Test difference between mean of a var before and after an event.

[Tomas Jankelevich]

I have a data like this:

Firm Year Event time_event Capital
1 | 1 | 0 | -1 | 100
1 | 2 | 1 | 0 | 150
1 | 3 | 0 | 1 |125
1 | 4 | 0 | 2 |100
2 | 1 | 0 | . | 200
2 | 2 | 0 | . | 200
2 | 3 | 0 | . | 200
3 | 1 | 1 | 0 | 300
3 | 2 | 0 | 1 | 300
3 | 3 | 0 | 2 | 300

And I want to calculate de Mean of the Capital for the years after an event (1) and for the years before an event (-1), and test its difference no matter which firm it is.

Thanks!

[Carlo Lazzaro]

Tomas:
welcome to this forum.
You might be interested in:

. g flag=0 if time_event==-1
(9 missing values generated)

. replace flag=1 if time_event==1
(2 real changes made)

. ttest Capital, by(flag) unequal

Fot the future, please use CODE delimiters and/or -dataex- to share codes/outputs/excerpts/examples with interested Stata listers (see the FAQ about these and other posting-related topics). Thanks.