Matching 1% IPUMS Sample to Full Count census


Dear Joe,

I have access to the full count version of the 1930 Census of IPUMS through NBER servers. I am trying to match the 1% sample of the 1930 census that is obtained from the website to the full count census using household and serial identifiers, but this is not working. I was wondering if you could provide some insight on how to match these two datasets.




SERIAL and PERNUM are only meant to uniquely identify individuals within a given sample. To link between the full count 1930 file from NBER and the 1930 1% sample I would recommend using REEL, MICROSEQ, PAGENO, and LINE, available in the IPUMS-USA Household->Historical Technical variable group. The IPUMS-USA 1930 samples also include NAMEFRST and NAMELAST (in the Person->Historical Technical variables group) which should be useful for verifying or making matches as well. These variables should allow you to match almost all of the cases from the 1930 1% file

I hope this helps.