Bershok Housing First Analysis

November 17, 2009

Councilman Rick Garcia Denver City Council, District 1 2727 Bryant Street, Suite 200 Denver, CO 80211 Dear Councilman Garcia, Please find attached my policy analysis summarizing available research on the program outcomes of Housing First homeless initiatives. My review of the available research indicates that Housing First programs do have documented outcomes pertaining to reduced public and emergency services usage and costs. These reductions include hospital visits, admissions, and duration of hospital stays, detox admissions, days incarcerated, and days in emergency shelters. Chronically homeless individuals also have been shown to be stably housed at a rate of 80% to 90% several years into the Housing First program versus 47% to 66% in other programs. Housing First programs also demonstrate significant cost savings in cost benefit analyses both in Denver and nationally. In Denver, public and emergency services were reduced by $31,545 per person, per year when they were involved in Housing First and had stable housing. After this cost savings was adjusted to account for housing costs of the program, the net savings were $4,745 per person, per year. Annually, this amounts to savings of $2,424,131. Policymakers have four alternative policy options available to them: (1) expanding funding for the Housing First programming in Denver, (2) expand funding for Continuum of Care programs already in place in Denver, (3) terminate all homeless program funding, or (4) leave the status quo unchanged with programming and current expense funding fixed as they are currently. Option 1 appears most likely to reduce the burden of cost on Denver public and emergency services, use tax dollars efficiently, increase the use of outpatient, preventative programs, and end chronic homelessness. Sincerely,

Lisa Bershok Lisa Bershok MSW Candidate Graduate School of Social Work University of Denver

ANALYSIS OF HOUSING FIRST POLICIES IMPACT ON HOMELESSNESS

Submitted to:

Councilman Rick Garcia Denver City Council, District 1

by:

Lisa Bershok Submitted in partial fulfillment for the requirements of Social Work 4670: Policy Analysis Development – Graduate School of Social Work University of Denver 2148 South High Street Denver, CO 80208 303-518-3886 [email protected] November 17, 2009

2

and

Executive Summary Do Housing First Policies Work and Will They Benefit Denver? The federal government and the City of Denver have created goals of eliminating chronic homelessness. The reasons that chronic homelessness is such a priority is the fact that these individuals often end up costing the state and municipalities significantly due to their utilizing so many public and emergency services. To achieve this goal, there are two major programming initiatives that are often presented to policymakers: Continuum of Care and Housing First. Continuum of Care programs require clients to address their drug misuse and mental health issues first, before they have access to housing. It expects that homeless individuals will enter and graduate from a sequence of programs (shelter, transitional housing, and permanent housing) that are aimed at helping individuals to attain self-sufficiency only after they have recovered. Housing First programming prefers a rapid and direct placement of homeless individuals into permanent housing with supportive services available, but without service or treatment utilization as a requirement to receive housing. In this type of program, housing is seen to be a component of a person’s recovery, not an end goal to be achieved at the completion of treatment. Research Outcomes The provision of housing has been shown to reduce public and emergency service costs. Housing First programs have demonstrated that more individuals are more likely to stay housed over several years time when compared to Continuum of Care programs. The average percentage of chronically homeless individuals stills housed in Housing First programs after two years ranged from 80 to 90%, depending upon the study. The range for Continuum of Care programs was 47% to 66%. Additionally, Housing First programs documented that there was a higher retention in treatment programs, even though it is not required in this program as it is in Continuum of Care programs. Cost benefit analysis put a dollar amount on the savings of Housing First programs. In Denver, public and emergency services were reduced by $31,545 per person, per year that they were involved in Housing First and had stable housing. This reduction was garnered from decreases in detox, emergency department visits, inpatient admissions, days admitted inpatient, day’s incarcerated, and emergency shelter days used. After this cost savings was adjusted to account for housing costs of the program, the net savings were $4,745 per person, per year. Annually, this amounts to savings of $2,424,131. Policy Options Policymakers have four alternative policy options available to them: (1) expanding funding for the Housing First programming in Denver, (2) expand funding for Continuum of Care programs already in place in Denver, (3) terminate all homeless program funding, or (4) leave the status quo unchanged with programming and current expense funding fixed as they are currently. Option 1 appears most likely to reduce the burden of cost on Denver public and emergency services, use tax dollars efficiently, increase the use of outpatient, preventative programs, and end chronic 3

homelessness. Table of Contents Executive Summary

3

Preface 6 I. DO HOUSING FIRST POLICIES WORK AND WILL THEY BENEFIT DENVER?

7

Definition of Homelessness

7

Why is Homelessness a Problem?

8

Causes of Homelessness

8

Demographics

9

National Statistics Denver Statistics

9 10

Homeless Programs

11

Continuum of Care Housing First

11 11

Research Outcomes

12

Cost Benefit Analysis

13

Denver Nationally

13 14

Limitations of Research

15

II. POLICY OPTIONS

15

Option 1: Option 2: Option 3: Option 4:

16 16 16 16

Expand Housing First Program Funding Expand Continuum of Care Programs Funding Terminate All Homeless Programs Funding Leave Status Quo Unchanged

4

III. EVALUTING OPTIONS

17

Options Evaluation Criteria

17

Goal 1: Reduce Burden of Cost on Denver Public and Emergency Services Goal 2: Use Tax Dollars Efficiently Goal 3: Increase Use of Preventative Programs Goal 4: End Chronic Homelessness Decision Matrix

17 17 18 18 19

IV. CONCLUSIONS

20

References

21

5

Preface This paper was prepared for partial fulfillment to meet the requirements of Social Work 4670 – Policy Analysis and Development at the Graduate School of Social Work, University of Denver. The paper reviews and summarizes the research outcomes of Housing First homelessness initiatives in comparison to Continuum of Care housing programs. This paper does not offer any new research on this subject. It only summarizes and discusses what other researchers have already discovered and concluded. The purpose of this paper is to increase awareness of the relevant research and outcomes, and to help policy makers consider alternatives and future homeless policy decisions. I wrote this paper for individuals who do not have advance training in research methods. I try to cite and interpret selected studies in plain language. I try to avoid technical and programmatic jargon. I do not discuss the methodological issues in detail. I do discuss limitations of the studies as a whole, but not on an individual study basis. Expert readers may prefer to review the cited studies and make their own methodological judgments. Opinions expressed in this paper do not necessarily represent the Graduate School of Social Work, the University of Denver, or my professor Dr. Michael Cortes. Any errors or omissions are solely my own.

6

ANALYSIS OF HOUSING FIRST POLICIES IMPACT ON HOMELESSNESS I. DO HOUSING FIRST POLICIES WORK AND WILL THEY BENEFIT DENVER? In the last decade, the federal government has begun to place a major emphasis on ending chronic homelessness (Kertesz, Crouch, Milby, Cusimano, & Schumacher, 2009). The 109th United States Congress set an official goal of ending chronic homelessness within ten years, and began developing proposals for federal interventions that can help to accomplish this goal (McCarty, 2005). Around this same time, mayors from major metropolitan cities also began devoting time and money into developing ten year goals of the own (McCarty, 2005). To date, more than 350 American communities have created plans to end chronic homelessness in their cities (Kertesz et al., 2009). Policymakers have been debating which homelessness interventions to implement in their cities. Some prefer the model of Continuum of Care, which believes that homeless individuals should enter a series of graduated housing, mental health, and substance programming until they are able to achieve independent tenancy (Kertesz et al., 2009). Others believe a newer program called Housing First, which emphasizes permanent housing be offered to homeless individuals first and supportive services second, is the right intervention to proceed forward with (Kertesz et al., 2009). This analysis will: (1) discuss underlying causes to homelessness to better analyze the costs and benefits of the proposed interventions, (2) provide further information on the differences between Continuum of Care and Housing First programs, (3) provide a summary of the research outcomes and costs benefit analyses on these two different programs, and (4) provide policy options for considering the homeless problem in Denver. Definition of Homelessness While there is no sole definition of homelessness, most federal programs use the following definition: (1) a person who lacks a fixed and night-time residence or (2) whose primary residence is a supervised public or private shelter designed to provide temporary living accommodations, (3) an institution accommodating persons intended to be institutionalized, or (4) a public or private place not designed for, or ordinarily used as, a regular sleeping accommodation for human beings (McCarty, 2005, p.2). The U.S. Department of Housing and Urban Development definition for chronically homeless individuals is that an individual must be an unaccompanied adult with a disabling condition who had been continuously homeless for a year or more, or has experienced at least four episodes of homelessness in the past 3 years (The Metropolitan Denver Homeless Initiative & The Colorado Department of Human Services, 2009).

7

Why is Homelessness a Problem? “Chronically homeless individuals are costly to the public through the high use of publicly funded health and criminal justice systems resources. Typically, interventions such as shelters, abstinence based housing, and treatment programs fail to reverse these patterns for this population” (Larimer et al., 2009, p.1349). High public system costs that are incurred by chronic homeless populations have been the trigger for nationwide efforts on a federal, state, and local level to eliminate homelessness (Larimer et al., 2009). Because there are more barriers for homeless individuals to obtain healthcare, they often use only acute care services that carry higher rates (Larimer et al., 2009). Other studies have shown that homeless individuals with concurrent mental health and substance abuse disorders have much higher frequency rates of emergency department and inpatient hospitalization services (Martinez & Burt, 2006). These costs disproportionately are placed on municipal and state governments to fund (Martinez & Burt, 2006). Homeless individuals with alcohol problems use more publicly funded healthcare, as well as criminal justice systems resources (Larimer et al., 2009). Communities incur steep costs when severely debilitated homeless persons cycle repeatedly through treatment programs, hospitals, and jails, often in quick succession (Kertesz & Weiner, 2009, p. 1822). Rates of chronic medical illness are high among homeless adults (Sadowski, Kee, VanderWeele, & Buchanan, 2009). Expect for obesity, stroke, and cancer, homeless adults are more likely to have chronic medical illnesses like HIV, hypertension, and diabetes. They often will experience more complications that result in more emergency department and hospital admissions because they lack adequate housing and don’t have access to regular, uninterrupted treatment (Sadowski et al., 2009). Mortality rates are three or more times higher for homeless adults than those in the general population (Larimer et al., 2009). For the chronic homeless population that suffers from alcohol problems, the mortality rates are similar to those of developing countries with average age of death between 42 and 52 years of age and 30% to 70% of deaths related to alcohol (Larimer et al., 2009). Causes of Homelessness According to the U.S. Conference of Mayors study, lack of affordable housing was the number one cause of homelessness (McCarty, 2005). In Denver, the number one cause of homelessness is the loss of job, with 34.7% of individuals listing this as the primary reason for their homelessness (The Metropolitan Denver Homeless Initiative & The Colorado Department of Human Services, 2009). Housing costs and the inability to pay rent is the second highest reason for homelessness at 31.2% (The Metropolitan Denver Homeless Initiative & The Colorado Department of Human Services, 2009). Other reasons documented include: alcohol or substance abuse 29.5%, 8

family problems 19.1%, illness 10.5%, being asked to leave residence 12.2%, domestic violence 9.9%, and being released from jail or other legal problems 16%. The data on Denver’s homeless population is similar to national data that has been collected. Nationally, other reasons for homelessness, in order of frequency, include mental illness and lack of needed services, substance abuse and lack of needed services, low-paying jobs, unemployment, domestic violence, poverty, and prison release (McCarty, 2005). Demographics National Statistics Because of the transient nature of homelessness individuals, it can be difficult to track demographic data. The current existing data estimates that the United State’s homeless population ranges from 600,000 to 2.5 million individuals (McCarty, 2005). It is estimated that approximately 18% of this number meets the definition for chronic homelessness (Kertesz & Weiner, 2009). The 2004 U.S. Conference of Mayors Study was conducted in an attempt to obtain more extensive data on the homeless population, and included 27 major metropolitan areas, including Denver (McCarty, 2005). The following data was obtained from that study.  Requests for emergency shelter increased by an average of 6%, with 70% of the cities involved in the study showing an increase.  Homeless families request for shelter increased 7%, with 78% of cities documenting this increase.  The average length of time individuals remained homeless was for eight months, which is an increase of five months from 2003 data.  Demographics comprising homeless population include: 

Single men, 41%; families with children, 40%; Single women, 14%; unaccompanied youth, 5%.



Homeless populations broken down by race include: African-American, 49%; white, 35%; Hispanic, 13%; Native American 2%; Asian 1%.



Also reported was that 23% of homeless individuals were considered to have mental illness, 30% had substance abuse issues, 17% were employed, and 10% were veterans.

 Access to housing assistance was reported by cities to have long waits, averaging 20 months for public housing and 35 months for Section 8 vouchers. 

More than 50% of cities surveyed had stopped accepting applications for at least one local housing program because of the length of waiting lists. 9

Denver Statistics A current homeless point in time study conducted on January 27, 2009 in a collaborative effort by The Metropolitan Denver Homeless Initiative and The Colorado Department of Human Services found the following data. A total of 11,061 persons in the seven county Denver metropolitan area were counted as homeless.  Single individuals accounted for 53.5% of this total, and persons in families (those with children and couple without children) accounted for 46.5% of this total.  Children and teens under the age of 18 comprised 26.8% of the total individuals counted.  Of family households counted, single parent households comprised 75% of the households surveyed.  58.2% of the individuals surveyed were male, and 41.8% were female.  Black/African Americans and American Indian/Alaska Natives are disproportionately represented within the homeless community in the Denver metropolitan area 

U.S. Census Bureau reports that African American individuals comprise just over 11% of the population in the Denver Metropolitan area but they represent 27% of the homeless population. American Indians only comprise 1.3% of the population but represent 8.5% of the homeless population



The homeless Hispanic population for the Denver metropolitan area, 30.7%, compares similarly to the U.S. Census Bureau’s reported 31.7% Hispanic or Latino population currently residing in the Denver metropolitan area.

 44.5% of homeless individuals surveyed were reporting homelessness for the first time; many citing inability to pay rent and losing a job as the major contributing factors  The total count of persons chronically homeless was 497. The chronic homeless population comprised 4.5% of the total homeless population and 8.4% of the single individual population  Males comprised 75.6% of the chronically homeless population  Veterans comprise 13% of the adult homeless population. Ages ranged from 20 to 84 with the largest percentage between the ages of 45 and 54. The average age of a homeless veteran was 50. 10

 Denver County comprised 60.2% of the homeless population. Other counties are listed as follows for percentage of homeless individuals: Jefferson 11.2%; Adams 10.9%; Boulder 9.5%; Arapahoe 6.1%; Broomfield 1.4%; and Douglas 0.2%.

Homeless Programs Continuum of Care Continuum of Care programs were created and favored by federal funding after the McKinney Homeless Assistance Act of 1987 (Kertesz et al., 2009). This type of programming expects that homeless individuals will enter and graduate from a sequence of programs (shelter, transitional housing, permanent housing) that is aimed at helping individuals to attain selfsufficiency only after they have recovered (Kertesz et al., 2009). Continuum of Care homeless programs require clients to address their drug misuse and mental health issues first, before having access to housing (Atherton & Nicholls, 2008). This process requires individuals to progress up a “staircase of transition” in which each step is the successful resolution of a problem like drug dependence or mental health stability (Sahlin, 2005 as cited in Atherton & Nicholls, 2008, p. 290). Failure to complete steps or sliding down the staircase often means independent house becomes more of a distant prospect to attain (Atherton & Nicholls, 2008). The U.S. Department of Housing and Urban Development found no reduction in homelessness from the mid-1990’s to 2000’s when this type of programming was the main intervention being used (Kertesz et al., 2009). Housing First Housing First programs are considered to be a new homelessness intervention even though Pathways to Housing, the first Housing First program in the nation, was started in 1992 (Pearson, Montgomery, & Locke, 2009). The term Housing First began to be used in 1999 by the National Alliance to end homelessness (Kertesz et al., 2009). Housing First programs address the complexities of the association between homelessness, mental illness, and addictions (Atherton & Nicholls, 2008). Housing First refers to rapid and direct placement of homeless individuals into permanent housing with supportive services available, but without service or treatment utilization as a requirement to receiving housing (Kertesz, 2009). This type of program assumes that people with mental health or substance misuse are capable of dealing with tenancy (Atherton & Nicholls, 2008). Normally, a considerable amount of services and support are offered to clients, and while they do not have to accept it, the services are “assertively provided” (Atherton & Nicholls, 2008, p. 291). When involved in this program, clients are usually periodically visited by case managers (Kertesz et al., 2009). Unlike Continuum of Care programs, successful completion of services or programs are not necessary to obtain or keep housing (Atherton & Nicholls, 2008). In these types of programs, housing is seen to be a component of a person’s recovery, not an end goal to be achieved at the completion of treatment (Atherton & Nicholls, 2008). Housing First programs consider housing needs paramount and a separate issue from treatment (Pearson, Montgomery, and Locke, 2009). 11

Housing First programs break from more traditional models that focus on “fixing” client to make them “housing ready” (Kertesz et al., 2009, p. 497). Housing First is promoted by the U.S. Interagency Council on Homelessness, and is considered to be the only homelessness intervention that is “evidence-based practice” by the U.S. Conference of Mayors in 2008 (Kertesz et al., 2009, p. 497). Individuals accepted into most programs are individuals with severe mental health and substance abuse issues; those individuals who are most vulnerable out of the homeless population and the least likely to success in group rehabilitation programs (Kertesz et al., 2009). The provision of housing has been shown to reduce hospital visits, admissions, and duration of hospital stays (Larimer et al., 2009). Overall, most research has shown that overall public system spending is reduced by nearly as much as is spent on housing (Larimer et al., 2009). Research Outcomes Comparison of Rates of Success for Housing First versus Continuum of Care Data from multiple studies on Housing First Programs suggests that individuals who have experienced chronic homelessness due multiple problems, such as mental illness and substance abuse can maintain stable tenancies even if their other problems are not resolved. It has been demonstrated that willingness to participate in and complete drug abuse treatment is correlated with access to housing (Gulcur, Stefancic, Shinn, Tsemberis, & Fischer, 2003). More traditional programs, like Continuum of Care, are thought to have worked well for individuals peripherally or recently homeless, but often have failed repeatedly for those who are chronically homeless (Kertesz & Weiner, 2009). The following are data and results from multiple studies that have researched and compared these two homeless programs.  A study conducted in 2005 found that homeless individuals with serious mental illness who had stable housing were more likely to stay enrolled in mental health treatment (Pearson, Montgomery, & Locke, 2009).  Housing First programs that incorporate case management services have greater reductions in hospital admissions and substance use, and higher retention in treatment, even though these services are not required of the individual (Gulcur et al., 2003).  In a two year study that compared Housing First to Continuum of Care programs, individuals in the Housing First group spent fewer days hospitalized and spent more time stably housed. This study also found that homelessness was virtually eliminated by the second half of the first year and remained at very low levels through the end of the two year study (Gulcur, et al., 2003). .  A study that examined Housing First outcomes with chronically ill homeless individuals found that there was a 29% reduction in the number of days spent in the hospital and a 24% reduction in emergency room visits. Therefore, for every 100 homeless adults there would

12

be 49 fewer hospitalizations, 270 fewer hospital days, and 116 fewer emergency department visits (Sadowski et al., 2009).  Housing outcomes for Continuum of Care programs have only been moderate. One study found that only 47% of individuals remained housed five years later compared with a Housing First program in which 88% of individuals remained housed (Kertesz et al., 2009).  Another study in Chicago found that only 61% of Continuum of Care individuals were housed after three years, and of those who were literally homeless at the beginning of the study, only 50% were still housed (Kertesz et al., 2009).  Tsemberis (2004), founder of the original Housing First program Pathways to Housing, found a greater time of housing stability in Housing First compared with Continuum of Care clients (Atherton & Nicholls, 2008).  In a randomized controlled trial comparing a Housing First program (Pathways to Housing in NY) to a control group that received Continuum of Care programming, the following data was found (Kertesz et al., 2009) 

80% to 90% of individuals in the Housing First program remained in housing versus the 40% of individuals in the Continuum of Care programming.



Both the Housing First participants and the Continuum of Care participants were found to be the same in regard to substance misuse

 In Phoenix, a Housing First program called Washington House tracked 99 homeless individuals who had received housing before graduating out of the program in the prior 12 months before the study was conducted. They found that 91 individuals out of this group were still successfully living independently a year later. From 2003 to 2006, 365 clients had graduated from the Washington House program and 90% had continued to live independently versus returning to homelessness (Legander, 2006). Cost Benefit Analysis Housing provision through Housing First programs has been shown to reduce hospital visits, admissions, and duration of hospital stays among the homeless population (Larimer et al., 2009). Public system spending has also been shown to be reduces by as much that is spent on housing (Larimer et al., 2009). For the very sickest homeless individuals, who have repeated hospitalizations or incarcerations, housing and supportive services may in fact generate an overall reduction in costs to society” (Kertesz et al., 2009). However, it has also been found that for homeless individuals who incur few public service costs on the street, a cost comparison could be less favorable to Housing First programs” (Kertesz et al., 2009). Denver 13

A cost benefit analysis and program outcomes report on Denver’s Housing First Collaborative was completed in 2006 (Perlman & Parvensky). It contrasted costs accumulated by a sample group of 19 chronically homeless individuals 24 months prior to entering the Housing First program, and 24 months after entering the program. The study documented the following findings.  An overall reduction of 72.95% in emergency services related costs amounting to almost $600,000 in the 24 months of participation in the housing program. This is a total emergency costs savings average of $31,545 per participant.  Emergency room visits were reduced by an average of 34.3%, inpatient visits by 40% and inpatient nights by 80%. Overall inpatient costs reduced by 66%.  Detox visits were dramatically reduced by 82%. This is an average cost savings of $8,732 per person.  Incarceration days and costs were reduced by 76% and emergency shelter costs were reduced by $13,600.  While emergency room care, inpatient medical and psychiatric care, detox services, incarceration, and emergency shelter were significantly reduced by participation in the program, outpatient health costs did increase. This was due to the participants being directed to these services as there are a more costs efficient and treatment effective.  In 2006 there were projected to be 513 chronically homeless individuals in Denver who would be eligible for the Housing First program versus the 497 counted by the 2009 point in time homeless study. The following numbers are based on 513 individuals. 

If all 513 chronically homeless individuals were involved in the Housing First program, the savings to Denver would amount to be $16.1 million.



After the investment costs of providing the comprehensive supportive housing and services through the Housing First Program are factored in, there would be a net savings of $4,745 per person. The projected net savings for all 513 chronically homeless persons, if they were enrolled in the Housing First Program, would be $2,424,131.

Nationally Nationally, cost benefit analyses have found similar results as the Denver cost benefit analysis. Below are the results of the other studies that were conducted.  A Housing First analysis in New York demonstrated similar reductions in emergency service usage and arrests. Prior to a housing intervention, homeless individuals with mental illness used $40,451 of services a year. This cost was reduced by $16,281 when provided 14

with housing. These figures included the cost of providing housing (Atherton & Nicholls, 2008).  A straight cost comparison completed in 2007 in New York City showed annual costs of supportive housing being $20,410 and the costs of 365 days of shelter occupancy costing between $24,269 and $43,530. However, these figures may only be relevant to communities that attempt to ensure shelter access every day of the year as New York City does (Kertesz et al., 2009).  One study found yearly costs for a Continuum of Care programs in 15 jurisdictions to be $14,000 for emergency shelter programs and $13,100 for transitional housing programs. Permanent supportive housing was found to only be $11,580 (Kertesz et al, 2009).  A research study conducted over two years in Seattle found the following findings when comparing individuals in a Housing First program to a control group who received Continuum of Care programming (Larimer et al., 2009): 

After 12 months, the 95 housed individuals had reduced their total public costs by $4 million from the prior year.



This amounts to $42,964 per year per person



Total cost to administer housing was $13,440 per person



Length of time in housing was significantly related to reductions in the cost of services.



This study also found that individuals in the housed group were able to reduce their alcohol use. Limitations of Research

There are several limitations to the research and data collected recently on the Housing First programs that are noteworthy to include. Most of this research has only been conducted for one to two years before publishing the data, which is a short time given that long term problems are being analyzed (Atherton & Nicholls, 2008). Many of the studies include smaller sample sizes with many groups below 100 individuals. It is of note that the cost benefit analysis for Denver only contained a sample size of 19 individuals. The programs studied have for the most part only been aimed at chronically homeless individuals, so assumptions have to be made about the impact of cost reductions on public services for those individuals who are not chronically homeless (Atherton & Nicholls, 2008). This is especially pertinent as the fasted growing segment of the homeless population is single parent families (Kertesz & Weiner, 2009). Also, Housing First programs do not resolve any structural issues that lend itself to homelessness, such as poverty, foreclosure crises, mental illness service provision and funding, and substance abuse issues (Kertesz et al., 2009). 15

II. POLICY OPTIONS Policy makers in Denver have several policy options they can consider after examining the above information about Housing First and Continuum of Care homeless programs. These include (1) expanding funding for the Housing First programming in Denver, (2) expand funding for Continuum of Care programs already in place in Denver, (3) terminate all homeless program funding, or (4) leave the status quo unchanged with programming and current expense funding fixed as they are currently. In the following discussion, these four options will be treated as mutually exclusive options. However, what must be considered when evaluating all four options is the federal government’s documented emphasis on ending chronic homelessness, and the best policy that Denver can adapt to fulfill this initiative. Option 1: Expand Housing First Program Funding This option is to expand Housing First program funding, as it has already been implemented in Denver to a certain degree. However, this option would include continuing to create housing stock for chronically homeless individuals. Right now, Denver’s Road Home organization states that 60% of the housing stock has been created, leaving 40% of capital expenses to build, remodel, purchase, or lease additional housing space (www.denversroadhome.org). It would also require continuing to build staff or re-train staff from other homeless programs to efficiently work in this system. As the prior research demonstrates that Housing First programs work best when case management is offered in conjunction with housing, training a quality staff will be of importance. This option includes the greatest net benefits to society in that it houses all chronically homeless individuals with lower public service costs in criminal justice and emergency services fields. Housing First research data demonstrates that it is more effective than Continuum of Care programs in keeping people stably housed over a several year period of time, thus coming closer to eliminating chronic homelessness. Option 2: Expand Continuum of Care Program Funding Option 2 would also require additional expenses for the homeless programs that are already in place in Denver. They would require additional budgetary funds to expand staff, programming, outreach, and housing. Since it is assumed by the Continuum of Care programming theory that after advancing through different treatment and housing programs an individual would be able to achieve stable tenancy, Denver must be prepared to increase beds in these programs and have enough long term housing available to still get approximately 500 individuals off the streets and into service provision. Option 3: Terminate all Homeless Programs Funding All homeless programming funded by the City and County of Denver would be terminated. This would have a significant financial impact on those organizations and programs that currently support individuals who are homeless and receive some sort of funding from the city or county. It would not facilitate either the federal or the City of Denver’s goals of ending homelessness. It 16

could potentially jeopardize federal funding that the City or County of Denver receives in the future if the objective of ending homelessness fails here. The City and County of Denver could expect that costs on public and emergency services would rise substantially. Option 4: Leave Status Quo Unchanged All funding for current homeless programs would continue at their current levels. All perceived benefits and costs would remain fixed. It is also assumed that Denver County would continue to have 60% of the 11,000 homeless individuals in the Denver metro area living on its streets. III. EVALUATION OF OPTIONS Design and evaluation of these policy options are beyond the scope of this paper. The following discussion utilizes a general approach to evaluating the four prior policy options summarized above. A thorough evaluation of options was not performed for this paper, and instead a conceptual discussion is utilized instead that hopefully models the results of what a rigorous analysis would hopefully show. As each of the four policy options has strengths and weaknesses, this paper assumes that policymakers want to compare their options with a number of goals in mind. Assumed goals are to (1) reduce the burden of cost on Denver public and emergency services, (2) use tax dollars efficiently, (3) increase use of preventative services, and (4) end homelessness. Options Evaluation Criteria Goal 1: Reduce Burden of Cost on Denver Public and Emergency Services Costs have been documented to be extremely high for public and emergency service provision. Finding ways to reduce these costs will free up dollars budgeted for these types of programs that could be moved to funding homeless programs, outpatient services, healthcare, or other needed programs in the City and County of Denver. Each of the following criteria can be developed into measures for future use by evaluation researchers to conclude whether this goal is being achieved. 1.a. 1.b. 1.c. 1.d. 1.e. 1.f.

Reduce use of emergency departments Reduce number of inpatient admissions Reduce the number of days spent in inpatient settings Reduce number of incarcerations and days incarcerated Reduce detox admissions Reduce shelter admissions and length of stays

Goal 2: Use Tax Dollars Efficiently The following criteria assume that placing chronically homeless individuals in Housing First programs versus Continuum of Care programs will be a better investment of tax dollars. 17

2.a. Increase time spent in stable housing 2.b. Increase number of chronically homeless individuals who can maintain tenancy in stable housing 2.c. Increase number of individuals who can participate in outpatient, preventative programs versus public and emergency services 2.d. Fulfill criteria from Goal 1 Goal 3: Increase Use of Preventative Programs While costs and funding needed for outpatient, preventative programming is documented to rise when Housing First programs are successfully implemented; these costs are minimal in comparison to public and emergency service costs. Emergency and public service treatment programs have also been shown to make no impact on reversing patterns amongst the chronically homeless. Having access to preventative care for physical health, mental health, and substance abuse will help decrease higher systems costs, and hopefully increase a person’s life span, happiness, and involvement in their community. Employment should also be easier to obtain and maintain when these aspects of an individual’s health are maintained. 3.a. Increased numbers of persons seeking outpatient, preventative programs 3.b. Increase in number of person who are able to show reduction in substance misuse and an increase in maintaining physical and mental health 3.c. Increase in number of individuals who are able to “graduate” from outpatient service programs 3.d. Increase in individuals able to seek and maintain employment Goal 4: End Chronic Homelessness Ending homelessness is a federal goal adopted by the City of Denver. This goal will require ingenuity, as well as outreach and embracing this policy change as a public goodwill goal. 4.a. Increase awareness of chronic homelessness issues and impact in the Denver metro community 4.b. Increase public support for use of funds to be deviated from public and emergency services, which are substantially higher, to homeless initiatives and outpatient, preventative services 4.c. Increased outreach to chronic homeless population in Denver to educate and inform individuals about programming and housing available to them 4.d. Increase training of staff in programs that serve the chronically homeless population so they can provide the support and services that these individuals need 4.e. Increase awareness that ending chronic homelessness is a long term plan that must maintain public and budgetary commitment

18

DECISIO N MATRIX Reduce Public & Emergency Services Costs

Expand Housing First Funding

Reductions estimated at: Total = $31,545/person Factoring in Housing Costs= $4,745/person Yearly Net Savings= $2,424,131

Expand Continuum of Care Programs Funding

Terminate All Homeless Programs Funding

Status Quo

Would continue to maintain current costs of $43,239 per person per year

Would predict costs to increase significantly

Would continue to maintain current costs of $43,239 per person per year Could potentially raise as costs for these services continue to rise

Could potentially raise as costs for these services continue to rise

Use Tax Dollars Efficiently

Would shift what programs tax $ spent in away from higher cost emergency services

Research shows less housing stability for $’s spent

More $ would be spend on public and emergency services at a higher cost

Would stay the same as costs currently

Increase Use of Preventative Programs

Increase of approximately $900$1,000 per person/per year

Would continue to stay relatively the same per $ person, but potentially more people added to programs increasing costs

Funding not available to increase outpatient/ preventative programmin g

Would not expect to see growth in preventative programs

19

End Chronic Homelessness

Estimates range from 80-90% of individuals staying in stable housing programs

No reduction predicted on past data

Not supportive of this goal

Would continue to see rates of chronic homelessness stay the same

IV. CONCLUSIONS “The overall quality of life for the community can be significantly improved as the negative impacts of individuals living and sleeping on the streets are reduced” (Perlman & Parvensky, 2006, p. 11). Available research that documents program outcomes for housing chronically homeless individuals shows strong support for the conclusion that Housing First programs provide more public benefits. This type of housing program significantly reduces public spending on emergency services and incarceration; more chronically homeless individuals remain stably housed for longer periods of time; and individuals were shown to utilize outpatient, preventative programs that are more cost effective on a more regular basis. Housing First research seems to confirm the theory that an individual does not need to address mental health and substance misuse problems to remain stably housed, and that permanent housing may in fact have a high correlation with entering and remaining in these types of service programs. The economic argument for Housing First programs is most powerful when individuals enrolled in the program are frequent users of health and judicial services. When other subgroups of the homeless population are enrolled, cost savings become more uncertain (Kertesz & Weiner, 2009). This means that Housing First is most effective for individuals who are chronically homeless, which leaves a majority of the homeless population out of this type of programming. It also appears that the longer an individual stays in stable housing, the better the benefits are to society with public and emergency services continuing to decrease. While Housing First programs have not been linked to increased substance use, even though there are few restrictions on use or requirements for treatment, neither has this intervention show to reduce substance use (Kertesz et al., 2009; Kertesz & Weiner, 2009). Policy responses should be framed not simply as “what works?” but as “what works for whom?” (Kertesz et al., 2009, p.498). Denver has already started to implement a Housing First initiative and has Continuum of Care programs in place as well. Additional programming for the non-chronically homeless should continue to be researched and best practices implemented. However, significant costs are affixed to Denver ignoring the needs of their chronically homeless population. While there may be some additional capital costs in locating housing, with net benefits of over $2 million a year, Denver cannot not afford to house its most frequent users of costly public services. 20

References Atherton, I. and Nicholls, C. M. (2008). ‘Housing First’ as a means of addressing multiple needs and homelessness. European Journal of Homelessness, 2, 289-303. Denver’s Road Home (n.d.). Current State. In Learn about DRH. Retrieved November 5, 2009, from http://www.denversroadhome.org/state.php?id_cat=1. Gulcur, L., Stefancic, A., Shinn, M., Tsemberis, S., and Fischer, S. N. (2003). Housing, hospitalization, and cost outcomes for homeless individuals with psychiatric disabilities participating in continuum of care and housing first programs. Journal of Community & Applied Social Psychology, 13, 171-186. Kertesz, S. G. and Weiner, S. J. (2009). Housing the chronically homeless: High hopes, and complex realities. Journal of American Medical Association, 301(17), 1822-1824. Kertesz, S. G., Crouch, K., Milby, J. B., Cusimano, R. E., and Schumacher, J. E. (2009). Housing first for homeless persons with active addiction: Are we overreaching? The Milbank Quarterly, 87(2), 495-534. Larimer, M. E., Malone, D. K., Garner, M. D., Atkins, D. C., Burlingham, B., Lonczak, H. A., Tanzer, K., Ginzler, J., Clifasefi, S. L., Hobson, W. G., and Marlatt, G. A. (2009). Health care and public service use and costs before and after provision of housing for chronically homeless persons with severe alcohol problems. Journal of American Medical Association, 301(13), 1349-1357. Leff, H. S., Chow, C. M., Pepin, R., Conley, J., Allen, I. E., and Seaman, C. A. (2009). Does one size fit all? What we can and can’t learn from a meta-analysis of housing models for persons with mental illness. Psychiatric Services, 60(4), 473-484. Legander, S. (2006). Housing first: A program to help people move off the streets and into treatment. Behavioral Net, May. Retrieved October 20, 2009, from http://www.behavioral.net. Martinez, T. E. and Burt, M. R. (2006). Impact of permanent supportive housing on the use of acute care health services by homeless adults. Psychiatric Services, 57(7), 992-999. 21

McCarty, M. (2005). Homelessness: Recent statistics, targeted federal programs, and recent legislation. Washington, D.C.: Congressional Research Service-The Library of Congress. Padgett, D. K., Gulcur, L., and Tsemberis, S. (2006). Housing first services for people who are homeless with co-occurring serious mental illness and substance abuse. Research on Social Work Practice, 16(1), 74-83. Pearson, C., Montgomery, A. E., and Locke, G. (2009). Housing stability among homeless individuals with serious mental illness participating in housing first programs. Journal of Community Psychology, 37(3), 404-417. Perlman. J. and Parvensky, J. (2006). Cost benefit analysis and program outcomes report. Denver: Colorado Coalition for the Homeless. Sadowski, L. S., Kee R. A., and VanderWeele, T. J. (2009). Effect of housing and case management program on emergency department visits and hospitalizations among chronically ill homeless adults: A randomized trial. Journal of American Medical vAssociation, 301(17), 1771-1778. The Metropolitan Denver Homeless Initiative and The Colorado Department of Human Services (2009). Denver metropolitan area 2009 homeless point-in-time key findings. Retrieved October 27, 2009, from http://www.denversroadhome.org/files/2009%20 Point%20in%20Time%20Key%20findings%209-14-09.pdf.

22