Do Something Different
(Originally published October 2, 2017)
The recent report from the Auditor General of Pennsylvania, State of the Child: A look at the strengths and challenges of Pennsylvania’s child welfare system (2017), could have been written about most state and large county child welfare systems in the United States during any of the past three or four decades. Pennsylvania’s Auditor General, Eugene DePasquale, describes what has become a standard child welfare story following high profile child deaths, for example the recent Seattle Times story (September 21), ‘Everyone Failed Him’: Boy’s aunt accused of murder, DSHS accused of critical errors, regarding events leading to the death of a 5 year old child, Gary Blanton, who had been placed in an aunt’s home by a Children’s Administration (CA) caseworker. This nearly universal child welfare story in English speaking countries has these common elements:
poorly paid caseworkers with overwhelming workloads
inadequate training programs
high turnover and lack of experienced staff in critical positions
acute and chronic shortages of critical resources, especially foster homes, but also shortages of poverty related services such as low income housing and mental health services
steadily expanding policy and procedural frameworks
Fatality reviews and news stories usually find that some agency policies were not followed by assigned caseworkers and supervisors, internal warnings were ignored, and children were left in circumstances which endangered their lives. Twenty-twenty hindsight usually leads advocates and concerned citizens to ask the question, “what could they (i.e., decision makers) have been thinking?” Depending on the circumstances, following a child fatality one or more caseworkers, supervisors and middle managers may be fired or pressured to resign in order to assure policymakers and concerned citizens that someone is being held accountable. Agency leaders can be expected to quickly announce a comprehensive reform initiative that often includes new assessment tools and new requirements for monitoring child safety, mandatory training, a modest number of additional casework positions and, sometimes, new laws that expand state or county definitions of child maltreatment and standardize investigative requirements. The cycle of high profile fatalities resulting in a media firestorm which includes variations of the standard child welfare story, followed by comprehensive child welfare reform initiatives, has been repeated again and again in many states and large counties. This cycle used to occur every few years in most child welfare systems, but in recent years has become a nearly continuous process in many states and large counties.
After reading variations of the standard child welfare story applied to multiple jurisdictions, a naïve observer might conclude that public child welfare systems are pretty much the same, but this is not true. For example, State of the Child asserts that entry level caseworker salaries in 18 Pennsylvania counties (including some large urban counties) average about $30,000. In many other states, including Washington, entry level caseworker salaries are closer to $40,000 and top out at a little more than $60,000 per year. In a few states and counties, top out salaries of caseworkers average $75-80,000, while in some states caseworkers earn a living wage only by working large amounts of overtime on a regular basis. In fact, state and county child welfare systems vary greatly on every element of the standard child welfare story: overwhelming workloads, inadequate pay, poor training, high turnover of new caseworkers and inexperienced caseworkers in critical positions, foster home shortages and service gaps, especially in rural areas. Child welfare systems vary enormously, yet almost every state and large county child welfare system has had to periodically cope with every element of the standard child welfare story; and some systems have faced all elements of this story in an unremitting fashion for decades, for example Florida, and (until recently) Texas. The longer that any or all of these elements (overwhelming workloads, poor pay, high turnover of caseworkers, foster home shortages, etc.,) have persisted, the harder they are to undo. In some states, including Washington, effective child welfare reform initiatives will take at least a decade or more to effect significant changes once policymakers and child welfare managers begin doing the right things, rather than digging an even deeper hole by continuing to pursue failed policies and ineffective management practices.
Adding programs and policies will not help
The Seattle Times story regarding the death of 5 year old Gary Blanton quotes the child’s former foster parent: “All of the policies, procedures and laws are in place to prevent bad things from happening. And yet they keep happening. That’s the crime here.” The typical child welfare response to a high profile child fatality is to add more policies and procedures, more assessment tools, case planning formats, practice models and levels of internal review, actions that increase workload pressures. Inevitably, caseworkers and supervisors will follow these new policies erratically, or not at all, thereby increasing the agency’s exposure to future tort claims when children on open cases are seriously harmed or killed. Bureaucracies, faced with intense public pressure to improve their programs, typically engage in bureaucratic solutions, which, in turn, create intolerable working conditions for caseworkers and supervisors. Nearly continuous reform initiatives have severely harmed the practitioners who must cope with overwhelming policy and procedural frameworks. A highly regarded child welfare leader once informed me that the large public child welfare agency he had managed received more than 800 recommendations for agency improvements in child fatality reviews and other reports over a couple of decades. This agency continues to struggle with much the same conditions (e.g., erratic use of assessment tools and high turnover of caseworkers) from decade to decade with occasional improvements in performance resulting from brief periods of excellent leadership.
Newly hired caseworkers and experienced caseworkers and supervisors from whom so much is expected often choose to leave agencies in which they count for so little. Caseworkers and supervisors in mismanaged agencies often personalize what has occurred in their offices, i.e., they feel persecuted by bad managers. In truth, every large organization has its share of bullies and incompetent managers. However, it is mainly bad ideas rather than bad people which have made public child welfare agencies places from which practitioners seek to escape. The bureaucratic hierarchies and management practices of public child welfare agencies are much the same as they were 30-40 years ago. Nevertheless, child welfare agencies have become much more difficult environments in which to work because of the cumulative effects of routine bureaucratic practices such as steadily adding to policy requirements and developing computer systems that prevent shortcuts. As a rule, it is far easier for child welfare managers to rethink – and change – programs and practices with children and parents rather than to seriously reflect on bureaucratic approaches to management that are as routine (to them) as breathing, and that incorporate assumptions and goals usually accepted without reflection and which are rarely discussed.
More resources are necessary but not sufficient
Some advocates and child welfare staff believe that underfunded agencies and programs is the main problem facing public child welfare systems, and that a large increase in staffing and services would fix these agencies. It’s true that most child welfare agencies need sizeable enhancements beginning with additional and better paid staff at all levels, including case aides, administrative assistants, home support specialists, etc., greater financial support for foster parents/unlicensed kinship care and more and different types of in-home services. However, resource enhancements during a single legislative session or two rarely have long term effects on child welfare systems because (a) they’re not large enough to change working conditions and, therefore (b) the effects of enhancements are short lived, and (c) child welfare systems continue to be funded from bi-ennium to bi-ennium without regard to updated workload standards or flexible funding formulas.
Currently, news stories regarding Washington State’s child welfare system usually reference the Council of Accredidation (COA) workload standard of 18 cases. This is an outdated standard that does not reflect the increase in workload requirements during recent years. Public child welfare systems need to engage in a comprehensive workload study every 5-6 years to establish the hours per month required by agency policies for various types of cases. Workload standards should include not just averages but also lids on case assignment for CPS investigators, Alternative Response (AR) caseworkers, foster care caseworkers, adoption staff and licensors. When lids are exceeded for more than 60 days, child welfare offices should be empowered by law to bring on additional CPS investigators, contract with private agencies for case management services and prioritize assignment of new investigations/assessments according to explicit guidelines. Cases inactive for longer than 60 days should not be counted in caseload statistics. Fraudulent misrepresentation of caseload numbers by caseworkers, supervisors or managers should be grounds for immediate dismissal.
Investing in the workforce is the only path to effective reform
State of the Child tells the standard child welfare story but, there is something different in this report than similar reports disseminated a decade or two ago. This report zeroes in on workforce issues: “assessing caseworkers’ ability to do their jobs effectively provides a basis for determining whether Pennsylvania’s children can be kept safe.” And
“Child welfare is a complex, nuanced profession that requires skilled professionals working with other human service agencies to provide the services troubled families need.” There is very little in this report regarding assessment tools, practice models, case planning formats, etc., not because these practices are unimportant but because
no child welfare practice or practice model can be effective without a cadre of well trained, adequately supported and reasonably experienced professionals committed to child welfare careers. The most immediate need of public child welfare agencies is to improve working conditions for caseworkers and supervisors, to elevate their status and give these staff a greater voice in agency decision making. A large reduction of turnover in caseworker positions and among newly hired caseworkers is the main indicator of whether public agencies have improved working conditions for practitioners.
In Washington State, caseworker and supervisor salaries need to be increased by at least 20% over and above the salary increase provided in the recent legislative session. Child welfare systems need to be managed through use of reasonable and regularly updated workload standards. It is also important to provide incentives for innovation at the local level, and to support professional development of staff at all levels with modest salary increases. Every public child welfare agency should be funding certification programs in substance abuse, mental health, DV, child development and cultural competence. It is also possible for child welfare managers, freed from a narrow focus on compliance with agency policies and procedures, to utilize their offices as learning laboratories in which innovative practices are developed and tested before being taken to scale in statewide implementation initiatives.
Child welfare agencies must develop an array of credible child safety measures
Readers of this commentary may have noticed that news stories and reports that contain the standard child welfare story rarely mention any widely used measure of CPS practice or trend line in child maltreatment fatalities or serious injuries resulting from abuse or neglect. News stories and reports usually focus on specific child maltreatment related deaths or other perceived failures of child protection. State of the Child contains information regarding the number of recent child maltreatment deaths in Pennsylvania, along with the percentage of deceased children known to child welfare offices prior to death (almost 50%), but without a trend line that puts these statistics in context.
Public child welfare agencies utilize one main outcome measure of their CPS programs included by the federal government in periodic Child and Family Service Reviews (CFSR’s) of state child welfare systems, i.e., the percentage of children with a substantiated or indicated allegation of maltreatment in the first 6 months of a reporting period who are not victims in another substantiated/indicated report within a 6 month period. There is a widespread understanding among both child welfare scholars and child welfare staff that this recurrence measure is easily “gamed”, i.e., manipulated. In addition, state and county child welfare systems engage in widely varying substantiation practices, for example, by using substantiation as a marker for criminal action or other legal action (Pennsylvania) or, by determining that almost every child victim has been neglected (several states in the Northeast) or, by applying an implicit harm standard (see Drake and Jonson-Reid, 2000). In addition, several research studies have found that a variety of factors other than information regarding alleged incidents of child maltreatment affect substantiation decisions. Child welfare scholars tend to be skeptical regarding this measure; the strongest defense of the measure I’ve heard in recent years is that “it’s all we’ve got.” In addition, prevention researchers have periodically reminded child welfare advocates and scholars that no one stand-alone measure of child maltreatment can ever be sufficient due to the multi-dimensional nature of child safety (see Olds, et al, 2005).
Public child welfare managers are in a peculiar position: these agencies have huge administrative data bases, and they regularly engage in rhetoric regarding data driven practices. However, their agencies have no credible measures of their child protection programs. Public child welfare agencies in the U.S. don’t know whether their CPS programs are doing better, worse or about the same in protecting children compared to a decade or two ago. These agencies typically do not measure serious inflicted or neglect related injuries following a screened in CPS report, and they do not measure chronicity, i.e, percentage of screened in reports with 4 or more prior reports. Public agencies commonly count and report child maltreatment deaths in a way that cannot be credibly used as performance indicator. The unwillingness to use an array of credible child safety measures cannot be blamed on inadequate funding or on bureaucracy. It is a risk aversive survival strategy intended to protect child welfare agencies from any measures that might further discredit their CPS programs. Child welfare leaders must find the political courage to change unacceptable measurement practices that have made it impossible to track multi-year trends in child protection.
DePasquale, E., State of the Child; a look at the strengths and challenges of Pennsylvania’s child welfare system and the safety of at risk children, September 2017,
available on line.
Drake, B. & Jonson- Reid, M. (2000), “Substantiation and Early Decision Making Points in Child Welfare,” Child Maltreatment, 5 (3), 227-235.
‘Everyone Failed Him’: Boy’s aunt accused of murder, DSHS accused of critical errors,” Seattle Times, September 21, 2017.
Olds, D.,Eckenrode, J. & Kitzman, H. (2005), “Clarifying the Impact of the Nurse-Family Partnership on Child Maltreatment,” Child Abuse and Neglect, 29, 229-233.