Married couples in the CPS

I can attach spouse characteristics for my selected sample, but do those partners also show up in the individual data? That is, suppose I have the following situation: Individual 1 is married to individual 10 and individual 2 is married to individual 20. Is it the case that individuals 10 and 20 are also in the data as observations? eg

id     id_sp     var1    var1_sp
1      10        100     1000
2      20        200     2000
10     1         1000    100
20     2         2000    2000

or is it just like below?

id     id_sp     var1    var1_sp
1      10        100     1000
2      20        200     2000

So it seems the data like in the first table. But then, given that asecwth is the same for both partners, wouldn’t any analysis using that weight be double counting each household?

Your first example is correct. If you have not restricted your sample using select cases, no records will be excluded from the data. Attach characteristics simply saves you the effort of writing code to attach a spouse’s characteristic to a person record, but it does not remove the spouse’s record from the data file so the spouse will be retained in the data as a separate observation.

ASECWTH is a household-level variable; it will be the same for all household members, not just persons flagged as spouses by SPLOC. For household-level analyses, restrict your sample to one person per household (e.g., PERNUM = 1).

1 Like