[Jeff Wooldridge]

ABW and AFG are country identifiers. As Carlo says, that's a panel data set at the country level. You can use pooled OLS with time dummies or two-way fixed effects with country as well as time dummies. That's your choice. If the regression is descriptive, maybe not including country FEs makes sense.

With such a large T, you'll likely want to use xtscc without (or with) the FE option, but including i.year.

[Martinо Cоmelli]

Thank you so much for your help Jeff!
I will present to the reader / peer reviewers a pooled OLS and xtscc.

Yes, the regression and my study is is purely descriptive.
thank you again!

[Jeff Wooldridge]

Just to be clear: The estimator doesn't change using "reg" and "xtscc." It's still pooled OLS. The calculation of the standard errors is different.