Creating Learning Organizations

(Originally published February 2015)

A model of child welfare reform that has gained wide acceptance among policymakers and advocates begins with strong leadership. In this model, a new child welfare director brings both vision and organizational capacity, quickly wins the trust of subordinates, the Governor or Mayor, key legislators and stakeholders, and energizes demoralized staff at all levels to recommit to the agency mission, as well believe in their talents and capacity to make dramatic improvements in policy and practice. Within a few months, a child welfare agency depicted in media accounts as inept and clueless begins to make rapid progress toward some measurable goals, e.g., reducing caseworkers’ workloads, completing permanent plans, shrinking the foster care system, increasing family support services, implementing a safety oriented practice model. In a few instances, exceptional leaders have maintained and built on these early gains so that within 5-10 years the state or county child welfare system has been transformed into a system widely viewed as effective and innovative.


In other instances, strong child welfare leaders have been replaced within a few years by managers who continued to build on successful initiatives, albeit in a critical and creative way, for example in Illinois following Jess McDonald’s departure. However, in a large number of states, reform agendas advanced by recently appointed directors and their management teams have run out of steam before new policies and practices have been fully implemented, either because these reforms lack a critical mass of internal “champions”, or because they are inadequately funded, or because other challenges and initiatives have led to a loss of focus and an attempt to do too much with too little “buy in” and persistent effort. All too frequently, an infusion of new leadership effects organizations in the way that steroids affect health, i.e., a temporary increase in energy and strength is purchased at the cost of long term damage to well-being. The repeated failure of system wide initiatives to improve child welfare agencies leads to cynicism and a deeply ingrained belief among both agency staff and advocates that “transformative” reforms are impossible due to organizational culture, resource deficits or for a variety of other reasons.


In my view, the single most common reason that reform initiatives fail is because long standing workforce issues are ignored or dealt with in superficial ways. Recently appointed child welfare leaders can usually convince governors and legislators to give the child welfare agency a modest number of new positions, but not enough positions to bring workloads to manageable levels. However, in recent years some legislatures have given child welfare agencies hundreds of new positions without these enhancements achieving any discernable effect on workload demands. If a child welfare system has an annual turnover rate of 30-50% in CPS investigators, or in newly hired caseworkers, then large staff enhancements may not reduce overwhelming workloads because of high vacancy rates. No set of practice improvements is likely to improve agency performance in these circumstances. Practice models, whatever they are, are unlikely to be implemented as designed in agencies which lose one third to one half of their casework staff annually.


The Stephens Report, commissioned by the Texas child welfare system and released in 2014, addresses workforce issues in an unusually forthright and insightful way. The report asserts that “direct care workers need to identify CPS as a place to stay and grow

if they are going to remain employed with the agency ...” With this goal in mind, the authors recommend:


  • Maximizing the time spent between field staff and children and families

  • Creating an environment where staff want to come and grow their skills

  • Building a culture of excellence that is focused on quality, and

  • Empowering staff with the responsibility, tools, and accountability for success.


The Stephens Report recognizes that creating work environments in which casework staff want to stay and develop their skills requires more than reducing workloads to reasonable levels, and increasing salaries and benefits, as essential as it is to do these things: child welfare agencies must be managed differently, and in ways that generate commitment to a challenging and dangerous profession.


Creating Learning Organizations


Child welfare agencies have vast amounts of experience working with high risk children and families. Why do these agencies have so much difficulty learning how to protect children and strengthen families? Why aren’t agencies steadily improving their policies, practices, information systems, performance indicators, training programs, in-home services and safety planning, foster homes and residential care facilities? Why do so many agencies around the country seem to need fundamental reform initiatives every few years? These are questions whose answers have multiple dimensions, but there is one

overarching answer: child welfare agencies utilize a managerial paradigm that provides a poor soil in which to grow a learning organization. The main elements of this paradigm (which is rarely questioned) include:


  • Top-down approaches to system change

  • Steadily expanding prescriptive frameworks that fill hundreds of pages of policy and procedural manuals and that attempt to establish rules and procedures for every element of practice.

  •  An insistence that units and offices strive for consistency rather than innovation and efficiencies.

  • Quality assurance systems focused on compliance with these myriad policies and rules.

  •  Low organizational rank and status for “workers”, i.e., practitioners, who are viewed as disposable tools for delivering services.

  •  A concern with creating a virtual world of documentation that is as important, or even more important, than what caseworkers and supervisors actually do in their work with children and families, i.e., “if it’s not documented, it didn’t happen.”

  • Highly specialized units that transfer cases as quickly as possible, without feedback on case outcomes from receiving units to transferring units.

  • Information systems that often produce huge amounts of aggregate data at the office, region and state level, but not at the unit and caseworker level.

  • A lack of infrastructure for testing programs and practices.

  • Training programs that provide massive amounts of information on specialized subjects and “one off” trainings with limited opportunities for coaching of skill development.

  • Distant, superficial relationships of units and offices with university faculty and other researchers.


Some child welfare systems have, of course, altered their quality assurance and training systems, and data systems to enhance opportunities for learning. Nevertheless, even in  child welfare agencies which have developed outstanding quality assurance and training programs and performance data which goes to the unit and caseworker level, managers may be uninterested in meeting regularly with caseworkers and supervisors whose experience with and understanding of practice issues is often far greater than their own. There may be no encouragement, incentives, or permission for innovations that have not been endorsed in the agency’s policy framework. It may be possible to conduct investigations, and develop services and safety plans without being informed by intake units or voluntary services units of outcomes such as re-reports, child injuries, or out-of- home placements. Programs, including assessment tools and practice models, are unlikely to be evaluated in a rigorous way, e.g., using control groups, following implementation.


Child welfare systems that operate according to this paradigm are mainly interested in delivering a standardized set of services in a completely consistent way. If a new practice model is adopted, then the agency managers and quality assurance staff will be mainly interested in evaluating fidelity to the model, not its limitations, and uninterested in ideas for improving the model. Critical questions will be viewed as resistance to change, something to be discouraged and sanctioned, if need be.


Finding a Better Way


In my view, the changes in managerial practices that will make public child welfare agencies more rewarding places to work for practitioners are also the changes that will lead to more stimulating and productive learning environments.


  1. The Stephens Report recommends “stripping away the layers of policies that have taken CPS focus in different directions…” in order to free up caseworkers for a much greater amount of contact with children and parents. In my experience, this is more difficult to do than one might imagine, even when tools and models are clearly serving no useful purpose. Public agencies are often loath to acknowledge that they have made large investments in system reforms that have achieved little or nothing, either because of poor implementation or because of the inherent limitations of these tools and models. It is often the case that policies and practices that could be “stripped away” have adherents in key positions so that it is easier to defeat a useless requirement by ignoring it rather than changing agency policy.

  2. Child welfare management teams should become far less willing to implement tools and models system-wide that have not been rigorously tested in experimental research by non-developers or researchers who have not had an active role in their implementation. Policymakers and managers need to be far more skeptical regarding the claims of developers and advocates that assessment tools and models are “evidenced based”, claims that are frequently exaggerated when they are not complete inventions. There has been very little experimental research of practices and models widely used in child welfare by non-developers, and no good purpose is served by exaggerating the extent of knowledge regarding “what works”. 

  3. Initiative and innovation at the unit and office level should be encouraged and supported with small amounts of funding. Middle managers and top managers should be explicit regarding the parameters of innovation, especially those policies and practices that must be done consistently and strictly according to agency rules and the areas in which innovations are permitted. Child welfare managers should seek to steadily expand the potential for innovations, instead of restricting the scope of unit and office initiative. “Best practices” will be discovered in this process, and management teams will need ways of identifying and disseminating these practices. “Bottom-up” reforms should become as common as “top-down” changes implemented by expanding policy requirements.  

  4. Top managers and middle managers should meet regularly (at least quarterly) with caseworkers and supervisors, and these discussions should begin with a few key questions taken from solution based practice models: “What’s working?” “What’s not working as well as it should?”  “What can we (managers) do to make your jobs more doable and rewarding?” Every child welfare manager should understand the basic tenets of servant leadership.  

  5. Child welfare agencies should make major investments in increasing staff expertise and skills in substance abuse and mental health assessment, DV, infant mental health and early childhood development, cultural competence, developmental disabilities through 90-100 hour certification programs. Staff with job relevant certifications should receive a modest salary increase.

  6.  Caseworkers and supervisors should be given a 90 day sabbatical from case assignment and case carrying every four years. Staff on sabbatical should be required to pursue a course of study or practice that will increase their child welfare skills and knowledge.

  7. Master practitioners (as identified by their peers and by key external stakeholders) should be paid a modest stipend to mentor new caseworkers and supervisors.

  8. Feedback loops should be developed to provide timely information regarding the outcome of case plans to any staff who had a role in creating these plans.

  9. Management information systems should provide process and outcome data to the unit and caseworker level.

  10. Child welfare agencies should develop ongoing conversations with birth parents who have recently had open child welfare cases. Agency managers should be as well informed regarding qualitative measures of agency functioning as about quantitative measures of performance.




For decades, public child welfare agencies in the U.S. have been managed in ways

that have made the jobs and working conditions of caseworkers and supervisors

increasingly intolerable. There have been relentless efforts to limit the discretion of practitioners by steadily increasing prescriptive requirements, limiting or prohibiting

initiative and creativity at the office, unit and caseworker levels and by devaluing

practitioners’ ideas and feedback regarding system improvements. In addition, in

recent years, demoting or firing caseworkers and supervisors following high profile child deaths has become increasingly common. Not surprising, many self-respecting experienced caseworkers and newly hired staff have concluded that public child welfare agencies are not supportive work environments in which to develop their knowledge and skills, and many able and or promising staff with opportunities to leave have done so.


Many child welfare agencies currently face a workforce crisis that is of their own making, but which can be “unmade” by valuing and empowering line staff and supervisors. Child welfare managers and policymakers must change what has become a dysfunctional and destructive set of practices that cannot be redeemed by

a leadership model that depends on extraordinary individuals to mobilize and motivate

a beleaguered workforce.




Stephens Group CPS Operational Assessment: Findings, completed April 28, 2014.    

© 2020 by Dee Wilson Consulting. Proudly created with