| Family Size | VLI Limit |
|---|---|
| 1 | $43,500 |
| 2 | $49,700 |
| 3 | $55,900 |
| 4 | $62,100 |
U.S. Rental Assistance Eligibility and Access Across the Lifespan: A Repeated Cross-Sectional Analysis of Current Population Survey Microdata
Anthony Traver
OSU College of Social Work
Aims
Link Current Population Survey (CPS) microdata and local HUD income limits to:
Estimate the probability of rental assistance eligibility across age and time.
Estimate the probability of rental assistance access among the eligible across age and time.
Why Link?
Common data sources have limited capacity to estimate probability.
American Housing Survey
- Unit of analysis is housing units, not individuals.
- Public use data excluded necessary variables.
- Internal Use Files may provide best data for this analysis.
Comprehensive Housing Affordability Strategy
- Households tabulated at county level.
- Reports age only by presence of an older adult.
American Community Survey and Decennial Census
- Rental assistance data only available in ACS 2000 - 2003.
Data Sources and Management
CPS March Annual Social and Economic Supplement (ASEC)
Conducted jointly by the U.S. Census Bureau and the Bureau of Labor Statistics.
Includes approximately 145,000 individuals in 90,000 households (U. S. Census Bureau 2023).
Microdata accessed through IPUMS (Flood et al. 2023).
IPUMS CPS ASEC
Variables
- Age (AGE)
- Household income (HHINCOME)
- Home ownership status (OWNERSHP)
- Housing assistance (PUBHOUS and RENTSUB)
- Person-level weight (ASECWT)
- Geographic indicator (METFIPS)
Samples
- 2000, 2010, 2023
HUD Income Limits
HUD defines income limits for:
- Metropolitan areas
- Areas within some metropolitan areas
- Non-metro countries
Very Low Income (VLI) Limit:
- 50% of Area median income
- Adjusted for family size
HUD Income Limits
Minneapolis-St. Paul-Bloomington, MN-WI HUD Metro FMR Area
Data Management
CPS ASEC
All cases: 489,645
Metropolitan area cases: 389,736
Drop cases without a metropolitan FIPS code (6%)
Drop cases with age over 79 (3%) (inconsistent top codes)
Drop cases with family size over 8 (1%)
Analytic sample: 346,427
HUD
Used lagged year (effective May - May)
Take the average when multiple VLI limits exist within a metropolitan area (33%).
Data Management
Link
VLI limits linked to CPS cases by metropolitan FIPS code, family size, and year.
Defining New Variables
Very Low Income: household income below linked VLI limit
Eligible for Rental Assistance: VLI and renter
Rental Assistance Access: public housing or rental subsidy
Estimation Approach
Eligiblity Among Metropolitan Cases
Survey-Weighted Generalized Linear Model with Logistic Link Function
Age Polynomial:
\(P(VLI\ Renter) = Age + Age^2 + Age^3 + Year + (Age + Age^2 + Age^3) * Year\)
Where Year is is a factor with three levels: 2000, 2010, 2023
Age Groups:
\(P(VLI\ Renter) = Age + Year + Age * Year\)
Where Year is a factor with three levels: 2000, 2010, 2023 and Age is a factor with four levels: 0-17, 18-39, 40-61, 62-79
Housing Assistance Among VLI Renter Metropolitan Cases
Survey-Weighted Generalized Linear Model with Logistic Link Function
Age Polynomial:
\(P(Assistance) = Age + Age^2+ Year + (Age + Age^2) * Year\)
where Year is is a factor with three levels: 2000, 2010, 2023
Age Groups:
\(P(Assistance) = Age + Year + Age * Year\)
where Year is a factor with three levels: 2000, 2010, 2023 and Age is a factor with four levels: 0-17, 18-39, 40-61, 62-79
Packages
tidyverse (Wickham et al. 2019)
ipumsr (Ellis, Burk, and Roberts 2024)
survey (Lumley 2023)
emmeans (Lenth 2024)
multcomp (Hothorn, Bretz, and Westfall 2008)
Results
Weighted Metropolitan Sample
| Year | Total | VLI Renter (%) | Assistance Among VLI Renter (%) |
|---|---|---|---|
| 2000 | 206,146,125 | 12 | 26 |
| 2010 | 235,204,438 | 15 | 20 |
| 2023 | 262,261,014 | 11 | 26 |
Age Distribution of Metropolitan Sample
VLI Renter Probability
VLI Renter Probability
VLI Renter Probability
VLI Renter Probability
Difference in VLI Renter Probability Across Years
Older Adults (62 - 79)
| Years | Odds Ratio | Lower | Upper | P Value |
|---|---|---|---|---|
| 2000 - 2010 | 1.01 | 0.92 | 1.10 | 0.87 |
| 2010 - 2023 | 1.12 | 1.03 | 1.21 | 0.01 |
| 2000 - 2023 | 1.13 | 1.03 | 1.23 | 0.01 |
Housing Assistance Among VLI Renters
Housing Assistance Among VLI Renters
Housing Assistance Among VLI Renters
Housing Assistance Among VLI Renters
Difference in Housing Assistance Among the VLI Renters Across Years
Older Adults (62 - 79)
| Years | Odds Ratio | Lower | Upper | P Value |
|---|---|---|---|---|
| 2000 - 2010 | 0.96 | 0.80 | 1.15 | 0.66 |
| 2010 - 2023 | 1.19 | 1.02 | 1.39 | 0.03 |
| 2000 - 2023 | 1.14 | 0.96 | 1.36 | 0.13 |
Conclusion
Summary of Findings
The age-based probability of a metropolitan area resident being a VLI renter has become more uniform.
The age-based probability of a of a metropolitan area VLI renter receiving assistance has become less uniform, tilting towards older adults.
Implications
As housing assistance programs shift their focus to older adults:
Programs: Services within public and projected-based housing will need to become more diverse and intense to promote healthy aging for older residents.
Tenant Rights: More modifications (reasonable accommodations) will be needed for rental subsidy users living in private-market units. Such requests could spur housing insecurity.
Policy: Housing assistance should be expanded to address the diminished access among VLI renter households with children.
Limitations
Sample
Excludes all rural cases and 9% of metropolitan cases.
Measurement
Eligibility determination is often more intricate than VLI renter status.
Averaging VLI limits across metropolitan areas reduces accuracy of the data link.
Modeling
Analysis does not account for clustering at the family or geographic level.
Thank you!
References
Traver| DIRC | 2024
Ellis, Greg, Derek Burk, and Finn Roberts. 2024. Ipumsr: An r Interface for Downloading, Reading, and Handling IPUMS Data.
Flood, Sarah, Mariam King, Renae Rodgers, Steven Ruggles, J. Robert Warren, Daniel Backman, Annie Chen, et al. 2023. “Integrated Public Use Microdata Series, Current Population Survey: Version 11.0 [Dataset].” Minneapolis, MN: IPUMS. https://doi.org/https://doi.org/10.18128/D030.V11.0.
Hothorn, Torsten, Frank Bretz, and Peter Westfall. 2008. “Simultaneous Inference in General Parametric Models.” Biometrical Journal 50 (3): 346–63. https://doi.org/10.1002/bimj.200810425.
Housing, U. S. Department of, and Urban Development. 2023. “Income Limits [Database].” https://www.huduser.gov/portal/datasets/il.html.
Lenth, Russell. 2024. Emmeans: Estimated Marginal Means, Aka Least-Squares Means. https://CRAN.R-project.org/package=emmeans.
Lumley, Thomas. 2023. Survey: Analysis of Complex Survey Samples. https://cran.r-project.org/web/packages/survey/index.html.
U. S. Census Bureau. 2023. “Current Population Survey 2023 Annual Social and Economic (ASEC) Supplement.” Washington. https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar23.pdf.
Wickham, Hadley, Mara Averick, Jennifer Bryan, Winston Chang, Lucy D’Agostino McGowan, Romain François, Garrett Grolemund, et al. 2019. “Welcome to the Tidyverse.” Journal of Open Source Software 4 (43): 1686. https://doi.org/10.21105/joss.01686.