Panel data 3 rounds: how do I count number of unique IDs for which a dummy=1 in none, one, two or all of the rounds

[Annie Ggg]

Hello,
I am struggling with a simple bit of summary statistics with panel data and wonder if someone can help.

I have a balanced panel data set with time variable: round (1, 2, 3), and panel variable: id (generated by egen group(householdID))
I have have a dummy variable: shock (1/0) , reported for each id for each round

I would like to count and produce a table that reports the percentage of households/unique ids that experience 0 shocks, 1 shock, 2 shocks and 3 shocks, across the 3 rounds (I do not care whether shocks are consecutive, e.g. a household experiencing a shock in rounds 1 & 3 only I would count as having '2 shocks').

Many thanks
Annie

[Marcos Almeida]

Welcome to the Stata Forum / Statalist

Please read the FAQ. There you'll find important advice on sharing data/command/output.

In short, you may use code delimiters or - dataex - for that matter.

That being said, you'll probably need to use - egen - command.

The reply was quite generical. A more precise reply demands workable data.