if we shouldn’t be using hrhhid for identifier, what is its purpose in the dataset? Does this confuse researcherse
You are correct that in IPUMS CPS data the CPSID and CPSIDP variables are essentially always preferable to the HRHHID and HRHHID2 variables. We provide these original household identification variables primarily for users who are interested in linking IPUMS CPS data to the original CPS files (available via the NBER website). Sorry about any confusion.