CITY AND COUNTY OF SAN FRANCISCO
BOARD of SUPERVISORS

OFFICE OF THE LEGISLATIVE ANALYST

TO: Honorable Members of the Board of Supervisors
FROM: Jesse Martinez, Legislative Analyst
DATE: May 9, 2002
SUBJECT: Cash Assistance Programs (file # 012179)

SUMMARY OF REQUESTED ACTION

A motion (introduced by Supervisor Newsom) requesting the Office of the Legislative Analyst (OLA) to prepare a report comparing the best practices and/or models on Cash Assistance Programs utilized in Alameda County and other surrounding Bay Area counties.

EXECUTIVE SUMMARY

Our office surveyed 5 jurisdictions on their cash assistance programs as they address the homeless issue. We discovered that most have shifted from an emphasis on cash assistance to in-kind services. In-kind services can include shelter and a variety of underlying needs such as clothing and food.

In addition, we learned that San Francisco pays the largest amount of cash assistance of the jurisdictions surveyed. For instance, San Francisco pays $395.00 per month in general assistance cash aid as compared to $336.00 paid out in Alameda County, $167.00 in Contra Costa County, and $58.00 in San Mateo County.

Finally, we found that the jurisdictions that shifted from cash assistance to in-kind services have done so because of budget constraints and cost-saving goals. Specifically, 4 of the 5 jurisdictions surveyed cited cost-efficiency as the reason for the adjustment.

BACKGROUND AND PRACTICE

Overview

Homelessness is a key topic confronting San Francisco for some time¹. The City has focused on this issue with the understanding that it is not caused merely by the lack of shelter, but involves a variety of underlying, unmet needs. This process of identifying, assessing, and meeting the service needs of the homeless began some time ago.

In fiscal year 2001-02, the City and County of San Francisco spent about $9.5 million in cash grants in the County Adult Assistance Programs (CAAP) for individuals who declare themselves homeless². This represents about 2,500 clients (30% of the CAAP population). These programs include General Assistance (GA), Personally Assisted Employment Services (PAES), Supplemental Security Income Pending (SSIP), and Cash Assistance Linked to Medi-Cal (CALM).

The General Assistance program is intended to meet the mandates under the State of California Welfare and Institutions Code (Section 17000), which requires counties to "relieve and support all incompetent, poor, indigent persons." In San Francisco, this mandate is met through the provision of monthly cash grants in the amount $320 and through the provision of emergency medical care at San Francisco General Hospital.

In addition to state law, GA is governed by local ordinance³ that states that the purpose of GA is:

(1) "To provide short-term financial or in-kind instance and other services to indigent residents of the City and County who are unable to support themselves and have exhausted their own means of support."

(2) The program is further intended to "enable and encourage persons aided to: find employment if employable; receive support from other federal or State sources; and to reduce or eliminate the conditions that have led to indigency and dependency."

(3) In addition, "assistance is to be administered in a manner which is consistent with and will help achieve basic program purposes and which respects individual privacy and personal dignity and to encourage self-respect, self-reliance and the desire to be a good citizen, useful to society.

The purpose of PAES is to provide supportive services and to assist the individual in obtaining paid employment, The purpose of SSIP is to assist disabled individuals in applying for and securing Supplement Security Income (SSI-federal disability payments). The purpose of CALM is to provide cash assistance to individuals who because of age or disability are eligible for state disability payments, but who are ineligible for federal disability payments.

Evaluation of Cash Assistance vs In-Kind Services

Researchers have discovered that government benefit checks to the poor are generating accidents, homicides and substance abuse leading to deaths⁴. The evidence lies in the fact that the federal government mails out the checks at the start of every month and deaths from these causes are higher in the first week of each month, compared with the last seven days of the prior month. The author, D. P. Phillips, points out that care for the poor comes from both the public and private sector, with the latter providing mostly "in-kind" help -- such as food, clothing or shelter. Public-benefits programs rely on cash, which can easily be spent on illegal drugs and alcohol. Phillips suggested that the federal government should consider acting more like a private charity, to keep addicts from indulging in the first few days of ever month.

However, some scholars take issue with the view that cash assistance should be decreased for supportive services.

Indeed, they argue those most disabled persons without sufficient resources have very little freedom of choice. Rather than being provided with a cash supplement which they, too, could spend to maximize independence, the poor in general either are restricted to choices from a narrow list of services paid for by public monies, or, more commonly, are limited to the type and quantity of services they can receive as dictated by agents of a public authority. According to this view, society is implicitly saying that the financially needy disabled cannot manage their resources as well as all other disabled persons and should not be trusted to do so⁵.

Nonetheless, increasing research does appear to support the notion that "public support..may stimulate increased substance abuse among beneficiaries who suffer from addictive disorders."⁶

POLICY ISSUES ANALYSIS

Problem

According to most of the jurisdictions surveyed, years ago, the cash grant was sufficient to enable homeless recipients to obtain permanent housing. With the increase of housing costs in San Francisco and the Bay Area, this is apparently no longer the perception by many of the Bay Area County providers. As a result, programs now provide an individual with a cash grant that is apparently insufficient to secure consistent, permanent housing.

Proponents of in-kind services believe that by providing a monthly cash grant that is insufficient to meet homeless clients" housing needs, the provider or City is in turn allowing them to remain homeless. The argument continues in that the cash provides them with sufficient support to live on the street thus creating further harm to themselves and potentially to others. This population utilizes the shelter system, lives part-time in SROs, and lives on the street. Some are also mentally ill and/or have substance abuse problems that prevent them from progressing toward self-sufficiency. Cash provided to those recipients who have alcohol and drug addictions certainly perpetuates these conditions by "enabling" them to continue to use alcohol and/or drugs. Ultimately the argument ends by claiming the City is certainly not fulfilling its locally legislated requirement to "administer a program as to encourage self-respect, self-reliance and the desire to be a good citizen, useful to society."

The Appendix to this report (see Table I) contains data comparing the caseloads of Bay Area counties and details their aid payment levels and General relief policies as reflected directly by the Bay Area County welfare agencies surveys of November 2001 and January 2002. The amount of cash provided by San Francisco is far higher than other Bay Area counties. Alameda, Contra Costa, and San Mateo rely on housing vouchers or other in-kind services to comprise part of their general relief payment. Most of these counties have policies for limiting cash aid to homeless persons.

The level of San Francisco"s commitment to General relief is much larger when compared to other jurisdictions statewide. In fact, San Francisco offers the highest maximum cash grant in the state. The City also spends the most when you consider the amount per capita in comparison to other counties (see Chart I in the Appendix).

Survey Methodology

Most of the 5 Bay Area counties we surveyed and who have shifted their general assistance emphasis from cash assistance to in-kind services have done so recently. Thus, they have not had sufficient time to conduct an adequate baseline evaluation to describe a prospective best practice model⁷.

Promising practices are programs and strategies that have some quantitative data showing positive outcomes in delaying substance abuse over a period of time, but do not have enough research or replication to support generalizable outcomes.

Thus, our methodology required that we get beyond the initial question: Does the shift from cash assistance to in-kind services work in promoting self-sufficiency in the homeless population? Does this policy `shift" decrease the potential for self-sufficiency.

A more useful question at this juncture is: What are the characteristics of this policy that matter most for good practice and for good homeless program outcomes? The counties have not agreed on a measure for identifying whether an in-kind services program is or is not a viable program.

Therefore, this report focused upon the shift from cash-assistance and its objectives, incentives, and finally, preliminary results. In addition to our survey of these elements, we also asked about their prospective research into related concerns raised by Dr. Phillips, which suggests that there might be a correlation in cash vs. in-kind help.

We specifically inquired whether the counties had commenced utilizing the death registry to determine if the average death per day on the first of each month is larger than the rest of the month?

For instance, Dr. Katz opined that by using San Francisco"s death registry and looking at a 48 month period from say November 1, 1997 to October 31, 2001, the average deaths per day on days 1-7 of the month was 19.4 average deaths per day. On subsequent days of the months the average number of deaths was 18.8.

Based on this rather preliminary review of data, Dr. Katz speculated that the average number of deaths is higher at the beginning of the month. A likely interpretation is that this is in association with receipt of public assistance. The difference is not very big, and may or may not be statistically significant. To complete an extensive study similar to Dr. Phillips, cited earlier, would be costly and time intensive. All the jurisdictions surveyed had not undertaken this type of research.

JURISDICTIONS

Summary of the jurisdiction survey:

SUMMARY OF JURISDICTIONS AND USE OF IN-KIND SERVICES

ALAMEDA

Incentives for Change--A main impetus for the change in the County"s approach to the homeless issue was an effort in cost efficiency. According to county representatives, within a 12 month period July 1995 and July 1996, the county approved Community Housing and Shelter Services (CHASS), reduced the grant to $221 (through mandate relief), and stopped paying the full grant to those who refused City provided housing.

Subsequently, the caseload fell from around 11,200 to 7,200. Because so many changes happened at the same time, it was difficult for the County to determine which element was directly the cause for the change. The reason for the decline remains elusive requiring more research.

Another stated impetus for the shift was to provide shelter for the homeless and stabilize the target population (adults), and ultimately decrease homelessness by treating the causes.

Decision Outcome-According to Alameda County Social Services representatives, GA still remains too costly at $580 monthly per person. In addition, there is a significantly large percentage of mentally ill in the target population where the treatment is relatively costly compounded by the associated substance abuse problems which many of the clients have.

The County stated that when they instituted a ruling that individuals requesting GA without a home address would be offered placement plus $17 for supplementary items, there was concerns that critically needy residents would be denied basic essentials and suffer accordingly. However, they have not seen statistical evidence of deteriorating outcomes, let alone death, due to this policy. Initially, the County did not have the capacity to refer, but since then, they generally have referral sources for all requests. The County believes that there are some individuals, mostly the mentally ill with substance abuse and, with behavioral problems that can"t be managed in County facilities.

This segment of the population is given GA even though they generally have no address, however, to avoid leaving them with nothing, the County allows for GA. Ironically, the County stated that most single adults without an address refused GA rather than accept shelter. The current number of individuals in this voucher vs. direct payment program is 115 with a total capacity of about 150. This number does not include those individuals placed in alcohol and other drug services that use GA to help offset their residential costs.

The County is most cautious in proceeding on this shift to in-kind services and believes that this approach is more effective where GA clients reside in stable housing.

SAN MATEO

Incentives for Change--San Mateo County shifted to a vendor-based program in August 1, 1993. According to the San Mateo County Human Services, they felt that many counties had already implemented some sort of a vendor/voucher system with varying caseload reductions from 5 to 35 percent. They estimated that with implementation in San Mateo there would be a resulting 35 percent reduction in caseloads. One of the arguments in favor of an emphasis on a vendor/voucher system was that it would more readily assure that the basic needs of housing, utilities and food of recipients would be met. It was noted that an internal survey reflected that 20 percent of San Mateo"s GA clients at that time had substance abuse problems and that most of this population did not have an adequate proficiency in financial literacy. In addition, homeless clients are able to save money in an escrow account from their GA funding that would enable them to meet housing move-in costs. Cost was one of the reasons for the change to in-kind services but it also provided more client and case stability. Landlords were now being more assured of getting their rent and were more willing to rent to GA clients.

Decision Outcome-According to the County, there was an estimated cost savings subsequent to the shift to in-kind services but the number reflected is an estimate and would not be accurate in reflecting the actual reduction in costs based on the implementation of a vender/voucher system.

There has been a reduction in cases since the implementation of the plan.

One of the additional costs to the system is multiple voucher issuance. For instance, one warrant for rent, one for utilities, one for food, and one to a client for cash as opposed to the previous system of just one issuance to a client.

CONTRA COSTA

Incentives for Change--In 1996 Contra Costa County instituted a "vendor-pay" for housing. The objective was to allow the County to pay directly to landlords as in-kind assistance to meet client housing needs in-lieu of cash assistance. The process requires that these payments for housing issued to a landlord was to be reported by the parties to the United States Internal Revenue Service and the California Franchise Tax Board if the amounts were larger than $600.00 annually. Another major reason for the change was to curtail the cost of GA.

Decision Outcome-According to the County Employment and Human Services, since the change to in-kind services as an emphasis in the homeless policy, the average grant is approximately $225-$250, notwithstanding the maximum grant allowed of $352.00 per month.

SANTA CLARA

According to the Santa Clara County Social Services, they had not shifted from cash assistance to in-kind services. In addition, they indicated that if their GA clients receive any item of need from a charitable organization, which is received in full, they do not count it as income to the GA budget.

SACRAMENTO

Incentives for Change--The Sacramento County Human Assistance refers to their homeless policy shift to in-kind services as Aid-in-Kind. This was initially a pilot program for a period of time and then was made permanent in 1990. According to the County, since 1990 they have been offering residence at an Aid-in-Kind shelter for homeless individuals and cash benefits to domiciled individuals.

In-kind services in Sacramento County consist of admission to one of two Aid-In-Kind residence facilities,

which provide housing, food, employment and housing services, a monthly bus pass and $15 in cash - all in a group living shelter situation. The in-kind services are mandatory for any GA applicant who states he/she is homeless and optional for recipients who are homeless and not listed as being on their GA roles. The process dictates that if a homeless individual is not appropriate for group living then they are eligible to receive cash assistance.

The primary objective was to address the homeless problem and secondarily - cost-efficiency. The County believed that by allowing cash to the homeless who were largely incapable of successfully managing money, was simply not addressing their basic needs as well as not correcting the homeless cause. That is, they stated that to address the "homeless problem" means to provide the basic needs, e.g., shelter, food, treatment, etc., so that eventually the public sector can empower the homeless to move out of this condition.

Decision Outcome--In any given month 120 individuals are living at the Aid-In-Kind shelters instead of on the street. A number of these individuals, accompanied with the services provided, are able to secure permanent housing and employment and leave the shelter or at the very least move into Transitional Housing which is a longer-term housing program.

The County is convinced that it is much cheaper for Sacramento County to run the Aid-In-Kind program for these 120 individuals each month than to have them receiving a monthly GA grant indefinitely without any supportive services to help them transition from their homeless cycle to a more stable living situation.

CONCLUSION

Based on our preliminary research and analysis, our office concludes that adopting in-kind services over cash assistance would aligned San Francisco with other Bay Area jurisdictions. Also, the suggestions by the Department of Public Health indicates that there may be a correlation between the number of homeless deaths and cash assistance pay out days.

Recommendation--This study suggests that policies that are developed on the basis of anecdotal evidence, even though they are responsive to the hardening tide of public opinion, may lead to inefficient and ineffective use of public resources, or to punitive withdrawal of funds that are desperately needed and that have been shown to contribute to achieving their desired goals. Indeed, researchers in the field agree that requisite and adequate research and analysis be conducted before any distinct conclusions are made on this issue.

Whether San Francisco chooses to reduce cash assistance and replace it with in-kind services, or adopt a combination is a policy matter for the Board of Supervisors.

APPENDIX

TABLE I

BAY AREA GENERAL RELIEF CASELOADS AND HOMELESS POLICIES

Data Source: Bay Area County welfare agencies survey, November 2001 and January 2002.

APPENDIX-CHART I

Source: California Health and Human Services, Department of Social Services, Data Systems and Survey Design Bureau, General Relief 237

¹ November 13, 1976, File 446-76-1, Ordinance No. 451-76; "Cash Assistance Linked to Medi-Cal]

Ordinance amending Administrative Code to establish a Cash Assistance, June 1, 1998. File 97-97-70, Ordinance No. 149-98

² "Survey of the Cost of Direct Services Provided to the Homeless Population," Budget Analyst, Board of Supervisors, December 31, 2001

³ San Francisco Administrative Code, Chapter 20 Social Services, article vii general assistance program.

⁴ "An Increase in the Number of Deaths in the United States in the First Week of the Month," D.P. Phillips, The New England Journal of Medicine, vol.341, number 20, November 11, 1999.

⁵ "An Evaluation of the Veterans Administration Housebound and Aid and Attendance Allowance Program," John M. Grana and Sandra M. Yamashiro, Project HOPE Center for Health Affairs April 15, 1987.

⁶ "Do public support payments encourage substance abuse?"

Health Affairs; Chevy Chase; Fall 1996; Rosenheck, Robert; Frisman, Linda; Vol 15, Issue 3.

6

⁷ Best practices are those strategies, activities, or approaches which have been shown through research and evaluation to be effective at preventing and/or delaying substance abuse; "best practices" are those strategies and programs which are deemed research-based by scientists and researchers: Center for Substance Abuse (CSAP),Western Center for the Application of Prevention Technologies, U. S. Health and Human Services