Family involvement interventions in child protection: learning from contextual integrated strategies.

The use of family group conferencing and related family involvement interventions in child protection is rapidly increasing in the United States and many other countries. There is some concern that the child welfare field will travel down the same road as it did with intensive family preservation services; that is, tremendous enthusiasm later derailed by rigidly designed evaluations that showed unimpressive effects. The work of John Braithwaite suggests an alternative path for finding justifiable excitement about these interventions. Drawing upon Braithwaite's writings and ongoing evaluation research, this article suggests afar steps we can take towards an integrative strategy for developing effective family involvement interventions.

Key words: group decision making, program evaluation, child welfare, child protection, family involvement, C.P.S. interventions

**********

The use of family group conferencing and related family involvement interventions in child protection is a rapidly growing practice around the world. For example, the number of communities in the United States trying Family Group Decision Making (FGDM) grew from five in 1995, to over one hundred by 2000; similarly, in 1994, four pilot programs began in England and Wales, and now fifty-five local authorities or nongovernmental groups have FGDM programs in those countries (Nixon, Merkel-Holguin, Sivak and Gunderson, 2001). Even as child welfare practitioners are eagerly implementing these programs, researchers are more cautious. For example, Whittaker asks: "While enthusiasm runs high, many questions remain: Will family group conferencing meet the ultimate test of empirical validation in rigorous studies with appropriate controls?" (Whittaker, 1999, p. xv). While clearly stating his preference for more rigorous clinical trials, Barth concedes that: "the assumptions of family group conferencing are so compelling that variations on this practice will undoubtedly continue to develop without evaluation endorsements" (Barth, 2002, p. 201). Gelles is more critical in saying: "This service is being widely touted as effective and widely adopted without a shred of scientifically reputable evidence that this intervention actually works. This is an echo of what happened with Intensive Family Preservation Services" (Shirk, 1999, p. 18). Many researchers would agree that the potential of Intensive Family Preservation Services was hurt by an early push for a specific family preservation services model, called Homebuilders, when there was no evidence (pro or con) to suggest that this specific program model was effective (Adams, 1994). Therefore, with family involvement interventions it may be prudent to more quickly involve evaluation in the development of the intervention. On the other hand, some FGDM proponents are wary of evaluation research: "Research has, for the most part, been done by someone, to someone else, to produce data that was used by yet someone else. It was experienced as having mystical importance and complexity but very little practical value at best and at worst was a tool to justify the continued oppression of others" (Nixon, Merkel-Holguin, Sivak and Gunderson, 2001, p. 29). What is needed is an approach to family involvement research that is consistent with the intervention's values of community and family empowerment. John Braithwaite's work on restorative justice may provide some theoretical concepts (Braithwaite, 2002a) and a method for developing theory (Braithwaite, 1993) that could be useful in current efforts to use family involvement interventions in child protection. In this article, I describe some key ideas from Braithwaite's work and then illustrate their applicability using evaluation research of family involvement programs. Family Group Decision Making and Team Decisionmaking are discussed under a rubric I call family involvement interventions. Both of these models focus on a plan for the care and protection of a child that is developed through a meeting of child welfare professionals and the child's extended family in cases of child abuse and neglect. I chose to discuss these models together because I believe they share many procedures and values with each other and with Braithwaite's restorative justice. However, I acknowledge that some proponents of Family Group Decision Making and Team Decisionmaking believe that these models are more different than they are similar. A person who is supportive of TDM and a person who is supportive of FGDM both read a previous draft of this article and both people questioned the validity of discussing these models together. My argument is that we should follow Braithwaite's contextual integrated strategy and explicitly identify differences and best practices in family involvement intervention both theoretically and empirically in order to determine the most effective ways to involve families in child protection. Comparing the models and discussing the differences helps us learn and improve our practice.

John Braithwaite and Restorative Justice

There is some concern in Family Group Decision Making that the practice has outrun the development of relevant theory (Burford and Hudson, 2000). One possible solution to this perceived problem is to look towards theory from related practices such as mediation and restorative justice. Following this strategy, there is considerable interest in the work of John Braithwaite within the Family Group Decision Making field. For example, he was asked to deliver the Closing Address at the 2002 Family Group Decision Making Roundtable held in Monterey, California (Braithwaite, 2002b). This special issue of the Journal of Sociology and Social Welfare further suggests that his work has broad appeal within social welfare.

Braithwaite has written extensively on both business regulation and criminology and recently brought these two areas of research together (2002a). In this latest work, Restorative Justice and Responsive Regulation, he suggests a theoretical approach to addressing a wide range of social problems and to improving democracy itself. Simply put, restorative justice is a process in which stakeholders come together to resolve a dispute. The specific organization of this process is less important than its core values, which include healing rather than hurting, moral learning, community participation and community caring, respectful dialogue, forgiveness, responsibility, apology, and making amends (Braithwaite, 2002a). The key point here is that the values emphasize restoring whatever was disrupted in the dispute. In the case of a crime, ideally the victim's sense of control is restored through an apology from the offender. However, Braithwaite insists that this does not mean restorative justice should require the offender to apologize. It is the values rather than the mechanisms which must be implemented. Similarly, responsive regulation is not a clearly defined program or a set of prescriptions concerning the best way to regulate. Braithwaite argues that the best regulatory strategy depends on context, regulatory culture, and history (Ayres and Braithwaite, 1992). Braithwaite illustrates his approach with a regulatory pyramid. At the base of the pyramid are those regulatory approaches that should be used first and most often. When these restorative approaches fail more confrontational means are necessary. Regulators should move from the means at the base of the pyramid up towards the point only when restorative processes are not working to elicit reform and repair.

[FIGURE 1 OMITTED]

Braithwaite then combines restorative justice and responsive regulation to suggest that, in a wide range of legal and policy concerns, we should begin with a restorative approach, and move to more confrontational practices only when restorative practices are not producing resolutions. For example, he applies this framework to world peacemaking.

Following arguments made by Desmond Tutu and others, Braithwaite suggests that securing peace in regional conflicts such as in the Middle East, South Africa, and the former Yugoslavia requires the use of restorative processes that convince people that their human rights are respected and that give them an opportunity to mourn and forgive (Braithwaite, 2002a). He cites the South African Truth and Reconciliation Commission as an example of bringing restorative practices to the grassroots level where they must be employed in order to promote sustainable peace. The pyramid illustrates that economic sanctions or military interventions should only be used when restorative processes have failed.

[FIGURE 2 OMITTED]

For those who are concerned that current family involvement interventions in child protection are under-theorized, Braithwaite's work provides both useful theory and a useful model for thinking about theory development and evaluation research. The next section describes Braithwaite's approach to evaluation research and its congruence with the evaluation of family involvement interventions. The paper concludes with a discussion of how to use Braithwaite's theory to advance the use of family involvement interventions.

The Role of Theory in Program Evaluation

Modern social program evaluation emerged during the Great Society of the 1960s, when there was a dramatic increase in social program spending and a corresponding demand for evaluation of these programs (Shadish, Cook and Leviton, 1995). This demand supported the development of evaluation as a profession. Social scientists who have taken on this work come from a wide variety of training and backgrounds, which has resulted in a lack of unified theories of how to evaluate programs. Early program evaluation guides focused on outcomes and quantitative analysis, while more recent ones have included process and qualitative methods. This shift has resulted in debates in the evaluation field about whether quantitative methods are superior to qualitative methods and whether positivist theories are superior to interpretivist theories. There are similar debates in the field of social work regarding which methods and epistemologies are best suited for social work inquiry (Allen-Meares and Lane, 1990; Ristock and Pennell, 1996). Lin (1998) suggests that a positivist approach seeks to identify those details with propositions that can be tested or identified in other cases, while an interpretivist approach seeks to combine those details into systems of belief whose manifestations are specific to a case. Lin further suggests that both approaches are useful in program evaluation because we need to demonstrate that policies have their desired effects with positivist analysis and demonstrate how they work using interpretivist analysis.

Braithwaite's theories are developed in a broad array of empirical investigations of occupational safety, nursing home regulation, consumer protection, and criminal justice. In a 1993 article, he describes the contextual integrated strategy he uses to contribute to better public policy in these numerous domains (Braithwaite, 1993). Braithwaite's approach to the positivist vs. interpretivist debate described above is eclectic (Braithwaite, 1993). He agrees with interpretivists who believe a theory developed in one context cannot be willy-nilly applied in another and that we are unlikely to discover universal theories of say, crime prevention. However, he also argues that theories developed in another context can be useful metaphorically for thinking about a new problem in a new context in different ways. He suggests that his approach to theory is similar to those of Allison (1971) and Morgan (1986). An example of Braithwaite's use of metaphoric theory is the pyramid described above.

While using theory as interprevitists do, Braithwaite still sees a role for positivistic quantitative methods in testing key claims of a theory: "one should definitely be discouraged in one's support of a particular element of an integrated long-term strategy if all the evaluation studies show that in the short term, this element never makes any difference" (Braithwaite, 1993, p 388). Or to put a positive spin on this, if one finds a statistically significant relationship between certain offender behaviors during a restorative justice conference (e.g., expressing remorse) and reductions in their future criminal behavior, this information is very helpful in improving the intervention and the theory behind it (Morris, 2002).

Finally, Braithwaite's contextual integrated strategy combines metaphoric theories and positivistic research through the engagement of key stakeholders from the context in which one is developing the intervention. He suggests we take the theories and the results of previous research and discuss them with the stakeholders from the community that is developing an intervention and learn from the community members the extent to which the theories and findings are relevant to their own community context. Then we develop a specific intervention strategy that is responsive to the stakeholder discussions of the theories, the previous research findings, and the specific community context. Lastly, the stakeholders would participate in the implementation of the intervention and the researcher's efforts to monitor the implementation.

Applying These Ideas to Family Involvement Interventions in Child Protection

Following the methodology described above, those aspiring to develop family involvement interventions for child protection would begin with Braithwaite's theory of restorative justice and responsive regulation. The restorative values he describes can be used to discuss the values that will drive the effort. Each community considering the use of family involvement interventions in child protection should review Braithwaite and his critics to think about which restorative justice values are applicable to child protection in their own community. For example, what is the role of apology, remorse, and shame in this process (Van Stokkom, 2002)? Family meetings are an emotional process and it is useful to discuss which emotions may emerge and how they can be used to facilitate child and family well being.

Braithwaite's pyramid can be used to think about the context of the intervention and under what circumstances it will be used. His pyramid also highlights the importance of not simply starting yet another new program. As Adams and Krauth point out: "There is a strong tendency in American human service systems for innovations such as family-based services to take the form of discrete packages of services produced for sale on the human services market. Innovative approaches to practice tend to become reduced to specific programs. Although there are attempts to reform whole systems toward family-based practice, the stronger tendency is for such work to be isolated in specialist units while the rest of the system continues largely unchanged" (1995, p. 92). Instead, family meetings can be used in a broad context of community- and neighborhood-based accountability. For example, Team Decisionmaking is one part of the Annie E. Casey Foundation's Family to Family Initiative which also includes strategies to recruit and support resource families and community partnerships as well as self evaluation procedures for determining how well the entire reform effort is working (DeMuro and Rideout, 2002).

By placing family meetings in the middle, rather than the base of the pyramid, I hope to clarify the role of family meetings in child protection. Prantis (2000) suggests that conferencing can strengthen or weaken community cohesiveness and sense of efficacy. Family meetings can be part of a broader effort to reinforce mutual accountability and responsibility or it can become another way professionals undermine community by creating dependence on formal services to solve community problems. Prantis is concerned that, for example, "community members have increasingly removed themselves from taking responsibility for the behavior of children and youth in public places" (p. 46) and that we need to ensure that family meetings reinforce rather than undermine a restoration of this sort of community accountability for children and families. Similarly, Braithwaite and Strang (2002) suggest that there should be synergy between public and private regulation and that we should simultaneously strengthen the regulatory capabilities of families, communities, non-governmental organizations and the state with respect to family violence. The American Humane Association's Front Porch Project is an example of an attempt to bring back this community response to child welfare (Wilmot, 2002). TDM is another example in that it is always implemented along with other Family to Family strategies that include strengthening community involvement.

Consistent with Braithwaite's pyramid metaphor, families and communities can move up and down the pyramid in Figure 3 as they demonstrate their capacity to care for children. Ideally, over time, professionals would be convening fewer meetings because the community would be organizing the meetings and other forms of support themselves. Family meetings would be convened when community accountability is not sufficient to keep a child safe. Initially, these family meetings could be used to keep children safely with their current caregivers (parents or guardians). Under more serious circumstances, family meetings can be used to place children with their extended family or possibly into formal foster care. In limited (ideally) circumstances, formal foster care services would be used to make placement decisions and provide services that could include the termination of parental rights and adoptive placement. Following Braithwaite's responsive regulation, the level of state intervention would depend on the ability of the families and communities to "regulate" themselves. Communities considering the implementation of family meetings can use this pyramid to discuss how these interventions could be useful in their own contexts.

[FIGURE 3 OMITTED]

Following Braithwaite' suggestions for positivistic inquiry, quantitative studies of other family involvement programs and ongoing analysis of the implementation of a new family involvement program can inform the development of the intervention. For example, many communities question the appropriateness of family involvement interventions in families that have a history of domestic violence or child sexual abuse. An evaluation of a FGDM program intervention in the eastern Canadian province of Newfoundland & Labrador suggests when and how family meetings can be used in these cases (Pennell and Burford, 2000). This is not to suggest that family meetings should always be used in cases of domestic violence or child sexual abuse, but the study may help communities consider whether they can make it work. Simple quantitative analysis can also be used to evaluate the implementation of a program. My own work includes the evaluation of a FGDM program that initially served African American, Asian American, Hispanic, and Native American families (Crampton, 2001). When the program was expanded to serve all families regardless of ethnicity, the level of program participation dropped. The FGDM staff suggested that Caucasian families were less willing to try FGDM than other families. A simple bivariate analysis confirmed that African American families were more willing to try FGDM than Caucasian families. However, in a multivariate analysis, race and ethnicity were no longer significant. Whether families had extended family members who were willing to participate was a better predictor of FGDM participation than race and ethnicity. This simple positivistic analysis helped the staff look for alternative explanations for the drop in program participation. An alternative theory suggested that the expansion of the population served without a corresponding increase in program resources, prevented the staff from fully exploring family resources and therefore limited the effectiveness of the program. Recent preliminary analysis of Team Decisionmaking in Cuyahoga County, Ohio (which includes Cleveland) suggests that children are more likely to be placed with relatives rather than in foster care when relatives attended the family meeting. While this finding may seem self evident, the results highlight for staff the importance of getting relatives to the meeting.

Braithwaite's approach to involving stakeholders in the evaluation of business regulation and criminal justice is consistent with many of the early evaluations of Family Group Decision Making. For example, the evaluators of the FGDM program in Newfoundland-Labrador described their approach as follows:

   The study used a collaborative action research approach which was
   viewed as congruent with its philosophy of forming partnerships.
   The study was designed: (a) collaboratively by involving a range of
   project participants as well as external consultants, (b)
   sequentially by drawing upon learning from earlier phases of the
   project, and (c) formatively by revamping procedures on the basis
   of feedback from participating families, community representatives,
   government officials and project staff (Burford and Pennell, 1995,
   p. 7).

Family involvement evaluators can combine quantitative and qualitative data and review it with program participants in this process to develop the program theory, test how well the program works under the theory, and then use the results to build support for resources for the program. Given the emphasis that Family Group Decision Making places on "widening the circle," it is not surprising that this approach to evaluation is often used in FGDM and Team Decisionmaking. Pennell, following her Canadian FGDM work is now taking these lessons to her North Carolina FGDM project and following the same process (Pennell and Weil, 2000). Communities that are using Team Decisionmaking as part of the Casey Foundation's Family to Family Initiative also have a self evaluation process that helps the key stakeholders focus on how Team Decisionmaking works and how it can improve outcomes for families (DeMuro and Rideout, 2002).

Next Steps

Given the extensive enthusiasm for Family Group Conferencing and related family involvement interventions, there will be some suggestions to begin rigorous clinical trials. I believe this would be a mistake. There is a need for additional work in developing the theory behind these interventions and understanding how they should be adapted in different contexts following the writings of Burford, Braithwaite, Pennell and others. After using the contextual integrated strategy described above, we can move towards randomized trials just as Braithwaite and his colleagues began randomized trials of restorative justice programs in Australia after they completed contextual integrated theory building research on these programs.

   Evaluating the Essential Elements of
   Family Involvement Interventions

A contextual integrated strategy for developing family involvement interventions should focus on evaluating the benefits of what are perceived to be the key elements of the intervention. Some Family Group Decision Making advocates have identified three essential elements of the practice: quality preparation time prior to the meeting, private family time during the meeting, and not prescribing the decision prior to the meeting (Mirsky, 2003). While there does not appear to be anything inherently incongruous in making these prescriptions, there is again an unfortunate parallel with the experience of Intensive Family Preservation Services. The Homebuilder's model of Intensive Family Preservation Services has a very explicit intensity (2 cases per caseworker) and duration (four to six weeks). Evaluations of these services suggest that this is not necessarily an appropriate programmatic design. For example, evaluators of the Illinois family preservation program concluded that the program design was unrealistic due to the difficulties with targeting and the severity of the issues faced by the some of families served (Schuerman, Rzepnicki, and Littell, 1994). Although families in the treatment group in this study received more intensive services compared to families in the comparison group, the short-term nature of the services often prevented the families from getting what they needed. The evaluators pointed out that while the staff was supposed to involve the extended family in case planning, in practice workers said they were reluctant to do so because it would heighten complex family dynamics that could not be managed with time-limited services (Schuerman, Rzepnicki, and Littell, 1994). Following Braithwaite's approach, this finding would not suggest that family preservation services should not be limited to six weeks, but it would suggest that stakeholders discussing the use of these services should consider the trade-off in potential benefit of intensive services vs. the time demands of involving extended family members in case planning. In a similar way, we should examine the research related to preparation time, private family time, and decision making authority in family involvement programs to see what lessons can be drawn and reviewed in places that are considering implementing similar programs. In the following sections, I review each of these three key practices using Braithwaite's strategy: what is the theory behind the practice? What does the research show about this practice? What is the perception of the stakeholders of this practice?

Preparation Time

FGDM advocates are understandably concerned that there will be attempts to start FGDM programs with insufficient resources, that these programs will then not produce the potential benefits of FGDM, and that therefore the reputation of FGDM will suffer. One way they attempt to ensure program integrity is to insist that FGDM must include quality preparation time, which is described as 20-25 hours on average (Mirsky, 2003). Preparation time is therefore a key distinction between these models: "Without thorough and intensive preconference planning, the FGC approach reflects more traditional case-planning methods" (Merkel-Holguin and Ribich, 2001, p. 203).

Research on FGDM suggests that preparation time is often extensive and that participants believe that preparation time is important for exploring family resources and beginning to change the relationships between family members and child welfare professionals (Marsh and Crow, 1998). To my knowledge, there is yet no research which demonstrates that preparation time produces these benefits or that preparation time improves outcomes for children and families. As stated above, my research showed that expanding the population served by an FGDM program without a corresponding increase in program resources, prevented the staff from fully exploring family resources and therefore limited the effectiveness of the program (Crampton, 2001). Clearly, program resources are important in making family meetings work. However, I did not find that preparation time itself was correlated with the primary goal of the program which was to divert children from foster care into kinship care (Crampton, 2001). In that study, preparation time was measured in terms of both the number of days from referral date to the meeting date, and the number of hours FGDM staff logged as preparation for the specific case between those dates. The average number of days was about eighteen and the logged preparation time averaged eight hours. Preparation time was slightly higher and number of days lower for meetings that developed a diversion plan, but these differences were not statistically significant. The Calgary FGDM pilot found that preparation time averaged seven hours (Sieppert, Hudson and Unrau, 2000). The Calgary researchers noted that Ban (1996) suggests that preparing for conferences takes approximately four times as long as actually having them. In my study, meetings averaged two hours and preparation time was eight hours, so these findings are consistent with Ban's observation. This would suggest that preparation time in the program was consistent with some standard FGDM practices, but preparation time was not significant in predicting which families developed a diversion plan.

Undoubtedly, preparation time is important in family involvement interventions. However, prescribing a specific amount of preparation time, without empirical support, is misleading. In a review of a Cedar Raids, Iowa application of a family-centered practice called Patch, the researchers points out that: "Patch need take no more time than conventional practice, but it does require that time be used differently" (Adams and Krauth, 1995). It may be possible that family meetings, organized one at a time, are very time intensive and require 20 hours to adequately prepare. However, if family meetings are organized in a larger context, such as the one illustrated in Figure 3, a structure of community and family support may be readily available thus making the organization of family meetings much faster. Family involvement interventions explored through contextual integrated strategies, may suggest whether this is a viable approach to organizing family meetings.

Private Family Time

The second essential element cited by some FGDM advocates is private family time (Mirsky, 2003). During private family time, after hearing from the professionals what their concerns are, the family is left alone to develop a plan. Many FGDM proponents regard private family time as an essential element: "Without private family time, research shows and history documents that power will not be shared, and the imbalance will persevere (Merkel-Holguin and Ribich, 2001, p. 211). Insisting that family involvement interventions must include this element is a potentially contentious requirement because two widely used family involvement interventions, Family Unity Meetings in Oregon and Team Decisionmaking in the Casey Foundation's Family to Family Initiative, do not routinely use private family time. Does this mean that these interventions do not empower families? In order to answer this question, we need to compare programs with and without private family time. While definitions of empowerment are illusive, some FGDM researchers have suggested evaluating FGDM's ability to empower families by whether programs can produce successful outcomes (Lupton and Nixon, 1997). Thus far, we have no outcome evidence that proves the value of private family time. Proponents of private family time say that it turns the decision making over to the family, gives them a sense of control, and symbolizes that the family is in charge (Mirsky, 2003). It would be useful to ask family members who participate in family meetings if they agree with these stated benefits.

An ongoing evaluation of Team Decisionmaking in Cuyahoga, County, Ohio is examining a related issue. In this case, staff are concerned about a tendency for some staff to step out of the room and consult with each other about a case without the family's participation. To evaluate this practice, the staff is asked to record how often this occurs and write a brief explanation of why it was necessary for someone to leave the room. Preliminary results suggest that someone left the room in only about eight percent of the meetings. The explanations show that sometimes people left the room for legitimate reasons unrelated to staff discussions (e.g. 'went to the bathroom,' 'feeding the meter,' 'infant being born'). Interestingly, it also includes examples of the family being left to discuss the case on their own without the professionals in the room. As suggested by Braithwaite, collecting this information, discussing it with staff, and linking it to outcomes will help us understand whether variations in attendance during the meeting are potentially important to the outcomes for children and families.

Prescribing Outcomes

Finally, FGDM advocates believe it is important that the facilitator of the meeting not prescribe the outcome (Mirsky, 2003). The question of who is making the decisions and when the decisions are made is a controversial issue in the family involvement field. For example, Team Decisionmaking consultants have written: "While team decisionmaking shares the same fundamental philosophy and values and is similar in participants and process, it differs significantly from family group conferencing. In team decisionmaking, the group is convened for the specific purpose of making an immediate placement related decision-and the process is used for each and every such decision faced by the public agency in its daily work. The public agency shares but does not delegate its responsibility to make critical placement decisions. Team decisonmaking therefore tends to be a high-volume and emotionally charged process which requires highly skilled agency staff to serve as facilitators" (DeMuro and Rideout, 2002, p. 12, emphasis added). In Team Decisionmaking, a high value is placed on involving families in decision making, however the primary concern is child safety. Bartholet, author of Nobody's Children: Abuse and Neglect, Foster Drift, and the Adoption Alternative, is explicit in saying: "If social workers are approving most of the family plans, I find it very troubling, because it says to me that what FGDM is about is near total delegation of decision making by the state to the family' (Shirk, 1999, p. 18). Usually in an FGDM process, the professionals are not supposed to be directing the family decision making; however, the social workers have an essential and significant role in both the preparation and the follow-up to the meeting. In addition, contrary to Bartholet's statement, typically the referring social worker must approve the family's plan, based on safety and permanency criteria, so it is not an abdication of decision making to the family, but a sharing of it.

While FGDM proponents typically clarify that social workers can veto a family's plan, they also suggest that families are making the decisions and that it is problematic to use FGDM to achieve a prearranged outcome. For example, Merkel-Holguin (2000) argues that:

   While family group confereneing provides families of origin and
   communities with a process to share decision-making authority
   with formal child welfare systems, it does not prescribe an outcome.
   In the United States, a troubling phenomenon is occurring in an
   increasing number of communities that, to receive much needed
   political support to initiate family group conferencing, are casting
   this approach as a cost-savings or containment mechanism, or one
   that results in certain outcomes. When this occurs, the intent of
   family conferencing is lost (emphasis added, p. 229).

Before an FGDM program makes the claim that they are not prescribing the outcomes, they should carefully review their referral process and examine which families are referred to an FGDM and which are not. In the United States, typically the public agency social worker most intimately involved in the case refers the case to an FGDM program (Merkel-Holguin and Ribich, 2001) and therefore, presumably also has the power to not make a referral. Although this FGDM selection process has not been fully explored in British studies of FGDM (called Family Group Conferencing or FGC), researchers from England suggest that there is evidence "that professionals retain considerable control over whether and which families are offered the choice of a FGC" (Lupton and Nixon, 1999, p. 119). Even if social workers are not prescribing outcomes in FGDM, if they have the power to decide who gets an FGDM, they do have considerable control of the outcomes. As Stevens (2003) points out: "Family Group conferencing involves a professionally initiated arena where families contribute to decisions about the future of their children, which represents a particular balance of control over decision making. While families make plans and decisions, professionals define and raise the specific issues to which families are responding" (p. 34).

Although the concern about who is making the decision may be important, it is difficult to develop a means of measuring it. In my study, we asked the FGDM staff to record the family's reaction to the child maltreatment report and the recommendation to remove the children from their caregivers (Crampton, 2001). Generally, this section is a few sentences long and indicates whether the family asked any questions about the investigation findings and whether they agreed with what was said. Family members typically agreed that the children needed to be removed, but they often asked questions and had different perspectives on the specific concerns. The study found that families who agreed with the initial removal decision but asked clarifying questions were more likely to develop a diversion plan. This is consistent with other studies showing that information sharing can be important. In Newfoundland-Labrador, they found that information sharing is critical: "The impact on the family of hearing the facts with everyone present in the room was regarded as a significant milestone in the reunification process.... the few exceptions being where the presentation was "preachy" in the words of one researcher and in one case where a presenter was described in the evaluation by several participants as "arrogant" (Burford, Pennell, MacLeod, Campbell and Lyall, p. 43).

In my study, families completed a seven-question survey immediately after the family meeting. A convenience sample of 156 surveys from forty-one meetings showed that the families were positive about the FGDM staff, fairly positive about the FGDM program and process, less enthusiastic about the role of the investigating Child Protection worker, and positive but sometimes disappointed by the outcome of their case. When social workers and families were engaged with each other in the process, they were more likely to develop a diversion plan. For example, those families who developed a plan were more likely to strongly agree that the coordinator shared necessary information and that they were given the information they needed to make a decision.

These observations of family members' reaction to the process and their responses to surveys seem to suggest that it is the sharing of information back and forth between the facilitator and the family that is important in reaching a collective decision. This may mean it is less important to determine who made the decision than to determine whether everyone felt they contributed to the decision. Burford and Pennell (1995) developed an instrument to measure meeting participants' perception of the amount of influence or "say," all the participants had in the meeting. This instrument was also used in a study of Family Unity Meetings in Oregon (Rockhill and Rodgers, 1999). The Oregon study suggested that a significant number of family members felt they had sufficient involvement in the decisions and the study identified a number of family meeting practices that seemed to improve the amount of say participants felt they had. For example, when a series of family meetings were held, the participants felt that their amount of say increased (Rockhill and Rodgers, 1999).

The findings summarized above begin to provide some ideas regarding the significance of preparation time, private family time, and decision making authority. Following Braithwaite's approach, communities developing family involvement interventions should review the theories related to these and other family involvement practices, the research findings on these practices and consider the applicability of this theory and research to their own community. While some stakeholders groups may reach similar conclusions about previous family involvement work, we would continue to expect to see variations in family involvement practice across communities. When we achieve some consensus around which practices are critical, we can then consider rigorous clinical trials that can demonstrate the efficacy of this approach. This follows the suggestions made by Wells (1994) that family preservation services research should place greater emphasis on connecting program design with theory and then, after the theory is developed, evaluate the links between the conceptualization of child welfare problems, causes, and their treatments.

Family Group Decision Making and related family involvement interventions are part of a larger debate about the role of the community and the role of the state in addressing cases of child abuse and neglect. In the United States, the child welfare system does not have sufficient resources to respond to all calls received from people who are concerned about the treatment of children by their parents (Faller, 1985; Lindsey, 1994; Schorr, 1997). The consequence of this dilemma for public child welfare managers is either to focus limited resources on the most severe cases or attempt to secure additional resources by developing partnerships with organizations and people who are willing to help in cases of child maltreatment. Adams (2000) describes these alternatives as a debate between the neostatist approach and the community-based approach to child welfare services reform. The neostatists believe state intervention should place a greater emphasis on adoption as the best option for children who are maltreated (Bartholet, 1999). The community-based approach emphasizes partnerships between families and communities to address child maltreatment (Waldfogel, 2000). Adams (2000) cites Family Group Decision Making as an example of this community-based approach. FGDM aspires to fundamentally change current child welfare practice and change it in a way that is very different from neostatist reform proposals. Not surprisingly, neostatists, like Elizabeth Bartholet (1999), are highly critical of FGDM and there is already a "backlash" against FGDM. Research following the methods of Braithwaite and others, can help provide a response to this backlash. Family involvement interventions appear to be a useful way to bring more community resources to the problem of child maltreatment. In order to demonstrate and promote their use, we should avoid the difficulties posed by the family preservation services movement and follow the path suggested by Braithwaite.

I would like to thank Pat Rideout, Kathleen Wells, Judy Wildfire and two anonymous reviewers for their comments on a previous draft of this paper. I would also thank colleagues from the Cuyahoga County Department of Children and Family Services, the Family Outreach Center in Grand Rapids, the Grand Rapids Community Foundation, the Kent County Family Independence Agency, the Mandel School of Applied Social Sciences, and the University of Michigan. While I received a great deal of help from them in understanding family involvement interventions in Cleveland and Grand Rapids, I alone am responsible for the views expressed in this paper.

References

Adams, P. (2000). "Bringing the Community Back In: Patch and Family Group Decision Making." In G. Burford and J. Hudson (Eds.). Family Group Conferencing: New Directions in Community-Centered Child and Family Practice. Hawthorne, NY: Aldine de Gruyter.

Adams, P. (1994). "Marketing Social Change: The Case of Family Preservation." Children and Youth Services Review, 16: 5-6, 417-431.

Adams, P. and Krauth, K. (1995). "Working with families and communities: The patch approach." In P. Adams and K. Nelson (eds.), Reinventing human services. Hawthorne, NY: Aldine de Gruyter. 87-108.

Allen-Meares, P. and Lane, B.A. (1990). "Social Work Practice: Integrating Qualitative and Quantitative Data Collection Techniques." Social Work, 35:5, 452-58.

Allison, G. T. (1971). The Essence of Decision: Explaining the Cuban Missile Crisis. Boston, MA: Harper Collins.

Ayres, I. and Braithwaite, J. (1992). Responsive Regulation: Transcending the Deregulation Debate. New York, NY: Oxford University Press.

Ban, P. (1996). "Implementing and Evaluating Family Group Conferences with children and families in Victoria Australia." In A. Morris, G. Maxwell and B. Gallaway (Eds.). Family Group Conferences: perspectives on Policy and Practice. Monsey, NY: Willow Tree Press.

Barth, R. P. (2002). "Review of Gale Burford and Joe Hudson (Eds.). Family Group Conferencing: New Directions in Community-Centered Child and Family Practice. Hawthorne, NY: Aldine de Gruyter." Journal of Sociology and Social Welfare, 29:3, 199-201.

Bartholet, E. (1999). Nobody's Children: Abuse and Neglect, Foster Drift and the Adoption Alternative. Boston, MA: Beacon Press.

Braithwaite, J. (2002a). Restorative Justice and Responsive Regulation. New York, NY: Oxford University Press.

Braithwaite, J. (2002b). Rebuilding a Compassionate Republic: The Family Group Conference as a Democratic Experiment. Closing Address presented at the 2002 Family Group Decision Making Roundtable in Monterey, California.

Braithwaite, J. (2000). "Democracy, Community, and Problem Solving." In Gale Burford and Joe Hudson (Eds.). Family Group Conferencing: New Directions in Community-Centered Child and Family Practice. Hawthorne, NY: Aldine de Gruyter.

Braithwaite, J. (1993). "Beyond Positivism: Learning from Contextual Integrated Strategies." Journal of Research in Crime and Delinquency, 30:4, 383-99.

Braithwaite, J. and H. Strang (2002). "Restorative Justice and Family Violence." In H. Strang and J. Braithwaite (Eds). Restorative Justice and Family Violence. New York, NY: Cambridge University Press.

Burford, G. (2001). "Advancing Innovations: Family Group Decision Making As Community-Centered Child and Family Work." Protecting Children, 16:3, 4-20.

Burford, G. and Hudson, J. (2000). Family Group Conferencing: New Directions in Community-Centered Child and Family Practice. Hawthorne, NY: Aldine de Gruyter.

Burford, Gale and Pennell, Joan (1995). Family Group Decision Making: New Roles for "Old" Partners in Resolving Family Violence: Implementation Report Summary. St. John's, Newfoundland: Institute of Social and Economic Research, Memorial University of Newfoundland.

Burford, G., Pennell, J., MacLeod, S., Campbell, S., and Lyall, G. (1996). "Reunification as an extended family matter." Community Alternatives: International Journal of Family Care, 8(2), 33-55.

Crampton, D. (2001). "Making Sense of Foster Care: An evaluation of Family Group Decision Making in Kent County, Michigan." Dissertation Abstracts International 62/10 (p. 3573). UMI publication number AAT 3029324.

DeMuro, P. and Rideout, P. (2002). Team Decisionmaking: Involving the Family and Community in Child Welfare Decisions. Baltimore, MD: The Annie E. Casey Foundation.

Faller, K.C. (1985). "Unanticipated Problems in the United States Child Protection System." Child Abuse and Neglect, 9, 63-69.

Lin, A. C. (1998). "Bridging Positivist and Interpretivist Approaches to Qualitative Methods." Policy Studies Journal, 26:1, 162-40.

Lindsey, D. (1994). The Welfare of Children. New York: Oxford University Press.

Lupton, C. and P. Nixon (1999). Empowering Practice: A critical appraisal of the family group conference approach. Bristol: The Policy Press

Marsh, P. and G. Crow (1998). Family Group Conferences in Child Welfare. Oxford: Blackwell Science Ltd.

Merkel-Holguin, L. (2000). "Diversions and Departures in the Implementation of Family Group Conferences in the United States." In G. Burford and J. Hudson (Eds.) Family Group Conferencing: New Directions in Community-Centered Child and Family Practice. Hawthorne, NY: Aldine de Gruyter.

Merkel-Holguin, L. and K. Ribich (2001). "Family Group Conferencing: An Extended Family Process to Safeguard Children and Strengthen Family Well-Being." In E. Walton, P. Sandua-Beckier, and M. Mannes (Eds.). Balancing Family-Centered Services and Child Well-Being. New York, NT: Columbia University Press, 197-218.

Mirsky, L. (2003). "Family Group Conferencing Worldwide: Part One in a Series." Restorative Practices EForum, February 20, 2003, The International Institute for Restorative Practices, available from http://www.restorativepractices. org

Morgan, G. (1986). Images of Organizations. Beverly Hills, CA: Sage.

Morris, A. (2002). "Critiquing the Critics: A Brief Response to Critics of Restorative Justice." The British Journal of Criminology, 2002:42, 596-615.

Nixon, P., Merkel-Holguin, L., Sivak, P., and Gunderson, K. (2001). "How Can Family Group Conferences Become Family-Driven? Some Dilemmas and Possibilities." Protecting Children, 16:3, 22-33.

Pennell, J. and Burford, G. (2000). "Family Group Decision Making: Protecting Children and Women." Child Welfare. 79:2, 131-58.

Pennell, J. and Weil, M. (2000). "Initiating Conferencing: Community Practice Issues." In Gale Burford and Joe Hudson (Eds.). Family Group Conferencing: New Directions in Community-Centered Child and Family Practice. Hawthorne, NY: Aldine de Gruyter.

Prantis, Kay (2000). "Conferencing and the Community." In Gale Burford and Joe Hudson (Eds.). Family Group Conferencing: New Directions in Community-Centered Child and Family Practice. Hawthorne, NY: Aldine de Gruyter.

Ristock, J. L. and Pennell, J. (1996). Community Research as Empowerment: Feminist Links, Postmodern Interruptions. Toronto: Oxford University Press.

Rockhill, A. and Rodgers, A. (1999). Family Decision Meetings: Final Report. Portland, OR: Child Welfare Partnership, Portland State University.

Schorr, L. B. (1997). Common Purpose: Strengthening Families and Neighborhoods to Rebuild America. New York, NY: Anchor Books.

Schuerman, J. R., Rzepnicki, T. L. and Litten, J. H. (1994). Putting Families First: An Experiment in Family Preservation. New York: Aldine de Gruyter.

Shadish, W. R. Jr., Cook, T. D., and Leviton, L. C. (1995). Foundations of Program Evaluation: Theories of Practice. Thousand Oaks, CA: Sage Publications.

Shirk, M. (1999, October). "New Zealand Social Work Tactic Hits Shore, Makes Waves: Is "Family Group" an exciting tool for abused kids and juvenile offenders, or an expensive fad?" Youth Today, Volume 8 (9), p. 1+18.

Sieppert, J. D., J. Hudson, and Unrau, Y. (2000). "Family Group Conferencing in Child Welfare: Lessons from a Demonstration Project." Families in Society: The Journal of Contemporary Human Services, 81:4, 382-391.

Stevens, M. (2003). "Reconstruction Works? Constructing Family Perspectives of the Outcomes of Family Group Conferences." Protecting Children, 18:1&2, 30-41.

Van Stokkom, B. (2002). "Moral Emotions in Restorative Justice Conferences: Managing Shame, Designing Empathy." Theoretical Criminology, 6:3, 339-60.

Waldfogel, J. (2000). "Reforming child protective services." Child Welfare; 79:1, 43-57.

Wells, K. (1994). "A Reorientation to Knowledge Development in Family Preservation Services: A Proposal." Child Welfare; 73:5, 475-488.

Whittaker, James K. (1999). Forward to Connolly, Marie and Margaret McKenzie Effective Participatory Practice: family group conferencing in child protection. Hawthorne, NY: Aldine de Gruyter.

Wilmot, Leslie (2002). "About AHA's Front Porch Project." Protecting Children, 17:2, 1.

DAVID STUART CRAMPTON

Case Western Reserve University

Mandel School of Applied Social Sciences