Need help to recreate a chart from Harvard University on Cost Burdened Ratio

Since the geography only goes down to the state level you can use PUMS data.

Dave Dorer

Julien,

I agree with David's comment. To get the 'Severely Burden' category for your analysis, you'll need to use the Public Use Microdata Sample (PUMS), 1-Year Estimate.

There are two main approaches:

Option 1: Straightforward Method

1. Download the PUMS data: https://www.census.gov/programs-surveys/acs/microdata.html
2. Use the variable "Gross Rent as a Percentage of Household Income (Past 12 Month)" (GRPIP) to generate your chart. You might need to filter for "Rented" tenure (TEN).

Note: This method doesn't account for inflation adjustment in GRPIP and might exclude utilities.

Option 2: More Detailed Method (Likely how they did it in the reference Chart)

1. Download the PUMS data: https://www.census.gov/programs-surveys/acs/microdata.html
2. Adjust household income (HINCP) using the CPI-U for All Items (https://www.bls.gov/cpi/). You could technically use the ADJINC variable in the ACS, but it's unclear why they didn't. Use CPI-U to match their approach.
3. Filter for "Rented" tenure (TEN).
4. Create a "rent and utilities variable" by summing variables like gross rent (FRNTP), water cost (WATP), electricity cost (FELEP), etc.
5. Create a rent burden variable by dividing the "rent and utilities variable" by the adjusted household income variable you created in step 2. Finally, use this new variable to generate your chart.

Important Note: I'm not sure how they did it, but technically, utilities should also be adjusted for inflation too.

Hope this helps!