Number of own children: CPS IPUMS vs CPS Basic from BLS website

I have been working with original files from the BLS (Basic Monthly CPS) and with IPUMS. Focussing on the share of individuals with own children, I got a significant large difference between two CPS versions. Focussing on the age group 18-64, in IPUMS slightly less than 59% of individuals do not have any own children (nchild) vs. about 68% (prnmchld) in the BLS. I cannot find similar discrepancies for other variables (gender, age, race/ethnicity/education/marital status/region).

I found this old link (I would like to know how the NCHILD variable compares to the variable PRCHLD on the CPS Basic Monthly.). 9pp differences seems still like a very large difference. Do you have any thoughts/insights on this?

As was described in the thread that you linked, NCHILD includes children of any age, while PRNMCHLD only counts unmarried children 18 or younger. I suspect this is causing the difference you are finding. I’d recommend tabulating the number of own children 18 and under in IPUMS (using MOMLOC or POPLOC) to better compare with PRNMCHLD. You can also get closer to this by limiting only to parents who have ELDCH<19.