Charting a New Path for Child Welfare

(Originally published June 2015)


The movie director, Stanley Kubrick, once said to an interviewer that “the only significant question in life is what to do next.” This sounds like an exaggeration even for creative artists with seemingly limitless possibilities, but it’s surely an important question for persons and organizations in extreme circumstances. Child welfare systems that have been discredited by a steady stream of high profile child deaths and other perceived failures of child protection have frequently had multiple reform initiatives and changes in leadership, as well as system reviews and evaluations of all types, none of which have made much of a positive difference and may even have made delivering effective services more difficult for caseworkers and supervisors.

What should be done by public agencies in these circumstances?


Typically, newly appointed leaders of state and county child welfare agencies want to engage in highly visible programmatic initiatives that will have immediate positive effects on the agency’s CPS program, and will convince policymakers and advocates that the child welfare agency is on the right track. Leadership teams may undertake intensive consideration of widely used assessment tools, practice models,

family meeting formats and training initiatives with the goal of adopting best practices in child protection. Legislatures will often provide modest increases in casework positions and support staff to reduce workloads in the first stages of these reform initiatives. This is a scenario that has repeatedly failed to make lasting changes in child welfare systems. Nevertheless, new leaders who have little if any understanding of the history of child welfare reform initiatives persist in doing what their predecessors have done because of the need to demonstrate immediate improvements and because of the inability to imagine a different course of action.


Child welfare systems in extreme circumstances characterized by overwhelming workloads, annual turnover rates in key positions of 30-50%, low salaries or frozen salaries undermined by inflation and decreases in benefits, vacancies that frequently cannot be filled and a demoralized workforce that has given up on policymakers’  ability to appreciate the reality of child protection are unlikely to benefit from programmatic initiatives that do not address the needs of the workforce: greatly reduced workloads, administrative support staff and case aides to assist with clerical tasks and transportation, increased salaries (especially top- out salaries) opportunities for professional development that go well beyond one-off training programs, a greater voice in both policy development and in implementation planning, a ‘safety’ culture that protects staff who must make life and death decisions without benefit of 20/ 20 hindsight, and leaders who can be trusted to tell the truth and listen attentively to what practitioners have to say on a regular basis. Until public agencies in extremis make these investments which will likely take 5-10 years to pay off in major ways, programmatic investments are likely to have little or no lasting effects, “dust in the wind” which will be quickly forgotten in most jurisdictions.


Fortunately, some child welfare systems are not in this extreme condition; their workforces are better paid, better educated and more experienced, and turnover rates in key positions are no higher than 10-20% per year. Programmatic initiatives have the potential to improve child protection and out-of-home care services in these agencies. However, because most child welfare systems have recent histories of implementing too many initiatives, these agencies are more likely to benefit from a reduction in assessment tools and practice models rather than adding something more to current requirements. When engaging in for real (as opposed to show and tell) child welfare reform, more is not better. Implementing a single practice model well is a better strategy than stacking one model on top of the other in policy and procedural manuals. State and county child welfare systems that have stuck with

imperfect practice models through leadership changes and periodically recommitted to improving their implementation of these models have had more success in my view than jurisdictions that have continued to layer one more tool/ model on top of others.


Nevertheless, any child welfare system which makes significant programmatic improvements will need to also invest in workforce development to sustain these reforms. Every child welfare system needs more expertise at the unit level in substance abuse, mental health, domestic violence, early childhood development,

Indian child welfare, cultural competence, and other subjects as well. Investments in certification programs that increase the availability of expert practitioners with specialized knowledge is an affordable way of developing expertise where it counts the most— in day in day out decisions at the unit level. In addition to funding these 90-100 hour programs, child welfare staff who have or acquire these certifications should receive modest salary increases of 5-10%.


Child welfare agencies also need an elite group of investigators who can do far more than apply standardized assessment tools and models to easily identified abuse and neglect cases. Some of the most distressing failures of child protection in recent years have involved the apparent inability of CPS investigators to recognize instances of starvation and torture of children, combined with grotesque emotional maltreatment, despite multiple opportunities. Child welfare systems that have implemented differential response systems in recent years have been far more interested in developing their staffs’ parent engagement skills than in honing the forensic skills of investigators. However, both skill sets are needed in child protection. Developing a higher level of investigative expertise will require an intentional effort to create the capacity for holistic pattern recognition, a talent which requires repeated exposure to common patterns in child protection combined with expert commentary regarding what these patterns mean. At this point, few child welfare training programs even grasp the concept of pattern recognition or its relevance to child welfare, much less how to train staff in its development.

 Joining Research and Practice in R&D Sites


One of the most curious features of U.S. child welfare systems is their commitment to statewide implementation of assessment tools and practice models despite wafer thin evidence that they will result in system wide improvements. In my experience,  child welfare leaders often have dismissive attitudes toward pilots of promising practices, which may be viewed as an indicator of weak commitment and indecision. The need to act decisively in the midst of crisis combined with a credulous acceptance of the overstated claims of developers and advocates regarding “evidenced based” and “evidence influenced” claims has repeatedly led to statewide implementation of practices that have had little or no effect on hoped for outcomes.

Statewide initiatives may require huge investments of limited resources and have  large opportunity costs that continue for years. Furthermore, once these models and “best practices” are implemented (a process that may take 2-3 years at least), it is unlikely in most states that they will be carefully evaluated. Child welfare leaders who have “bet the bank” on practice models or specific evidenced based practices are often disinclined to discover that outcomes have not improved.   


One main reason state and county child welfare leaders depend more on word of mouth of their peers and the opinion of early adopters than on research in choosing practice models, assessment tools and other “best practices” is that the federal government and state governments have not made sustained financial investments in practice based research; as a consequence child welfare managers may resort to shopping for evidenced based practices off of authoritative websites or pay consultants to tell them what to do in a particular area of practice. This is another area in which child welfare policy and practices have not improved in recent decades. There was a time in the early history of modern child welfare systems when the federal government made large ongoing commitments to a child welfare research agenda, but those days are long past. Foundation grants and sporadic federal funding of research projects have supported child welfare research to a limited degree, but to a remarkable extent state and county child welfare systems gravitate to practices already adopted by their peers in other jurisdictions independent of evidence of results. Fundamentally, U.S. child welfare systems have not organized themselves to learn from their vast experience with children and families and to develop knowledge that can be used by decision makers at all levels ( including courts), for example be developing strong partnerships with universities and other academic entities.


This state of affairs must change. The federal government should begin to reimburse states for the cost of practice based research in the same way that child welfare training is reimbursed with IV-E funds. In cooperation with academic entities, state and county child welfare systems can develop research and development sites for the testing and evaluation of programs and practices, perhaps by creating one urban based site and one rural site which will allow comparison of process indicators and outcomes in R &D sites with similar sized offices conducting business as usual.


The idea that selected child welfare offices could become research laboratories part of whose mission is to test promising practices and evaluate their outcomes seems not to have occurred to either researchers or child welfare managers. Some researchers have been blocked from conducting research on open child welfare cases, and child welfare managers consider adapting to research protocols a nuisance or ethically questionable. However, once child welfare managers understand the power of continuing partnerships with researchers interested in child welfare practice, obstacles to comparison group research in child welfare settings will be easily overcome. Someone will have to pay for this partnership between researchers and practitioners (and the federal government needs to fund a large percentage of the costs), and there must be enlightened child welfare managers who understand the potential benefits of practice based research. Neither of these conditions will be easy to satisfy, but once even a few child welfare systems demonstrate the value of researcher/ practitioner partnerships, other systems will seize the opportunity to become active participants in knowledge development.

© 2020 by Dee Wilson Consulting. Proudly created with