In Massachusetts a few years ago, data from homeless shelters revealed that a majority of residents had entered the system almost directly from foster care, prison or hospitals. As a result, then-governor Paul Cellucci committed to zero tolerance for discharge from government institutions into homelessness. In New York City, a recent assessment of a large, publicly-funded supportive housing initiative (New York/New York Agreement to House Homeless Mentally Ill Individuals) showed that the cost of permanent housing and services for disabled homeless people was only marginally more than the public cost of leaving them homeless. In this time of scarce resources (at least for those most needy), continuing affordable housing crisis and growing homeless numbers, we must demand more resources and make the best use of those that we have.
In 2001 Congress asked the U.S. Department of Housing and Urban Development (HUD) to take the lead in requiring communities to develop an unduplicated count of the homeless. It did this for two reasons. First, it was persuaded that cities with good data systems (Columbus, New York City and Philadelphia) did a better job of addressing the problem. Second, it was frustrated by the almost total lack of reliable national data about homelessness and the impact of federal spending on the problem. It wanted communities to have accurate data on homelessness and also use this information to evaluate patterns of program use and effectiveness. As good data emerged from communities, Congress asked HUD to use it to paint a more accurate picture of what was happening nationally.
To meet these goals, HUD required federally funded public and nonprofit organizations to implement a homeless tracking system – Homeless Management Information Systems, or HMIS. HUD provided technical assistance to HMIS and allowed for funding from the Supportive Housing Program (SHP). In July 2003 HUD released draft HMIS standards that laid out the information it expects agencies to gather through HMIS. Final standards are anticipated in early 2004, the deadline for agencies to implement HMIS.
Creating a useful tracking system is a formidable and challenging task. Issues of privacy, expense, participation and usage are just some of the concerns that must be addressed.
Confidentiality. While this is critical for all homeless people, special issues arise for victims of domestic violence and people with HIV and AIDS. Further, there are laws requiring protection of individual information (such as the federal Health Information Portability and Accountability Act). Confidentiality must be carefully maintained in information gathering, analysis protocols and data security.
Resources. The hardware, software, transmission, training and staff costs associated with implementing HMIS can be costly. Although HUD has allowed communities to apply for SHP funds to cover these costs, such requests compete with those for direct services and housing.
Analysis. One real benefit of HMIS is the information it can generate about the cost of homelessness, program outcomes and linkages with mainstream systems (government programs like foster care, employment training, health care and corrections that assist low-income people more broadly, or have custodial responsibilities). Obtaining such information requires resources and a sophisticated approach to both research design and data analysis. While few homeless assistance organizations possess these resources or skills, they can partner with agencies or communities that do.
Coverage. HMIS can contribute valuable information if they “cover” the maximum number of homeless people. Grantees of other federal programs (Health Care for the Homeless, PATH) and programs that do not receive government funding (missions, soup kitchens, clothing banks) should be included in order to gather the most complete set of data.
Use of Data. While HMIS data can tell us many things about the homeless, there are only certain kinds of information that HMIS should provide. Aggregating anonymous data to examine numbers and trends is relatively uncontroversial. More troublesome are possible uses of personal data, including sharing client files among programs. HUD has stipulated that it wants only aggregated, not client level, data.
These issues will have to be addressed at the community level. The real challenges of HMIS are not technical, they are civic. Communities must grapple with how they will handle information sharing, right to privacy and program evaluation. HUD should not tell communities how to deal with these matters, but should help with standards and protocols, best practices and technical assistance.
There are at least seven opportunities for well-run data systems to make a tremendous difference in the lives of homeless people and communities.
1. Plan to End Homelessness. Communities across the nation are developing plans to end homelessness. HMIS can help measure the success of their implementation.
2. Encourage Mainstream Participation. Data has shown that homelessness is extremely costly to public health care and other systems. Identifying the costs can convince those systems to more actively address homelessness.
3. Program Management. All homeless programs collect information for a number of reasons – to assess capacity, manage staff, allocate resources or prepare budgets and reports. HMIS can be used to manage and simplify these tasks.
4. Attract Resources. In a climate of intense competition for resources, homelessness organizations can use HMIS data to make a more compelling case for funding.
5. Assess Costs. Agencies can use HMIS to assess their costs and the cost effectiveness of their programs. States and localities can use HMIS to assess the cost effectiveness of various programs or interventions.
6. Prevention. HMIS can help identify where the homeless come from and who is most likely to become homeless.
7. Measure Outcomes. HMIS can be used to assess the impact of service and housing interventions on meeting immediate needs and the long-term goal of ending homelessness.
Creating good homeless management information systems is difficult work, with many and significant challenges, but making the best use of our existing resources, and building a strong case for new ones, will require more than a big heart and an apt anecdote. HMIS is not, of course, the solution to homelessness. It can, however, help us move past managing the problem, to ending it.