Basic question on Time-series–cross-section
Posted: Wed Oct 16, 2013 12:05 pm
Hello!
I believe mine is a relatively basic question: my sample consists of time-series-cross-sectional data, i.e., tax revenue data from 300+ municipalities, for two different years.
From a simple graphical analysis, and also by performing the White's test to the data of each of those two years, I know there is (strong) heteroskedasticity which I must account for, and certainly, given the nature of the time-serie, there must be some autocorrelation.
My question is how to proceed in the estimation. I expected to use the Newey-West estimator, which deals both with heteroskedascity and autocorrelation, but such option does not appear. I'm a user of Eviews 7. What should I do? :(
Thank you in advance for your time.
I believe mine is a relatively basic question: my sample consists of time-series-cross-sectional data, i.e., tax revenue data from 300+ municipalities, for two different years.
From a simple graphical analysis, and also by performing the White's test to the data of each of those two years, I know there is (strong) heteroskedasticity which I must account for, and certainly, given the nature of the time-serie, there must be some autocorrelation.
My question is how to proceed in the estimation. I expected to use the Newey-West estimator, which deals both with heteroskedascity and autocorrelation, but such option does not appear. I'm a user of Eviews 7. What should I do? :(
Thank you in advance for your time.