Inconsistency with PUMS data


As a basis for some more complex work, we are trying to calculate the number of households (using HHWT) in NYC (Bronx County, Richmond County, Kings County, New York County, Queens County). There seems to be a discrepancy between the number of households that the 5-year 2019 IPUMS data shows – 3.34 Million and the PUMS data from Census (3.16 Million). I am concerned that this discrepancy at the outset will affect other portions of our analysis, so please let me know what the basis of this discrepancy might be!


I suspect the discrepancy here is that you have not excluded group quarters in your IPUMS estimates. I used the SDA online analysis tool to look at the household counts of these counties (restricting my analysis to only the first person in each household and applying HHWT); my household-only estimates are on par with your estimates using PUMS (and individual county estimates using the 2019 5-year from, but you can see that including persons in group quarters brings the total up to 3.35 million (what you were getting using IPUMS).

Thank you!!

1 Like