Targeting Interventions for Homeless Families and Individuals: Program

1
Targeting Interventions for Homeless Families and
IndividualsProgram System Approaches to
Better Targeting
Tom Albanese July 28, 2008

• National Alliance to End Homelessness
• Annual Conference

2
Program System Change Process Steps

• Step 1 Assess current client characteristics/need
  s
• Step 2 Assess current program/system model,
  resources and performance
• Step 3 Investigate other approaches, best
  practices
• Step 4 Design future program/system and outline
  goals
• Step 5 Implement and evaluate a pilot to test
  new approach and assumptions
• Step 6 Implement program/system change

3
Step 1 Assess current client characteristics/need
s

• Admitted/served
• Not admitted/served due to
• barriers to access
• choose not to access
• needs greater than program/system capacity
• ineligible or denied
• pending-waitlist
• Demographics, resources, service utilization,
  housing history, etc.
• Q How do those admitted/served differ from those
  not admitted/served?
• Q Whos not being served and why?
• Q Are clients being linked to the appropriate
  service and/or housing resource(s)? How is this
  done? Is there an assessment tool currently used
  to determine the best service/housing linkage?

4
Step 2 Assess current program/system model,
resources, and performance

• Current Program/System Model
• mission/purpose
• target population
• referral sources
• services
• housing
• partnerships
• staffing
• Q Does the mission/purpose reflect community
  values and client needs?
• Q Is there a clear system mission/purpose?
• Q Is the target population efficiently
  identified and linked to the program/system? How
  do individuals/families access and flow through
  the program/system?
• Q How are system resources-services and
  housing-organized? Are they organized in a way
  that most efficiently and effectively links
  clients with differing needs to the appropriate
  intervention?
• Q How do persons without shelter access the
  system? Is there a single point of access,
  multiple points, centralized administration/decent
  ralized access, other means?
• Q What works well and should be retained?

5
Step 2 Assess current program/system model,
resources, and performance

• Program/System Resources
• current cash/in-kind services, operating
• potential cash/in-kind services, operating
• Q Are program resources sufficient to achieve
  program objectives and outcomes?
• Q How flexible are program resources? Can
  program resources be used in different ways?
• Q Do funders have an interest in better
  targeting/service matching?
• Q What other resources are available to support
  program/system objectives and outcomes?

6
Step 2 Assess current program/system model,
resources, and performance

• Program/System Performance
• Outputs number served, nights of service,
  average length of stay, etc.
• Outcomes housing, graduation rate,
  income/benefits, educational attainment,
  health/recovery, recidivism, etc.
• Q Whats different between those deemed
  successful and those not successful? Are those
  who most need the service also the most likely to
  not be successful?
• Q Does the program/system model facilitate or
  hinder achievement of program outcomes?
• Q Is housing stability a primary goal? Why or
  why not?
• Q How are clients who dont progress toward
  housing stability (i.e. long-stayers,
  chronically homeless) identified?
• Q How are outcomes monitored? What is the
  program/system consequence for non-achievement?
• Q Do program outcomes align with client, funder,
  community expectations and requirements?

7
Example Analysis of Family Homeless Episodes
of Homeless Episodes Among Families
of Homeless Episodes Among Individuals
Source Montgomery Co., MD HMIS data analysis,
2005-2006
8
Example Analysis of Family Homeless Episodes
of Days per Episode of Homelessness
Families
Individuals
Q What are the characteristics of families with
one episode? Multiple episodes? Whats
different between these groups?
Source Montgomery Co., MD HMIS data analysis,
2005-2006
9
Step 3 Investigate other approaches, best
practices

• Benchmarking
• The American Productivity Quality Center
  (www.apqc.org) defines benchmarking as the
  process of identifying, learning, and adapting
  outstanding practices and processes from any
  organization, anywhere in the world, to help an
  organization improve its performance.
  Benchmarking gathers the tacit knowledge--the
  know-how, judgments, and enablers--that explicit
  knowledge often misses.
• programs locally or in other communities that
  serve the target population and have good
  outcomes
• direct contact (e.g. interviews) with key
  contacts or review of related documentation or
  research
• Q Who else is serving the target population and
  doing it well? How do they identify and engage
  the target population?
• Q What works in other programs/communities that
  can be replicated?
• Q How are they funded?
• Q What key partnerships support their model?
  How are the partnerships managed? At system or
  program level or both?
• Q What tools, forms, policies, procedures
  support their model? Can these be
  shared/modified?

10
Step 4 Design future program/system and outline
goals

• Program/System Plan
• mission/purpose
• target population prioritization/selection
  criteria
• referral sources
• client pathway first contact to service
  termination
• program/system services, housing
• partnerships, on-site or off-site
• resources needed
• client outcomes
• Q How can interventions (services/housing
  offered) and outcomes be differentiated according
  to differing client needs?
• Q Who identifies the target population? Whats
  the most efficient approach?
• Q Do partner expectations align with the program
  model?
• Q Does one or more funding sources need to
  approve a modified program model?
• Q Does program/system quality and performance
  monitoring need to be changed to better inform
  management and funders?

11
Step 5 Design and implement a pilot

• Pilot the proposed program/system model to test
• assumptions
• triage/assessment tools, other forms
• policies/procedures
• process efficiency
• client outcomes
• resources needed
• staff skills needed
• Develop evaluation plan before implementing What
  is being evaluated? What data needs to be
  collected? How/when/where/who will collect data?
• Retain expert evaluator if feasible
• Use findings to inform full program/system
  implementation What was learned? What needs to
  change? Is further testing needed?

12
Step 6 Implement program/system change

• Take successful aspects of pilot to scale-i.e.,
  full program/system implementation
• Develop evaluation plan to test full
  implementation at regular intervals
• Retain expert evaluator if feasible
• Use findings to inform future program/system
  improvements

13
Example Columbus Rebuilding Lives Plan

• Research concluded that
• permanent supportive housing successfully ended
  homelessness for chronically homeless individuals
• Multiple and diverse PSH options
• No single, easy to navigate access point for
  clients and advocates clients with greatest
  needs not always admitted
• Because supportive housing is scarce and valuable
  resource, need to prioritize admission for those
  with the greatest needs
• RL Plan Strategy Create a unified system for
  supportive housing which includes centralized
  eligibility determination, prioritization and
  placement, periodic review of tenant needs, and
  move up incentives to encourage tenants to be
  more independent.
• 2008-2010 pilot aspects of future system
• Centralized referral, assessment, matching at two
  new PSH projects
• Periodic tenant review/assessment and move-up
  assistance with existing PSH projects
• 2011 full system implementation

14
Example Columbus Rebuilding Lives Plan

• Targeting to focus on individuals and families
  who
• Meet local Rebuilding Lives criteria (long-term
  homeless, disabled)
• Are identified using administrative data (e.g
  HMIS) as high service utilizer and/or meet other
  high risk/vulnerability criteria
• System level administrative process to
  screen/link prospective clients to PSH options
• Clients not deemed eligible and priority referred
  to less intensive assistance (i.e., transitional
  assistance such shallow rent subsidies, etc.)

15
Other key considerations

• No one size fits all approach to targeting and
  assessment
• Keep it simple-for clients, staff, other
  stakeholders
• Easily administered scales to identify level of
  need, appropriate intervention
• Readily accessible process for clients/advocates
• Does everyone need to agree? Who needs to
  buy-in?
• Success may be contingent on buy-in and support
  from providers, funders, consumers, advocates and
  other key stakeholders. Design a process that
  gathers and uses input from key stakeholders from
  the onset.
• Intentional, thoughtful communication key local
  government, stakeholders, providers, users
• How will other providers react? Other systems of
  care? Will changes create stress for other
  providers/systems or require them to make
  adjustments?
• What is the cost of change and how much are you
  willing to spend in dollars and political
  capital?
• Good quality data are essential to both program
  design (knowing who youre targeting) and
  implementation (knowing when youve successfully
  identified and served intended target population)
  
• Is the process (current or proposed) consumer
  friendly (e.g., easily navigated, client choice,
  due process if denied, etc.). If not, make
  changes.

16
Tom Albanese Abt Associates Inc. tom_albanese_at_abta
ssoc.com