Housing Homeless People with Serious Mental Illness in California September 13-14, 2005 St. Louis, Missouri Sponsored by the U.S. Department of Housing and Urban Development Martha R. Burt
Jan 14, 2016
Housing Homeless People with Serious Mental Illness in California
September 13-14, 2005St. Louis, Missouri
Sponsored by the U.S. Department of Housing and Urban Development
Martha R. Burt
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
2
Overview
• Describe data analysis of clients and outcomes in a California program for homeless people with serious mental illness
• Use to illustrate some of the ways that data in an HMIS, if it were the right data and covered the right programs, might produce similar analyses
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
3
What Is AB 2034?
• California legislation providing flexible funding (“whatever it takes”) to counties
• Focus—people with serious mental illness, with or without co-occurring substance abuse, who are homeless or high risk
• Goals are to • End homelessness for this population,
especially long-term homelessness• Prevent returns to homelessness or first-time
homelessness for people with SMI leaving institutions
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
4
Where Does It Operate; How Many Does It Serve?
• First legislation passed in 1999; funded 3 pilot counties to start services in 2000
• Expanded in 2001 based on preliminary evidence of success
• Now 52 programs—18 in Los Angeles County, 34 in other counties (1 county has 2 programs)
• Over 11,000 people served to date; current enrollment around 5,000
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
5
How Does It Document Its Activities?
• Extensive systematic data collection using common system (CAMINAR)
• Components:• Circumstances at enrollment (demographics,
homelessness, housing status, hospitalizations, incarcerations, employment, income sources)
• Experiences while enrolled (same categories)• Disenrollment timing, reasons, and last known
housing status
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
6
Potential for HMIS
• HMIS could collect much of the same data that CAMINAR does:• Circumstances at enrollment (demographics,
homelessness, housing status, hospitalizations, incarcerations, employment, income sources)
• Experiences while enrolled (same categories)• Disenrollment timing, reasons, and last known
housing status
• To do this, HMIS would have to cover PSH and also SSO programs
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
7
Why Is It Important to Know About AB 2034?
• Important in own right, as model legislation, model programs, and significant evidence of success in housing long-term homeless people
• New money coming—• Proposition 63 passed in November 2004,
provides masses of service money to Department of Mental Health to expand the types of programs that AB supports
• Also Prop 43 in 2003, provided new capital resources for housing for disabled people (although most already committed)
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
8
Why Do I Have These Data?
• Passage of Prop 63 meant DMH was going to have to figure out how to distribute money, what to ask for, what outcomes to expect/require
• Needed more information on AB programs, about their outcomes, their housing strategies, the challenges they have faced, and how they have responded
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
9
Why Do I Have These Data?
• Corporation for Supportive Housing (CSH) offered to assemble the relevant data
• Evaluation sponsors were:• California Institute of Mental Health
(association of county mental health agencies, which mostly have responsibility for AB programs)
• Conrad N. Hilton Foundation
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
10
Purpose of Evaluation
Identify the range of housing strategies implemented by programs in the 34 counties that have received AB2034 funding Data that would be external to HMIS
Determine association between being able to house consumers and being able to retain consumers (lower dropout rate) Should be able to get relevant data from HMIS
Determine if the housing strategies or range of strategies that a county (or program within a county) employed made a difference for successful housing outcomes Analysis that would combine HMIS and other data
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
11
• Analyzed housing outcome data available from the website www.ab34.org
• Mental Health Association of Los Angeles provided further analysis of the CAMINAR data
• These two data sources are most similar to HMIS
• “Housing Strategies Survey”• Developed survey and conducted phone interviews with
AB2034 Coordinators or designated staff to gather survey responses
• Completed 41 surveys covering 44 programs (of 53) in 28 different counties, including 2 programs in two counties and 13 of Los Angeles’ 19 programs
• Response rate = 83% of all programs
Data Sets Used
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
12
• Set up 6 criteria for “hard to serve” (HtS)• 70+% homeless at enrollment• 60% street homeless at enrollment• 180+ average homeless days in 12 months
before enrollment• 40+% incarcerated in 12 months before
enrollment• 50+ average incarcerated days in 12 months
before enrollment• 60+% with co-occurring substance abuse
• Could do with HMIS data
Who Are AB Programs Serving?
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
13
% of Programs Meeting HtS Criteria (of 28 state and 18 LA programs with data)
0
2
4
6
8
10
12
14
16
70% HL 60%street
HL
180+days HL
40%incar
50+daysincar
60%dual
State (28)
LA (18)
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
14
• % retained (currently enrolled), of everyone ever enrolled
• Median: 42% (state); 44% (LA) • Range: 17-82% (state); 27-69% (LA)
• % housed, of everyone ever enrolled• Median: 37% (state); 39% (LA)• Range: 12-62% (state); 25-56% (LA)
• Relationship between % housed and % retained:
•.929• Or, “if you house them, you keep them”
AB Program Outcomes(all potentially calculable with HMIS data)
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
15
• Programs that enroll “more challenging” consumers (those that have had longer histories of homelessness or more barriers to housing stability) are not getting worse results in terms of housing outcomes
• In fact sometimes the results are better—higher average homeless days in year before enrollment is associated with higher proportion housed
• If client difficulty indicators suggest “Housing Unreadiness” to those who adhere to this concept, AB data show the concept is not a good predictor of housing outcomes
Housing Outcomes
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
16
• How quickly do consumers leave AB2034 programs?
• 56% of those who have enrolled in AB2034 programs have dis-enrolled. Among programs, the range is 17% to 83%. Of the 56% who have dis-enrolled:
• 25% dis-enrolled by 6 months• 51% dis-enrolled by 12 months• “dropouts” account for 51% of the reasons for dis-
enrollment• Other reasons include rehospitalization, death,
movement to a lower level of care, and being inappropriate for the program (not having SMI, once substance abuse symptoms have subsided)
Dis-enrollment(all calculable with HMIS data, if you collect it)
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
17
• There is wide variation among counties/programs in the proportion of ever-enrolled consumers who are now in stable housing. The range is 12% to 62%.
• There is a very strong correlation (.929) between having a high proportion of ever-enrolled consumers who are in housing and having a very low proportion of consumers who dis-enroll.
• Also, the faster a program is able to move clients into housing from homelessness, the higher the retention rate.
Housing Makes a Very Big Difference
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
18
Some strategies are being used by nearly every county/program – and offered to virtually every consumer. These include:
• Advocacy on behalf of individuals to help them find and get housing
• Supportive services to help people keep housing
• Back-up problem-solving help for landlords
Housing Strategies
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
19
Other widely implemented housing strategies include:
• Assist consumers to apply for housing subsidies (all do)
• Provide short-term subsidies or help with move-in costs (all do, & most consumers get)
• Provide long-term rent subsidies to some consumers (all do, but few consumers get)
• Provide temporary or transitional housing to get people off the streets and/or for respite/crisis
Housing Strategies, cont’d
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
20
Some strategies are used in about half of the counties/programs:
• Actively recruiting landlords, systematically finding available units, making arrangements with landlords to secure the next vacant unit
• Master-leasing buildings or apartments within buildings and sub-leasing units to consumers
• Use AB2034 funds to secure dedicated or set-aside units for consumers
Housing Strategies, cont’d
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
21
Strategies more likely to be implemented in large/urban counties:
• Administer Section 8 or Shelter + Care rent subsidies that are available to consumers
• Partner with Housing Authority or other public agencies that control rent subsidies
• Work with PSH providers; partner to supply services if the other partner does development and/or operations
Housing Strategies, cont’d
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
22
Strategies more likely to be implemented in smaller/rural counties:
• Offering maintenance or cleaning – either ongoing to help tenants keep housing or when tenants move out to keep landlords willing to rent to other consumers
Housing Strategies, cont’d
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
23
Effective Strategies
• Many counties/programs offer a range of housing strategies including: Partnering with housing providers Recruiting landlords Securing rent subsidies or set-aside units
• No single strategy appears to be more effective than others• Nor do the total number of strategies used make much
difference• However, working with PSH providers appears effective in
retaining clients for 24+ months• And using your AB funds to pay rents long-term appears to
be the least effective strategy• Counties/programs that have been most successful have
focused on expanding the supply or range of housing available to consumers, rather than just helping each individual find housing
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
24
What AB Providers Say Is Needed Most
• More affordable housing• Permanent supportive housing• Landlords and property managers who are
tolerant and understanding of consumers• “Wet/Damp” and “Harm Reduction” housing
models for people with substance use problems• Support for both landlords and tenants• Wide range of housing options
September 13-14, 2005 St. Louis, MissouriSponsored by the U.S. Department of Housing and Urban Development
25
Implications for HMIS
• Assess the types of analyses that will be most useful to your community for:• Tracking outcomes• Convincing funders that they should invest in particular types
of programs
• Determine how much of the relevant data you could collect through HMIS
• Determine what other data you will need (in the AB case, information about the interventions)
• The AB study was done very quickly; was possible because the CAMINAR data were already available
• You could set up your HMIS system to do the same