Legislative Analyst's Office, May 1998

California Counties

A Look at Program Performance

Part II

Social Services and Health California counties administer many social service and health programs for low-income families and individuals. These programs include: cash grants to families with children (California Work Opportunity and Responsibility to Kids [CalWORKs] program), cash aid to indigent adults without children (general assistance), mental health, drug and alcohol treatment services, child support collections, public health, indigent health, and Medi-Cal eligibility determinations.

For many of these programs it is difficult to gauge county performance because cross-county information is not available. In other cases, the level of county performance is difficult to ascertain because program outcomes are largely determined by policies or decisions made by the state or federal government--or other factors beyond the county's control. As we discussed earlier, we have sought to exclude measures where most of the cause of the variation was beyond a county's control.

In reviewing county social service and health program responsibilities, we found three programs where reasonably good performance information is available: CalWORKs, child support collections, and drug and alcohol addiction treatment. While limited in scope, the information presented in this chapter provides an interesting perspective on county success in maintaining the state's safety net for those in need.

Where Are Welfare Rates Declining Fastest?

Federal and state welfare reforms have focused increasing attention on moving families from welfare to self-sufficiency. To measure the effectiveness of county efforts to promote this goal, ideally information would be available on the "work readiness" (education and employment history) of each family on the CalWORKs program, and each family's change in income after leaving the program. Because such data do not exist, Figure 5 provides a "second-best" measurement of county success. Specifically, Figure 5 shows for the 20 largest counties, the decline in welfare dependency rates between April 1995 (the statewide peak in CalWORKs caseload) and October 1997. During this period, CalWORKs cases dropped by 20 percent statewide, but there was variation among counties.

Ideally, the state would like to know how much of the differences shown in Figure 5 reflects the relative effectiveness of county programs, rather than other factors outside county control. The limited information available prevents a precise assessment. Nevertheless, our statistical analysis found that changes in the local economies did not account for the differences in caseload reductions. In addition, although differences in funding for state employment services explained about 40 percent of Figure 5's variation, accounting for these funding differences seldom changed the rank order in which counties are listed in Figure 5. Accordingly, we conclude that a significant portion of the variation among county declines in welfare dependency rates probably results from differences in (1) relative work-readiness of county caseloads and (2) county program management.

Recently, the state made major changes to this welfare program, including giving counties new fiscal incentives and increased funding for employment services. Given the magnitude of these changes, future county performance may well be different than the recent past.

Collecting Child Support

When a noncustodial parent fails to pay court-ordered child support, the custodial parent may turn to the county district attorney for assistance. Such child support collection services are available to welfare families and nonwelfare families. Child support collections on behalf of welfare recipients are used to reduce state, county, and federal public assistance costs, as well as help move a family off welfare. Collections made on behalf of nonwelfare clients are distributed directly to the clients.

Our review of this program found that county methodology for closing collection "cases" varies throughout the state, as does the income of noncustodial parents. In order to develop a fair index of county program success, therefore, we limited our review to cases in which the family receives CalWORKs aid. Limiting our review offered two advantages: (1) a more reliable data base and (2) there tends to be less variation among noncustodial parent incomes.

Figure 6 compares the 25 largest counties on an index of their performance in child support collection in 1996-97. The index measures each county's collections per case on behalf of families receiving CalWORKs, relative to the collections of the best performing county. For example, the county with the highest collections (Sonoma) receives a score of 100 percent, and a county with a score of 75 percent would indicate that the county had 75 percent of the collections of the best performing county. Figure 6 indicates that there was a wide range in county performance. Four counties (Sacramento, Riverside, San Bernardino, and Los Angeles) had less than half the rate of collections of the highest performing county (Sonoma). While the findings in Figure 6 pertain to CalWORKs cases, our review suggests that county success in collecting child support for nonwelfare families is likely to be similar.

Helping People with Addictions

Counties offer numerous substance abuse treatment services, ranging from intensive short-term detoxification to longer-term outpatient counseling and treatment. The state Department of Alcohol and Drug Programs collects program data from each county, including the number of publicly funded treatment "slots" available. A client occupies a slot when receiving treatment services.

Figure 7 shows the number of treatment slots per 10,000 population for the 15 largest counties in April 1997. A variety of factors affect the number of treatment slots, including the amount of state and federal funds allocated to a county for drug treatment, the amount of county funds allocated for drug treatment, the availability of services, the level of demand for certain types of treatment, and the willingness of providers to create new treatment programs in a county or geographic area. We note that some of these factors are beyond the control of county drug and alcohol programs. Furthermore, while the number of treatment slots does not measure program performance, it is an important "input" variable that may be related to program outcomes.

The figure shows San Francisco County with the highest treatment capacity among the large counties, with 54 slots per 10,000 population. In contrast, Orange and San Mateo Counties each have 11 slots. The average among the large counties is 22 slots, while the statewide average (excluding the three smallest counties) is 28.

Between 1995 and 1997, 14 of the 58 counties reduced the number of treatment slots per 10,000 residents. Of the counties shown in Figure 7, Orange reduced its treatment capacity the most, cutting nearly a third of its slots. Santa Clara, Contra Costa, and San Mateo Counties also reduced treatment slots, all by less than 10 percent. In contrast, Fresno County more than doubled its treatment capacity, while Los Angeles County increased slots by 78 percent.

Waiting for Substance Abuse Services

When people needing help with drug or alcohol addictions must wait lengthy periods of time for treatment, their health may worsen and some may engage in criminal activity. Factors affecting wait time in a county include the number of treatment slots, the demand for each type of treatment, and the length of time clients remain in treatment. Thus, while the length of time waiting for treatment services does not measure the effectiveness of county treatment programs directly, it is a useful indicator of the accessibility of county treatment programs.

During April 1997, 3,800 people statewide moved into treatment programs after waiting an average of 20 days for services. As Figure 8 shows, wait times ranged from 102 days in Kern County to less than 6 days in Orange County. It is important to note that the length of waiting lists varies by treatment types. People who moved into a residential detoxification program waited an average of four days for services. In contrast, the wait for a narcotic treatment program slot was much longer, which in some counties contributed to a longer overall wait time for services. Kern County, for example, moved 7 people (who waited an average of nearly two years) into the narcotic treatment program, and 72 people (who waited an average of 42 days) into other types of treatments.

Finally, we note that while this measure is useful in considering county program performance, it does not accurately gauge county program outcomes. It could be, for example, that a county's waiting list is long because it has a good outreach program and very successful treatment results that tend to attract more clients.

Drug and Alcohol Deaths

Because of the detrimental effects of drug and alcohol addictions on people's health, capacity to earn a living, and ability to maintain shelter, some people die from untreated drug and alcohol addictions. Figure 9 illustrates an outcome measure for county drug and alcohol programs: the average annual number of drug- and alcohol-related deaths per 100,000 population from 1993 through 1995.

The figure shows that drug-related death rates ranged from over 20 deaths per 100,000 in San Francisco to less than 5 deaths in Santa Clara. For alcohol-related deaths, San Francisco had the highest rate, 18 deaths per 100,000 population, while Santa Clara again had the lowest death rate.

We examined the data regarding drug-related deaths in more detail (time limitations precluded a similar review of alcohol-related deaths). We tested whether factors such as poverty, treatment funding, and the number of treatment slots influence the drug-related death rate. We found that per capita treatment funding and the number of county-provided treatment slots had a positive relationship with the death rate—in other words, the greater the number of treatment slots and per capita funding, the higher the drug-related death rate. The most likely explanation for this finding is that counties increase treatment capacity in response to problems caused by substance abuse, such as a high death rate due to drugs or alcohol. We did not find evidence that the availability of treatment reduced the death rate, but because data on death rates were only available through 1995, we were unable to assess the effects of increased treatment capacity over time. Much more analysis (including studies that cover several years and control for other variables) would be needed before drawing any conclusions.

Social Services and Health Conclusion

The preceding figures displayed information on three important county social service and health program responsibilities (CalWORKs, child support, and drug and alcohol treatment). In reviewing the information, we note that some counties, particularly Santa Clara, San Mateo, and Orange, regularly score well, whereas Sacramento, Kern, San Bernardino, and Los Angeles Counties rate lower. It is important to recognize, however, that the data are not adequate to delineate how much of these differences reflect variations in county program management, demographics, and other factors. Similarly, the data collected often are measures of "inputs" rather than program outcomes.

What information does California need to collect to allow a better examination of program success? In general, we find that counties (or the state) already collect detailed information on the number of people receiving services, but little information on program outcomessuch as the change in income of people leaving welfare, or the number of relapses experienced by people provided drug or alcohol treatment. In addition, because the information on people receiving services seldom is presented in detailed form, it is difficult to determine the extent to which differences result from variations in county administration, or the characteristics of the caseload. In the case of welfare programs, for instance, it would be helpful to have data on recipients' work readiness. In the case of drug and alcohol programs, it would be helpful to have data on the recipients' level of drug or alcohol use, level and severity of criminal activity, and ability to work. Absent increased efforts to collect data, it will continue to be difficult to gauge how well counties are performing at their social service and health tasks.

Return to California Counties Table of Contents

Continue to California Counties Part III--Criminal Justice

Return to LAO Home Page