The Future of Children, Princeton - Brookings: Providing research and analysis to promote effective policies and programs for children.

There are at least six critical issues that must be addressed more fully and consistently as school-linked services are planned and implemented. They are: systemic change in the schools and child-serving agencies, targeting, financing, evaluation, state and federal leadership, and alternatives to the school-linked service approach.

As discussed in the preceding section, one emerging criterion for these efforts is that participating schools and other agencies change how they provide services to children and families and how they interact with each other. For school-linked services to be effective, proponents believe that these agencies' daily processes must be altered to facilitate the collaboration. Although this general proposition is often stated, the specifics of the change required are not typically discussed. We believe that this assertion requires more clarification as the school-linked service approach is further implemented.

The above criteria and the Levy and Shepardson article in this journal issue suggest a beginning definition of the necessary change. For effective school-linked services, the schools and the other participating agencies will need to create an ongoing process to identify common goals and to plan the implementation and evaluation of their efforts. Schools and agencies will need to redirect their existing funding streams and to increase the authority of personnel who work directly with children and families to allow personnel to make decisions about services without having to wait for district or county approval. They will need to establish internal procedures by which their staff can identify problems in children and follow up with staff from the other agencies. Within each institution these needs will require changes in the roles and responsibilities of selected personnel. Teachers, health providers, and social workers will need the time, training, and authority to participate in the school-linked service collaboration. New standards of accountability will need to be developed for every level of staff in the participating agencies. The new standards must emphasize the achievement of positive outcomes rather than dictate specific inputs.

These and other changes necessary for school-linked services may be encouraged as part of broader reform movements within each of the education, health, and social services sectors. As one report about school-linked services states: "[The] convergence of reform in all of the people-serving systems presents a moment of unique opportunity in which to pursue a collaborative agenda."²⁵

Proponents of school-linked services should be in dialogue with those involved in broader reform of the education, health, and social services systems. Some of their agendas may well be compatible with school-linked services. For example, in education reform, one element of restructuring has been for school districts to give more decision-making authority and discretion to local schools and their staffs.²⁶ In the health field, there has been increased interest in creating health centers that can link patients with other services they need.²⁷ In the child welfare field, there is more discussion of focusing on the needs of the child in the context of the family and emphasizing preventive services.²⁸ Each of these developments is consistent with the goals and strategy of school-linked services.

The effect of other reform efforts on school-linked services may not be so clear. For example, the introduction of choice and/or a school voucher system may affect the schools' ability to participate with other agencies and may also affect their interest in collaborating to meet the social and health needs of students (see the discussion by Tyack in this journal issue). Making students' academic outcomes a more significant part of teachers' evaluations could also impact the attention teachers pay to school-linked health and social services.

In sum, there is a need for more precision in defining how schools and other agencies must reorganize to implement school-linked services. It is also important that these systemic changes are considered in relationship to broader reform efforts.

One of the most important questions for determining what services should be offered, how they should be governed and evaluated, and what outcomes should be sought is whether a school-linked service effort is designed to serve all children at a school or in a community or whether it is targeted to a particular population of children with specific, identified needs. To date, most efforts target services—either by selecting only those schools where a high percentage of the student body is from low-income families or by providing services at a school only to those students with particular needs.

The preference for targeting is understandable. In fact, limited funds may make it essential. Also, for both moral and practical reasons, it makes sense that the children with the most severe needs receive first priority. To make the system available to these children, there must be special outreach to them.

There is, however, some danger to targeting. The special attention focused on multiproblem students can be stigmatizing. Also, when a program focuses only on those most in need, there is increased risk that the program will begin treating the children and families served as somehow deficient; the program may emphasize only the clients' weaknesses and not their strengths. Finally, targeting may undermine a basic goal of prevention: to make services available to children and their families before a crisis develops or a problem becomes acute.

We believe a better approach is to ensure some level of universal access to school-linked services to all children, even if special emphasis is directed toward high-risk students. In this regard the New Beginnings program in San Diego offers guidance. The program assesses every family when the son or daughter enters school. Some students are identified as in immediate need of targeted services. All others, however, can still use the school-linked service center through self-referral. This approach carries the advantages of both a universal and a targeted system. The fact that everyone can receive some level of benefit may produce broader community buy-in and political support.

The question of targeting is a difficult one, but it needs to be explicitly addressed in any new endeavor, with full consideration of the options and their consequences.

The financing of school-linked services is the area that most critically needs attention before school-linked services can be broadly implemented. As Farrow and Joe write in this journal issue: "[Many] of the issues that surround the financing of school-linked services are really issues of priorities, authority, and control over resources. For this reason, financing strategies require careful thought and, in the long run, can greatly affect whether these new services succeed or fail."

Farrow and Joe recommend a fiscal strategy that includes some new core funding for the additional staff and services necessary for the collaboration; redeployment of existing funds, including relaxing categorical requirements for these funds; and maximum use of the federal Medicaid entitlement funds. Apart from the handful of efforts at decategorization and redirection discussed in the articles by Farrow and Joe, by Levy and Shepardson, and by Jehl and Kirst, most current efforts have not made significant progress in developing such a comprehensive strategy.

The difficulty of this task cannot be overestimated. Today's school-linked service programs are only minuscule experiments in the massive health, education, and social services systems that have been built up over the last several decades. Morrill's article in this journal issue cites an estimate that $278.4 billion is spent annually on children through the education, health, and social services systems. To change such large systems, or even a portion of them, to allow more flexible funding for school-linked services is an extremely difficult task—one that most of the current efforts at school-linked services have failed to accomplish.

On the other hand, the massive nature of the system also offers incentive. If proponents of school-linked services are correct and collaborative, integrated services do produce better outcomes for children, the potential gains due to increased efficiency and effectiveness from large-scale reorganization of these systems could be very great.

Evaluation of current school-linked service efforts has been sporadic both in type and in quality. Although some consensus is emerging that integrated service efforts generally—and school-linked services specifically—should be evaluated both for process and outcomes, this has rarely happened (see the Gomby and Larson article in this journal issue for a discussion of evaluation strategies). Many of the most recent efforts have collected some information about what services were provided and for whom, but few have gone beyond this to learn more from the process and to determine whether the efforts made any difference.

For example, little is known to date about the optimum governance structure or organization of services for successful implementation. Comparative process evaluations may assist in answering these questions. In such evaluations, activity logs, staff time sheets, case records, interagency memoranda of understanding, logs indicating requests for services, and interviews with staff and program participants could be used to help trace how students and families interact with the new systems that have been established.

Perhaps the greatest need in evaluation, however, is the development and implementation of appropriate procedures for assessing outcomes. There should be more rigor in evaluations of school-linked services, including more use of closely matched comparison groups and/or random assignment to create genuine control groups. There needs to be more attention to behavioral outcomes and to changes in knowledge and attitudes. Researchers and others must develop common measures and approaches for measuring outcomes in a way that allows comparison among sites.

Evaluation is all the more important because school-linked service efforts are still in a stage of experimentation. We need to learn as much as we can from these initiatives. Toward this end, every school-linked service effort, no matter the size, should engage in some level of evaluation. For some programs, this might consist simply of a clear statement about what changes in behavior, knowledge, or attitudes the program is designed to affect and the collection of data and other information to measure these targeted outcomes. Other efforts will be able to incorporate some elements of experimental design to look more closely at the issue of causation (see the discussion in the Gomby and Larson article).

Large state or federal multisite initiatives should include well-designed outcome evaluations. Funding these evaluations will require a new appreciation of and commitment to evaluation by both legislators and private funders such as foundations. The investment in both the initiatives and their evaluations must be undertaken with a long-term view.

The fifth issue that emerges from many of the articles in this journal is the need for more state and federal leadership in the development and testing of school-linked service efforts. Too many critical issues simply cannot be resolved without greater participation at these levels of government. To date, even in those states where there are major initiatives under way, there is a tendency for the state to narrowly limit its involvement in scope, financial commitment, and time.

The Gerry and Certo article in this journal issue expresses the federal government's desire to assist state and local community-based efforts to integrate services for children and families. To fulfill this intent, federal leadership will be necessary to change federal eligibility requirements, funding restrictions, and program goals. More than technical assistance is needed if the federal government is to exercise leadership.

Furthermore, the state and federal governments can do much to alleviate some of the confidentiality requirements that block the sharing of information among agencies involved with school-linked services (see Appendix A). Finally, only the states and the federal government can fund some of the large-scale evaluations that are necessary to test this approach.

Although a few states have shown leadership in one or more of these tasks, most have not. And although the federal government has expressed considerable interest and funded some discrete projects, its efforts have been limited. The movement toward school-linked services will not be successful, or even successfully evaluated, without a decision at these levels of government to make a deeper commitment to testing this approach.

Finally, it is critical to compare the school-linked service approach with others. As the article by Chaskin and Richman in this journal issue discusses, this approach will not be effective if the school seeks to dominate or control the planning and governance of services. Furthermore, in some communities, a significant number of residents may view the school as antagonistic to their interests; the young people most in need of services may be for the most part out of school; or busing or the use of magnet schools may mean that students do not live near their schools. Some of these barriers can be overcome through special outreach programs and through flexibility as to where services are provided. But in some communities, political and historical barriers may make the school-linked approach unworkable. Service collaborations that are planned and governed by the broader community or neighborhood may be more successful in developing a service system based on the needs of children and their families rather than the needs of an established institution like the school. In their article in this journal issue, Chaskin and Richman briefly discuss some of these alternative, community-based approaches.

In addition to these concerns about the effectiveness of schools as a central focus for integrated services, there is also concern about their appropriateness. Some believe that schools will better meet academic objectives if they are not also asked to be primarily responsible for students' nonacademic needs.²⁹ This caution, however, is consistent with the contention of many supporters of school-linked services that schools should only attempt to address the health and social services needs of children if they have the resources, expertise, and collaboration of other community agencies.

Given such concerns about school-linked services, experimentation with alternative approaches for integrating services for children and families is appropriate.³⁰ The implementation of effective school-linked services will be facilitated if planners know in what circumstances this approach would have the best chances for working.