Homelessness Data in Health and Human Services Mainstream Programs

Final Report

Winter 2009

Prepared for:Office of Human Services PolicyOffice of the Assistant Secretary for Planning and EvaluationHealth Resources and Services AdministrationU.S. Department of Health and Human Services

Prepared by: Michelle Wood, Lauren Dunton, Brooke Spellman, Michelle Abbenante, and John Griffith Abt Associates Inc.

Under Contract No: 233-02-008806TK004

This report is available on the Internet at:http://aspe.hhs.gov/hsp/09/HomelessnessData/

Printer friendly version in PDF format (164 pages) [Free PDF reader ]

Contents

Executive Summary

Acknowledgements

Chapters

1. Introduction
   1. Background
   2. Overview of the TANF and Medicaid Programs
   3. Study Activities
      • Initial Application Collection and Review
      • Survey Implementation
      • Analysis of Survey Data
      • Research on Other Mainstream Programs
   4. Organization of the Report
2. Homelessness Data Collected from TANF and Medicaid Applicants
   1. Types of Applications Used in the States
   2. Collection of Housing Status, Homelessness Indicators, and Homelessness Risk Factors
   3. Housing Status Information
   4. Homelessness Indicators
   5. Homelessness Risk Factors
   6. Application Process
      • Updates to the Medicaid and TANF Applications
      • Completion of Medicaid and TANF Applications
3. Use of Homelessness Data
   1. Defining Homelessness
   2. Use of Homelessness Data
4. Potential for Using Homelessness Data to Measure Access to TANF and Medicaid for Homeless Individuals
   1. Reasons for Collecting Homelessness Data
   2. Reasons for not Collecting Homelessness Data
   3. Analyzing Homelessness Data Collected Through Medicaid and TANF Applications
      • New York
      • California
5. Investigating Homelessness Data Collection in Other Mainstream Programs
   1. Methodology
   2. Data on Homelessness Status
      • National Client-level Data
      • Aggregate Data on Number of Homeless Beneficiaries
      • No Data on Homelessness
   3. Approaches to Estimate the Number of Homeless Beneficiaries
      1. Matching Program Data with Other Administrative Data Sets
      2. Conducting Research with a Sample of Program Beneficiaries
   4. Summary
6. Summary

References

Appendices

1. State Descriptions
2. Medicaid and TANF Program Names
3. Interview Procedures and TANF and Medicaid Survey Instruments
4. Use of Homelessness Data Collected in TANF and Medicaid Applications
5. List of Program Contacts HHS Mainstream Program Profiles for Nine Mainstream Programs Reviewed Under Task 5

Endnotes

Acknowledgments

The authors of this report wish to acknowledge the assistance provided to the study by many individuals.  We are especially grateful to the Medicaid and TANF officials in the 50 states and DC who participated in the interviews and shared their data collection practices and insights.  We are also grateful to administrators of the nine other HHS mainstream programs included in this exploratory study for their assistance in understanding the data collection practices in place regarding homelessness. These additional programs are: Community Services Block Grant (CSBG); Head Start; Social Services Block Grant (SSBG); State Childrens Health Insurance Program (SCHIP); the Health Center Program; Maternal and Child Health Services Title V Block Grant (Title V);Ryan White HIV/AIDS Program (Ryan White); Community Mental Health Services Block Grant (CMHSBG); and Substance Abuse Prevention and Treatment Block Grant (SAPTBG).

At Abt Associates, a number of staff members have played important roles in designing the study and in collecting, assembling, and analyzing the interview data.  Dr. Howard Rolston, as the studys Project Quality Advisor provided thoughtful and constructive comments on early versions of this report and has provided technical review throughout all phases of the research. Michelle Abbenante, Lauren Dunton, Tara Smith, and Caryn Nagler conducted the telephone interviews.  Carissa Climaco assisted with the development of data entry protocols and John Griffith and Bulbul Kaul assembled the SAS datasets, cleaned and verified the data, and prepared the tables presented in this report.  Daniel Smith, Charmayne Walker, and Missy Robinson did an excellent job producing this document.

Additional information regarding this project can be obtained from the Federal Project Officers: Flavio Menascé (202-260-0384, Flavio.Menasce@hhs.gov), Anne Fletcher (202-690-5739, Anne.Fletcher@hhs.gov) and Lynnette Araki (301-443-6204, LAraki@hrsa.gov).

[ Go to Contents ]

Executive Summary

In 2006, the U.S. Department of Health and Human Services (HHS) undertook a study to explore the extent to which states collect data on housing status and homelessness from applicants for Medicaid and Temporary Assistance for Needy Families (TANF), the two largest HHS mainstream programs that may serve individuals or families experiencing homelessness.  The Office of the Assistant Secretary for Planning and Evaluation (ASPE) managed the project in partnership with the Health Resources and Services Administration (HRSA).  Abt Associates Inc. was selected as the Contractor to conduct the study.  This project is part of a larger Departmental effort to increase access to HHS mainstream resources for persons experiencing homelessness.  Specifically, the study begins the exploration of available data that could be used to identify the number of homeless persons currently accessing HHS mainstream programs.

The main activity of the study was to design and administer a survey to all 50 states and the District of Columbia to document which states collect data on housing status and homelessness from individuals and families applying for Medicaid and TANF.  The focus of the survey was on state practices related to the collection of housing status and homelessness data from program applicants, not the collection and/or analysis of state data.  The study also included a review of practices in nine other HHS mainstream programs related to homelessness data collection.

Given the challenges that exist in developing baseline data on the number of homeless individuals or families served in mainstream HHS programs, this study focuses primarily on two of the eleven identified mainstream programs.  TANF and Medicaid were selected for study not only because they are the two largest mainstream programs relevant to homelessness in terms of budgets, but also because both programs have individual or family eligibility criteria that determine access to enrollment and both programs require the completion of a state-administered application.  While not a federal requirement, some states do collect data on housing status on state Medicaid and/or TANF application forms.

Between February and April 2008, interviewers from Abt Associates Inc. completed interviews with TANF and Medicaid directors in the 50 states and the District of Columbia.  In total, Abt Associates conducted 70 interviews with program representatives.  Of the completed interviews, 30 were with TANF program representatives and 40 were with Medicaid program representatives.  It was possible to collect information about all states with fewer than 102 interviews (one for each state and DC with Medicaid and TANF representatives) because of the prevalence of combined applications that covered both programs.  Combined applications are used by many states, and include multiple benefit programs, including TANF and/or Medicaid, on a single application.  In each state, for both Medicaid and TANF, the same application is used throughout the state.  This is true even in states where the TANF program is administered at the county level.

Findings

Collection of Housing Status, Homelessness Indicators, and Homelessness Risk Factors

The primary goal of the study was to ascertain whether housing or homelessness data is collected from TANF and Medicaid program applicants by the states administering the programs.  For those states that do collect such data, a second objective was to learn about the specific data collected and how the information is used by the state, as well as the incentive for and cost of collecting the data.

The study found that states collect three general categories of information related to housing status and homelessness on TANF and Medicaid applications:

• general information on housing status
• indicators of homelessness
• risk factors often associated with homelessness (these are not indications of actual homelessness but may identify precarious housing situations that can lead to homelessness).

Housing Status Items

The states collect a variety of general housing status items in seven main categories:

• Home address;
• Mailing address;
• Directions to home;
• Residence in public or subsidized housing;
• Intention to remain in state;
• Residence in long-term care facility; and
• Residence in medical or rehabilitation facility. 

All states ask for home address on their TANF, Medicaid, or combined applications, and most also ask for mailing address.  Only one state (Michigan) includes all seven of these housing status items on their combined application.  Less frequently asked questions about housing status deal with residence in long-term care facilities or a medical or rehabilitation facility.  Several states report including other types of housing status items, such as alternative mailing address, amount spent on housing each month, residence in assisted living facility, group home, previous address, interest in receiving housing assistance, receipt of housing assistance, type of housing situation (rent or own home), adult foster care, residence in a group home or half-way house, reason moved to state, and incarceration.

These housing status items by themselves do not provide evidence of homelessness or risk of homelessness.  However, in some cases housing status is explored in greater detail during interviews with case workers during the application process, and homelessness or housing instability may be ascertained, even if not formally recorded on the application. Understanding the types of housing status items included on the application is a good first step in identifying states that may have potential for measuring homelessness among program applicants.

Homelessness Indicators

The study found four questions included on TANF, Medicaid, and combined applications that constitute indicators of homelessness.  These are:

• Are you homeless?
• Do you reside in a shelter?
• Are you staying in a domestic violence shelter? and
• Do you have a permanent home? 

The study found that more than half (28) of states collect at least one of these homelessness indicators on their TANF, Medicaid, or combined applications.  In addition, the District of Columbia, North Carolina (TANF), and Wyoming stated that they expect to add a question Are you homeless? to their applications sometime during 2008.  Arizona, which currently asks if applicants reside in a shelter, plans to add Are you homeless? to their combined application before the end of 2008.

Homelessness Risk Factors

We also found that some states include two items on their TANF, Medicaid, and combined applications that do not explicitly indicate homelessness, but may indicate risk factors for homelessness.  These two items are:

• Do you live with friends or relatives? and
• Do you have an eviction notice?

Thirteen states collect at least one of these items.  Most states that collect data on one of the homelessness risk factors also collect information on the homelessness indicators discussed above.  The exceptions are Idaho and Mississippi (Medicaid application).  In these states, information on one of the risk factors for homelessness is collected, but none of the homelessness indicators are included on the application.

It is important to note that living with friends or relatives is not an unambiguous risk factor for homelessness.  Under some circumstances a doubled up living situation can be considered unstable and can be a precursor to homelessness for families who may exhaust options for sharing housing with other family or friends[1]. In other cases, however, living with friends or family members could offer stability to a family.  As a result, while this study discusses this type of living situation as a potential precursor to homelessness, it is important to note that not every family living in a doubled up situation with friends or relatives will experience homelessness.

In analyzing the type of homeless indicators and risk factors collected on TANF, Medicaid, and combined applications, we concluded that 30 states can be considered homelessness-data collecting states, defined as collecting at least one of the homelessness indicators or risk factors on at least one of the applications used for TANF or Medicaid.

Even if homelessness is not explicitly addressed on the application, our telephone survey revealed that eleven other states gather information about homelessness or risk factors in other ways during the application process.  These states provide potential for measuring homelessness among applicants and we consider them to be potential homelessness-data-collecting states.  Finally, we found 10 states that do not collect homelessness data or risk factors on any of their TANF or Medicaid applications and do not use other means to determine if program applicants are experiencing homelessness.

Use of Homelessness Data

The study also explored how data on homelessness indicators and risk factors are used by states.  We found that homelessness data are used for the following purposes: to document living arrangements and applicant home address to process benefits and maintain contact with the applicant; to facilitate eligibility determination and administration of the Supplemental Nutrition Assistance Program (SNAP)[2] benefits (one of the programs covered on the combined applications); and to ascertain service needs and make referrals to needed services.  Most respondents did not know how complete the homelessness data are or whether the state has procedures in place to improve the quality of the items collected.  It appears that the data collected on homelessness indicators and risk factors are entered into statewide databases in many states, but the data are not routinely confirmed or verified, making it unclear without testing how reliable the data might be for analysis of homelessness.

Only five states (California, Colorado, Connecticut, New York, and Washington) reported having conducted any analysis of the available data and the analysis has been fairly limited.  For example, in Washington, the TANF and Medicaid officials have prepared reports to the Governor about the number of homeless persons applying for benefits using the data collected from TANF and Medicaid applicants.  In Connecticut, state officials have done some analysis of the information collected as part of a pilot program across the state to examine homelessness.  This information has been used in analysis of impacts of state efforts to alleviate homelessness and to link people discharged from prison to services.  The homelessness information collected from TANF and Medicaid applicants serves as a benchmark against which the impacts of other state efforts can be measured.  The respondents in Colorado were not able to provide details about the analysis conducted in that state.

In New York, data collected on homelessness indicators on the combined application are entered into the TANF management system and can be queried to produce reports.  Reports are usually prepared by districts or counties and the only reporting done thus far of homeless information has been done on an ad hoc basisthere is no established format, schedule, or protocol for producing reports on homelessness.  Customized reports based on data available in the system, including data on homelessness, could be generated, and state officials appear willing to consider this.

In California, families who are eligible for Californias TANF Program (CalWorks) and who are homeless can apply for a special need payment called a Homeless Assistance Payment to meet their costs for temporary and permanent housing.  Families who receive a homeless assistance payment are noted in the CalWorks data system and the state reports on the number of payments made each month.  Generating customized reports on homelessness using data collected on the combined Medicaid/TANF application has not been undertaken because of concerns about the resources needed to do so.

States that collect homelessness data were asked whether doing so is burdensome to local or state staff.  Overwhelmingly, respondents indicated that collecting the homelessness data does not take more time than other items on the application, does not involve additional costs to collect, nor does it take more time to verify or record than other application items (this may, however, be a result of the fact that few states have done extensive verification or analysis of the data).

Most states indicated a willingness to grant access to the homelessness data collected on TANF and Medicaid applications, even if no previous analysis has been conducted of the data.  Even among the willing states, however, there were some concerns about resource constraints for responding to customized data requests and concerns about the reliability of the data available since no previous assessment of the information has been done.  Several states were unsure whether they could provide data extracts, and some stated that they would only do so if required.

Reasons for Not Collecting Homelessness Data

States that do not collect homelessness data were asked about the reasons this type of information is not included on the applications.  The most frequent reason cited is that the information is not needed to determine eligibility for the TANF and Medicaid programs.  Only one of the respondents indicated that they thought collecting homelessness data would be too burdensome to administer.  More commonly, reasons for not collecting homelessness data were related to a desire to keep the application simple and focused only on information absolutely critical for processing eligibility determination.  Some also responded that it was unnecessary to collect homelessness data explicitly on the application because other processes like the application interview are used to confirm address information and allow case workers to ascertain homelessness.

Investigating Homelessness Data Collection in Other Mainstream Programs

A separate task under the study explored the extent to which it may be feasible for HHS to use existing program data to estimate the number of homeless people currently served by its smaller mainstream programs.  The objectives of this task were to explore the types of information collected by these programs and to identify possible approaches that HHS might use to obtain information on the number of homeless persons using these programs.  Nine mainstream programs that may serve persons experiencing homelessness were the subject of this review:

• Community Services Block Grant (CSBG)- Administered by the Administration for Children & Families (ACF)
• Head Start- Administered by the Administration for Children & Families (ACF)
• Social Services Block Grant (SSBG)- Administered by the Administration for Children & Families (ACF)
• State Childrens Health Insurance Program (SCHIP)- Administered by the Centers for Medicare & Medicaid Services (CMS)
• Health Center Program- Administered by the Health Resources & Services Administration (HRSA)
• Maternal and Child Health Services Title V Block Grant (Title V)- Administered by the Health Resources & Services Administration (HRSA)
• Ryan White HIV/AIDS Program (Ryan White)- Administered by the Health Resources & Services Administration (HRSA)
• Community Mental Health Services Block Grant (CMHSBG)- Administered by the Substance Abuse & Mental Health Services Administration (SAMHSA)
• Substance Abuse Prevention and Treatment Block Grant (SAPTBG)- Administered by the Substance Abuse & Mental Health Services Administration (SAMHSA)

Data Collection in the Nine Programs

Only one of the nine programs reviewed for this study currently provides individual-level data on homelessness status to HHS.  Grantees of the Substance Abuse Prevention and Treatment (SAPTBG) Block Grant are required to report admission and discharge data for all persons served by the program using a national database known as the Treatment Episode Data Set (TEDS).  Initially, TEDS was designed to generate annual statistics on the number and characteristics of persons admitted to public and private substance abuse treatment programs receiving public funding, including living arrangements at admission and discharge.  The living arrangement variable collected in TEDS includes three response codes:  1) homeless indicating no fixed address (including shelters); 2) dependent living (this refers to living in a supervised setting such as a halfway house or group home); and 3) independent living (this refers to living alone or with others without supervision and includes cases where children live with parents, relatives, guardians, or in foster care).[3]

Four of the mainstream programs report aggregate data to HHS on the number of homeless program beneficiaries served: Health Center program, Head Start, Ryan White, and Community Mental Health Services Block Grant (CMHSBG).  Grantees of these four programs use their own locally-managed administrative data systems to generate the required reports, and grantees all report beneficiary housing/homelessness status using this model.  The reports from each of these programs can help HHS estimate the number of homeless persons served by each program; however, they each have limitations.  Since the figures are aggregate, HHS cannot ensure that the numbers are unduplicated across grantees, or whether the questions are interpreted and calculated consistently.  In addition, information is limited to the tables or questions asked in a standard report template.  Nonetheless, estimates from aggregate reports can help HHS understand the extent to which their programs are accessible to persons who are homeless.

The four remaining programs the Community Services Block Grant (CSBG), Social Services Block Grant (SSBG), Maternal and Child Health Block Grant, and the State Childrens Health Insurance Program (SCHIP) also provide HHS with regular aggregate reports.  However, these reports do not include information on the number of program beneficiaries who are homeless or other data related to housing status.  Of the four programs in this category, CSBG, SSBG, and the Maternal and Child Health Block Grant programs have very limited quantitative reporting requirements.  Instead of using standardized report templates or tables, states submit annual narrative reports on the activities funded by the grants.  Alternatively, the SCHIP program has quarterly and annual reports in which grantees report data on the unduplicated count of new enrollees and data used to determine program eligibility.

Potential Approaches for Measuring Incidence of Homelessness Among Mainstream Program Participants

We considered several approaches that HHS might use to understand the incidence of homelessness among mainstream program beneficiaries. Two options appear most promising for further consideration.  The first would be to match individual program beneficiary identifiers with other benefit programs that collect homelessness data.  An example of this would be to search for mainstream program participants in a data system, such as SNAP records[4], where homelessness data are collected.  These data would require further extrapolation to account for people who are not represented in the other data system.  Another approach would match program beneficiary identifiers with local Homeless Management Information System (HMIS) data.  The U.S. Department of Housing and Urban Development (HUD) requires every community that receives HUD McKinney-Vento funding to participate in an HMIS and collect a set of standardized data elements on each individual served.

A second option would entail surveying directly a sample of program beneficiaries.  This would involve conducting a special research activity to collect data directly from a representative sample of program beneficiaries.  Using this approach, HHS could work with grantees to identify standard types of services delivered, and to select a sample of sites and program beneficiaries from each site from which to collect data.  If grantees or service providers collect more comprehensive data than are reported to HHS, it may be possible to identify homelessness based on case file reviews for the sample.  Alternatively, it may be necessary to work with grantees to collect data from a sample of beneficiaries over a specified period of time.

These approaches may be feasible, but they each offer advantages and disadvantages, and there is not a single option that would work unambiguously for all of the programs.  With each approach, HHS will need to carefully weigh the desire to understand better the population enrolled in the benefit programs with the desire to minimize reporting burdens for the states and entities that operate the programs, as well as for the recipients of the programs themselves.

[ Go to Contents ]

Chapter One. Introduction

This report presents the findings from an exploratory study requested by the U.S. Department of Health and Human Services (HHS) and managed by the Office of the Assistant Secretary for Planning and Evaluation (ASPE) in partnership with the Health Resources and Services Administration (HRSA).  The study investigates which states (including the District of Columbia) currently collect housing status data from applicants for Medicaid and/or Temporary Assistance for Needy Families (TANF), the two largest HHS mainstream programs that may serve individuals or families experiencing homelessness.  This project is part of a larger Departmental effort to increase access to HHS mainstream resources for persons experiencing homelessness.  Specifically, the study begins the exploration of available data that could be used to identify the number of homeless persons currently accessing HHS mainstream programs.  Understanding more about the extent to which these programs collect information on homelessness is an important first step in assessing the number of homeless individuals and families who access mainstream programs.

The main activity of the study was to design and administer a survey to all 50 states and the District of Columbia in order to document which states collect data on housing status from individuals and families applying for Medicaid or TANF.  The survey examined such topics as data collection cost and usability, and the reasons why states do or do not collect housing status data.  The focus of the survey was on state practices related to the collection of housing status data of program applicants and enrollees; not the collection and/or analysis of state data.  The study also included a review of practices in nine other HHS mainstream programs related to homelessness data collection.

Given the challenges that exist in developing baseline data on the number of homeless individuals or families served in mainstream HHS programs, this study focuses primarily on two of the eleven identified mainstream programs.  TANF and Medicaid have been selected for study not only because they are the two largest mainstream programs relevant to homelessness in terms of budgets, but also because both programs have individual or family eligibility criteria that determine access to enrollment and both programs require the completion of a state-administered application.  While not a federal requirement, some states do collect data on housing status on state Medicaid and/or TANF application forms.

The study examined only whether states collect information on homelessness from applicants for TANF and Medicaid.  The objective of the study was not to obtain details  on whether states track changes in housing status over time for program participants and whether data are available that would permit an analysis of trends in homelessness among program participants or effects of efforts to reduce homelessness.  It is important to bear this in mind in considering the types of analyses that could be done with the data being collected by the states and the research questions they can and cannot answer.

1.1    Background

According to a report by the Government Accountability Office (GAO; 1999), homeless individuals and families can be served by two types of federal programs: targeted programs and mainstream programs (sometimes referred to as nontargeted programs).  Targeted programs fund services designed specifically for persons experiencing homelessness (e.g., Projects for Assistance in Transition from Homelessness). Mainstream programs, on the other hand, are designed to serve low-income individuals or families who meet specific eligibility criteria, such as income, disability, or family composition (e.g., Medicaid or TANF), but can be accessed by persons experiencing homelessness provided they meet the eligibility criteria.  In a report to Congress in 2003, ASPE reported that, targeted programs in HHS were originally created to provide an emergency response and have provided extensive services.  However, they have not resulted in long term solutions to homelessness (Department of Health and Human Services [HHS], 2003^a). Furthermore, targeted programs receive considerably less funding in comparison to mainstream programs.  In fiscal year 2006, the five HHS targeted homelessness programs combined received roughly $342 million[5], whereas the eleven HHS mainstream programs most relevant to homelessness received a combined total of about $231 billion.[6]

Given this context, the Department of Health and Human Services has been pursuing a strategy over the past several years of increasing access to mainstream resources for eligible homeless individuals and families. In the March 2003 report, Ending Chronic Homelessness: Strategies for Action, the HHS Secretarys Work Group on Ending Chronic Homelessness outlined sixteen strategies to reduce chronic homelessness, one of which was to improve the transition of clients from homeless-specific programs to mainstream service providers (HHS, 2003^b).  A cornerstone activity under this strategy has been the development and implementation of nine Homeless Policy Academies that were designed to bring together state-level program administrators and homeless service providers in order to develop state-specific action plans designed to increase access to mainstream resources for persons experiencing homelessness.  However, the key policy question, Has HHS been successful at improving access to mainstream service programs? cannot yet be answered because no baseline data are available.  At the federal level, most mainstream programs are not required to collect data related to the number of homeless clients served.  This lack of baseline information about the number of homeless individuals and families served in HHS mainstream programs makes it difficult, if not impossible, for HHS to document improvements in access.

There are a number of challenges in developing this kind of baseline data, particularly due to the fact that homelessness is a dynamic state; a person may be homeless today but housed tomorrow, thus causing fluidity in the number of program participants experiencing homelessness at any given point in time.  However, further exploration is warranted to improve the Departments ability to develop measures related to increasing access to mainstream resources.

1.2    Overview of the TANF and Medicaid Programs

TANF

The Temporary Assistance for Needy Families (TANF) program is the nations primary cash assistance program for low-income families with children.  In 1996, with the passage of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), TANF replaced the Aid to Families with Dependent Children (AFDC) program.  TANF was reauthorized by the Deficit Reduction Act of 2005.  The current program, administered by HHSs Administration for Children and Families (ACF), is a product of substantial transformation that has reflected changing views of the purpose of assistance and the responsibilities of parents to support their families primarily through work.

TANF is structured as a flexible block grant and states have discretion to establish eligibility thresholds and benefit levels.  Federal law imposes work requirements for TANF recipients and a five-year limit on the receipt of TANF benefits.[7] As of March 2008, 1.6 million families were receiving TANF assistance, according to monthly caseload data maintained by ACF.

The purpose of TANF is to assist low income families to achieve self-sufficiency, so that children can be cared for in their own homes and to reduce their dependency by promoting job preparation, work, and marriage.[8] TANF cash assistance, child care assistance, and case management can provide important assistance to families experiencing homelessness, but at the same time homelessness can pose barriers to meeting the TANF work requirements.  Most states that have programs in place address family homelessness through their homeless assistance service systems and identify improving access to mainstream benefits as one of their goals.  Some states have established goals to improve access to TANF among homeless families, solicited help in benefits screening from community and homeless assistance providers, and established cooperative working relationships with homeless and housing providers to prevent homelessness or minimize the frequency and duration of homelessness.  Many family homeless assistance providers assess client eligibility for TANF, the Supplemental Nutrition Assistance Program (SNAP- or the program formally know as the Food Stamp Program), and other mainstream benefits as part of their case management process.

Medicaid

The Medicaid program is administered at the federal level by the Centers for Medicare and Medicaid Services (CMS).  Medicaid provides access to medical services to individuals and families satisfying financial standards established by the states.  Medicaid eligibility is also subject to categorical restrictions and is generally available only to low-income persons who are: aged; blind; receiving Supplemental Security Income benefits (SSI); members of families with dependent children; and certain other pregnant women and children.[9] As of December 2006, more than 42 million people were enrolled in Medicaid.[10]

Funds are allocated to states to provide medical benefits (with payments directly to vendor) to low-income people who have no medical insurance or inadequate medical insurance.  States receive formula grants that are matched at a rate ranging from 50 to 83 percent, based on the state average per capita income relative to the national per capita income (wealthier states receive lower federal matches).  Each state determines specific eligibility criteria and services, as long as they meet the minimum requirements.  As a result, a person that qualifies for Medicaid in one state may not qualify in another state.  States also determine the application process for obtaining Medicaid benefits.

Prior to TANF, families who qualified for cash assistance under the former AFDC program were automatically eligible for, and in most states, were automatically enrolled in Medicaid.  TANF eligibility does not provide this categorical eligibility for Medicaid, but under TANF, Medicaid entitlement was retained for individuals who meet the requirements of the previous AFDC program in effect as of July 1996.  For low-income parents these AFDC-related rules provide the primary pathway to Medicaid assistance.[11]

Depending on the state context, Medicaid-funded services can provide a significant level of assistance to persons who are homeless.  This frequently happens in two ways.  First, homeless service providers help clients assess whether they appear to be eligible for benefits, navigate the application process, and follow-through with the steps associated with enrollment, appeals (if necessary), and re-certification.  Once enrolled, these clients have access to services that may help address the underlying factors that contributed to their homelessness.

Secondly, some homeless assistance providers are forming direct partnerships with Medicaid service providers to address the primary and behavioral healthcare needs of homeless persons who are diagnosed with a chronic mental illness.  For instance, the homeless assistance provider may receive funding from the U.S. Department of Housing and Urban Development (HUD) to support transitional or permanent housing for residents, and the behavioral health provider uses Medicaid and other behavioral health funding to provide clinical supportive services for the residents.  Although these clients still need to be enrolled in Medicaid, the program is structured to use Medicaid to support specific, clinical services for its residents that help leverage housing grants.  The clients also received expanded access to a wide-range of other services, which is a positive by-product of Medicaid enrollment.

Regardless of the model, both the eligibility requirements and application process may pose challenges to homeless people who need medical assistance, particularly to those who are considered chronically homeless. While many homeless people have disabling conditions that would make them eligible for SSI, they may not have had the continuity of medical care and access to services needed to document the disability required to apply for SSI (and thus Medicaid coverage), among other mainstream programs.  The application process for SSI can also be lengthy.[12] The transience associated with homelessness raises additional barriers to receiving information about, applying for, and maintaining benefits.  Once the homeless client is enrolled, performing the annual financial eligibility re-determination can be equally challenging.

There are no requirements for either the Medicaid or TANF programs to track or report the number of homeless individuals or families who apply for or receive services from these programs.  This study explored the extent to which program applicants are asked about their housing status or homelessness, and if so, how the state does or could use the data to understand more about the number of homeless individuals and families who access these programs.

1.3    Study Activities

Initial Application Collection and Review

The first task under the study was to collect and review TANF and Medicaid applications from the 50 states and the District of Columbia to determine whether information on housing status and/or homelessness are included in the applications.  Assembling a complete set of state applications required a state-by-state search, since no centralized repository for these forms was available for either the TANF or Medicaid programs.  Some applications were located online on state websites and others had to be obtained from state Medicaid and TANF officials.  When an application was located online, we conducted follow-up phone calls with the contact person listed on the program offices website to verify that the application was up to date and to confirm the contact information of the state directors for each program.  This contact information was used as the starting point for contacts for the telephone survey.

One of the major findings during this initial application review was that many states have consolidated applications for multiple benefits programs, including TANF and Medicaid.  We refer to these as combined applications.  The use of combined applications influenced the design of the telephone survey instrument because it was important to clarify how the combined application was used and whether questions on these applications are asked of applicants for all programs covered.  We found three general practices regarding the type of applications used:  1) States with separate Medicaid and separate TANF applications with no combined application; 2) States using a separate Medicaid application (for applicants only seeking Medicaid assistance) and a combined application for TANF and Medicaid; and 3) a combined application only.

Through the application review, we also found that the information on the applications related to housing and homelessness could be grouped into three categories:  1) general housing status information; 2) indicators of homelessness; and 3) possible risk factors for homelessness. These categories were used to structure the telephone survey instrument.

Two versions of the telephone survey instrument were developed in collaboration with HHSone survey was targeted to states that collect information related to homelessness, and the second survey was targeted to states that did not collect housing or homelessness data.  For the states that do collect housing or homelessness data, the purpose of the interviews was to learn more about the information collected on the application related to housing status, homelessness indicators, and homelessness risk factors, and how such information is used in the states.  For those states that do not collect housing or homelessness data, the interviews were conducted to learn about the reasons why homelessness information is not collected on applications in some states and whether there are plans to add such information to the applications in the future.

A pre-test conducted in seven states provided valuable information about how states use data collected in application forms and the nature of the application process.  In the majority of the pre-test sample, the states used the application form as a first step in eligibility determination.  In these states, applicants first complete the application without assistance from program staff (after obtaining the application online or requesting one by mail), and subsequently come into the program office for an in-person interview with an intake worker to complete the application process and determine eligibility.  During the in-person interview, the intake workers typically enter data collected in the application into a computer database while they interview the applicant.  Caseworkers probe on certain questions in an effort to collect additional information from applicants.  Learning about this practice from the pre-test interviews led us to adopt procedures in the full survey implementation to determine whether homelessness data are collected at any point during the application/eligibility determination process, even if this information is not explicitly requested on the initial application form.

Survey Implementation

To encourage state participation in the study and establish the legitimacy of the study for respondents, two letters were sent to state TANF and Medicaid directors prior to the interviews, one from HHS and one from Abt Associates.  The letters briefly described the objectives of the study and emphasized that the information collected during the interview would not be used for monitoring purposes.  In addition, these letters provided a contact person at Abt to field any questions or concerns about the study from the state representatives.

After thorough interviewer training, four Abt staff members contacted state representatives by telephone to schedule and conduct interviews.  Prior to the interview, the Abt staff member reviewed the previously collected application to see if the state had or had not collected homeless data.  This information determined which version of the interview protocol to use for the interview.  The interview responses were entered into an Access database.  The data were then converted to Excel and SAS files for analysis.

Between February and April 2008, we completed interviews with TANF and Medicaid directors in the 50 states and the District of Columbia. In total, we conducted 70 interviews with program representatives.  Of the completed interviews, 30 were with TANF program representatives and 40 were with Medicaid program representatives.  It was possible to collect information about all states with fewer than 102 interviews (one for each state and District of Columbia with Medicaid and TANF representatives) because of the prevalence of combined applications that covered both programs.  (For more information about the interview procedures see Appendix C.)

Analysis of Survey Data

The analysis in Chapters 2, 3, and 4 of this report is based on the data collected during the interviews with state Medicaid and TANF staff.  The analysis is descriptive and intended to summarize the range of data collection practices and opinions found across the states.

In general, there is a great deal of consistency between the information collected for the TANF and Medicaid programs within each of the 50 states and District of Columbia.  Because of this, most of the analysis presented in the report focuses on the state as the unit of analysis, providing lists of states with various types of data collection practices and varying approaches to collecting homelessness data from Medicaid and TANF applicants. This approach identifies clearly which states have practices in place to record homelessness among Medicaid and TANF applicants.  In other cases, clearly denoted in the tables, we summarize responses by type of program to show how an issue is covered in TANF and Medicaid.  In each section of the report, we describe the analyses conducted and how the information was assembled.  Appendix A provides a description of each states data collection practices for readers desiring a state-by-state summary of the key interview findings.

Research on Other Mainstream Programs

A separate task under the study explored the extent to which it may be feasible for HHS to use existing program data to estimate the number of homeless people currently served by its smaller mainstream programs.  The objectives of this task were to explore the types of information collected by these programs, identify approaches that HHS might use to obtain information on the number of homeless persons using these programs and the relative costs of these approaches.

HHS identified nine other mainstream programs that may serve persons experiencing homelessness that were the subject of this review:

• Community Services Block Grant (CSBG)- Administered by the Administration for Children & Families (ACF)
• Head Start- Administered by the Administration for Children & Families (ACF)
• Social Services Block Grant (SSBG)- Administered by the Administration for Children & Families (ACF)
• State Childrens Health Insurance Program (SCHIP)- Administered by the Centers for Medicare & Medicaid Services (CMS)
• Health Center Program- Administered by the Health Resources & Services Administration (HRSA)
• Maternal and Child Health Services Title V Block Grant (Title V)- Administered by the Health Resources & Services Administration (HRSA)
• Ryan White HIV/AIDS Program (Ryan White)- Administered by the Health Resources & Services Administration (HRSA)
• Community Mental Health Services Block Grant (CMHSBG)- Administered by the Substance Abuse & Mental Health Services Administration (SAMHSA)
• Substance Abuse Prevention and Treatment Block Grant (SAPTBG)- Administered by the Substance Abuse & Mental Health Services Administration (SAMHSA)

Abt staff collected background information for each of the nine programs about the services funded, eligibility requirements, whether programs are administered at the state or sub-state level, and whether the programs report homelessness data at the federal level.  We supplemented the above information with interviews of federal officials who administer each of the programs.  The results of this exploratory review are presented in Chapter 5.

1.4    Organization of the Report

The remainder of the report is organized as follows. Chapter 2 discusses the type of application used by each state for TANF and Medicaid programs, as well as the type of housing or homelessness data they collect.  Chapter 3 describes the information collected by states about homelessness or homelessness risk factors on their TANF or Medicaid applications.  It also describes how the states use homelessness information that is collected.  Chapter 4 reviews the reasons cited by the states for collecting or not collecting homelessness data, as well as states plans for collecting these data or using them for measuring program performance.  In Chapter 5, other mainstream programs are examined to determine whether they collect information on homelessness and the approaches employed to collect homelessness data.  Chapter 6 offers a summary of the findings as well as suggestions for next steps for further analysis.  Appendix A provides a summary of the findings for each state.  Appendix B provides a list of program names for Medicaid and TANF in each state and Appendix C includes a description of the interview procedures and copies of the survey instruments.  Appendix D includes supplementary data tables regarding the use of homelessness data and Appendix E provides profiles of the nine other mainstream programs examined under the study.

[ Go to Contents ]

Chapter Two. Homelessness Data Collected from TANF and Medicaid Applicants

This chapter describes the types of applications used in states for the TANF and Medicaid programs, as well as findings from the telephone interviews about the type of housing status, homelessness indicators, and homelessness risk factors collected on the applications.  The chapter also presents information on the nature of the application process and practices in the states that may yield information on homelessness even if explicit questions are not asked on the application.

2.1    Types of Applications Used in the States

The majority of states and the District of Columbia use consolidated applications for multiple benefit programs, including Medicaid and TANF.[13] We refer to these as combined applications.  Some states use both a combined application and a separate Medicaid application.  Applicants for Medicaid fill out either the combined application or the Medicaid-only application, but are not required to complete both.  The Medicaid-only applications are used when an applicant only wishes to apply for Medicaid assistance.  In other states, only the combined application is used for all benefits programs, even when the applicant is requesting only one form of assistance.

Benefits of a combined application.  The use of combined applications is considered to simplify the application process for benefits and to ensure that applicants are given the opportunity to apply for all benefits for which they may qualify.  In a 2003 study of six states, Holcomb et al. notes that such applications are a result of earlier efforts to integrate the eligibility systems for TANF, the Supplemental Nutrition Assistance Program (SNAP), and Medicaid into a single eligibility package.  Programs typically included on the combined application are TANF, SNAP, and Medicaid.  Other programs like energy assistance, some state disability benefits, and SCHIP are also included in many combined applications.

In most cases, applicants completing a combined application first indicate which type of benefits they wish to apply for and are then instructed to answer questions pertinent to those programs.  This means that not all individuals using a combined application necessarily answer all questions.  During our interviews, we confirmed which program applicants answer each type of housing and homelessness question.  When applicants complete online versions of the combined applications, they are automatically advanced to the questions they must answer depending on the type of benefits applied for. Florida, Kentucky, Maine, and Maryland are examples of states where this practice is followed.

Implications of combined applications.  Given the wide use of combined applications that cover both Medicaid and TANF applicants, there is a great deal of overlap and consistency between the data collected by Medicaid and TANF programs.  For the current study, this meant that one interview with either a Medicaid or TANF respondent was often sufficient to provide information about both programs.  For subsequent data collection that might be pursued to analyze the number of Medicaid and TANF applicants who are homeless, this may imply that data can be collected for both types of applicants from a single source.  However, we would expect specific practices to vary from state to state and further investigation would be necessary to develop approaches for collecting these data in each state.

Prevalence of combined applications.  At the time of the interviews, we found that 45 states use a combined application for their Medicaid and TANF programs.  Twenty-one use combined applications in addition to a separate Medicaid application and 24 states use only a combined application for Medicaid and TANF programs (see Table 1).  We found that no state uses both a combined application and a TANF-only application, and that no states use separate TANF and separate Medicaid applications in addition to a combined application.  Only six states (Alabama, Louisiana, Massachusetts, Mississippi, North Carolina and South Carolina) reported at the time of the interview not using combined applications at all.  One state, North Carolina, was not using a combined application at the time of the interview, but is reportedly considering doing so in the future.

Another important finding from the telephone survey is that in all states, for both Medicaid and TANF, the same application is used throughout the state.  This is true even in states where the TANF program is administered at the county level.

We found that when homelessness data are collected on a combined application they are almost always collected for both Medicaid and TANF applicants.  The exception to this is Virginia, where a combined application is used and includes several homelessness-related questions.  The homelessness questions are asked of SNAP applicants and TANF applicants, but not of Medicaid-only applicants.  Since responses differ for the TANF and Medicaid programs in Virginia, we treat this state in our analysis as if there were separate TANF and Medicaid applications.

Of the 21 states where a separate Medicaid application is used in addition to a combined application, six states (Colorado, Missouri, New Mexico, New York, Rhode Island, and Utah) collect information on homelessness through their combined application but not through their Medicaid-only application.  As a result, data are available on homelessness for applicants applying for more than one type of program but not for Medicaid-only applicants.  The states were not able to provide estimates of the proportion of Medicaid applicants completing the combined or separate application that would be necessary to gauge the coverage of Medicaid applicants based on the use of the combined application.  The main reason given for not including the homelessness information on the Medicaid-only applications is that the information is not required for determining eligibility for Medicaid and is thus considered unnecessary.

2.2    Collection of Housing Status, Homelessness Indicators, and Homelessness Risk Factors

The primary goal of the study was to ascertain whether housing or homelessness data is collected from TANF and Medicaid program applicants by the states administering the programs.  For those states that do collect data, a second objective was to learn about the specific data collected and how the information is used by the state, as well as the incentive for and cost of collecting the data.

The application review revealed three general categories of information that are collected on TANF and Medicaid applications:

• general information on housing status (home and mailing address; directions to home; residence in public or subsidized housing; intention to remain in state; residence in long-term care facility; residence in medical or rehab facility)
• indicators of homelessness (self-reported homelessness; residence in shelter; residence in domestic violence shelter; self-reported lack of permanent home)
• risk factors often associated with homelessness (residing with family or friends[14]; being subject to an eviction notice).  These risk factors are not indications of actual homelessness but may identify precarious housing situations that can lead to homelessness.

The application form itself is not the only vehicle that states may use to ascertain the housing status of program applicants.  As a result, reviewing the applications is not sufficient to answer the questions of interest for the study.  The telephone interviews were used to learn more about the application process and to learn whether other procedures are followed to document homelessness of program applicants.

In this section we describe items included on the applications in the three categories (housing status, homelessness indicators, and homelessness risk factors).  This review identified three groups of states: 1) homelessness-data-collecting states with homelessness indicators or risk factors included on the applications; 2) potential homelessness-data-collecting states that do not include questions on homelessness indicators or risk factors on the application but ascertain homelessness from applicants in other ways; and 3) states that do not collect homelessness data by any means.

2.3    Housing Status Information

Table 2 shows which states collect each of seven housing status items on Medicaid and TANF applications.  Notes on the table indicate if an item is included only on the combined application in states that use both a combined and separate Medicaid application.  Information in Table 2 is shown separately for TANF and Medicaid in states using separate applications for each program.

As the table indicates, all states ask for home address and most also ask for mailing address.  These are the most common housing status items included on the applications.  Only one state (Michigan) includes all of the housing status items on their application.  Less frequently asked questions about housing status are about residence in long-term care facilities or a medical or rehab facility.  Several states, most of which include sparsely populated rural areas, include a question asking for directions to the applicants home, to facilitate home visits that may be needed and to help confirm home address.  These states are Alaska, Alabama (TANF application), Kansas, Michigan, Montana, Nevada, New York, North Carolina, North Dakota, Oklahoma, South Dakota, Texas, Vermont, Virginia, and West Virginia.  Several states report including other types of housing status items.  These included items such as:

• alternative mailing address (DE);
• housing costs (CA);
• residence in assisted living facility, group home, and previous address (NE);
• interest in receiving housing assistance (PA);
• amount paid for rent (WY);
• receipt of housing assistance (CT);
• type of housing situation (rent or own home) (ID);
• safe delivery address (ME);
• adult foster care (MI);
• residing in group home, or half-way house (MO and RI);
• reason moved to state (MT), and;
• residence in group home or incarcerated (UT).

These housing status items by themselves do not provide evidence of homelessness or risk of homelessness, but in some cases, when housing status is explored in greater detail during interviews with case workers during the application process, homelessness or housing instability may be uncovered.  For example, applicants may be asked to confirm their home address or to provide more information about their living situation and this may reveal that the applicant is homeless.  Understanding the types of housing status items included on the application is a good first step in identifying states that may have potential for measuring homelessness among program applicants.

2.4    Homelessness Indicators

In the application review and telephone survey, we also asked interview respondents which homelessness indicators are included on the application.  Twenty-eight states are collecting some level of homelessness indicators, as shown in Table 3.  Notes on the table identify situations where an item is not included on all applications.  In addition to the states shown in Table 3 that were collecting the homelessness indicators at the time of the interview, DC, North Carolina (TANF), and Wyoming stated that they expect to add a question Are you homeless? to their applications sometime during 2008.  Arizona, which currently asks if applicants reside in a shelter, plans to add Are you homeless? to their combined application before the end of 2008.

2.5    Homelessness Risk Factors

In the interviews we found that 13 states ask questions that do not explicitly indicate homelessness, but could be considered risk factors for homelessness.  The two items were living with friends or relatives and having an eviction notice. Table 4 displays the states that include these questions on their Medicaid and TANF applications.  One of these items, living with friends or relatives may or may not indicate risk factors for homelessness.  On the one hand, under some circumstances a doubled up living situation can be considered unstable and can be a precursor to homelessness for families that exhaust options for sharing housing with other family or friends.  In other cases, living with friends or family members could offer stability to a family.  As a result, while we present this living situation as a potential precursor to homelessness, not every doubled up family becomes homeless.

All states but two that ask about homelessness risk factors (Table 4) also ask about homelessness indicators (Table 3).  Idaho and Mississippi (Medicaid application only) ask about a risk factor but do not include a homelessness indicator on the application.  Sixteen states include homelessness indicators but not risk factors.

We consider states that collect either homelessness indicators or risk factors to be homelessness-data-collecting states.  Any state that explicitly collects at least one of the homelessness indicators or homeless risk factors on an application meets the standard established for this categorization.  As shown in the Tables 3 and 4, the interviews show that 30 states collect at least one of the homeless indicators or risk factors on at least one of the applications used for TANF or for Medicaid.  However, this information does not necessarily cover all applicants for Medicaid and TANF in some states[15], since the items are included on only the combined application or only the application for Medicaid or TANF.

Even if homelessness is not explicitly addressed on the application, our telephone survey revealed that several states gather information about homelessness or risk factors in other ways during the application process.  These states provide potential for measuring homelessness among applicants and we consider them to be potential homelessness-data-collecting states.  These states are listed in Table 5 along with the procedures they use to obtain homelessness data.

Table 5: States that collect homelessness data using procedures other than the application

Alaska:  Applicants are asked about homelessness at an eligibility interview conducted after the application is completed.  If an applicant indicates being homeless, this information is entered into the state database.

Arkansas:  When eligibility workers verify addresses provided on the application, they can often tell if an applicant is homeless by the address provided.  This is because eligibility workers are familiar with the addresses of local shelters.  However, even if a worker realizes that a program applicant is homeless, this information is not recorded in the TANF or Medicaid databases.

Delaware:  The combined application does not explicitly ask applicants if they are homeless.  However, applicants who complete the combined application online are given a menu of response options to the question of home address.  The response options include homeless, staying in a shelter, and staying in a domestic violence shelter.  When these responses are selected they are entered into the statewide database and could be queried for analysis.  People who fill out a paper version of the application are not asked about homelessness on the application but meet with an eligibility worker after completing the application and homelessness is discussed during the interview.

D.C.:  TANF applicants meet with an eligibility worker after the application is completed and housing status and homelessness are discussed as part of that interview.  For TANF, being homeless is considered a barrier to achieving the work requirements associated with TANF and case workers work with participants to make referrals to services to address homelessness. Information on homelessness is not recorded in a database though and is thus not available for analysis.  DC officials expect a question on homelessness to be added to the combined application in 2008.

Georgia:  Homelessness information is not collected on the application but if an applicant volunteers that (s)he is homeless, the usual practice would be to enter the TANF or Medicaid office address into the home address field in the database.  As a result this could be used as an indicator or homelessness.  But this practice is not standardized across the state and may not provide a reliable count of homelessness among program applicants.

Hawaii:  TANF applicants meet with an eligibility worker after completing the application and the applicants living arrangement is reviewed during this interview.  A code for homelessness can be entered into the database if an applicant appears to be homeless on the basis of the interview.

Indiana:  During an interview with eligibility workers after the application is completed TANF and Medicaid applicants are asked whether they receive housing assistance and if they are homeless.  This information is not recorded in a state database and the state has not attempted to analyze the information obtained during the interviews.  These questions have been asked as part of the application interview for at least 18 years.

Massachusetts (TANF application):  The Medicaid application collects homelessness indicators; however, the TANF application does not.  For the separate TANF application, caseworkers use an informal process to confirm the home address provided by the applicant, and in this process they often become aware if an applicant is homeless.  This information is used to make appropriate referrals to services.  Data on homelessness gathered in this way are not entered into a database and are not considered reliable for measuring homelessness among program applicants.

Mississippi (TANF application):  Eligibility workers ask TANF applicants about their homelessness situation during an interview as part of the application process even though explicit questions about homelessness are no longer part of the application (these were eliminated to make the application shorter).

Nevada:  TANF applicants participate in an interview after completing the application and the home addresses listed on the application are reviewed.  Caseworkers can usually tell if an address given on the application is for a shelter.  Caseworkers record homelessness for applicants who provide a shelter address or report no home address in the state TANF database.

New Jersey:  Applicants who complete the combined application meet with an eligibility worker after completing the application and housing status is discussed in the interview.  There are no standard procedures for collecting information on homelessness during the interview nor is information on homelessness entered into the TANF or Medicaid databases.

Washington:  TANF applicants are interviewed by a case worker after completing the application and case workers ask applicants if they are homeless during that interview.  This information is entered into the state TANF database and has been analyzed in the past to prepare reports to the Governor about the number of homeless persons applying for benefits.

Wisconsin (TANF):  TANF applicants and those enrolled in TANF meet with case workers who ask about housing status and homelessness.  The reason for this is to make sure that TANF participants receive assistance and referrals to needed services to help remove barriers to employment like homelessness.  The state TANF database includes a code for unstable housing and homelessness which could be queried to analyze homelessness among program applicants.  Such an analysis has not been conducted in the past.

Source:  2008 Interviews with TANF and Medicaid state officials

A final group was identified during the telephone interviews which we call states that do not collect homelessness data.  This group (see Table 6) does not collect any of the homelessness indicators or risk factors on their applications and do not use other methods to document homelessness.  However, as noted previously, Wyoming reported that their state plans to add a question about homelessness to the combined application, and North Carolina respondents indicated that their state may transition to a combined application in the future with a question about homelessness.

States that have eliminated homelessness questions on their application.  We also found that Hawaii, Iowa, Mississippi (TANF application), and Oklahoma collected homelessness data on applications in the past but have subsequently eliminated these questions.  Iowa and Oklahoma are states where no homelessness data are collected, and questions regarding homelessness were eliminated from the applications to streamline the application to focus only on items needed for determining eligibility for the Medicaid and TANF programs.  In Hawaii, an interview conducted with TANF applicants is used to ascertain homelessness in order to refer needy families to housing services.  Homelessness questions were eliminated from the application because the interview was considered a better method for finding out about homelessness and taking steps to address it.  The situation in Mississippi for the TANF program is similar.  In that state, efforts have been made to move to a paperless application process and to reduce the size of the TANF application.  Questions about homelessness have been removed from the application but eligibility workers continue to address homelessness with program applicants in an interview.  In both Hawaii and Mississippi the information gleaned from the interview regarding homelessness is entered into the TANF program database and could be queried.  In Oregon the question Are you homeless? was previously included on the separate Medicaid application but was later eliminated.  However, the combined application in Oregon asks if the applicant has a permanent home and whether the applicant has an eviction notice.

2.6    Application Process

In this section we discuss the application process in greater detail to describe other ways in which homelessness data are obtained from TANF and Medicaid applicants, even if not on the application form itself.

Updates to the Medicaid and TANF Applications

The telephone interviews were conducted about 15-17 months after we collected copies of the application forms.  By the time of the interview, more than half of states responded that they had made changes to their applications since the initial review in 2006.  We asked the interview respondents how frequently applications are typically updated.  This information is shown in Table 7 for Medicaid and TANF applications.  Responses for states that do collect homelessness data has been combined with those who do not collect homelessness data because we did not find differences in patterns of application updates across the two types of states.  Well over half of all states (65 percent of Medicaid programs and 73 percent of TANF programs) report that they update their applications as needed, in response to policy or regulatory changes or to make other necessary adjustments to the application. Most who reported as needed updates indicated that applications are updated at least every three years.  No states indicated that their applications are updated more often than quarterly, but 10 percent of Medicaid and 4 percent of TANF applications were reported to be updated twice per year.  In Florida, the application is updated 2-3 times per year in response to changes in eligibility requirements and to ensure that the form continues to be user friendly.  Annual updates to applications are the practice for just over a quarter of all Medicaid programs and just under a quarter of TANF programs.

Despite these relatively frequent application updates, we found only one case in which a change had been made specifically regarding the collection of homelessness data from the time of the initial review in 2006 and the interview in 2008.  In Oregon (where both a separate Medicaid application and combined application are used) at the time of the initial application review, the separate Medicaid application included a question Are you homeless?  This question was eliminated from the application by the time of our interview in 2008.  However the combined application asks Do you have a permanent home? and Do you have an eviction notice?

Completion of Medicaid and TANF Applications

For this study, the interview instrument asked respondents to tell us the most common ways that applications are completed.  Response categories were: 1) completed online; 2) completed in person at a TANF or Medicaid office; 3) completed and then mailed to a TANF or Medicaid office; or 4) other.  The responses to this question are shown in Table 8 for Medicaid and TANF programs.  The interviews provided information on the types of activities conducted during the application process and whether data on homelessness is obtained during any of these steps.  However the purpose of this study was not to develop a comprehensive assessment of the TANF and Medicaid application process or eligibility determination so interview respondents were not asked to give step-by-step descriptions of the application and eligibility determination processes.  A 2003 study (Holcomb et al.) examined the application process for TANF and Medicaid in detail in six states using data gathered from site visits and interviews with program officials.

Across all programs, completing the application in person at a TANF or Medicaid program office was the most frequent method used. Completing an application online was reported infrequently, but many respondents indicated that applicants can obtain a copy of the application online and then print and complete it and take it to a TANF or Medicaid office.  For the most part, TANF and Medicaid applicants complete the application form, at least initially, without assistance from an eligibility worker.  Table 8 also shows that about half of all programs collect housing status information from applicants in venues other than the application form itself, usually at an application interview.  We also asked states that collect homelessness data whether homelessness is addressed at points other than the initial application.  A quarter of Medicaid programs and a third of TANF programs reported that they collect information prior to eligibility determination, 10 percent of Medicaid and 16 percent of TANF collect such information after eligibility is determined, and 27 percent of Medicaid programs and 33 percent of TANF programs collect it again at annual recertification.

During the interviews, several states provided more detailed information about the application process used for TANF and Medicaid.  Several states reported that after an application form is complete, the applicant meets with an eligibility worker who reviews the application and confirms and verifies the information, and enters data into the state system to determine eligibility.  This interview typically includes a discussion of living arrangements, household members, and often reveals if an applicant is homeless.

Some states described efforts to move to a paperless system where applicants are not required to fill out an application form on paper at all.  For example, in Maryland, applicants can access the Service Access Information Links (SAIL) system online and can apply for TANF, Medicaid Child Care Subsidy, SCHIP, Energy Assistance, and other programs through this system (https://www.marylandsail.org).  An individual interested in TANF or Medicaid benefits can complete the application and submit it electronically for consideration.  An eligibility worker then contacts the applicant and schedules an interview to review the application and complete eligibility determination.  Applicants can complete the application online and submit it in a similar process in Florida and New Jersey (for those states Medicaid-only applications).  Other paperless systems are used in Missouri using the states Family Assistance Management Information System (FAMIS); Wisconsin, via the Client Assistance for Re-employment & Economic Support (CARES); Kentucky, using Kentucky Automated Management and Eligibility System (KAMES); and West Virginia, using Recipient Automated Payment Information Data System (RAPIDS).  In these states, an applicant meets with an eligibility case worker who enters application data into the states system where information is confirmed and discussed, and eligibility can be determined in real time- no paper forms are completed.

The application interview seems to be more prevalent for TANF applicants and for combined applications than it is for Medicaid-only applicants who complete a separate Medicaid application, but this varies from state to state.  In most cases, homelessness is determined in these interviews through the course of verifying the home address, confirming residence in the state, discussing living arrangements and reviewing household composition.  Most respondents report that they dont follow a strict protocol for obtaining the information during the interview, and dont ask uniform, specific questions. If an applicant is determined to be homeless during the interview, it is sometimes entered into a portion of the program database called case notes or into a comment field, but in some states it is not entered into the database at all.  In Delaware, the interview is an opportunity for the intake worker to probe for more detail on the living arrangement.  In that state, the eligibility worker enters data from the application into the state system and there are response codes related to living situation in the database to indicate homelessness, even though applicants are not asked directly about their homelessness status on the application.

In Hawaii, an interview is required for TANF applicants, and during the interview the applicants living situation is reviewed.  If an applicant is found to be homeless during the interview, a code is entered into the TANF database.  Homelessness is also discussed and recorded in the database at annual recertification in Hawaii.

Interviews were reported to be required for those completing a combined application in Montana, New Jersey, Oklahoma and Oregon.  TANF applicants were reported to require an in-person interview in Georgia, Hawaii, Idaho, Mississippi, Nebraska, Nevada, New York, Pennsylvania, South Carolina, South Dakota, Tennessee, Utah, and Vermont.  Since we did not ask each state to describe the step-by-step process for applying for benefits, it is possible that such interviews are used in other statesthese are the states that mentioned it without prompting.  On the other hand, requiring an interview as part of the application process does not necessarily mean that homelessness is addressed.  Montana, Vermont, South Dakota, and South Carolina conduct interviews during the application process but do not collect information on homelessness.

In summary, telephone interviews with TANF and Medicaid program officials indicate that more than half of all states (30) collect homelessness data from people applying for Medicaid and TANF benefits on at least one of the applications.  An additional 11 states ascertain homelessness of applicants at some point in the application process, even if not explicitly addressed on the application.  Twelve states do not collect homelessness data using any method from Medicaid applicants and ten do not collect these data from TANF applicants.

Chapter Three. Use of Homelessness Data

3.1    Defining Homelessness

There is no single common definition of homelessness used by the federal government.  Rather, different departments and agencies have developed their own programmatic definitions to determine eligibility for assistance.  HUD programs funded by the McKinney-Vento Act define homelessness based on the lack of a fixed, regular, and adequate nighttime residence; and the use of a nighttime residence that is a supervised public or privately owned shelter, an institution that provides a temporary residence for individuals intended to be institutionalized, or a public or private place not designed for, nor ordinarily used as a regular sleeping accommodation.[16]  This definition excludes individuals who are temporarily living with others and those staying in motels due to a lack of housing alternatives.

Other agencies like the U.S. Department of Education use a broader definition that includes individuals living in doubled up situations.[17]  The U.S. Department of Agriculture also uses a broader definition than HUD for the purposes of the SNAP program, considering an applicant homeless if he or she 1) has no place to sleep; 2) lives in a shelter or halfway house; 3) lives in someone elses home temporarily (fewer than 90 days); or 4) lives in a doorway, lobby, bus station, or some other place where people do not usually live.[18]

In the telephone survey, we asked whether the states had adopted a formal definition for homelessness when considering applications for Medicaid and TANF.  Even though homelessness is not a condition for eligibility for TANF or Medicaid, 14 states (covering 11 Medicaid programs and 14 TANF programs) responded that the state has adopted a formal definition of homelessness that is used to document homelessness for a variety of programs.  Many of these states that use a combined application that also covers the SNAP have adopted the SNAP definition of homelessness since the information is used for SNAP eligibility determination.  It appears that most states that report using a formal definition of homelessness tend to use a broader definition encompassing doubled-up situations.  Use of a formal definition may indicate that states are making an effort to coordinate with homeless assistance service providers to address homelessness.  The states that report using a formal definition of homelessness are:

• Alaska
• Arizona
• California
• Connecticut
• Delaware
• Illinois
• Kentucky
• Louisiana (for the TANF program)
• Maryland
• Massachusetts (for the TANF program)
• New York (for the TANF program)
• Texas
• Washington
• West Virginia

3.2    Use of Homelessness Data

Once we determined which homelessness indicators and risk factors are collected, our next task was to identify how the information is stored and used by the state.  In the telephone interviews, we asked the following questions for each homelessness indicator and risk factor collected on the application:

• Is a response to the question required for the application for assistance to be considered?
• Is the response used for any of the following purposes: program eligibility, determining benefit amounts, reporting; referral to services; program outcome measures; used to match to other datasets?
• Is the response entered into a statewide database?
• For all applications received in 2006, approximately what percent of cases in the state database have missing data for the item?
• Are there procedures in place to improve the quality of data for this item?
• When did the state begin asking this question?
• Has the wording of the question or the response format changed since that time?
• Do you expect to make changes to the wording or response format?
• Do you foresee eliminating this item?

For the most part, we found that homelessness data tend to be collected on applications for the following purposes: to document living arrangements and applicant home address in order to process benefits and to know how to reach the applicant; to facilitate eligibility determination and administration of SNAP benefits (one of the programs covered on the combined applications); and to understand service needs and make referrals to needed services.  At the end of this section we include a discussion of the SNAP and homelessness.  Most respondents had difficulty answering questions about the completeness of the homelessness data and most did not know whether there were procedures in place to improve the quality of the items collected.  It appears that these items are entered into statewide databases in many states, but the data are not routinely confirmed or verified, making it unclear without testing, how reliable the data might be for analysis of homelessness.

In this section, we discuss each homelessness indicator and risk factor, along with information provided about how each item is used; Table 3 provides details on which states ask each question.  Table 9 provides information on how the question Are you homeless? is used and detailed tables for the other items can be found in Appendix D.

Are you homeless?  This question is asked by 20 states, covering 17 Medicaid programs and 19 TANF programs.  Medicaid-only applicants using the separate Medicaid application are not asked this item in three of these states (Missouri, New York, and Rhode Island), but Medicaid applicants in those states who use the combined application are asked this question.  Table 9 displays information on how this item is used and stored by states that collect it.  The item is required by some but not all of the states that collect it and the data is reported to be used to determine benefits, program eligibility for SNAP, and for reporting.  Most states that ask the question Are you homeless enter the responses into a statewide database and about half report having procedures in place to improve the quality of the data collected in response to this question.  Most states that ask this question have been doing so for more than 10 years, with only a couple having added the question in the past 3 years.  Most states do not anticipate changing the wording of the question or eliminating it from the application, indicating that the collection of these data is a well established practice that is not subject to change in the future.  This may indicate that it would be feasible to obtain data from the subset of states that collect and electronically record this homelessness indicator.

Do you reside in a shelter?  This question is asked in 13 states, covering 12 TANF programs and 11 Medicaid programs; (the question is not asked on the Louisiana or Virginia Medicaid applications; and is not asked on the Massachusetts TANF application).  In four states (Missouri, New York, Rhode Island, and Utah) Medicaid-only applicants who use the separate Medicaid application are not asked this question.

Do you reside in a DV shelter?  A total of ten states ask this question on their applications, covering ten TANF programs and nine Medicaid programs (the question is not asked on the Virginia Medicaid application).  In four states (Missouri New York, Rhode Island, and West Virginia), Medicaid-only applicants who use the separate Medicaid application are not asked this question.

Do you have a permanent home?  Seven states include this question on their applications, covering seven TANF and seven Medicaid programs.  In two states (New Mexico and Rhode Island), Medicaid-only applicants who use the separate Medicaid application are not asked this question.

Do you live with friends or relatives?  Nine states ask if applicants live with friends or relatives at the time of program application, covering eight TANF programs (the question is not asked on the Mississippi TANF application) and eight Medicaid programs (the question is not asked on the Virginia Medicaid application).  In two states (New Mexico and Rhode Island), Medicaid-only applicants who use the separate Medicaid application are not asked this question.

Do you have an eviction notice?  Five states ask if an applicant currently has an eviction notice at the time of application, covering five TANF and five Medicaid programs.  In two states (New York and Oregon), Medicaid-only applicants who use the separate Medicaid application are not asked this question.

Several states reported collecting homelessness indicators for use in the Supplemental Nutrition Assistance Program (SNAP- formerly known as the Food Stamp Program).  How is homelessness data used for SNAP?  The Supplemental Nutrition Assistance Program is administered by the U.S. Department of Agriculture.  It is the nations largest nutrition assistance program serving low-income families and it provides assistance in purchasing food.  Eligibility for the program is determined by income.  In many states, the combined application is also used to determine eligibility for SNAP.  The reason cited in many cases for including questions on the combined application related to homelessness was to facilitate eligibility determination or other administrative requirements for SNAP.  Under federal law, SNAP offices are required to assist homeless persons in receiving benefits.  Federal requirements also require SNAP caseworkers to verify a persons address, unless (s)he is considered homeless.  Homeless persons are eligible for SNAP benefits if they meet the SNAP definition of homelessness.  The manner in which the state agency verifies and documents a persons homeless status is up to the state.  In many states, caseworkers will accept a written or oral statement from a collateral contact such as an employer, a homeless shelter worker, or the individual with whom the applicant is temporarily residing.[19]

Homeless persons may use SNAP benefits in participating shelters and soup kitchens, as well as in grocery stores and farmers markets. In calculating benefits amounts, SNAP allows homeless persons a deduction for shelter costs.  For instance, some homeless persons may incur housing costs by paying to stay temporarily with friends or at a hotel.  If the homeless person can document those shelter costs, (s)he may be able to deduct this amount from her/his net monthly income used to determine benefits.  If a homeless person cannot document shelter expenditures, (s)he may be able to use the shelter deduction.[20]  The homeless shelter deduction allows homeless persons who have incurred some type of housing costs to deduct $143 from their monthly income.  The Farm Bill of 2002 simplified the requirements for states to utilize the homeless shelter deduction.  Previously, states were required to document the average amount spent by every homeless person on shelter for the person to use this deduction.  Currently, states are only required to inform the USDA that they want to utilize the shelter deduction for a homeless person; there is no need for documentation.  As a result of this change to the legislation, it is likely that more states will permit homeless persons to make use of the shelter deduction.[21]

Chapter Four.Potential for Using Homelessness Data to Measure Access to TANF and Medicaid for Homeless Individuals

4.1    Reasons for Collecting Homelessness Data

This chapter examines the reasons identified by states for collecting (or not collecting) homelessness data from Medicaid and TANF program applicants, and program administrators opinions about the burden of collecting this information.  The chapter also includes a discussion of the extent to which states have analyzed homelessness data.  Table 10 shows the reasons states gave for gathering this type of information on their TANF and Medicaid applications.

As shown in the table, the most common response was the other category.  The kinds of things mentioned under other purposes included: using the information to satisfy the interests of local homelessness advocates, to make referrals for services, to assist in contacting the applicant, and because of a desire to understand the personal circumstances of applicants.  As discussed in the previous chapter, several respondents also report that the combined application had been developed to accommodate eligibility rules for SNAP, and the homelessness questions were included on the application for the SNAP.  The respondent from New York stated that the homelessness items had been added to the combined application in that state with the goal of understanding causes of homelessness and to help develop homelessness prevention strategies.

Two states (covering two TANF programs and one Medicaid program) responded that collecting homelessness data is used to support the development of the states 10-year plan to end homelessness.  One of these states was Ohio, but it was not clear from the interviews how the data are being used to serve this purpose.  In Ohio, it is possible that the application data may help to support efforts underway through the state Interagency Council on Homelessness and Affordable Housing though the interview respondents we spoke to did not make this connection.  Ohio is a participant in the SOAR (SSI/SSDI Outreach, Access, and Recovery) Initiative, a Federally-sponsored effort to streamline access to SSI/SSDI benefits for eligible persons experiencing homelessness.  As part of this initiative, the Ohio Department of Job & Family Services, in partnership with the State of Pennsylvania, has committed funds to upgrade the Ohio Benefits Bank (OBB), an online public benefits and EITC screening and application tool, to include Medicaid and Social Security benefits screening/application.  This is an example of the ways that the application data could perhaps be used to support other efforts to reach and serve people who are homeless.

States that collect homelessness data were asked whether doing so is burdensome to local or state staff (see Table 11).  Overwhelmingly, respondents indicated that collecting the homelessness data does not take more time than other items on the application, nor do they take more time to verify or record than other application items (this may, however, be a result of the fact that few states have done extensive verification or analysis of the data).  Only one state (Wisconsin) reported that there are additional costs associated with the collection of homelessness data from TANF and Medicaid applicants, in terms of the time and effort required by case workers to obtain and record the information. This may be because in this state a formal question is not included on the application, but instead case workers meet with program applicants to discuss their housing status, determine whether the applicant is homeless, and record it in the database, something that may require more time than would a standardized question on the application.

4.2    Reasons for not Collecting Homelessness Data

States that do not collect homelessness data were asked about the reasons for not including this type of information on the applications, and Table 12 shows these responses.  The most frequent reason cited for not collecting homelessness data is that the information is not needed for eligibility determination for the TANF and Medicaid programs. Several states reported that they only include essential information on the application.  As one Medicaid program official explained, Im not sure what we would do with the information if we collected it.  Only one respondent indicated that he thought collecting homelessness data would be too burdensome to administer.  More commonly, reasons for not collecting homelessness data were related to a desire to keep the application simple and focused only on information absolutely critical for processing eligibility determination.  Some also responded that it was unnecessary to collect homelessness data explicitly on the application because other processes (like the application interview to confirm address information) allow case workers to ascertain homelessness.

The state of New York provides an interesting case example. The combined application for TANF and Medicaid includes three homelessness indicators (self-reported homelessness, residence in a shelter, and residence in a domestic violence shelter), and one risk factor (residing with friends or relatives).  The separate Medicaid application, completed by people who only want to apply for Medicaid assistance, does not include any of these questions.  The Medicaid program representative stated that the reason that homelessness questions are not part of the separate Medicaid application is to keep the focus of the program on health insurance.  They described a desire to maintain a distinction between the combined application, which is used to apply for both TANF and Medicaid, and the separate Medicaid application and have made a concerted effort to treat Medicaid as a health insurance program rather than a poverty alleviation program.  There are five additional states (Colorado, New Mexico, Missouri, Rhode Island, and Utah) that have similar arrangements, collecting homelessness data on the combined application, but not collecting homelessness data on the separate Medicaid application.

4.3    Analyzing Homelessness Data Collected Through Medicaid and TANF Applications

In most cases, homelessness data that are collected by states through Medicaid and TANF applications are stored in statewide databases.  However, only five states (California, Colorado, Connecticut, New York, and Washington) report having conducted any analysis of the available data.  Moreover, in most cases the analysis has been fairly limited.  For example, in Washington, the TANF and Medicaid officials have prepared reports to the Governor about the number of homeless persons applying for benefits, using the data collected from TANF and Medicaid applicants.  In Connecticut, state officials have done some analysis of the information collected as part of a pilot program across the state to examine homelessness.  This information has been used in analysis of impacts of state efforts to alleviate homelessness and to link people discharged from prison to services.  The homelessness information collected from TANF and Medicaid applicants serves as a benchmark against which the impacts of other state efforts can be measured.  In Colorado, the TANF and Medicaid contacts we spoke to were not able to provide details about the nature of the analysis conducted with the homelessness data collected from TANF and Medicaid applicants.  More detailed information was available from California and New York and this is summarized below, after Table 13.

When asked if the state would be willing to grant access to the homelessness data collected through the applications, most states (53 percent of Medicaid programs and 60 percent of TANF programs that collect homelessness data) indicated that they would be willing to provide de-identified extracts of homelessness data to HHS.  (See Table 13, noting that percentages shown in the table include missing responses and programs that do not collect homelessness data).  Even among the willing states, however, there were some concerns about resource constraints for responding to customized data requests and concerns about the reliability of the data available since no previous assessment of the information has been done.  Several states were unsure whether they could provide data extracts, and some stated that they would only do so if required.

Table 13: Potential for using Homelessness data to analyze access to Mainstream programs for homeless individuals

Potential for using Homelessness data to analyze access to Mainstream programs for homeless individuals	Medicaid (N=51)		TANF (N=51)
	N	%	N	%
Can the TANF and Medicaid databases be queried to provide statistics about homelessness among program applicants?
Yes	9	18%	11	22%
No	4	8	3	6
Missing or not asked (question only asked of states that collect homelessness data)	38	75	37	73
Would the state consider providing de-identified data extracts or aggregate reports to HHS on homelessness among TANF and Medicaid applicants?a
Yes, or if requested	16	31%	20	39%
No	4	8	3	6
Unsure, not sure it is possible, or no reports built into the system at this time, or no ability to aggregate data	10	20	10	20
Missing or not asked (question only asked of states that collect homelessness data)	21	41	18	35
a Percentages shown in this table include missing responses.  The questions were asked only of states that report collecting homelessness data as part of the application for Medicaid and TANF benefits but not all of them responded.  In the second panel of the table, 30 Medicaid programs and 33 TANF programs responded to the question with 53 percent of Medicaid programs and 60 percent of TANF programs indicating that they would consider providing extracts or aggregated data to HHS. Source:  2008 Interviews with TANF and Medicaid state officials.

New York

On the combined application for TANF and Medicaid, the State of New York asks applicants if they are homeless, if they reside in a shelter, and if they reside in a domestic violence shelter.  This information, as well as information gathered during interviews with TANF applicants and during intake, can be used to assess the number of TANF applicants and recipients whose residential status indicates homelessness.  This information is entered into the TANF management system and can be queried to produce reports.  Reports are usually prepared by districts or counties and any reporting of homeless information would be done on an ad hoc basis  there is no established schedule or protocols for producing reports on homelessness.  The TANF agency often produces reports in response to requests from the TANF policy office or from the state legislature or governors office.  The TANF system is used to generate regular bi-monthly reports on TANF applications, but the reports are not related to homelessness.

Occasionally reports are generated from the TANF system to assess changes in service needs or to make projections about future service needs and budget implications.  For example, given increases in the price of heating oil, the state anticipates an increase in demand for its energy assistance program for the coming winter.  The governors office needs to determine how much state funding will be needed to supplement federal funds for the program and reports from the TANF system can help project demand for the program.  Similar reports to predict demand for homelessness services have not been generated, but one could foresee a desire to do so in the future.  Recent increases in the number of mortgage foreclosures could, for example, result in increased demand for housing services, and there may be data in the system that could be tapped to analyze the potential need for services among program applicants.  Typically, reports produced from the TANF system are not made available to the public.  Customized reports based on data available in the system could be generated, and state officials appear willing to consider this.

California

Families who are eligible for Californias TANF Program (CalWorks) and who are homeless can apply for a special need payment called a Homeless Assistance Payment to meet their costs for temporary and permanent housing.  The maximum Homeless Assistance Payment is based on the size of the family.  Homeless assistance is available only once in a lifetime unless the family is homeless as a result of domestic violence, natural disaster, uninhabitability of a residence, or a physical or mental disability.  Families who receive a homeless assistance payment are noted in the CalWorks data system. The state reports on the number of payments made each month.  This information is an indicator of which TANF recipients have experienced homelessness at some point but cannot be used to track trends in homelessness for individual families.  The information collected on the combined application is not analyzed separately, but is stored in the TANF applicants record in the system and is used in determining eligibility for the homeless assistance payments. The interview respondent reported that generating other more customized reports on homelessness would be difficult given the cost of doing so.

Chapter Five. Investigating Homelessness Data Collection in Other Mainstream Programs

In addition to Medicaid and TANF, HHS identified nine other mainstream programs that may serve persons experiencing homelessness.

• Community Services Block Grant (CSBG)  Administered by the Administration for Children & Families (ACF)
• Head Start  Administered by the Administration for Children & Families (ACF)
• Social Services Block Grant (SSBG)  Administered by the Administration for Children & Families (ACF)
• State Childrens Health Insurance Program (SCHIP)  Administered by the Centers for Medicare & Medicaid Services (CMS)
• Health Center Program  Administered by the Health Resources & Services Administration (HRSA)
• Maternal and Child Health Services Title V Block Grant (Title V)  Administered by the Health Resources & Services Administration (HRSA)
• Ryan White HIV/AIDS Program (Ryan White)  Administered by the Health Resources & Services Administration (HRSA)
• Community Mental Health Services Block Grant (CMHSBG)  Administered by the Substance Abuse & Mental Health Services Administration (SAMHSA)
• Substance Abuse Prevention and Treatment Block Grant (SAPTBG)  Administered by the Substance Abuse & Mental Health Services Administration (SAMHSA)

We examined each of these programs to determine what information is collected from or about program participants, with a focus on data related to homelessness.  This chapter describes the results of this review, and explores the extent to which it may be feasible for HHS to use existing program data to estimate the number of homeless people served in these nine programs.  We also identified three possible approaches that HHS might use to obtain information on the number of homeless persons using these programs and the relative costs of each of these approaches.

The chapter is organized as follows.  Section 5.1 provides a brief description of the methodology used for this task and the sources of information explored.  Section 5.2 summarizes how the nine mainstream programs currently collect information about the homelessness status of program beneficiaries, and the extent to which HHS may be able to use these data to understand the accessibility of its mainstream programs for people who are homeless.  Section 5.3 describes additional approaches that HHS might use to estimate the number of homeless persons using each of these programs in cases where a program does not collect data on beneficiary homelessness status.  Appendix E includes profiles of each of the nine mainstream programs.

5.1    Methodology

The first step was to collect background information (primarily through internet searches) for each program, including the services funded, eligibility requirements, whether programs are administered at state or sub-state level, and whether the programs report homelessness data in state or federal reports.  The information collected is referenced throughout this chapter and is summarized in program profiles for each program in Appendix E.  We augmented information collected via internet searches with interviews of federal officials who administer each of the programs.  In some cases, the federal officials provided recommendations of other state and non-Federal agencies for Abt to interview.[22] For example, officials at the Administration for Children and Families (ACF) who administer the Community Services Block Grant program (CSBG) referred us to the National Association of State Community Services Programs (NASCSP) to learn more about the CSBG program.  NASCP collects and maintains data on CSBG grantees activities using a system called Results Oriented Management and Accountability (ROMA) and prepares reports on performance outcomes for the CSBG program.  The Abt team reviewed the extensive information about CSBG on NASCSPs website and interviewed their staff about ways in which states use CSBG funds to support services for homeless persons, as well as the type of data states collect and report.

We pursued similar information gathering strategies with the remaining eight programs.  For instance, SAMHSA has adopted an extensive process of aligning policy, program, and resource allocation with priority areas and cross-cutting management principles,the process is collectively referred to as The Matrix. [23] As part of the Matrix, SAMHSA adopted an Action Plan for Homelessness and established a Homeless Matrix Workgroup to coordinate and track Matrix efforts related to homelessness.  SAMHSA has adopted National Outcome Measures (NOMs), which track the Departments progress.  The long-term Housing Stability performance measure for SAMHSA related to homelessness is to increase the number of individuals who are receiving mental health, substance abuse prevention, or substance abuse treatment services, have entered into stable living conditions, and show improved functioning in the community.[24]  Materials on SAMHSAs website, such as the Homeless Action Plan, NOMs, and other data sets were an important part of our research on the SAMHSA mainstream programs.

In developing possible approaches that HHS could use to understand the incidence of homelessness among mainstream program beneficiaries, we considered several options, based on information gathered in interviews with mainstream program administrators, Abts knowledge of these options from other research studies, and internal discussion.  Two appear most promising for further consideration.  The first option would be to match individual program beneficiary[25] identifiers with other benefit programs that collect homelessness data.  An example of this would be to search for mainstream program participants in a data system, such as SNAP records,[26] where homelessness data is collected.  These data would require further extrapolation to account for people who are not represented in the other data system. Another approach would match program beneficiary identifiers with local Homeless Management Information System (HMIS) data.  A second option would entail directly surveying a sample of program beneficiaries.[27]

5.2    Data on Homelessness Status

All nine of the mainstream programs reviewed for this project serve, or potentially serve, persons who are homeless.  However, all of the programs are administered by state or other non-Federal agency applicants, and frequently the applicant for the funds contracts with other entities at the community level to actually deliver the services.  In many cases, HHS does not interact directly with the population served by the programs.  To understand who is served by a program, the Department relies on program beneficiary data submitted by grantees, who may in turn obtain the data from another entity that is delivering the services at the community level.  Thus, the extent to which HHS can learn about the homelessness status of program beneficiaries depends on the specificity and quality of data collected on homelessness by grantees or their subgrantees.

The nine programs can be grouped into the following three categories based on the nature of the data reported to HHS:

1. programs that report client-level data on homelessness status;
2. programs that report aggregate data on the number of beneficiaries who are homeless, and;
3. programs that do not report any data on beneficiary homelessness status.

Each category and the potential uses of these data are described.

National Client-level Data

As we reviewed each program, we documented whether the program uses a national client-level dataset.  A system that records data on the homelessness status of program beneficiaries would allow HHS to understand the magnitude of homelessness among the program beneficiaries, and also to analyze homelessness status in relation to other beneficiary characteristics and outcomes.  Client-level datasets that include personal identifiers that can be matched to other data systems would also allow HHS administrators or researchers to understand whether the same persons are receiving services from more than one provider and/or in multiple communities over time.  The availability of national client-level data would allow HHS to develop a more accurate understanding of treatment needs, utilization, and unduplicated[28] numbers of persons served.  Another advantage of national client-level datasets is that HHS can use the data to answer questions that were not anticipated at the time of data collection, since researchers would have access to the full set of client-level data.

Only one of the nine programs reviewed for this study provides client-level data on homelessness status to HHS.  Grantees of the Substance Abuse Prevention and Treatment (SAPT) Block Grant are required to report admission and discharge data for all persons served by the program using a national database known as the Treatment Episode Data Set (TEDS).  Initially, TEDS was designed to generate annual statistics on the number and characteristics of persons admitted to public and private substance abuse treatment programs receiving public funding, including living arrangements at admission and discharge.  The living arrangement variable collected in TEDS includes three possible response codes:  1) homeless indicating no fixed address (including shelters); 2) dependent living (this refers to living in a supervised setting such as a halfway house or group home) and; 3) independent living (this refers to living alone or with others without supervision including children living with parents, relatives, guardians, or in foster care).[29]  TEDS currently provides descriptive information to HHS about SAPT beneficiaries and serves as a source for SAMHSAs National Outcome Measure (NOM) for Housing Stability.  It is unclear whether TEDS includes identifiers that could be used to assess patterns of service utilization across SAPT-funded providers.

Aggregate Data on Number of Homeless Beneficiaries

Four of the mainstream programs report aggregate data to HHS on the number of homeless program beneficiaries served: Health Center program, Head Start, Ryan White, and Community Mental Health Services Block Grant (CMHSBG).  Aggregate reports present responses to predefined questions about program beneficiaries and program results.  For instance, aggregate reports frequently provide the number of people served, their demographic characteristics, their housing/homelessness status, and the types of services they received.  Aggregate reports protect client confidentiality and provide consistent information from one grantee to another, but are limited only to those questions that were defined on the report and rarely correlate client characteristics with services received or outcomes.  An additional limitation that should be noted is that information reported across programs may not be consistent.

Grantees of these four programs use their own locally-managed administrative data systems to generate the required reports, and grantees all report beneficiary housing/homelessness status using this model.  For example, it is our understanding that Health Centers enter aggregate patient-level data in the Uniform Data System (UDS) that includes data on patient characteristics, housing status, revenue/insurance, and services provided.  Head Start programs report on the number of homeless families served at any point during the school year.  Head Start grantees are also expected to track whether the families were able to obtain housing.  The Ryan White HIV/AIDS Program grantees report aggregate data on the number of homeless persons served each year; however, homelessness status is not required or applicable to all services funded by the Act.  Therefore, the data on the number of homeless beneficiaries served may not represent all program beneficiaries.  HRSA staff indicated that Ryan White grantees will be required to report client-level data, including a question on homelessness, beginning in 2009.  More information was not available on what will be feasible with the new system, since the client-level reporting format is not yet finalized. Additionally, CMHSBG grantees annually submit aggregate implementation reports that include the number of homeless persons served by community mental health programs by age and gender.

The reports from each of these four programs help HHS estimate the number of homeless persons served by each program; however, they each have limitations.  Since the figures are aggregate, HHS cannot ensure that the numbers are unduplicated across grantees, or whether the questions are interpreted and calculated consistently.  In addition, information is limited to the tables or questions asked in a standard report template.  In other words, SAMHSA can learn the number of homeless persons served by age and gender, since they specify that cross-tabulation in the report template; however, staff cannot understand homelessness in relation to any other characteristic or outcome, nor can it look at periods of time other than the period defined for the complete report.  Nonetheless, estimates from aggregate reports can help HHS understand the extent to which their programs are accessible to persons who are homeless.  Additional instructions could be provided to grantees to increase the likelihood that responses represent similar information, and standard report templates could be modified to add additional cross-tabulations; however, these modifications would need to be phased in over time to allow adequate time for grantees to train staff and to change their data collection systems.

It may be possible to use the data maintained by grantees at the state or local level to learn more about the homeless beneficiaries served.  Local data systems offer the same opportunities described above under the national client-level reporting database.  Data can be analyzed beyond the predefined reports, and homelessness status can be correlated with other patient characteristics and outcomes.  For instance, if HHS were interested in learning more about Health Center beneficiaries, the Department might be able to enter into a data use agreement that allows UDS data from a statistically reliable sample of grantees to be compiled as part of a protected national research database.

No Data on Homelessness

The four remaining programs: the Community Services Block Grant (CSBG), Social Services Block Grant (SSBG), Maternal and Child Health Block Grant, and the State Childrens Health Insurance Program (SCHIP), also provide HHS with regular aggregate reports, however, these reports do not include information on the number of program beneficiaries who are homeless or other data related to housing status.  Of the four programs in this category, CSBG, SSBG, and the Maternal and Child Health Block Grant programs have very limited quantitative reporting requirements.  Generally speaking, block grant programs are designed to give maximum flexibility to the state or community to determine how to utilize the funding to best suit their regional needs. Instead of using standardized report templates or tables, states submit annual narrative reports on the activities funded by the grants.  Alternatively, the SCHIP program has quarterly and annual reports in which grantees report data on the unduplicated count of new enrollees and data tied to program eligibility.

The CSBG, SSBG, and Maternal and Child Health Block Grant programs do not appear well-suited to collect meaningful data on homelessness because of their block grant format, wide-ranging uses, and localized delivery mechanisms (see Appendix E for details about program activities and services provided).  Though many activities funded by each of these programs are directed to specific clients, including people experiencing homelessness, others may be offered more generally in the community, such as prevention or system coordination services.  Reporting requirements for direct service activities could be modified to collect data on homelessness; however, more detailed requirements may not be easy to implement at the local level without major changes for grantees and substantial burden for beneficiaries.  For instance, it may not be reasonable for a family that receives lunch at a soup kitchen funded by one of the above programs to provide complete demographic and household status data.

5.3    Approaches to Estimate the Number of Homeless Beneficiaries

Based on the information we have gathered from the nine mainstream programs included in this study, we conclude that existing reporting and data collection practices in four programs; CSBG, SSBG, SCHIP, and the Maternal and Child Health Block Grant, are not sufficient for HHS to determine the homelessness status of participants in these programs.  Substantial effort would be required to construct estimates of the number of persons experiencing homelessness who are accessing these programs.  For the five mainstream programs that do collect data on homelessness, HHS would need to consider approaches to validate the estimates of persons experiencing homelessness, as the numbers reported in the existing data may not be entirely reliable or comprehensive.  For instance, program administrators reported that definitions and methods for documenting homelessness often varied from state to state and even within states, so homelessness may not be recorded consistently from one aggregate report to another.  In addition, program data may only reflect fixed points of time, such as the housing status of program participants at the time of application, and therefore, would not convey information about the numbers of program beneficiaries who experience homelessness at some point during the reporting period or upon ending program participation.

Based on our review of the nine mainstream programs, we have identified two primary approaches that HHS could consider for understanding the homelessness status of program participants:

1. matching program data with other administrative data sets; and
2. conducting research with a sample of program beneficiaries.

The advantages and disadvantages of each of these are described below.  A third option, involving revising data collection and reporting requirements for program grantees, was considered but is not suggested. Revising grantee data collection and reporting requirements is something that would most likely require legislative or regulatory changes.

Approach 1: Matching Program Data with Other Administrative Data Sets

The first approach would allow HHS to match program data gathered from the nine mainstream programs with other administrative data sets.  Using this approach, HHS or its grantees would link identifiable client-level data from the mainstream program administrative data set with identifiable data from another database that is considered a more reliable source of data on homelessness.  Even if reliable data are not available in all states, there may be data from a sufficient number of states that HHS could generate reasonable national estimates of the number of homeless beneficiaries served by each mainstream program.

In the telephone interviews we conducted with TANF and Medicaid program officials, the study focused only on data collected about program applicants at the point of initial application, even though recipients of the programs will receive benefits over a period of time, during which the housing status of the applicant could very well change.  Homelessness is typically a transient state, meaning that a program applicant could be homeless at the time of application, but housed shortly thereafter.  Likewise, a program applicant may be housed at the time of application, but become homeless at some period of time while receiving benefits.  If the application for different programs were completed at different points of time in the same individuals life, the housing status response could very well be different for both programs.  The relative efficacy of each of these suggested approaches depends in part on how well this issue is addressed.

There are at least three sources of matching data that might be viable for this approach.  The first is data collected by states for the SNAP, administered by the USDA.  For example, a query of SNAP and TANF records in the Commonwealth of Virginia in 2007 revealed that approximately 83 percent of current TANF recipients in Virginia also received assistance through SNAP. In this example, matching individual-level SNAP data (which contain information on homelessness) with TANF records would allow for a reasonably reliable assessment of the incidence of homelessness among TANF recipients since there is such a high degree of overlap between TANF and SNAP participants.

There may be similar overlap between persons participating in the other HHS mainstream programs.  If so, HHS can link records between the mainstream program and the SNAP data and can check the homelessness status of mainstream program beneficiaries by analyzing the appropriate field within the SNAP data.  With this approach, HHS would only be able to confirm the homelessness status for those mainstream beneficiaries who are found in the other programs data set (i.e. an individual receiving both TANF and SNAP).  Methods would need to be developed to impute homelessness status for individuals who participate in only one program, and thus do not appear in the linked data set.  These methods could include applying the same proportion of homelessness rates for those identified to those that are not, or taking into account the homelessness status of SNAP recipients with similar demographic and income characteristics, or other statistical pairings.[30]  However, individuals receiving only one type of benefits may have quite different probabilities of homelessness given their ineligibility for one of the programs and any method developed to examine this issue would need to take this into consideration.  For those states that collect data on the homelessness status of their TANF and Medicaid applicants, it may be possible to match mainstream program beneficiary data against the TANF or Medicaid datasets.[31]  This type of administrative system matching is becoming more readily achievable as states, such as South Carolina and Michigan, establish state-wide data warehouses that integrate client-level data across benefit programs.

Other potential sources to match against are the local Homeless Management Information System (HMIS) databases that are maintained by local communities across the country to record identifiable client-level information on an ongoing basis about all homeless persons who access homeless assistance programs within a city, county or region.  The U.S. Department of Housing and Urban Development (HUD) requires every community that receives HUD McKinney-Vento funding to participate in an HMIS and collect a set of standardized data elements on each individual served.  The HMIS may be administered at the local level or may be implemented at a regional or state-wide level.  By definition, every person entered in the HMIS is homeless; thus, by matching program data with the HMIS the grantee can get an accurate count of program beneficiaries who are homeless during the reporting period, and possibly use the data to understand whether the homeless episode overlapped with the receipt of benefits or services from the mainstream program.  That is, the matched data can indicate the percentage of beneficiaries who were homeless before receiving services, during the period in which they received services, and following the receipt of HHS-funded services.

The matching approach offers many advantages.  Matching provides the data on homelessness without imposing additional data collection requirements on grantees.  There would not be a lag in collecting data, since this analysis can be conducted retrospectively.  Once the data are obtained, matching datasets is a relatively efficient, low-cost method of estimating homelessness, and the data should be available for most, if not all, parts of the country.  Matching will also produce strong estimates, assuming the source data are reliable.  Since the SNAP and HMIS datasets contain data on homelessness for purposes that directly relate to grant management or program eligibility, the source data are likely to be more reliable than descriptive program data collected by one of the mainstream programs.  Furthermore, HMIS data offers the advantage of tracking homelessness over time, yielding longitudinal estimates of homelessness rather than the static point-in-time data collected at program application by HHS-funded programs.

There are several disadvantages to using this approach.  Obtaining access to administrative data on homelessness can be time-consuming and onerous, since datasets may be managed by different parties.  As well, privacy and confidentiality concerns may hamper access to data, though these are generally surmountable with time and continued negotiation.  In some communities, HMIS databases are still in their infancy, and may not yet comprehensively represent all homeless persons in the community.  Matching administrative data also relies on the existence of identifiers for program beneficiaries (e.g., Name, Date of Birth, Social Security Numbers), which may not be collected consistently by mainstream programs, especially block grant programs.

Before pursuing this approach, HHS would need to determine whether the mainstream program of interest collects personal identifiers, and whether it is feasible to identify and obtain appropriate data sources to match against the program database.  If HHS pursues this approach, the Department may want to consider using a sample of grantees (e.g., states) and extrapolating to represent program beneficiaries from all grantees.  This approach appears to be the most efficient and lowest cost.

Approach 2: Conducting Research with a Sample of Program Beneficiaries

The second option would involve collecting data directly from a representative sample of program beneficiaries.  Using this approach, HHS could work with grantees to identify standard types of services delivered, and to select a sample of sites and program beneficiaries from each site from which to collect data.  If grantees or service providers collect more comprehensive data than is reported to HHS, it may be possible to identify homelessness based on case file reviews for the sample.  Alternatively, it may be necessary to work with grantees to collect data from a sample of beneficiaries over a specified period of time.

This approach would likely produce reliable estimates, and both the approach and data collection could be tailored to answer the research questions of most interest more precisely, as compared to other approaches that would require the Department to work with existing data or data that can be reasonably collected by grantees and reported in aggregate.  For example, a research approach would allow HHS to explore the relationship of beneficiaries homelessness status to other client demographics or outcomes, and could support follow-up with beneficiaries to understand changes in housing status over time.  One-time data collection directly from beneficiaries may also be the most feasible approach for block grant programs that may not be able to accommodate more intensive, permanent reporting requirements.

The primary disadvantages to using this approach are that it could be resource-intensive and would impose a burden on grantees and beneficiaries.  The study procedures and consent process would require review by an Institutional Review Board (IRB) and Office of Management and Budget (OMB) clearance.  In addition, this approach would be limited to a one-time or periodic study; whereas other approaches could be replicated more readily over time.

HHS may want to consider this approach if it wants flexibility and has resources available to gain a point-in-time estimate of the homelessness status of program beneficiaries.

5.4    Summary

This chapter explores two methods for understanding the number of mainstream program beneficiaries who are homeless.  Five of the nine programs examined (SAPTBG, Head Start, the Health Center Program, Ryan White, and CMHSBG) already report data annually on the number of homeless persons served.  Each of these current data sources yield a limited amount of information about the homelessness status of program beneficiaries, and there may be an opportunity to conduct further analysis on these data sources to learn more.  The SAPTBG TEDS data offers the most promise, since HHS already receives client-level data; however, HHS could explore the possibility of obtaining access to client-level source data for a representative sample of grantees to learn more about the beneficiaries of other programs.  This strategy seems both feasible and fruitful to explore further, since the data are already being collected.  The other approaches presented in this chapter could be used to validate or improve the quality of these data.

The merits and limitations of the approaches are discussed in Section 5.3 as they relate to collecting data on program beneficiaries of the other four mainstream programs (CSBG, SSBG, Maternal & Child Health Block Grant, and SCHIP) that do not currently report data on homelessness status to HHS.  These approaches may be feasible, but they each offer advantages and disadvantages, and there is not a single option that would work unambiguously for all four of the programs.  As was briefly discussed above, the CSBG, SSBG, and Maternal & Child Health Block Grant programs fund an extremely diverse set of activities.  Some activities are targeted to homeless persons, but others are not.  The fact that it would not be feasible to remove duplicate entries for the same individual from one program to another (since data for these programs are available only in aggregate) limits the utility of having such data.  In contrast, programs like SCHIP, that collect client-level data on insured beneficiaries, are attempting to enroll all persons who are eligible within the grantees jurisdiction.  Therefore, it seems more feasible to consider collecting new aggregate or beneficiary-level data from the SCHIP program, and presumably this data would be meaningful to understand the extent to which SCHIP reaches beneficiaries who experience homelessness, in conjunction with Medicaid and the Health Center program.

With each approach, HHS will need to carefully weigh the desire to better understand the population enrolled in the benefit programs with the desire to minimize reporting burdens for the states and entities that operate the programs, as well as for the recipients of the programs themselves.

Chapter Six. Summary

This study investigated which states (including the District of Columbia) currently collect housing status data from applicants of Medicaid and/or Temporary Assistance for Needy Families (TANF).  The study is a first step in the exploration of available data that could be used to identify the number of homeless persons currently accessing HHS mainstream programs.

Data were collected from a survey of the 50 states and the District of Columbia regarding application data collected for the TANF and Medicaid programs, and a review of practices in nine other HHS mainstream programs related to homelessness data collection.  The focus of the survey was on state practices related to the collection of housing status data of program applicants and enrollees; not the collection and/or analysis of state data.

Interviews with Medicaid and TANF program officials indicate that more than half of the states and the District of Columbia collect information on homelessness indicators and risk factors for applicants to these programs.  In addition, several states that do not explicitly collect this information on the application nonetheless address homelessness with program applicants through interviews with eligibility workers that are part of the application process.  Only 13 states do not collect any homelessness data at all from Medicaid applicants and 10 do not collect homelessness data from TANF applicants.

Despite the fact that many states have procedures in place to ask program applicants about their housing and homelessness status, questions remain as to whether these data are readily available for analysis or could easily be accessed to benchmark the participation of homeless individuals and families in the TANF and Medicaid programs.  Questions also exist about whether such data are adequate to answer the policy questions of most interest to HHS.  For instance, some states that collect homelessness data either do not record it in state databases or, if it is recorded, have not taken steps to verify or confirm the information and have not used the data for reporting or analysis.  Only five states (California, Colorado, Connecticut, New York, and Washington) report having conducted any analysis of the homelessness data collected.  Nonetheless, this group of five includes the two largest TANF and Medicaid programs, which represent a significant portion of the nations TANF and Medicaid applicants, and analysis of all five may provide potential for learning about the feasibility of gaining access to and the quality of homelessness data for other states.

It is likely that data on program applicants is inadequate for examining access to mainstream programs in a rigorous manner, as the data provide no information on the individuals who do not ever apply for benefits and the proportion of the homeless population they may represent.  Data on program applicants is also not expected to be adequate for assessing the effectiveness of program services for families who are homeless since housing status can be expected to change over time and data on program applicants cannot be used for longitudinal tracking of housing status.

States that collect homelessness data typically do so in order to confirm the living arrangement and address for applicants in order to process benefits; to facilitate eligibility determination and administration of the SNAP (one of the programs usually covered on combined applications in use in the states); and to understand the service needs of program applicants in order to make referrals to needed services.  For TANF programs, homelessness is often assessed as a barrier to work and case workers attempt to help families experiencing homelessness get housing assistance and services to alleviate their homelessness and make it possible to meet the TANF work requirements.  The data tend not to be collected for the purposes of reporting on homelessness. However, we hypothesize that because the data are collected primarily for operational purposes, the quality of the data may actually be better than data collected for reporting purposes only, but this hypothesis would require further investigation.

Only a small number of states have done some ad hoc analysis of the data to produce estimates of the number of homeless individuals applying for and then receiving TANF and Medicaid benefits.  HHS could focus efforts on those states that have done some preliminary analysis to pursue additional data assessment and data collection.  One encouraging finding from the study was that the majority of states do not appear to consider it burdensome or costly to collect homelessness data, and for the most part states appear to be willing to consider making extracts of the homelessness data collected available for research purposes.  Nevertheless, many states expressed concerns about available resources for the creation of data extracts and competing priorities, suggesting that providing data extracts might be difficult.  A preferable approach might be to focus on states already conducting analysis of homelessness data to explore the feasibility of larger scale analysis of homelessness in TANF and Medicaid programs.

Understanding more about the extent to which the eleven HHS mainstream programs chosen for study in this project collect information on homelessness is an important first step in assessing the number of homeless individuals and families who access mainstream programs.

[ Go to Contents ]

References

Administration for Children & Families.  Fact Sheet: Office of Family Assistance (OFA).  U.S. Department of Health and Human Services.  http://www.acf.hhs.gov/opa/fact_sheets/tanf_factsheet.html (accessed August 5, 2008).

Center for Medicaid Services.  FirstStep, An Introduction to the Food Stamp Program, http://www.cms.hhs.gov/apps/firststep/content/foodtips.html (accessed August 5, 2008).

Cline, Terry L.  A Life in the Community for Everyone: SAMHSA.  SAMHSA. http://www.samhsa.gov/Matrix/matrix_brochure.pdf (accessed August 7, 2008).

Coverage of the TANF Population Under Medicaid and SCHIP.  Hertz, Elicia J. CRS Report RS22035.  Washington: Library of Congress.  Congressional Research Service, 2005.

U.S.C. Title 42, Chapter 119, Subchapter 1.

Food Research and Action Center.  Homeless Persons Rights Under the Food Stamp Program.  FRAC, http://www.frac.org/html/federal_food_programs/FSP/Homeless_Rights_Under_FSP.html (accessed August 5, 2008).

Holcomb, Pamela A., Karen Tumlin, Robin Koralek, Randy Capps, and Anita Zuberi.  The Application Process for TANF, Food Stamps, Medicaid, and SCHIP: Issues for Agencies and Applicants, Including Immigrants and Limited English Speakers.  Washington, DC: Urban Institute, prepared for the Department of Health and Human Services, 2003.

National Coalition for the Homeless. McKinney-Vento Act: NCH Fact Sheet #18,  NCH. http://www.nationalhomeless.org/publications/facts/McKinney.pdf (accessed August 5, 2008).

National Association for the Education of Homeless Children and Youth.  FACT CHECK:  Updating HUDs Definition of Homelessness in the Reauthorization of HUD McKinney-Vento Act Programs.  NAEHCY, http://www.naehcy.org/dl/defmyth.pdf (accessed August 5, 2008).

Parry, Elaine.  SAMHSA Action Plan: Homelessness, Fiscal Years 2006 2007. SAMHSA.  http://www.samhsa.gov/matrix/SAP_homeless.aspx (accessed August 7, 2008).

Sara Simon Tompkins.  Homelessness and Food Stamps: Rights and Barriers.  National Law Center on Homelessness & Poverty.  http://www.nlchp.org/content/pubs/Homelessness%20and%20Food%20Stamps%2020031.pdf (accessed August 5, 2008).

StateHealthFacts.org. State Health Facts.  The Henry J. Kaiser Family Foundation. http://www.Statehealthfacts.org (accessed August 1, 2008).

Social Security Advisory Board, May 2006 SSAB Report.

U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation.  Homelessness:  Barriers to Access in HHS Nontargeted Programs, 2003^a.

U.S. Department of Health and Human Services.  Ending Chronic Homelessness:  Strategies for Action, 2003^b.

U.S. Department of Housing and Urban Development, Office of Community Planning and Development.  The Second Annual Homeless Assessment Report to Congress, 2008.

U.S. Government Accountability Office, GAO/RCED 99-178: Homelessness:  State and Local Efforts to Integrate and Evaluate Homeless Assistance Program, 1999.

U.S. Government Accountability Office, GAO/03-119: High-Risk Series: An Update, 2003.

[ Go to Contents ]

Appendix A State Descriptions

[ Go to Contents ]

Appendix B Medicaid and TANF Program Names

[ Go to Contents ]

Appendix C Interview Procedures and TANF and Medicaid Survey Instruments

Interview Procedures

We collected information from the 50 states and the District of Columbia regarding data collection practices for both TANF and Medicaid applications.  For states that use separate TANF and Medicaid applications, we conducted interviews with representatives of each program.  For states using both separate Medicaid and combined applications, we typically conducted one interview with a Medicaid representative.  During the interview, the staff person was asked to confirm whether the responses also applied to the TANF program.  If the Medicaid representative was unable to confirm this, Abt interviewers also conducted an interview with the TANF representative. Generally for states with combined applications, only one interview with either the TANF or Medicaid representative was needed to collect the survey data for both programs.  During the interview with the initial agency, we confirmed with the respondent that the information provided applied to both the Medicaid and TANF program.  If the respondent was unsure, an interview was also conducted with the other program representative.  If only one interview was conducted in a state, the program representative of the other program received a courtesy call to let them know we conducted an interview and collected all the necessary information from their counterpart.  In some cases, both the Medicaid and TANF representatives participated in the interview in a conference call setting. And in one case (District of Columbia) one individual was responsible for both TANF and Medicaid program application and provided information for both programs in one interview.

Identical interview instruments were used for the Medicaid and TANF interviews, and the data from the interviews were entered into a database for each type of program.  In cases when interviews were conducted with both TANF and Medicaid program officials (for example, when separate TANF and Medicaid applications are used in a state), separate records were entered for TANF and Medicaid.  In cases where only one interview was conducted with either a TANF or Medicaid representative (this was done when a combined application was used in a state and the responses pertaining to Medicaid and TANF applicants are identical), data were entered in to the database corresponding to the program providing the interview data, and then replicated in the other program database.  In some cases, when states use both a combined application and a separate Medicaid application, responses regarding the separate Medicaid application differ from the combined application.  In those cases, data pertaining to the separate Medicaid application were recorded in the Medicaid database and responses pertaining to the combined application were recorded in the TANF database.

Interview Guide for TANF State Directors or other TANF application experts

This version to be used for programs that do include questions about homelessness or risk factors associated with homelessness on the TANF or combined applications

State: _____________________________________________________

Name of Respondent:________________________________________

Phone number: _____________________________________________

Email address of Respondent:_________________________________

Interview conducted by:_______________________________________

Date of interview:     ______/______/______    

                                   Month Day       Year

According to the Paperwork Reduction Act of 1995, no persons are required to respond to a collection of information unless it displays a valid OMB control number. The valid OMB control number for this information collection is 0990-0320. The time required to complete this information collection is estimated to average 1 hour per response, including the time to review instructions, search existing data resources, gather the data needed, and complete and review the information collection. If you have comments concerning the accuracy of the time estimate(s) or suggestions for improving this form, please write to:  U.S. Department of Health & Human Services, OS/ocio/PRA, 200 Independence Ave., S.W., Suite 537-H, Washington D.C. 20201,   Attention: PRA Reports Clearance Officer. Alice Bettencourt

NOTE to Interviewer: Review the TANF and/or Combined Application prior to the interview to review the type of homeless data or risk factors for homelessness collected on the application.   Also note the date when Abt obtained the application to confirm that it is the most recent version. 

Introduction: My name is ___________ and I work for a private, independent research firm called Abt Associates.   The office of the Assistant Secretary for Planning and Evaluation at the Department of Health and Human Services is sponsoring this study.  The purpose of the study is to conduct interviews with each states TANF and Medicaid program staff to collect information about the type of housing status and/or homeless status data that is collected on TANF and Medicaid application forms.  Someone from the Abt study team contacted you in October or November 2006 to obtain a copy of your TANF and/or combined application. Today Id like to ask some more detailed questions about your applications and the information you collect from program applicants about whether they are homeless and their housing status in general. Neither the study nor this interview is being used to monitor program performance in any way.  The information you provide will only be used for the purposes of this study.

A. Confirmation that application on hand is current

1.       We collected a TANF application from you (online/via email/copy sent to us) in October 2006.  Is that application still in use?

__Yes

__No  When was the application updated? [ENTER DATE].  Could you please fax me a copy of the application that is now in use or direct me to a copy on your website?  What has changed on the application from the version we collected in October 2006? [ENTER WEB ADDRESSINTERVIEWER: INTERVIEW CAN CONTINUE IF CHANGES TO THE APPLICATION DO NOT INVOLVE QUESTIONS REGARDING HOMELESSNESS.  INTERVIEW SHOULD BE SUSPENDED AND RESCHEDULED IF CHANGES REGARDING HOMELESSNESS HAVE BEEN MADE TO THE APPLICATION.]

2.       [IF WE IDENTIFIED A COMBINED APPLICATION] We collected a Combined benefit application from you (online/via email/copy sent to us) at the same time.  Is that application still in use?

__Yes

__No  When was the application updated? [ENTER DATE]. Could you please fax me a copy of the application that is now in use or direct me to a copy on your website?  What has changed on the application from the version we collected in October 2006? [ENTER WEB ADDRESSINTERVIEWER: INTERVIEW CAN CONTINUE IF CHANGES TO THE APPLICATION DO NOT INVOLVE QUESTIONS REGARDING HOMELESSNESS.  INTERVIEW SHOULD BE SUSPENDED AND RESCHEDULED IF CHANGES REGARDING HOMELESSNESS HAVE BEEN MADE TO THE APPLICATION.]

3.       [IF WE DIDNT IDENTIFY A COMBINED APPLICATION] As we collected application forms from other states, we learned that many states use a combined application covering several programs, either in addition to or in place of a separate TANF application.  Does your state use a combined application for the TANF program?

___No (skip to Section C)

___Yes

B. Clarification regarding use of combined application

4.       Is the combined application used instead of a separate application for TANF or in addition to the TANF application?

___Combined application used instead of a separate TANF application (Ask 4a)

a.       Does that mean that when a family applies for TANF assistance they complete only the combined application?

___Yes

___No (please explain under what circumstances the separate application is used) ________________________________________________________________________________________________________________________________________________________________________________________________

___Combined application used in addition to a separate TANF application (Ask 4b)

b.       Does that mean that when a family applies for TANF assistance they must complete both applications?

___Yes

___No (please explain under what circumstances the combined application is used) ________________________________________________________________________________________________________________________________________________________________________________________________

___Dont know  Is there someone else we could talk to, to find out more about how the combined application is used?  [ENTER CONTACT INFORMATION].

C. Collection of Housing Status, Homeless, or Risk Factors Data

These questions are used to confirm that the correct set of questions is being asked.  

The purpose of this study is to find out whether states routinely collect information about homelessness or risk factors often associated with homelessness for TANF applicants through the regular application process.  Examples of items on an application that might indicate homeless ness are things like living in a shelter, or having no permanent residence.  Risk factors associated with homelessness include things like living with friends or relatives, having an eviction notice, or others.

5.       a. What is the typical process for completing the application? Which is the most common way applications are completed? (CHECK ONLY ONE)

____ application completed online

____ application completed in-person at a TANF office

____ application filled out and mailed to the TANF office

____ Other (SPECIFY)

b.       Do applicants usually fill out the applications themselves without assistance, or do they receive assistance from an intake worker to complete the application? Which is the most common approach? (CHECK ONLY ONE)

____ fill out with assistance

____ fill out without assistance

c.       Is the same application used throughout the state?

____ Yes

____ No

d. Other than in the application form, is additional information on housing status collected at other points in the eligibility determination process? If yes, when?

____ Yes (SPECIFY______________________________________________)

____ No

Wed like to ask you some questions to confirm that our understanding of the types of information included on your TANF/combined application is correct and to find out more about how these data are collected and used, and reasons why your state collects this information.

[ENTER RESPONSES FOR 6-8 IN THE TABLE BELOW.  REVIEW APPLICATON REVIEW DATABASE AND MARK PRELIMINARY RESPONSES FROM THE DATABASE FOR QUESTIONS 6 AND 7 IN THE TABLE BELOW.]

6.       When we reviewed your combined benefit application, we found that you collect [READ DATA ELEMENTS FROM Q6 COLUMN] on housing status, homelessness, and homeless risk factors.  Is our understanding accurate?  Is it correct that you do not collect [READ REMAINING DATA ELEMENTS] on the combined application?  [INTERVIEWER: IF STATE DOES NOT USE A COMBINED APPLICATION, CODE AS NOT APPLICABLE.]

7.       We also determined that you collect [READ DATA ELEMENTS FROM Q7 COLUMN] on your separate TANF application.  Is our understanding accurate?  Is it correct that you do not collect [READ REMAINING DATA ELEMENTS] on the separate TANF application? [INTERVIEWER: IF STATE DOES NOT USE A SEPARATE TANF APPLICATION, CODE AS NOT APPLICABLE]

8.       If a combined application is used, do applicants for all programs covered by the combined application answer each item? Or are there items that are only required for some of the programs included on the combined application?  Please explain.  [INTERVIEWER: IF STATE DOES NOT USE A COMBINED APPLICATION, CODE AS NOT APPLICABLE]

The responses to these questions will be used to determine which set of detailed questions should be asked.  For those programs collecting neither homeless items nor items considered risk factors for homelessness, the no-homeless data questions will be asked.  For those collecting either homeless or homeless risk factor items, the homeless data questions will be asked.

D. Homeless Data Questions

Id like to understand more about the data that you collect for each of the questions that specifically relate to homelessness. If there are different answers for the combined application and separate TANF application, please let me know.  [INTERVIEWER: ONLY ASK ABOUT THE RELEVANT COLUMNS, AS CONFIRMED IN Q6-7.  PROBE FOR WHETHER RESPONSES DIFFER FOR COMBINED APPLICATION AND SEPARATE TANF APPLICATION.]

18.   Do you collect any data on the homeless status of applicants at any time other than in the application?

____  No (Skip to 20)

____  Yes

19.   Do you collect homeless status information at [READ ITEM]?  At each of the subsequent times you mentioned, do you update the information (that is confirm or update the same data items again), or do you collect new information on housing or homeless status?  Is the new or updated information entered into the database?

INTERVIEWER: ASK QUESTIONS 19a, 19b, 19c and 19d FOR EACH TIME PERIOD

[ASK 20-22 ONLY IF RESPONSE TO 6 and 7i: Are you homeless IS NO.  If YES, skip to 23.]

20.   Even if you do not explicitly ask program applicants or participants if they are homeless, do you use any other standard conventions for noting that an applicant is homeless in your database or on application materials?  Examples of this might be noting than an applicant is living with family or friends, has not had a permanent address in recent months, note whether an address given is for a shelter, enter 99999 for zip code when homelessness is suspected, etc.?

____  No (SKIP TO 23)

____  Yes confirm all addresses (would know if address was a shelter)

____  Yes intake workers are instructed to follow certain protocol (e.g., enter 99999 for zip code) for applicants who are thought to be homeless. PLEASE DESCRIBE:

____  Yes Other (DESCRIBE)

21.   Are these conventions formalized in such a way that you would say that they have been adopted throughout the state?  [PROBE: E.g., are they written down in training instructions?  Are they written on the computer screen where intake workers enter the data?]

____  No

____  Yes (SPECIFY)                                                                                       

22.   Do you have the ability to query the program database based on these protocols to produce statistics about homelessness among program applicants?

____  No

____  Yes (please describe how, if you have done it before, and limitations of the results)                                                                                                                                                                                                                                                                                                                                                                                                                                       

23.   Has your state developed a formal definition for homelessness for purposes of this application/program?

____  No (SKIP TO 25)

____  Yes (please tell me what the definition is and how it is used by the state)                                                                                                                                                                                                                                                                       

24.   Is the definition of homelessness clearly documented or widely known to intake workers or others involved with helping applicants complete the application forms?

____  No

____  Yes (DESCRIBE)                                                                                                

                  If yes, SKIP TO 26

25.   Has homelessness been informally defined in the state?

____  No

____  Yes (specify how and please provide the definition)                                                                                                                                                              

26.   Many states have not chosen to include questions on homelessness on their applications for TANF benefits.  We would like to understand more about why leaders in your state believe that it is important to do so.  What are the main reasons?  [INTERVIEWER: RECORD VERBATIM AND THEN RECODE TO ONE OF THE CODES BELOW. IF MORE THAN ONE REASON CHECK ALL THAT APPLY. PROBE AS NEEDED].

____  Program and reporting purposes (DESCRIBE)

____ Collected as part of State effort to enhance access to mainstream benefits for people who are homeless

____ Collected to support development or implementation of State Plan to End Homelessness

____  State Policy Academy requested data or uses it

____  Other (SPECIFY)

27.   Has the state conducted any analysis on the homelessness data yet?

____  No (SKIP TO 32)

____  Yes

28.   Describe what was done, produced, and learned (CHECK ALL THAT APPLY).

____  Estimates of homeless persons applying for benefits

____  Estimates of homeless persons receiving benefits

____  Estimates of impact of benefits on alleviating persons homelessness

____ Studies on costs of mainstream services for people who are homeless

____ Other? (SPECIFY)

29.   How has the state used the results of the study?

30.   Did you alter your data collection requirements, training of intake workers, or analysis/use of data as a result of this study? PLEASE DESCRIBE

31.   Are the data on homelessness accessible to the public? [CHECK ALL THAT APPLY]

____  No (SKIP TO 33)

____  In written reports

____  In state-level data sets

32.   Who has access to the data sets?

33.   Would you consider providing periodic de-identified data extracts or aggregate reports to HHS on homelessness on applicants to your TANF program?

34.   Do you think it takes more time to collect the questions on homelessness than other items on your application?

____  No

____  Yes (DESCRIBE)                                                                                     

35.   Does it take more time to verify, record, or analyze the responses than for other items on the application?

____  No

____  Yes (DESCRIBE)                                                                                     

36.   Are there any additional costs incurred to collect this information?  For instance, does it add to the verification time or effort?

____  No

____  Yes (DESCRIBE)                                                                                     

37.   How frequently is your TANF application updated?

____ Monthly

____ Quarterly

____ Semi-annually (every 6 months)

____ Annually

____ Other (SPECIFY)

38.   Are these updates seen as an opportunity to add or refine questions on homelessness or housing status?

____ Yes

____ No

39.   That is all the questions I have about homelessness data and your TANF and/or combined application. Is there anything else you would like to add that you think is important for this study to take into account?

Thank you very much for your time and help with this study.

---

Interview Guide for TANF State Directors or other TANF application experts

This version to be used for programs that do not have questions about homelessness or risk of homelessness on the TANF or combined applications

State: _____________________________________________________

Name of Respondent:________________________________________

Phone number: _____________________________________________

Email address of Respondent:_________________________________

Interview conducted by:_______________________________________

Date of interview:     ______/______/______

                                   Month Day       Year

According to the Paperwork Reduction Act of 1995, no persons are required to respond to a collection of information unless it displays a valid OMB control number. The valid OMB control number for this information collection is 0990-0320. The time required to complete this information collection is estimated to average 1 hour per response, including the time to review instructions, search existing data resources, gather the data needed, and complete and review the information collection. If you have comments concerning the accuracy of the time estimate(s) or suggestions for improving this form, please write to:  U.S. Department of Health & Human Services, OS/ocio/PRA, 200 Independence Ave., S.W., Suite 537-H, Washington D.C. 20201,   Attention: PRA Reports Clearance Officer. Alice Bettencourt

NOTE to Interviewer: Review the TANF and/or Combined Application prior to the interview to review the type of homeless data or risk factors for homelessness collected on the application.   Also note the date when Abt obtained the application to confirm that it is the most recent version.

Introduction: My name is ___________ and I work for a private, independent research firm called Abt Associates.   The office of the Assistant Secretary for Planning and Evaluation at the Department of Health and Human Services is sponsoring this study.  The purpose of the study is to conduct interviews with each states TANF and Medicaid program staff to collect information about the type of housing status and/or homeless status data that is collected on TANF and Medicaid application forms.  Someone from the Abt study team contacted you in October or November 2006 to obtain a copy of your TANF and/or combined application. Today Id like to ask some more detailed questions about your applications and the information you collect from program applicants about whether they are homeless and their housing status in general. Neither the study nor this interview is being used to monitor program performance in any way.  The information you provide will only be used for the purposes of this study. 

A. Confirmation that application on hand is current

1.       We collected a TANF application from you (online/via email/copy sent to us) in October 2006.  Is that application still in use?

__Yes

__No  When was the application updated? [ENTER DATE].  Could you please fax me a copy of the application that is now in use or direct me to a copy on your website?  What has changed on the application from the version we collected in October 2006? [ENTER WEB ADDRESSINTERVIEWER: INTERVIEW CAN CONTINUE IF CHANGES TO THE APPLICATION DO NOT INVOLVE QUESTIONS REGARDING HOMELESSNESS. INTERVIEW SHOULD BE SUSPENDED AND RESCHEDULED IF CHANGES REGARDING HOMELESSNESS HAVE BEEN MADE TO THE APPLICATION.]

2.       [IF WE IDENTIFIED A COMBINED APPLICATION] We collected a Combined benefit application from you (online/via email/copy sent to us) at the same time.  Is that application still in use?

__Yes

__No  When was the application updated? [ENTER DATE]. Could you please fax me a copy of the application that is now in use or direct me to a copy on your website?  What has changed on the application from the version we collected in October 2006? [ENTER WEB ADDRESSINTERVIEWER: INTERVIEW CAN CONTINUE IF CHANGES TO THE APPLICATION DO NOT INVOLVE QUESTIONS REGARDING HOMELESSNESS. INTERVIEW SHOULD BE SUSPENDED AND RESCHEDULED IF CHANGES REGARDING HOMELESSNESS HAVE BEEN MADE TO THE APPLICATION.]

3.       [IF WE DIDNT IDENTIFY A COMBINED APPLICATION] As we collected application forms from other states, we learned that many states use a combined application covering several programs, either in addition to or in place of a separate TANF application.  Does your state use a combined application for the TANF program?

___No (skip to Section C)

___Yes

B. Clarification regarding use of combined application

4.       Is the combined application used instead of a separate application for TANF or in addition to the TANF application?

___Combined application used instead of a separate TANF application (Ask 4a)

a.       Does that mean that when an individual applies for TANF assistance they complete only the combined application?

___Yes

___No (please explain under what circumstances the separate application is used) _________________________________________________________________________________________________________________________________________________________________________________________________________

___Combined application used in addition to a separate TANF application (Ask 4b)

b.       Does that mean that when an individual applies for TANF assistance they must complete both applications?

___Yes

___No (please explain under what circumstances the combined application is used) _________________________________________________________________________________________________________________________________________________________________________________________________________

___Dont know  Is there someone else we could talk to, to find out more about how the combined application is used?  [ENTER CONTACT INFORMATION].

C. Collection of Housing Status, Homeless, or Risk Factors Data

These questions are used to confirm that the correct set of questions is being asked.  

The purpose of this study is to find out whether states routinely collect information about homelessness or risk factors often associated with homelessness for TANF applicants through the regular application process. Examples of items on an application that might indicate homelessness are things like living in a shelter, or having no permanent residence.  Risk factors associated with homelessness include things like living with friends or relatives, having an eviction notice, or others.

5.       a. What is the typical process for completing an application? Which is the most common way applications are completed? (CHECK ONLY ONE)

____ application completed online

____ application completed in-person at a TANF office

____ application filled out and mailed to the TANF office

____ Other (SPECIFY)

            b. Do applicants usually fill out the applications themselves without assistance, or do they receive assistance from an intake worker to complete the application? Which is the most common approach?

            ____ fill out with assistance

            ____ fill out without assistance

c. Is the same application used throughout the state?

            ____ Yes

            ____ No

d. Other than in the application form, is additional information on housing status collected at other points in the eligibility determination process? If yes, when?

____ Yes (SPECIFY______________________________________________________)

____ No

Wed like to ask you some questions to confirm that our understanding of the types of information included on your TANF/combined application is correct and to find out more about how these data are collected and used, and reasons why your state collects this information.

[ENTER RESPONSES FOR 6-8 IN THE TABLE BELOW.  REVIEW APPLICATON AND REVIEW DATABASE AND MARK PRELIMINARY RESPONSES FROM THE DATABASE FOR QUESTIONS 6 AND 7 IN THE TABLE BELOW.]

6.       When we reviewed your combined benefit application, we found that you collect [READ DATA ELEMENTS FROM Q6 COLUMN] on housing status, homelessness, and homeless risk factors.  Is our understanding accurate?  Is it correct that you do not collect [READ REMAINING DATA ELEMENTS] on the combined application?  [INTERVIEWER: IF STATE DOES NOT USE A COMBINED APPLICATION, CODE AS NOT APPLICABLE.]

7.       We also determined that you collect [READ DATA ELEMENTS FROM Q7 COLUMN] on your separate TANF application.  Is our understanding accurate?  Is it correct that you do not collect [READ REMAINING DATA ELEMENTS] on the separate TANF application? [INTERVIEWER: IF STATE DOES NOT USE A SEPARATE TANF APPLICATION, CODE AS NOT APPLICABLE]

8.       If a combined application is used, do applicants for all programs covered by the combined application answer each item? Or are there items that are only required for some of the programs included on the combined application?  Please explain.  [INTERVIEWER: IF STATE DOES NOT USE A COMBINED APPLICATION, CODE AS NOT APPLICABLE]

Item on application	6. Confirmation that [READ ITEM] is included on your combined application. [Fields should be highlighted in advance based on app review. Code Yes if it is included; Code No if it is not included.]	7. Confirmation that [READ ITEM] is included on your separate TANF application.	8. If combined application is used, do applicants for all programs (TANF, FS, Medicaid, etc) answer the question about [READ ITEM] (if no, please explain)
Housing Status Items
a. Home Address	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
b. Mailing Address	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
c. Directions to home address	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
d. Do you live in public/subsidized housing?	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
e. Do you intend to stay in State?	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
f. Do you live in a long-term care facility or nursing home?	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
g. Do you live in a medical or rehab facility?	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
h. Any other questions on housing status? If yes, PLEASE SPECIFY	__Yes SPECIFY__ __No __NA	__Yes SPECIFY__ __No __NA	__Yes SPECIFY__ __No __NA
Homeless Items
i. Are you homeless?	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
j. Do you reside in a shelter?	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
k. Are you staying in a domestic violence (DV) shelter?	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
l. Do you have a permanent home?	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
m. Other Homeless item? If yes, PLEASE SPECIFY	__Yes SPECIFY__ __No __NA	__Yes SPECIFY__ __No __NA	__Yes SPECIFY__ __No __NA
Homeless Risk Factors
n. Do you live with friends or relatives?	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
o. Do you have an eviction notice?	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
p. Other risk factors? SPECIFY	__Yes   __NA __No	__Yes   __NA __No	__Yes   __NA __No
q. Are there any other items on your application that we have not talked about that are related to housing status, homelessness, or risk factors associated with homelessness? If yes, SPECIFY	__Yes SPECIFY__ __No __NA	__Yes SPECIFY__ __No __NA	__Yes SPECIFY__ __No __NA

The responses to these questions will be used to determine which set of detailed questions should be asked.  For those programs collecting neither homeless items nor items considered risk factors for homelessness, the without-homeless data questions will be asked.  For those collecting either homeless or homeless risk factor items, the homeless data questions will be asked.

INTERVIEWER: PLEASE CHECK ONE TO INDICATE TYPE OF AGENCY. DO NOT ASK A QUESTION, BUT SELECT THE CORRECT IDENTIFIER BASED ON QUESTIONS ASKED PREVIOUSLY:

            Is this agency a homeless data collecting or No homeless data agency?

___Homeless data collecting agency (homeless data or homeless risk factors are collected.

___No homeless data agency. No data on homelessness or homelessness risk factors are collected.

D. Questions for States without Homeless or Homeless Risk Factor Data

9.       Do you collect any data on the housing status of applicants at any time other than in the application?

            ___No (skip to 11)

      ___Yes

10.    At what point(s) do you collect housing status information? Do you collect this information at [READ ITEM]?

Item	Yes	No
a. Prior to eligibility determination (during subsequent interview or home visit with applicant)
b. After eligibility is determined (during an interview or on another form)
c. At recertification
d. Other (SPECIFY)

11.   Even if you do not explicitly ask program applicants or participants if they are homeless, do you use any other standard conventions for noting that an applicant is homeless in your database or on application materials?  Examples of this might be noting than an applicant is living with family or friends, has not had a permanent address in recent months, note whether an address given is for a shelter, enter 99999 for zip code when homelessness is suspected, etc.)?

____  No  [PROCEED TO Q12 BELOW]

____  Yes  confirm all addresses (would know if address was a shelter)

____  Yes  intake workers are instructed to follow certain protocol (e.g., enter 99999 for zip code) for applicants who are thought to be homeless PLEASE DESCRIBE:

____  Yes  other (SPECIFY)

IF YES to Q 11, THEN SKIP TO QUESTION 21 ON HOMELESS DATA VERSION OF QUESTIONNAIRE

12.   What would you say are the main reasons your program does not collect such information on the application? [INTERVIEWER: RECORD VERBATIM AND THEN RECODE TO ONE OF THE CODES BELOW.  IF MORE THAN ONE REASON CHECK ALL THAT APPLY. PROBE AS NEEDED.]

________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

___Homeless data or homeless risk factors are not needed to determine eligibility for TANF program

___We have made an effort to streamline application; collecting only necessary information

___Collecting homeless data would be too burdensome (financial or administrative burden is too great)

___Other (SPECIFY)_____________________________________________________________________________________________________________________________________________________________________________________________________________________

13.   Has your application (either the combined or separate TANF) ever included any questions on homelessness or homeless risk factors?

            ___No (SKIP to 18)

            ___Yes

14.   What questions were included? LIST ALL. [PROBE: Examples might be things like are you homeless? do you live in a shelter?]

________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

15.   When were the questions dropped from the application?

________/_________/___________

mm             dd               yyyy

16.   What is the main reason the questions were dropped from the application? [INTERVIEWER: RECORD VERBATIM AND THEN RECODE TO ONE OF THE CODES BELOW.  IF MORE THAN ONE REASON CHECK ALL THAT APPLY. PROBE AS NEEDED.]

________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

_____Not needed to determine eligibility

_____Not needed to administer the program

_____Too burdensome to collect

______ Other (SPECIFY)

17.   When you collected these questions, how were they used in your state?

____  Program and reporting purposes (DESCRIBE)

____  Collected as part of State effort to enhance access to mainstream benefits for people who are homeless

____  Collected to support development or implementation of State Plan to End Homelessness

____  State Policy Academy requested data or uses it

____  Other (SPECIFY)

18.   How frequently is your TANF application updated?

____ Monthly

____ Quarterly

____ Semi-annually (every 6 months)

____ Annually

____ Other (SPECIFY)

19.   Are these updates seen as an opportunity to add or refine questions on homelessness or housing status?

____ Yes

____ No

20.   Do you think that questions related to homelessness or risk factors for homelessness may be added to your application in the future?

            ___No (SKIP to 26)

            ___Possibly but not certain

            ___Yes

21.   What are the most likely questions that would be added to the application? LIST ALL.

__________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

22.   When do you think such questions would be added to the application?

23.   Are other questions being considered?

___No (SKIP to 26)

___Yes (DESCRIBE)

24.   How would these questions likely be used?  [INTERVIEWER: RECORD VERBATIM AND THEN RECODE TO ONE OF THE CODES BELOW.  IF MORE THAN ONE REASON CHECK ALL THAT APPLY. PROBE AS NEEDED.]

________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

____  Program and reporting purposes (DESCRIBE)

____  Collected as part of State effort to enhance access to mainstream benefits for people who are homeless

____  Collected to support development or implementation of State Plan to End Homelessness

____  State Policy Academy requested data or uses it

____  Other (SPECIFY)

25.   What is the main reason your state is considering adding questions to the application about homelessness or risk factors for homelessness? [INTERVIEWER: RECORD VERBATIM AND THEN RECODE TO ONE OF THE CODES BELOW.  IF MORE THAN ONE REASON CHECK ALL THAT APPLY. PROBE AS NEEDED.]

________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

______ To satisfy new reporting requirements

______To increase our involvement with state policy academy (EXPLAIN)

______To address goals (state or federal) regarding ending homelessness and access to mainstream programs for homeless persons

______To facilitate determination of categorical or expedited eligibility for homeless persons

_____ Other benefits or incentives (SPECIFY)___________________________________________________________________________________________________________________________________________

SKIP TO Q27

26.   What are the main reasons this type of question is not likely to be added to your application?

_____Not needed to determine eligibility

_____Not needed to administer the program

_____Too burdensome to collect

_____ Other (SPECIFY)

27.   Is there anything else you would like to add about your TANF application form and information collected from applicants about housing status or homelessness?

Thank you very much for your time and help with this study.

  Interview Guide for MEDICAID State Directors or other MEDICAID application experts

This version to be used for programs that do include questions about homelessness or risk factors associated with homelessness on the MEDICAID or combined applications

State: _____________________________________________________

Name of Respondent:________________________________________

Phone number: _____________________________________________

Email address of Respondent:_________________________________

Interview conducted by:_______________________________________

Date of interview:     ______/______/______

                                    Month Day       Year

According to the Paperwork Reduction Act of 1995, no persons are required to respond to a collection of information unless it displays a valid OMB control number. The valid OMB control number for this information collection is 0990-0320. The time required to complete this information collection is estimated to average 1 hour per response, including the time to review instructions, search existing data resources, gather the data needed, and complete and review the information collection. If you have comments concerning the accuracy of the time estimate(s) or suggestions for improving this form, please write to:  U.S. Department of Health & Human Services, OS/ocio/PRA, 200 Independence Ave., S.W., Suite 537-H, Washington D.C. 20201,   Attention: PRA Reports Clearance Officer. Alice Bettencourt

NOTE to Interviewer: Review the MEDICAID and/or Combined Application prior to the interview to review the type of homeless data or risk factors for homelessness collected on the application.   Also note the date when Abt obtained the application to confirm that it is the most recent version.

Introduction: My name is ___________ and I work for a private, independent research firm called Abt Associates.   The office of the Assistant Secretary for Planning and Evaluation at the Department of Health and Human Services is sponsoring this study.  The purpose of the study is to conduct interviews with each states TANF and Medicaid program staff to collect information about the type of housing status and/or homeless status data that is collected on TANF and Medicaid application forms.  Someone from the Abt study team contacted you in October or November 2006 to obtain a copy of your MEDICAID and/or combined application. Today Id like to ask some more detailed questions about your applications and the information you collect from program applicants about whether they are homeless and their housing status in general. Neither the study nor this interview is being used to monitor program performance in any way.  The information you provide will only be used for the purposes of this study.

A. Confirmation that application on hand is current

1.       We collected a MEDICAID application from you (online/via email/copy sent to us) in October 2006.  Is that application still in use?

__Yes

__No  When was the application updated? [ENTER DATE].  Could you please fax me a copy of the application that is now in use or direct me to a copy on your website?  What has changed on the application from the version we collected in October 2006? [ENTER WEB ADDRESSINTERVIEWER: INTERVIEW CAN CONTINUE IF CHANGES TO THE APPLICATION DO NOT INVOLVE QUESTIONS REGARDING HOMELESSNESS.  INTERVIEW SHOULD BE SUSPENDED AND RESCHEDULED IF CHANGES REGARDING HOMELESSNESS HAVE BEEN MADE TO THE APPLICATION.

2.       [IF WE IDENTIFIED A COMBINED APPLICATION] We collected a Combined benefit application from you (online/via email/copy sent to us) at the same time.  Is that application still in use?

__Yes

__No  When was the application updated? [ENTER DATE]. Could you please fax me a copy of the application that is now in use or direct me to a copy on your website?  What has changed on the application from the version we collected in October 2006? [ENTER WEB ADDRESS INTERVIEWER: INTERVIEW CAN CONTINUE IF CHANGES TO THE APPLICATION DO NOT INVOLVE QUESTIONS REGARDING HOMELESSNESS.  INTERVIEW SHOULD BE SUSPENDED AND RESCHEDULED IF CHANGES REGARDING HOMELESSNESS HAVE BEEN MADE TO THE APPLICATION.]

3.       [IF WE DIDNT IDENTIFY A COMBINED APPLICATION] As we collected application forms from other states, we learned that many states use a combined application covering several programs, either in addition to or in place of a separate MEDICAID application.  Does your state use a combined application for the MEDICAID program?

___No (skip to Section C)

___Yes

B. Clarification regarding use of combined application

4.       Is the combined application used instead of a separate application for MEDICAID or in addition to the MEDICAID application?

___Combined application used instead of a separate MEDICAID application (Ask 4a)

c.       Does that mean that when an individual applies for MEDICAID assistance they complete only the combined application?

___Yes

___No (please explain under what circumstances the separate application is used) ________________________________________________________________________________________________________________________________________________________________________________________________

___Combined application used in addition to a separate MEDICAID application (Ask 4b)

d.       Does that mean that when a family applies for MEDICAID assistance they must complete both applications?

___Yes

___No (please explain under what circumstances the combined application is used) ________________________________________________________________________________________________________________________________________________________________________________________________

___Dont know  Is there someone else we could talk to, to find out more about how the combined application is used?  [ENTER CONTACT INFORMATION].

C. Collection of Housing Status, Homeless, or Risk Factors Data

These questions are used to confirm that the correct set of questions is being asked.  

The purpose of this study is to find out whether states routinely collect information about homelessness or risk factors often associated with homelessness for MEDICAID applicants through the regular application process.  Examples of items on an application that might indicate homeless ness are things like living in a shelter, or having no permanent residence.  Risk factors associated with homelessness include things like living with friends or relatives, having an eviction notice, or others.

5.       a. What is the typical process for completing an application?   Which is the most common way applications are completed? (CHECK ONLY ONE)

____ application completed online

____ application completed in-person at a MEDICAID office

____ application filled out and mailed to the MEDICAID office

____ Other (SPECIFY)

b. Do applicants usually fill out the applications themselves without assistance, or do they receive assistance from an intake worker to complete the application?  Which is the most common approach? (CHECK ONLY ONE)

__fill out with assistance

__fill out without assistance

c. Is the same application used throughout the state?

____ Yes

____ No

d. Other than in the application form, is additional information on housing status collected at other points in the eligibility determination process? If yes, when?

____ Yes (SPECIFY)

____ No

Wed like to ask you some questions to confirm that our understanding of the types of information included on your MEDICAID/combined application is correct and to find out more about how these data are collected and used, and reasons why your state collects this information.

[ENTER RESPONSES FOR 6-8 IN THE TABLE BELOW.  REVIEW APPLICATON REVIEW DATABASE AND MARK PRELIMINARY RESPONSES FROM THE DATABASE FOR QUESTIONS 6 AND 7 IN THE TABLE BELOW.]

6.       When we reviewed your combined benefit application, we found that you collect [READ DATA ELEMENTS FROM Q6 COLUMN] on housing status, homelessness, and homeless risk factors.  Is our understanding accurate?  Is it correct that you do not collect [READ REMAINING DATA ELEMENTS] on the combined application?  [INTERVIEWER: IF STATE DOES NOT USE A COMBINED APPLICATION, CODE AS NOT APPLICABLE.]

7.       We also determined that you collect [READ DATA ELEMENTS FROM Q7 COLUMN] on your separate MEDICAID application.  Is our understanding accurate?  Is it correct that you do not collect [READ REMAINING DATA ELEMENTS] on the separate MEDICAID application? [INTERVIEWER: IF STATE DOES NOT USE A SEPARATE MEDICAID APPLICATION, CODE AS NOT APPLICABLE]

8.       If a combined application is used, do applicants for all programs covered by the combined application answer each item? Or are there items that are only required for some of the programs included on the combined application?  Please explain.  [INTERVIEWER: IF STATE DOES NOT USE A COMBINED APPLICATION, CODE AS NOT APPLICABLE]

The responses to these questions will be used to determine which set of detailed questions should be asked.  For those programs collecting neither homeless items nor items considered risk factors for homelessness, the no-homeless data questions will be asked.  For those collecting either homeless or homeless risk factor items, the homeless data questions will be asked.

D. Homeless Data Questions

Id like to understand more about the data that you collect for each of the questions that specifically relate to homelessness. If there are different answers for the combined application and separate MEDICAID application, please let me know.  [INTERVIEWER: ONLY ASK ABOUT THE RELEVANT COLUMNS, AS CONFIRMED IN Q6-7.  PROBE FOR WHETHER RESPONSES DIFFER FOR COMBINED APPLICATION AND SEPARATE MEDICAID APPLICATION.]

18.   Do you collect any data on the homeless status of applicants at any time other than in the application?

____  No (Skip to 20)

____  Yes

19.   Do you collect homeless status information at [READ ITEM]?  At each of the subsequent times you mentioned, do you update the information (that is, confirm or update the same data items again), or do you collect new information on housing or homeless status?  Is the new or updated information entered into the database?

INTERVIEWER: ASK QUESTIONS 19a, 19b, 19c and 19d FOR EACH TIME PERIOD

[ASK 20-22 ONLY IF RESPONSE TO 6 and 7i: Are you homeless IS NO.  If YES, skip to 23.]

20.   Even if you do not explicitly ask program applicants or participants if they are homeless, do you use any other standard conventions for noting that an applicant is homeless in your database or on application materials?  Examples of this might be noting than an applicant is living with family or friends, has not had a permanent address in recent months, note whether an address given is for a shelter, enter 99999 for zip code when homelessness is suspected, etc.?

____  No (SKIP TO 23)

____  Yes confirm all addresses (would know if address was a shelter)

____  Yes intake workers are instructed to follow certain protocol (e.g., enter 99999 for zip code) for applicants who are thought to be homeless PLEASE DESCRIBE:

____  Yes Other (DESCRIBE)

21.   Are these conventions formalized in such a way that you would say that they have been adopted throughout the state?  [PROBE: E.g., are they written down in training instructions?  Are they written on the computer screen where intake workers enter the data?]

____  No

____  Yes (SPECIFY)                                                                                                   

22.   Do you have the ability to query the program database based on these protocols to produce statistics about homelessness among program applicants?

____  No

____  Yes (please describe how, if you have done it before, and limitations of the results)                                                                                                                                                                                                                                                                                                                                                                                                   

23.   Has your state developed a formal definition for homelessness for purposes of this application/program?

____  No (SKIP TO 25)

____  Yes (please tell me what the definition is and how it is used by the state)                                                                                                                                                     

24.   Is the definition of homelessness clearly documented or widely known to intake workers or others involved with helping applicants complete the application forms?

____  No

____  Yes (DESCRIBE)                                                                                                

                  If yes, SKIP TO 26

25.   Has homelessness been informally defined in the state?

____  No

____  Yes (specify how and please provide the definition)                                                                                                                                                              

26.   Many states have not chosen to include questions on homelessness on their applications for MEDICAID benefits. We would like to understand more about why leaders in your state believe that it is important to do so.  What are the main reasons?  [INTERVIEWER: RECORD VERBATIM AND THEN RECODE TO ONE OF THE CODES BELOW. IF MORE THAN ONE REASON CHECK ALL THAT APPLY. PROBE AS NEEDED].

____  Program and reporting purposes (DESCRIBE)

____ Collected as part of State effort to enhance access to mainstream benefits for people who are homeless

____ Collected to support development or implementation of State Plan to End Homelessness

____  State Policy Academy requested data or uses it

____  Other (SPECIFY)

27.   Has the state conducted any analysis on the homelessness data yet?

____  No (SKIP TO 32)

____  Yes

28.   Describe what was done, produced, and learned (CHECK ALL THAT APPLY)

____  Estimates of homeless persons applying for benefits

____  Estimates of homeless persons receiving benefits

____  Estimates of impact of benefits on alleviating persons homelessness

____ Studies on costs of mainstream services for people who are homeless

____ Other? (SPECIFY)

29.   How has the state used the results of the study?

30.   Did you alter your data collection requirements, training of intake workers, or analysis/use of data as a result of this study? PLEASE DESCRIBE

31.   Are the data on homelessness accessible to the public? [CHECK ALL THAT APPLY]

____  No (SKIP TO 33)

____  In written reports

____  In state-level data sets

32.   Who has access to the data sets?

33.   Would you consider providing periodic de-identified data extracts or aggregate reports to HHS on homelessness on applicants to your MEDICAID program?

34.   Do you think it takes more time to collect the questions on homelessness than other items on your application?

____  No

____  Yes (DESCRIBE)                                                                                     

35.   Does it take more time to verify, record, or analyze the responses than for other items on the application?

____  No

____  Yes (DESCRIBE)                                                                                     

36.   Are there any additional costs incurred to collect this information?  For instance, does it add to the verification time or effort?

____  No

____  Yes (DESCRIBE)                                                                                     

37.   How frequently is your MEDICAID application updated?

____ Monthly

____ Quarterly

____ Semi-annually (every 6 months)

____ Annually

____ Other (SPECIFY)

38.   Are these updates seen as an opportunity to add or refine questions on homelessness or housing status?

____ Yes

____ No

39.   That is all the questions I have about homelessness data and your MEDICAID and/or combined application. Is there anything else you would like to add that you think is important for this study to take into account?

Thank you very much for your time and help with this study.

 

---

Interview Guide for MEDICAID State Directors or other MEDICAID application experts

This version to be used for programs that do not have questions about homelessness or risk of homelessness on the MEDICAID or combined applications

State: _____________________________________________________

Name of Respondent:________________________________________

Phone number: _____________________________________________

Email address of Respondent:_________________________________

Interview conducted by:_______________________________________

Date of interview:     ______/______/______

                                   Month Day       Year

According to the Paperwork Reduction Act of 1995, no persons are required to respond to a collection of information unless it displays a valid OMB control number. The valid OMB control number for this information collection is 0990-0320. The time required to complete this information collection is estimated to average 1 hour per response, including the time to review instructions, search existing data resources, gather the data needed, and complete and review the information collection. If you have comments concerning the accuracy of the time estimate(s) or suggestions for improving this form, please write to:  U.S. Department of Health & Human Services, OS/ocio/PRA, 200 Independence Ave., S.W., Suite 537-H, Washington D.C. 20201,   Attention: PRA Reports Clearance Officer. Alice Bettencourt

NOTE to Interviewer: Review the MEDICAID and/or Combined Application prior to the interview to review the type of homeless data or risk factors for homelessness collected on the application.   Also note the date when Abt obtained the application to confirm that it is the most recent version.  

Introduction: My name is ___________ and I work for a private, independent research firm called Abt Associates.   The office of the Assistant Secretary for Planning and Evaluation at the Department of Health and Human Services is sponsoring this study.  The purpose of the study is to conduct interviews with each states TANF and Medicaid program staff to collect information about the type of housing status and/or homeless status data that is collected on TANF and Medicaid application forms.  Someone from the Abt study team contacted you in October or November 2006 to obtain a copy of your MEDICAID and/or combined application. Today Id like to ask some more detailed questions about your applications and the information you collect from program applicants about whether they are homeless and their housing status in general. Neither the study nor this interview is being used to monitor program performance in any way.  The information you provide will only be used for the purposes of this study.

A. Confirmation that application on hand is current

1.       We collected a MEDICAID application from you (online/via email/copy sent to us) in October 2006.  Is that application still in use?

__Yes

__No  When was the application updated? [ENTER DATE].  Could you please fax me a copy of the application that is now in use or direct me to a copy on your website? What has changed on the application from the version we collected in October 2006? [ENTER WEB ADDRESSINTERVIEWER: INTERVIEW CAN CONTNUE IF CHANGES TO THE APPLICATION DO NOT INVOLVE QUESTIONS REGARDING HOMELESSNESS. INTERVIEW SHOULD BE SUSPENDED AND RESCHEDULED IF CHANGES REGARDING HOMELESSNESS HAVE BEEN MADE TO THE APPLICATION.]

2.       [IF WE IDENTIFIED A COMBINED APPLICATION] We collected a Combined benefit application from you (online/via email/copy sent to us) at the same time.  Is that application still in use?

__Yes

__No

When was the application updated? [ENTER DATE].  Could you please fax me a copy of the application that is now in use or direct me to a copy on your website?  What has changed on the application from the version we collected in October 2006? [ENTER WEB ADDRESSINTERVIEWER: INTERVIEW CAN CONTNUE IF CHANGES TO THE APPLICATION DO NOT INVOLVE QUESTIONS REGARDING HOMELESSNESS.  INTERVIEW SHOULD BE SUSPENDED AND RESCHEDULED IF CHANGES REGARDING HOMELESSNESS HAVE BEEN MADE TO THE APPLICATION.]

3.       [IF WE DIDNT IDENTIFY A COMBINED APPLICATION] As we collected application forms from other states, we learned that many states use a combined application covering several programs, either in addition to or in place of a separate MEDICAID application.  Does your state use a combined application for the MEDICAID program?

___No (skip to Section C)

___Yes

B. Clarification regarding use of combined application

4.       Is the combined application used instead of a separate application for MEDICAID or in addition to the MEDICAID application?

___Combined application used instead of a separate MEDICAID application (Ask 4a)

a.       Does that mean that when an individual applies for MEDICAID assistance they complete only the combined application?

___Yes

___No (please explain under what circumstances the separate application is used) _________________________________________________________________________________________________________________________________________________________________________________________________________

___Combined application used in addition to a separate MEDICAID application (Ask 4b)

b.       Does that mean that when a family applies for MEDICAID assistance they must complete both applications?

___Yes

___No (please explain under what circumstances the combined application is used) _________________________________________________________________________________________________________________________________________________________________________________________________________

___Dont know  Is there someone else we could talk to, to find out more about how the combined application is used?  [ENTER CONTACT INFORMATION].

C. Collection of Housing Status, Homeless, or Risk Factors Data

These questions are used to confirm that the correct set of questions is being asked.  

The purpose of this study is to find out whether states routinely collect information about homelessness or risk factors often associated with homelessness for MEDICAID applicants through the regular application process. Examples of items on an application that might indicate homelessness are things like living in a shelter, or having no permanent residence.  Risk factors associated with homelessness include things like living with friends or relatives, having an eviction notice, or others.

5.       a. What is the typical process for completing an application? Which is the most common way applications are completed? (CHECK ONLY ONE)

____ application completed online

____ application completed in-person at a MEDICAID office

____ application filled out and mailed to the MEDICAID office

____ Other (SPECIFY)

b. Do applicants usually fill out the applications themselves without assistance, or do they receive assistance from an intake worker to complete the application? Which is the most common approach? (CHECK ONLY ONE)

____ fill out with assistance

____ fill out without assistance

c. Is the same application used throughout the state?

            ____ Yes

            ____ No

d. Other than in the application form, is additional information on housing status collected at other points in the eligibility determination process? If yes, when?

____ Yes (SPECIFY)

____ No

Wed like to ask you some questions to confirm that our understanding of the types of information included on your MEDICAID/combined application is correct and to find out more about how these data are collected and used, and reasons why your state collects this information.

[ENTER RESPONSES FOR 6-8 IN THE TABLE BELOW.  REVIEW APPLICATON AND REVIEW DATABASE AND MARK PRELIMINARY RESPONSES FROM THE DATABASE FOR QUESTIONS 6 AND 7 IN THE TABLE BELOW.]

6.       When we reviewed your combined benefit application, we found that you collect [READ DATA ELEMENTS FROM Q6 COLUMN] on housing status, homelessness, and homeless risk factors.  Is our understanding accurate?  Is it correct that you do not collect [READ REMAINING DATA ELEMENTS] on the combined application?  [INTERVIEWER: IF STATE DOES NOT USE A COMBINED APPLICATION, CODE AS NOT APPLICABLE.]

7.       We also determined that you collect [READ DATA ELEMENTS FROM Q7 COLUMN] on your separate MEDICAID application.  Is our understanding accurate?  Is it correct that you do not collect [READ REMAINING DATA ELEMENTS] on the separate MEDICAID application? [INTERVIEWER: IF STATE DOES NOT USE A SEPARATE MEDICAID APPLICATION, CODE AS NOT APPLICABLE]

8.       If a combined application is used, do applicants for all programs covered by the combined application answer each item? Or are there items that are only required for some of the programs included on the combined application?  Please explain.  [INTERVIEWER: IF STATE DOES NOT USE A COMBINED APPLICATION, CODE AS NOT APPLICABLE]

The responses to these questions will be used to determine which set of detailed questions should be asked.  For those programs collecting neither homeless items nor items considered risk factors for homelessness, the without-homeless data questions will be asked.  For those collecting either homeless or homeless risk factor items, the homeless data questions will be asked.

INTERVIEWER: PLEASE CHECK ONE TO INDICATE TYPE OF AGENCY. DO NOT ASK A QUESTION, BUT SELECT THE CORRECT IDENTIFIER BASED ON QUESTIONS ASKED PREVIOUSLY:

            Is this agency a homeless data collecting or No homeless data agency?

___Homeless data collecting agency (homeless data or homeless risk factors are collected.

___No homeless data agency. No data on homelessness or homelessness risk factors are collected.

D. Questions for States without Homeless or Homeless Risk Factor Data

9.       Do you collect any data on the housing status of applicants at any time other than in the application?

            ___No (skip to 11)

            ___Yes

10.    At what point(s) do you collect housing status information? Do you collect this information at [READ ITEM]?

Item	Yes	No
a. Prior to eligibility determination (during subsequent interview or home visit with applicant)
b. After eligibility is determined (during an interview or on another form)
c. At recertification
d. Other (SPECIFY)

11.  Even if you do not explicitly ask program applicants or participants if they are homeless, do you use any other standard conventions for noting that an applicant is homeless in your database or on application materials?  Examples of this might be noting than an applicant is living with family or friends, has not had a permanent address in recent months, note whether an address given is for a shelter, enter 99999 for zip code when homelessness is suspected, etc.)?

____  No  [PROCEED TO Q12 BELOW]

____  Yes-- confirm all addresses (would know if address was a shelter)

____  Yes intake workers are instructed to follow certain protocol (e.g., enter 99999 for zip code) for applicants who are thought to be homeless PLEASE DESCRIBE:

____  Yes other (SPECIFY)

IF YES TO Q11, THEN SKIP TO QUESTION 21 ON HOMELESS DATA VERSION OF QUESTIONNAIRE

12. What would you say are the main reasons your program does not collect such information on the application? [INTERVIEWER: RECORD VERBATIM AND THEN RECODE TO ONE OF THE CODES BELOW.  IF MORE THAN ONE REASON CHECK ALL THAT APPLY. PROBE AS NEEDED.]

________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

___Homeless data or homeless risk factors are not needed to determine eligibility for MEDICAID program

___We have made an effort to streamline application; collecting only necessary information

___Collecting homeless data would be too burdensome (financial or administrative burden is too great)

___Other (SPECIFY)_____________________________________________________________________________________________________________________________________________________________________________________________________________________

13.   Has your application (either the combined or separate MEDICAID) ever included any questions on homelessness or homeless risk factors?

            ___No (SKIP to 18)

            ___Yes

14.   What questions were included? LIST ALL. [PROBE: Examples might be things like are you homeless? do you live in a shelter?]

________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

15.   When were the questions dropped from the application?

________/_________/___________

mm             dd               yyyy

16.   What is the main reason the questions were dropped from the application? [INTERVIEWER: RECORD VERBATIM AND THEN RECODE TO ONE OF THE CODES BELOW.  IF MORE THAN ONE REASON CHECK ALL THAT APPLY. PROBE AS NEEDED.]

__________________________________________________________________________

__________________________________________________________________________

__________________________________________________________________________

__________________________________________________________________________

_____Not needed to determine eligibility

_____Not needed to administer the program

_____Too burdensome to collect

_____ Other (SPECIFY)

17.   When you collected these questions, how were they used in your state?

____  Program and reporting purposes (DESCRIBE)

____  Collected as part of State effort to enhance access to mainstream benefits for people who are homeless

____  Collected to support development or implementation of State Plan to End Homelessness

____  State Policy Academy requested data or uses it

____  Other (SPECIFY)

18.   How frequently is your MEDICAID application updated?

____ Monthly

____ Quarterly

____ Semi-annually (every 6 months)

____ Annually

____ Other (SPECIFY)

19.   Are these updates seen as an opportunity to add or refine questions on homelessness or housing status?

____ Yes

____ No

20.   Do you think that questions related to homelessness or risk factors for homelessness may be added to your application in the future?

            ___No (SKIP to 26)

            ___Possibly but not certain

            ___Yes

21.   What are the most likely questions that would be added to the application? LIST ALL.

__________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

22.   When do you think such questions would be added to the application?

23.   Are other questions being considered?

___No (SKIP to 26)

___Yes (DESCRIBE)

24.   How would these questions likely be used?  [INTERVIEWER: RECORD VERBATIM AND THEN RECODE TO ONE OF THE CODES BELOW.  IF MORE THAN ONE REASON CHECK ALL THAT APPLY. PROBE AS NEEDED.]

____________________________________________________________________________________________________________________________________________________

__________________________________________________________________________

____  Program and reporting purposes (DESCRIBE)

____  Collected as part of State effort to enhance access to mainstream benefits for people who are homeless

____  Collected to support development or implementation of State Plan to End Homelessness

____  State Policy Academy requested data or uses it

____  Other (SPECIFY)

25.   What is the main reason your state is considering adding questions to the application about homelessness or risk factors for homelessness? [INTERVIEWER: RECORD VERBATIM AND THEN RECODE TO ONE OF THE CODES BELOW.  IF MORE THAN ONE REASON CHECK ALL THAT APPLY. PROBE AS NEEDED.]

____________________________________________________________________________________________________________________________________________________________________________________________________________________________

_____ To satisfy new reporting requirements

_____ To increase our involvement with state policy academy (EXPLAIN)

_____ To address goals (state or federal) regarding ending homelessness and access to mainstream programs for homeless persons

_____ To facilitate determination of categorical or expedited eligibility for homeless persons

_____ Other benefits or incentives (SPECIFY)___________________________________________________________________________________________________________________________________________

SKIP TO Q27

26.   What are the main reasons this type of question is not likely to be added to your application?

_____Not needed to determine eligibility

_____Not needed to administer the program

_____Too burdensome to collect

_____ Other (SPECIFY)

27.   Is there anything else you would like to add about your MEDICAID application form and information collected from applicants about housing status or homelessness?

Thank you very much for your time and help with this study.

[ Go to Contents ]

Appendix D Use of Homelessness Data Collected in TANF and Medicaid Applications

[ Go to Contents ]

Appendix E List of Program Contacts HHS Mainstream Program Profiles for Nine Mainstream Programs Reviewed Under Task 5

Program Contacts Interviewed

1. ACF Community Services Block Grant

• Peter Thompson
• Marsha Werner

National Association for State Community Service Programs (NASCSP)

• Jenae Bjelland
• Jovita Tolbert

2. ACF Head Start

• Amanda Bryans

3. ACF Social Services Block Grant (SSBG)

• Peter Thompson
• Marsha Werner

4. CMS State Childrens Health Insurance Program (SCHIP)

• Maurice Gagnon

5. HRSA Health Center Program

• Marquita Cullom-Stott

6. HRSA Maternal and Child Health Services Title V Block Grant (Title V)

• Cassie Lauver

7. HRSA Ryan White HIV/AIDS Program (Ryan White)

• Alice Kroliczak
• Gettie Butts

8. SAMHSA Community Mental Health Services Block Grant (CMHS)

• Sue Becker
• Valerie Mills

9. SAMHSA Substance Abuse Prevention and Treatment Block Grant (SAPT)

• Sue Becker
• Valerie Mills

HHS Mainstream Program Profiles

1. ACF Community Services Block Grant (CSBG)
2. ACF Head Start
3. ACF Social Services Block Grant (SSBG)
4. CMS State Childrens Health Insurance Program (SCHIP)
5. HRSA Health Center Program
6. HRSA Maternal and Child Health Services Title V Block Grant (Title V)
7. HRSA Ryan White HIV/AIDS Program (Ryan White)
8. SAMHSA Community Mental Health Services Block Grant (CMHS)
9. SAMHSA Substance Abuse Prevention and Treatment Block Grant (SAPT)

ACF - Community Services Block Grant (CSBG)

Administering Agency: ACF, Office of Community Services

Short Description: The program's purpose is to reduce poverty, revitalize low-income communities and empower low-income families and individuals to be self-sufficient.  Grantees receive funds to provide services and activities under the following nine categories: employment, education, income management, housing, nutrition, emergency services, linkages, self-sufficiency and health.

Services Provided: CSBG provides flexible core or foundational funding to more than 1,000 community-based organizations (Community Action Agencies, or CAAs) in almost every county in the nation to promote innovative, community-generated and location-specific actions to reduce the incidence and severity of poverty.

In a summary prepared by NASCSP of the uses of CSBG funds to address homelessness, we found a broad range of efforts and activities to serve homeless persons currently in place.  Some of these include:

• Funding for Project Homeless Connect, a collaborative regional effort to link homeless persons with needed services;
• Outreach to homeless persons;
• Homeless prevention efforts, including emergency rental assistance, utility assistance, eviction prevention, housing counseling, and landlord-tenant mediation;
• Operation of emergency shelters and transitional housing for homeless persons;
• Rental assistance and security deposits;
• Building homeless coalitions, including providing technical assistance and training;
• Providing meals and emergency food assistance to homeless persons;
• Operating second-hand clothing stores;
• Supporting Continuum of Care operations, including covering the costs of information systems, annual point-in-time counts, and other grant administration activities; and
• Providing supportive services to homeless persons, including healthcare, case management, mental health services, substance abuse services, benefits acquisition assistance, job training, and education.

Eligibility:

Applicant: State and U.S. Territory Governments and Tribal Governments or Tribal Organizations (including the District of Columbia, the Commonwealth of Puerto Rico, Guam, the Virgin Islands, American Samoa, and the Commonwealth of the Northern Mariana Islands).

Beneficiary: The CSBG Act mandates that States pass through 90% of the funds allocated to the eligible entities.  States make grants to qualified locally-based nonprofit community antipoverty agencies and other eligible entities which provide services to low-income individuals and families.

Administered: Formula block grant awarded to states.

Collect Data on Homelessness: No specific data on the homelessness status of program participants is collected or reported nationally.  As a block grant program, each state manages the program according to its individual needs.  While many states may use CSBG funds to serve homeless persons, and may track their efforts and activities, they are not required to report these data at the national level.  The CSBG program requires states to submit annual reports to the Office of Community Service detailing their use of CSBG funds; however these reports are in narrative format and the Office of Community Service does not aggregate the data or look across state efforts.

The National Association for State Community Service Programs (NASCSP) also collects and maintains data on CSBG grantees activities and uses of funding.  NASCSP uses the Results-Oriented Accountability and Management (ROMA) system, a performance-focused system for assessing and reporting the effectiveness of the CSBG networks antipoverty strategies in terms of results achieved among low-income families and neighborhoods.  The CSBG network compiles information from state-level databases which incorporate ROMA reporting, however the databases differ between states and usually between CSBG network agencies.  ROMA is primarily used as a management tool for the local agencies, their boards of directors and for the State CSBG office.  It does not collect client-level data on persons served by CSBG funds, and therefore could not be used to determine the homelessness status of program participants.

NASCSP collects data on 6 strategic goals:

1. Low-income people become more self-sufficient.
2. The conditions in which low-income people live are improved.
3. Low-income people own a stake in their community.
4. Partnerships among supporters and providers of service to low-income people are achieved.
5. Agencies increase their capacity to achieve results.
6. Low-income people, especially vulnerable populations, achieve their potential by strengthening family and other supportive systems.

While none of these goals directly relate to homelessness, states that do spend CSBG funds to serve homeless persons generally fall under one program objective (objective #6).

NASCSP also conducts an annual survey of CSBG grantees (all 50 states, as well as the District of Columbia and Puerto Rico).  The survey collects information on the level and uses of CSBG funds, on the sources and uses of other funding administered by the CSBG local network, on the program activities of the network, and on the number and characteristics of clients served.  The most recent report available is for FY2005.  NASCSP reported that in the CSBG program and its network partners served approximately 15 million people in FY2005. Although the survey does not explicitly collect information on the homelessness status of persons served with CSBG funds, it does collect information on income level and sources, barriers to self-sufficiency, disability, poverty level, and other demographic characteristics.

Program Website: http://www.acf.hhs.gov/programs/ocs/csbg/

Other Relevant Websites:

• http://www.acf.hhs.gov/programs/ocs/csbg/fact_sheets/fs_csbg.html
• http://www.acf.hhs.gov/programs/ocs/csbg/documents/mission.html
• http://www.nascsp.org

ACF - Head Start

Administering Agency: ACF, Office of Head Start

Short Description: Head Start is a national program that promotes school readiness by enhancing the social and cognitive development of children through the provision of educational, health, nutritional, social and other services to enrolled children and families.

Services Provided: Services are provided under two initiatives: the Education and Early Child Development program and the Community Partnerships program.  The objective of Education and Early Childhood Development is to provide all children with a safe, nurturing, engaging, enjoyable, and secure learning environment, in order to help them gain the awareness, skills, and confidence necessary to succeed in their present environment, and to deal with later responsibilities in school and in life.  The varied experiences provided by the program support the continuum of children's growth and development, which includes the physical, social, emotional, and cognitive development of each child.

The objective of Community Partnerships is to ensure that grantee and delegate agencies collaborate with partners in their communities, in order to provide the highest level of services to children and families, to foster the development of a continuum of family centered services, and to advocate for a community that shares responsibility for the healthy development of children and families of all cultures.

Eligibility:

Applicant: Any local government, federally-recognized Indian tribe, or public or private nonprofit or for profit agency which meets the requirements may apply for a grant.  However, applications will be considered only when submitted in response to a specific announcement, published on the Internet at the following website address: http://www.acf.hhs.gov/grants/grants_hsb.html, which solicits proposals to establish new Head Start Programs.  Grantee agencies may subcontract with other child-serving agencies to provide services to Head Start children.

Beneficiary: Head Start/Early Head Start programs are for children from birth up to the age when the child enters the school system; however, Head Start programs only serve pre-school age children (i.e. children three or four years old) while Early Head Start programs serve children from birth through age three.

Administered: Administered locally by non-profit organizations and local education agencies

Collect Data on Homelessness: All Head Start programs are required to report the number of homeless families served in the annual Program Information Report.  (The number of homeless families using Head Start has remained at 2% since 2002.)  Head Start also tracks whether families were able to obtain housing during the school year.  These data include families that participate at any point, including those who do not complete the year.

Head Start programs vary considerably by community; each community conducts an annual needs assessment and targets grant resources to locally-identified needs.  Data are collected at the program-level by each grantee in their own system.  At the end of each funding year, each grantee reports aggregate data to ACF in the Program Information Report (PIR).  ACF then aggregates the PIRs to look at the overall profile of Head Start programs and their participants.

The PIR collects detailed information about the Head Start grantee, funding sources, program approach, services offered, program staff, and demographic characteristics of families participating in the programs.  Some of the required data elements include:

• Enrollment by age;
• Enrollment by eligibility status (including homelessness);
• Ethnicity;
• Race;
• Language spoken at home;
• Turnover of program staff;
• Services provided to children and families;
• Staff qualifications;
• Health insurance status of children; and
• Disability status.

Specifically, under the Enrollment by eligibility status category, the PIR asks programs to report on the number of children and families enrolled in the program who are homeless. Beginning in December 2007, the program defined homeless using the McKinney-Vento Act definition.  Program staff make their own determinations regarding whether families meet this definition; ACF staff were unsure if there are national training efforts or other standards in place to ensure the definition is being applied consistently.

Under the Services provided category, grantees report the number of families who received housing assistance services.  ACF staff explained that this category could include a range of housing assistance services including referrals to housing counseling agencies, assistance in obtaining housing, utility or other financial assistance, landlord mediation, or other services.  Because the term is defined so broadly, it may even include referral to or placement into emergency shelter in some cases.  When asked how program staff know that families actually obtained housing, ACF staff said that staff are totally supposed to do follow-up but it is impossible to know whether any follow up case management or contact occurred by looking at the PIR.

Program Website: http://www.acf.hhs.gov/programs/hsb/

Other Relevant Websites:

• http://www.clasp.org/publications/hs_brief8.pdf
• http://homelessed.net/legislat/Information%20Memorandum%20-%20Head%20Start.pdf

ACF - Social Services Block Grant

Administering Agency:  ACF

Short Description: Social Services Block Grant funds enable states to provide social services best suited to meet the needs of the individuals residing within the state.  Services funded by the SSBG are directed at one or more of these five goals:

1. Achieving or maintaining economic self-support to prevent, reduce or eliminate dependency;
2. Achieving or maintaining self-sufficiency, including reduction or prevention of dependency;
3. Preventing or remedying neglect, abuse or exploitation of children and adults unable to protect their own interest, or preserving, rehabilitating or reuniting families;
4. Preventing or reducing inappropriate institutional care by providing for community-based care, home-based care or other forms of less intensive care; and
5. Securing referral or admission for institutional care when other forms of care are not appropriate or providing services to individuals in institutions.

Services Provided:  Service categories include, but are not limited to: daycare for children or adults, protective services for children or adults, special services to persons with disabilities, adoption, counseling, case management, family planning, health-related services, transportation, foster care for children or adults, substance abuse, legal, housing, home-delivered meals, congregate meals, independent/ transitional living, special services for youth, employment services or any other social services found necessary by the state for its population.

Eligibility:

Applicant:  The 50 States, the District of Columbia, Puerto Rico, Guam, the Virgin Islands, the Commonwealth of the Northern Mariana Islands, and American Samoa.  Funds are awarded in proportion to each state or territorys population.  States have flexibility to determine which of the 29 eligible service categories are funded, who will provide the services, and who services will target as long as they meet the programs five strategic goals (listed above).

Beneficiary:  Under Title XX, each eligible jurisdiction determines the services that will be provided and the individuals that will be eligible to receive services.  In order to qualify for this benefit program, financial situation must be low income.  Funding under the grant is not targeted to persons who are homeless, or any other specific subpopulations.

Administered:  Funds are allocated at the state-level (including the District of Columbia, the Commonwealth of Puerto Rico, and the Territories of Guam, American Samoa, the Virgin Islands, and the Northern Mariana Islands) in proportion to each states population.

Collect Data on Homelessness: No specific data on housing or homelessness status is collected or reported nationally.  National reporting requirements on SSBG beneficiaries are limited to pre- and post-expenditure reports that detail state spending and numbers served (separated by age categories) within 29 eligible service categories.  Persons who are homeless are most likely served within at least one service category, Independent/Transitional Living Care, which targets youth transitioning to self-sufficiency.  Unfortunately, this service category is the least commonly implemented and least popular among states.

Further investigation of annual state-specific reports did not reveal any explicit descriptions of state efforts to use SSBG funds to serve homeless persons or evidence to suggest that individual states are collecting data on homelessness status.

Program Website: http://www.acf.hhs.gov/programs/ocs/ssbg/

Other Relevant Website: http://www.acf.hhs.gov/programs/ocs/ssbg/fact_sheets/fs_ssbg.html

CMS - State Childrens Health Insurance Program

Administering Agency: CMS

Short Description: Created by the Balanced Budget Act of 1997, this children's health insurance program gives each state permission to offer health insurance for children, up to age 19, who are not already insured.  SCHIP is designed to provide health insurance coverage to "targeted low-income children" who are not eligible for Medicaid or other health insurance coverage.  Under Title XXI of the Social Security Act, states were given the option to set up a separate child health program, expand Medicaid coverage, or have a combination of both a separate child health program and a Medicaid expansion.

Services Provided: For little or no cost, this insurance pays for doctors visits, immunizations, hospitalizations, and emergency room visits.

Eligibility:

Applicant: All States and Territories may apply.

Beneficiary: Targeted low-income children residing in a family with income below the greater of 200 percent of poverty or 50 percentage points above the state's Medicaid eligibility threshold.  Most states have an upper eligibility limit of 200 percent of the federal poverty level, however, some states have amended their SCHIP plans to expand coverage to include children with family incomes above 200 percent federal poverty level.

Administered: SCHIP is a state administered program.  Each state sets its own guidelines regarding eligibility and services.

Collect Data on Homelessness: The program does not collect data on homeless persons served at the federal level.  SCHIP is targeted to low-income children, which includes children 150%-200% of the poverty level.  These children are not eligible for Medicaid or other health insurance.  Although beneficiaries may be homeless, housing status is not considered as an eligibility criteria, and homelessness would not affect a familys ability to obtain/maintain benefits.

Since each states population and priorities differ, HHS staff speculate that states with large urban populations may target homeless persons, and therefore, may collect data on homeless persons served.  However, staff do not know if this is the case, as they only collect aggregate enrollment data on beneficiaries in annual and quarterly reports not state-level efforts to target specific subpopulations. In fact, the program does not collect any data on the housing status of persons served, as housing status does not affect program eligibility.  Staff speculated that it is more likely that states access other databases (for example, Medicaid) to obtain housing status information, if necessary.

States are required to submit quarterly and annual reports in which they report on progress toward strategic objectives that they set.  All data collected relate to health, insurance, and progress toward meeting state-specific strategic objectives.  All information collected is required by statute.

Some of the data elements in the reports include:

• Strategic objectives
• Unduplicated count of new enrollees
• Eligibility information
• Process information about program administration
• % of enrollees below poverty level
• Number of people immunized

Program Website: http://www.cms.hhs.gov/home/schip.asp

Other Relevant Websites:

• http://www.whitehouse.gov/omb/expectmore/detail/10000306.2003.html
• http://12.46.245.173/pls/portal30/CATALOG.SEARCH_PGM_TXT_DYN.SHOW?p_arg_...

HRSA - Health Center Program

Administering Agency:  HRSA

Short Description: Health centers are community-based and patient-driven organizations that serve populations with limited access to health care. These include low income populations, the uninsured, those with limited English proficiency, migrant and seasonal farmworkers, individuals and families experiencing homelessness, and those living in public housing.  Health Centers that receive grant funding under the Health Center Program (section 330 of the Public Health Service Act) are routinely referred to as Federally Qualified Health Centers (FQHCs).  FQHCs are public and private non-profit health care organizations that meet certain criteria under the Medicare and Medicaid Programs (respectively, Sections 1861(aa)(4) and 1905(l)(2)(B) of the Social Security Act.  Some Health Centers also receive specific funding to focus on certain special populations.

Federally Qualified Health Center Look-Alikes (FQHCLAs) are health centers that have been designated by the Centers for Medicare and Medicaid Services, in consultation with HRSA, as meeting the definition of health center under Section 330 of the PHS Act, although they do not receive grant funding under Section 330.

FQHCs and FQHCLAs are eligible to receive enhanced reimbursement from Medicare and Medicaid and to participate in the 340B Program that enables them to purchase drugs at reduced prices.  Federally Qualified Health Centers also have access to medical malpractice insurance through the Federal Tort Claims Act.

Services Provided: Health Centers are community based providers that must provide all required primary, preventive, enabling health services and additional health services as appropriate and necessary, either directly or through established written arrangements and referrals per program requirements.  Applicants requesting funding to serve homeless individuals and their families must provide substance abuse services among their required services.

Eligibility:

Applicants: Public and private non-profit entities, including tribal, faith-based and community-based organizations who can demonstrate need and have the capacity to effectively manage the grant.

Beneficiaries:  People of all ages; people without and with health insurance; people of all races and ethnicities; and special populations (including people who are homeless).

Administered:  Project grants are administered by HRSA. Awards are negotiated based on the availability of funds and grant activities proposed by the applicant.

Collect Data on Homelessness:  Yes.  The Health Center Program collects limited data on the homelessness status of program participants.

For each patient seen at a Health Center, clinical staff assess eligibility for services targeted specifically to homeless populations.  The program defines homeless as: "an individual who lacks housing (without regard to whether the individual is a member of a family), including an individual whose primary residence during the night is a supervised public or private facility (e.g., shelters) that provides temporary living accommodations, and an individual who is a resident in transitional housing.  Program staff explained that a homeless person is an individual without permanent housing who may live on the streets; stay in a shelter, mission, single room occupancy facilities, abandoned building or vehicle; or in any other unstable or non-permanent situation.  An individual may also be considered homeless if that person is "doubled up," or in a situation in which they are unable to maintain their housing situation and are forced to stay with a series of friends and/or extended family members.  In addition, previously homeless individuals who are to be released from a prison or a hospital may be considered homeless if they do not have a stable housing situation to which they can return.  Staff maintain that a recognition of the instability of an individual's living arrangements is critical to the definition of homelessness.

In most cases clinical staff determine whether a patient meets the programs definition of homeless; however staff suspect that in many cases homelessness is self-reported.

Grantees are required to report data on clients served by the Health Center Program into the Uniform Data System (UDS).  The system collects information on patient revenue, insurance, demographic characteristics and other patient characteristics, including housing status.  On Line 22 of Table 4 in the UDS, grantees report the total number of patients known to have been homeless at the time of any service provided during the reporting period. This could include living in transitional housing, living in doubled up situations, or living on the streets.  Program staff commented that HRSA has not spent a great deal of time analyzing the data that emerge from UDS. Additionally, while the Health Center program recently reviewed and implemented performance improvement measures, they were not specifically related to the issues of homelessness or housing status.

Program Website: http://bphc.hrsa.gov/about/healthcenters.htm

Other Relevant Websites:

• http://12.46.245.173/pls/portal30/CATALOG.SEARCH_PGM_TXT_DYN.SHOW?p_arg_...
• http://bphc.hrsa.gov/uds/manual/table4.htm
• http://intranet.hrsa.gov/Communications/factSheets.asp

HRSA - Maternal and Child Health Services Title V Block Grant

Administering Agency:  Maternal and Child Health Bureau, HRSA

Short Description: This 70-year old federal program focuses on improving the health of all mothers,  infants, children (including those with special health care needs), women and families.  Title V is a partnership with State Maternal and Child Health (MCH) and Children with Special Health Care Needs (CSHCN) programs that supports core public health functions, such as resource development, capacity and systems building, population-based functions such as public information and education, knowledge development, outreach and program linkage, technical assistance to communities, and provider training, in addition to direct care and enabling services.

Services Provided:   The block grant has three components: formula block grants to 59 states and jurisdictions, grants for Special Projects of Regional and National Significance, and Community Integrated Service Systems grants. 

The largest portion of Title V goes to the States through a formula-based block grant process, which includes a matching funds requirement. (States match $3 in funds or resources for every $4 in Federal funds they receive.)  This $5 billion Federal/State partnership develops service systems in communities to meet critical challenges in maternal and child health, including:

• Significantly reducing infant mortality and incidence of handicapping conditions;
• Providing and ensuring access to comprehensive care for women;
• Promoting the health of children by providing preventive and primary care services;
• Increasing the number of children who receive health assessments, diagnostic and treatment services; and
• Providing family-centered, community-based, coordinated care for children with special health care needs.

Eligibility

Applicant: Title V MCH Block Grants are limited to States and insular areas.  Title V of the Social Security Act requires $3 of every $4 Federal dollars to be matched by states.

Beneficiary:  Mothers, infants, children, including CSHCN, and their families, particularly those of low-income.  Specific eligibility requirements to receive benefits, as well as the type and scope of services provided, are determined by each individual participating State or jurisdiction.

Administered:  Each States health agency is responsible for the administration (or supervision of the administration) of programs carried out with allotments made to the State under Title V.  Allotments are made based on a formula that provides basic funding to each state that submits an application.  The formula is based on the relationship between the number of children in poverty in a State relative to the total number of children in the United States.  MCHB's Division of State and Community Health (DSCH) oversees the day-to-day administration of the Maternal and Child Health Services Title V Block Grant Program and its staff serve as Project Officers  to the States and jurisdictions.

Collect Data on Homelessness: No specific data on housing or homelessness status is collected or reported nationally.  All states that receive Title V funds are required to submit annual reports and complete a comprehensive, statewide needs assessment every five years.  In their applications/annual reports and needs assessments, states describe priority areas for funding, vulnerable populations that will be served with Title V funds, and performance against established program outcome measures.  While the National performance measures and those developed by the States may not directly target the homeless population or those who lack affordable housing, States may include this population in their efforts to address health issues, such as teenage pregnancy, prenatal care, suicide prevention, domestic violence, substance abuse, and access to health care.   The program targets women, children, children with special needs (including those with low income) and therefore may  serve homeless persons.  In a search of the fiscal year (FY) 2009 State Title V MCH Block Grant applications and FY 2007 annual reports on the Title V Information System (TVIS), 30 States make reference to the homeless population in the Overview, Agency Capacity and State Agency Coordination sections of their application or in describing their efforts around specific National and State performance measures and Health System Capacity Indicators.  The search did not yield any states that explicitly target funding to this population or track efforts to serve homeless persons.

Program Website: http://mchb.hrsa.gov/

Other Relevant Websites:

• https://perfdata.hrsa.gov/mchb/mchreports/Search/search.asp (TVIS)
• ftp://ftp.hrsa.gov/mchb/titlevtoday/UnderstandingTitleV.pdf
• http://www.commonwealthfund.org/usr_doc/rosenbaum_titlev_481.pdf?section=4039
• http://mchb.hrsa.gov/programs/blockgrant/overview.htm
• http://www.hhs.gov/homeless/grants/index.html#MainstreamPrograms
• http://www.whitehouse.gov/omb/expectmore/detail/10000268.2002.html
• http://12.46.245.173/pls/portal30/CATALOG.PROGRAM_TEXT_RPT.SHOW?p_arg_na...

HRSA - Ryan White HIV/AIDS Program

Administering Agency:  HRSA

Short Description: In 1990, Congress enacted Title XXVI Public Health Service (PHS) Act, commonly referred to as the Ryan White Comprehensive AIDS Resources Emergency Act (RWCA), mainly for the purpose of the improvement of quality and availability of care for low-income uninsured and underinsured individuals and families affected by HIV disease. Almost two decades later, the continuum of care services had changed tremendously from that which was contemplated in 1990.  Since the 2006 reauthorization, the RWCA title is now referred to as the Ryan White HIV/AIDS Treatment Modernization Act (Ryan White HIV/AIDS Program).  The amended Ryan White HIV/AIDS Program was written to (a) reflect the shift in treatment for HIV towards a chronic disease model, (b) changes in the demographics of persons needing CARE Act services, and (c) changes in the health care financing environment.  Ryan White HIV/AIDS programs are the "payer of last resort.  They fill gaps in care not covered by other resources.  Ryan White clients include people with no other source of healthcare and those with Medicaid or private insurance whose care needs are not being met.

Services Provided:  The Ryan White HIV/AIDS Program includes:

• PART A: Grants to Eligible Metropolitan Areas and Transitional Grant Areas:  Part A of the Ryan White HIV/AIDS Program provides emergency assistance to Eligible Metropolitan Areas (EMAs) and Transitional Grant Areas (TGAs) that are most severely affected by the HIV/AIDS epidemic
• Part B: Grants to States and Territories:   Part B of the Ryan White HIV/AIDS Program provides grants to all 50 States, the District of Columbia, Puerto Rico, Guam, the U.S. Virgin Islands, and 5 U.S. Pacific Territories or Associated Jurisdictions. Part B grants include a base grant, the AIDS Drug Assistance Program (ADAP) award, ADAP supplemental grants and grants to States for Emerging Communities - those reporting between 500 and 999 cumulative reported AIDS cases over the most recent 5 years.
• PART C: Early Intervention Services:  The Part C Early Intervention Services (EIS) program of the Ryan White HIV/AIDS Program funds comprehensive primary health care in an outpatient setting for people living with HIV disease.
• PART D: Services for Women, Infants, Children, Youth and Families:  Part D grantees provide family-centered care involving outpatient or ambulatory care (directly or through contracts) for women, infants, children, and youth with HIV/AIDS. Grantees are expected to provide primary medical care, treatment, and support services to improve access to health care.
• PART F: Special Projects of National Significance Program:  The Special Projects of National Significance (SPNS) Program advances knowledge and skills in the delivery of health and support services to underserved populations diagnosed with HIV infection. SPNS grants fund innovative models of care and support the development of effective delivery systems for HIV care.
• PART F: AIDS Education and Training Centers Program:  The AIDS Education and Training Centers (AETC) Program of the Ryan White HIV/AIDS Program supports a network of 11 regional centers (and more than 130 local associated sites) that conduct targeted, multidisciplinary education and training programs for health care providers treating people living with HIV/AIDS. The AETCs serve all 50 States, the District of Columbia, the Virgin Islands, Puerto Rico, and the 6 U.S. Pacific Jurisdictions.
• PART F: Dental Programs:  Funds from all grant programs of the Ryan White HIV/AIDS Program can support the provision of oral health services. Two programs, however, specifically focus on funding oral health care for people with HIV: the Dental Reimbursement Program (DRP) and the Community-Based Dental Partnership Program (CBDPP).
• PART F: Minority AIDS Initiative:  The Minority AIDS Initiative grants provide funding to evaluate and address the disproportionate impact of HIV/AIDS on women and minorities.

Eligibility:

General Beneficiary: The Federal government establishes general guidelines for the administration of these services. However, specific eligibility requirements to receive benefits, as well as the type and scope of services provided, are determined by each individual participating grantee, state or jurisdiction.

Specific Beneficiary: Individual or family/household members must have HIV or AIDS and financial situation must be low income or affected family members.  In many states, program eligibility is related to eligibility for Medicaid.  As a result, eligibility criteria for Ryan White HIV/Aids Program funds vary from state to state since each state has different poverty levels and Medicaid and Medicare coverage thresholds.

Administered: Local and State-level grants, except for Part C grants, which are awarded directly to service providers.

Collect Data on Homelessness:  The Ryan White HIV/AIDS Program collects limited aggregate data on the number of homeless persons served each year by grantees. Annually, all Ryan White HIV/AIDS Program Grantees and service providers are required to show how those funds have been used to provide services to low-income and underserved individuals and families living with HIV/AIDS.  Using the Ryan White HIV/AIDS Program Annual Data Report (RDR) grantees describe: 1) characteristics of their organization; 2) the number and characteristics of clients they served; 3) the types of services provided; 4) the number of clients receiving these services; and 5) the number of client visits by type of service.  The Housing/Living Arrangements indicator asks whether the individual is permanently or non-permanently housed.  The Non-Permanent response categories include homeless, transient, or transitional housing. The data on homelessness and housing status as reported in the RDR are limited to this variable.  HRSA will phase-in implementation of the Ryan White HIV/AIDS Program Services Report (RSR), a client-level data reporting system beginning in 2009.  Ryan White HIV/AIDS Program grantees and service providers will use a new semiannual data collection and reporting system to report information on their programs and the clients they serve.  As part of the RSR, all grantees and service providers will report housing/living arrangement for each client at the end of the reporting period.

Program Website:  http://hab.hrsa.gov/livinghistory/

Other Relevant Websites:

• http://hab.hrsa.gov/aboutus.htm
• http://hab.hrsa.gov/reports/data2b.htm#1
• http://www.whitehouse.gov/omb/expectmore/detail/10000296.2007.html
• http://naeh2stage.forumone.com/files/1216_file_RyanWhiteCareActFactSheet...
• http://www.gao.gov/new.items/d06646.pdf

SAMHSA - Community Mental Health Services Block Grant

Administering Agency: SAMHSA

Short Description: Through the Community Mental Health Services Block Grant (CMHSBG), a joint Federal-State partnership, SAMHSA supports existing public services and encourages the development of creative and cost-effective systems of community-based care for people with serious mental disorders.  CMHS supports grassroots involvement of the major stakeholders in mental health services at both the State and the Federal levels.  At the Federal level, mental health planning council members-including family members, consumers, State officials, and service providers, serve as CMHS reviewers of State Mental Health Plans and Implementation Reports.  They often share their expertise at regional and national technical assistance meetings.  Each State or Territory is required to have a mental health planning council to review the State Mental Health Plan.

Services Provided: CMHS works in close collaboration with each State or Territory to develop and implement its own State Mental Health Plan for improving community-based services and reducing reliance on hospitalization.  The program stipulates that case management be provided to individuals with the most serious mental disorders and encourages appropriate partnerships among a wide range of health, dental, mental health, vocational, housing, and educational services.  The program also promotes partnerships among Federal, State, and local government agencies.

Eligibility:

Applicant: State and U.S. Territory Governments; or Tribal Organizations.

Administered: Grants are awarded to the States to provide mental health services to people with mental disorders.

Collect Data on Homelessness: Yes.  Community Mental Health Services Block grantees collect data on homeless persons served by funded programs.

Grantees (state mental health agencies or SMHAs) are required to submit annual Community Mental Health Services Block Grant Implementation Reports to SAMHSA using the Uniform Reporting System (URS).  The URS consists of a number of standardized tables, which are grouped into the following categories:

• Four tables address issues of Access to serviceshow many consumers received mental health services in states and in particular service settings.
• Four tables focus on the Appropriateness of care or levels of treatment.  These tables include numbers of homeless persons served, admission rates, length of stay in inpatient programs, and treatment of persons with co-occurring mental health and substance abuse disorders.  The Appropriateness Domain: Table 1: Homeless Persons Served by Community Mental Health Programs by Age and Gender reports the numbers of homeless persons served in community mental health programs in each state, by age and gender.  The numbers of homeless persons served and the rate of homeless persons served per 1,000 persons is shown for selected states.
• Three tables focus on Outcomes, including employment status of consumers served by SMHAs.  This section also includes outcomes reported by consumers after treatment.
• Three final tables focus on Structure, providing information on the expenditures and funding sources of State mental health agencies, as well as how States expended their mental health block grant funds.

The URS also includes a consumer living situation indicator that has a response of homeless or shelter.  Homelessness is defined as homeless or living in a homeless shelter at some assessment during the reporting year.

By FY 2008, each state will report on all CMHS National Outcome Measures (NOMs).  Stability in Housing is one of the new NOMs that is being collected and recorded.  For mental health, the NOM profiles include changes in the clients living arrangement, including homelessness.

Program Website:  http://mentalhealth.samhsa.gov/publications/allpubs/KEN95-0022/

Other Relevant Websites:

• http://download.ncadi.samhsa.gov/ken/pdf/URS_Data05/AL.pdf
• http://www.samhsa.gov/Matrix/matrix_brochure.pdf
• http://www.samhsa.gov/matrix/SAP_homeless.aspx

SAMHSA - Substance Abuse Prevention and Treatment Block Grant

Administering Agency: SAMHSA

Short Description: Funds provided to prevent and treat the abuse of alcohol and other drugs.

Services Provided:  This formula block grant to states provides substance abuse treatment and prevention services to individuals in need.  The grant is intended to provide maximum flexibility to states in determining allocations of the block grant to all populations within the states, dependent on state needs and priorities, including vulnerable and underserved populations, such as the homeless and those at risk of homelessness.  The authorizing legislation does not, however, specify homeless services and current policy does not encourage set-asides for specific populations. 

Eligibility:

Applicant:  States and U.S. Territories are eligible entities for formula-based funding; states contract with or award grants to providers and/or local government.

Administered: Block grant; services administered at the state-level.

Collect Data on Homelessness: Yes.  Grantees are required to report admission and discharge data through a client-level database called the Treatment Episode Data Set (TEDS).  TEDS is an administrative data system providing descriptive information about the national flow of admissions to providers of substance abuse treatment.  The TEDS series was designed to provide annual data on the number and characteristics of persons admitted to public and private substance abuse treatment programs receiving public funding.  The unit of analysis is treatment admissions.  TEDS provides information on living arrangements at admission and discharge, in addition to service setting, number of prior treatments, primary source of referral, employment status, whether methadone was prescribed in treatment, diagnosis codes, presence of psychiatric problems, source of income, health insurance, expected source of payment, substance(s) abused, route of administration, frequency of use, age at first use, pregnancy and veteran status, health insurance, demographic characteristics, and days waiting to enter treatment. The living arrangement data element specifies whether the client is homeless, living with parents or in a supervised setting, or living on his or her own. Homeless is defined as living with no fixed address, including in a shelter.

Additionally, SAPTBG grantees report data to the National Survey of Substance Abuse Treatment Services (N-SSATS).  This national survey began in the 1970s, and is designed to collect data on the location, characteristics, and utilization of alcohol and drug treatment facilities and services throughout the 50 States, the District of Columbia, and other U.S. jurisdictions.  The Office of Applied Studies at SAMHSA plans and directs N-SSATS.  N-SSATS collects information from all facilities in the United States, both public and private, that provide substance abuse treatment.  It provides the mechanism for quantifying the character and composition of the U.S. substance abuse treatment delivery system.  Two questions on the survey relate to housing/homelessness status: the first asks whether the facility provides assistance to clients in locating housing or provides residential beds for clients children, and the second asks whether the facility operates a halfway house or other transitional housing program for substance abuse clients at the location.

SAMSHA is also currently working to link data across block grants and discretionary programs by implementing National Outcome Measures (NOMs), though this is not yet completed.  In FY2005, states began reporting on performance.  Stability in Housing is one of the new NOMs that is being collected and recorded.

Program Website:  http://www.samhsa.gov/Matrix/programs_treatment_SAPT.aspx

Other Relevant Websites:

• http://www.hhs.gov/homeless/grants/index.html#MainstreamPrograms
• http://www.nationaloutcomemeasures.samhsa.gov/./outcome/index_2007.asp
• http://www.icpsr.umich.edu/cocoon/SAMHDA/DAS3/00056.xml
• http://wwwdasis.samhsa.gov/05nssats/NSSATS2k5Chp1.htm

---

[ Go to Contents ]

Endnotes

[1]     A recent study of homelessness based on data obtained from a nationally representative sample of communities found that about 38 percent of homeless persons in households with children stayed with family or friends prior to entering a shelter.  (U.S. Department of Housing and Urban Development, Office of Community Planning and Development, Third Annual Homeless Assessment Report to Congress, July 2008.)

[2]       Formerly the Food Stamp Program.

[3]       Information about TEDS is available at http://www.icpsr.umich.edu/SDA/SAMHDA/04626-0001/CODEBOOOK/4626.htm (accessed on October 14, 2008).

[4]       This would require an agreement for data access with the Food and Nutrition Service of the U. S. Department of Agriculture, which administers SNAP.

[5]       HHS targeted homelessness programs include: Grants for the Benefit of Homeless Individuals, Health Care for the Homeless, Programs for Runaway & Homeless Youth, Projects for Assistance in Transition from Homelessness, and the Title V Federal Surplus Property Program.

[6]       HHS mainstream programs relevant to homelessness include: Community Mental Health Services Block Grant, Community Services Block Grant, Consolidated Health Center Program, Head Start, Maternal & Child Health Services Block Grant, Medicaid, Ryan White HIV/AIDS Program, Social Services Block Grant, State Childrens Health Insurance Program, Substance Abuse Prevention and Treatment Block Grant, and Temporary Assistance for Needy Families.

[7]       Coverage of the TANF Population Under Medicaid and SCHIP, Elicia J. Hertz, CRS Report RS22035, Washington: Library of Congress, Congressional Research Service, 2005.  Also, Administration for Children & Families, Fact Sheet: Office of Family Assistance (OFA), U.S. Department of Health and Human Services, http://www.acf.hhs.gov/opa/fact_sheets/tanf_factsheet.html (accessed August 5, 2008).

[8]       ACF, FACT SHEET.

[9]       CRS Report RS22035, Hertz, 2005.

[10]     StateHealthFacts.org, State Health Facts, The Henry J. Kaiser Family Foundation, http://www.Statehealthfacts.org (accessed August 1, 2008).

[11]     CRS Report RS22035, Hertz, 2005.

[12]     In 2003 the GAO reported an average processing time of 4 months for initial decisions and one year for appeals and put the disability programs on its list of high-risk programs.  The Social Security Advisory Board (SSAB) reviewed even more recent data in a May 2006 report, suggesting that processing times for SSI applications have continued to rise, with times for appeals hearings averaging 422 days in 2005.

[13]     The use of these consolidated or integrated applications was also found in a 2003 study conducted by the Urban Institute about the application process for TANF, Medicaid, food stamps, and SCHIP in six states.  However, that study found that a consolidated TANF and Medicaid application was in use in Raleigh NC at the time of that study but as of the time of our interviews in 2008 the state reported using separate applications for Medicaid and TANF.  See Holcomb, et al., The Application Process for TANF, Food Stamps, Medicaid, and SCHIP: Issues for Agencies and Applicants, Including Immigrants and Limited English Speakers, Washington, DC: Urban Institute, prepared for the Department of Health and Human Services, 2003.

[14]     A recent study of homelessness based on data obtained from a nationally representative sample of communities found that about 38 percent of homeless persons in households with children stayed with family or friends prior to entering a shelter.  (U.S. Department of Housing and Urban Development, Office of Community Planning and Development, Third Annual Homeless Assessment Report to Congress, July 2008.)

[15]     Massachusetts, Missouri, New Mexico, New York, Louisiana, Oregon, Rhode Island, South Carolina, Utah, Virginia, and West Virginia

[16]   Title 1 of the McKinney-Vento Act defines homelessness.  See United States Code Title 42, Chapter 119, Subchapter 1 (42 U.S.C 11302).

[17]     For example, Subtitle B of Title VII of the McKinney-Vento Act addresses education for homeless children and youth programs administered by the US Department of Education.  It uses a broader definition of homelessness.  See 42 U.S.C. 11431 (section 725).

[18]     7 C.F.R. § 271.2, definition of homeless individual.

[19]     FirstStep, An Introduction to the Food Stamp Program, Center for Medicaid Services, http://www.cms.hhs.gov/apps/firststep/content/foodtips.html (accessed August 5, 2008).

[20]     See 7 C.F. R. § 273.10(e) (G).  Additional detail is available from the Food Research and Action Center, Homeless Persons Rights Under the Food Stamp Program, FRAC, http://www.frac.org/html/federal_food_programs/FSP/Homeless_Rights_Under_FSP.html (accessed August 5, 2008).

[21]     Sara Simon Tompkins, Homelessness and Food Stamps: Rights and Barriers, National Law Center on Homelessness & Poverty, http://www.nlchp.org/content/pubs/Homelessness%20and%20Food%20Stamps%2020031.pdf (accessed August 5, 2008).

[22]     A complete list of persons interviewed is included in the Appendix E.

[23]     Terry L. Cline, A Life in the Community for Everyone: SAMHSA, SAMHSA, http://www.samhsa.gov/Matrix/matrix_brochure.pdf (accessed August 7, 2008).

[24]     Elaine Parry, SAMHSA Action Plan: Homelessness, Fiscal Years 2006 2007, SAMHSA,  http://www.samhsa.gov/matrix/SAP_homeless.aspx (accessed August 7, 2008).

[25]     We use the term program beneficiary to refer to the individual or household that receives the service or benefit funded by the mainstream program.  For instance, an individual served by a program funded with CSBG grants would be a beneficiary, as would a TANF recipient or a patient of a Consolidated Health Clinic.  However, this notion applies to TANF or Medicaid in a way that it wouldnt apply to at least some of the expenditures under CSBG or the other block grants, which are used to fund broader types of programs that do not necessarily provide assistance directly to individuals or families.

[26]     This would require an agreement for data access with the Food and Nutrition Service of the U. S. Department of Agriculture, which administers SNAP.

[27]     A third option, which would involve modifying the data collection and reporting requirements for program grantees, was considered, but such modifications would likely require legislative or regulatory changes.

[28]     Unduplicated numbers served means that the count of people served reflects the number of unique households or persons served across all service providers and grantees.  That is, if an individual receives substance abuse inpatient treatment at one facility and outpatient treatment from another, an unduplicated count would report that one individual received two services, whereas duplicated data would report that two individuals received one service each.  Unduplicated numbers are much more meaningful for HHS, since they will provide a more accurate measure of the extent and intensity of program utilization.

[29]     Information about TEDs is available at http://www.icpsr.umich.edu/SDA/SAMHDA/04626-0001/CODEBOOOK/4626.htm (accessed on October 14, 2008).

[30]     There are many methods that could be explored for extrapolation or imputation if HHS is interested in considering this approach.

[31]     We will comment more on the possibility of matching mainstream program data with TANF or Medicaid datasets in the final report once we have completed the analysis of the availability and usability of these data.

[32]     Alabama Medicaid Agency, Welcome to the Alabama Medicaid Agency, State of Alabama, http://www.medicaid.state.al.us/> (accessed July 17, 2008).

[33]     Office of Public Assistance, Medicaid, State of Alaska, http://www.hss.state.ak.us/dpa/programs/medicaid/> (accessed July 17, 2008).

[34]     AHCCCS, About AHCCCS, State of Arizona http://www.azahcccs.gov/Site/AboutUs.asp (accessed July 17, 2008).

[35]     Arkansas Department of Human Services, Program Overview: State Fiscal Year 2007, State of Arkansas, http://www.medicaid.state.ar.us/Download/general/OverviewSFY07.pdf> (accessed July 17, 2008).

[36]     California Department of Health Care Services, Medi-Cal, http://www.medi-cal.ca.gov/> (accessed July 17, 2008).

[37]     Colorado Department of Health Care Policy and Financing, Welcome to the Department of Health Care Policy and Financing, http://www.chcpf.state.co.us/default.asp> (accessed July 17, 2008).

[38]     Connecticut Department of Social Services, Health Care, State of Connecticut, http://www.ct.gov/dss/cwp/view.asp?a=2353&q=305218#MAID> (accessed July 17, 2008).

[39]     Delaware Medical Assistance Program, Delaware Medical Assistance Program, State of Delaware, http://www.dmap.state.de.us/home/index.html> (accessed July 17, 2008).

[40]     Agency for Health Care Administration, Florida Medicaid, State of Florida, http://www.fdhc.state.fl.us/Medicaid/flmedicaid.shtml> (accessed July 17, 2008).

[41]     Georgia Department of Community Health, Medicaid, State of Georgia, http://dch.georgia.gov/00/channel_title/0,2094,31446711_31944826,00.html> (accessed July 17, 2008).

[42]     Med-QUEST Division, Eligibility: Assistance Programs, State of Hawaii, http://www.med-quest.us/eligibility/EligPrograms.html> (accessed July 17, 2008).

[43]     Idaho Department of Health and Welfare, Idaho Medicaid Program, State of Idaho, http://www.healthandwelfare.idaho.gov/site/3629/default.aspx> (accessed July 17, 2008).

[44]     Illinois Department of Healthcare and Family Services, Medical Programs Listings State of Illinois, http://www.hfs.illinois.gov/programs/> (accessed July 17, 2008).

[45]     Indiana Medicaid Indiana Health Coverage Programs, State of Indiana, http://www.indianamedicaid.com/ihcp/index.asp> (accessed July 17, 2008).

[46]     Iowa Department of Human Services, HEALTH CARE - Division of Financial, Health and Work Supports, State of Iowa, http://www.dhs.iowa.gov/Consumers/Health/Medical_Insurance/WhatsAvailable.html> (accessed July 17, 2008).

[47]     Kansas Health Policy Authority, Medicaid and SCHIP (HealthWave Program) Fact Sheet, State of Kansas, p. 1.

[48]     Cabinet for Health and Family Services, Department for Medicaid Services, Commonwealth of Kentucky, http://chfs.ky.gov/about/about.htm> (accessed July 17, 2008).

[49]     Louisiana Department of Health and Hospitals, Management & Finance, State of Louisiana, http://www.dhh.state.la.us/offices/?ID=78> (accessed July 17, 2008).

[50]     Office of MaineCare Services, Office of MaineCare Services, State of Maine, http://www.maine.gov/dhhs/bms/> (accessed July 17, 2008).

[51]     Maryland Medical Programs, About our Programs, State of Maryland, http://www.dhmh.state.md.us/mma/mmahome.html> (accessed July 17, 2008).

[52]     Health and Human Services, MassHealth, Commonwealth of Massachusetts, http://www.mass.gov/?pageID=eohhs2agencylanding&L=4&L0=Home&L1=Government&L2=Departments+and+Divisions&L3=MassHealth&sid=Eeohhs2> (accessed July 17, 2008).

[53]     Medical Services Administration, Medicaid Programs, State of Michigan, http://www.mdch.state.mi.us/msa/mdch_msa/msahome.HTM> (accessed July 17, 2008).

[54]     Department of Human Services, Medical Assistance, State of Minnesota, http://www.dhs.state.mn.us/main/idcplg?IdcService=GET_DYNAMIC_CONVERSION&RevisionSelectionMethod=LatestReleased&dDocName=id_006254> (accessed July 17, 2008).

[55]     Mississippi Division of Medicaid, Medicaid, State of Mississippi, http://www.dom.state.ms.us/> (accessed July 17, 2008).

[56]     Missouri Department of Social Services, Medical Services MO HealthNet, State of Missouri, http://www.dss.mo.gov/fsd/msmed.htm> (accessed July 17, 2008).

[57]     Department of Public Health and Human Services, Montana Medicaid Provider Information, State of Montana, http://medicaidprovider.hhs.mt.gov/> (accessed July 17, 2008).

[58]     Department of Health and Human Services, Nebraska Medicaid Program: Homepage, State of Nebraska, http://www.hhs.state.ne.us/med/medindex.htm> (accessed July 17, 2008).

[59]     Division of Health Care Financing and Policy, Nevada Medicaid State Plan, State of Nevada, http://dhcfp.state.nv.us/MSP%20Table%20of%20Contents.htm> (accessed July 17, 2008).

[60]     Department of Health and Human Services, Medicaid Program, State of New Hampshire, http://www.dhhs.state.nh.us/DHHS/MEDICAIDPROGRAM/default.htm> (accessed July 17, 2008).

[61]     Division of Medical Assistance and Health Services, New Jersey Medicaid, State of New Jersey, http://www.state.nj.us/humanservices/dmahs/dhsmed.html> (accessed July 17, 2008).

[62]     Human Services Department, Medicaid Public Information, State of New Mexico, http://www.hsd.state.nm.us/mad/medpublicinfo.html> (accessed July 17, 2008).

[63]     Department of Health, Medicaid in New York State, New York State, http://www.health.state.ny.us/health_care/medicaid/index.htm> (accessed July 17, 2008).

[64]     Division of Medical Assistance, North Carolina Medicaid, State of North Carolina, http://www.dhhs.state.nc.us/dma/> (accessed July 17, 2008).

[65]     Department of Human Services, General Information: About Medicaid, State of North Dakota, http://www.nd.gov/dhs/services/medicalserv/medicaid/info.html> (accessed July 17, 2008).

[66]     Department of Jobs and Family Services Welcome to Ohio Medicaid, State of Ohio, http://jfs.ohio.gov/ohp/> (accessed July 17, 2008).

[67]     Oklahoma Health Care Authority, What is SoonerCare?  State of Oklahoma, http://www.ohca.state.ok.us/individuals.aspx?id=52&parts=152R_7451_7449> (accessed July 17, 2008).

[68]     Department of Human Services, Division of Medical Assistance Programs (DMAP, State of Oregon, http://www.oregon.gov/DHS/aboutdhs/structure/dmap.shtml> (accessed July 17, 2008).

[69]     Department of Public Welfare, Health Care/Medical Assistance, Commonwealth of Pennsylvania, http://www.dpw.state.pa.us/ServicesPrograms/MedicalAssistance/> (accessed July 17, 2008).

[70]     Rhode Island Department of Human Services, Rhode Island Department of Human Services, State of Rhode Island, http://www.dhs.state.ri.us/> (accessed July 17, 2008).

[71]     Department of Health and Human Services, South Carolina Healthy Connections (Medicaid) Program, State of South Carolina, http://www.dhhs.state.sc.us/dhhsnew/InsideDHHS/Bureaus/EligiblityPolicyAndOversight/Partners%20for%20Health%20(Medicaid)%20Program.asp> (accessed July 17, 2008).

[72]     South Dakota Department of Social Services, Medical Eligibility, State of South Dakota, http://dss.sd.gov/medicaleligibility/index.asp> (accessed July 17, 2008).

[73]     Bureau of TennCare, Welcome to TennCares Website, State of Tennessee, http://www.state.tn.us/tenncare/> (accessed July 17, 2008).

[74]     Health and Human Services Commission, Texas Medicaid Program, State of Texas, http://www.hhsc.state.tx.us/medicaid/> (accessed July 17, 2008).

[75]     Utah Department of Health, Utah Medicaid Program, State of Utah, http://hlunix.hl.state.ut.us/medicaid/> (accessed July 9, 2008).

[76]     Green Mountain Care, Vermont Expands Health Coverage, State of Vermont, http://www.greenmountaincare.org/> (accessed July 17, 2008).

[77]     Virginia Department of Social Services, Virginia DDS Vol. XIII: Medicaid Eligibility Manual, p. 5.

[78]     Health and Recovery Services Administration, Medicaid State Plan: What is the State Plan, State of Washington, http://fortress.wa.gov/dshs/maa/medicaidsp/> (accessed July 17, 2008).

[79]     Mountain Health Choices, BMS Medicaid Redesign: Mountain Health Choices, State of West Virginia, http://www.wvdhhr.org/bms/oAdministration/Medicaid_Redesign/MedRedesign_main.asp> (accessed July 17, 2008).

[80]     Department of Health and Family Services, Wisconsin Medicaid: Who is Eligible?  State of Wisconsin, http://dhs.wisconsin.gov/medicaid/> (accessed July 17, 2008).

[81]     Wyoming Department of Health, EqualityCare Org Chart, State of Wyoming, http://wdhfs.state.wy.us/healthcarefin/equalitycare/index.html> (accessed July 17, 2008).

[82]     DC Department of Health, Medicaid HIPAA, District of Columbia, http://doh.dc.gov/doh/cwp/view,a,1371,q,581500,dohNav_GID,1841,.asp> (accessed July 17, 2008).

[83]     Administration for Children and Families, Names of State TANF Programs, U.S. Department of Health and Human Services, http://www.acf.hhs.gov/programs/ofa/states/tnfnames.htm (accessed July 17, 2008).