I am an undergraduate student working on environmental economics research. The econometric specification I am using is the following:

ln(Per.Capita.Dollars) = ln(Per.Capita.Dollars.lag) + Average.Annual.Temp + Average.Annual.Prcp + Average.Annual.Temp.lag + Average.Annual.Prcp.lag + factor(Year) + factor(Unique.County)


Temp = Temperature; Prcp = Precipitation; Unique.County = Unique US county

.lag is indicating a new 1 year lagged variable.

If I am aggregating my weather variables annually does it still make sense to have a Year fixed effect (I know it does when using a smaller time step)? Similarly I will be aggregating over counties to a national scale (looking at just the US), will it make sense to have some sort of national fixed effect?

  • $\begingroup$ Hi Dan, welcome to Economics.SE. To me those are many questions at once. This is not the best suited for the format. It's better to focus on one single answerable question. And possibly asking follow-up (separated) questions later. I recommend to read the tour to be more familiar with the format. $\endgroup$ Jul 8 '16 at 5:56
  • $\begingroup$ @bilbo_pingouin There is one question about the annual fixed effect and one, which may well be answerable with reference to the same principles, about the national fixed effect. The sentence beginning "Any ideas ..." is just a summary of the above. So overall the question looks fine to me. $\endgroup$ Jul 8 '16 at 8:43
  • $\begingroup$ Are you asking if you should account for individual fixed effects or are you asking if you should use clustered SE? $\endgroup$
    – 123
    Jul 15 '16 at 3:04

Your Answer

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

Browse other questions tagged or ask your own question.