urban institute press

Child Welfare: The Challenges of Collaboration | Chapter One


Child Welfare: The Challenges of Collaboration Cover

1. Cross-Cutting Issues in Child Welfare

Many child welfare issues cut across traditional human service agency boundaries. Child welfare agencies often have responsibility for a problem but do not have the authority or capacity to solve the problem without cooperation from other government departments. These types of concerns create frustration for frontline staff and managers not only in child welfare, but also in many other government agencies and nonprofit organizations. These concerns are costly, often absorbing staff time, duplicating services, and using resources inefficiently. Most important, cross-agency issues create injustices and engender distrust among the children, youth, families, and communities that the child welfare system aims to serve. Despite these consequences, other branches of government and the nonprofit sector are often unable, uninterested, or unwilling to work with child protective and foster care staff to solve cross-agency problems.

The book draws on research studies conducted at the Vera Institute of Justice, a 47-year-old nonprofit public policy organization in New York City, to analyze various cross-cutting issues. These studies have several purposes. They describe many topics, primarily those that cut across child welfare to include juvenile and criminal justice agencies. They show how social science research can inform managers and government officials in a child welfare world often driven by anecdotes and high-profile cases instead of empirical data. In many cases, the studies demonstrate how innovations can mitigate or solve cross-agency difficulties and improve the quality of services and justice government provides to those in need. These studies also provide the basis for a discussion of what is needed to solve cross-cutting issues in child welfare and other human services.

This book aims to inform officials, practitioners, funders, advocates, scholars, students, and the general public about cross-cutting problems in child welfare and possible solutions to these problems. This opening chapter lays the groundwork for what follows. It starts by explaining why cross-cutting issues exist and presents a typology of these problems that references many of the chapters to come.

1The second part of the chapter explains why this type of research is rare, discusses the origins of the research studies that are the basis of the chapters, and summarizes the strengths and weaknesses of the approach used to conduct these studies.

Understanding the Problem: Why Cross-Cutting Problems Exist

Many people who have worked in or encountered child welfare refer to it as “the system” and families who are the subject of child protective investigations as “system involved.” This language conjures up visions of an all-encompassing and elaborately engineered machine. A closer look at child welfare agencies and the environment in which they operate suggests fragile webs. These webs have numerous connections within and across each other, but they also have many weak and broken strands, and they need constant re-spinning to maintain their strength. Child welfare agencies, like other specialized human service bureaucracies, are often internally divided and embedded in fragmented, complex organizational structures with multiple stakeholders competing for resources and advantage. The specialization and fragmentation of service delivery creates and magnifies the issues that cut across child welfare and other agencies.

Organizational Factors

Child welfare’s primary mission is sometimes narrowly construed as keeping children physically safe, but even this limited definition connects child welfare with many other divisions of government. Children and youth must be safe in different locations—safe at home, safe at school, and safe in their neighborhoods—and responsibility for these locations falls not only on parents, but on many government staff: teachers, school safety agents, police officers, juvenile probation officers, and child welfare workers. In every jurisdiction, government staff from different departments must work with parents and each other to ensure child safety. Expanding the mission to ensuring children’s physical, educational, financial, and mental health involves that many more agencies.

Coordinating any two complex systems is a challenge. Government departments are specialized—they focus on one set of services—and they employ large numbers of people. Even in medium-sized counties, child welfare, education, and criminal justice agencies each have dozens of employees focused on their agency’s particular mission. In major urban centers, human service agencies have hundreds or even thousands of employees. Because most agencies face an almost limitless demand for help, these departments develop procedures that organize their employees to focus on their core mission. Agency managers usually try to prevent staff from straying into the specialization areas of other agencies because doing so takes away resources from the core mission and risks offending the other agency.

Specialization, however, is hardly the sole origin of cross-agency problems. Competition and culture clashes prevent collaboration between any two agencies, and incorporating additional agencies adds exponentially to the obstacles such efforts encounter. The byproduct of a decentralized government, these challenges often lead to bewildering laws, regulations, policies, and court decisions that further complicate developing cross-cutting solutions.

The executives and managers of government agencies, including those involved in human services, often view other agencies as competitors for scarce resources. They may vie with each other for budget authority, for the attention of elected or appointed officials to their policy positions, or for the lead role in important new initiatives. At the same time, they may jockey to avoid budget cuts, the addition of onerous responsibilities, or “turf” encroachments by other agencies. By their nature, attempts to coordinate activities involve ceding some authority and control to another agency as well as complicating the lines of responsibility—activities that executives often try to avoid. The short tenures of many people who work in child welfare, from senior managers to frontline staff, undermine the development of personal relationships that can mitigate conflict between organizations.

Every bureaucracy, moreover, has its own culture that often differs from and may conflict with those of other departments. Child welfare agencies, for example, are dominated by trained social workers. Social work training emphasizes empathy and resolving conflict through negotiation. In contrast, police departments are paramilitary organizations authorized to resolve conflict with force if necessary. Social workers and police often use different language, have different values, and may come to contrary conclusions when presented with the same facts. Even when the leaders of police and child welfare organizations agree on a course of action, coordinating the staff of two organizations with such dissimilar orientations is a challenge.

Competition among local agencies and conflicting agency cultures are major reasons for cross-cutting issues. Every agency involved in these issues, however, is embedded in a bewildering set of demands created by federalism—the division of responsibilities among federal, state, and local governments which are themselves divided into executive, legislative, and judicial branches. A quick look at the stakeholders involved in child welfare serves as a good example, though a similar description can be created for law enforcement, juvenile justice, education, and most other government departments.

The Role of Federalism

Child welfare authorities have an awesome power—the right to separate children from their parents, in some cases permanently. High levels of government should and are involved in the regulation of this power. Sitting on top of the governmental pyramid is the federal Administration for Children and Families, which is part of the Department of Health and Human Services. Federal officials work to ensure compliance with federal child welfare laws and distribute a significant portion of child welfare funding. For example, the federal government pays for half of all child welfare placement costs, and children in foster care are automatically eligible for Medicaid health insurance.

In most of the country, state government plays a dominant role in child welfare policy. In 40 of 50 states, child welfare is state administered—meaning most child welfare workers are state employees who are hired by and report to state officials. These officials ultimately report to a commissioner appointed by the governor. The other 10 states are state supervised—meaning county or municipal child welfare staff are hired by and report to local agencies whose commissioners are appointed by local elected officials. Even in state-supervised states, however, state officials have the responsibility to hold localities accountable for following state laws and regulations.

Child welfare policy also involves the legislative and judicial branches of the various levels of government. Congress has long played a role in establishing the legal framework and funding of child welfare. Federal laws, including the Child Abuse Prevention and Treatment Act of 1974 (CAPTA) and the Adoption and Safe Families Act of 1997 (ASFA), establish the broad legal parameters of child welfare policy.

2In the many policy areas without federal mandates, state legislatures influence child welfare practice by passing laws, providing funding, and conducting oversight hearings, as do city councils in many large urban jurisdictions. Family Court judges must review and approve significant interventions—such as child removals or mandatory services. In many jurisdictions, advocates have used the courts to advance their causes by suing child welfare agencies over specific practices or failure to abide by laws and regulations. These suits may be filed in state or federal courts. Court decisions have established case law that influences child welfare policy.

The fragmentation of the child welfare field extends beyond government. More than many other child- and youth-serving agencies, child welfare relies heavily on the nonprofit sector. Nonprofit organizations, often including those with religious roots, provide the bulk of preventive services and foster care placements in many jurisdictions. Other nonprofits vigorously advocate for child welfare issues. Public employee unions are also important stakeholders in many areas. While unions, nonprofits, and advocacy groups cannot make policy in the same fashion as government, each can have substantial influence through contract negotiations and lobbying activities.

The diffusion of power in the child welfare field and other human service agencies further complicates coordination. First, keeping up with the different laws, regulations, and court decisions emanating from various levels of government means that child welfare officials rarely have time to deal with issues that may be perceived as outside their core mission. Second, any new initiative usually has winners and losers—and with power so diffuse, those with an interest in the status quo have many options available to them in trying to derail changes they see as detrimental or onerous.

While unrelated to federalism or organizational behavior, a final factor deserves mention. Child welfare agencies are often internally fragmented. Units responsible for child protection, for example, decide whether families are left alone, receive preventive services, or have one or more children placed in foster care. Decisions made by child protective workers ripple through other parts of the agency, affecting budgets, staffing, and workloads, and creating internal divisions among units. In addition, staff inside child welfare agencies often have deep and passionate differences toward the work. Some workers firmly believe that child welfare agencies should work primarily to preserve families through preventive services, while others believe that child welfare should focus primarily on removing children from dangerous situations. These ideological cleavages often exacerbate divisions along race, religion, and class lines among child welfare staff.

In sum, child welfare agencies are shaped by and interact with networks of stakeholders that include departments within each branch of local, state, and national government as well as unions, advocacy groups, and nonprofit organizations. They are rife with internal divisions. The same point applies to the other agencies that work with children and youth, including schools, police, and juvenile justice. Federalism’s checks and balances have many advantages: democratic accountability, oversight, forums for debate, flexibility for adapting to local conditions, and other benefits associated with American government. But the fragmentation of service provision and the multiple nodes of influence also contribute to the development of cross-agency issues. As we will see, when such elaborate systems have clients, responsibilities, or problems that overlap, coordination is the exception, not the rule. The typology below provides a more concrete understanding of how these issues develop.

A Typology of Cross-Cutting Issues

The Tunnel Problem, or “Problem Solved!”

A child or youth’s entry point into the series of systems that serve this population, not the youth’s underlying problems, usually determines how government responds. Each of the many systems that serve youth has fixed services or solutions to offer. Because most agency staff think primarily of the solutions within their system, they usually send youth down one of these “service tunnels.” The tunnel may be the most appropriate choice among the agency’s options, but it may or may not be an effective course of action. Once a young person starts down a particular tunnel, it is often hard to reverse course and take a different path.

A clinically depressed youth who often skips school and occasionally self-medicates with marijuana, for example, will receive markedly different treatment depending on which tunnel she enters. If the teen receives a referral to the mental health system, she will likely receive some counseling. The school system, however, may demand that the parent file a status offense petition to address the truancy problem (chapter 7 discusses cross-cutting issues in the status offender system). This may result in services that focus on truancy or other educational issues, or merely a warning from a probation officer. Depending on the relationship between the youth and his or her parent(s), a status offense case might lead to a court appearance and placement in foster care. School officials might bypass the status offender system and file an educational neglect petition that triggers a child protective services investigation. If police or school safety officials catch the youth smoking marijuana, she will enter the juvenile justice system. Juvenile justice officials may end up putting the youth on probation, sending her to an alternative program, or ordering her to “placement”—usually a secure facility that is essentially a juvenile prison.

The tunnel problem can lead to perverse and unintended consequences. In one case, police arrested a mother and her early adolescent daughter working together to sell marijuana.

3Though the daughter might be seen as a child welfare case, the arrest sent the daughter into the juvenile justice tunnel. Juvenile justice officials felt uncomfortable giving the daughter probation because of her unstable home life and recommended that she be sent to placement. The judge concurred and sentenced her to a secure placement for 12 months (the minimum time for placements according to New York State law). In contrast, the mother received a 30-day jail sentence.

In another example, a study of the implications of a new status offender law finds that several New York City schools routinely threaten to file educational neglect charges if a parent does not file a status offender petition (Souweine and Khashu 2001). Yet a study on the educational impact of child welfare placement shows that school attendance rates for status offenders who enter the child welfare system decline (Conger and Rebeck 2001). Another study finds that over half of all status offenders placed with child welfare leave after two months, and 90 percent of those who leave return to their families (Ross, Wamsley, and Khashu 2001). These quick turnarounds mean that youth do not have time to form bonds or mentoring relationships with child care workers. Indeed, child care workers have more of an incentive to invest their time and energy working with other youth, leaving status offenders without appropriate services targeted to their needs. This may be one reason status offenders leave care without permission at far higher rates than youth who enter foster care for other reasons (Finkelstein et al. 2004; Ross et al. 2001). Attempting to resolve a truancy problem may lead youth down a tunnel that exacerbates the issue and creates further obstacles to his or her healthy development.

Any discussion of the services that youth receive would be incomplete without highlighting issues of cultural competency and institutional racism. Youth of color, especially African Americans, are more likely to receive harsher treatment when involved in school discipline proceedings, child welfare cases, or the juvenile justice system.

4Indeed, widespread disproportionate minority confinement (DMC) led to a federal mandate requiring states to assess the extent of DMC and take steps to address the problem. Tunneling, then, is not only a function of a youth’s presenting problem, but is also influenced by conscious and unconscious biases on the part of individuals and institutions.

These examples demonstrate how tunneling can lead to government responses that are either arbitrary or inappropriate. And once youth enter a particular tunnel, the fractured systems that serve youth may cause additional problems for both children and agencies. The constricted flow of information among agencies, including child welfare, often prevents the identification and management of cross-cutting issues.

Information Flow, or “I Didn’t Know about the Problem”

Many cross-cutting issues revolve around the difficulties of sharing information across agencies. Agency databases rarely “talk” to each other for bureaucratic, legal, budgetary, and technological reasons, and this lack of communication is reflected in the practices of caseworkers and other staff in different agencies. Many agency managers are eager to solve cross-cutting issues once they identify and understand them. But the fragmentation of policymaking and service delivery means that these issues rarely rise to a managerial level where they can be addressed.

Project Confirm, an effort to bridge the gap between the child welfare and juvenile justice systems discussed in chapters 8 and 9, arose out of a common feeling among New York City Family Court judges, juvenile justice officials, and child welfare staff that the lack of a release resource caused arrested foster kids to be unnecessarily detained in juvenile detention facilities. None of these actors knew the extent of the problem. Initial research showed that foster youth accounted for 2 percent of all youth in New York City but 15 percent of detained youth, even though foster youth were not arrested for more serious delinquency charges than their nonfoster peers (Conger and Ross 2001; Ross and Conger 2002). This information persuaded officials to take action that led to the creation of Project Confirm: a relatively simple intervention that eliminated “the foster care bias” in detention decisions for low-level first-time delinquents.

Information-sharing issues may also complicate efforts by law enforcement and child protective workers to coordinate responses to severe child maltreatment, a challenge discussed in chapter 2. Child protective workers, police, and district attorneys are all trained to share information only when necessary, as much of the work they do involves confidential data that are either protected by law or whose release might compromise investigations, prosecutions, or a third party’s safety. In many instances, however, these agencies legitimately need information held by their counterparts, especially when investigating the same situation. For example, child protection investigations may benefit from checks on prior domestic violence convictions, while police investigations may benefit from learning of prior indicated child welfare cases. Without memoranda of understanding or other protocols for information sharing, the missions of law enforcement and child welfare agencies are compromised.

Information-sharing challenges make it difficult for child welfare and corrections agencies to do their jobs as well, as discussed in chapter 4. In many states, parents have the right to visit with their children in foster care even when they are incarcerated. Yet information sharing between child welfare and corrections rarely takes place—resulting in case workers not knowing that parents are incarcerated and often assuming that they are unresponsive to case planning efforts. Children in care may feel abandoned because they have not seen their parents, and parents cut off from their children often experience frustration and anxiety. Advocates, children, and inmates report that the lack of communication can lead to destabilizing behavior.

5Diffusion of Responsibility, or “It’s Not My Problem”

Information sharing alone rarely ameliorates cross-agency problems. Even when agencies know that they have a common client, tension around appropriate roles and responsibilities—especially in times of high caseloads—often results in a lack of follow-through by frontline staff and their managers. When responsibility is diffuse, kids fall through the cracks.

In Project Confirm, many child welfare caseworkers presumed that once police arrested a foster youth, their responsibilities ended because “the kid belonged to juvenile justice.” To solve this problem, Project Confirm routinely faxed a letter to caseworkers and their supervisors from the commissioner of child welfare reiterating caseworkers’ responsibility to go to court on delinquency cases. Program planners made numerous presentations to agency staff and foster care providers to educate them on the overlap problem and their responsibility in resolving it.

Status offender systems also encounter diffusion of responsibility problems, examined in chapter 7. Status offenders are youth who have not committed delinquent acts, but their behavior concerns government because of their status as minors. Status offenses include running away from home, not obeying parents, truancy, and drinking alcohol. Child welfare staff may see youth running away from home as a missing persons issue that ought to be handled by the police, while police may view such events, especially repeat runaways from the same home, as a child welfare or juvenile probation issue. The status offender “system” is often an office with relatively little service capacity of its own; thus, it relies on other agencies, including the Family Court, to make decisions for youth and to provide services. Staff in each organization may also look to the parent to take responsibility for the youth’s issues as well as city agencies or their contracted service providers.

Similarly, responsibility for the educational performance of foster children is spread among many people, including foster parents, biological parents, group home staff, caseworkers, and teachers. Even when information is shared and each player knows about a youth’s status, confusion over roles and responsibilities may result in no one working with a foster youth on educational issues (Finkelstein, Wamsley, and Miranda 2002). After officials have assigned responsibilities, successful implementation requires that frontline employees are trained to carry out their assignments. Many people working with kids in foster care are eager to help after their roles are clarified, but they may be inexperienced in navigating the educational system or intimidated by working with teachers and school officials. To counter this problem, some programs use “train the trainer” sessions and develop materials for caseworkers that explain how best to engage the school system (Armstrong and Mandelstam 2003).

Unloading Cases and Shifting Burdens, or “It’s Your Problem Now”

As the examples above show, officials and staff often want to solve cross-agency problems when they have the tools to do so. In some instances, however, agencies, staff, and parents act to rid themselves of troubled youth. Hard data on this phenomenon are rare, but there are enough anecdotes and stories to make this pattern noteworthy.

While planning Project Confirm, program designers heard many stories of foster care staff who called police to have youth in their care arrested—often for minor incidents that biological parents resolve without police interference. Indeed, some frontline staff see calling the police as a way to assert authority or to have responsibility for a kid transferred to juvenile justice.

6In some cases, these youth may have suffered from mental health problems, but the lack of access to mental health services led to voluntary placements with child welfare. According to the Rochester Youth Study, one in five male delinquents and one in three female delinquents has a diagnosable mental health problem (Huizinga et al. 2000).

The Family Courts in New York City place hundreds of status offenders in traditional congregate care settings filled mostly with youth in foster care (Ross 2001; Ross et al. 2001). Courtroom observations found that if parents demanded placement—in effect, unloading the youth onto the child welfare system—judges almost always acquiesced. Focus groups with parents showed that they usually cared about their children and thought placement was a “boot camp” where their kids would learn discipline, though no such facilities exist in New York.

Given the mismatch between what parents want, what kids need, and what congregate care facilities provide, it is not surprising that child welfare discharges status offenders quickly. Status offenders cycle back into foster care more often than other foster youth as neither parents, nor the status offender system, nor the child welfare system want to take responsibility for these problem cases. In the worst cases, kids bounce between home and placement, and bounce around within the child welfare system—often having to transfer to new schools with each movement. And, like youth with delinquency records, they may face extraordinary hurdles in their attempts to register in a new school despite regulations that require officials to enroll these children.

Detained youth in New York City are automatically enrolled in an in-facility educational program. Attendance is mandatory. Ironically, students taking courses while in detention receive only half credit for their efforts. Many court-involved youth are already behind educationally, and this policy makes catching up that much harder. Even if delinquent youth manage to register at a school, school policies and administrators may push a youth to drop out (Fine 1991).

Not surprisingly, a review of case files in New York City and other studies finds a strong link between poor school attendance and recidivism. In essence, the education system is unloading problem cases onto the delinquency system. Youth in foster care who have established records of behavioral problems face many of the same challenges, bouncing from placement to placement and school to school.

The examples in this typology show the difficulty of addressing cross-cutting issues. Many professionals in child welfare and other fields have recognized the bureaucratic behavior that creates or amplifies these problems, and some may be familiar with the specific problems mentioned. Given the harm to children, youth, and families as well as the cost and the injustice of cross-agency problems in child welfare, it might be thought that a well-developed research literature describes these issues and some solutions. Such is not the case. The following section discusses why relatively little research on cross-cutting issues exists, describes how the research in this book originated, and discusses the strengths and weaknesses of the approach used.

Studying Cross-Cutting Issues

Researchers face serious challenges in conducting studies of cross-agency issues in child welfare and youth services generally. Research is expensive. Many funders, including government agencies and the legislators that appropriate their budgets, face pressure to fund programs rather than research. Gaining access to child welfare data is and should be challenging—information about abuse, neglect, family health, and relationships is extraordinarily sensitive and legally protected. Many of the same confidentiality protections also apply to data collected by justice system agencies—broadly defined to include police, prosecutors, juvenile justice, corrections, courts, and probation—that are the focus of this book. The quality of child welfare and other administrative data is often poor. Research involving foster children, juvenile delinquents, and incarcerated populations draws extra scrutiny from government agencies and institutional review boards.

7Research designs that use some form of random assignment of research subjects are rarely approved and difficult to execute in the child welfare and justice fields.

These obstacles make conducting research in any one arena—be it child welfare, juvenile justice, or some other justice field—difficult. Studying cross-agency issues exacerbates these difficulties. Instead of persuading officials in one agency that a study is worthwhile, researchers must persuade two or more sets of officials about the value of the study, develop two or more data-sharing protocols, learn the intricacies of two or more datasets, and gain expertise in additional policy areas. The time this takes adds to the cost of the study, and many skills needed for this work fall outside the bailiwick of researchers and into the domain of lawyers and policy entrepreneurs. For academically based researchers, policy-oriented studies that include areas outside the tradition of the discipline receive few rewards.

Taken together, conducting cross-cutting research on child welfare issues presents challenges that most researchers avoid due to the cost, the aggravation, or the lack of capacity. Funders may also be reluctant to invest in expensive research that has a greater risk of failure. The research presented in this book took place in a setting designed to tackle many of these obstacles.

A Model for Cross-Cutting Research

The New York City Administration for Children’s Services (ACS), the city’s child welfare agency, paid the Vera Institute of Justice to conduct most of the research presented in this book.

New York City created ACS in 1996 in the wake of widely publicized child deaths and widespread concern about the child welfare system’s ability to keep the city’s children safe. The city split the child welfare agency from a larger agency, the Human Resources Administration, with the hope that a higher profile, a narrower focus, more resources, and direct access to the mayor’s office would enable improvements to the long-troubled agency.

8Since its inception, ACS has received funding to support external research that helps implement and evaluate reforms and informs other issues that the agency encounters. Most of the research in this book is a result of this initiative, though none of it could have been accomplished without the assistance and cooperation of numerous other agencies.

The Vera Institute of Justice (Vera) is a New York City–based nonpartisan nonprofit organization that conducts research, plans and operates programs, and offers technical assistance on social justice issues. Established in 1961, Vera serves as an incubator for innovation and a source of new ideas to make government more fair, humane, and efficient. Since its inception, the Institute has launched dozens of demonstration programs and spun off 17 nonprofit organizations. Each nonprofit continues to serve vulnerable populations, such as crime victims, inmates, parolees and probationers, immigrants, court-involved youth, the homeless, and families involved with child welfare. These programs usually arise to fill in gaps in services or to tackle unaddressed problems.

Vera’s research applies transparent social science methodologies to public policy issues. The Institute employs researchers from a range of disciplines that have included sociology, public policy, experimental and clinical psychology, political science, criminology, anthropology, and history. The capacity to draw on diverse substantive and methodological expertise for internal peer review helps researchers produce studies that have a reputation for credibility and independence. Vera values its “just the facts” reputation and has a track record of producing reports that present both positive and negative aspects of programs and processes.

The different disciplines and orientations at Vera reduce the barriers to conducting cross-cutting research studies. Researchers have ready access to experts—both researchers and program staff—who know the issues facing justice agencies; the organization, strengths, and weaknesses of administrative data; the details of memoranda of understanding for sharing data; and the people at agencies with the knowledge or authority to “get things done.” Lawyers, researchers, and institutional review board members know the issues in conducting research involving child welfare–involved families and incarcerated populations.

Vera’s New York City location also helps reduce the barriers to conducting cross-cutting research. The city’s budget exceeds that of any other jurisdiction in the country save

New York State, California, and the federal government. Combined with the city’s many foundations and history of philanthropy, New York City can support research that often exceeds the financial capacity of other jurisdictions.

New York’s strong progressive tradition, with its emphasis on using government to address social ills and rational public decisionmaking, provides an impetus for research that is not always present in other jurisdictions.

Vera’s research model addresses a primary problem in conducting sensitive research that involves multiple vulnerable populations: trust. Trust on the part of clients, frontline staff, and agency managers can help solve numerous problems that might sink a research project otherwise. The Institute’s research model builds trust in many ways.

A distinctive element of Vera’s approach to research is creating open partnerships with government. Engaging researchers is a leap of faith for government agencies, as evaluators usually need access to nonpublic data, use up valuable staff time, and enter facilities that are often not open to the public. To help gain the trust of government officials, researchers actively solicit the input of government staff and may ask government managers to guide research questions to ensure that studies will be useful to them, not just the research community. The premise is that government officials are more likely to support and act upon research if they understand how conclusions are made and feel they are part of the work the research team undertook. Finding individuals in government who have political capital and a track record of acting on empirical findings further increases trust at all levels.

Vera’s research model strives to incorporate four sources of information in each study: administrative data collected by government agencies, conversations with government and nonprofit managers that implement and direct programs, discussions with frontline staff, and interviews with clients or their representatives. Systematically combining multiple data sources produces new knowledge, and taking this approach helps build trust by assuring different stakeholders that their voices have played a role in the research.

9Vera is uniquely positioned to conduct research on cross-cutting issues in child welfare and other justice issues. Even with these advantages, however, there are limits to the type and sophistication of the research produced. In addition, the Vera research model has strengths and weaknesses that readers should keep in mind when examining the ensuing chapters.

Strengths and Weaknesses of the Model

The strengths of this approach to research include access, transparency, and practicality. Access includes not only access to administrative data that are not commonly available, but also access to government staff, facilities, and clients. In most cases, Vera researchers have access to staff meetings and other nonpublic events that allow a fuller understanding of agency operations. Transparency refers to the use of replicable social science methods. For example, the findings from the cohort study methods described in chapters 3, 4, and 5 could be duplicated by government researchers or others with access to the same data. Another strength of the model is practicality: by taking an empirical approach, Vera’s research aims to draw lessons that government officials can use in carrying out their responsibilities.

The model also contains some weaknesses—weaknesses that are often the flip side of strengths. The price of access is that government agencies usually retain the right to approve reports before they are issued publicly—though often on the condition that approval is not unreasonably withheld. Agencies insist on this right because they need to ensure that no individually identifiable, confidential information is released. In the Institute’s experience, the right to approve reports has rarely been used as a way to influence the content of the report. Most government officials know that such efforts will be vigorously resisted and that biased research has less substantive and political value. Still, Vera researchers do not have the unfettered right to publish that university-based researchers enjoy.

Another perceived weakness of the model is the focus on one city. The research is limited to one city because of funding constraints and because all the difficulties of conducting cross-cutting research would be that much more complicated by researching multiple jurisdictions. It takes years to understand how government organizations in one jurisdiction really work, how and why one part of government interacts with another in a particular way, and the strengths and weaknesses of the administrative data that each agency collects. Learning the same information in another jurisdiction or investing in a partnership with another organization further complicates an already complex undertaking.

This drawback is offset by the depth and detail of the research. And, New York is an extraordinarily diverse city. The city’s five boroughs are built at different population densities, have demographically distinct populations, and have varying access to resources and opportunities. Borough offices serve client populations larger than most other cities in the country and in many cases have developed distinct organizational cultures.

10These differences give the research the flavor of multicity studies in some cases, such as the differences in court culture and probation case processing noted in chapters 8 and 9. Most important, the origins of cross-cutting issues cited above are not unique to New York. Other urban child welfare systems exhibit much of the same complexity as New York and are embedded in the same system of service specialization and federalism.

A final critique is that the model produces studies that rarely address broader theoretical understandings of social justice issues. Indeed, there is little discussion of such issues as the socioeconomic structures that contribute to the placement of children in foster care, racial and gender bias, or the social control functions of the criminal justice system in a capitalist society. These are important issues that readers should keep in mind. The technocratic nature of the work, however, often makes the discussion relevant to many areas of public administration. All governments, regardless of their ideology or the demographics of the populations they serve, must deal with coordination and overlap among different bureaucratic units. The research in this volume contributes to our understanding of how these issues play out in numerous ways, whether in American urban jurisdictions, U.S. suburbs and rural areas, or other countries.


Cross-cutting issues in child welfare lead to many problems. These problems have their origins in the organizational structure of providing human services and are made more challenging by the fragmentation of American government and policymaking. Most of these problems can be classified in one or more types that include service tunnels, information flow, diffusion of responsibility, and unloading problem cases. Despite the inefficiencies and injustices that cross-cutting issues often create, the cost and complexity of studying them has limited the research available to guide policymakers. This book presents unique research that informs policymakers about specific cross-cutting issues in child welfare. Despite this sobering introduction, these studies show that in many cases, cross-agency issues can be resolved.

Readers of this volume, especially those not familiar with New York City, are strongly urged to read the appendix to this chapter to have a better sense of the government agencies discussed in the book, their responsibilities, and the data they collect.


Child Welfare: The Challenges of Collaboration, by Timothy Ross, is available from the Urban Institute Press (ISBN 978-0-87766-756-8, paper, 268 pages, $29.50)

2011 Urban Institute | Contact Us | Privacy Policy