EViews.com

I have data that is unbalanced panel data, ie, the number of obs in the cross-section varies from period to period. What is the effect of choosing from the various balancing options? Is there any reason (in terms of consistency or unbiaedness of parameter estimates) why I would want to insert missing obs? In other words, why would I want to do this? If I leave it was unbalanced, won't EViews just do the estimation at each period with whatever number of cross-section obs that is available? I am wondering if it is common to balance an unbalanced panel since this is the default option in the dialog box. Sorry if its a stupid question - I don't normally work with panel data. BTW, I am planning to run OLS with fixed effects. Thank you.

Hey,

I ran into the same question the other day, i'm also doing fixed effect OLS regression with an unbalanced panel set. anyway i just tried out both options and running regressions they appeared to be the same for both options since eviews just used the available obserations for each year, which. since the dataset was unbalanced the observations obviously differed per year.

Pitchfork