I took a look at your most recent IPUMS USA extract, and I see that it is an extract from the 1% sample of the 1950 decennial census. This sample includes only 1% of the records in the full 1950 census. There are zero person records in the sample that match the criteria you specified using the select cases feature: STATEFIP=53 and RACED=200 and OCC1950=(071 or 075). If you use the full count 1950 census data instead, which are also available from IPUMS USA, there are a total of six person records meeting those three criteria. There may have been more African American doctors in Washington in the 1950 census who are not identifiable using these filters; there are over 5 million individuals with OCC1950 “not yet classified” and some of these could be doctors. When someone’s occupation is coded as “not yet classified” in historical data, it usually means their occupation response was unclear, illegible, or a very low frequency (rare) occupation that IPUMS has not yet coded. This forum post provides more information on the “not yet classified” code in occupation.