Presentation on theme: "Rapid Re-Housing Research Evidence and Beyond"— Presentation transcript:
1 Rapid Re-Housing Research Evidence and Beyond Jamie TaylorCloudburst Consulting GroupConnecticut Coalition to End Homelessness Training InstituteMay 8, 2014
2 Objectives – Rapid Re-Housing Overview 1) RRH - Results from across the nation 2) RRH - The Philadelphia Story 3) RRH – Promising Practices 4) RRH – Local RRH Evaluation
3 Did RRH help decrease risk of homelessness in CT 2010–2013? 2013: 222,197 people in families were homeless on a single night, 36 percent of all homeless people counted.That estimate represents a 7 percent decline compared to HUD's 2012 estimate and an 11% decline compared to its 2007 estimate.Source: HUD CoC Reports
4 RRH Success Across the Country Region/ Program# of RRH Households (HH) servedTime- frame of analysisReturn to Homeless Rate (date assessed)Support Services for Veterans & Families (SSVF)13,766(2011 and 2012 SSVF RRH HHs )12 mos. after exit RRHSingles-15.7%Families-10.1%(Feb.2014)State of Michigan RRH program4,25112%(Dec. 2013)Philadelphia HPRP RRH program1,28610/09 – 5/1213.6%(Aug.2013)Utah – The Road HomeHPRP RRH program1,10013%(2013)Connecticut HPRP RRH program3,1008%(avg. over 3 yrs)(Sept. 2013)D. C. Community of Hope RRH program1178%In Connecticut, three years after re-housing, 5% return for families,
5 NAEH Evaluation of 7 CoC Programs, Average RRH cost = $4,000/family
6 Georgia Study of Reoccurrence Rates – Rigorous method to control for differences, found factors most correlated with a return to homelessnessResearch question: Which client, program, geographical characteristics exert greatest influence on the likelihood that someone returns to homelessness?Results: 9000 enrollments, 28% return to homelessness. Return Risk Factors:Was not in a Rapid Re-Housing programHad a history of homelessnessWent to a “temporary” destinationWas Non-Hispanic / Non-LatinoWas Non-WhiteHad a disabling condition at program exitProgram was in a non-rural countyWas maleWas unaccompaniedWas not with a teenage maleSource: Jason Rodriguez, GA Dept of Community AffairsKey Finding:Exits fromShelter 4.7 times;Tran. Housing 4.0 times more likely to return to homelessness than exits from Rapid Re-HousingJason assessed over 9,000 enrollments in one year of Georgia HMIS shelter data28% of enrollees re-entered a homeless program in Georgia. Regression analysisto control for background differences, found highest risk factor for return to homelessnesswas not being in RRH program
7 Research Aims for Rapid Re-Housing Can we answer the counterfactual? RESEARCH AIM for RRH Policy:Research for RRH policy goal is to estimate whether RRH is the specific element responsible for decreasing homelessness. Counterfactual:What would have happened to RRH households if there was no RRH?WHY RESEARCH DESIGN IS NECESSARY:When households who participate in RRH are different from households who do not, need to control for differences using research design. Differences in RRH and non-RRH households show up as confounders: i.e. RRH enrollment strategies differences by case manager, by program; length of RRH assistance; Housing market variabilityGold Standard = Random Control Trial = assess causal effect of RRHRESEARCH DESIGN WITHOUT RCTWith no RCT, matching methods can be used to create comparison groups that look alike, controlling for confounding differences. Propensity score matching now widely applied, probability of participation estimated using observable variables.,
8 Specific Research Questions for Philadelphia Rapid Re-Housing Study Does Rapid Re-Housing improve housing stability for formerly homeless households by decreasing the risk of a return to homelessness?Does RRH help to improve household income?Was the HPRP RRH policy effective in decreasing the risk of homelessness?
9 Dataset: All Households that entered Philadelphia shelters 10/2009-5/2012 Propensity Score Match 4716 cases discarded1,286 Non-RRH Households8 cases discarded1,169 RRH HouseholdsPropensity Score Match - to control for observable characteristics / differences between RRH and non-RRH households, propensity score matching used, where the probability of participation in RRH is estimated using risk of return to homelessness variables and individual households are matched based on their predicted propensity. Each RRH household was matched to “nearest neighbors” with same or similar probability to have received RRH intervention.
10 PSM Result–households in each group similar, standard means balanced RRH Treatment……….1169 householdsNon-RRH Control… householdsEach variable included in PSM represents HMIS data indicator correlated with risk of homelessness. (Disabling condition excluded based on high correlation with SSI-SSDI) Standard means comparison, t-tests performed on PSM matched groups, strong PSM model, households similar
11 PSM Analysis: Return to Homelessness Results Comparison Group# Households% Returned to HomelessnessRapid Re-Housing Group1,169 households13.6%Non- RRH Group1,286households39.4%Total2,455 casesResults show returns to shelter after varying lengths of time “at risk” for return. Those exiting in 2010 were measured over 2 years post-exit while those exiting in 2012 were measured some months post-exit.Odds ratio: The odds of returning to homelessness were 42% higher for households that did not receive RRH compared to households that did receive RRH
12 Washington State Evaluation – Robust matching model RRH and employment Washington State 2010 Evaluation - Rapid Re-Housing Impacts on Employment*Washington Study conclusion: RRH stops the trajectory of downward employment for homeless households*RRH clients were 1.25 times more likely to be employed, and, on average, earned $422 more annually than their counterparts who did not receive RRH.
13 RRH Promising Practice: King County RRH Pilot Goal – To move 350 homeless families in King County into rental housing by December 31, 2014Assessment: Short-term financial assistance and temporary housing-focused supports, including employment and training services,RRH funding: $3.1 million over Funders and planning partners include King County DCHS, City of Seattle Human Services Department, United Way of King County, Building Changes and the Seattle and King County Housing Authorities.RRH partnerships: Employment Navigator program. The navigators will provide critical supports to assist in gaining employment. Families may continue working with the employment navigator after rapid re-housing assistance
14 RRH Promising Practice: Massachusetts Fireman Foundation Secure Jobs Pilot Goal – Offer employment assistance to families transitioning from shelter into housing with Rapid Re-housingAssessment: Participating agencies enrolled 506 formerly homeless parents in the Secure Jobs program from a pool of 5,400 Massachusetts families receiving rental subsidiesRRH funding: Fireman Foundation awarded $1.5 million in grants to encourage housing, employment, and other agencies to work together provide comprehensive services to help low-income families regain financial independence and stay out of the shelter system.RRH partnerships: Collaboration with workforce-training organizations with employer partners. Secure Jobs participants employed by large retailers, hospitals and nursing facilities, hotels and hospitality industries, social service agencies, and manufacturing,
15 RRH Promising Practice: Tacoma Housing Authority Goal – Serve Homeless households with children. Housing Authority launching pad for family successAssessment: Tailor the availability, type, amount, and duration of assistance to the need for family housingRRH funding: Use Tacoma Housing Authority Moving to Work flexible demonstration status (HUD) for RRH assistance$80.00 for 19 families$650,000$1millionRRH partnerships: Schools and the child welfare system
16 RRH Promising Practice: Utah - The Road Home Goal – Exit family households out of shelter to stable housing as soon as possibleAssessment: Of 659 families entered Salt Lake County shelter families moved out: % of all families move out with RRH % families moved into supportive housing, % of families moved out of shelter with no financial assistanceReassessment: Progressive EngagementRRH funding: Utah uses state TANF $$ for first four months of RRH, then ESG and other RRH funding if household still needs RRHRRH partnerships: TANF, State Department of Workforce Services to increase employment income
17 Recommendations for the Hennepin County Family Shelter System 2013 Summary of Recommended PracticesCollaboration and communication are key to providing not only a positive environment for families experiencing homelessness, but also provide better outcomes for families. Streamlining the movement for a family from the point in time in which they seek out shelter to the point that they are stably housed reduces inefficiency and better serves our community. Using existing resources provides the largest area of opportunity to make immediate changes and see an immediate reduction in family shelter use.Targeting services based on individualized needs of the family is a more efficient use of resources, and provides the best outcomes for families.
18 RRH appears to effectively decrease risk of a return to homelessness RRH appears to effectively decrease risk of a return to homelessness. Why?Maybe….RRH housing case management services access landlord partnerships, find new viable housing opportunities not previously on the radar for very poor households with housing barriersMaybe….time-limited housing stabilization assistance provides a self-determination boost, motivating efforts to do “whatever it takes” to stay out of homelessnessMaybe… RRH works on the same fundamental principle as Housing First - -CLIENT CHOICE By putting housing first in the service equation, clients access all three critical aspects of self-determination: autonomy, competence, and connectedness
19 Multiple factors in every region impact RRH outcomes Variable influencing factors in every RRH region:Housing market – % affordable rentsNetwork of Landlord partnershipsCapacity to leverage TANF / HOME/ other Rental Assistance FundsESG funding levelsBelief in RRH approachCoordinated Assessment ToolsMass movement out of state or HMIS regionGrowing need for additional RRH research evidence AND additional investment in affordable housing RRH does not end poverty.
20 Local HMIS RRH Evaluation – Five Steps 1. Define Rapid Re-Housing Success in own community2. Use HMIS data indicatorsReturn to Homelessness by cohort/groupLength of stay/time homelessReduction in shelter households over timeAverage Shelter costs per dayAverage RRH assistance costs per day3. Establish comparison group using matching method4. Analyze Data – Courageously accept data shortcomings5. Add results to emerging RRH evidence
21 Strong Performance Measurement Driver Diagram Mapping out a theory of change is key to monitoring RRH performance and continuous quality improvement Three questions:1) What is the aim of your RRH intervention?What are you seeking to improve?2) What are the necessary conditions for achieving RRH aimWhat strategies will be necessary to achieve your RRH aim?How will you know you are successful with each strategy?3) What will it take to implement each primary strategy?
22 Driver Diagram for Expanded RRH Theory of Change - Change Metrics Expand RRH subsidies to 1000 households/yearby 12/2014Educate and recruit RRH providers - Increase RRH providers 25% by 10/14Increase RRH funding sources beyond ESG $$ by 9/14Educate community and stakeholders on RRH success by 6/14How to track nuanced RRH impacts?Educate stakeholders: proportion of CoC programs that understand RRH success; proportionof RRH programs that set expanded RRH goalsEnsure that RRH providers and organizations havepractice transformation support in order to makes changesneeded to achieve RRH aims
23 Driver Diagram – Housing Stability Theory of Change Support RRH households long-term housing stability goals -by 12/2015 decrease mobility500 households/yearExpand use of long-term housing subsidies and RRH bridgeLandlord / Tenancy Support NetworkHousing Tenancy Improvement Fund
24 Cloudburst Consulting Group Thank you!Jamie TaylorCloudburst Consulting GroupPhone #