I’m working with some years of the June CPS and noticed that in 1990 and 1992, the UHRSWORKORG variable has hours information for workers that PAIDHOUR indicates are not hourly workers. It seems that UHRSWORKORG is identical to UHRSWORK1 in those samples. Potentially other months/years as well? But I have only looked at the 1990 and 1992 June CPS.
From the “Hours Worked Variables Notes” page in the technical documentation section, I understand that there are times where UHRSWORKORG will not have hours information for workers that PAIDHOUR indicates are hourly workers, but I don’t see information anywhere on the above.
For 1989-1993 samples, the information contained in UHRSWORK1 and UHRSWORKORG are identical. The question is asked from all employed ORG respondents, regardless of their pay schedule, and the wording refers to their main job. I agree the documentation about this is unclear, and I am working with the IPUMS CPS team on improving the documentation. For simplicity, I recommend using UHRSWORKORG only for 1994 and later, and use UHRSWORK1 for 1993 and earlier (with the understanding that they are not entirely comparable in the two periods, since UHRSWORK1 asks about usual hours worked at the respondent’s main job, while UHRSWORKORG–in the post-1993 period–asks about the usual hours worked at the pay rate specified in the Earner Study questions).