Indicator for more than one marital change

[Lydia Palumbo]

Dear Stata Users,

I have a panel dataset and I would like to identify marital changes (var marital).
Suppose that 1 represents single, 2 marriage and 4 divorce, I would like to create a unique indicator that
considers that the individual 1002352* had a transition from single to marriage AND, then, to divorce.
So far, I used the lagged operator because I was interested in just one transition and not in the subsequent one.

But I am not so sure on how to add the second change, Do you have any suggestion?

ID t marital
1002352* 1 1
1002352* 2 2
1002352* 3 4
1002352* 5 4
1002352* 6 4
1002352* 7 4

Thank you, Lydia

[Andrew Musau]

input str10 ID float(t marital)
"1002352*" 1 1
"1002352*" 2 2
"1002352*" 3 4
"1002352*" 5 4
"1002352*" 6 4
"1002352*" 7 4
end

bys ID (t): gen change= sum(marital!= marital[_n-1]) if _n>1 
bys ID (t): egen wanted = max(change)
replace wanted= wanted>1

Result:

 l, sep(6)

     +------------------------------------------+
     |       ID   t   marital   change   wanted |
     |------------------------------------------|
  1. | 1002352*   1         1        .        1 |
  2. | 1002352*   2         2        1        1 |
  3. | 1002352*   3         4        2        1 |
  4. | 1002352*   5         4        2        1 |
  5. | 1002352*   6         4        2        1 |
  6. | 1002352*   7         4        2        1 |
     +------------------------------------------+