How to determine number of persons in a household born after a given year


Hello, new here. I know that there are variables for youngest and oldest children in a household. I’m wondering whether there’s a way for me to determine number of children in a household born after, say, 2003.

Is the following possible?

(1) Group persons into households, which seems possible because data is rectangularized.

(2) Make a new variable that indicates whether person was born after 2003 (year of survey - age will give me year born, and from there I can determine Yes/No born before 2003)

(3) Make a new variable again, summing up the number of persons born after 2003

Should this work?

Thanks for any insight!


The method you propose above sounds perfectly reasonable. In principle, performing this method using the year of survey will induce more error in your measure than using the month and year of the survey. This is because, everyone ages up once within each calendar year. So, people who happen to be surveyed before their birthday within a calendar year will mechanically look younger than those who are surveyed after their birthday. Using the month and year cuts down on this source of error somewhat because in 11 out of 12 months within each year people do not age up.