of Medicaid Prior Authorization Requirements and Buprenorphine Treatment for Opioid Use Disorder

Christine,

Larochelle,

Lin

et al. 2023

JAMA Health Forum

View full text Add to dashboard Cite

ImportanceBuprenorphine treatment for opioid use disorder (OUD) is associated with decreased morbidity and mortality. Despite its effectiveness, buprenorphine uptake has been limited relative to the burden of OUD. Prior authorization (PA) policies may present a barrier to treatment, though research is limited, particularly in Medicaid populations.ObjectiveTo assess whether removal of Medicaid PAs for buprenorphine to treat OUD is associated with changes in buprenorphine prescriptions for Medicaid enrollees.Design, Setting, and ParticipantsThis state-level, serial cross-sectional study used quarterly data from 2015 through the first quarter (January-March) of 2019 to compare buprenorphine prescriptions in states that did and did not remove Medicaid PAs. Analyses were conducted between June 10, 2021, and August 15, 2023. The study included 23 states with active Medicaid PAs for buprenorphine in 2015 that required similar PA policies in fee-for-service and managed care plans and had at least 2 quarters of pre- and postperiod buprenorphine prescribing data.ExposuresRemoval of Medicaid PA for at least 1 formulation of buprenorphine for OUD.Main Outcomes and MeasuresThe main outcome was number of quarterly buprenorphine prescriptions per 1000 Medicaid enrollees.ResultsBetween 2015 and the first quarter of 2019, 6 states in the sample removed Medicaid PAs for at least 1 formulation of buprenorphine and had at least 2 quarters of pre- and postpolicy change data. Seventeen states maintained buprenorphine PAs throughout the study period. At baseline, relative to states that repealed PAs, states that maintained PAs had lower buprenorphine prescribing per 1000 Medicaid enrollees (median, 6.6 [IQR, 2.6-13.9] vs 24.1 [IQR, 8.7-27.5] prescriptions) and lower Medicaid managed care penetration (median, 38.5% [IQR, 0.0%-74.1%] vs 79.5% [IQR, 78.1%-83.5%] of enrollees) but similar opioid overdose rates and X-waivered buprenorphine clinicians per 100 000 population. In fully adjusted difference-in-differences models, removal of Medicaid PAs for buprenorphine was not associated with buprenorphine prescribing (1.4% decrease; 95% CI, −31.2% to 41.4%). For states with below-median baseline buprenorphine prescribing, PA removal was associated with increased buprenorphine prescriptions per 1000 Medicaid enrollees (40.1%; 95% CI, 0.6% to 95.1%), while states with above-median prescribing showed no change (−20.7%; 95% CI, −41.0% to 6.6%).Conclusions and RelevanceIn this serial cross-sectional study of Medicaid PA policies for buprenorphine for OUD, removal of PAs was not associated with overall changes in buprenorphine prescribing among Medicaid enrollees. Given the ongoing burden of opioid overdoses, continued multipronged efforts are needed to remove barriers to buprenorphine care and increase availability of this lifesaving treatment.

show abstract

Evaluation of US Hospital Episode Spending for Acute Inpatient Conditions After the Patient Protection and Affordable Care Act

et al. 2020

View full text Add to dashboard Cite

IMPORTANCE Under the Patient Protection and Affordable Care Act (ACA), US hospitals were exposed to a number of reforms intended to reduce spending, many of which, beginning in 2012, targeted acute care hospitals and often focused on specific diagnoses (eg, acute myocardial infarction, heart failure, and pneumonia) for Medicare patients. Other provisions enacted in the ACA and under budget sequestration (beginning in 2013) mandated Medicare fee cuts. OBJECTIVE To evaluate the association between the enactment of ACA reforms and 30-day pricestandardized hospital episode spending. DESIGN, SETTING, AND PARTICIPANTS This policy evaluation included index discharges between

show abstract

Generalized Synthetic Control Method: Causal Inference with Interactive Fixed Effects Models

Cited by 667 publications

References 54 publications

An exact and robust conformal inference method for counterfactual and synthetic controls

An exact and robust conformal inference method for counterfactual and synthetic controls

Removal of Medicaid Prior Authorization Requirements and Buprenorphine Treatment for Opioid Use Disorder

Evaluation of US Hospital Episode Spending for Acute Inpatient Conditions After the Patient Protection and Affordable Care Act

Contact Info

Product

Resources

About