The cost of renting a home remains high in across the United States (Joint Center for Housing Studies of Harvard University 2026), and while much is known nationally about what factors are driving the nation’s crisis in housing affordability (Iqbal, Brdedthauer, and Decker 2025), little information is available about rents in the Nashville area and which demographic groups have the most difficulty paying them. This project will develop code for extracting insights from current Census data into rent affordability in the Nashville area and produce a news story package to be pitched to a Nashville news outlet.
Rent eats up 35% or more of total household income for between a third and nearly half of renters in the Nashville area’s three largest counties, the latest U.S. Census data show.
An estimated 33.9% of renters in outlying areas of Rutherford County send that much of their income to a landlord every month, as do about 37.5% of renters in the outlying areas of neighboring Williamson County.
But in more urban areas, especially in Davidson County, gross rent as a percentage of income, or GRAPI, runs 35% or more in as many as 43.7% of renting households. Most financial experts agree that, all else being equal, paying more than 30% of total household income on gross rent - meaning rent plus basic utilities - leaves too little money for other household budget essentials.
Figure 1 summarize 35-plus GRAPI estimates from the Census Bureau’s 2024 five-year American Community Survey (US Census Bureau, n.d.a) for each of the 11 Public Use Microdata Areas across Davidson, Williamson, and Rutherford counties. The 2024 five-year dataset is the latest-available dataset (US Census Bureau, n.d.b). The 2025 data likely will be released in September 2026.
Note: See live, interactive map and graphic, with code, at: https://rpubs.com/drkblake/Fig1
While informative, these data offer few insights about who these cost-burdened renters are, or what might be done make housing more affordable for them.
American Community Survey results such as those summarized in Figure 1 are designed for comparing geographic areas using aggregate measures of population characteristics. Often, such results can offer clues about the individuals within the aggregate measures. For example, if high GRAPI estimates were more common in in areas with low household incomes than in areas with high household incomes, a conclusion that high GRAPI is associated with low household income would be reasonable.
But as the error bar chart in Figure 1 shows, the estimated 35-plus GRAPI percentages are fairly uniform across the Nashville region’s sub-county areas. The error margins for these estimates, while generally small, tend to overlap, meaning that most of the estimates are statistically equivalent to one another. For example, 35-plus GRAPI estimates for western and northern Davidson County are statistically indistinguishable from each other. As a result, even though aggregate measures of traits like income, race, age, and education differ substantially for these two areas, the differences can’t help researchers understand why some renters in these areas end up using so much of their income to afford housing.
To address exactly this problem, the American Community Survey regularly releases Public Use Microdata Sample (PUMS) files shortly after the release of each aggregate ACS dataset. While a typical aggregate data file from the ACS contains one row of data per geographic area, a PUMS data file from the ACS contains one row of data per individual, with all personal identifying information about the individual excluded. Working with PUMS files allows researchers to examine individual-level associations among variables, overcoming the limitations inherent in data aggregated by geography.
The PUMS file for the 2024 five-year ACS was released on March 5, 2026. My faculty mentor, Dr. Ken Blake, data journalism professor in the School of Journalism and Strategic Media, will be developing computer code over the summer to extract GRAPI-related data from the latest PUMS file and use it to investigate predictors of high housing cost burdens among Nashville-area renters. I will help Dr. Blake develop and test the code. I also will produce the news story package and pitch it to a Nashville news outlet. I already understand fundamentals of working with and adapting R code, and I am working on a degree in journalism.
Dr. Blake and I will delve into the tidyverse package documentation (Walker 2023) to develop R code capable of extracting, analyzing and visualizing PUMS-level data on rent affordability and rent cost burden distribution among demographic segments of the Nashville-area renter population.
After helping Dr. Blake develop and test the code, I will use the code’s results to anchor a news story package that I will produce using standard journalistic interviewing and research methods.
By Friday, May 29: Plan the analysis and produce a working R script
By Friday, June 19: Produce a well-documented story pitch and a rank-ordered list of three Nashville-area news outlets to approach about publishing the story.
By Friday, July 10: Finalize a publishing arrangement with a Nashville news outlet.
By Friday, Aug. 7: Publish the story and submit all required URECA materials.
I will assist Dr. Blake with developing and testing the code. I have been learning fundamentals of data analysis with R in Dr. Blake’s JOUR 3841 Data Skills for Media Professionals course this spring. I already know how to extract, analyze and visualize U.S. Department of Housing and Urban Development Small-Area Fair Market Rent estimates for ZIP codes in the Nashville area, and how to supplement the data with estimates of rental and total household counts from the American Community Survey. I also am proficient with R Markdown, a language for publishing annotated R code and output on the web. Learning how to work with PUMS data will significantly expand my expertise with R code and Census data. Meanwhile, Dr. Blake will incorporate what we learn about PUMS data analysis into his data journalism courses and also into a book he is writing to help media students and professionals understand and use R code.
With Dr. Blake’s guidance, I also will produce the news package and pitch it to a Nashville news outlet. Possibilities include WPLN News, Nashville’s public radio news service; the Nashville Banner, a locally owned, community-supported civic news organization; or the Nashville Scene, Nashville’s alternative newsweekly. Depending on the needs of the news outlet, I may produce the story package entirely on my own or work collaboratively with one of the outlet’s reporters or producers.