EViews.com

Dear all, Panel unit root testing shows me that most of my variables suffer from non-stationarity, and are first-difference stationary. A couple of my variables however are second-difference stationary. In have included a screenshot below for reference. Here you can see that I have twice differenced...

With fixed effects you can control for variables that are unobservable or for which you don't have data. It's a way to avoid omitted variable bias in your estimation. You use cross-section fixed effects if you think there are omitted variables that vary across entities (countries, companies, persons...

Anyone?
I tried several econometrics textbooks, but couldn't find anything on it.

Thanks in advance for your help!

Regarding logs: in its basic form it depends on what kind of (interpretation of) output you want. Using logs both sides will give you elasticities, which may be more realistic than levels both sides as log-log would e.g. say that if GDP grows with 1%, the gold price increases with 0.5% per ounce as ...

The Maddison Project could help you find some of the data you are looking for: http://www.ggdc.net/maddison/maddison-project/home.htm

Regarding your financial crisis dummy, a very helpful thread on dummy variables is the following: http://forums.eviews.com/viewtopic.php?t=3840&f=4

So for your model you could create a dummy using the following code:

Dear all, I'm aware that I cannot use non-time-varying variables such as gender in cross-section fixed effects. I would, however, still like to check for such differences in my panel (in my case checking for differences between country subgroups like Latin America and Sub-Saharan Africa). Would it b...