Hello,
For the CPS variable whyunemp, I am wondering what the difference between the categories “Job Loser/On Layoff” and “Other job Loser” is? Does “Job Loser/On Layoff” include both those who are on temporary and permanent layoff, and is there a way to differentiate between permanent and temporary layoff in the IPUMS data?
Thanks.
Job Loser/On Layoff includes only those who are on temporary layoff, which means they expect to be recalled back to work either at a specific date or within six months (see the Unemployment Concepts section of the CPS Interviewer Manual). Those who are permanently laid off will generally be included in one of two places:
Those who are searching for work will be counted as unemployed and have WHYUNEMP = Job Loser/Other
Those who are not searching will be counted as out of the labor force, and will be NIU for WHYUNEMP
I don’t believe there is a reliable way to identify all of those who are permanently laid off and currently NILF in the microdata, versus those who were fired. You may be able to get at this partly using the unharmonized variable UH_WHYLFT_B2. You can find additional detail on layoffs in the Displaced Worker supplement, but this is only fielded once every two years.
Hello,
I have a follow-up question on this. Are these definitions applied consistently over the entire time span that the variable WHYUNEMP is available over (i.e., 1976-present)? If not, how have they evolved?
It seems that the universe of the variable has changed in 1989; prior to that, it included “unemployed (looking for work)”; only afterwards does it include "unemployed (looking for work) or on layoff ". What does this imply for the category “Job loser/On Layoff”?
Thank you in advance for any clarification.
The original documentation is somewhat unclear about this. In 1988 and earlier, the only identifiable question similar to WHYUNEMP is only asked for those who are looking for work. But it is likely that this question was combined with other information on layoffs to produce the source variable for WHYUNEMP. If you look at the crosstab of WHYENUMP and WHYABSNT for unemployed individuals in 1988 versus 1989, the only noticeable difference is that those who were classified as “on temporary layoff” for WHYABSNT have WHYUNEMP=“temporary job ended” in 1988 and earlier, and they switch to “job loser/ on layoff” in 1989 and later (see screenshot - this only includes a selection of months for 1988 and 1989). So it seems those on layoff are in-universe for WHYUNEMP both before and after 1989. I’ll let the IPUMS CPS team know that the documentation should be corrected.
Thank you very much for your answer.
Yes, I have noticed that regularity. In fact, it seems that prior to 1989, (almost) all workers who are classified as “on temporary layoff” in the WHYABSNT variable are assigned “temporary job ended” in the WHYUNEMP variable. Between 1989 and 1993, they are instead classified as “Job loser/layoff” in the WHYUNEMP, whereas the “temporary job ended” category effectively ceases to exist in those years. However, it reappears again 1994.
I understand that it may be difficult to treat layoffs in a manner that is perfectly consistent across time, given that the questions posed to respondents have changed (at least so in 1994). In particular, I imagine it may be difficult to ascertain whether those classified as being on an indefinite layoff expect to be recalled within 6 months or no. Nevertheless, the category “on temporary layoff” in WHYABSNT contains workers who have been given a definite date to return within 30 days of when they were laid off. Thus, it would seem to me that they should fit squarely within the definition of “Job loser/layoff” in the WHYUNEMP variable, which you have given in your original answer. That is indeed the case between 1989 and 1993. Yet, as you have noticed, that is not the case prior to 1989 - instead, they land in a different category in WHYUNEMP. Do you believe that this is an inconsistency, or is there some deeper logic to it?
To me it looks like an inconsistency in how the Census Bureau or BLS decided to categorize these individuals. There might be some discussion of this in documentation about the 1994 CPS redesign, but I haven’t seen anything about it. Thanks for bringing this up. I’ll mention this inconsistency to the IPUMS CPS team so they can consider adding a note in the documentation for WHYUNEMP.
No problem at all. Thank you for clarifying this.
