Does CPS Help Families?

(Originally published October 2010)

“Household, Family, and Child Risk Factors After an Investigation for Suspected Child Maltreatment: A Missed Opportunity for Prevention,” by Kristine Campbell, Lawrence Cook, Bonnie LaFleur and Heather Keenan, Archives of Pediatric and Adolescent Medicine, October, 2010 presents a study that has already been written up by  a national news magazine and occasioned newspaper articles around the country. Both the media and professionals who have various roles in child welfare systems have been interested in the study’s finding that there were no significant differences in seven modifiable risk factors between investigated and non – investigated families, or in families substantiated for child maltreatment, with the exception of maternal depression (higher in investigated families) a few months to 4 years following a CPS investigation. This study raises the question whether CPS investigations and services provided to families during or following an investigation help families, harm them or have little or no effect on family functioning.


This study is a secondary analysis of 595 children included in Longitudinal Studies of Child Abuse and Neglect (LONGSCAN) in five different widely dispersed sites. These children were initially interviewed by LONGSCAN investigators at age 4 between July 1991 and March 2000. They were then followed until age 18 with interviews every 2 -4 years. Slightly more than a quarter of the sampled children (27.8%) lived in families investigated for child maltreatment between the first LONGSCAN interview at age 4 and the follow up interview at age 8. It is important to note that investigated and non – investigated families were not equivalent in important respects at the baseline interview. Families investigated by CPS were poorer, and the mothers had less education and were more depressed than mothers in non – investigated families. The 4 year old children in the investigated families had more externalizing behavior problems than the 4 year olds in non – investigated families. In summary, investigated families were poorer and more troubled at baseline than non – investigated families. Not surprisingly, the investigated families continued to be poorer, and the maternal caregivers in these families more depressed than the mothers in uninvestigated families in the months and years following CPS investigation. A CPS investigation was associated with higher levels of risk in social support, family functioning and child behavior in investigated families at the follow up LONGSCAN interview, though only elevated maternal depression was significantly higher for investigated families in the researchers’ linear mixed effects statistical model.


Before commenting on these research findings, there are other issues related to the study that deserve comment:

  • Initial baseline interviews with the 4 year old children occurred between 1991 and March 2000. This is an old data set with which to evaluate the effects of CPS interventions. Furthermore, it is no longer true (if it ever was) that CPS programs within states or among states are pretty much the same. The increased use of differential response systems around the country has been an important development during the past decade; and a number of child welfare jurisdictions have made large investments in family support services in recent years. A number of  counties and states use solution based practice models, but most do not.

  • Substance abuse and domestic violence were not measured in these families; consequently, any conclusions regarding the risk of child maltreatment in investigated families compared to non- investigated families can only be tentative at best.

  • This study did not track the effects of services on families; and the study lacks information regarding the types and amounts of services provided during or following a CPS investigation. These researchers acknowledge that nationally, less than 40% of families investigated by CPS receive services.

  • 45% of investigated families were substantiated for child maltreatment; however, a large percentage of substantiated families do not receive services for various reasons, sometimes because they don’t want them, or because services that are urgently needed (for example, housing services or child care) are not available. Again, this study provides no information regarding services provided to families substantiated for child maltreatment.  


Given the limitations of this study, why has it attracted so much media attention and interest among persons deeply knowledgeable about child welfare? I have yet to encounter a child welfare practitioner, scholar, or professional who works in any capacity with families involved with public child welfare agencies who is surprised or upset with the study’s findings. Knowledgeable professionals within and outside child welfare agencies understand that most CPS investigations are not substantiated and most are quickly closed without services or any other official action. Why would anyone expect that a CPS investigation absent any other agency intervention would have a positive effect on family functioning, social support, child behavior or any other risk factor or family problem? Furthermore, if a family receives services during or following an investigation, services are often brief and limited with the goal of preventing the need for out- of- home placement. Nevertheless, the reaction to the study suggests a deep unease regarding this  approach to child protection (i.e., investigations combined with brief crisis oriented services or removal from the home in extreme cases) that cannot be dispelled by an analysis of the study’s research design or limitations.


Many parent advocates believe that CPS investigations harm families by increasing parents’ stress and presenting a threat of child removal cased on individual caseworkers’ biases. A large number of advocates, practitioners and scholars question a CPS system organized around the investigations of overwhelmingly poor families – usually for neglect – when poverty related services are thin “on the ground” or non- existent. The racial disproportionality of children referred to CPS and placed in out- of- home care, and inequities in services and outcomes for African American and Native American families, has led to persistent concerns about institutional racism in child welfare. Child advocates are often furious at the inability of CPS to prevent the worst outcomes, including child abuse/neglect related deaths. For years, I have been appalled at the large number of families who are referred over and over again (usually for neglect) to CPS with no apparent benefit to either children or parents and at enormous costs to child welfare agencies. Scholars often point out the lack of evidence that services provided by child welfare agencies, such as most parent education programs, reduce child abuse and neglect. This is a disparate set of concerns that do not necessarily lead to a common agenda; but the list suggests a pervasive discontent that provides fertile ground for new research that raises fundamental questions about the goals and effectiveness of child protection.


Campbell, et al, have a straightforward and provocative argument: CPS investigations in the aggregate, even when combined with services, do not appear to lead to long term reductions in important and modifiable risk factors. Nevertheless, CPS investigations provide an opportunity to engage families (voluntarily in most cases) in services and programs that could achieve this end. Campbell, et al, comment on the possible benefits of poverty related services,  improved social support, Parent- Child Interaction Therapy (PCIT), Trauma Focused CBT in sexual abuse cases; and the need for research to identify effective interventions that can improve long term outcomes (presumably related to child development). In addition, one of the most distressing findings in the study is the elevated level of maternal depressive symptoms at both baseline and follow up LONGSCAN interviews. There is strong evidence that depression plays a key causal role in neglect, physical abuse and emotional maltreatment; and there are a variety of well tested treatments for depression that are under utilized in child welfare cases.


It would seem a “no brainer” that child welfare agencies should add tested effective services and programs as resources allow. Curiously, however, there is far from universal agreement with this position among policy makers, experts and practitioners. There is a widespread view within child welfare that CPS programs should seek to reduce or eliminate immediate threats to child safety, usually defined as threats of serious physical harm or potentially traumatic events; children’s development outcomes lie outside the scope of safety oriented CPS programs according to this view. A more common perspective is that CPS should be concerned with both the immediate safety of children and risk of recurrent abuse or neglect. As in Campbell, et al’s study, children’s behavior problems are viewed as a CPS concern when they are a risk factor for future child abuse or neglect, for example, conduct disorder often leads to physical abuse. However, in most child welfare agencies, recurrent child maltreatment is tracked over a period of months, not years, because agencies use the inadequate CFSR measure of recurrence, i.e., the percentage of “founded” cases substantiated a second time within 6 months.


 If CPS programs embrace the goal of helping families in a much larger percentage of investigations – or assessments in states with differential response systems -- weak family support systems must be greatly strengthened in most child welfare jurisdictions.   I have stated my view in a recent Sounding Board that child welfare agencies in the U.S. are slowly but erratically moving toward an approach to child protection oriented around family support services rather than child removal, much like child welfare programs in Western Europe. Child welfare agencies should consider the following guidelines in developing and (hopefully) testing family support programs:


  • An expanded array of poverty related services should be available to families. These services should include educational assistance for parents motivated to  use it as a means of offering low income parents an avenue out of poverty.   

  • Parents’ views of their families’ needs should be given serious and ongoing consideration at the program level. More use should be made of parent surveys and parent advisory groups by public agencies and by private agencies engaged in service delivery to child welfare populations.    

  • Services that lead to improved child development outcomes (as in some prevention programs) also reduce child maltreatment, for example, the Chicago Parent Child Centers and David’s Olds’ Nurse Home Visitation Program. Child welfare agencies should rethink the strict distinction between interventions aimed at child safety and those targeted at child well being.    

  • Evidence based parenting programs are skill based; but there is a group of troubled families who cannot benefit from skill based programs until substance abuse, mental health problems or family violence have been addressed. Effective targeting of services requires expert family assessments and the thoughtful sequencing of services.

  • A large fraction of families can be served in time limited programs (3 -9 months); but there is a population of mentally ill and/ or cognitively impaired parents who require case management services for many years, especially when they are parenting children.

  • Foster care can be used to provide respite care to birth families as well as in traditional ways. Some foster parents can serve as mentors to birth parents.

  • Public health nurses can be utilized in child welfare settings to work with parents of infants and toddlers. Much greater expertise regarding early childhood development is needed in child welfare agencies to adequately serve babies and other young children.

  • DV services for both victims and batterers can be provided in child welfare settings or can be offered as a part of family support centers or agencies.

  • Neighborhood based empowerment programs organized around family support centers is a promising approach to prevention that warrants further evaluation.


One of the interesting findings of the Campbell, et al study is that the large group (almost three quarters) of non - investigated families had levels of risk factors that were not significantly different from those of investigated families with the exception of maternal depression. News stories have emphasized the apparent lack of benefits to family functioning resulting from CPS investigations. However,  the reality that a large number of poor families never involved with CPS have a multitude of risk factors similar to investigated families suggests that (a) it may be a bit arbitrary which of these families are investigated by CPS and (b) the level of need among “at risk” families is far greater than child welfare programs can hope to meet. Perhaps other LONGSCAN studies have answered the question, “Is there a difference in developmental outcomes for children served by child welfare agencies vs. other LONGSCAN children in late adolescence and young adulthood?” Unfortunately, the demise of some or all LONGSCAN studies may prevent this question from being fully answered.


© 2020 by Dee Wilson Consulting. Proudly created with