In support of the Federal Strategic Plan to Prevent and End Homelessness, the U.S. Department of Housing and Urban Development (HUD) is sponsoring a research effort to explore and document how public housing agencies (PHAs) currently serve and interact with homeless households. The goal of this study is to establish a baseline level of PHAs’ current engagement in serving homeless households and to better understand the current opportunities provided by PHAs that either have an explicit preference for homeless households or that target assistance to homeless households in other ways. The study will also seek to explore barriers encountered by PHAs or concerns perceived by these agencies about increasing the number of homeless households served or targeting homeless households. HUD intends to use this information to identify mechanisms to address or eliminate barriers to serving homeless households in mainstream housing assistance programs, with a focus on the housing choice voucher (HCV) program and public housing. This study will be completed by Abt Associates Inc. on behalf of HUD.
In order to assess how the preferences, set-asides, and targeted programs for homeless people are implemented by PHAs, and to identify the barriers to targeting homeless households for assistance among public housing agencies that have not adopted such a strategy, this study will use a two-part data collection process. First, HUD will conduct a census of all PHAs to determine the extent and type of activities that PHAs are using to serve homeless households. The census will provide information about PHAs that use preferences or other approaches to serve homeless households and about PHAs that do not make special efforts to serve homeless households. The second part of data collection will involve a follow-up telephone survey with a purposive sample of 125 PHAs. This in-depth phase will allow HUD to gather more information on several issues that can only be touched upon in the census survey, including why (or why not) a PHA has implemented preferences for admission of homeless applicants and a further exploration of the barriers a PHA may perceive in targeting homeless households for assistance. The findings of the census survey will inform the sampling approach to be used to select the sample of the follow up telephone survey.
This Submission covers only the first part of the data collection effort, the census instrument. A modification to this Submission will be submitted at a later date that will include a detailed description, and seek approval for the use of, the follow-up telephone survey.
The Department of Housing and Urban Development’s (HUD) Study of Public Housing Agencies' Engagement with Homeless Households seeks to examine the extent to which public housing agencies (PHAs) have developed and implemented preferences for homeless households. The public housing and Housing Choice Voucher (HCV) programs administered by local PHAs are critical housing resources for low-income households across the country. However, because these resources are in such high demand, there are often waiting lists that are years in length or are closed to new applications altogether. To address this shortage relative to need, some PHAs have enacted preferences or set-asides for homeless households, who are the most vulnerable among the low-income population.
This study will:
examine how many PHAs have adopted preferences and set-asides in their Housing Choice Voucher and public housing programs, to help homeless households access housing assistance more quickly and use it to prevent future homelessness;
document PHA administration of other programs that serve homeless households, including HUD McKinney-Vento programs and allocations of resources from additional mainstream programs, such as HOME, that may be targeted to homeless households;
explore how the preferences, set-asides, and targeted programs for homeless households are implemented; and
identify the barriers to targeting homeless households for assistance among public housing agencies that have not adopted such a strategy.
The housing available through PHA programs such as public housing and HCV falls far short of the existing level of need in many communities. Most PHAs must place households that apply for one of their housing programs on a waiting list that may span years in length. Prior to 1998, the Department of Housing and Urban Development (HUD) mandated that PHAs target their available public housing units or Housing Choice Vouchers to applicants with the most serious housing issues. These included households that were displaced by government action, living in severely substandard housing or homeless, and households with an unsustainable rent burden (paying more than half of their income for rent). Homeless applicants were given priority for admission because they were deemed to be living in substandard housing. However, in 1998 the federal government rescinded the mandated federal preferences and replaced them with broad PHA discretion to establish preferences based on their own local conditions and policy priorities.
PHAs currently enact a variety of preference systems. Some PHAs continue to provide preferences to homeless households within a framework similar to the former federal preferences. Others have shifted to a first-come, first-served system. Many PHAs give preference to local residents, and some give priority to veterans, to people with disabilities, to homeless people, or to other groups. PHA preference systems vary by level of complexity and explicit target populations; some work in favor of homeless households and some to their disadvantage.
PHAs must complete an Annual PHA Plan that describes the agency's overall mission and plan for serving low-income and very low-income families, including any local preferences for selecting applicants from their waiting lists. While HUD reviews all PHA plans to ensure that the preferences chosen by the PHA do not exceed the discretion given to them, there exists no central database of the preferences used by PHAs nationally. PHA Plans could be reviewed for information on preferences, but they are difficult to search electronically, and most of the plans do not provide a level of detail sufficient to capture the complexity of PHA preference systems or to discover whether a PHA has implemented set-asides of units for the full range of programs that could serve homeless households.
There are approximately 1.2 million public housing units and 2 million housing choice vouchers administered by over 4,000 PHAs nationwide. To identify the extent to which these 3.2 million housing “units” intentionally serve homeless households, HUD will conduct a census of all 4,085 PHAs. This section details the approach HUD will take to accomplishing the national census of PHAs.
Since the termination of federal preferences for homeless households, great variation has developed in approaches to prioritization of access to PHA housing resources. The first goal of this study is to obtain a clear understanding of the extent and nature of PHA engagement with homeless households, including whether vouchers and/or public housing units are set aside and whether PHAs administer other programs that serve homeless households. To do this, HUD will conduct a census of PHAs, developing a comprehensive description of PHA preferences nationally. By defining "preference" broadly, to make it inclusive of set-asides of vouchers and public housing units and targeting of other programs PHAs may administer, the census can take a snapshot of the current engagement of PHAs with homeless households. Specifically, the PHA census aims to answer the following research question: to what extent are PHAs serving homeless households?
The objective of this question is to explore whether PHAs are prioritizing homeless households through preferences or set-asides within their public housing or HCV programs. A number of sub-questions will be asked to identify the types of preferences and set asides that exist for homeless households and other special populations. In addition, HUD will document how a PHA defines homelessness and whether the PHA operates any other programs besides the HCV program and public housing that serve homeless households.
HUD, the primary beneficiary of the planned data collection, will use the information from the study to enhance its understanding of the extent and type of activities PHAs are using to serve homeless households, as well as the barriers reported by PHAs that may be preventing some agencies from targeting homeless households. The census will provide information about PHAs that use preferences or other approaches to serve homeless households and PHAs that do not make special efforts to serve homeless households. Through the study, PHAs themselves will learn about different approaches to assisting homeless households. The information collected here may also be of interest to community agencies that serve homeless populations, researchers, and local policy makers seeking to address homelessness in their communities.
The census instrument is a web-based survey that will be released to the executive directors of all 4,085 PHAs nationwide. This self-administered survey consists mainly of closed-ended questions based on the research question summarized in Section A.2.2 above. There are a few questions where the respondent is asked to type in words that add detail to a response of "other." The instrument was designed so that it is clear and easily used by PHAs that administer public housing alone, housing choice vouchers alone, or that administer both programs.
The web survey includes questions that confirm the size of the PHA and the types of programs it administers, and questions that ask about existing preferences and set-asides for homeless households and how they fit into the PHA's overall preference system. A draft of the census instrument is included as Appendix A.
The survey also includes questions about other programs administered through PHAs. Many of these agencies use other programs and resources to target and assist homeless households, including Section 811, permanent supportive housing funded by HUD's McKinney-Vento programs (such as Shelter Plus Care), HOME, and Homeless Prevention and Rapid Re-housing (HPRP). Some PHAs have received HOME TBRA funding or rental assistance funded from state or local tax revenue to provide transitional housing to homeless families, particularly around the concept of self-sufficiency and often in coordination with TANF. Information on whether PHAs administer these programs will help establish how engaged the PHA is in preventing and ending homelessness in the community already, and could inform the PHAs chosen for the more intensive follow-up survey.
There are a total of seven sections to the census instrument:
Section A includes questions about Housing Choice Vouchers and Project-Based Vouchers;
Section B includes questions about Public Housing;
Section C includes questions about other programs operated by PHAs;.
Section D includes questions about how the PHA identifies applicant households as homeless;
Section E includes questions about partnerships between the PHA and community service providers serving homeless people;
Section F includes questions about barriers perceived by the PHA in engaging with homeless households; and
Section G includes questions for Moving to Work (MTW) agencies only and seeks to understand how their discretionary authority under MTW may (or may not) be used to assist homeless households.
Appendix B contains the item-by-item justification for all items in each section of the PHA census.
Improved information technology will be used in this study in two distinct ways:
To pre-fill some data on the census using HUD administrative data; and
To facilitate collection of the data in standardized and accurate way.
The study will generate a substantial amount of data on the universe of PHAs. Where possible, data from existing HUD databases will be pre-filled into the instrument, to avoid asking questions for which data are already available. Improved technologies will also support a receipt control system, containing PHA information such as agency size and geographic location; this will allow us to generate frequent status reports on response by different categories of respondents. The receipt control system will enable the study team to gauge progress in relation to achieving full response from different types of PHAs and to identify the agencies that require telephone follow-up.
It is also important to note that the web-based technology that will be used to collect information from PHA staff for the PHA census reduces respondent burden, as it allows respondents to complete the census in stages, at their own pace. It also allows for multiple responders–important for this study because in many PHAs more than one administrator may need to contribute to the census responses. Further, web technology allows respondents to proceed more quickly and accurately through the survey instruments, minimizing the interview length and the need for subsequent call-back. Web technology also improves data quality through more uniform administration of the survey questions, more accurate implementation of the skip patterns, and immediate application of range checks, edit checks, and consistency checks of item-by-item responses. For example, agencies that do not operate an HCV program will be automatically routed past the screens with questions about the HCV program.
In order to avoid duplicate data collection and data entry, the study team will use HUD administrative data regarding which housing programs are administered by each PHA, coupled with information on program size. Rather than ask PHA staff to report these data items, the survey will instead instruct the respondent to review and confirm that the pre-populated administrative data are correct. This minimizes burden on the PHA staff and allows them to correct information if needed.
The research team may work with small community agencies that administer HCV programs. Given that a large proportion of housing agencies are small in size, the research team has designed the approach with these agencies in mind. The research team will take great care to ensure minimal burden of participating in this study.
The census survey will be administered a single time. HUD intends to use the data collected through this survey to identify mechanisms to address or eliminate barriers to serving homeless households in mainstream housing assistance programs, with a focus on the housing choice voucher (HCV) program and public housing.
PHAs must complete an Annual PHA Plan that describes the agency's overall mission and plan for serving low-income and very low-income families, including any local preferences for selecting applicants from their waiting lists. While HUD reviews all PHA plans to ensure that the preferences chosen by the PHA do not exceed the discretion given to them, there exists no central database of the preferences used by PHAs nationally. PHA Plans could be reviewed for information on preferences, but they are difficult to search electronically, and most of the plans do not provide a level of detail sufficient to capture the complexity of PHA preference systems or to discover whether a PHA has implemented set-asides of units for the full range of programs that could serve homeless people. Further, an analysis of the Plans would not provide any insight into the barriers that may be preventing PHAs from targeting housing assistance resources to homeless households. Therefore, this data collection from the universe of PHAs is critical to help inform HUD about the various ways in which PHAs are engaging with homeless households, or the agencies that serve them. The census data will be provided to HUD in electronic format—making future searches and analysis feasible.
The proposed data collection activities are consistent with the guidelines set forth in 5 CFR 1320.6
(Controlling Paperwork Burden on the Public, General Information Collection Guidelines). There are no circumstances that require deviation from these guidelines.
In accordance with the Paperwork Reduction Act of 1995, the Department of Housing and Urban Development (HUD) published a notice in the Federal Register on November 14, 2011. The docket number is FR-5486-N-24. The Federal Register Notice appeared on pages 64367 and 64368. A copy of the Federal Register Notice is included in Appendix C.
The instrument was reviewed by staff from several departments within HUD and one external agency. The specific departments within HUD included the Offices of: Public and Indian Housing (Housing Choice Voucher and Public Housing Operations); Community Planning and Development (SNAPS – Office of Special Needs Assistance Programs); and Policy Development & Research. HUD also sought feedback from the U.S. Interagency Council on Homelessness. In addition, HUD reached out to several industry organizations, including the National Association of Housing and Redevelopment Officials (NAHRO), Council of Large Public Housing Agencies (CLPHA) and Public Housing Authorities Directors Association (PHADA), for feedback and input on the instrument.
HUD received two sets of public comments in
response to Federal Register Notice FR-5486-N-24. Comments were
submitted by the HCV Program Manager of the Brazos Valley Council of
Governments public housing agency and a Senior
Project Manager with Shelter Partnership, Inc., a non-profit
organization located in Los Angeles, CA which operates numerous
homeless assistance programs. Both sets of comment were excellent
and provided sound suggestions to the research team for improvements
to the survey instrument. The suggestions submitted via public
comment were incorporated into the final survey instrument, with the
exception of one comment that suggested that HUD add an additional
component to the web survey regarding cost. HUD chose not to address
this topic in the web survey due to time constraints, however,
questions related to the costs of targeted efforts to engage homeless
households will be addressed in the follow-up sample survey which
will be a longer and more detailed survey instrument.
The respondents to the PHA web-based census will not receive payments for completing the survey because the response burden is minimal and the respondents are agency staff.
Because of the nature of the information collected from study participants, and more significantly the method being used (via the internet) to gather information, strict confidentiality procedures will be followed for this evaluation. The information requested under this collection will be used for research purposes only and will not be used for compliance monitoring.
Prior to the start of the PHA web-based census, the study protocols will be reviewed in detail by Abt Associates’ Institutional Review Board (IRB). The IRB review process will review all study protocols, including the consent process and confidentiality protections. Detailed plans for data security procedures are described below.
The research team has established rigorous data security and confidentiality provisions that are documented as part of the study’s Data Security Plan. First, all data users will be made aware of and trained on their responsibilities to protect agency information, including the limitations on uses and disclosures of data. The research database will be designed to limit access to data to authorized users, with levels of access commensurate with each person’s role on the project.
The web survey platform allows for tight control over respondent recruitment and data acquisition while addressing both data security and confidentiality concerns. Respondents will access the PHA census through a secure web site, where they are protected by a strict data security system. Only those given the seven-digit personal identification numbers (PIN) can access the survey. Upon entering the PIN, the respondent moves to a non-public directory inaccessible through the Internet. As data are entered, they are stored on a second non-public directory accessible only to the system administrator. Partial responses are saved in this way. Once respondents finish the census and press the "Submit" button on the screen, the ID number used to access the survey becomes invalid and the instrument cannot be accessed again with that number.
Every precaution will be taken to ensure that data collected on the Internet remain both secure and confidential. All data collection servers are housed in an AT&T Network Operations Center (NOC) with redundant power, expandable bandwidth, and a high level of physical security.
All study staff are required to sign a confidentiality pledge stating that no data will be released to unauthorized personnel. In addition, all electronic data for the study are stored on strong, encrypted, password-protected servers and accessible only by authorized research and programming staff.
The web-based census of Public Housing Agencies does not include any questions of a sensitive nature.
The study’s intention is to gather data from all PHAs, as described in Section A.2.1 above. All 4,085 PHAs will be invited to participate, and follow-up will be conducted to maximize the response.
Exhibit A.12.1 shows the respondent burden with 100 percent participation to be 1,362 hours. However, it is more likely that around 80 percent of the PHAs will actually complete the census, in whole or in part. The total burden for the study is 1,089 hours, based on an 80 percent response to the PHA Census.
Exhibit A.12.1 Burden of Hours
Respondents |
Projected Response |
Number of Entities |
Responses per Entity |
Average Time to Complete in Minutes |
Burden (hours) |
Maximum burden based on 100% response |
100% |
4,085 |
1 |
20 min. |
1,362 |
Likely burden based on projected 80% response |
80% |
3,268 |
1 |
20 min. |
1,089 |
This data collection effort involves no recordkeeping or reporting costs for respondents, other than the time burden to respond to questions on the data collection instruments as described in item A.12 above. There is no known cost burden to the respondents.
The total contract amount for this research effort is $251,078.
This submission to OMB is a new request for approval; there is no change in burden.
This information collection request focuses on a web-based survey census of all PHAs—those that operate a Housing Choice Voucher program, a public housing program, and/or other housing programs directed toward homeless households. Data will be collected through a web-based survey from the universe of PHAs (4,085 individual PHAs). The specific use of the data collected through this PHA census is described below.
Following data cleaning and production of an analysis file of the web census data by Abt SRBI, the research team will produce frequencies and descriptive statistics on the answers to each of the survey questions, including percentages of all PHAs selecting each response category and percentages selecting the response category among PHAs to whom the question pertained (i.e., adjusted for skip patterns). For numerical answers, the team will produce means, medians, and distributions.
After reviewing those tables, the research team will define categories and data filtering rules in order to produce cross tabulations that provide a comprehensive picture of PHAs’ engagements with homeless households. The analysis will show the percentage of all PHAs that report that they have:
Set-asides of HCVs or Public Housing: the PHA designates a specific number of vouchers or public housing units for homeless households.
Waiting List Preferences: the PHA’s policy gives homeless households priority for tenant selection from waiting lists
Other Programs Serving Homeless People: the PHA administers rental assistance through programs other than vouchers or public housing that benefit homeless households (such as Shelter Plus Care, HOME, or CDBG).
Within each category, cross-tabulations will provide additional detail on the sizes of set-asides, the type of preference (e.g., does a homeless household go to the top of the waiting list or compete with other preferences given equal weight?), whether set-asides or preferences are for HCVs or public housing, and the numbers of PHAs reporting that they serve homeless households through particular “other” programs.
The analysis then will cross-tabulate the basic results on PHAs that do/do not have particular types of engagements with homeless households against key characteristics of PHAs and their service environments, such as:
Census regions and sub-regions
Type of service area (MSA-central city, MSA-outside of central city, non-MSA)
PHAs with HCV program vouchers only, public housing only, and both HCVs and public housing
Type of PHA (local/regional/statewide)
Size of the PHA (sum of HCV program vouchers and public housing units)
Size of the PHA’s HCV program
Size of the PHA’s public housing program
Numbers of homeless people reported by the Continuum of Care
Rates of homelessness reported by the CoC (homeless people/total population)
The analysis may also characterize the PHAs by type of service population—e.g., elderly-only, high percentage of elderly, high percentage of people with disability—based on HUD administrative data or public use data such as Pictures of Subsidized Households. (Answers to the census will tell us only whether this is a PHA that serves exclusively elderly people—e.g., its entire program is a single public housing development that has been designated for elderly use.)
The research team will also explore the feasibility of characterizing PHAs by the extent of competition for their housing assistance slots. For example, this could be done using a question on the survey that asks for the size of the PHA’s waiting list for HCVs, or it could be measured as a ratio between the size of the PHA’s program(s) and the number of poor renters in the metropolitan area (or in the non-metropolitan county).
Exhibit A.16.1 provides one of many possible examples of cross-tabulations that might be produced for the analysis. The exhibit shows primary categories of preferences based on the responses to the census along the left-hand column and a categorization of the homeless population in the CoC that includes the PHA in the remaining columns.
The analysis will also provide cross-tabulations that report on how different types of PHAs define homelessness in the context of their preferences for homeless people—that is, whether they define homelessness as on the street or in shelter or in some other way.
Exhibit A.16.1 PHA Engagement with Homeless Households by Size of Homeless Population
Types of Preference (HCV or public housing) |
Number of homeless people in the Continuum of Care that Includes the PHA’s Service Area |
|||
1 to 250 |
251 to 1,500 |
1,501 to 5,000 |
>5,000 |
|
Set asides |
||||
No set-aside for homeless |
% |
% |
% |
% |
Set aside of 1-25 units |
% |
% |
% |
% |
Set-aside of 26-100 units |
% |
% |
% |
% |
Set-aside greater than 100 units |
% |
% |
% |
% |
Waiting list preferences |
||||
Strong preference for homeless |
% |
% |
% |
% |
Moderate preference for homeless |
% |
% |
% |
% |
No preference for homeless |
% |
% |
% |
% |
Other Programs |
||||
Shelter Plus Care |
% |
% |
% |
% |
HOME Tenant-based Assistance |
% |
% |
% |
% |
SRO Mod Rehab |
% |
% |
% |
% |
Other programs |
% |
% |
% |
% |
Total |
% |
% |
% |
% |
Under the current schedule, the PHA census will span a 13-week period. Upon receipt of the census data, analysis and reporting will occur over an 11-week period. The data analysis will produce numerous cross-tabulations from which the research team will select those that together provide a description of the way in which PHAs are engaging with homeless households. The goal is to provide the tabulations that will be most informative and interesting to federal policy-makers and to those at other levels of government, including the PHAs themselves. The research team will then prepare a draft of the Sketch of PHA Preferences and Activities Serving Homeless Households and submit the draft to HUD for review. The “Sketch” will be a set of exhibits with brief narrative explanations. HUD will have an opportunity to comment on the draft table shells and on draft tables and to suggest additional cross-tabulations and different ways of displaying the data. The data analysis and reporting will conclude 24 weeks after the data collection is initiated.
All data collection instruments will prominently display the expiration date for OMB approval.
This submission describing data collection requests no exceptions to the Certificate for Paperwork Reduction Act (5 CRF 1320.9).
A.
Justification ▌pg.
| File Type | application/vnd.openxmlformats-officedocument.wordprocessingml.document |
| File Title | Abt Single-Sided Body Template |
| Author | Nichole Fiore |
| File Modified | 0000-00-00 |
| File Created | 2021-01-31 |