Abstract: This article explores Texas Child Protective Services (CPS) caseworker turnover, utilizing a systems thinking approach to exemplify possible applications of systems thinking in the public policy realm. Caseworker turnover inhibits the success of child services agencies in most states, and despite expenditures of state resources, efforts to lower turnover have failed. Initially, the article uses systems thinking methods to examine identified problems and historical solutions to high caseworker turnover, then the analysis explores the systems involved in the cycles of turnover. The impact of public perception on turnover is used as an example of an overlooked but potentially profound variable in combating turnover. Lastly, the results of applying systems thinking to a public policy problem is summarized.


Systems thinking strives to reform the methods used to think about complex, wicked problems. Before any exposure to systems thinking, I understood that complex problems needed creative solutions, but I viewed this creativity as ephemeral and creative solutions more akin to alchemy. I knew, based on previous experiences, that creative solutions happened through assessment, brainstorming, evaluation, and working in groups, but I assumed that creative thought itself was just a byproduct of those activities. Since my exposure to systems thinking, I view creative thought as a compound made up of elements. These thoughts can be broken down into four building blocks: Distinctions, Systems, Relationships, and Perspectives (DSRP). All four interact in countless ways to articulate a complex system; by allowing systems thinkers to build a more realistic and comprehensive mental model of the problem at hand, these processes form the basis for creative solutions.

Systems thinking using DSRP can be applied to any discipline, including public policy, to formulate innovative methods for addressing multivariate, or wicked, problems. Child services agencies are wicked problems for most state governments in the US. Agencies face a myriad of interlocking difficulties that need comprehensive solutions, but so far, there is no model for an effective child services agency. Success measures are challenging for state agencies to determine and adhere to, societal perceptions of handling abuse are constantly shifting, and the structure of agencies is often convoluted. Government organizations are often slow to adapt to these shifting perceptions, further inhibiting the agencies from effectively keeping children safe. States regularly expend resources to reform child services agencies, only to find agencies continue to struggle to successfully fulfill their mission of keeping children safe from abuse and neglect. The problem of organizing and operating effective child services agencies is vast and complex, with no easily observable solution. In short, the problem is wicked.

My exploration of this problem using systems thinking operated according to many limitations, assumptions, and biases. First, I assumed child services casework conducted by a governmental entity is beneficial for the US. Second, due to the timeframe to explore the topic, I opted to focus on one problem of child services agencies: caseworker turnover. I assumed, based on past reforms, that excessive caseworker turnover impairs child services agencies because the quality of the workforce principally defines the quality of casework. Third, I used Texas Child Protective Services (CPS) as a reference point for historical statistics and previous attempts at reform. While each state has a particular agency climate, most face common problems and most state agencies have attempted common solutions to caseworker turnover. I opted to concentrate on Texas because of familiarity with the particular agency: I worked for three years in a related department to CPS under the umbrella of Texas Department of Family and Protective Services. I regularly interacted with CPS staff and witnessed numerous caseworkers leaving their jobs for a variety of reasons. Finally, I did not collect any data on agencies, instead relying on previous reports proposing reforms. I assume these reports represent an accurate assessment of the problems resulting in caseworker turnover. These limitations and assumptions should not impede the goal of this paper: to explore a topic through the lens of systems thinking.

Child services agencies have identified turnover as a concern since first they first began collecting data on turnover. For example, in 2003, the national average turnover of child services caseworkers was 30-40% nationwide, and that number has not diminished since. Texas Child Protective Services follows the national trend. In 2013, a report by the Stephen Group found that 43% of CPS investigation caseworkers, those who first investigate abuse reports, left their jobs within 24 months. The repercussions of turnover include expense and, most importantly, diminished success of cases. A survey conducted by the Sunset Commission estimated the cost of each caseworker loss to be $54,000. Child permanency rates, the rate at which a child is either placed in a foster home or reunited with their family without further social services intervention, are higher in regions where turnover is lower. Although there are many possible explanations for the correlation, research has indicated that high turnover rates are likely a partial cause.

The State of Texas has attempted many times to curb caseworker turnover, with little success. In 2009, the Center for Public Policy Priorities recommended revisions such as lowering caseloads (specifically for Family Based Services and Conservatorship caseworkers), further increasing salaries, creating a more appealing career path, hiring ahead of vacancies, and raising public awareness. In 2014, reports from Sunset Commission and the Stephen Group both recommended policy streamlining, increased pay, better performance evaluation tools, better training, mentoring, and simply hiring better-equipped workers. All of these attempted solutions appeared appropriate to combat turnover, but despite attempts to implement these changes, turnover remains high. For example, in 2006, the state provided a $5000 stipend for all direct caseworkers, including those in Family Based Services (FBSS), Conservatorship Caseworkers (CVS), and Investigations (INV), and reduced the caseload of each worker. The result was a 7% increase in turnover the following year.

Figure 1. A Metamap diagramming the interaction between problems identified as contributing to child services caseworker turnover

I began my examination of child services caseworker turnover by determining the problems from the solutions proposed by the Stephen’s Group and the Sunset Commission aimed to fix, then diagramming the relationships between each identified problem using Metamaps (a tool designed to aid in systems thinking through DSRP). I found that each problem created a feedback loop between the problem and turnover. For example, high caseload increased turnover and, due to a decrease in workforce, increased caseloads on remaining agency workers. Each problem also interacted with other problems. For instance, high caseload results in poor performance on cases and lowers morale, increasing turnover. In 2006, reforms addressed some causes of turnover, but they did not take into account how the lack of experience in the workforce perpetuated each cause. Turnover was already high in 2006, and about one third of the caseworkers were newly hired. Because caseworkers often require a great deal of experience to get satisfactory results within an allotted timeframe, the slower work of less experienced staff increased the overall workload of the entire workforce. Both new and experienced caseworkers might have ended up feeling dissatisfied with their performance, resulting in more turnover. Each set of proposed solutions has attempted to solve a particular set of issues that has been deemed most important, overlooking the intertwining nature of the set of problems. After exploring these relationships, I realized that a comprehensive, time-sensitive approach is crucial to any lasting reform.

However, focusing on the problems without accounting for the entities involved does not accurately model the causes of turnover. Any systemic solution must account for all of the distinct actors involved in casework. State government defines regulations, overall goals, and funding, while administration defines field policies, hiring practices, and caseloads. The direct work environment supervises and trains workers, and each worker brings his or her own experience level, background, and education. These distinct actors each have unique relationships among each other and differing perspectives on turnover. Most previous solutions have not involved two other critical actors: the public and the clients. With the exception of raising public awareness, the above solutions focused on the administration of CPS, the work environment, caseworkers themselves, or the state legislature. While these four entities heavily influence caseworker turnover, clients and the public shape the relationships and perspectives of all other actors, directly or indirectly. I realized that any set of solutions, no matter how comprehensive, would not result in real reform if the solutions ignored all the relationships among and the perspectives of these six constituencies.

Figure 2. A Metamap of entities influencing in Child Protective Services.

After establishing the primary actors, I plotted the system of actors surrounding caseworker turnover. Each actor constitutes a system in itself, including many interrelated procedures connected to other actors and their respective systems. These actors form a cycle articulating the chain of command, leading from state government to administration, direct caseworker environment, caseworkers, clients, and finally the public perception of the outcomes. Public perception then influences state government to take action. The cycle serves as a guide for studying more nuanced relationships, like the interaction between clients and caseworkers; both influence overall success, defining the perception of agency efficacy, and ultimately affect all actors in the cycle. A lasting solution to the problem of caseworker turnover requires not only a comprehensive solution to the existing problems derived from all these relationships, but also a coordinated approach among all actors involved in the overall system, including clients and the public.

To better convey the nuances of the cycle, I will dissect the multifaceted impact of public perception of caseworker turnover as an example of how thinking systemically leads to more robust understanding. Before systems thinking, my perspective aligned with the previously proposed agency action to increase general public awareness about the positive results of child services agencies, but I did not realize the widespread influence of public perception. Public perception sways decisions made by state government and the agency, but it also affects every other actor in the system. The relationships revealed through the network of relationships and perceptions promoted by systems thinking demonstrate that the negative public perception of CPS has the potential to drastically impact turnover in a variety of ways.

The public often does not have direct interaction with CPS, so media coverage defines common beliefs about the agency. Public perception also shapes the narrative of stories in the media, so each influences the other. Generally, the media focuses on child welfare services negatively because tragedy makes more compelling narrative. The Stephen’s Group report cited this as a cause for concern, but did not go into great detail about potential solutions. In my experience, media outlets publicize disastrous cases or agency failures as examples of a broken system, and the public reacts by pressuring state government; officials react to this pressure by implementing changes to policies or administrative staff. While some changes or staff adjustments are beneficial, the disruption results in a chaotic agency environment. The new administration creates a different plan of action for the agency and the caseworkers are given another set of policy changes promising reform. Caseworkers become frustrated by this chaos and untrusting of future administrative attempts at reform, further contributing to turnover.

The media also influences caseworkers and clients directly. Negative public perception affects caseworkers’ interactions with family, friends, co-workers, strangers, and clients. Prospective caseworkers might view CPS only based on the failures presented in the media and ultimately decide to not go into child services welfare. Current workers lose trust in their co-workers, diminishing the overall solidarity of the workforce. Clients become more hesitant to trust CPS, increasing the difficulty of achieving a successful outcome and perpetuating the negative feedback loops throughout the other catalysts of turnover.

If the public witnessed instances of positive outcomes of casework on a regular basis, public perception might shift from disdain for to admiration of caseworkers. The effect of this shift in the public’s mental models could potentially have a profound impact on turnover specifically and casework outcomes more generally. Caseworkers might benefit from their status as civil servants through personal interaction and pride. Clients might view caseworkers more as partners, as opposed to an antagonistic relationship, helping clients through a difficult time in the family. Child services might attract more social workers, and experienced workers might feel empowered to help shape the habits of new caseworkers, providing a more positive and consistent work environment. A favorable public opinion of caseworkers would not address all the factors involved in turnover, but it is an important factor to consider in an overall plan of action.

Systems thinking dramatically changed my impression of how best to approach child services caseworker turnover. Understanding the interrelated nature of each problem is crucial to finding solutions. Whereas the piecemeal quality of previous reforms has meant they provide only temporary relief, a coordinated and holistic strategy is a viable method might more permanently alleviate the issues concerning child protective services agencies. I also realized that public perception and the clients’ perspectives must be more thoroughly taken into account when developing a holistic strategy. More generally, this exercise provided a practical example of how systems thinking could shape the procedures for assessing problems and appropriate solutions in public policy.


About the Author: Harrison Speck is a 2017 CIPA Fellow and an associate editor for the Cornell Policy Review. Before attending Cornell, he worked for the State of Texas in public assistance eligibility and, most recently, in child welfare with the Texas Department of Family and Protective Services. He is an avid musician as well. His varied background reflects his primary interests, including social policy and music royalty regulation, although he appreciates many topics outside of his experience.


Bernadette Wright

Nice example of systems thinking! To take the analysis of your map a step further, you might look at the structure of your map and see what concepts are the best explained (has two or more arrows pointing to them — in the first map, culture of fear, poor performance, caseworker turnover, lack of experience, and high caseloads). Also, how many concepts (boxes) are on the map and which ones are less well-explained? That can show how your map improves as you gain new information. For more information, see our white papers on strategic knowledge mapping at http://meaningfulevidence.com/skm

Paul Taylor

So in a nutshell, previous analyses had used an artificially small scope, missing some of the negative feedback loops in the wider system. By expanded the context and including all of the actors, these loops emerge. Like most systems thinking stories, there are insights to be had, but no proof. Potential intervention — address case worker turnover with an old/new media marketing campaign to install a positive reinforcement loop for workers. Spend money or advertising, not case workers or the agencies. I love it when systems thinking leads you to a counter-intuitive intervention.


Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.