I am doing an analysis where I would ideally like to distinguish between biological coresident fathers and social fathers among cohabiting couples. Is there a way to do this with the new parent pointers (in the ACS 2018)? If the reference person is the social father, I can quite easily determine the relationship - using relate. If the father does not identify the “child” but the child’s pointer points to the reference person (i.e., pernum ==1) then I can assume it is not a biological child. In the 2018 ACS this is the case for nearly 3K children so it’s not an unusual situation. What I can’t figure out is how to identify these cases if the mother or another person is the reference person. In the old parent pointers I don’t think social fathers were linked to children - so I could go back to 2015 where the old pointers are available but is there a way to do this with the 2018 ACS data? Thanks!
Your assessment is correct. Unfortunately, there is no easy way to do this with the family interrelationship variables in IPUMS USA. This is due to (a) a lack of information about this specific detail in the raw ACS PUMS data and (b) the method used by IPUMS to probabilistically identify family interrelationships, which primarily focuses on social relationships.