Connecting person/household data across IPUMS USA and IPUMS CPS?


#1

Hello. I’m new to IPUMS and I’m trying to determine whether my use case is feasible. I’m struggling to understand person codes.

I’d like to examine a relationship pertaining to a household’s children and the receipt of TANF welfare subsidies. It seems like useful variables may exist in both USA and CPS. Is it possible to connect persons and/or households between USA and CPS sets, or are they entirely independent?

Furthermore: I know that because the data is rectangularized, one can connect a person to a household. How, exactly, do I do that?

Thank you so much for any help!


#2

The source data for IPUMS USA and IPUMS CPS are entirely independent. The former provides data from the US census and American Community Survey, the later provides data from the Current Population Survey. Since both surveys are only administered to a small share of the total population, it is very unlikely that any household will be included in both surveys in one particular reference period.

That being said, it sounds like your study could use IPUMS CPS data. Similar to IPUMS USA, IPUMS CPS has a host of family interrelationship variables that provide information about children within the household. Additionally, the IPUMS CPS variables SRCWELFR and MTHWELFR both identify TANF receipt.

In rectangularized data, as is provided by IPUMS CPS, household information is added directly onto the person-level records and there is no separate household record. So, in these files each observation is an individual and each individual within the same household will have identical household-level information.