Mission accomplished: A reply to Reuveny and Keshk
Cullen F. Goenner
Abstract
Cullen F. Goenner
Abstract
Reuveny and Keshk (“Reconsidering trade and conflict simultaneity: The risk of emphasizing technique over substance,” this issue, 2013) argue that the econometric techniques used by Goenner ( Conflict Management and Peace Science 28(5): 459–477, 2011) to test and control for endogeneity when estimating the relationship between trade and conflict lack substance. Both sets of authors propose the use of instrumental variable methods, which are known by econometricians to be the natural remedy for estimation with potentially endogenous regressors. Where Goenner (2011) and Reuveny and Keshk (2013) agree is that theory should guide variable selection and the model’s specification. Yet they differ in that, while econometric tests cannot replace theory, one should not trust the appropriateness of the model’s specification based on theory alone – one should also verify. Otherwise, as Goenner (2011) notes, attempts to control for endogeneity may fail.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Reuveny and Keshk (“Reconsidering trade and conflict simultaneity: The risk of emphasizing technique over substance,” this issue, 2013) argue that the econometric techniques used by Goenner ( Conflict Management and Peace Science 28(5): 459–477, 2011) to test and control for endogeneity when estimating the relationship between trade and conflict lack substance. Both sets of authors propose the use of instrumental variable methods, which are known by econometricians to be the natural remedy for estimation with potentially endogenous regressors. Where Goenner (2011) and Reuveny and Keshk (2013) agree is that theory should guide variable selection and the model’s specification. Yet they differ in that, while econometric tests cannot replace theory, one should not trust the appropriateness of the model’s specification based on theory alone – one should also verify. Otherwise, as Goenner (2011) notes, attempts to control for endogeneity may fail.
Key concepts: Endogeneity, Simultaneity, Instrumental variable, Economics, Econometrics, Control (management), Specification, Estimation