Abstract

In a study of the effect of civil war exposure on local collective action outcomes in Sierra Leone, Bellows and Miguel (2009) employ a coefficient stability approach to assess the importance of omitted variable bias building on Altonji et al. (2005a). Here we clarify the econometric assumptions underlying Bellows and Miguel (2009), and extend their analysis using data on dependent variable reliability ratios and the method developed in Oster (2015). (C) 2015 Elsevier BM. All rights reserved.