How is AGEMARR calculated in IPUMS-USA?


I was looking for an explanation for how AGEMARR is calculated in, say, the 1970 and 1980 decennial census samples. In IPUMS-CPS, the description for AGEMARR is quite clear in that it uses data on the form but not in the public-use data (namely, birth month and month of first marriage). Is this also the case for AGEMARR in IPUMS-USA?



The Comparability Tab hints to how this variable is calculated. For samples between 1960 and 1980 this variable was not asked directly, but calculated by the reported date at first marriage and date of birth.

I hope this is helpful.

Thanks Jeff. I had read the Comparability Tab as well, but it still leaves an ambiguity. My question is about the inputs into that calculation.

If you look at the form (pages 1 and 2 in the cited), individuals recorded the month (as text) and year of birth and the month (as text) and year of first marriage, and then were asked to bubble in the quarter that corresponded to that month in each case. IPUMS has the quarter recorded as public-use data, but not the month (for the 1970 and 1980 census samples). So my question is: when AGEMARR is calculated, is the calculation a function of monthly data or quarterly data?

Thanks again.

Yes, I agree the documentation could be a bit more clear about the specific calulation of this variable. Based on the Questionnaire Text, however, I beleive for the 1970 and 1980 samples, this variable was calculated using the monthly data rather than the quarterly data.

Thanks Jeff, that was my guess as well.

As one additional piece of somewhat corroborating evidence for this interpretation, the Comparability notes for AGEMARR in IPUMS-CPS (as opposed to IPUMS-USA) is suggestive (albeit vague) that the calculation is the same; that is, that IPUMS-USA also uses the monthly frequency data for calculating AGEMARR.

1 Like