Can I link 1880 NAPP sample to 1880 1% IPUMS sample?


Is there a way to link individuals between these two samples? The serial and pernum identifiers do not translate between the two files. But could there be a link based on microfilm reel and line number instead?

I ask because I have used the 1% sample to do some follow-on linking the 1880 Ag Census and now I would like to merge those hand-constructed links into the NAPP 100% sample.

Thank you.


There are several key identifiers in both the NAPP 1880 samples and the 1880 1% sample in IPUMS USA that you may be able to link on. For example, you could look into using NAMEFRST and NAMELAST as well as some demographic information such as AGE/BIRTHMO and SEX. Additionally, there are several historical technical variables available in IPUMS USA for the 1880 1% sample that are also available in the NAPP 1880 samples as either harmonized or source variables. For example, REEL is available in both, and STREET (IPUMS USA) is also available in the 1880b NAPP sample as a source variable. This information may also help to confirm links using first and last name.

Just a few additional notes: the sample and the 100% file data entry were performed separately by different teams so some strings (like names and streets) may not match. REEL, MICROSEQ, PAGENO, and LINE are fairly stable for linking households between the sample and the complete count file, however household compositions may change slightly between the two (fewer or more people in the household).


Thank you very much!