Looking at same-sex marriages in 2012, I noticed a lot of "0" for the YRMARR variable. What does this mean?

I have created a same-sex couple dataset using 2012 IPUMS ACS data and specifically used the qrelate variable to identifiy same-sex marriages. A lot of the heads and spouses have “0” for the YRMARR but I wanted to know why this is and if there was a code for “0”? I am trying to modify my dataset but I feel as if I am losing too many same-sex marriages in 2012 because of the “0” in the YRMARR variable.

The 2013 ACS is the first sample to retain unedited sex and marital status variables for same-sex married couples. Unfortunately, this means I would strongly advise using caution with the YRMARR variable for same-sex couples (identified with QRELATE=9) in the 2012 ACS sample. A high percentage of those respondents with QRELATE=9 have been recoded as MARST=“6:Single/Never Married” and have a corresponding YRMARR value of zero. It is unclear why some QRELATE=9 cases have a non-zero value for YRMARR; thus, it is certainly not a guarantee that these cases are an unbiased sample of same-sex couples.

Hope this helps.