When I import the 1940 100% census data to Stata, I find that there are non-zero observations for the variable migcounty only for those individuals whose 1940 state of residence is Alaska, Hawaii or Wyoming. The total amount of non-zero observations is less than 0.5% of the population. Is this a problem in how I am reading the IPUMS-generated data file, or are there no data for individuals residing in other states?
You have found an error in the 1940 dataset that we did not know about. Thanks for pointing this out! We plan to release a corrected version of the MIGCOUNTY variable in the next couple months. In the meantime, since you identified a significant data error you are now entitled to a shiny, new IPUMS Mug! Please email your address to ipums@umn.edu so that we can send you your prize.
Thank you for helping us maintain these important data resources!
1 Like
Thank you Tim. Any idea when the updated dataset will be available? I created an extract on 4/13/2023 and found the same problem with MIGCOUNTY.
I didn’t see the date on the former post…sorry. Only at 13mm records (of ~131mm) have MIGCOUNTY with a value not equal to 9999. Is this as expected??
I am not able to replicate what you are seeing. I requested a 1940 full count data file with the variables MIGCOUNTY and MIGRATE5 (which reports if the person lived in a different house five years ago). I see 48,980,536 individuals with MIGCOUNTY values of 9999; the same number of people who report living in the same house five years ago, which I would expect. This is 37% of the observations in the full count file. Please let me know if I have misunderstood the issue you are describing.