Merge dated paneldata in a gravity framework

[Dde100]

Hey,

For my thesis I'm using a gravity framework for 5 different countries, for which I have 5 different excel files.
For my main analysis, I need to merge these 5 files to run pooled regressions.
I used Eviews for all the files seperately and it works perfectly fine. For my research, I need to use the 'fixed effects' option.
Now I merged my 5 files in Excel and sorted them on importing country, however, now Eviews does not recognize it as 'dated panel' anymore.
It says its 'unstructered/undated' and when I click the 'dated panel' option, it says it has repeated values and does not let me continue.
I need the 'dated panel' recognition in order to use 'fixed effects' which is of vital importance in my gravity framework.
Is there any way around this? Do I need to merge it in a different way? Or do I have to code in the fixed effects manually?
I added my merged excel file for convenience.
Thanks in advance already for your help.

Cheers.

[EViews Gareth]

What are your panel identifiers?

[Dde100]

The cross sectional is: Importer.
The time variant is: year.

Importer is the country which imports from my 5 countries of interest.
Year is the years of interest to me. I think the problem lies in the fact that they overlap in the cross section.
Example:
I have multiple observations for Italy in 1980.

[EViews Gareth]

Yes. A panel has a unique observation for each cross-section/year combination. You do not have a panel.

If you just want to estimate with country effects, structure it as unstructured, then use @expand(industry, @dropfirst) as a regressor.

[Dde100]

Thanks! Works perfect now.

[Dde100]

Hey,
Can you maybe also tell me the command for random effects and the hausman test?
Again for the unstructured data I cannot really find that on the forums.
Thanks

[EViews Glenn]

Use the LS estimation method, with the "CX=R" option for random effects. From the dialog, use least squares and just look in the options tab for the effects settings. Once estimated, you'll have an option for the Hausman test as a view off of the equation.

[Dde100]

I don't really understand. I cannot really figure out where to put in the command 'cx=r'. I added a screenshot of what the options tab looks like for me. Where do I exactly put this command for the random effects in?
My panel data is unstructured/undated.

[EViews Gareth]

You do not have a panel.