Econometrics: Is this an instance of simultaneity?

Econometrics: Is this an instance of simultaneity?




I'm looking at a dataset, and here's the situation:



a rise in x --> a rise in y

a rise in y --> a fall in x



I want to regress y on x. Is this a situation in which x is correlated with the error term? If so, do I need to introduce an instrumental variable?



Thank you!



(PS Don't worry; it's not homework. It's a lonnng-term project, and I can't talk with my advisor until next week. Thank you!!)


Update 1: Anjaree - thank you so much!!



It is panel data, which I think is time-series data... would you mind explaining again what effect that has?



Thank you!!





No Answers Posted Yet.