Panel data not balancing for some reason?

[stats321]

My panel isn't balancing and I don't know why! I've followed the instructions in the handbook- but then it just reverts back to the default setting and won't let me estimate with two-way random effects or mixed effects.

Also, on a side note which doesn't really need to be answered- when I imported my data from excel, it just copied the first observations in for all the cross sections and I had to manually ammend them. Am i doing something wrong?

As you can tell, this is the first time I am using Eviews!

log3.wf1

(120.1 KiB) Downloaded 202 times

[EViews Gareth]

I don't follow. That workfile is balanced.

[stats321]

This is what I thought, but then when I loaded it into a new workfile, I selected the balanced panel option yet afterwards when I had imported my data (and tried inserting it manually too) it keeps telling me it is unbalanced.

I then proceeded to 'balance' it per the corrections and it still told me it was unbalanced

[EViews Gareth]

You have 20 years and 57 cross-sections. You have 1140 observations. 20x57 = 1140. It is balanced.

[stats321]

See attached for my print screens. It doesn't allow me to do so with the number of observations in the background- but I promise i haven't done anything strange!

[EViews Gareth]

The actual data may be unbalanced due to NAs. You can't have non-symmetric NAs

[stats321]

Sorry do you mean that if I have an NA for one country in one year but I do have an observation for another country in the same year, it is unbalanced?

[EViews Gareth]

Yes, that's what unbalanced means.

[stats321]

Do you have any reccomendation of how I can balance this- as I assume the inserting observations option won't be relevant here?

Or to balance it, do I need to delete the other observations if I don't have them for the full data set? I have quite a few missing pieces of data so this may not be the best option.

Thank you so much.