In the past, I have been using the Census’s dataFerrett to access CPS data. Now I am using IPUMS to access CPS data. The problem is that IPUM variable names are not the same as the CPS/BLS variable names. For example, the composite weight used when working with IPUMS is WTFINL. However, the original CPS data refers to the same variable as PWCMPWGT.
Is there anything like a dictionary explaining the correspondence between the IPUM variable names and the original CPS variable names?
Thanks.
First, it should be noted that WTFINL corresponds with the variable PWSSWT in the raw CPS data. IPUMS CPS currently does not have an analog variable corresponding to PWCMPWGT in the raw data.
Second, the IPUMS CPS project does not currently have a direct correspondence between the raw CPS variable names and the IPUMS variable names. Each variable within IPUMS CPS, however, does have a detailed variable description that is often useful in identifying the codes and comparability of each variable.
I hope this is helpful.
First, thanks for your reply. Your answer brings up a several questions.
>> First, it should be noted that WTFINL corresponds with the variable PWSSWT in the raw CPS data.
>> IPUMS CPS currently does not have an analog variable corresponding to PWCMPWGT in the raw data.
First, I take it that the difference between WTFINL (PWSSWT) and PWCMPWGT is not that great especially given that the public data is a perturbed version of the official version.
Second, given how the CPS recommends using PWCMPWGT when constructing various measures associated with the labor force why isn’t there a corresponding weight in IPUMS? It’s not a trivial weight!
Third, clearly the IPUMS staff knows which raw CPS vairables are used to create the IPUMS variables. Why can’t the recipee the IPUMS staff uses be documented and made available to the public. For example, add a tab explaining the variable’s relationship to the raw CPS data. Better yet would be a CPS to IPUMS and IPUMS to CPS dictionary. I would bet that having this type of dictionary would improve IPUMS usage.
Thanks.
I’ll address your questions one at a time.
(1) That is correct, the difference between these variables is not very large. Some users are interested in the differences when they are comparing their calculations with official statistics, but the difference is very small.
(2) The IPUMS CPS project aims to add value to the data that is already publically available. This typically takes the form of making the CPS data comparable over time and useable as a panel. One option, if you want to use the PWCMPWGT weight with IPUMS data, is to attach the PWCMPWGT to the Non-March IPUMS basic monthly CPS samples, using a sequential merge in STATA. This is possible because NBER and IPUMS-CPS data share a common sort order and a sequential merge uses this ordering to match records together. Note, that this is impossible if you’ve used the “Select Cases” tool in the IPUMS data extract system.
(3) This is a good idea and is one that would certainly assist users who are more accustomed to the raw CPS data. I’ll pass this recommendation along to the IPUMS CPS Team.
I hope this has been helpful.
Hi. In response to item #3 above, has there been any progress in making a CPS to IPUMS and IPUMS to CPS dictionary? If so, where can I find it? Thanks.
There is still no such crosswalk between original CPS and IPUMS CPS variables.
Hi, Checking in to see if there is now a crosswalk of CPS to IPUMS CPS variable names?
There is still no crosswalk and no immediate plans to release one. The IPUMS CPS team is aware that there is demand for such a file, but it is not currently a top priority. If you have questions about how particular variables in the raw data are represented in IPUMS, feel free to email ipums@umn.edu for clarification.