The 2007-2008 Civil Grand Jury for the City and County of San Francisco the Homeless Have Homes, but They Are Still on*

lowest of its two benefit levels) declined from 2,632 to 642. It has since hovered at about 600. From 2004 through December 2007, 4,317 formerly homeless single adults have been housed. Since Care Not Cash went into effect, supportive housing units have increased from about 2,000 to about 4,500. Of the current total, 880 are sponsored by the Health Department under a "Direct Access to Housing" program that provides more intensive medical services to the elderly, disabled and the mentally or physically ill. The rest are sponsored by the Human Services Agency. Projects that are currently planned or under 3 "Statistics show that the care of one chronically homeless person using Emergency Room services, and/or incarceration, cost San Francisco an average of $61,000 each year. On the other hand, permanent supportive housing, including treatment and care, would cost $16,000 a year." Ten-Year Plan, p. 8 development will add about 450 units by early 2010. IV. DISCUSSION Α. HOMELESSNESS SPENDING – A BIG BUSINESS In many ways San Francisco has a Homeless Industry. Homeless spending is big business. For its current fiscal year, the City budgets about $186 million for direct spending on the homeless or those at risk of becoming homeless. The Jury has not attempted to quantify the generous contributions from individuals, charities, and foundations. This figure does not include the money represented by CAAP welfare grants, emergency medical response, hospitalization, jail costs, most city management and overhead functions, and much else. Although estimates of homeless spending differ widely and include different costs, it is clear that homeless spending has mushroomed since the City began to track it. The Budget Analyst's previous reports estimated direct spending on the homeless to be $31.1 million in fiscal year 1993-1994, and $73 million in fiscal year 2001-2002 (with administrative overhead and capital costs raising the latter total to $104.3 million). Thus, the City's budget for direct spending on the homeless or those at risk of becoming homeless increased nearly 600% since fiscal year 1993-1994. í z^{\frac{1}{2}} The 2001-2002 Budget Analyst Report is now six years old, and predates the Ten-Year Plan and Care Not Cash. Updating the report would provide worthwhile data to aid in . . assessing whether the trend in ever-increasing spending on homelessness is changing. --- While more limited audits have been done, such as the Controller's 2008 audit of Care Not Cash, there has not been a comprehensive assessment since the early part of the decade. Money spent on homeless services comes from a variety of federal, state, and local sources. Federal funds include: $18.8 million of McKinney-Vento Homeless Assistance Act HUD funding for building permanent supportive housing; $8.6 million under "Shelter Plus Care," a federal supportive housing program for people with disabilities; $8.1 million in Housing Opportunities for People with AIDS Act ("HOPWA") funding for those with HIV; Veterans Administration funds; SSI payments to individuals (or to third party agents on their behalf), and Medicare. State funds include Medi-Cal payments, $1.8 million in "Prop 63 Funds" for the mentally ill, proceeds of a 2002 bond measure for affordable housing, and even a pinch of tobacco tax revenue. Local funds include general revenues, proceeds of development bonds, developer fees, hotel tax revenue, property "tax increment" (increases in property taxes in redevelopment areas), and a few private grants. Much of this funding is uncertain and highly variable in amount. For example, recent Affordable and Homeless Housing Bond ballot measures have passed or failed by narrow voter margins. The sheer number of funding sources and agencies involved, and the need for constant coordination and cooperation among agencies, create unique challenges for budgeting and planning. The City has been adept at dealing with these challenges so far. For example, it has made up for cutbacks in HOPWA and other AIDS funding by tapping general funds and by adding AIDS homeless units in mixed buildings, or by developing self-sustaining businesses as part of its housing program. Some uses or combinations of funding sources are proscribed; for example, funds may be available for capital costs of constructing a building, but not for subsequent operating expenses. The City has recently created a local operating subsidy fund to pay ongoing operating costs of supporting housing, so that in a project like the soon-to-open Salvation Army community center (where private philanthropy funded the construction), the City can subsidize operating costs and avoid the gradual erosion of services that occurs when new supportive housing is created c, without a stable long-term source of operating funds. В. ROLE OF NON-PROFITS - WHERE MOST OF THE MONEY GOES Most of the City's spending on supportive housing (at least $90 million) goes to over 80 private, mostly non-profit organizations. They are community based, usually formed around a particular neighborhood, demographic group, or issue. Some are large and 1 highly sophisticated, others small and dependent on collaborations with others to run their programs. Most depend largely or entirely on City contracts for their funding. . Perhaps because of this dependence, politics infuse their relationship with the City. On the one hand, City officials occasionally said things like: "Non-profits spend too much time on advocacy and not enough time running their programs," or "Some non-profits have a strong sense of entitlement," or that they would anticipate "political fallout" from any attempt to terminate a contractor. On the other hand, one non-profit manager feared (rightly or wrongly) that a visit he got from a Controller's Office auditor was the result of his having publicly expressed views on a policy issue that differed from the Mayor's. He spent months unnecessarily gathering data to defend his program. Objective cost/benefit : 5 measures of success could help ease these tensions. On the whole, however, the Jury observed that relationships seem harmonious. The City recognizes the value of working through organizations with a strong cultural and community base, and the non-profits welcome opportunities to collaborate with City staff, who they find experienced, supportive, and willing to work closely with them. į To the extent that some non-profits may, in fact, feel a sense of entitlement, the Jury believes this can be remedied to a great extent by strengthening performance measures as recommended later in this report. C. ECONOMIC MODELS FOR SUPPORTIVE HOUSING: MASTER LEASE AND DEVELOPMENT i Supportive housing in San Francisco has been created in two main ways -- a master lease program whereby existing buildings are leased from private owners and re-let to the homeless, and a development program whereby specially-built housing is owned and developed by non-profits using private contractors and city-backed financing. The master lease approach typically involves the City's funding master leases between non-profits and residential hotels or SRO's (i.e., buildings for "Single Room Occupancy," once called "flophouses"), usually with small rooms and shared bathrooms. These can be quickly renovated and converted to homeless use, but the housing quality is uneven. These master leases are usually for 10-year terms with a 10-year renewal option. This approach has been criticized for helping to shrink the pool of private SRO's and forcing up rents on the lowest rung of the private housing market ladder. The development approach often costs more and typically takes 1 to 5 years. But the newly built housing has amenities like kitchens, baths, and common rooms that greatly enhance the quality of life. The development corporation owns the land, develops it through the use of city-backed financing, and operates the project thereafter. It gets a development fee (during construction), management fees (once opened), and has right of first refusal to purchase the building though with many covenants designed to assure the ÷, building remains dedicated to housing purposes. One major such developer is the Tenderloin Neighborhood Development Corporation ŀ which operates at least 12 buildings, mostly in the Tenderloin. Other such buildings are currently being developed in the Mission and South of Market. The typical development model uses property purchased by the non-profit with ŀ improvements being constructed with the aid of loans from private capital sources such as banks, with additional loans and/or loan guarantees being made by the City directly or through the San Francisco Redevelopment Agency. Private financing at attractive rates is available to these projects because of low-income housing or other federal tax credits. A typical arrangement uses cooperating corporations and limited and general partnerships arrayed, under City direction, to finance and operate the project. The loans are theoretically repayable in 55 years, but are typically forgiven sooner, when tax benefits have been exhausted. Although the sums involved are large and the financial arrangements complex, the Jury found no evidence that excessive profits are being made. Nor is there any evidence of profiteering in non-profit owned development, where fees and loan terms are standard, and non-profit lenders must often partner with commercial banks to make projects viable. City officials involved in master lease negotiations indicate that the owners of residential 1 hotels that enter into master leases are able to make more money renting their rooms to the City at a lower rate based on 100% occupancy than they could renting them out directly at a 30-40% occupancy rate. Although many of these deals were negotiated at a time when the City was scrambling to ramp up supportive housing as fast as possible and saw master leasing as the best short-term approach, the hotel owners did not have the upper hand. The transactions occurred against a backdrop of increasing regulation of residential hotels in the 1990's that spawned some major litigation and left the hotel owners little choice, since they were precluded from going out of business or converting residential hotel rooms to other uses. There is a wide range of housing provided to the homeless. The two most common types, however, are master leased hotels and new or reconstructed purpose-built facilities. The differences between these two dominant approaches are described elsewhere in this . . report. Even though the two approaches to developing a housing unit, the time-to-market and the gross development costs are very different, many housing advocates assert that 1. unit costs for development and operations, amortized over time, are comparable. Since new construction costs have increased significantly in recent years, and are likely to continue increasing, it seems unlikely that this real or perceived comparability will continue. Little or no analysis of these competing approaches has been done. A comprehensive, cost benefit analysis is in order to guide future decisions as to whether one or the other of the two alternatives makes the most sense as a model for additional housing development. D. COSTS OF SUPPORTIVE HOUSING – AN INCENTIVE TO MIGRATE? 1 Supportive housing is not cheap. The recent Controller's Office: Performance Audit of ř Care Not Cash (April 30, 2008) concluded that the program was meeting its goals, but ì noted that the housing and supportive services received were worth about $1300 per month, much more than the former maximum cash payment of $410 per month. The Controller said that it was likely that Care Not Cash had reduced but not eliminated the incentive for the homeless to flock to the City. Migration is easy because all that the City requires in order for someone to qualify for general assistance is that the person spend 15 days in the City and declare an intent to reside here permanently. High-quality homeless programs might thus exert their own magnet effect. The supportive housing services administered by the Human Services Agency for the Care Not Cash population of single homeless adults are typically limited to a 24-hour security desk and minimal casework help. Supportive housing through the Health Department programs costs more because services for the sick and disabled are more intensive. They may include on-site doctors, psychiatrists, nurses, and a more intensive caseworker-to-client ratio. Health Department staff estimate that services account for about half the total cost of their supportive housing. The City cannot afford free housing for all comers. It is essential that supportive housing . continually demonstrate its cost effectiveness by reducing both the overall homeless population and the costs of local services that the homeless would otherwise consume, including shelters, hospitals, nursing homes, jails, and police and ambulance costs. The next bi-annual homeless count in January 2009 should show whether the initial decline in homelessness when Care Not Cash began is a continuing trend. The upcoming count should expand its efforts to conduct a survey to provide demographic data and help identify current needs. Ε. STUDIES OF SUPPORTIVE HOUSING'S COST EFFECTIVENESS Continual monitoring is needed to demonstrate the cost effectiveness of supportive į. housing programs. To date, some studies have shown a reduction in health care costs in the Health Department's supportive housing programs: a) A Goldman School of Public Policy, University of California, Berkeley study of the homeless entering two City sites between 1994 and 1998 (Canon Kip Community House and the Lyric Hotel, early innovators in supportive housing for the mentally ill) showed marked reductions in emergency room visits and hospitalizations among the "dually diagnosed" homeless (in this case, those diagnosed with mental illness and at least one other condition, usually addiction, HIV, or other physical disability). The study compared their medical expenses for the two years before they moved in with the two years after and found that hospitalizations and emergency room visits had about halved.4 Ł b) An internal Health Department review of housing and health costs for 98 formerly homeless adults entering its Plaza Apartments site when it first opened in December 2006 demonstrated a net saving in one year of about $20,000 per resident, after factoring the cost of building and operating homeless housing into its "housing as health care" equation. The savings were, however, largely attributable to 10 residents who moved in from Laguna Honda Hospital, a long- ١, term nursing care facility. Substantial if less dramatic savings were seen in data from two other sites operated under Health Department contract, Folsom Dore and the Empress. The factors that seemed to contribute most to housing stability and reduced health care usage were a mixed population building, a high concentration of seniors, and a neighborhood with less drug activity. However, the savings seen with the mentally ill or those with physical disease or disability have yet to be demonstrated among the chronic homeless or in the broader supportive housing population. Tia Martinez and Martha Burt, "Impact of Permanent Supportive Housing on the Use of Acute Care Health Services by Homeless Adults" (Psychiatric Services, July 2006, Vol. 57, No. 7: 992-999) http://ps.psychiatryonline.org/cgi/reprint/57/7/992. F. MEASURES OF PROGRAM SUCCESS - TIME TO GO BEYOND 'RETENTION RATE' AND TO FOSTER INNOVATION Measures of cost effectiveness also need to be developed for judging specific programs and organizations. The key measure of success in supportive housing programs is the retention rate, i.e. the percentage of residents who stay or leave for other permanent housing in a given period. The Human Service Agency's contracts typically set a target of a 90% retention rate over one year. Most of their non-profits achieve 97%. The Jury believes that more rigorous measures of success are needed. Success measures could focus on outcomes relative to intake criteria. These might stress . cost inputs ("Per Unit Per Annum" operating costs + capital costs or lease costs of bringing buildings online) and benefit outcomes (housing retention rates + savings on : public service costs like emergency services, jails and hospitalizations). This analysis ۲. would also have to include controls such as by population served, size of building and intake criteria (since some programs are selective and others take all referrals). Innovation and experimentation should be encouraged. This could be achieved by a framework that provides the same per unit resources to each organization, but sets few if any rules about how its programs are run. There is evidence, for example, that the caseworker to client ratio is irrelevant in the typical Human Services Agency single adult site. The Tenderloin Housing Clinic is experimenting with two smaller "step up" SROs: in place of the on-site caseworker and security desk staff, there are an on-site hotel owner/manager and a locked garden gate to which every tenant has a key. The owner is å allowed to select his tenants, many of whom work or are in job training programs. With Health Department clients the requirement might call for an intensive caseworker/client ratio, or caseworkers who can be contacted round the clock. The residential treatment ٠,٠ facility might require a licensed social worker, whereas para-professionals, peer counselors or volunteers could suffice in some supportive housing programs. The Jury believes that more variation and innovation should also be encouraged in the tenant mix and types of housing arrangement -- for example, by allowing communal living, or transitional housing programs tied to acceptance of specific kinds of services. While the "harm reduction" principle generally requires that services be voluntary, there is room for differences in approach. Several interviewees commented to the Jury that expectations were unnecessarily low for the homeless who entered supportive housing programs. There are currently a few housing programs which have selective intake criteria and require their residents to stay "clean and sober" and/or to be enrolled in treatment programs. One of these is the soon-to-open Salvation Army housing project in the Tenderloin, which offers 40 apartments to the chronically homeless who "are actively 5 Human Services Agency internal data working toward rehabilitating their lives." If tenant rights preclude such conditions in permanent housing, an expansion of transitional housing programs with incentives tied to participation in treatment programs should be considered. G. COST/BENEFIT COMPARISONS – LET'S FIND OUT WHAT WORKS BEST There are wide differences in operating costs. Some can be explained by the size of the building and the number of rooms over which the cost of the front desk security is spread. Some may be attributable to the relative lack of experience most non-profits have in property management, so that buildings run smoother with professional property managers. The Community Housing Partnership has an innovative approach in this area- -a supportive employment program which trains its formerly homeless residents in property management. It now manages six buildings for other non-profits with its own home-grown staff. 1 i The Human Services Agency is currently developing a "best practices" working group to explore the reasons for the variations in service and operating costs - whether tenant mix, service mix, staffing patterns or building conditions -- and to help define cost and outcome measures that will be used to evaluate future contract performance. It hopes to develop outcomes measures beyond retention rate (mentioned above), such as employment, income (either earned or from a mainstream benefit program such as SSI) or a move to other permanent housing. As the supportive housing system matures and residents turn into long-term tenants, such outcomes will become more common, although retention rate will probably remain the major measure of success. The Jury fully supports this working group's efforts, and believes this sort of cost/benefit analysis and emphasis on objective input/output measures of success should be undertaken in other homeless programs too, including supportive housing for seniors, families, and Health Department clients. There is great potential for cross-fertilization of ideas between programs. For example, rent subsidy vouchers now available to families and for AIDS patients might be considered for the chronically homeless adults, especially if they can be used outside San Francisco where rents are lower (as vouchers for families now can as long as they are under case management in the City). H. CONTRACTOR TRACKING AND REPORTING REQUIREMENTS - INCREASE THEIR USEFULNESS The same objective performance measures would improve oversight and accountability without stifling innovation and creativity. They could also reduce reporting requirements. Currently each contractor files periodic reports, varying greatly in form and content, to one of seven departments administering the funds. Non-profits that operate supportive housing programs are required to file quarterly and annual reports along with backup financial and other data. Some of the data relate to the key measures of housing retention and increased client incomes (whether through employment or qualification for SSI or other mainstream benefits) or reduced costs. But others are tied to looser objectives or to process measures (such as number of caseworker contacts, or attendance at coffee hours). A required annual client survey is limited to four ٠. narrow questions (e.g. whether posted notices are in a language the client can read) rather than being a customer satisfaction survey of a kind that a non-profit might devise and use . for itself. These particular reporting requirements are probably HUD's, but to the extent the City has reporting requirements of its own it should ask for information which is simple to provide and uniform across agencies, so that a non-profit with funding from various sources can use the same report for all. Ideally, it would ask for information that the non- profit finds useful to track for its own internal purposes, such as client housing and service history. The City's supportive housing programs expanded rapidly to meet the 3,000 additional housing unit target of the Ten-year plan. It now expects to meet this goal in 2010. Although no goals beyond this point have been set, projects currently in the pipeline will increase homeless housing units at a slower pace thereafter. Now it is time to take stock, set new goals, and increase cost effectiveness. While there is no computerized system for tracking and analyzing this large loose network of homeless housing and service providers, federal permanent housing funds are being used to pay for two staff positions to develop one. The Jury believes that such a tracking system, properly designed and maintained, will be an invaluable tool for establishing the effectiveness and cost effectiveness of these entities. I. THE HOMELESS AND SAN FRANCISCO'S QUALITY OF LIFE - A COMPLEX 1 CONNECTION ٠, The Jury has focused in this discussion primarily on the economic aspects of the City's initiatives to address homelessness. The amounts of money and the scope of the challenges are vast, and these have to a great extent received the appropriate amount of attention by City leadership and departments. The Jury is aware, however, that many San Franciscans equate the homeless with conditions they encounter every day as they travel virtually anywhere in town: panhandling, public drunkenness, public drug use, sleeping in doorways, and the like. It is likely that San Francisco residents and visitors have not noticed improvement in these conditions since the Ten-Year Plan and Care Not Cash began. They are understandably skeptical of counts of the homeless population that show sharp reductions. To some extent, the success of these programs has been invisible to the public, because many of the formerly homeless people who now have places to live still spend much of their day on the street. It is important that public support be maintained for worthwhile programs to house the homeless. This can be achieved only if the City formulates an equally aggressive, long-term plan to address quality of life issues. į ŀ V. FINDINGS 1. Homeless spending is large scale, complex and dispersed, but the City has no mechanism for regularly monitoring it across departments. ř 2. Updating the six-year old 2001-2002 Budget Analyst's Report of direct and indirect homeless spending would provide worthwhile data to aid in assessing whether the trend in ever-increasing spending on homelessness is changing. 1 3. Current performance measures are not useful in comparing the overall performance of non-profit providers.

Efforts to monitor the cost effectiveness of supportive housing program have not been comprehensive, and have not included all cohorts of the population at risk of homelessness.

The costs of different homeless housing programs vary greatly for reasons which are poorly understood. Little analysis of these differences has been done.

It is unclear whether current plans to create one type of building over another or to develop new models are informed by past experience or by any cost/benefit analysis.

The City is just beginning to develop simple outcome measures of success relative to cost in its supportive housing programs.

Greater freedom to experiment with different approaches and levels of service could facilitate greater cost effectiveness in supportive housing programs.

While the City needs to monitor its contracts closely, it could simplify reporting requirements and provide a computerized method for tracking and comparing outcomes among non-profits that non-profits themselves could use in collaborations among themselves as well as for their own internal uses.

San Franciscans' continued support of efforts to house the homeless may depend on their observing an improvement in quality of life attributes such as aberrant street behaviors. VI.