Generate variable with differences between observations

[Hallo Du]

I have a panel data set on Mexican regions and agegroups, respectively divided by high and low income groups. For every case I calculated a numeracy Index (abcc). I want to calculate abcc differences between the income groups for each state and agegroup. Can anybody help me with the command? I tried < by agegr (state), sort: gen abccc = abcc if med_high_skill == 1 - abcc if med_high_skill == 0 >, but it does not work with this one.

[Carlo Lazzaro]

Hallo:
welcome to this forum.
Please note:
1) as per FAQ example/excerpt (via -dataex-) are really welcome on this forum. Thanks;
2) I'm not clear with what your main by "calculate";
3) as per FAQ again, "does not work" is totally unhelpful to interested listers.

[Hallo Du]

Hello, "calculate" here meant i collapsed the data and obtained an index for numeracy, called abcc. this abcc value was collapsed among states, agegroups and a dummy for income group.
Now I want to generate a new variable containing the differences in this index for each state and agegroup between the two income groups.

[Carlo Lazzaro]

Hallo:
point 1), #2 still applies.