Legislative Analyst's Office, May 1998

California Counties

A Look at Program Performance

Part I

Table of Contents

Introduction

Children's Programs

Social Services and Health

Criminal Justice

Roads and Libraries

Conclusion

Appendix: Information Needed to Assess Performance

Data Supplement

Acknowledgements

Introduction

Californians depend on counties for many important services. Counties prosecute, jail, and supervise most criminals; operate libraries; work to protect children from abuse; help residents with health, mental health, and substance abuse problems; fix potholes; and serve their residents in many other ways. For many programs, counties serve as local agents to implement the state's priorities. Counties frequently administer their programs under guidelines in state law.

Despite the importance of county programs, there is little information on their results, or "outcomes." This shortage of information makes it difficult for policymakers, residents, or county administrators to gauge a county's performance, or to observe changes in county performance levels over time.

This report compiles the best information available on county programs--and gives Californians a snapshot view of county performance. Whenever possible, this report displays information comparing the outcomes of county programs. For some programs, however, information on program outcomes is not available, and the report instead displays information on key county inputs (such as staffing levels or treatment slots) that we believe are associated with program success. Finally, for some programs, such as mental health, adult protection, and public health, we could not obtain satisfactory information on county outcomes or inputs. In these cases, the report does not provide information on the program.

In selecting the program outcome and input measures included in the report, we focused on measures where differences in county management or policies explained at least part of the results. In many cases, we performed statistical analyses to examine the relative influence of demographic and economic factors, and other factors beyond a county's control.

This report's size reflects the small amount of information currently available on county performance. Given the data limitations and other extenuating factors, we caution the reader from assuming that differences in county management fully explain every county's performance. Specifically, a county's relative ranking under some of the report's performance measures may be influenced by the county's demographic characteristics and access to revenues. As we described in a companion report, Why County Revenues Vary: State Laws and Local Conditions Affecting County Finance, counties differ greatly in their revenues and program demand. In addition, because of information limitations, some important county activities are not included in this report, or are measured imperfectly.

While recognizing these limitations, we publish this report in the hope of launching a dialogue about the outcomes of important county programs. It is also our hope that someday Californians may have a more complete report on county program performance.

The report is presented in a graphical format, with chapters summarizing county program performance in four program areas: children's programs, social services and health, criminal justice, and roads and libraries. The charts typically summarize data for the state's 15 to 20 most populous counties, or the most populous counties for which reliable data are available. In addition, the data supplement in the back of the report includes available data for the remaining counties. The report's final chapter presents our findings on county performance as a whole.

Children's Programs

California's county-administered Children's Services and Foster Care programs are designed to protect children from abuse and neglect. When a case of abuse is confirmed, counties work to help the family in crisis and may place the child temporarily--or permanently--in another home.

While county child protection programs operate under many federal and state laws and guidelines, counties have some discretion over the level of services provided and how their programs are administered.

In our analysis, we relied primarily on data compiled by the Child Welfare Research Center at the University of California at Berkeley (in Performance Indicators for Child Welfare Services in California: 1996). We note that these measures pertain only to program activities after the county has made the decision to remove the child from the parents' home. If there were data on those cases where the county did not remove the child from the home after a report of abuse or neglect and then there was subsequent maltreatment, a more complete picture of county performance in child protection would be available.

The next three figures compare county performance in (1) moving children from temporary foster care to a permanent home within four years, (2) providing children in foster care with stable placements, and (3) reducing the likelihood that children who are reunified with their families will again be victims of abuse or neglect (that is, the rate of recidivism). A fourth figure combines county performance scores in the aforementioned areas of permanence, stability, and recidivism into an overall index of performance in child protection.

A Permanent Home Within Four Years

One of the priorities of child protection programs is to move children expeditiously from temporary foster care to a secure, stable ("permanent") home. This could be either reunification with the family, or placement with an adoptive family or guardian. Figure 1 compares county scores on a permanence index (as developed by the University of California at Berkeley Child Welfare Research Center) for children who first entered foster care in 1991-92 in 24 of the largest counties. Los Angeles County was dropped from the sample because of incomplete data. The index is constructed from data on children whose time in placement was primarily spent in nonrelative foster care. This is because many children placed in long-term foster care with relatives are in secure, stable homes.

Figure 1 indicates that there is a wide range in county performance in moving children from foster care to a permanent home, ranging from a high of 91 percent in Merced County to a low of 60 percent in San Francisco County. The statewide average is 75 percent.



How Often Do Children in Foster Care Change Homes?

It is widely accepted in the child development field that multiple foster care placements create instability, which can be detrimental to a child's development. Figure 2 reports the proportion of children experiencing no more than two different placements in their first spell in foster care, among those children who entered foster care in 1989-90 and remained in care for at least six years. The figure shows 14 large counties for which reliable data are available. Los Angeles County was dropped from the sample because of incomplete data.



San Francisco had the most stable foster care placements among these large counties, with 70 percent of children experiencing no more than two different placements in the first six years in foster care. By contrast, San Diego County had a rate of 40 percent, meaning 60 percent of their foster children experienced three or more placements in the first six years of their first spell in foster care.

We note that there may be an inverse relationship between the measures of permanence and stability. That is, counties that score well on the permanence index will have relatively few children remaining in foster care at the end of six years, but those children who do remain may be particularly difficult to maintain in a stable placement. This may result in a number of changes in their home placements. Finally, we note that the stability measure is calculated using data on both relative and nonrelative placements, and that children placed in nonrelative foster care tend to have less stable spells in foster care than is implied by the overall measure reported here.

How Often Do Children Reenter Foster Care?

Figure 3 shows the "recidivism" rates for the 25 largest counties. Specifically, this figure reports for those children who left foster care due to family reunification or placement with a guardian from 1993 through 1996, the proportion who reentered foster care within three years, or by the end of the sample period (December 31, 1996).



San Mateo County has the lowest recidivism rate among the 25 largest counties in California (8 percent); Shasta's recidivism rate is more than three times greater (25 percent). The statewide average is 16 percent.

Some care should be used in interpreting the recidivism measure. First, the extent to which the event that precipitated reentry involved abuse to the child is not known. Second, the measure reported here understates the true rate at which children reenter care within three years of reunification. This is because, for some children in the sample, three years had not elapsed between reunification and the end of the sample period. Finally, we note that the recidivism measure is calculated using data on both relative and nonrelative foster care placements. The reentry rates for children placed in nonrelative foster care tend to be higher than the rates for children in relative placements, so the measure reported here understates the reentry rate for children in nonrelative placements.

Overall Child Protection Index

It is useful to compare simultaneously, county performance on the measures of permanence and recidivism. A few counties--in particular San Mateo, but also Merced, Santa Clara, and San Diego--do well on both measures. These counties appear to have particularly effective child protection programs based on these measures: they move children quickly from foster care to a permanent home, and relatively few children reunified with their families return to foster care. On the other hand, Sacramento and Kern Counties are below average on both measures: children in these counties move to a permanent home relatively slowly, and a relatively high proportion of those children reunified with their families return to foster care.

Figure 4 compares county scores on a child protection index that combines the permanence, stability, and recidivism performance measures. A county's index score is the weighted average of its scores in each performance area, relative to the highest performing county. In other words, if a county had the highest score on each of the three performance measures, then its child protection index score would be 100 percent; if the county's score in each area were 75 percent of the highest county score, then its index score would be 75 percent.



We weighted performance in recidivism twice as heavily as performance in each of the areas of permanence and stability because we believe that recidivism--which is related directly to the act, or threat of, abuse or neglect--represents a particularly important program outcome.

San Mateo County had the highest overall level of performance in child protection and clearly stood out among the 14 counties shown in Figure 4, scoring at or near the top in permanence, stability, and recidivism. Sacramento County had the lowest overall level of performance.

Children's Programs Conclusion

Measuring differences in child well-being due to county service interventions is a very challenging task. The best available performance measures reflect the actions of a county after it decided to remove a child from the parents' home. Currently, there is relatively little information about the quality of county "front-end" decisions. If there were data on subsequent maltreatment in the cases where the county did not remove the child from the home after a report of abuse or neglect, then a more complete snapshot of county performance in child protection would be available.

We note that county performance may be influenced, to some extent, by demographic variables that are beyond the control of the county. However, in our statistical analysis we found no significant relationship between county performance in child protection and demographic variables, including poverty, unemployment, crime, unwed births, ethnicity, and median county income.

Return to California Counties Table of Contents

Continue to California Counties Part II--Social Services and Health

Return to LAO Home Page