Multidisciplinary Research Team (MDRT)

This report has two key components. The first component is a descriptive analysis of housing market factors in communities with different trends in their unsheltered homeless counts. Consistent with previous work, the authors found that communities with increasing trends had tighter housing markets. The second component provided case studies of efforts to address unsheltered homelessness in three communities – Richmond, VA; Montgomery County, MD; and San Diego, CA.

This report merges FHA’s single-family mortgage insurance portfolio—section 203(b) program assisted mortgages only—with National Flood Insurance Program (NFIP) records and flood maps from the Federal Emergency Management Agency (FEMA) by property in North Carolina and Florida. The results show many FHA-insured mortgages are located in high-risk flood zones—Special Flood Hazard Areas (SFHA). Even though program regulations require nearly all properties located in such high-risk flood zones to be covered by flood insurance, the data indicate that a substantial share of these loans do not have or do not continuously maintain such insurance coverage. However, it should be noted that some of these properties may not require mandatory flood insurance due to property improvements, such as elevation, and could be in compliance with FHA’s regulations without carrying flood insurance. Further, HUD does not regularly gather information about the status of each property’s flood insurance coverage. The analysis was made possible only through a special data sharing agreement with FEMA that allowed HUD to gain access to a single point-in-time dataset of flood insurance policies.

This study develops and explores four metrics that could potentially support HUD’s goal of increasing economic opportunity for renter households assisted by public housing, assisted multifamily housing, and Housing Choice Voucher programs. The proposed metrics apply to work-able individuals and households in each program and address aspects of economic opportunity, self-sufficiency, and financial stability. Two of the metrics use comparisons with local job markets to provide context for assessing work participation of assisted tenants. The study estimates metropolitan-level values for the three programs and analyzes the variations using quartile means. This preliminary analysis reveals substantial variation in tenant levels of work participation relative to their metropolitan areas or non-metropolitan portions of the state. The work suggests that HUD could use estimates from these or similar metrics to target interventions or technical assistance toward lagging metropolitan areas or housing providers, thereby reducing disparities in economic opportunity and improving average levels.

This report investigates heterogeneity in community-level rates of sheltered and unsheltered homelessness, separately and combined, and provides insight into underlying community-level factors associated with homelessness across the United States. This study (1) identifies and describes market variables associated with sheltered and unsheltered homelessness, (2) constructs and evaluates empirical models of community-level homelessness, and (3) analyzes relationships within subgroup populations of local markets. Findings provide insights into predicting homelessness across different community types and market factors to consider as policy interventions are developed. The study finds that housing factors, such as rental costs, crowding, and evictions, are most consistently associated with higher rates of community-level homelessness. This demonstrates that housing market dynamics and the availability of affordable housing are closely tied to homelessness at the Continuum of Care (CoC) level even when controlling for a range of economic, demographic, safety net, and climate factors.

This work explores ways to refine HUD’s Fair Market Rent (FMR) methodology, particularly as it relates to calculating a trend factor. Phase I of the study presents a statistical approach for deriving local trend factors for selected areas where local Consumer Price Index (CPI) data is available for use in the calculation of FMRs by expanding on HUD’s existing methodology for estimating a national trend factor and applying it to CPI data for the 13 metropolitan and 4 regional areas of two different size classes produced by the Bureau of Labor Statistics. Phase II of the report presents alternative approaches to augmenting local market conditions into calculations of FMRs for a broader range of metropolitan areas by utilizing alternative sources of data and empirical frameworks.

The primary objectives of this study are: 1) to provide insights from Public Housing Authority (PHA) staff on the factors associated with landlord decisions about whether to participate in the Housing Choice Voucher (HCV) program; and 2) to identify a collection of promising and innovative practices that PHAs have used to increase landlord participation. The study provides key insights into landlord participation in the HCV program and the perspectives of PHA staff on factors influencing landlord decisions on whether to participate. The study also identifies a diverse collection of innovative activities adopted by PHAs to mitigate financial concerns among landlords, make the HCV program simpler, and alleviate landlord concerns about HCV tenants. The study finds that a majority of PHA staff interviewed identified financial reasons as the most important factor affecting landlord participation - with payment standards and fair market rents, damage costs and security deposits, and profit motivations cited as key determining factors.

This study examined the length of time that households participate in assisted housing programs administered by HUD. The research uses HUD administrative data from 1995 through 2015 to examine program and household variations and trends in length of stay and factors affecting length of program participation. The typical household in assisted housing stays for about 6 years, but the average length of stay varies by household type. Elderly households stay about 9 years and non-elderly families with children stay about 4 years. The average length of stay in assisted housing is generally increasing for most cohorts of assisted households, influenced by factors such as household characteristics and market conditions.

This report examines trends in housing cost burden for Housing Choice Voucher (HCV) participants between the years of 2003 and 2015. Cross-sectional data in each of these years is analyzed and a cohort analysis is conducted of those participants who initially leased a unit in 2003 or 2008. The study finds that housing cost burdens for HCV participants have risen since 2003, and the year-to-year changes in housing cost burden roughly approximate trends in the recent housing market cycle and that housing cost burdens have been particularly high for those earning the lowest incomes.

The goal of this study was to develop a housing quality index using data from the 2011 and 2013 American Housing Survey (AHS), and apply indices to compare both assisted housing and comparable unassisted housing stock. Because there is no consensus about the features that should be included in the definition of a dwelling’s quality and how each should be weighted in determining overall quality, three alternative indices of housing quality are examined: market value, consumer rating, and normative standards. This study uses the consumer rating index and normative index for the housing quality analysis. Consistent with previous AHS housing quality research, the prevalence rate is low for housing quality problems of almost all types. The quality of assisted housing is comparable to the quality of unassisted housing. Additionally, the proportion of housing problems persisting over two years, 2011 and 2013, is very low, and repairs appear to be made promptly.

This study updates the HUD report Multifamily Properties: Opting In, Opting Out and Remaining Affordable, prepared by Econometrica and Abt Associates in 2006. The original report examined the loss of affordable housing units associated with HUD’s Section 8 project-based rental assistance and Section 236 and 221(d)(3) subsidized mortgage programs. The authors analyzed property-level data to identify the physical, financial, location, ownership, and tenant characteristics associated with opt-outs from rental assistance programs, prepayment of subsidized mortgages, and enforcement actions by HUD.

This report examines the characteristics of federally-assisted housing designated for disabled households living in the United States.

In 2011, the American Housing Survey (AHS) added a topical module on accessibility. The module asked about the presence of accessibility features in housing units, including wheelchair accessibility features, and whether the accessibility features were used or not. Together, these data provided an opportunity to examine the accessibility of the U.S. housing stock and to ask whether people with disabilities reside in accessible homes. This report presents summary measures of housing accessibility based on the 2011 American Housing Survey (AHS). To develop these summary measures, researchers examined U.S. and international standards and regulations regarding housing accessibility, reviewed the relevant literature, and conducted interviews with a set of disability and housing design experts. Based on these summary measures, the report describes how accessibility varies by housing market characteristics as well as resident characteristics such as age, disability status, and income. The report also presents evidence on the relationship between the need for and availability of accessible housing units, taking affordability of accessible units into account.