Not in the labor force that want a job using CPS basic monthly files

CPS
1

Hi,

I have a question about deriving estimates of the number of people not in the labor force that want a job using basic monthly files. As a crosscheck, I tried to derive the 2018 annual average CPS national published estimate for not in labor force, want a job (5,249,000). It appears that the variable WANTJOB is only available in the ASEC IPUMS files, though, not the monthly. Also, it does not seem that WNLOOK covers all not in the labor force people that want a job in the basic monthly files; when I sum the non-999 responses from WNLOOK (US 2018 annual average), I get 4,472,509, much less than the 5,249,000.

A note in Table 35 referenced below says that the 5,249,000 figure “includes some persons who are not asked if they want a job.” Are you aware if we can derive this 5,249,000 number using the IPUMS basic monthly files? I know the Census modifies some of the individual records, so the totals between published figures and those using PUMS may differ slightly, but it seems like something else is going on here to have that large of a difference. Using annual averages would allow for a bigger sample than the ASEC, especially helpful at the state level.

Note – the 5,249,000 figure comes from Table 35 here: CPS Tables

Thank you for any assistance you can provide.

-Tom Pethan

2

There is an unharmonized variable available in IPUMS back to 1994 called UH_DWWANT_1. This is the closest variable that exists in IPUMS basic monthly samples. Using this variable, I was not able to exactly replicate the numbers in the table you referenced, but it was fairly close (4.9 million in 2018 and 5.2 million in 2017). I don’t know who the “some people who weren’t asked if they wanted a job” are but you could contact BLS to ask for details: Contact the Current Population Survey (CPS)

3

Thank you for response, Matthew. I had not been looking at unharmonized variables. While browsing through some of those, I did end up finding UH_NLFWANT_1, which does appear to be what I was looking for based on a few checks. Thanks again for taking a look!