Dear all,
I am surprised that in the IPUMS CPS unemployment duration in weeks (DURUNEMP) of a person reporting to be unemployed is only available from 1994 on for the basic monthly data. In the original CPS, this variable is also available before the CPS redesign in 1994 for the basic monthly data. Does somebody know something about this? Maybe there is another variable available that I am missing here …
I am actually unaware of a duration of unemployment within basic monthly samples prior to 1994. If you have some documentation on this, could you send it our way? Of course, if this data is indeed available in these samples, we’d like to make it available via IPUMS CPS.
Of course there is. In the January 1989 files this is item 22c, for example. Note that there is an important difference in the coding of this variable before/after 1994. This is explained in more detail, for example, in the appendix of Shimer (2012), “Reassessing the ins and outs of unemployment”.
Best
It seems that the variable you are referring to in January 1989 is not the same as the variable used for DURENEMP. One difference involves the universe for the variables, since item 22C doesn’t seem to be restricted to only those who are not working. This could be addressed with some internal programming on our end. Additionally, it also seems that different people are asked different questions as a part of item 22C based on the response to A$WHYLK. The documentation does not seem to be all that clear as to how these different questions are assigned, so it would take a bit of work on our end to figure this out.
Therefore, although these variables clearly measure similar concepts, they are also quite different. That is why DURENEMP is currently only available back to 1994 in the basic monthly surveys. The IPUMS CPS team will look into the possibility of extending this variable back to pre-1994 basic monthly samples. At this point, however, I am not able to forecast when this work will be completed.
@JeffBloem Hi,
are there news on this variable? I have found the variable UH_UNDUR_B1 which seems to work for the period before 1994 but this is only available between 1976 and 1988.
Do you know of other variables for the period between 1988 and 1994?
Thank you very much! Best
Jeff successfully defended his dissertation this spring, and has wrapped up his time with IPUMS. However, I am able to answer your question.
You may also be interested in the unharmonized variable UH_WKSUN_B1, which reports the number of weeks unemployed and looking for a job. Please note–as with all unharmonized variables–there is less documentation than variables harmonized by IPUMS. Specifically for UH_WKSUN_B1, there are a number of unusual and undocumented negative codes in 1988, and there are definitely cases that don’t adhere to the universe definition in the original documentation. We will be updating the universe tab soon to reflect that the respondent needed to be a civilian aged 14 or older in all samples. The IPUMS CPS team has made a note that this unharmonized variable may be appropriate for expanding DURUNEMP into basic monthly data for earlier years, but has no firm timeline for this work.