Legislative Analyst's Office Analysis of the 2002-03 Budget Bill

Foster Care

Foster care is an entitlement program funded by federal, state, and local governments. Children are eligible for foster care grants if they are living with a foster care provider under a court order or a voluntary agreement between the child's parent and a county welfare department. The California Department of Social Services (DSS) provides oversight for the county-administered foster care system. County welfare departments make decisions regarding the health and safety of children and have the discretion to place children in one of the following: (1) a foster family home (FFH), (2) a foster family agency (FFA) home, or (3) a group home.

The 2002-03 Governor's Budget proposes expenditures totaling $1.5 billion from all funds for foster care payments. The budget proposes $426 million from the General Fund for 2002-03, which is an increase of $9.8 million, or 2.3 percent, compared to 2001-02. The caseload in 2002-03 is estimated to be approximately 70,800, a decrease of 6 percent compared to the current year. Most of this decrease is due to child exits from foster care to the Kinship Guardianship Assistance Program, which is part of the California Work Opportunity and Responsibility to Kids program.

Supportive Transitional Emancipation Program Is Overbudgeted

We recommend reducing General Fund support for the Supportive Transitional Emancipation Program by $4.6 million in 2002-03 to account for (1) implementation delays and (2) a reduced caseload. The program is overbudgeted in the current year by $1.1 million for the same reasons. (Reduce Item 5180-101-0001 by $4.6 million.)

Background. Chapter 125, Statutes of 2001 (AB 427, Hertzberg) created the Supportive Transitional Emancipation Program (STEP), which provides cash assistance--approximately $600 monthly--to eligible former foster youth. The purpose of this program is to provide former foster youth with cash assistance for a limited period of time while they transition from foster care to independent living. To be eligible, former foster youth must (1) have been in either the Foster Care or Kin-GAP Program on their 18^th birthday, (2) have been placed in foster care by a county that is participating in STEP, (3) be under 21 years of age, and (4) be following a county-approved independent living plan.

Chapter 125 became operative in January 2002. The budget proposes $3.7 million ($1.5 million General Fund) to support STEP in 2001-02 and $33.5 million ($13.4 million General Fund) in the budget year. Costs are shared by the state and counties 40 percent and 60 percent, respectively. County participation is voluntary. The proposed funding would provide cash assistance to an average of 1,000 youth per month in the current year and 4,700 youth per month in 2002-03.

Flawed Budgeting Assumptions. The budget proposal overestimates the amount needed to fully fund STEP because of three flawed assumptions that all tend to overestimate costs. Specifically, the assumptions are: (1) timely implementation by DSS and the counties, (2) participation by all 58 counties, and (3) participation by all emancipated foster youth.

State and County Implementation Delayed. In order for STEP payments to reach eligible youth, both DSS and the counties need to be fiscally and programmatically prepared to implement the program. Thus far, DSS has failed to meet key deadlines for a January 2002 start date. For example, neither written statewide guidance nor county fiscal claiming directions have been provided. We note that even after such key steps are completed by DSS, counties may need up to several months to implement changes to local policies and practices. Finally, due to the current economic environment and the significant county share of cost in STEP, counties have expressed concerns over the potential cost of STEP. At the time this analysis was prepared, only eight of the 58 counties had expressed a firm commitment to current-year participation in STEP.

Budget Assumes Universal Youth Participation. The budget assumes that all 58 counties will agree to implement STEP and that all eligible emancipated foster youth will choose to participate in the program, even though participation is voluntary. In our view, this assumption is unrealistic. We believe that some former foster youth will choose not to participate, perhaps on the basis that they are receiving adequate financial support from employment or other sources of public assistance. Other emancipated youth may choose not to participate because they dislike government programs or may object to the requirements of the independent living plan.

LAO Estimate of STEP Budget. We have developed an estimate of STEP costs taking into account (1) the delayed state and county implementation, (2) the degree to which fiscal concerns will prevent certain counties from participating, and (3) the degree of voluntary participation by youth. Our estimate assumes (1) a three-month delay in implementation and (2) that ultimately, only 70 percent of eligible youth would participate in 2002-03. Under our assumptions, the savings would be $1.1 million from the General Fund in 2001-02 and $4.6 million General Fund in 2002-03, relative to the Governor's budget. This assumes that cash assistance would be paid to an average of approximately 300 youth per month in the current year and 3,100 youth per month in 2002-03. Even our estimate may overstate program costs because the Governor's budget may overestimate the universe of eligible foster youth. Specifically, data from Los Angeles County suggest that the newly eligible 18 year-old population may be significantly smaller than assumed in the Governor's budget. Nevertheless, at this time we would recommend that General Fund support for this program be reduced by $1.1 million in the current year and $4.6 million in the budget year.

Examining the Role of Foster Family Agencies

Based on our review of the Foster Family Agency (FFA) program, we conclude that (1) children stay longer in FFAs than in foster family homes (FFHs); (2) neither child nor family background differences explain the longer FFA stay; and (3) youth in FFHs are reunified with biological families and adopted at a much higher rate than FFA youth. Accordingly, we provide three options for limiting the number of FFA placements: (1) hold funding to current-year levels; (2) decrease FFA treatment placements by 20 percent; and (3) reduce FFA rates over time.

Background

Following the investigation of child abuse or neglect, county welfare departments make decisions regarding the health and safety of children and have the discretion to place a child in one of the following: (1) a foster family home (FFH), (2) a foster family agency (FFA) home, or (3) a group home. The FFHs must be located in the residence of the foster parent(s), provide services to no more than six children, and be licensed by DSS. The FFAs are nonprofit organizations that recruit foster parents, certify them for participation in the program, and provide them with training and support services. Group homes may vary from small, family-like homes to larger institutional facilities and generally serve children with greater emotional or behavioral problems who require a more restrictive environment.

In theory, the respective foster care rates were designed to reflect the needs of children. These rates are: (1) FFH$425 to $597 monthly plus "specialized care increments" for children needing special support services; (2) FFA$1,589 to $1,844 monthly; and (3) group home$1,454 to $6,371 monthly. Figure 1 compares FFH and FFA rates, including the approximately $1,000 additional per child paid monthly to FFAs for services and administration.

History of Foster Family Agencies

Why Were Foster Family Agencies Created? In the mid-1980s, the Legislature became concerned about (1) the increased use of group homes; (2) the associated cost increases in the foster care program; and (3) county difficulties recruiting and retaining foster parents. The FFAs were created to address some of these concerns. The FFAs were intended to: (1) provide an alternative placement in a family setting to the more expensive group homes; (2) increase the availability of foster care placement resources; and (3) provide an enhanced level of service to foster children and families.

Rapid Growth of Foster Family Agencies. Since the creation of FFAs as an alternative placement: (1) the number of children placed in FFAs increased almost twentyfold between 1988 and 2000from 2 percent to 36 percent of first time foster care placements; (2) the proportion of first time FFH placements declined significantly during the same time period, from 87percent to 55 percent of first time placements; and (3) the proportion of children placed in group homes remained relatively stable. We note that these trends are statewide and, therefore, vary among counties.

While the total statewide foster care caseload has declined in the past two years, FFA placements have continued to increase. According to counties, the growth in FFA placements is due to a shortage of nonrelative FFHs, not increased needs of the youth being served. In addition, the longer FFA length of stay, discussed below, has also contributed to the continued growth of FFA placements.

Foster Care Length of Stay

Youth Stay Longer in FFA Homes. For several years, FFAs have been the focus of concern regarding children's length of stay in foster care. This is because the longer a child is in foster care, the less likely it is that he or she will be reunited with his or her family of origin, placed permanently with relative caregivers, or adopted. In a report submitted to the Legislature by DSS in 1997, the department concluded that FFAs were "holding onto the children much longer than an average foster family home placement and there is a resulting loss of time for a child to be permanently placed, if at all." In our Analysis of the 2001-02 Budget Bill, we examined foster care length of stay and concluded that children stay in FFAs twice as long as children in nonrelative FFHs. The median length of stay for children in FFAs was two years, while the median length of stay for children in FFHs was one year.

In response to our earlier analyses, advocates and providers have suggested that county social work practices may contribute to the longer FFA length of stay. They argue that because FFA children have their "own" social workers (associated with the FFA), these children may receive less attention from county social workers. Less contact with a county social worker could result in a longer stay in foster care, as the county social worker's assessment is a key factor in returning a child to his or her home or placing the child for adoption. When we asked county child welfare administrators about this, they indicated that while this phenomenon may sometimes occur, it was not widespread. Therefore, while we believe county social work practice could influence FFA length of stay, we do not believe that it explains a significant portion of the variation between the time in foster care for FFH and FFA children.

Extended Length of FFA Stay Persists in New Analysis. This year, we analyzed data that include only children who were in foster care 30 days or longer, thereby excluding children who were in an emergency placement. Although the length of stay for FFA children is still greater than for children in FFHs, the difference is not as great as in our earlier analysis. Those children for whom a FFA was their primary placement stayed in foster care almost two years, or 25 percent longer than youth in nonrelative FFHs, who were in care for just under a year and a half. As discussed above, increased time in foster care is generally considered undesirable, as children are less likely to be reunified with their family of origin or adopted.

Do Child or Family Characteristics Explain Length of Stay Differences?

New Research Conducted. Recent research from the University of California (UC) at Davis examined similarities and differences between FFH and FFA children, families, and services. The study included (1) several hundred phone interviews with foster parents, adolescent foster youth, and local program staff; (2) focus groups with each of these cohorts; and (3) a case file review of several hundred foster youth. Generally, the research findings indicate that the children in each placement type are similar to one another. Below, we discuss several findings of the new research.

Foster Youth Similar With Respect to Mental Health Needs. Longer stays in FFA homes might be justified if research indicated that the children in FFAs need more services before returning home or being adopted than do children in FFHs. The UC Davis study examined whether the children in the two placement types differed with respect to "special needs"--including psychological, physical, medical, and other problems. Notably, the two groups of foster children were actually quite similar with respect to foster parent reporting of psychological and abuse-related problems. This finding supports earlier findings that the special needs of children do not explain the increase in the number of FFA placements or the extended FFA length of stay.

FFH Youth Face Greater Physical and Medical Challenges. While the groups of children were similar in terms of mental health needs, they did differ in two other special needs areas. The UC Davis research suggests that FFH children face greater medical and physical challenges than their FFA counterparts. According to the reports of several hundred foster parents surveyed, the FFH children had more medical and physical problems than FFA children did. Medical problems included asthma, drug exposure, and HIV. Physical problems included motor, vision, and hearing disabilities. These findings, paired with the lack of differentiation in mental health needs, run contrary to the notion that children in FFAs face greater challenges. This research indicates, instead, that FFH children may actually have more special needs, at least in terms of physical and medical issues, than FFA children.

Parental Backgrounds Suggest That FFA Children Should Return Home at a Higher Rate. Child reunification with the biological family is one of the preferred permanency outcomes for children in foster care. Longer stays in foster care might be justified if research indicated that the biological families of children in FFAs face more challenges that result in delayed reunification. However, research indicates that the opposite is true.

The UC Davis research shows that children in the FFA homes have more stable family backgrounds-- suggesting a greater likelihood of reunification--than children in FFHs. For example, the biological parent(s) of FFA children were almost twice as likely to (1) have housing and employment and (2) be "law-abiding." Sibling bonding was also stronger for youth in FFA placements. Finally, biological parents of children in FFA placements visited their children in foster care more often than the parents of FFH children.

Children in FFHs Achieve More Stability in Living Situations. The UC Davis researchers surveyed foster parents about outcomes such as reunification, adoption, guardianship, or subsequent foster placement for the two groups of children. The study indicated that FFH children achieve more stability in their living arrangements, regardless of whether they return home or remain in foster care to be adopted later by foster parents.

First, FFH children were more likely to reunify with their families, while FFA children were more likely to move to another foster placement. Second, for those FFH children who were not reunified with their families, their foster parents were approximately twice as likely to pursue adoption than FFA foster parents. Finally, of those foster parents who pursued adoption or guardianship, FFH parents were twice as successful in finalizing adoption or guardianship than the FFA parents.

Conclusion

Based on our review of research and data on FFA placements, we conclude (1) children stay longer in FFAs than FFHs; (2) neither child nor biological family differences explain the longer stays; and (3) youth in FFHs have more positive permanency outcomes than their FFA counterparts. Finally, we would note that while the outcomes for children in FAAs are less positive than for children in FFHs, the costs for FFA placements are more than twice that of FFH placements. Accordingly, we believe that limiting the use of FFA placements could both improve foster youth permanence and result in savings. Below, we outline three options for the Legislature to consider for the budget year.

Earlier, we acknowledged the possible role that FFA and county social work practices may play in prolonging children's length of time in FFAs. According to the UC Davis research, there is little standardization across the state in FFA and county cooperation regarding case planning. Therefore, unclear responsibilities or poor communication between county and FFA social workers may result in children remaining in foster care longer than is necessary. In addition to considering the options listed be low, we suggest the Legislature clarify the roles and responsibilities of the respective county and FFA partners in expediting children's movement toward more permanent living situations.

Options for Reducing FFA Placements and Costs

While we conclude that there are significant problems associated with the use of FFA placements, such as extended length of stay, decreased permanency for children, and increased costs, we believe that FFAs do meet a need in California's continuum of foster care placements. In other words, FFAs provide an enhanced level of service relative to FFH placements and FFAs provide a less expensive placement option than more expensive group homes. However, we conclude that FFAs are relied on too heavily for children who do not require this enhanced level of service. Below, we present three options for reducing FFA placements and costs.

Hold Funding to Current-Year Levels. The first option for the Legislature to consider would be to limit the growth of FFA placements by holding funding to current-year levels. Halting projected year-over growth in FFA placements in this manner would result in an estimated $1.1 million in General Fund savings in the budget year.

Decrease the Number of FFA "Treatment" Placements. Alternatively, the Legislature could reduce the statewide number of funded FFA "treatment" placements from the Governor's budget. The FFA treatment placements are those for which approximately $1,000 per month per child is paid to the agency for treatment and administration. As several studies (including the most recent by UC Davis) indicate, many children placed in FFAs may not need these enhanced services. Therefore, we suggest the use of "nontreatment" rates, which currently allow foster families to receive the same monthly grant to care for the child, but would eliminate the monthly per child rate paid to the FFA.

For example, if legislation were enacted to require each FFA to reduce its treatment placements by 20 percent statewide by January of 2003, approximately 3,000 fewer of these placements would be funded. This reduction, midway through the budget year, would result in savings of approximately $4 million General Fund. In 2003-04, these savings would grow to approximately $8 million General Fund, relative to the 2002-03 Governor's budget.

Adjust FFA Treatment Rates. A final option, which we discussed in our Analysis of the 2001-02 Budget Act, is to adjust the FFA rates over time to encourage the movement of children toward reunification or adoption. While the rate paid to the FFA foster family would remain the same over time, the portion of the rate paid to the FFA organization for services and administration would decrease the longer a child remained in care. For example, the monthly services and administration component per child could be reduced by one-quarter (between approximately $240 and $270), incrementally, after each six-month period.

Figure 2 shows an example of this incremental reduction in the treatment rate. Under this example, treatment and administrative costs would be funded at the full rate for the first six months a child is in placement. The funding would continue, at a reduced rate, for up to two years while a child remains in care. A similar step down of the treatment and administration component would be applied to all of the age-adjusted rates. (We note that many of the youth in FFAs are either reunified with their family of origin or adopted before two years has passed.)

This tapering of the treatment and administration component of the rates could create an incentive system by encouraging FFAs to move children toward reunification or adoption more quickly. However, a decrease in rates could reduce the number of participating FFAs. This option would result in approximately $5 million General Fund savings in 2002-03. We note that savings would increase to at least $15 million General Fund in 2003-04.

Analyst's Recommendation

Above, we outlined three options to address concerns related to FFA length of stay, child outcomes, and costs. These options are: (1) holding finding to current-year levels; (2) decreasing the statewide number of FFA treatment placements; and (3) adjusting the treatment rates over time. While each of these options has merit, we believe that adjusting the FFA treatment rates over time may be the best option. This option should correct fiscal incentives, encourage the movement of children toward permanency, and control costs, while providing enhanced services to children and families for up to two years.

Assessing California's Foster Care Program Performance

New federal performance reviews of state child welfare services and foster care programs will be conducted in California for the first time in the fall of 2002. These reviews will rely, in part, on quantitative measures of state performance. Preliminary analysis indicates that California may fail to meet national standards on a number of these measures. Such failure could result in the loss of federal funding. We (1) describe the new federal review process; (2) examine California's performance on selected measures; and (3) make recommendations to improve California's performance.

Federal Government Shifts to New Review System

The federal Adoption and Safe Families Act (AFSA) of 1997 made the most sweeping changes to state child welfare services (CWS) and foster care programs since 1980. The principles of AFSA were to achieve child safety, permanency, and well-being. One significant requirement was that the federal Department of Health and Human Services develop a set of outcome measures and overhaul state performance review processes in the CWS and foster care programs.

The new Child and Family Service Reviews, resulting from AFSA directives, are a departure from prior federal evaluations of states in several ways. These changes include: (1) a focus on outcomes for children and families; (2) the use of multiple quantitative and qualitative measures to evaluate outcomes and performance; and (3) joint federal and state review teams. The federal Child and Family Service Reviews began in 2001. Figure 3 shows the schedule of state Child and Family Service Reviews over the next several years.

Of the 15 states reviewed in 2001, all have had to submit a performance improvement plan, indicating that none have "passed" all components evaluated during the reviews. California's review is scheduled for the fall of 2002.

Components of the Reviews

The federal review process has three major components: (1) the statewide data assessment, (2) county case file reviews, and (3) county interviews on seven "systemic" factors.

Statewide Assessment. For the statewide assessment, the federal government provides analysis of state data to indicate whether federal standards have been met. Each state then responds to the data review in narrative form to interpret federal findings on these quantitative measures.

County Case File Reviews. For the on-site component of the review, a team composed of federal and state representatives will review about 50 case files in each of three California counties to determine whether child safety, permanency, and well-being goals are being met. Los Angeles County will be reviewed in addition to two other counties. The two additional counties will be notified several months in advance of the on-site review.

County Interviews on Seven Factors. In addition to the initial case reviews, the review team will also interview children, staff, providers, and other stakeholders to examine seven systemic factors that impact the quality of services provided to children and families. These factors are the (1) statewide automation system; (2) array of services available; (3) case review system; (4) staff training; (5) quality assurance system; (6) agency responsiveness to the community; and (7) foster and adoptive parent recruitment, licensure, and retention.

How Performance Is Measured On the Statewide Assessment

In General, What Are Outcomes and Their Indicators? Outcomes and their indicators aid in measuring program performance by providing a comparison of program results consistently and over time. In the context of child welfare, outcomes are a condition of children expressed as a goal. For example, "children are safe;" "children are living in stable living situations;" and "children are succeeding in school" are outcomes. An outcome measure, or indicator helps quantify the degree to which a desired outcome has been reached. An indicator is expressed using data that measure a specific condition.

Consider, for example, the outcome "children are safe." One indicator is the number of substantiated child abuse or neglect referrals per 1,000 children in a particular community, county, or state. Local variation in this measure could indicate underlying variation in local policies or demographics, or changes in family factors associated with child abuse/neglect, such as substance abuse or poverty.

What Are the New Federal Outcomes and Their Indicators? The goals (articulated in the Adoption and Safe Families Act) of child safety, permanency, and well-being provided the framework for the development of the federal outcome measures to be used in the Child and Family Service Reviews. The goal of safety is defined as the protection of children from abuse or neglect in their own homes or foster care. The goal of permanency is defined as children having stable and consistent living situations (such as living with their families of origin, adoptive families, or legal guardians). The goal of well-being is defined as children receiving education and physical and mental health services adequate to meet their needs.

The federal government identified six outcome measures ("indicators") in order to determine the extent to which states are meeting the goals of AFSA. Figure 4 shows the six outcome measures to be used in the Child and Family Service Reviews. The figure also identifies the related national standards which a state must meet in order to pass. These standards, shown in Figure 4, were set by the federal government based on the performance of the top 25 percent of the 50 states during federal fiscal year 1998.

These federal measures reflect several key aspects of child safety and permanency by providing information on the recurrence of child abuse and/or neglect; the rate of reunification and adoption; and placement

stability. For California's first review in 2002, federal fiscal year 2000 data will be used to judge performance on the six indicators. We note that these six indicators do not address the outcome of child well-being, which will be measured using qualitative measures elsewhere in the federal reviews.

Because the 2000 data are not yet available, we cannot predict performance. However, historic trends indicate that California's performance are not likely to vary widely from the 1998 data we present later.

What if California Doesn't Meet Federal Review Standards?

If California fails to meet the national standards on the six indicators (discussed above) or the seven systemic factors, it could face fiscal penal ties in future years. These penalties would be levied as a percentage of federal CWS (IV-B) funds and a portion of foster care (IV-E) administrative funds. For each of the six indicators and each of the seven additional systemic factors that California fails to meet, the state could initially lose 1 percent per year of this federal funding.

According to DSS, the penalties in the first year could range from approximately $400,000 to over $5 million, depending on the number of indicators and systemic factors the state fails to meet. The penalty would increase to 2 percent per measure per year following failure during a second review and 3 percent per measure per year following failure of a third review.

In order to avoid penalties, states have the opportunity to design a "performance improvement plan" (PIP) which outlines how the state intends to reach national standards on the measures it failed. If the state and federal governments agree on the PIP, penalties are suspended until the next scheduled review, which would follow two years later.

How Would the State Fare Under Selected Measures?

Below, we examine California's performance, relative to other large states, on five of the six federal indicators that will be used in the upcoming federal review. Data for the final measure, the frequency of abuse or neglect that occurs in foster care, are not currently available for many states, including California. Therefore, we do not discuss this measure. The data used for the five measures discussed are generally drawn from the federal fiscal year 1998, because that is what has been published by the federal Department of Health and Human Services. As noted earlier, for California's first review in 2002, federal fiscal year 2000 data will be used to judge performance on the six indicators. Thus, given the two-year difference, California's performance could be different than the 1998 data we present below.

Child Abuse/Neglect Recidivism High in California. One measure of performance of the CWS and Foster Care Programs is the extent to which children served by the programs later return to these programs. Two of the federal measures are intended to capture this aspect of child safety. The first, the maltreatment recurrence indicator, measures the percent of children who have been reabused/neglected within six months of an earlier abuse/neglect incident. Included in this measure are children who were reabused after returning home from foster care and children who remained in their home following substantiation of earlier abuse/neglect. The second, the foster care reentry indicator, measures the percent of children who enter foster care more than once within a 12 month period. 

For each measure, a lower percentage is more desirable. The national standard for the maltreatment recurrence indicator is 6 percent. The national standard on the foster care reentry indicator is almost 9 percent. Figures 5 and 6 show large states' performance on these two measures.

As regards the reoccurrence of abuse, California fails the national standard with 11 percent of children having a second substantiated abuse/neglect report within six months. New York and Illinois also exceed the federal standard among large states. As regards reentry into foster care, California also exceeds the federal standard, with 14 percent of children reentering foster care within 12 months of a prior entry.

California Fails Federal Standard on Timely Child Reunification. Federal and state policies generally view foster care as a temporary, rather than a long-term, solution when children are removed from an abusive or neglectful home. Generally, the goal is to reunify the child and family as soon as is reasonably possible. This indicator measures, of those foster care children who were reuniting with their family of origin, the proportion who did so within 12 months of entering the foster care system. A higher percentage indicates more satisfactory performance than a lower percentage on this measure. The national standard for this measure is 76 percent. Figure 7 shows large states' performance on the child reunification indicator.

All the states shown in Figure 7 failed to meet the national standard. In California, only 55 percent of children who exited foster care to reunite with their families did so within 12 months of entering foster care. This performance places the state about 20 percentage points below the national standard for this measure. California ranks fifth among the eight states for which data were available.

California Provides Placement Stability for Most Foster Children. The next measure is the child placement stability indicator, which measures whether children have stable living arrangements. This indicator measures what proportion of children remained in either one or two foster care placements during the 12 months under review. A higher percentage indicates more satisfactory performance than a lower percentage on this measure. The national standard for this measure is about 87 percent. Figure 8 shows large states' performance on this measure.

Georgia, with 90 percent of foster youth having two or fewer placements in 12 months, is the only state in the figure that surpasses the national standard, although Pennsylvania and New Jersey are each only one percentage point from the standard. California, too, is close to the standard, with 85 percent of foster youth having two or fewer placements during the time period.

California Moves Children to Adoption in a Timely Manner. Another federal measure related to the permanency outcome is that children are adopted in a timely manner. A higher adoption percentage indicates more satisfactory performance than a lower percentage on this measure. The national standard for the adoption indicator is that 32 percent of children who are available for adoption are adopted within 24 months of placement in foster care. Figure 9 shows California's performance on the federal adoption indicator relative to other large states and the national standard. The graph shows that California, along with Texas and Florida, meets the national standard. This is the only federal measure, of the five we examine in this review, for which California meets or exceeds the national standard.

California Struggles With Federal Performance Measures

Of the five quantitative measures we examined above, California met or exceeded federal standards on only one measure. Although the 2002 reviews will rely on more recent state data, we have no basis upon which to expect that California's performance will vary widely from the measures reported here.

On the timely adoption measure (Figure 9), California exceeded the federal standard. In addition, the state has received national recognition for the number of adoptions finalized in recent years. These results are impressive, but not surprising, since the state has funded both one-time and ongoing initiatives to increase the number and timeliness of adoptions. In addition, a number of laws have been enacted that encourage adoption as an alternative to foster care.

On other measures, however, state performance is lackluster at best. For example, the state fails to return an adequate number of children home within 12 months of their entering foster care (Figure 7), which means children may remain in foster care longer than is necessary. When children return home from foster care, child reabuse/neglect (Figure 5) oc curs far more than is acceptable under federal standards, resulting in high rates of foster care reentry (Figure 6).

How Can the Legislature Aid in Improving Program Performance?

California faces significant challenges in attempting to improve the CWS and Foster Care Programs. There are potentially a number of strategies that can be employed to improve the state's performance. No single approach by itself is likely to result in significant improvement. At this time, we have identified several areas that may benefit from legislative focus: (1) state oversight of county programs; (2) decision making on child safety; (3) the role of early childhood interventions; and (4) replication of effective programs.

Capitalize on Improved State Oversight of County Programs. In the past, state reform efforts may have been limited in part by a lack of useful information about the progress of children and families through the CWS and Foster Care Programs. The federal Child and Family Service Reviews should provide new and useful statewide information to the Legislature and administration. However, because the on-site portion of the reviews will occur in only three counties, the reviews may fail to identify programmatic variation across the rest of the state.

Recently, Chapter 678, Statutes of 2001 (AB 636, Steinberg) was signed into law. This law, also known as the Child Welfare System Improvement and Accountability Act of 2001, requires DSS to (1) overhaul the county review process; (2) use the federal measures, discussed above, to measure individual county performance; and (3) choose additional measures to evaluate county performance. Ideally, the information that results from the Chapter 678 review process will provide new opportunities for legislative oversight. As these reviews are being designed, the Legislature should consider what kinds of information would be most useful to future initiatives and ensure that this information is included. As the reviews are implemented, the Legislature should monitor the review process to ensure that information needs are being met in a timely manner.

Improve Decision Making. Another way for California to improve its CWS and foster care program performance is to improve decision making regarding child safety. Many decisions, from whether to investigate child abuse/neglect to whether a child is a likely candidate for adoption, all influence California's performance on the federal measures examined above.

In our Analysis of the 2001-02 Budget Bill (please see page C-223), we examined Structured Decision Making (SDM), one approach to risk assessment that improves the consistency and accuracy of child welfare decisions. The SDM approach is a series of tools that help workers deter mine: (1) when to investigate abuse/maltreatment decisions; (2) the degree of child safety at the time of an investigation; (3) the risk of future child maltreatment; (4) the targeted services to be provided to families at the highest risk of reabuse; and (5) whether to remove a child to foster care.

Evaluations from other states such as Michigan and Wisconsin have concluded that SDM has significant value in predicting the likelihood of future abuse or neglect and that it improves child welfare outcomes. Currently, about one-quarter of California counties--including some large counties--use some or all of the SDM tools. Expanding SDM to all counties, including a significant number that remain on a waiting list, could improve the statewide accuracy and consistency of decision making. Such improvements could increase child safety; permanency; and well-being, as well as provide resources to those children and families most in need of services.

Strengthen the Role of Targeted Early Intervention Programs. Increasing the role of effective early childhood program interventions is another area that may improve CWS performance, by decreasing or preventing child abuse/neglect. There is evidence that targeted early childhood intervention programs may be effective in preventing or decreasing child abuse. For example, targeted home visiting programs like the Elmira, New York Prenatal/Early Infancy Project have shown positive results. Under this program, nurses trained in parent education made 32 home visits over a three-year period to new mothers. The nurses provided the mothers with information on available support and services. Studies of the program indicated that it resulted in a variety of short-term and long-term health and social benefits, including fewer reports of child abuse and neglect. The program was particularly effective for high-risk families (single mothers with low socioeconomic status). When targeted to high-risk families, research concluded that the home visiting program resulted in $18,500 in net savings to government per family. (For further discussion of such cost-effective programs, please see our report, Proposition 10: How Does it Work? What Role Should the Legislature Play in Its Implementation?, January 1999.)

We note that current funding streams, particularly those from the federal government, may encourage the placement of children in foster care, rather than encourage the use of effective early childhood programs. One fiscal resource that California has for funding early childhood intervention programs is the California Children and Families Program. Through this program, revenue generated by a per package surtax on cigarettes, pursuant to Proposition 10, will provide counties with over $650 million in 2002-03 to support local programs serving children through the age of five. This represents a source of potential funding for early childhood intervention programs.

Given the availability of these funds, the Legislature could consider establishing a state-funded voluntary matching program, which would fund (1) early childhood programs that have been shown to be cost-effective and/or (2) demonstration programs that are potentially cost-effective, based on existing research. (For further detail on a legislative matching program, please see our 2000-01 Analysis of the Budget Bill, page C-54.)

Facilitate Identification and Furtherance of Effective Programs. California's CWS and Foster Care Programs are administered by 58 counties. This results in a wide variety of programmatic approaches and makes it difficult to judge which programs are most effective. We believe that DSS could improve CWS and foster care program performance by facilitating expansion of effective programs once they are identified. The DSS could accomplish this by increasing information sharing, and validating county "best practices."

For example, several California counties are currently using the Family Group Decision Making (FGDM) model as an approach to reducing child abuse and neglect. In the child welfare context, FGDM offers children, families, and child welfare workers an opportunity to make decisions and develop plans together that are intended to protect children from future abuse and neglect. This model, in contrast to other more adversarial approaches, recognizes that families have the most information about resources at their disposal and how child safety can be made a priority--thereby preventing further child abuse or neglect.

The Santa Clara County FGDM program has been studied by an independent evaluator for over three years and early findings indicate favorable outcomes: (1) a decrease in the number of children living in foster care; (2) an increase in the number of children living with kin; and (3) a decrease in the number of court proceedings. If these positive outcomes are also found in the final report, which is due in the next few months, we believe DSS could provide all counties information on how to effectively implement FGDM.

The FGDM model is just one example of an innovative program that is currently used by some California counties and shows promising research results. We believe there are other strategies and practices that, if expanded, could improve statewide program performance. Such expansions could occur within the resources currently available to counties, especially if such programs result in offsetting savings elsewhere in the continuum of services available to children and families.

Conclusion

Our review indicates the need for improvement in California's CWS and Foster Care Programs. However, improving performance will not be an easy task. Given the competing demands for resources, it is important that available funding--whether new or existing--be used effectively. We believe that program performance can be improved by using better decision making tools (such as SDM); strengthening the targeted early intervention programs; and identifying and sharing "best practices" such as FGDM.