Unicon Research Corp used to provide this variable, but since they have shut down, I have no idea where to get this information. Is there any way to obtain the “suprec” variable (it denotes whether an observation from the March supplement has an imputed value)? It was part of the CPS utilities. More generally, will IPUMS eventually have all the materials Unicon cleaned or is the status of those materials unknown?
The ‘suprec’ variable was created by UNICON based on the variable FL_665 from the source CPS microdata. IPUMS CPS is currently planning on integrating this variable but I am not able to say when this would become available at this time.
In the mean time you can merge the variables from the NBER files onto the IPUMS-CPS files using the instructions found here: CPS has data available on multiple job holders. Where is the comparable variable in IPUMS' CPS?
I’m sorry I wasn’t able to provide the data you were looking for at this time. But I will pass on your interest in the variable. There are a number of variables that we would like to add to the IPUMS CPS collection and prioritizing is the most difficult part, so any input on variables you are aware of that would benefit your research is greatly appreciated.
Thank you for the guidance. I looked into FL_665. Helpful note, this is labeled as FL-665 for some years (1991-2000 at least). Follow up question, this variable is constructed from 1988 onwards, but for 1988-1991, FL_665/FL-665 is not listed in the data dictionary, is there another variable “suprec” is derived from in these years?
Sorry for the multiple postings. To anyone interested in recreating the “suprec” variable, you can do so from 1988 onwards using the CPS micro data at NBER as suggested. Here are the facts you need to know:
-
FL_665 is the variable name for the survey completion from at least (2010-2016)
-
FL-665 is the variable name for the survey completion from at least (1991-2000). From 2001-2009, it is one of those two designations, I just haven’t looked in detail yet as to when it changes.
-
For 1988,1989,1991, there is no singular variable that denotes this. If you look at the technical documentation files provided by NBER under the section entitled " Allocation Flags for March Supplement," you will see a list of dummy variables that denote whether a specific category was imputed. Based on my understanding, the variable “suprec” takes a value of 1 if any of these dummy variables is equal to 1 (i.e. allocated). Using this information and merging accordingly, you can recreate the “suprec” variable.
Thanks again for the initial guidance, it really helped me figure everything out.
Follow up question. “Suprec” is available in the CPS utilities provided by Unicon from 1988 onwards, and this is based on the variable FL_665, but FL_665 does not exist for the years 1988, 1989, and 1990. As I posted in a previous comment, I simply created the variable for those years using the allocated supplement flags. However, after further investigation, it seems like this is not equivalent.
Can you provide me with more information about how exactly FL_665 is constructed and/or how to create it for the years 1988, 1989,1990?
It is definitely possible since Unicon did so, I just seem to be missing something.
The FL_665 variable for the 1988-1990 ASEC samples was actually released separately by the Census Bureau. The files containing the variable as well as linking information can be found on the Census Bureau’s CPS Demographic Extract Files page under the “Miscellaneous data extract files” heading.