Number of Adults Living with Parents/Grandparents


I recently read an article from Zillow (see link below) that showed the number of adults living with a parent or grandparent is at an all time high due to the COVID-19 pandemic.

I was hoping someone could help me find this data using the Basic Monthly CPS samples?

Ideally, I would love to obtain historical monthly data going back to just before the Great Recession to determine how the number of adults living at home with their parents/grandparents today compares to what happened during the height of the financial crisis.

Any help would be greatly appreciated!


According to that article, “To be considered living in their parent’s home a person had to be the child or grandchild of the head of household.” This can be found using the IPUMS variable RELATE. For generational statistics, use AGE. For the labor force estimates, you’ll want to use EMPSTAT and DURUNEMP.

Thanks for the response! I was able to get a working dataset by filtering for people whose relationship to the head of the household is child/grandchild and who also are at least 18 years old. I went ahead and selected monthly CPS surveys going all the way back to 2000 because my overall goal is to construct a monthly time series of the data so I can see how the series behaved before/after the Great Recession as well as before/after this covid recession.

When I downloaded the dataset, however, I was surprised to see the final dataset only contains about 2.6 million cases. If I’m interpreting the data (and filter I applied) correctly, shouldn’t each observation/row correspond to one adult at least 18 years old still living at home (in other words, the individuals I’m looking for)? It seems like there should be more total observations in the dataset, since, for example, the Zillow article stated there were 32 million adults living with their parents in April 2020.

So with the way this data is structured, would that mean there should be 32 million rows in the dataset just for the month of April 2020 alone?

Thanks again!

Since the CPS is a sample survey, only a small fraction of the total population (about 60,000 households) is surveyed each month. In order to produce population-level estimates, it is necessary to apply sampling weights. For person-level analysis (e.g. calculating the number of adults living with parents), you’ll want to use WTFINL (this is automatically included in all IPUMS CPS extracts). Exactly how to apply the weight differs based on which statistical software you are using. Some examples of using weights with IPUMS data can be found here.