merged countries that were divided later on

[Tilo Gashi]

Serbia and Montenegro dissolved in 2006. Now they are separate countries. In my analysis, from 1990 until 2014 I include the trade flows of both countries, Montenegro and Serbia.
But, how to include those from 1990 to 2006 since they were merged into one country?

[Sebastian Geiger]

There is no perfect answer to that question. There are different alternatives. Which one fits best depends on your research question. Here are some alternatives:
1. Combine both countries even after they split up. In this case you may include a variable for post or pre split-up years of Serbia and Montenegro if your doing a regression analysis.
2. Treat all three entities (combined, separate) as countries (i.e. missings for the years where they did not exist). This will make your panel unbalanced (if it was not unbalanced from the start).
3. Exclude them altogether. I guess you do not have data on every country anyway. If you are not doing research on Eastern European countries specifically, it may be justified to exclude both countries because they might not be the focus of the analysis.

I tend to the first alternative, but it certainly depends on your research question.