2017 Federal Rental Assistance Factsheets Sources and Methodology
March 31, 2017
To create the federal rental assistance factsheets for the 50 states, District of Columbia, and the United States, CBPP analyzed administrative data from the Department of Housing and Urban Development (HUD) and the Department of Agriculture (USDA) concerning the characteristics of households that receive federal rental assistance. Section 1 outlines the sources and methods used to create charts and figures on these programs and their recipients. CBPP also estimated the unmet need for housing assistance among all low-income households using the American Community Survey. Section 2 outlines the methods used to generate charts and figures related to housing need in each state. CBPP used the most recent data available, typically from 2015 or 2016.
Section 1: Federal Rental Assistance
Federal rental assistance encompasses USDA Rural Rental Assistance (Section 521) and the following HUD-administered programs:
- Public Housing
- Section 8 Housing Choice Vouchers
- Section 8 Project-Based Rental Assistance (including Moderate Rehabilitation)
- Supportive Housing for the Elderly (Section 202)
- Supportive Housing for People with Disabilities (Section 811)
- Rent Supplement
- Rental Assistance Program
- McKinney-Vento Permanent Supportive Housing, Safe Haven, and Transitional Housing units and beds
- Housing Opportunities for Persons with AIDS (HOPWA)
CBPP relied on two types of data to analyze HUD-assisted programs. We collected data on the number of households and federal funding levels for each program from public sources published by HUD and USDA. In addition, CBPP relied on a non-public dataset from HUD’s Office of Policy Development and Research (available through a research agreement) to analyze demographic characteristics of households using HUD rental assistance. HUD collects demographic information on recipients through Form 50058 and the Tenant Rental Assistance Certification System.
Households using federal rental assistance: This number is the sum of all households in the federal rental programs listed above, rounded down to the nearest thousand. In some instances, CBPP used unit counts if household counts were not available. Unit counts represent all units in each federally assisted property, regardless of occupancy status.
CBPP used the following publicly available sources for each program:
- Section 8 Housing Choice Voucher household counts are from HUD’s Voucher Management System quarterly reports. CBPP calculated the monthly average of voucher households in 2016.
- Public Housing, Section 8 Project-based Rental Assistance, Supportive Housing for the Elderly, Supportive Housing for Disabled Persons, Rental Assistance Program, and Rent Supplement program data are from HUD’s 2016 Picture of Subsidized Households dataset. CBPP estimated the number of households in each program by multiplying total units by the occupancy rate. Total units were substituted for households if the occupancy rate was unavailable.
- McKinney-Vento unit counts for households with children and bed counts for households without children in Permanent Supportive Housing, Transitional Housing, and Safe Havens are from HUD’s 2016 Housing Inventory Count by State.
- HOPWA household counts are from grantee performance profiles for 2015-2016.
- USDA Rural Rental Assistance household counts (Section 521) are from the USDA’s fiscal year 2016 Multi-Family Fair Housing Occupancy Report.
Some funds authorized under the Native American Housing Assistance and Self Determination Act and HOME Investment Partnership program are used by some grantees for rental assistance. However, we exclude these households because HUD does not publish detailed demographic data for households receiving rental assistance through these programs.
Share of extremely low-income households: CBPP estimated the share of extremely low-income households using the non-public 2016 HUD dataset referenced above. HUD considers a household to be extremely low income if its income is at or below 30 percent of the local median or the federal poverty line. To calculate the share, CBPP divided the number of extremely low-income households on HUD assistance in each state by the total number of households in both HUD and USDA rental assistance programs. Because detailed income data was not available for households in the USDA’s rental assistance program, we used the language “at least” to describe the share of extremely low-income households. It is likely that many households on USDA assistance also meet HUD’s definition of extremely low income: the average income for a household receiving rural rental assistance was $10,504 in fiscal year 2016.
Households who were working, recently working, or subject to work requirements: CBPP used the non-public HUD dataset to estimate this figure. We considered households to be working, recently working, or likely subject to a work requirement if the household reported wage income in 2015 or 2016; received unemployment insurance in 2016; or received Temporary Assistance for Needy Families (TANF), which typically requires recipients to work, in 2016. For this calculation, CBPP relied on HUD’s definition of wage income: earnings from any job, self-employment, or military pay. USDA does not publish detailed income statistics on households receiving rural rental assistance, making it impossible to include those households in our analysis. For more details on how we determine work status, see “Most Rental Assistance Recipients Work, Are Elderly, or Have Disabilities.”
Households living in non-metropolitan areas: CBPP used the non-public 2016 HUD dataset and the USDA’s Multi-Family Section 514 and 514 Active Projects list to determine the location of assisted households. Both datasets contain latitude and longitude coordinates for households using rental assistance. CBPP considered a household or unit to be non-metropolitan if it was located outside a metropolitan statistical area.
Pie Chart: CBPP created the categories for this pie chart based on demographic characteristics of each household’s head, co-head, or spouse. Households that are headed by a non-disabled, non-elderly adult but have an elderly or disabled member in the home — e.g., a parent with a disabled child or a family with several generations living under one roof — are categorized as “Adults with children” if appropriate, or “Other.” As a result, this chart undercounts the total number of elderly and disabled people using federal rental assistance. Percentages may not add to 100 due to rounding.
We followed HUD’s definitions for elderly (age 62 or older) and disabled. HUD considers a person disabled if he or she has one or more of the following:
- a disability as defined in Section 223 of the Social Security Act
- a physical, mental, or emotional impairment that is expected to be of long-continued and indefinite duration, substantially impedes his or her ability to live independently, and is of such a nature that such ability could be improved by more suitable housing conditions
- a developmental disability as defined in Section 102 of the Developmental Disabilities Assistance and Bill of Rights Act
- Acquired Immune Deficiency Syndrome (AIDS) or any condition that arises from the etiologic agent for AIDS
Only certain federal rental assistance programs publish full demographic data on their recipients. For this calculation, CBPP made several assumptions to adjust for the limitations of the data. We treated all households assisted by HOPWA and all households without children assisted by McKinney-Vento programs as disabled adults without children, as programs rules typically require adults to meet disability criteria to receive assistance and nearly 80 percent of disabled households in other HUD programs do not have children. We also treated all elderly USDA households as childless, as 97 percent of elderly households in HUD programs do not have children. See above for a full list of sources used in this chart.
Bar Chart – What Major Types of Federal Rental Assistance Do Families Use: This chart includes households in the largest federal rental programs; it excludes a small number of households in the Rent Supplement, Rental Assistance Program, HOPWA, and McKinney-Vento programs. See above for a list of sources used in this chart.
Federal funds for rental assistance: This figure represents the sum of federal funding for all rental assistance programs in fiscal or calendar year 2016 (depending on program funding rules), rounded to the nearest million. We substituted 2015 funding data if 2016 figures were not available. Funding data is from the following sources:
- Housing Choice Vouchers, Public Housing, Section 8 Project-based Rental Assistance, Supportive Housing for the Elderly, Supportive Housing for Disabled Persons, HOPWA, Rental Assistance Program, and Rent Supplement fiscal year 2016 funding data are from HUD’s Community Assessment Reporting Tool.
- McKinney-Vento Homeless Assistance funds for Permanent Supportive Housing, Transitional Housing, and Safe Havens are from HUD’s Continuum of Care funding reports for fiscal year 2015 and Emergency Services Grant awards by state for fiscal year 2016. A share of these dollars includes some supportive service funding.
- USDA Rural Rental Assistance funds are from the Rural Housing program FY 2016 obligations report.
Median Incomes for Low- and Extremely Low-Income Families: Median family incomes for a low-income and extremely low-income family of three are from HUD’s fiscal year 2016 Section 8 Income Limits.
Section 2: Households in Need of Rental Assistance
CBPP used the 2015 American Community Survey Public Use Microdata Sample (ACS PUMS) to determine housing need in each state. We used HUD’s definition of low income for this analysis. HUD treats households as “low income” if their cash income is 80 percent or less of the local area median income (AMI), typically set at the metropolitan area or non-metro county. Area median incomes are based on HUD’s 2016 Section 8 Income Limits.
The ACS PUMS data uses Public Use Microdata Areas (PUMAs) as the primary geographic unit. PUMAs partition each state into non-overlapping areas containing about 100,000 residents. Since PUMAs do not always correspond to county boundaries, we needed to make approximate matches between PUMS households and counties. Using the Census Bureau’s Census tract-to-PUMA and Census tract-to-county geographic relationship files, we developed a PUMA-to-zip code crosswalk, which we weighted by the share of renters earning less than $25,000 in each county in each PUMA. We used this crosswalk to generate geographical relationships between PUMAs and counties, allowing us to assign ACS household records to counties and, subsequently, to HUD’s metro and county-based 2015 area median incomes.
Households paying more than half their income for rent: CBPP used existing variables in the ACS dataset to determine whether a household paid half or more of its income for housing. CBPP used the monthly gross rent variable (GRNTP) to estimate each household’s monthly housing costs. This variable includes the household’s monthly rent plus estimated average monthly cost of utilities (electricity, gas, and water and sewer) and fuels (oil, coal, kerosene, wood, etc.) if these are the responsibility of the renter. We used the annual household income variable (HINCP), adjusted for inflation and divided by 12, to determine each household’s monthly income. This variable includes the following income sources: wages, salary, commissions, bonuses, and tips from all jobs; self-employment income (net income after business expenses) from non-farm or farm businesses, including proprietorships and partnerships; interest, dividends, net rental income, royalty income, or income from real estates and trusts; Social Security or Railroad Retirement; Supplemental Security Income; any public assistance or welfare payments from the state or locality; retirement, survivor or disability pensions; and any other regularly received income (e.g., Veterans’ payments, unemployment compensation, child support, or alimony).
Households with negative income and households that did not pay rent were excluded from the analysis. Households with zero income and renters that pay no rent but do pay monthly utilities were included in the analysis. We considered households with zero income and positive housing costs to be paying more than half their income for rent.
This number may include some households receiving federal rental assistance. The ACS data does not include information on receipt of rental assistance, making it impossible to exclude such households.
Percent change in low-income households that pay more than half their income in rent: This figure is the change in severely housing cost-burdened low-income households from 2007 to 2015.
Box – Who are these households?: These percentages are based on CBPP tabulations of the 2015 ACS, described above.
- Households were considered to have children if there was at least one member under the age of 18.
- A household was considered elderly if the head, co-head, or spouse was age 62 or over.
- A household was considered disabled if the head, co-head, or spouse met one of the ACS’s six disability criteria.
- A household was considered to be working if at least one member reported earned income in the last year. Due to data limitations, we are not able to use the broader definition of labor force attachment — households who were working, recently working, or subject to work requirements — described above.
- A household was considered to be in poverty if its income was below the 2015 federal poverty guidelines.
These categories are overlapping, meaning that the same household could fall into more than one category. The denominator is the total number of low-income households paying more than 50 percent of their income for housing costs.
Ratio of assisted households to homeless or low-income households that pay more than half their income in rent: This chart relies on CBPP tabulations of households receiving rental assistance, homeless households from HUD’s 2016 Continuum of Care Homeless Populations and Subpopulations Reports, and low-income renter households paying more than half their income for rent in the ACS, described above.
Pie Chart – Rental Assistance Falls Far Short of Need: This chart relies on CBPP tabulations of households receiving federal rental assistance and unassisted low-income renter households paying more than half their income for rent and utilities in the ACS, described above. We subtracted a small number of households receiving HUD rental assistance and still paying half their income for housing from our ACS estimates in order to estimate the number of unassisted low-income renters paying too much for rent. (The ACS data does not include information on receipt of rental assistance.) We calculated the number of assisted low-income households paying over half their income for rent using our non-public HUD administrative dataset (see Section 1). Detailed household income and rent data was not available for several small HUD programs and for USDA households.
Homeless people: The numbers of homeless people living in each state, including homeless veterans and people in families, are from HUD’s 2016 Annual Homelessness Assessment Report to Congress. These numbers are based on a single night census of people living in shelters or other areas not designed for human habitation.
Children in unstable housing: Data on the number of school-age children in unstable housing are from the Education Department’s ED Data Express tables for homeless students enrolled during school year 2014-2015. The Education Department considers children to be living in in unstable housing if they are in shelters, transitional housing, or awaiting foster care; doubled up with other families; in hotels or motels; or in places not designed for human habitation.