Unobservable Selection and Coefficient Stability: Theory and Validation

By Emily Oster. 2013

290 citations recorded by the source corpus.

Subfields: Control and Systems Engineering

View Paper: https://www.semanticscholar.org/paper/d0f95fa4ba2f5b01d7e107b20a038ba44199c010

How researchers see this paper

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

In Control and Systems Engineering

Citations in the corpus

Internet and Politics: Evidence from U.K. Local Elections and Local Government Policies (2015, background): “The first one follows the insights of Altonji, Elder, and Taber (2005), as recently extended by Oster (2013), and constructs bounds on the effects of internet diffusion on voter turnout based on the correlation between observable controls and internet diffusion.”
Law Enforcement, Municipal Budgets and Spillover Effects: Evidence from a Quasi-Experiment in Italy (2017, methodology): “All calculations are made using the PSACALC stata module by Oster (2013).”
Pre-Event Trends in the Panel Event-Study Design (2019, unknown): “5See also Altonji et al. (2005) and Oster (2016), who propose an alternative way to use observed covariates to allow for unobservable confounds.”
Mining and Local Corruption in Africa (2017, background): “Hence, bias stemming from potential unobserved confounders must be extremely large relative to the estimated e↵ects for our results to stem from confounding (see, e.g., Oster, 2013).”
The long Term Effects of the Printing Press in Sub Saharan Africa (2016, methodology): “M is a set of observed controls that do not have a related unobserved component and are orthogonal to W and W ′(Oster, 2013).”
Parks versus PES: Evaluating direct and incentive-based land conservation in Mexico (2017, unknown): “We apply the rule of thumb suggested by Oster (2013). news for policymakers facing a choice between PAs and PES to achieve REDD, as they indicate that both can achieve conservation while not impoverishing local communities.”
Causal Inference in Urban and Regional Economics (2014, background): “Oster (2013) reformulates this assumption as the following proportional selection relationship: ν CovðT ,XδÞ VarðXδÞ ¼ CovðT ,WρÞ VarðWρÞ : That is, the correlation between observables and the treatment is proportional to the correlation between the unobservables and the treatment.”
US political corruption and firm financial policies (2016, methodology): “To assess this concern, I conduct a formal analysis of coefficient movements following Oster (2015). I find that selection on unobservables must be 2.”