ACS Data Users Group

I am currently working on a project to estimate the number of (renter) households by AMI ranges (such as 80, 50 and 30% of AMI) using micro data (PUMs). How can I add AMI to the PUMs data file to analyze them together? The geography in PUMs is PUMA (not matching the AMI geography) and AMI is based on family size. Once I solve the issue with the geography, do I need to make select the AMI based on the household size (from PUMs)?

Thank you

When I do this, I create the closest version of the MSA possible using PUMAs and then calculate the AMI for that created area. You can then adjust that AMI based on household size and place your renters in the correct AMI category.

I'd be happy to send you the Rental Housing Affordability in Louisiana 2021 report that does this. It has a section on methodology that explains how we place renters and units based on AMI.

Hello Douglas,

I found the report you referenced here. I will review and try to replicate with recent data for Tennessee. Did you create MSAs from PUMAs because you need the data for MSAs? Or is it because metro and non-metro distinction is necessary for AMI consideration?

Thank you

We use the MSAs because that's how HUD does it. Although there are some exceptions to this where HUD creates an AMI for specific counties in an MSA.

You could just take HUDs value and use that once you figure out which MSA/county the PUMA is in, but we wanted to use that actual calculated AMI from the ACS data to keep things data sources consistent.

I did a project with the Atlanta Fed that covers Tennessee. They have data that might be useful for you from the 2021 5 YR ACS data.

It's located here: www.atlantafed.org/.../southeastern-rental-affordability-tracker.aspx

I will check them both out. I am using the most recent ACS, but at least I can verify my calculations.

Thank you so much for your time!!