Crosswalk between census block to puma, Census or ACS 2020

Hi, I am looking for the crosswalk boundary data between the census block (15 digits) and PUMA with Census or ACS 2020. I can find the crosswalk data between these two in Census 2010, but I cannot find the one with the 2020 boundary. Do you have any idea how I can crosswalk between the census block and PUMA with 2020 data? Thank you.

The Census Bureau has not yet published definitions or data for 2020 PUMAs. The first data release to use them will be the 2022 ACS, due out in late 2023.

If you’d like to crosswalk from 2020 blocks to 2010 PUMAs (which are the PUMAs used in 2020 ACS data), you could use the NHGIS 2020-2010 block crosswalks to go from 2020 blocks to 2010 blocks, and then crosswalk to 2010 PUMAs using the 2010 block-PUMA relationships you found.

Very helpful. Thank you so much!

Hi Jonathan,

I am doing census tract level crosswalk between Census 2020 and 2010 data, and downloaded Relationship Files from census (Relationship Files). However, this data doesn’t have the variable like (parea) from NHGIS 2020-2010 block crosswalks. And NHGIS doesn’t provide census tract level crosswalks between 2020 and 2010. Do you have any idea where I can download the census tract level crosswalk file? Thank you for your help.

You’re correct that the Bureau’s Relationship Files don’t include area proportions (like the “parea” value in NHGIS crosswalks), but they do include both the numerator (area of intersection) and denominator (area of source tract) for that statistic, so if needed you should be able to use that.

We also plan to add 2010-2020 tract crosswalk files on NHGIS, which will include several interpolation weights in addition to area proportions, but we don’t have a expected date for that yet. It may be several weeks or months.

Much sooner, hopefully within days, we plan to add 2010-2020 crosswalks from block groups to tracts (and to block groups and to counties). I recommend you check back soon and use those crosswalks to the extent possible.

In general, we highly recommend starting from the smallest possible units when creating standardized data. Even if you’re going to analyze data at the tract level, there’s usually no need to start with tract data… Very often, the data of interest is also available for smaller units, like blocks or block groups, and if you allocate data from the smaller areas to your tracts of interest, the interpolation is guaranteed to be at least as accurate as, and often much more accurate than, interpolating from tract data. The NHGIS crosswalks page provides more explanation and illustrations about this.