Collaboration as a Means, Not an End: Serving Disadvantaged Families and Children

Olivia Golden

Kennedy School of Government

Harvard University

The purpose of this paper is to offer a framework for thinking about collaboration as a strategy for successful service delivery to children and families, particularly the most needy children and families. The paper aims to stimulate discussion by raising some central questions and offering a single perspective on how to resolve them; it does not aim to provide a comprehensive account of what we know or how we ought to proceed.

Discussions of collaboration often begin from an account of successful collaborations and the lessons they teach us. However, to start there begs three important questions:

  1. Why is collaboration a good thing? That is, why should we be interested in it in the first place?

  2. What is collaboration? What are the important types of collaboration for us to pay attention to?

  3. Why is collaboration so hard? That is, what did these programs have to overcome to become successful?

This paper starts with these questions, because my own experience of studying successful programs convinces me that we cannot do justice to the lessons from program experience if we ignore them, if we implicitly assume that collaboration in itself is a good and sufficient purpose, or that the barriers to collaboration are obvious. Furthermore, whether or not we have to answer these questions in order to understand program experience, we certainly have to answer them in order to persuade public officials to act on program experience, since those officials are unlikely to be interested in collaboration for collaboration's sake—and they are likely to be concerned about the likely obstacles.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 84
Effective Services for Young Children: Report of a Workshop Collaboration as a Means, Not an End: Serving Disadvantaged Families and Children Olivia Golden Kennedy School of Government Harvard University The purpose of this paper is to offer a framework for thinking about collaboration as a strategy for successful service delivery to children and families, particularly the most needy children and families. The paper aims to stimulate discussion by raising some central questions and offering a single perspective on how to resolve them; it does not aim to provide a comprehensive account of what we know or how we ought to proceed. Discussions of collaboration often begin from an account of successful collaborations and the lessons they teach us. However, to start there begs three important questions: Why is collaboration a good thing? That is, why should we be interested in it in the first place? What is collaboration? What are the important types of collaboration for us to pay attention to? Why is collaboration so hard? That is, what did these programs have to overcome to become successful? This paper starts with these questions, because my own experience of studying successful programs convinces me that we cannot do justice to the lessons from program experience if we ignore them, if we implicitly assume that collaboration in itself is a good and sufficient purpose, or that the barriers to collaboration are obvious. Furthermore, whether or not we have to answer these questions in order to understand program experience, we certainly have to answer them in order to persuade public officials to act on program experience, since those officials are unlikely to be interested in collaboration for collaboration's sake—and they are likely to be concerned about the likely obstacles.

OCR for page 84
Effective Services for Young Children: Report of a Workshop Having offered answers to these initial questions, the paper then goes on to suggest promising future directions. It tries to explain how successful collaborations have overcome the barriers to collaboration, providing a tentative list of elements these successes share. It concludes by proposing four opportunities for change, selected in the hope that the forum's recommendations could make a difference. WHY IS COLLABORATION A GOOD THING? For some writers on collaboration, its value is self-evident. The present system is fragmented, meaning that service deliverers and the agencies they work in do not know about each other's services, talk to each other, work together, or even share a common language or perception of the problem. Therefore, the solution is collaboration: service deliverers and program administrators ought to bridge these gaps. As Bruner (1990b) and Altman (in this volume), among others, have recently noted,1 this is not a very satisfying account of the reasons for collaboration, for several reasons. First, it is not obvious that collaboration always has good rather than bad effects on services for families and children. Collaboration might lead agencies to carry out their differentiated, precollaboration mission less well: for example, an eligibility worker asked to assess families' service needs and refer them to family support programs might get benefit checks out less quickly or take information gained from the assessment inappropriately into account in the benefit determination. Collaboration might lead a program that has been effective on Schorr's criteria to become less so, if it collaborates with a more rigid, bureaucratic program and its mission and culture are diluted. For example, staff in a teenage parent program I visited for recent research on welfare and children's services were very nervous about the emphasis on rules that their (ultimately unsuccessful) collaboration with a local welfare agency was, they thought, imposing on their services. And for some families, there seems to be nothing wrong with the present system of services and therefore little to be gained from collaboration: many middle-class families, for example, may be able to ensure that a child gets to school and gets to the doctor without seeing any particular reason why the school system and the health system ought to communicate (Bruner, 1990b:8). Second, collaboration has costs for staff. It requires time, often a great deal of time, for staff to learn about each other's purposes and activities, develop a common language, work on practical problems, and resolve conflicts (Bruner, 1990b:33). Collaboration may also be costly when it shifts the kind of skills workers 1    For an argument that goes even further to make the case that collaboration is not a desirable route for human services, see Weiss (1981).

OCR for page 84
Effective Services for Young Children: Report of a Workshop need—for example, by adding negotiation and problem-solving skills to narrower technical skills—and therefore requires retraining and perhaps redeployment or attrition of staff. And collaboration can be costly in terms of increased levels of staff tension, at least in the short run: questions that once could be decided simply by one or two people now require an extended process, and jobs that once were private turf are invaded by outsiders (see for example Morrill and Gerry, no date; Bruner, 1990b:7). Thus, we need a better justification for collaboration as a solution than simply that services are fragmented. In particular, to advocate for collaboration in services to the most needy families and children, we need to believe that collaboration will typically lead to better service delivery. Why should this be? The answer, it seems to me, lies in the findings of Schorr and others about the characteristics of disadvantaged families and the way those characteristics interact with the system that serves them (for a longer discussion, see Golden, 1990). For families and children who have multiple needs—needs that are related to each other in complex causal chains—and who lack knowledge, self-confidence, money, and skill in handling bureaucracies, the fragmented service system imposes enormous burdens. And these burdens may well fall most heavily on those least able to cope: the most fragile, those with the most complex needs, those with the fewest resources. For example, a case manager in a program for low-income teenage parents reported on the difficulties that the teenagers' lack of education posed for their dealings with the health system, in the case (which she reported was not rare) of teenage parents of babies with chronic illness (Golden, 1990:48): The one I have now without a bladder—this baby will have multiple surgery, and the mother has been tested at third grade and she doesn't understand what is going on. So I have to go along to interpret what the nurses are saying. For these families, fragmentation is indeed destructive, because it means that no matter which fragment of the system they enter, which door they go through, they will find someone who can respond only to a tiny piece of their problem. Without the capacity to work the system themselves, they will end up lost unless the system itself provides the glue to keep the pieces together. Therefore, service deliverers who are to succeed with these families must identify and respond to needs that cross system boundaries, or in Lisbeth Schorr's terms, "adapt or circumvent traditional professional and bureaucratic limitations when necessary to meet the needs of those they serve" (Schorr, 1988:258). Unless these service deliverers are able to do everything from educate a young child to care for a sick baby to counsel a suicidal mother within their own organizations, this demand requires that they collaborate.

OCR for page 84
Effective Services for Young Children: Report of a Workshop Charles Bruner (1990b) takes this analysis one step further, sorting out three different types of families whose needs particularly require collaboration. First is the family with many needs at a crisis or near-crisis stage and many human services providers already involved. For this family, the key benefit of collaboration is that the providers will talk to each other, stop conveying to the family contradictory or at least unrelated expectations, and think about the impact of their interactions with different family members on the family as a whole. An example from my own research, offered by case managers in Oklahoma's Integrated Services System, illustrates the nature and benefits of collaboration in such a case (Golden, 1990:49-50): A seven-year-old boy came to the attention of a school principal because of both physical and emotional health problems. The boy had long been prone to seizures and self-destructive behavior and was just starting to threaten other children. When the principal called IFS [Integrated Family Services], he found that IFS was already working with the family because the mother was on AFDC and herself had multiple problems. The IFS worker called a meeting of all the agencies who had contact with the family to talk about the child's needs. As a result, the boy was admitted and sent to a diagnostic center for several months of testing and treatment; the mother received needed services such as mental health treatment and literacy training; and the Child Protective Services worker changed her mind about the possible outcomes for the case and concluded that the mother had the potential to be an adequate parent. Second is the family with many low-level needs that add up to serious risk for the child. Such a family may have limited contact with three or four service systems—the public schools, the unemployment office, the city hospital, and perhaps the welfare department—but it probably won't meet anyone's criteria for intensive services. Yet a collaborative approach might well identify it as a family for whom there are real opportunities for intervention that could change a child's life. Bruner (1990b:10) illustrates this point with an example: Johnny, a seven-year-old first grader, is behind his fellow students in reading and other activities. He often is late to school, as his mother works nights and sometimes does not get up in time to get him off to school. There are no books in Johnny's home, and his mother, a dropout from ninth grade, views the school system with a sense of powerlessness and distrust. Johnny and his mother live together in a ten-year-old trailer, and Johnny frequently gets colds from the drafts through the trailer. Third, Bruner notes the role of collaboration in serving families whose problems are related to community-or neighborhood-wide concerns: unsafe housing, poverty, violence on the streets, limited access to health care

OCR for page 84
Effective Services for Young Children: Report of a Workshop (p. 10). In this kind of example, collaboration is important because of the interlocking problems in the neighborhood and because failures by a wide range of public service systems have contributed to those problems, not necessarily because of the interlocking problems of the family itself. Thus, the purpose of this kind of collaboration is to produce changes in various service policies and in the allocation of resources, not necessarily changes in the ways individual families are treated. Collaboration, then, is not an end in itself but a means to better services for families and children. I have argued that it is nonetheless a crucial means for many of the most needy families: those who cannot themselves integrate a fragmented system to get what they need, because of the complexity of their needs, their financial and other difficulties in access, their isolation or lack of self-confidence. And collaboration may also be a crucial part of the solution for families not themselves fragile or isolated who live in neighborhoods characterized by the failure of a whole interlocking set of public systems. WHAT IS COLLABORATION? In the account above of why we should care about collaboration, I have offered examples of collaboration but not a definition. One important reason for that choice is that the purpose we have in mind for collaboration, meeting the needs of fragile families and children, is difficult enough without artificially limiting the possible techniques by defining them away. For example, there is no way to know a priori whether collaboration at the street-level, where services to families are delivered, has more or less potential to change their lives than collaboration at the state level, which affects the decisions of top policy staff. Therefore, I propose that we define collaboration as simply as possible and concentrate not on its boundaries but on the multiple ways of carrying it out. A simple definition might be that collaboration involves separate organizations working together. The key point here is that collaboration is not the same as merger: a collaborative effort retains different organizations with their separate specialties and points of view. Therefore, for a collaboration to succeed, each organization must see a contribution from the collaboration to its own mission and purposes. A collaboration between a school and a health clinic, for example, does not imply that they can now substitute for each other, but rather that they are working together in a way that each (if it is a successful, ongoing collaboration) sees as helping it do its own job better. Not surprisingly, this feature of collaboration has implications both for barriers to collaboration and for the approaches taken by successful collaborators. An enormous variety of activities could fit under this simple definition. Classifying these activities into types has two purposes: to enable us to scan what is already going on or contemplated, and to help us think up new

OCR for page 84
Effective Services for Young Children: Report of a Workshop activities that seem promising. Three dimensions of collaborative activities seem to me useful in such a classification: mission, defined either as a vision to be achieved or a problem to be solved; level, such as service delivery, city, and state; and function, such as outreach and intake, evaluation, development of new resources, and policy design. Thus, a collaboration might have the mission of reducing teenage pregnancy, be organized at the city level, and focus on achieving its mission through several functions: joint outreach by service deliverers who see out-of-school teenagers (public health nurses, community youth workers), direct service delivery in the schools (life planning programs), improved and more personalized brokering of services (a nurse who will personally follow up on a young woman's referral to a local health clinic for a family planning appointment), and cross-training of service deliverers. Another collaboration might have the mission of serving children with multiple disabilities who fall through gaps in service, might be organized at the state level, and might plan to achieve the mission primarily through functions like the exchange of information on caseload and service characteristics, evaluation of services (for example, the identification of gaps), and the development of priorities for new programs and resources. Table 1 offers examples of collaborative activities now being carried out or proposed at each level, to suggest the broad range of possibilities. As the next section suggests, we will need to consider all the flexibility and all the opportunities this broad range offers if we are to overcome the many and fundamental obstacles. WHY IS COLLABORATION SO HARD? To the doctor, the child is a typhoid patient; to the playground supervisor, a first baseman; to the teacher, a learner of arithmetic. At times, he may be different things to each of these specialists, but too rarely is he a whole child to any of them. — From the 1930 report of the White House Conference on Children and Youth, cited in Harold Hodgkinson, The Same Client: The Demographics of Education and Service Delivery Systems, 1989. Ricardo and his family are being "served" by at least nine agencies—but no one agency truly takes responsibility for helping Ricardo, and separately they fail to treat him as a whole person. They are paid to treat each of a variety of "problems"—poor grades or absenteeism or child abuse, for instance—that add up to his being at risk. And they have no way to get beyond these symptoms to who he is as a whole person, and so he moves, lost, from one agency to another. — From ''Failure by Fragmentation" by Sid Gardner, Fall 1989.

OCR for page 84
Effective Services for Young Children: Report of a Workshop TABLE 1 Examples of Types of Collaborative Activities Level Activity Service delivery Case management: a worker collaborates personally with a range of other service deliverers in order to provide a broad menu of services to a family.   Case management within a structure of formal interagency agreements that guarantee access to resources.   Team staffings, to ensure that information is shared and decisions made jointly about individual cases.   Colocation of services, one-stop shopping.   Cross-training among service deliverers.   Regular team meetings of service deliverers from multiple agencies, to share information and perspectives on the system as a whole, not just individual cases.   Regular team meetings within a structure in which the service deliverers at the meeting have authority to change policies. Program Interagency agreements specifying access to resources, joint roles of staff.   Flexibility to use resources across program boundaries.   Changes in policies or structures to respond to needs identified by other systems, improve access, etc.   Regular meetings by program directors to discuss policies, needs of target populations, quality of services, etc.   Sharing of information about populations served.   Joint efforts at intake and case tracking.   Joint decisions about target populations to be served and services to be offered. Community or city Joint information collection about families, neighborhoods, and services.   Joint development of mission statement and accountability measures.   Joint decisions or recommendations about allocation of resources or at least priorities for new programs. State or federal Policy changes or waivers to produce pooled funding streams for local programs, probably with conditions for inclusiveness and accountability of local efforts.   Joint development of accountability measures.   Changes in regulation to eliminate major barriers to collaboration.   Grant funding made available with incentives to collaboration, or on condition of collaboration.   Mandates for joint planning.   Joint information collection about families and services, joint planning, joint development of priorities for resource allocation or for new programs.   Technical assistance to communities, cities, programs, service delivery workers to carry out all the new collaborative activities.   Dissemination of information about collaborative activities. Academic institution Development of joint curricula, across social work, education, public health, mental health.

OCR for page 84
Effective Services for Young Children: Report of a Workshop These two quotations 60 years apart point to exactly the same problem: a fragmented service system that fails to serve a child as a whole person, let alone a family as a whole unit. If anything, it looks as though the problem has gotten worse from the first to the second: at least the first child is seen by two of the service deliverers as someone with capacities, while the second child is seen solely as someone with problems. Why so little success over all that time? Many authors have identified a long list of reasons why collaboration is difficult in the world of child and family services (see for example Schorr, 1988; Weissbourd, 1990; Gardner, 1989; Weiss, 1981): Categorical funding streams. Federal and state funding sources tend to support programs defined in terms of substantive categories Second, multiple governmental jurisdictions, and therefore multiple accountability relationsheriment with collaboration in child welfare, the state put together what had originally been 30 different funding streams. Professional education and traditions. Teachers, social workers, pediatricians, mental health clinicians, and all the rest of the service deliverers for children and families have had different educational backgrounds, speak different languages, and look to their own professions for recognition, respect, and promotion. Organizational rigidities and loyalties. Reinforcing the professional loyalties are organizational loyalties that may lead staff to view other agencies suspiciously: they have an easy job or don't really care about children the way we do, for example. At the same time, staff may well be trained and monitored for their conformity to rigid organizational procedures, which cannot be varied to meet the needs of other agencies or of families who don't fit the categories. For example, one welfare agency did not succeed in collaborating with an early childhood development program because it monitored its workers for attention to ''mandatory" enrollees in work and training—whose youngest child was at least six—rather than "voluntary" enrollees, who might have a child the right age for the early childhood program (Golden, 1990:133-139). Inadequate resources to do the core job. As suggested above, collaboration has costs, which staff may be unable or unwilling to incur if they cannot carry out their core functions well. While it may be true over the long term that collaboration will help solve core problems, individual workers may not be in a position to take a chance in the short term. For example, when child welfare workers are stretched to their limit to investigate emergency cases, even simple collaborative activities like explaining to school personnel why they are unable to take action on an abuse and neglect referral seem impossible—and more demanding activities such

OCR for page 84
Effective Services for Young Children: Report of a Workshop as meeting with teachers to talk about common concerns can't even be contemplated.2 But this list of problems seems to me to raise another question: why, over 30 years or more, haven't we been able to solve them? Surely with a compelling analysis of the benefits of collaboration, we ought to have been able to make funding streams more flexible, break down the barriers of professional education and organizational rigidity, and persuade workers and administrators that core functions will be advanced, not hindered, by work with other agencies. The answer, I think, is that once again we cannot ignore the costs of collaboration or the benefits of the existing structure, the values that all of these barriers and obstacles are protecting. In particular, two important values protected by a fragmented system are differentiated expertise and accountability for functions that the public and political leaders know, understand, and believe in. If we are to develop recommendations for collaboration that can survive, we need to take these values seriously. The Value of Differentiated Expertise The definition of collaboration offered above emphasized that it involves different organizations with their own different outlooks and purposes. As many writers have noted (see for example Lawrence and Lorsch, 1967), this differentiation of mission, expertise, and outlook among organizations (or even parts of organizations) is precisely what makes it difficult to work together. At the same time, this differentiation is there for a reason: the different organizations are responding to different technologies, different environments, different problems. A terrific surgeon can't work in exactly the same way as a terrific classroom teacher: the tasks are too different, and so the approach is different. An organization made up of terrific surgeons will, as a result, operate quite differently from an organization made up of terrific classroom teachers. But what do you do when it is important for organizations that are legitimately very differentiated to work together in order to accomplish a crucial purpose? A famous study of private-sector corporations argues that this is precisely the situation in certain industries characterized by rapidly changing technologies, markets, and production and a need for constant and innovative response. In these industries, success requires high levels of both differentiation (that is, specialized expertise in very different fields—for example, in basic research and customer relations) and integration (that is, collaboration to further the common mission of the firm). The study argues 2    For this experience from the teacher's point of view, see Weissbourd (1989).

OCR for page 84
Effective Services for Young Children: Report of a Workshop that although there is a trade-off between differentiation and integration; excellent companies achieve a higher level of both than weaker companies, relying on a variety of techniques to improve communication and resolve conflicts (Lawrence and Lorsch, 1967:52-53). The implication for collaborative services to children and families, it seems to me, is that we may be in exactly the same position as companies in those difficult industries: our success, at least for the most needy children, may depend on excellence at both differentiation and integration. That is, more than other children, the most fragile children may need skilled teachers, counselors, and sometimes even surgeons—yet at the same time, more than other children, they need to be helped by these skilled specialists as whole children, seen in the light of other aspects of their lives. To take a slightly different example, responding successfully to the needs of a fragile family with a very ill baby may require both the highly differentiated skills of the surgeon and the ability to integrate many different services around the family's needs. Thus, to think effectively about collaboration, we may need to think about differentiation and maintaining separate standards of excellence at the same time. As I argue below, effective collaborations seem to take up this challenge in two ways: they try to redefine the mission of each constituent unit to some degree, in order to make vivid to each specialist how he or she fits into the overall picture, and they also anticipate continuing conflict even with the redefined missions and develop effective ways of resolving it. The Value of Accountability Besides the inherent value of differentiated expertise, public-sector agencies also need to take into account their political mandate. Public agencies are accountable to political overseers and ultimately to the public, who have a conception, implicit or explicit, of what those agencies are supposed to be doing. In many cases, public conceptions are defined in terms of the differentiated missions, not the integrated ones. Therefore, shifting toward collaboration may look to the public and the political overseers like shifting toward activities on the fringe of the agency's mission, at the expense of the core. For example, child protective agencies are expected by the public to prevent children from injury and death and to place children in safe foster homes that will not endanger them further. Public welfare agencies are expected to control the total amount of money they spend and spend it accurately. The director of one welfare agency that has engaged in successful collaboration reports that he is able to spend his time on external tasks only because his agency has a history of excellence on cost control, fraud control, and automation that gives him the freedom to move outward. Two features of accountability in the field of children's services may further complicate the task of the agency trying to move to a broader mis-

OCR for page 84
Effective Services for Young Children: Report of a Workshop sion. First, the narrow missions may well be much easier to measure in quantitative terms: it is, for example, easy to measure how accurately the welfare department is determining eligibility. Without equivalent benchmarks for the new, broader mission, the public may be suspicious, and the "harder" measurements may tend to drive out the fuzzier or harder-to-measure mission from the minds of agency workers and managers. Second, multiple governmental jurisdictions, and therefore multiple accountability relationships, characterize services to children and families in the United States. The agencies we might select to collaborate around services to families and children are generally accountable not to one but to two, three, or four different levels of government or elected bodies. Welfare and child protective services are generally located at the state or county level; the school system is likely to be more local, possibly but not necessarily coterminous with the city or county government and probably accountable to a separately elected school board; the city government may fund a variety of child and family services; Head Start programs are subgrantees to the federal government; and so forth. In one California county with an impressive record on collaboration, the collaborating agencies worked for four separate governmental jurisdictions: the city, the county, the unified school district, and the community college district. That list does not include the state and federal governments, which provided funds for major services such as welfare and child welfare and therefore demanded accountability to their own rules and purposes. The important point here is that accountability for an agency's purposes and effectiveness is not in itself a villain; it is part of a democratic system. But left to themselves, many of our existing mechanisms for accountability will operate against collaboration and for existing bureaucratic procedures. Effective collaboration, therefore, requires the deployment of political, analytic, and managerial talent to discover and carry out alternative approaches to accountability (see also Gardner, 1989; Morrill and Gerry, no date). ELEMENTS OF SUCCESSFUL COLLABORATION With that range of barriers, it seems extraordinary that recent reports identify so many collaborations on behalf of families and children that have been successful at least for some period of time. How do these successful collaborations overcome the range of barriers described above? Based on the programs that I have studied, supplemented by an incomplete review of the literature, I would offer two general answers. First, we should not forget that, although we can identify many successes, we can also identify many programs that don't succeed. For example, in the research on welfare and children's services, out of seven programs we selected for study because of their professional reputation as successes, two

OCR for page 84
Effective Services for Young Children: Report of a Workshop are no longer operating as collaborations, only a year after our site visits (Golden, 1990).3 The implication I draw from this point is not that we should give up but that we should understand what success in promoting collaborative service provision really means: creating an atmosphere where more, but still not all, attempts at collaboration could succeed over the long run. Second and perhaps more optimistically, the successful collaborations, while sharing many common themes, demonstrate a particular knack for skillful adaptation to their local situations. While they must overcome a common set of barriers and address some number of a common set of tasks, their success at overcoming the barriers and completing the tasks comes precisely from their ability to adapt to local characteristics and needs. For example, collaborations are able to develop a common mission despite the pressure for differentiated, specialized missions precisely because the collaborators are good at identifying important local problems that require cross-cutting activities for their solution. More specifically, in the remainder of this section, I identify five common elements of successful collaboration that come out of my own research on welfare agencies and children's services and that appear consistent with other findings in the literature. I add a sixth element that is frequently mentioned in the literature and seems centrally related to the underlying justification for collaboration. Redefined and overlapping missions. Collaborating organizations have succeeded in developing overlapping conceptions of mission. They have managed to enlarge their missions beyond the conventional limits associated with their agency, often because of managerial skill in identifying key local political problems or opportunities. Conflict resolution. In effective collaborations, people do not expect that conflict between the organizations will disappear or attribute such conflict solely to uncooperative personalities, power struggles, or turf battles. Instead, they develop ongoing mechanisms for conflict resolution. Commitment of managerial time. Managers devote considerable time and attention to collaboration, spending time looking out from their organizations rather than simply up or down within them. Role of personal relationships. Managers and service deliverers cultivate personal relationships as a basis for collaboration. Exchange relationships. Collaborating agencies have some basis for exchange—something they can do for each other. Involvement of families and a broad spectrum of the community. A number of writers have argued that involvement of those who are served at all stages of the collaboration—from planning to service delivery—is critical to success. 3    See Weiss (1981:24) for a pessimistic assessment of the failure rate of collaborations.

OCR for page 84
Effective Services for Young Children: Report of a Workshop Redefined and Overlapping Missions If one explanation for the difficulty of collaboration is agencies' differentiated missions, we might expect to see redefined and overlapping missions in successful collaborations. Several recent reviews of collaborative experience have developed this theme: for example, one recent report emphasizes the role of establishing values and sharing a vision, and another notes the value of overlapping administrative responsibilities and legal mandates (see for example Institute for Educational Leadership, no date; Levy, 1989). The evidence from seven sites I studied at which the welfare agency collaborated in the delivery of children's services is consistent with this theme. For example, a collaboration between a welfare agency and a nonprofit agency providing child care resource and referral services to welfare recipients illustrates a common conception of mission. In our interviews, eligibility workers and employment and training (ET) workers at the welfare department and parent counselors at the child care resource and referral agency told stories about their own aims and their own greatest successes that suggested a common mission: self-sufficiency for mothers on welfare, coupled with excellent child care to ensure a mother's peace of mind. The sites also illustrate the other side of this proposition: in cases in which the mission conflicts have not been effectively resolved, collaboration has not succeeded. For example, we visited a program in which the welfare agency was funding several sites in a family literacy program that provided a high school equivalency program for the parent and a preschool program for the child, with slots at those collaborative sites to be reserved for welfare recipients. The agencies began the project because it seemed to meet both their needs: the family literacy program had difficulty recruiting families, and the welfare department was interested in training programs for mothers, particularly given the convenient pairing with child care. But the collaboration did not survive early difficulties, because the local welfare office saw its primary mission as following the rules on benefit checks and processing clients who were mandated to register for training programs. Because the families with young children who were appropriate for the family literacy program were not mandated registrants, welfare workers did not refer enough families to fill the program, and local office administrators were unable to move beyond their mission and improvise an agreement to solve the problem. Although intervention from the welfare agency's central office filled the program in its first year, the welfare agency chose not to participate again in the second year. But how do agencies develop support for a broader sense of mission, given the pressure for narrower accountability described above? The answer suggested by the sites I studied is consistent with that suggested by several recent reports: successful collaborations are extremely responsive

OCR for page 84
Effective Services for Young Children: Report of a Workshop to local needs, to problems already defined or potentially definable by the local political system. As Levy (1989) points out, "systemic pain," meaning problems that no one can solve alone, is a powerful source of collaboration (see also Institute for Educational Leadership, no date). One of the collaborative managers in a site I studied expresses his belief that this motivation will eventually drive widespread collaboration (Golden, 1990:11): Economics are going to drive it. They will force people to get together. Everyone around the country has overwhelming caseloads and too few resources. The Federal deficit won't be resolved, taxes won't be enough to keep up with health and social problems. So collaboration makes sense from an economic point of view. But the nature of the "systemic pain" that drove collaboration in my study sites, and the mission that emerged from that pain, did not seem to me to be consistent across collaborations. Rather, agency managers were able to take advantage of particular political and organizational opportunities to develop their own unique problem and mission definition. The agencies have sustained political and organizational support through a match between the precise form of the collaborative mission and locally defined problems and needs. For example, a Detroit program involving collaboration between the welfare agency and the schools responded to local concern about school dropouts, a problem that was politically salient there in a way that "losing families between the cracks," one of the problems that drives another of the site programs, might not be (Golden, 1990:147-157). Conflict Resolution Defining collaboration as the effort to integrate units that must at the same time remain highly differentiated makes sharply evident the need for continuing processes of conflict resolution in even the most successful collaboration. Teachers, psychiatrists, and child protective workers will inevitably see different aspects of a child, no matter how respectful they are of each other's viewpoints, and some mechanism for settling the disagreements seems critical. Perhaps because of an idealized view of collaboration, this aspect has not been much studied, although several reports emphasize the value of skilled facilitators to collaborations, and one report notes that a mechanism for "voicing concerns" is crucial to successful collaborations.4 This finding is similar to the Lawrence and Lorsch (1967) finding in the private sector that effective conflict resolution in highly differentiated 4    The phrase "voicing concerns" is from Institute for Educational Leadership (no date: 42-44.) One source in which the facilitator suggestion is developed at length is Robinson and Mastny (1989).

OCR for page 84
Effective Services for Young Children: Report of a Workshop organizations involved more direct and explicit disagreement than ineffective conflict resolution. Again, the findings from my sites are consistent with the idea that effective conflict resolution matters to collaboration. For example, although ET workers and parent counselors in the child care agency agree on their broad mission, they reported disagreeing on priorities for particular families. For example, the parent counselors reported that they might disagree with ET workers about the timing of a mother's entry into training: Should she start right away, or should she wait until the next program cycle to take the time to feel more comfortable with her child care arrangements? They reported a clear sense of how they would resolve that conflict: they would encourage the client to delay (which she is able to do under a voluntary program), speak to the ET worker if necessary, and perhaps try to "educate" workers at the public welfare office as a whole about the nature of adjustment to child care. One of the other sites at which the collaboration has dissolved seemed to have a particularly ineffective mechanism for conflict resolution. In this collaboration, between a nonprofit organization and a county department of social services, there was no clear process for resolving conflicts that might have been easily anticipated, given the agencies' different missions and purposes. Conflicts appeared to drag on, without personal meetings or negotiation, and to be elevated through exchanges of high-level memos and sometimes testimony at public or legislative hearings. Commitment of Managerial Time In the sites I studied, the program managers in successful collaborations devote enormous time and attention to collaboration—to facing outward from their organization rather than merely up or down within it. For example, the founder of the San Diego Teen Parent Project reports that her proudest achievement from the program's first year was its strong reputation in the community, and she continues to reserve considerable time to meet with outside organizations, including a regular monthly meeting with two other agencies that focus on teenagers. Her advice to other jurisdictions is to make sure you have linkages to other organizations first. Similarly, the program director of a teenage parent program in New York State describes her long-standing and constructive relationships with a wide variety of community service providers as key in getting services to clients, and she says that "We make it our business to know how to get in the back door." At the service delivery level as well, caseworkers in many of the programs report a great deal of time spent in contact with other providers, ranging from formalized roles as on-site liaison to other agencies and convener of "team staffings" to less structured roles in other case management programs.

OCR for page 84
Effective Services for Young Children: Report of a Workshop This finding, that successful collaborators spend a lot of time and energy on collaboration, is not generally reported in quite this form, perhaps because it seems too obvious. But it is consistent with the process-oriented and clearly time-consuming approaches to collaboration recommended by a number of sources and with the exhortation not to expect immediate results (Levy, 1989:16). It is also consistent with the idea that successful collaborations have to be good at both differentiation and integration: one way to do better at both sides of a trade-off is to invest a great deal of managerial time and energy. Role of Personal Relationships The site visits I conducted are consistent with another finding widely reported in the literature on collaboration and case management: that ''institutions don't collaborate, people do" (Toby Herr, cited in Bruner, 1990b). That is, collaboration in the sites is associated with a network of personal relationships, built over time, between the major participants. Staff in the successful collaborations make a special effort to select strategies with a personal element: for example, face-to-face meetings instead of communications by phone or memo and colocation or outstationing of workers whenever possible so that people get to know each other. Among the examples of successful techniques using this approach: In San Diego, the heads of the GAIN Teen Parent Program, the major health center for pregnant teens, and the major community case management program meet in person once a month for a formal meeting and see each other informally much more often. As a result, the head of the health center describes their relationship as "much stronger than simply referrals . . . much more personal." In Oklahoma, one local Integrated Family Services case manager says that at first he doubted the value of the face-to-face meetings of service deliverers that the state requires him to convene, because he already talked to them frequently on the phone. He now thinks the meetings are important, because "face-to-face is different." In two instances of collaboration that did not last in the sites, personal links seem to have been missing: In the site mentioned earlier with weak conflict resolution between the community nonprofit and the county social services department, the relationship did not involve frequent personal contacts at the case manager level; at the time we were there, problems were being addressed through memos between the program supervisors. In the collaboration between the family literacy program and the welfare

OCR for page 84
Effective Services for Young Children: Report of a Workshop department, there was personal contact and collaboration at the state level but considerably less at the site level. At the time we were there, the literacy teachers reported having visited the welfare department once, but most communication seemed to be over the phone about particular problems. Exchange Relationships A number of reports on successful collaborations describe what the agencies are able to do for each other: for example, school-based collaborations typically offer outside agencies access to children they could not otherwise reach (see for example Petit, 1988; Sylvester, 1990). My sites suggest that, although the particulars of the exchange depend on the substantive domains of the agencies and their local environments, successful collaborations seem often to involve skill at creating exchanges, at noticing what one agency can do for another. In Oklahoma, when an Integrated Family Services (IFS) unit was first set up in a community to work with multineed families who might be falling through gaps in the system, the IFS case managers did not simply announce they were there to coordinate other service deliverers. Instead, they spent several months learning how they could offer concrete assistance that would build their credibility, for example by developing an automated resource directory and identifying training and service gaps they might be able to fill. In Detroit, a dropout prevention program run by the department of social services demonstrated in its first two years that it could meet specific needs of the schools— assistance in keeping kids in school, a capacity for home visits and connection to families, an ability to run interference in the bureaucracy—and the schools have responded with a commitment of resources for the planned expansion. Although there are common themes to the exchanges, there is also considerable local variation. An exchange that works for one school district or family services agency will not work for another, because the agencies' needs, the key problems they are focused on, and their resources are different. One school district may be eager to serve teenage parents and want to work with the welfare department around case management services and child care for them; another district may face resistance from citizens who view teenage pregnancy as someone else's problem and may therefore be more willing to work on reaching families whose children are in middle school. The structure of Oklahoma's IFS explicitly recognizes this multiplicity of needs in the agencies with which the welfare department needs to cooperate, by leaving great flexibility to the local IFS teams to identify needs and develop particular roles to meet those needs. Programs we heard about ranged from a rural site where the biggest issue is farm families in crisis to an urban site where a local community group wants the help of an IFS team to carry out health and housing-related projects. The state IFS

OCR for page 84
Effective Services for Young Children: Report of a Workshop director viewed both projects as consistent with the IFS focus on community capacity to serve troubled families with multiple needs. Involvement of Families Several writers emphasize the importance of including those who will receive services in the collaboration. Gardner (1989), for example, argues that effective collaborations can only come out of broad-based community planning efforts. Bruner (1990b) argues persuasively for the role of collaboration between family and service deliverer as the foundation for broader collaboration. If, as I have argued above, the whole reason for collaboration in programs to serve at-risk families and children is founded in the needs and characteristics of those families, then it makes sense that effective collaborations would have to go back over and over to those families to make sure they were in fact in line with real needs. NOTES TOWARD SOME RECOMMENDATIONS Having suggested that collaboration occurs in many different ways at many different levels, that it is extremely difficult, and that successful collaborations overcome those difficulties through ingenious local adaptations, I am not in a position to offer authoritative recommendations. The challenge I have left for myself (and the group) is to identify recommendations that are not so small in scope or incremental as to produce trivial effects and yet not so sweeping as to contradict the arguments offered above about the complex and essentially local nature of collaboration. Therefore, I have tried to select recommendations that influence large parts of the service system by changing the capacities, resources, or incentives of those within it, rather than by prescribing narrow models. I have also tried to select arenas in which the recommendations of a group like the forum might carry particular weight. Based on these criteria, I propose four recommendations as a starting point for discussion: Academic institutions should develop curricula across the professional specialties for street-level service deliverers (social workers, teachers, doctors, nurses) that prepare them for collaboration, as well as inservice training for supervisors, program managers, and high-level administrators. One advantage of this recommendation is that we in academia can work on it ourselves, rather than only exhorting others to act. It also fits the multiple and local nature of the tasks involved in collaboration: rather than prescribing rigid models, we could support those involved in collaboration in developing the skills to create their own models.

OCR for page 84
Effective Services for Young Children: Report of a Workshop States and the federal government should experiment with pooled funding streams and new accountability measures. Sid Gardner has proposed that state governments (after negotiating federal waivers) could decategorize a wide range of funding streams for selected localities. That is, the states could pool the money and offer localities the opportunity to spend the whole lump sum for the purpose of improving the lives of families and children, with complete discretion over how best to spend it to meet local needs (Gardner, 1990). The conditions for this offer would be an inclusive local planning process, probably with signoff by the important local interests and outcome measures agreed to by the community. The closest thing to a real example of this is probably Iowa's experiment with decategorizing child welfare funding in selected counties (Bruner, 1990a). States and cities should experiment with partnerships in which the city takes responsibility for services to children and the state agrees to support that responsibility. More concretely, the city would convene citizens to determine what the current service picture looks like, what the needs are, and what should change; the state would agree to help with identified needs (regulatory waivers, funding shifts, training and technical assistance, etc.). This is a variant on the previous plan that is not as dependent on freeing up large amounts of money from restrictions right away. The experience of cities that have developed children's policies (like Minneapolis and Seattle) or that are currently developing them (like Cambridge, Mass.) suggests that when mayors are eager to take some level of responsibility for children's outcomes, we should encourage their interest even if they control only a small part of service delivery, because of the critical role they can play in convening and influencing a community-wide process. Advocates and policy makers within each of the relevant children's services fields should seize upcoming legislative opportunities to encourage thinking and implementation across substantive boundaries. For example, the implementation of the Family Support Act, the likely implementation of the Act for Better Child Care, and the likely consideration over the next few years of new child welfare legislation offer opportunities to think about how best to build in collaboration. CONCLUSION Collaboration is an appealing strategy for improving services to poor children and families, but it is clearly not an easy strategy. To the extent that many teachers, nurses, social workers, school principals and superintendents, health clinic directors, welfare office managers, and state agency heads are concluding on their own that they cannot do their jobs without it,

OCR for page 84
Effective Services for Young Children: Report of a Workshop the timing may be right for efforts to support and channel the stream of collaborative experiments so that they will be most effective, most influential, and most likely to last. And, if enough experiments get started, supported from enough different levels at the same time, so that communities truly change the way they serve families, then the experience of success could create a powerful momentum for further change. As Director Richard Jacobsen of the San Diego County Department of Social Services argues, there is a certain simple power to the common mission that agencies can arrive at when they realize they are serving the same clients: ''to eliminate patients, clients, offenders and increase the number of leaders, parents, students." ACKNOWLEDGMENT This paper draws in part on research on welfare agencies and services to children funded by the Foundation for Child Development. It also draws heavily on the conversations and working papers of the Kennedy School of Government's Executive Session on Making the System Work for Poor Children, funded by the Carnegie Foundation. I want to thank all the members of the executive session for their ideas, which I have probably borrowed here. I want to thank particularly those who have written down their ideas, including Mary Jo Bane, Charles Bruner, Sid Gardner, Paul Jargowsky, and Rick Weissbourd. The interpretations, conclusions, and recommendations, though, are mine and do not claim to reflect the views of other members. REFERENCES Bruner, Charles 1991a State-Level Approaches to Foster Local-Level Collaboration: Iowa's Initiatives to Improve Children's Welfare. Unpublished paper prepared for NCSL. July. 1990b Twelve Questions State Policy Makers Should Ask About Fostering Collaboration in Meeting Children's Needs. Unpublished manuscript prepared for the William T. Grant Foundation. July 31. Gardner, Sid 1989 Failure by fragmentation. California Tomorrow (Fall):20-21. 1990 The Hugo System: A Storm That Might Blow Children Some Good. Unpublished paper prepared for the Executive Session on Making the System Work for Poor Children. January. Golden, Olivia, with Mary Skinner and Ruth Baker 1990 Welfare Reform and Poor Children: Collaboration and Case Management Approaches. Draft final report prepared for the Foundation for Child Development. September. Institute for Educational Leadership no date What It Takes: Comprehensive Service Delivery Through School and Human Service Collaborations. Unpublished paper.

OCR for page 84
Effective Services for Young Children: Report of a Workshop Lawrence, Paul R., and Jay W. Lorsch 1967 Organization and Environment: Managing Differentiation and Integration. Boston: Division of Research, Graduate School of Business Administration, Harvard University. Levy, Janet E., with Carol Copple 1989 Joining Forces: A Report From the First Year. Alexandria, Va.: National Association of State Boards of Education. Morrill, William A., and Martin H. Gerry no Integrating the Delivery of Services to School-Aged Children at date Risk. Unpublished paper prepared for the U.S. Department of Education, Conference on Children and Youth at Risk. Petit, Michael 1988 Issues Surrounding State-Level Collaboration on Services to At-Risk, Preschool Age Children. Paper presented to the Summer Institute of the Council of Chief State School Officers. August. Robinson, Estelle R., and Aleta You Mastny 1989 Linking Schools and Community Services: A Practical Guide. Center for Community Education, School of Social Work, Rutgers University. Schorr, Lisbeth, with Daniel Schorr 1988 Within Our Reach: Breaking the Cycle of Disadvantage. New York: Doubleday/Anchor Press. Sylvester, Kathleen 1990 New strategies to save children in trouble. Governing (May):32-37. Weiss, Janet A. 1981 Substance vs. symbol in administration reform: The case of human services coordination. Policy Analysis (Winter):20-45. Weissbourd, Rick 1989 Public Elementary Schools: A Ladder Out of Poverty? Unpublished paper prepared for the Executive Session on Making the System Work for Poor Children. October. 1990 Making the System Work for Poor Children. Unpublished paper prepared for the Executive Session on Making the System Work for Poor Children. October.