Counts of the healthcare workforce

I have estimates from the PUMS ACS one year sample for 2001, 2004, 2010, 2013 and 2017. I am currently generating counts for physicians and surgeons from the 2019 IPUMS. The crosswalk from 2010 census Occupational codes to 2018 codes shows that in 2018 Code 3060 was split in to four codes: 3065, 3070, 3090, and 3100. My question concerns the undercount of physicians and surgeons I get when I do a weighted count using the IPUMS 2019 ACS physician and surgeon codes 3090 and 3100. Do the iPUMS code include all of the four 2018 listed codes above? I do not understand why I am getting such a significant undercount in 2019.

Your characterization of the change between the 2010 and 2018 occupation codes for physicians and surgeons is accurate. However, I am not seeing evidence of an undercount in the 2019 data. I am wondering if you are restricting your analytical sample using the codes from the unharmonized occupation variable (OCC) but referencing a harmonized occupation measure (e.g., OCC2010). IPUMS includes all available codes; if you look at the “2018 ACS PUMS and 2018 SIPP Public Use Occupation Code List” on the Census Bureau’s Industry & Occupation page, you will see that the PUMS code 3090 includes occupations originally coded as 3065, 3070, and 3090. Below are some quick tabulations using the online tabulator of the OCC (for unharmonized OCC codes of 3060 through 3100) and the OCC2010 code of 3060 over time. I get the same totals with either approach and both appear to be in line with the overall trend. Please let me know if I have misunderstood your question.

Unharmonized OCC:

Harmonized OCC2010:

Thank you very much for your reply. I did find the Census Bureau’s Industry & Occupation page after writing my note to the forum. It appears that my undercount may have something to do with incorrect extraction code. However, not due to occupational code differences as you point out.
Again, thanks for your assistance.

I am not sure what you mean by “incorrect extraction code.” Please follow up if you think there is an error in the data, documentation, or syntax files from IPUMS.