I have a panel data which had autocorrelation problem. I used cochrane orcutt method to remove autocorrelation problem with this data. If so, does this also mean that I removed the causality problem as well?


The Cochrane-Orcutt FGLS estimator makes a number of strong assumptions, one of which is that the regressor is strictly exogenous ($x_t$ is uncorrelated with past, present and future values of the disturbance term). This means that you have not remedied the causality problem by using it, and that it relies on the absence of endogeneity to work.

| cite | improve this answer | |
  • $\begingroup$ Thank you so much for your kindly answer. Then I figured out I needed to use the instrumental variable test (IV) as one of the way to deal with causality problem. My data is panel but I run pooled panel regression since it fits better for my situation. However, I get the same answer whether I use the pooled panel regression or instrumental variable test over my data after I apply Cochrane Orcutt method on either way. If I get the same answers in either way, can this be a problem? If so, is there a way to solve this? Or is it ok to leave it as it is? $\endgroup$ – Eric Nov 23 '17 at 7:54
  • $\begingroup$ @Eric Panel and IV are both methods to deal with endogeneity, so if they give the same answer, that is reassuring. $\endgroup$ – Dimitriy V. Masterov Nov 23 '17 at 16:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.