It appears that the variables regarding rent and rental status were not included in the 1950 PUMS file created by the Census Bureau. So, while the data was collected (and is available in summary form through NHGIS), the variables were simply excluded from the PUMS files. It seems strange that the rent data would be included in 1940 and 1960-on but not in 1950, which also made me wonder if the variable was mistakenly omitted in the IPUMS files. However, if you look at the original codebooks for 1940 and 1950 and compare the Sample Subject Information (pg 21 of the 1940 pdf, and page 13 of the 1950 pdf) you can see that “Property Value/Monthly Contract Rent” is listed under Household record in 1940, but is not listed in 1950.
I hope this helps.