It appears that the unique person identifiers have been reconstructed for the IPUMS CPS dataset as opposed to the raw CPS ASEC files from the Census Bureau. Is there a way to crosswalk the CPSIDP from the IPUMS CPS data to the PERIDNUM in the raw CPS ASEC files?
I am not sure if you are interested in PERNUMID as a unique identifier in general or linking data over time; I will provide a bit of information on both. PERNUMID is not available in the IPUMS CPS extract system. CPSIDP is created by IPUMS to link records over time and leverage the panel aspect of CPS; I am linking to a paper that describes the creation of CPSIDP. If you are interested in a unique identifier to crosswalk IPUMS CPS to the original CPS data, you should use the IPUMS variables HSEQ and LINENO, which correspond to the original variables PH_SEQ and A_LINENO. Please let me know if I am not answering your question.