I’m looking at quite a few samples, from 2013 back to 2000. I understand how some of these will use different weights and how industry classification has changed. However, none of the samples should have a response be -1 for industry. I was wondering what this meant. From what I pulled, this was all from 2000-2002.
The -1 is a value from the source CPS data. It corresponds to a value of 0 in the IPUMS-CPS data. The code was simply missed in our recoding structure, but the correct data will be released with the next IPUMS-CPS update. In the mean time, you can simply recode the value of -1 to 0 to harmonize the variable across all samples.
As is the policy here at IPUMS, because you identified a significant error in the IPUMS data, thereby helping us maintain the reliability of our data, you are entitled to a shiny new IPUMS Mug. Just email an address where we can send the package to ipums@umn.edu and you will receive your much coveted prize.