How to use pumas and pwpumas with multiyear files


I’m analyzing data by region in CT using the 2010-2012 file and I just realized that the 2012 file is now using the 2010 census definitions while the 2010 and 2011 files are still using the 2000 census definitions. Is the 2010-2012 IPUMS file harmonized, or does this file has different pumas depending on the year of the dataset?

If it has different pumas and has not been harmonized, is there any recommendation you could give me to conduct my analisys?

I would really appreciate your help or any guidance you could give me.




You are correct, in multi-year data files PUMA is based on the year of the original survey (identifiable using MULTYEAR). The variables affected by the PUMA change, as well as a brief description of the change, can be found in the release note that accompanied the IPUMS-USA 2012 ACS release.

Since you are interested in a relatively specific geographic area, I would recommend trying to create consistent boundaries by grouping PUMAs so that a group of PUMAs from the 2010 and 2011 respondents covers the same area as a group of PUMAs from the 2012 respondents. You can do this by looking at the PUMA maps for the Census 2000 based PUMAs (available through IPUMS-USA) and the PUMA maps for the Census 2010 based PUMAs (available through the Census Bureau). You can also use the resources available through TIGER/web if you prefer interactive maps.

I hope this helps.