Problem with IPUMS pointer variable??? (Brazil 1980, 1991, 2000)


I have identified a non negligeable number of individuals indicating to be the child of the household head (relate=3) or of the subfamily head (subfrel=3) but who are not linked to any parent in the household (parrule=0).

This does not make sense to me. Do you have any explanation?

The identification of parents is a crucial element in my paper so thank you very much in advance for your help.



There are at least two possible causes. One is that some households may not have heads/spouses, which should not happen in Brazil, which used a de jure method of enumeration, but there could be data errors. The other possibility, which I believe explains these cases, is that there is an impossible age difference between the child and the parents (less than 10 years). There is a data error of some kind, and it is unknown whether age or relationship is at fault. IPUMS does not currently perform data editing, so such contradictions can be present in the data.

Matt Sobek