do the 2013 supplement data (volunteer-sept) and (civic engagement-nov), contain the same respondents?


#1

I am trying to analyze American’s volunteer and civic engagement behaviors. I would like to construct a dataset that contains variables for both the Sept and November supplements for 2008, 2009, 2010, 2011 and 2013. I am able to create a single dataset on IPUMS with both supplements but want to make sure these respondents are the same people. I am concerned that perhaps that Sept and Nov data are different respondents. Is there some way to check this to ensure the same respondent was interviewed in both Sept and Nov of a given year?


#2

Yes, the September and November supplements do contain some of the same households, but not all households are the same. Note that the CPS uses a sampling framework that includes the same household for 4 consecutive months, excludes them for 8 months, then includes them for 4 more months. Therefore, most households found in September supplement will be in the November supplement, but not all such households will be found in both months. You can use the variables CPSID and CPSIDP to link households and people across samples. For more details about how to take advantage of this sampling framework, see this paper.

I hope this helps. Let us know if you have any additional questions.