I am interested in identifying respondents who were in the labor force the previous year, but I do not see a variable that measures this directly. Does such a variable exist? Or do I have to use another work-related variable for last year as an indirect measure?
For example, there is a variable for “worked full or part time last year.” Would I assume that respondents who worked full-time and part-time last year were in the labor force? If so, would I also assume that those who are labeled “NIU” for this variable were not in the labor force last year?
Any suggestions or help on this matter would be greatly appreciated.
Since many people cycle in and out of the labor force over the course of a year (occasionally multiple times) there are many different ways to define “in the labor force last year.” You could decide that they had to be in the labor force for the entire year, or a majority of the year, or if they spent any time at all in the labor force. Probably the best source for generating any of these definitions would be WKSWORK1 and WKSUNEM1. It should be noted that in order to be asked this question respondents had to have indicated that they worked “at all” in the previous year. It is also important to recognize that for both of these variables, a response of ‘0’ represents both a real 0 (zero weeks worked/unemployed) and ‘Not in Universe’. However, by limiting the sample to only individuals who should be in the universe, all zeros will represent a real zero.
I hope this helps.
Is there a question about labor force participation last year in the original CPS and IPUMS just did not include this question? Or original CPS never asked a question about labor force participation during last year?
As Joe’s answer above suggests, defining “in the labor force last year” is tricky. Therefore, there is no specific variable in the CPS that cleanly identifies this concept. I’d say Joe’s suggestion about generating a variable to define a working concept of “in the labor force last year” with the WKSWORK1 and WKSUNEM1 variables seems like a good approach.