A Note on Endogenous Control Variables in Evaluation Studies
AbstractThe paper addresses the issue of potentially endogenous control variables in evaluation studies based on the assumption that selection bias can be avoided by conditioning on observed variables. It shows that the standard formulation of the conditional independence (CIA) or selection on observables assumption obscures the endogeneity problem. The paper clarifies the issue by suggesting a CIA based on exogenous (potential) outcomes as well as explicit exogeneity conditions. The refined CIA allows to asses the endogeneity bias as well as the plausibility of the CIA, because it allows a separate discussion of (i) the problems of missing variables that are necessary to avoid selection bias, and (ii) the problem that some of the control variables may be influenced by the treatment.
Download InfoIf you experience problems downloading a file, check if you have the proper application to view it first. In case of further problems read the IDEAS help page. Note that these files are not on the IDEAS site. Please be patient as the files may be large.
Bibliographic InfoPaper provided by Department of Economics, University of St. Gallen in its series University of St. Gallen Department of Economics working paper series 2005 with number 2005-16.
Length: 8 pages
Date of creation: Aug 2005
Date of revision:
Find related papers by JEL classification:
- C21 - Mathematical and Quantitative Methods - - Single Equation Models; Single Variables - - - Cross-Sectional Models; Spatial Models; Treatment Effect Models
- C31 - Mathematical and Quantitative Methods - - Multiple or Simultaneous Equation Models; Multiple Variables - - - Cross-Sectional Models; Spatial Models; Treatment Effect Models; Quantile Regressions; Social Interaction Models
This paper has been announced in the following NEP Reports:
- NEP-ALL-2005-11-09 (All new papers)
You can help add them by filling out this form.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Joerg Baumberger).
If references are entirely missing, you can add them using this form.