I am pulling monthly CPS data from 1990 to 2018 including the earnweek variable for analysis in STATA. According to its documentation, earnweek has a NIU code of 9999.99 and topcodes of 1923 (1990-1997) or 2885/2884.61 for 1998 onward. It then refers you to the user note on topcodes, but that user note and the accompnaying top codes table don’t include any reference to earnweek besides "With the exception of the variable, EARNWEEK, IPUMS code notation only identifies Top Codes that have been imposed by IPUMS coding itself.” While the data contains both the 9999.99 code and the topcodes such as 2884.61, the data also includes the number, 999999, which is not listed in the variable documentation. I am assuming that this also refers to NIU but wanted to confirm as it is not listed in the variable documentation. Thanks!
This could probably be explained better in the documentation, but yes the EARNWEEK code 999999 == NIU for basic monthly samples prior to 1994. You can always verify this by checking the values of ELIGORG and AGE, as these variables help define the universe for EARNWEEK.
Hi, I wanted to follow-up on this thread. I think it would be helpful if the documentation from EARNWEEK was updated to reflect the different NIU codes in 1989-93. Currently, the only NIU code for this variable is listed at 9999.99 in the variable description.
Thanks for the note on this topic. I’ve sent a note to the IPUMS CPS Team about updating the documentation about this detail. It should be reflected on the website with the next data release.