Structuring bilateral data on excel

[LKie]

Hello,

I have bilateral FDI inflows data varying across time and also bilateral data on distance between countries, and I want to import it to eViews 8. Other variables that I will use in the equation include host country GDP, source country GDP, host country wage and a number of dummies on EMU and EU.

I'm choosing from two ways of doing it, I have attached Excel file with short examples of them. Could you please advice me which one is more suitable or perhaps if there's some other way to do it?

Thank you!

[EViews Gareth]

Neither example seems to make much sense to me.

[LKie]

Perhaps I should have clarified what FDI flows data I have. I have FDI flows from 32 source countries (Austria and Belgium in the examples) to 11 host countries (Slovenia, Estonia in the examples) varying across time from 2000 to 2012 (I included only 2000 and 2001 in the examples). I have attached an excerpt of how it is structured straight from the Eurostat database (only for Belgium as a host country).

I'm not sure if my examples are clearer now but if not, could you please suggest a data structure forgetting my examples?

[EViews Gareth]

Really depends on what you want to do.

[LKie]

I want to examine how FDI flows in selected host countries are affected by number of variables, particularly by a dummy of euro introduction in the country.

[LKie]

EViews Gareth,

What do you think?

[EViews Gareth]

I think you need to detail exactly what you're trying to do in EViews.

[LKie]

What I'm trying to do exactly in eviews is run a regression. I sent you a pm asking to be more specific cause I'm a beginner but fair enough, I guess I won't get much help here...

[EViews Gareth]

You'll get lots of help, but you need to help people help you. Saying things such as "want to run a regression" is really no information at all.

What sort of regression do you want to run? What will be the variables involved? Do you want to run it as a panel regression? If so, what sort of options?

[LKie]

What I was trying to say is that my eViews knowledge is very basic and I simply do not know what information I need to provide to get help.

I have attached a .doc with listed variables and full equation. Yes, I would need to run a panel regression, I would estimate the equation with random effects and fixed effects.

[EViews Gareth]

What are the dimensions of the panel? i.e. what are the cross-sections?

[LKie]

I understand the same as the variables? FDI flows from source country to host country, host country GDP, source country GDP, host country wage, distance between host country and source country, varying over the period 2001-2012 + dummies.

[EViews Gareth]

I think you need to think/learn a bit more about the econometrics you're trying to do before you can think about doing it in EViews. You need to figure out exactly what you're trying to do before we can help you do it.

[LKie]

I'm left quite confused to be honest, could you give me some pointers what exactly do I need to learn? I mean, I know the equation I want to estimate and I know the estimators I will need to use. What else is there?

[EViews Gareth]

Well first thing to learn is what panel data is, then figure out how your data fits into a panel data structure. You need to be able to define the identifiers.

If you are going to do fixed effects, what are the dummies dummying for?