UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The fidelity of implementation of family development response in British Columbia Ji, Daniel 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2015_september_ji_daniel.pdf [ 769.14kB ]
JSON: 24-1.0166351.json
JSON-LD: 24-1.0166351-ld.json
RDF/XML (Pretty): 24-1.0166351-rdf.xml
RDF/JSON: 24-1.0166351-rdf.json
Turtle: 24-1.0166351-turtle.txt
N-Triples: 24-1.0166351-rdf-ntriples.txt
Original Record: 24-1.0166351-source.json
Full Text

Full Text

  THE FIDELITY OF IMPLEMENTATION OF FAMILY DEVELOPMENT RESPONSE IN BRITISH COLUMBIA  by  Daniel Ji  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SOCIAL WORK  in  The Faculty of Graduate and Postdoctoral Studies    THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  July 2015  © Daniel Ji, 2015   Abstract While previous research has explored the efficacy of differential response programs in child welfare, there have been no studies to date about coding decisions between designations by child protection service agencies. Research has explored client satisfaction with differential response as well as rates of recidivism and removal/placement but with limited attention paid to the rationales behind coding decisions and re-coding once an initial designation pathway is assigned.  This descriptive study uses data previously gathered by child protection social workers to qualitatively evaluate the fidelity of implementation of Family Development Response in British Columbia and the integrity of the program with regards to its stated objectives. Based on a random sample of intakes, decision-making fidelity to code as family development response or investigation was examined by exploring rationales behind coding at critical decision points and mechanisms for re-coding during family involvement with child protective services. Subsequently, this study examined whether cases that had been coded as Family Development Response differed substantially from investigations in terms of service provision, outcomes and appropriateness of family development response for high-risk cases.        ii   Preface This study required the approval of the University of British Columbia’s Behavioural Research Ethics Board. The project name was The Fidelity of Implementation of Family Development Response in Child Welfare in British Columbia, project number H12-03040. The Ethics Board granted approval on November 1, 2012.  Dr. Richard Sullivan was the Principal Investigator in this study. As a Master’s thesis student, I wrote the thesis proposal, ethics application, invitation to participate in research and questionnaire. I reviewed the secondary assessment data obtained from the Ministry of Children and Family Development via drafted research agreement and wrote the manuscript. Dr. Richard Sullivan reviewed all drafts and provided direction and feedback for the study.               iii   Table of Contents Abstract……………………………………………………………………………………..…....ii Preface…………………………………………………………………………………………...iii Table of Contents………………………………………………………………………………..iv List of Tables………………………………………………………….………………….…….viii List of Figures……………………………………………………………………………………ix Acknowledgments………………………………………………………………………………..x 1 Introduction………………………………………..…………...………………………..1 1.1 Definition of Problem……………………………………………………………..1 1.2 Defining Differential Response………………………………………………….11 1.2.1 Personal Interest……………………………………………………….…19 2 Literature Review………………………………………………………………….…...23 2.1 Implementation……………………………………………………….………….23 2.2 Rationale…………………………………………………………………………25 2.3 The Screening Process…………………………………………………………...28 2.3.1 Appropriate Investigation Referrals……………………………………...28 2.3.2 Appropriate Family Development Response Referrals………………….29 2.4 Concluding Statement……………………………………………………………30 2.5 Differential Response in Historical Context……………………………….…….30  2.5.1 Paternalism and Child Saving in Child Welfare………….……………..30  2.5.3 Protection and Segregation……………………………………….……...32  2.5.4 The Depression Era to the Welfare State……………………………..….33 iv    2.5.5 Neo-conservatism in Child Welfare……………………………….……..36  2.5.6 The Move to Managerialism……….…………………………….………38  2.5.7 Residualism: A Modern Expression of Neo-conservatism………………39  2.5.8 Threshold Systems……………………….………………………………41 2.6 Differential Response: A Post-modern Hybrid……………………………..……42 2.6.1 Managing Volume after Gove and Risk Assessment……………………42 2.6.2 Old Wine in New Bottles………………………………………………...44 2.7 Current Research and Key Findings……………………………………..………47 3 Organizational Theory……………………………………………….………………...63 3.1 From Child Welfare to Child Protection………………………………….……...63  3.1.1 Critiques of Threshold Systems……….…………………………………63  3.1.2 Rationalization of Resources……………….………………….....……...65 3.2 Defining Organizational Theory………………………………………………....66  3.2.1 Responding to External Demands and Pressures………………..………67  3.2.2 Internal Systems Adaptations……………………………………..……..69  3.2.3 Resource Implications Demanding Solutions……………………….…...72 3.3 Bifurcation into Family Development Response and Investigation…………..…73  3.3.1 Organizational and Contextual Factors………………………….……….73  3.3.2 Influences on Case Designation………………………………………….75 3.3.3 Organizational Theory and Contextual Factors as Explanatory Variables…………………….………………………………...…………76 3.4 Research Questions………………………………………………………………77 v   4 Methods………………………………………………………………………………….79 4.1 Sampling…………………………………………………………………..……..79   4.1.1 Intake Data………………………………………………………………79   4.1.2 Staff Interviews………………………………………………………….81 4.2 Data Analysis………………………………………………………….………...82 5 Results……………………………………………………………….……………..……84 5.1 Preliminary Discussion…………………………………………………………..84 5.2 Comparing Canadian Incidence Study Data and Research Question One………84 5.3 Overview of Larger Sample and Research Question Two………………………85 5.4 Re-coding after Initial Designation and Research Question Three……….……..92 5.5 Overview of the Four Team Sub-sample………………………………………...95 5.6 Comparison between Regions……………………………………….…….……..98 5.7 Staff Interviews and Research Question Four…………………...…...………...100   5.7.1 Training and Early Implementation………………………………….....100   5.7.2 Decision-making Challenges……………………………………..…….102   5.7.3 Policy Changes………………………………………………………….103    5.7.4 Contracting out Family Development Response……………………….106   5.7.5 Outcome Monitoring……………………………………………..……..106 5.8 Overall Results………………………………………………………………….107 6 Discussion……………………………………………………………………………...110 6.1 Results……………………………………………………………………..……110   6.1.1 Review of Research Questions…………………………….…………...110 vi   6.2 Comparisons with Existing Research…………………………………………..114 6.3 Study Limitations……………………………………………………………….114 6.4 Broader Implications and Future Research……………………………………..115 References……………………………………………………………………………………117 Appendices……………………………………………………………………………………127 Appendix A: Interview Guide…………………………………………..………………127                 vii   List of Tables  Table 5.1 Designations (Frequency and Percent), from January 2007 to March 2012 …… 85 Table 5.2 Type of Caller (Frequency and Percent), from January 2007 to March 2012 (Ordered from Most Frequent to Least Frequent) ……………………………… 85 Table 5.3 Caller and FDR or INV designation, from January 2007 to March 2012 (Sorted by Total Records per Caller Type)………………………... ……………86 Table 5.3b Mandatory vs. Voluntary Caller and FDR or INV designation, from January 2007 to March 2012 ………………………………………………………………….. 89 Table 5.4 Intake Designation Type, by Year …………………………………….…...…... 89 Table 5.5 Nature of Request (Frequency and Percent),  from January 2007 to March 2012 ..………………….………………………… 90 Figure 5.2 Chart to Track Sample Sizes Leading to Re-designations ……..………………. 91 Table 5.6 Intakes with Follow-up Calls, Follow-up Calls Within One Year, Change in Status, by Year …………………………...…….…………………………….… 92 Table 5.7 Pilot Versus Non-Pilot Sites and  FDR use from January 2007 to March 2012 …………………………………… 94 Table 5.8 Pilot Versus Non-Pilot Sites and FDR use by Year …………………..……...… 95 Table 5.9 Proportions of FDR and INV Between  Two Regions, January 2007 to March 2012……………………………..………96 Table 5.10 Contracted and In-house and FDR use by Year ………………………………... 97   viii   List of Figures Figure 5.1 Designation by Caller Type ……………………………………..………...…… 87                ix   Acknowledgments I offer my utmost gratitude to Dr. Richard Sullivan for providing unwavering support and guidance at every step along the research process. I wish to thank my thesis committee members Dr. Sheila Marshall and Dr. Carrie Yodanis for their enthusiasm and commitment to this thesis. I would also like to thank my field placement supervisor Don Miller for fostering the connections that helped make my placement and study possible and support provided through expertise and experience. I would like to acknowledge and thank the Ministry of Children and Family Development for its continued support of my education and professional development and the Team Leaders who were willing to participate in this research: Jay Chalkman, Lynn Schieder, Kathryn Smith and Laura-Dawn Matta. Thank you to Ally Butler and Marie Worden for being available throughout the process and last but not least, I would like to thank my Team Leader Paul Houle for his continuing support of my academic and professional development. To my family and friends, I could not have done this without you.         x 1  I Introduction I.I Definition of Problem Child protection social work is a challenging profession. Social workers who are on the front lines of protection must make clinical judgments every day about the safety of vulnerable children based on a timely, comprehensive assessment of risk. When a child protective services agency receives a report that a child has been harmed or is at risk, the report is screened for service according to eligibility criteria that are subject to influence from a variety of sources and time is of the essence once a decision is made that a child requires protective intervention. Any plans that are informed by these assessments must be made tentatively however, as new information may change the suitability of one plan over another. As a consequence, workers must formulate multiple possible plans simultaneously.  To be effective in child protection, social workers must possess a broad range of skills which include clinical interviewing, managing limited time and resources and navigating agency expectations and mandates. The organizational environment in which child protection social work is performed may be subject to instability due to changes resulting from unexpected child deaths (Douglas, 2013), high demands on workers or the agency (Bednar, 2003) and staffing limitations and caseload sizes (Smith &Donovan, 2003) along with changing legislation and practice guidelines (Buckley, 2000). Finally, workers may face organizational and public scrutiny in the event a child is harmed or worse (Buckley, & O’Nolan, 2014). Workers may even face legal or criminal repercussions for the choices they make in certain instances (Kanani, Regehr & Bernstein, 2002). In this context social workers face an array of obstacles that impede their capacity to properly address the needs of children and concerns for their safety (Bednar, 2003).  2  The necessity of a different means of responding to protection reports becomes clearer when examining the shortcomings of the current system. Waldfogel (1998) articulates concerns with the current state of child protective services (CPS) in five different categories. First, she argues that there are families who are over-included in child protective services that should not have been and as a result are exposed to unjustified public interventions. Children can come to the attention of protective services for multiple reasons based on calls from a myriad of sources and social workers at the screening phase must make the crucial decision of whether a response is required based on limited information. There will inevitably be families that are screened in for a response where one may not necessarily be warranted. Secondly, the number of families in the CPS system far outweighs the capacity of agencies to provide a satisfactory response. The welfare of children, particularly those who are living in poverty, has been a low priority for federal and local government funding (Ivanova, 2013; Wallace, Klein & Reitsma-Street, 2006).  Thirdly, there are families who would benefit from protective services that are under-included or never reached by CPS. Fourthly, CPS has always had a dual service orientation of investigating and remedying maltreatment authoritatively while being obliged to help keep families together whenever possible.  The tension between these competing goals has often led to an alternating pendulum swing between ‘saving the child’ (often via removal and placement) versus ‘preserving the family’ (providing safety with home-based intensive support and supervision). Finally, the delivery of apt services based on an accurate and timely assessment of child and family strengths and needs is an objective that is not always met in practice. Services may simply not be available, and language, culture and other barriers (location of service office, mental health barriers, lack of adequate childcare to attend services, financial restrictions) may lead to 3  compromised and inadequate service provision. It is within this context that the need for a different way of providing protective services for children and their families was recognized and developed. Recent research suggests that the number of reports being made to protection agencies is on the rise (Shusterman, Hollinshead, Fluke, & Yuan, 2005). This increase in the volume of reports may be the result of several factors including heightened awareness and reporting by the public and professionals (Parton, 1998), more reports being received about emotional maltreatment and intimate partner violence, a larger number of children being investigated per family, higher rates of substantiation and inclusion of investigations based solely on concerns of possible future risk of maltreatment (Trocmé, MacLaurin, Fallon, Daciuk, Billingsly et. al., 2008). Not only are the number of reports being made to child welfare agencies increasing, the types of maltreatment that are being reported are changing with more reports made about risk of maltreatment, neglect and exposure of children to domestic violence (Trocmé et. al., 2008; Sedlak et. al., 2010). According to the 2008 Canadian Incidence Study of Reported Child Abuse and Neglect, out of 235,842 child maltreatment investigations conducted in Canada, 174,411 child maltreatment investigations (74%) focused on possible abuse or neglect which may have already occurred while 61,431investigations focused on risk of future maltreatment (26%). Out of the total 235,842 investigations, 85,440 (36%) were substantiated, 71,053 were unfounded (30%) and eight percent (17,918) were those where there was insufficient evidence but suspected maltreatment of children by the time the investigation was complete. In five percent of investigations (12,018), risk of future harm was indicated while no risk was indicated in seventeen percent of investigations (39,289). In four percent of the investigations, the worker did not know if there was any future likely risk to the child(ren). 4  Of the 85,440 substantiated investigations of child maltreatment, physical abuse constituted approximately twenty percent (17, 212) of substantiated child maltreatment in Canada in 2008. Three percent (2,607) of substantiated maltreatment reports were for sexual abuse and nine percent (7,423) of substantiated reports were for emotional maltreatment. Interestingly, thirty-four percent of substantiated maltreatment reports were for neglect (28, 939) and another thirty-four percent for domestic violence (29, 259).  A comparison with the 2003 Canadian Incidence Study is telling. For the same categories used in the 2008 study, the 2003 CIS shows that physical abuse constituted twenty-four percent (25, 257) of substantiated maltreatment reports while neglect represented thirty percent (30, 366) of substantiated reports. Substantiated sexual abuse reports in 2003 remained consistent at three percent (2, 935) while substantiated emotional maltreatment reports represented fifteen percent (15, 369) of all substantiated reports.  Substantiated reports for exposure to domestic violence in 2003 represented twenty-eight (29, 370) of all substantiated maltreatment reports. We see by comparing these two studies a modest decrease in substantiated physical abuse reports and a six percent decrease in emotional maltreatment reports from 2003 to 2008. We also see modest increases in substantiated reports about neglect and an increase in substantiated exposure to domestic violence reports by six percent.  Again, these changes often occur in a context of limited resources, conflicting agency goals, high caseloads and increasing demands on child welfare workers and agencies (Buckley & O’Nolan, 2014).  Between the 1998 Canadian Incidence Study and the 2003 study there is a two-fold increase in the number of investigations conducted (approximately 135,261 in 1998 to 235,315 in 2003). By contrast, the rate of change between the 2003 and 2008 CIS has not significantly changed (approximately 235,315 to 235,842).  Placement rates have stayed relatively consistent 5  between all three phases of CIS data collection except for a moderate increase in the use of placement of children with relatives. Aboriginal children were determined as a key group to examine due to concerns of their over-representation in the foster care system. Approximately twenty-two percent (18,510) of substantiated maltreatment investigations were of children of Aboriginal decent (15% First Nations children with status, 3% First Nations children without status, 2% were Metis, 1% were Inuit and another 1% were classified as other Aboriginal).  In forty-six percent (39,460) of substantiated child maltreatment investigations there was at least one child functioning issue indicated. Academic difficulties were most common at twenty-three percent, followed by depression/anxiety/withdrawal (19%), aggression (15%), attachment issues (14%), intellectual/developmental disability and Attention Deficit Hyperactivity Disorder (11% respectively). In seventy-eight percent (66,282) of all substantiated maltreatment investigations, there was at least one primary caregiver risk factor indicated. The most frequent risk factors were being a victim of domestic violence (46%), having few social supports (39%), mental health issues (27%).  Alcohol abuse (21%), drug or solvent abuse (17%), being a perpetrator of domestic violence (13%), physical health issues (10%), history of foster care/group home (8%) and cognitive impairment (6%) were also identified risk factors. In terms of the household and social risk factors, being on social assistance, employment insurance or on other benefits was a risk factor in thirty-three percent of substantiated maltreatment investigations. Other risk factors included one move within the last twelve months (20%), at least one household hazard (12%), living in public housing (11%) and two or more moves in the past twelve months (10%).  If we were to use the CIS 2008 data to compile an average profile of the Canadian family where child maltreatment has been substantiated, we might begin by saying that there is a high 6  probability of domestic violence having occurred in the home and that mental health or substance abuse played a factor in the reported concern. We may also notice that there are few existing social supports for the family and that the family is receiving some form of social assistance and living in public housing or moved recently. It may be likely that the concern presented in the report was for neglect rather than abuse and it is also likely that the report also contained at least some concern for the functioning of the child, whether behavioural or academic.  In the United States, the National Incidence Study of Child Abuse and Neglect (NIS-4) provides data on the nation’s assessment of neglect and abuse at regular intervals (NICHD, 2010). The authors report an overall decrease in incidence of maltreatment since the previous cycle of the study in 1996 with changes in specific maltreatment categories. The study uses two separate standards of maltreatment: the harm standard where abuse or neglect has been substantiated and the endangerment standard where there may be no substantiation but reason to believe a child is in danger of harm.  An estimated 1,256,000 children experienced maltreatment during the NIS-4 study year of 2005 to 2006 overall. Of these children, forty-four percent (approximately 553,300) were abused while most of the children (sixty-one percent or approximately 771,700) were neglected. Having a total percentage of over one hundred is explained by having children classified in every category that applies (ie. if the same child fit more than one category). Most of the maltreated children (58%) experienced physical abuse (approximately 323, 000). Less than twenty-four percent were sexually abused (approximately 135,300), and approximately twenty-seven percent (148,500) were emotionally abused. In addition, the authors reported that forty-seven percent (360,500) of the children who had experienced neglect had experienced educational neglect, over one third (38% or 295,300) had experienced physical neglect and one-fourth (25% or 193,400) 7  children had experienced emotional neglect. Finally, the authors noted that since the NIS-3 there has been a significant decrease (26%) in the total number of children who had experienced harm (743,000 in the NIS-3 to 553,300 in the NIS-4) with decreases in rates in all categories including physical (23%), sexual (38%) and emotional abuse (27%). Cases of the most severe harm (cases where injury could be inferred) declined from 165,300 children in the NIS-3 to 71,500 in the NIS-4 representing a fifty-seven percent decrease in number and sixty percent decline in the rate in the population. It is interesting to note that the rate of neglect cases which met the standard for harm since the NIS-3 did not exhibit any statistically reliable changes overall or in any specific category (educational, physical or emotional).   The endangerment standard of maltreatment is more inclusive and “provides a very different picture of the incidence and distribution of child abuse and neglect” (NICHD, 2010). Nearly three million (approximately 2,905,800) children experienced maltreatment during the NIS-4 study year (2005-2006). The majority of children who experienced endangerment standard maltreatment were neglected (77% or 2,251,600) while the minority (29% or 835,000 children) were abused. The majority of abuse consisted of physical abuse (57% or 476,000) while thirty-six percent of children (302,600) were emotionally abused and twenty-two percent (180,500 children) were sexually abused.   Under the lower endangerment standards of neglect, the majority of maltreated children were physically neglected (53% or 1,192,200 children) but a similar number were emotionally neglected (52% or 1,173,800). Educational neglect made up only sixteen percent (approximately 360,500) of all neglected children. Between the NIS-3 conducted in 1993 and the NIS-4 in 2005 – 2006, there was no statistically reliable change in the incidence of children who had experienced maltreatment by the endangerment standard; however, within the endangerment 8  standard maltreatment, the authors reported that “Significant decreases in the incidence of abuse and all specific categories of abuse contrast with a significant increase in the incidence of emotional neglect” (NICHD, 2010).   Specifically, the number of children who experienced endangerment standard abuse (32% decrease in number and 38% rate decline), physical abuse (a 22% decrease in number from 614,100 to 476,000 corresponding with a 29% decline in rate), endangerment standard sexual abuse (decreasing from 300,200 from the NIS-3 to 180,500 in the NIS-4 indicating a 40% decrease in number and 47% decline in rate) and endangerment standard emotional abuse (532,000 to 302,600 reflecting a decrease in 43% and rate decline of 48%) all fell significantly. In contrast, the number of emotionally neglected children more than doubled from the last cycle of the study from 584,100 in the NIS-3 to 1,173,800 in the NIS-4. These numbers represents a 101% increase in number and 83% increase in rate. The number represents the actual number of children who have been maltreated while the rate represents the percentage per one thousand children who have been maltreated after accounting for the increase in numbers of children between study intervals. The authors are tentative about making conclusions however, stating that the increase in rates of emotional neglect since 1993 could represent a real increase in occurrence or a change in policy or focus towards recognizing the role of factors such as domestic violence or substance misuse.  In any discussion of child maltreatment, it is useful to consider the characteristics of the children who are being maltreated. Before exploring the characteristics of children who have experienced maltreatment in the 2008 CIS for the purposes of comparison and drawing implications, we can begin by exploring the demographic factors of children who are maltreated including race, age, sex, disability and school enrollment found to be relevant in the NIS-4 (NICHD, 2010). Of the more relevant findings of this study, the NIS-4 echoes previous findings 9  of a strong correlation between all categories of maltreatment and socioeconomic status and adds that these observations cannot be explained by the claim that such families are simply more visible to community professionals providing the data because the methods employed by the study observe a significant number of children and families at middle and upper income levels. Rates and severity of harm for all children increased after their second birthday. Girls were more vulnerable to sexual abuse and rates of serious harm and physical neglect decreased at a higher rate than for girls since the NIS-3. Finally, the authors reported that Black children were more likely in all cases to experience maltreatment than their White or Hispanic counterparts.  As with previous iterations of the NIS, the authors found that the majority of maltreated children do not receive a child protection investigation, adding that involvement was less likely for children living in jurisdictions where calls came through a centralized hotline, where new reports were combined into ongoing investigations, or where alternate response was offered to children and families referred for suspected maltreatment. The authors conclude by repeating earlier recommendations of better working relationships between child welfare agencies with schools and capitalizing on the unique role of front-line observers played by school professionals. When we examine the 2008 Canadian Incidence Study along with the 2010 Fourth National Incidence Study for trends we observe the following. First, there appear to be more protection reports made and substantiated about neglect and decreasing rates of reports and substantiations of physical abuse. Particularly in the United States with the two separate standards of maltreatment we see an increase in the number of reports that are at the lower risk or Endangerment standard rather than the higher risk Harm standard.  Secondly, domestic violence is flagged as an important factor in a high number of child protection reports in both countries. 10  Third, there were concerns raised about a range of social issues contributing to emotional neglect from mental illness, isolation and substance use to poverty-related concerns including receipt of social assistance and a strong correlation between socioeconomic status and child protection reports. Fourth, race plays a role as Aboriginal children are disproportionately represented in Canadian foster placements while Black children in the United States are reported as more likely to experience maltreatment than their White or Hispanic counterparts. Next, there appears to be a high proportion of protection reports that do not receive a response. Whether this is because the concern reported is not sufficient to meet the threshold of agency mandates or for other reasons, many children are simply not meeting the required definitions to be included for response. Conversely, the data indicates that there are many children who have been screened in for response where concerns were not substantiated or followed up on. Finally and perhaps most important, we can observe that there are many incidents where a report made to an agency responsible for child safety does not trigger an investigation or further involvement, despite there being a suspected risk to the child from the reporter or the agency itself.  From the trends we observe in the two studies, we can draw practical implications from the changing trends in child welfare reports. First and foremost, there is an indication in both studies of data collected from nation-wide samples that many of the cases where the risk level is not severe are simply not being followed up. As mentioned previously, the sheer volume of reports being received by protection agencies may preclude a satisfactory assessment and response for lower risk cases if workers are simply overwhelmed and lack the resources to adequately manage the flow of reports. There is clearly a need for a paradigm shift in child welfare that is designed to respond to the high volume of lower risk cases where the concerns indicated may be inclusive of problems of the family unit as well as social factors. The previous 11  findings in both Canada and the United States indicate that an approach that is suitable for lower risk cases may be necessary to manage volume and achieve the most efficacious response possible in light of these increasing demands. 1.2 Defining Differential Response Differential response is defined as the ability to respond to child protection reports with more than one pathway based on the type and severity of alleged maltreatment, age and vulnerability of the child, number of prior reports about the family and their willingness to collaborate in addressing identified concerns (Merkel-Holguin, Kaplan & Kwak, 2006). It is a secondary prevention paradigm in child welfare in which strength-based services are offered to families deemed to be at low to moderate risk for maltreatment (Conley, 2007). Sometimes referred to as ‘multi-track systems’ or ‘alternative response models’, differential response offers families the opportunity to engage with a worker cooperatively to address identified child welfare concerns before they reach levels that warrant more intrusive interventions (Merkel-Holguin, Kaplan & Kwak, 2006; Conley & Berrick, 2010). The definition was intended to shift the focus towards responding to the needs of the child and family instead of determining a perpetrator and victim.  Agencies struggling with limited resources, complex cases and a large volume of reports recognized the need to match interventions more closely to the severity of the reported concern and to engage families more in the assessment and service planning process (Dumbrill, 2006; McCroskey & Meezan, 1998). Jurisdictions that recognized the need for change in child welfare have implemented differential response programs in hopes of achieving more family-centered options and a nuanced approach to protection (Franke, Bagdasaryan & Furman, 2011). The model recognizes the need for a service system that targets families at lower levels of risk where there is a clear need for support but insufficient evidence of maltreatment to 12  meet statutory requirements for coercive intervention (Fallon, Trocmé & MacLaurin, 2011). The rationale behind multi-track systems is that if tailored, preventative services can be provided in a timely manner when risk is moderate to low, future involvement with child welfare agencies can be precluded (Conley, 2007). Ideally, the approach serves the purpose of providing timely service to address concerns without resorting to more intrusive measures. The differential response paradigm emerged from ongoing political concerns with the lack of response paradigms in child protection and the notion that “positive child welfare cannot be the responsibility of one sole agency” (Kyte, Trocmé & Chamberland, 2012) and attempted to answer the question of whether there could be a means of addressing the ever-broadening scope of child development and safety needs with tailored alternatives. A basic synopsis of historical trends in child welfare in the United States and Canada will establish the political foundation for understanding the introduction of differential response nationally. Eight core elements are consistent across jurisdictions that have implemented differential response (Merkel-Holguin, Kaplan & Kwak, 2006). First, differential response utilizes two or more discrete pathways of intervention, usually a traditional investigation, a differential response and often a third including some combination of the two. Second, differential response allows for multiple responses to child protection reports that are screened in for a response. Third, assignment of the response pathway is based on a determination of imminent danger, level of risk, number of prior reports on the same child or family, the source of the report and the presenting concerns such as type of alleged abuse or neglect or age of the subject child(ren). Reports that are coded at the screening phase as low- to moderate-risk are usually assessed as appropriate for differential response. Fourth, the designated response pathway can be changed based on information gathered during the assessment or investigation. When new information is 13  discovered about a family that lowers or heightens the level of risk to a child, a mechanism to allow for pathway re-assignment is built into practice guidelines so that workers can respond appropriately (BC Ministry of Children and Family Development, 2004). Fifth, there are clearly codified statutes, policy and/or protocols guiding the use of differential response for implementation to be effective. A sixth element of differential response is the ability for families who receive a non-investigative response to voluntarily accept or refuse the offered services and assessment without consequence. Since substantiation is the grounds on which services are provided in a traditional investigation and substantiation is not a focus of differential response, families can refuse when offered services. The voluntary nature of engagement with services may become less clear in practice if families choose to refuse since refusal might lead to a change in pathway designation. Seventh, there are no victims or perpetrators of the alleged abuse or neglect identified when a family receives a non-investigative response. Finally, the name of any alleged perpetrator is not entered into the central registry for individuals who receive a non-investigative differential response.  Beginning in 1974, the United States enacted the Child Abuse Prevention and Treatment Act (CAPTA) which formalized the regular use of intrusive interventions within child welfare. The Act defined maltreatment as “… at a minimum, any recent act or failure to act on the part of a parent or caretaker which results in death, serious physical or emotional harm, sexual abuse or exploitation, or an act or failure to act which presents an imminent risk of serious harm” (Stoltzfus, 2009, p.3 as cited in Hughes, Rycus, Saunders-Adams, Hughes & Hughes, 2013). This definition meant that protection social workers were obliged to act in situations of imminent risk but the main course of action at the time was removal. The removal and placement of at-risk children into foster homes has been the primary method of protecting children historically and 14  was supported by federal funding towards foster placements but not necessarily for services for children to remain within the family in the home. When research began to surface on the traumatic and unstable nature of foster placements, with children having to constantly change placements and having few supports for independent living without connections to family, a national permanency planning movement was born to bring permanence to the forefront for maltreated children (Wiltse, 1985).  By the time the Adoption Assistance and Child Welfare Act was passed in 1980, the importance of permanence was enshrined into legislation by requiring states to make efforts to prevent removal of children from their families and/or reunify with family members in a timely manner. In this effort, new services were developed to strengthen families and empower them to provide safety for their own children. These service models required collaboration between parents, child protective services and community agencies and included family preservation (McCroskey &Meezan, 1998) intensive services in the home (Robinson, 1985; Ziefert, 1985) wraparound services (Laird, 1985) and solutions-focused intervention (Berg, 2000). The Adoption and Safe Families Act passed in 1997 ensured that state child welfare agencies make efforts to preserve and promote the strengths of the family and put the health and safety of children at the forefront of all decisions for either removal or reunification. For the same time period in Canada (1974), the decline of the welfare state triggered major changes in the way that child welfare services were provided. Many of the institutionalized programs that constituted the welfare state were being systematically attacked and dismantled, leaving them vulnerable and resulting in a shift in focus from advocacy to monitoring and regulating recipients (Jennissen & Lundy, 2011, p. 269). Between 1974 and 1989, conservative and business forces attempted to use the ideal of the sanctity and 15  independence of the family to argue that governments were overspending on social programs, thereby creating reliance on the state. A solution to this ‘problem’ was to reduce public spending and place the responsibility back to the family or private sector. Thus, social and child welfare began to devolve programs in a way that had not been observed since the Depression era 1930’s. Program costs were transferred from the federal to provincial governments and municipalities. Privatization of services ushered in the dismantling of many social programs which were the foundation of the welfare state; child welfare agencies began to narrow their scope through limiting policy and service guidelines out of necessity.  Until the early 1960’s, Canadian child welfare agencies tended to respond to the most obvious cases of child maltreatment. Unfortunately, interventions were often directed towards families who belonged to marginalized groups possessing limited resources and education of the legal process (Bala, Zapf, Williams, Vogl & Hornick, 2004). This was the decade of the infamous “Sixties Scoop” of Aboriginal children along with groups like the socially marginalized Doukhobors in British Columbia beginning in 1953 when the Social Credit government under W.A.C.Bennett threatened to remove children if community members did not voluntary send their children to state-run schools (Androsoff, 2013). Limited attention was paid to children’s rights and they were rarely, if ever, involved in the child welfare proceedings where decisions were made about their futures. In the early 1960’s, social work and medical professionals began to use the term ‘battered child syndrome’ to identify cases in which child injuries lacked consistency with parental explanations. Legislative reforms resulted in mandatory reporting laws and child abuse registers and this in turn resulted in rising numbers of physical abuse reports. Professional and public awareness of child sexual abuse during the 1970s and 1980s resulted once again in increased numbers of reports as well as important reforms in the 16  treatment of child witnesses in the Canadian criminal system. Consequently, a high volume of perpetrators were convicted of current and past child sex crimes.  During the 1980s the stigma associated with childbirth out of marriage had largely faded from the public consciousness as more attention was being paid to sexual abuse cases and more support targeted towards single mothers. Definitions of family became more inclusive and an increase in mothers deciding to keep their newborns reduced the involvement of public child welfare agencies in adoption work. The introduction of the Young Offenders Act in 1984 (to replace the previous Juvenile Delinquents Act) meant that Canadian protection agencies were no longer responsible for delinquent youth and youth justice systems continued to evolve into an entity separate from child welfare. The introduction of the Canadian Charter of Rights and Freedoms in 1982 enshrined into legislation by the Supreme Court of Canada the notion that parents have an intrinsic interest in maintaining a relationship with their children, a right which is entitled to be protected under s. 7. Under this legislation, single mothers on social assistance were legally entitled to government-paid representation if a child in their care had been removed by child protective services. The shift towards parental rights occurring in Canada coincided with the American movement in favour of family preservation. The 1980s were a period focused on family preservation in Canada as in the United States but the 1990s would usher in a focus on protection, permanency and placement.  A number of high-profile cases where children had tragically died in the care of parents under the auspices of preserving the family resulted in sweeping reforms to the child welfare system in the 1990s. This involved a declining tolerance for allowing children to remain in homes where identified risk factors were present. These deaths also brought to the forefront the 17  destructive effects of emotional neglect and abuse from multiple factors including exposure to domestic violence, poor attachment, mental illness or substance use. The perception that protection agencies did not do enough to protect children was growing and resulted in frequent inquiries in multiple provinces. Perhaps one of the most well-known of these inquiries was the Gove Inquiry in 1995 which chronicled the tragic death of five year-old Matthew Vaudreil in the care of his mother and resulted in the creation of a new Ministry of Children and Families in British Columbia and a new piece of legislation with the renamed agency called the Child, Family and Community Services Act. After multiple inquiries into child deaths across the country, many provinces began to implement changes through legislation. The new legislation set out parameters for earlier intervention in child maltreatment cases and widened the scope of intervention for emotional abuse and neglect cases. Standardization of risk assessment was also an important outcome of many of the legislative reforms across jurisdictions in cases of child abuse or neglect such as the British Columbia Risk Assessment Model. Standardized assessment was intended to promote consistency in practice and provide structure to decision-making within investigations in terms of removal and placement. The end goal of the late 1990s was permanency for children to be achieved as quickly as possible once children had been removed.  It was a goal founded on the desire to maximize child safety but occurred within a context of reduced government spending based on the conservative political idea that the family should take economic responsibility for the care of their children. Self-sufficiency for individuals and families guided much of the social assistance re-structuring efforts at the time as evidenced by workfare programs and demonstration of attempts to find employment as necessary pre-requisites of receiving government financial support. When parents were the victims of conservative government 18  spending systems based on notions of independence and self-sufficiency and the child welfare system emphasized earlier intervention and permanency over supporting parental capacity, substantially more children ended up coming into care in North America as a result in the late 1990s into the 2000s. The high volume of children coming into care became problematic for child protection agencies already stretched to their limits in terms of resource, placements and available supports for children. The legislative reforms which were intended to reduce the likelihood of maltreatment to children had the inverse effect of adding burdens to an already overwhelmed system.  In 2004 the Supreme Court of Canada decided in a landmark case which required the balanced consideration of the rights of children versus the rights of parents with respect to the use of ‘reasonable’ corrective force on a child. The importance of this decision was that even though the courts recognized the vulnerability of children, the right and authority of parents to use physical discipline on their children was preserved in recognition of the importance of family context. The decision also opened the doors to discussion which emphasized rights of both parents and children in any state decision to intervene with families. The increasing legalization of child protection has also meant that proceedings are now subject to the same challenges associated with other parts of the legal system, such as delays in court processes, an adversarial system, and consequent hostility and insensitivity to the challenges faced by social workers in the child welfare system. If there are trials, then the children involved may be unfairly subjected to the delays of the system. In recognition of the adverse effects of the trial system on children, many jurisdictions including British Columbia have instituted different forms of mediation and alternative dispute resolution to assist in making timely decisions for the sake of the children.  19  Today’s child protection agencies now face a dilemma of how to keep children safe by reducing risk levels of abuse and neglect to children while limiting the use of costly court intervention or child placements as much as possible. Clearly, the family preservation approach of allowing children to remain in homes where there are identified chronic risk factors has resulted in tragic outcomes that have received much public scrutiny and, on the other hand, the adverse effects of drawn-out court proceedings as an impediment to permanence is equally problematic. It is within this context that we find differential response gaining attention in North America. The rising demands on child welfare agencies requires a response that involves coordination with knowledgeable professionals in the fields of education and child development, social assistance, affordable housing, domestic violence, mental health, immigration, addictions and law enforcement, among others. A compartmentalized approach to child protection that fails to recognize how the interactions of these systems may contribute to child maltreatment is no longer sufficient. 1.2.1 Personal Interest My interest in differential response comes from time spent meeting with children and families as a child protection social worker with the British Columbia Ministry of Children and Family Development. Many of these meetings consisted of parents telling me stories about how their family was struggling, usually in more than one area, and why they were asking for help. It was not uncommon for these family challenges to intersect with each other in unique ways to produce intricate hardships for parents that would in turn be experienced in some way by the children. Parents often attempted to explain during first meetings how their struggles were related to the child protection concerns. Issues ranging from chronic poverty, social isolation, difficulty navigating service systems, domestic violence, the immigration process, mental illness 20  and substance use were ongoing problems by the time child protective services became involved. The building of relationships in this context of uncertainty continues to be a great challenge, yet these relationships were the most memorable and continue to be, in my opinion, the most effective element of meaningful change.  I first received training in differential response under the mentorship of workers who were part of the first Family Development Response team in British Columbia and one of the only teams to provide the approach as a distinct service model. The opportunities to employ their teachings in practice with families were plentiful as children and families with whom I met were often facing multiple barriers in providing for their children’s needs.   The complex nature of the struggles faced by parents, combined with a constant reflection on what it means to protect children, fostered in me an appreciation of the need for a different approach to ensuring child welfare. Digging deeper with families tended to reveal more than was included in the initial report and while maltreatment cannot be ignored, protection work tended to take on a different tone when a willingness to understand became a necessary part of the child protection assessment process.  Of course, differential response did not sit favourably with all workers and I had heard criticisms from colleagues including how the approach was ‘too soft’, that it was not adequate to keep children safe from harm or abuse, or that workers who provided it did not focus primarily on the child’s needs. The most fascinating aspect to these criticisms was not necessarily in weighing their merit or content, but the extent to which opinions about FDR were polarized. I began to wonder how a different approach to keeping children safe could be so controversial. This controversy surrounding the capacity of differential response (or perceived lack thereof) to 21  keep children safe continues to interest me, given its ability to traverse the ideological spectrum with some flexibility.  Differential response can be made to fit within a range of political and philosophical orientations depending on how it is conceptualized. If one’s orientation is that children and families are inherently entitled to social services that will enable them to live comfortably and participate equitably in society, then differential response supports this aim by fostering partnership with clients through a strength-based approach inclusive of advocacy towards social justice, one family at a time. If, on the other hand, one believes that individuals and families ought to use whatever resources they have at their disposal to ensure that their children’s developmental needs are met, then a differential response empowers families by providing tailored referrals that will allow them to do so. It may have been this flexibility in its underlying philosophy that allowed differential response to continue as a distinct service model in the face of changes in political regimes in British Columbia.  Another point of interest in conducting this study was a recent development occurring specifically within British Columbia: the direction for workers to code protection reports in favour of differential response when possible. This coding presumption will be discussed further in later sections but there are aspects of this trend that I mention here as they relate to nurturing my interest in the topic. What was most appealing for me was attempting to find out how this presumption was informed. To code protection reports in presumption of differential response is no small decision, especially when considering the possible implications with respect to the original stated objectives of the program. This is where screening and coding plays such a critical role. If there are agency guidelines under which protection reports are supposed to be coded as 22  differential response, then how does a worker effectively balance current agency obligations with existing stated coding guidelines and policy?                      23  2 Literature Review 2.1 Implementation British Columbia’s Family Development Response (FDR) was implemented in November of 2003 by the Ministry of Children and Family Development (MCFD), the provincial government agency responsible for the protection of children (British Columbia Ministry of Children and Family Development, 2004). The Ministry introduced new Child and Family Service Standards to represent “shifting practice away from a reliance on investigation to more flexible responses emphasizing collaboration and building the strengths and capacity of vulnerable children and families” (BCMCFD, 2004). The Standards were accompanied by the Service Transformation Charter in January, 2004 which confirmed the Ministries’ commitment to “empower families, organizations and communities to share in the responsibility with government to protect and care for children” (BCMCFD, 2004).  Both the Charter and Service Standards were internal Ministry documents intended to rationalize the implementation of FDR along with several additional innovations designed to engage extended family and community members in child welfare assessment and planning. These innovations included the expanded use of family conferences, mediation and collaborative planning and decision-making practices outside of the court room. Alternatives to foster care were introduced to reduce the numbers of children coming into care by placing them with relatives or others who were connected to the child. The new Service Standards defined FDR as: An approach to child protection reports when, according to an assessment, the risk of harm can be managed through the provision of intensive, time-limited support services. It includes a strengths-based assessment of the family’s capacity to safely care for a child, and provision of support services, instead of a child protection investigation” (British 24  Columbia Ministry of Children and Family Development, 2004).    A province-wide working group composed of regional representatives and staff from Education Services Branch was formed in 2004 to meet the challenges of providing a consistent FDR response regardless of geographical location, encourage its appropriate use regardless of family ethnicity, and to standardize reporting and transfer practices between regions. By July of 2004, Regional and Deputy Directors reviewed the recommendations of the working group and made several recommendations of their own. Three of their recommendations included 1) focus implementation on the approach over the tools; 2) use selected assessment tools to aid in decision making and service delivery; and 3) emphasize clinical judgment and supervision to support practice. These recommendations were formally articulated into MCFD’s practice standards 12 and 14 with the hope that “strength-based components and detailed assessment criteria that are consistent with best practice principles, research and experience of other jurisdictions will assist with determining the most appropriate response to child protection reports” (BCMCFD, 2004).  Practice standard 12 guides workers who are screening child protection reports and helps them to determine which response pathway is most appropriate. To help in the assessment the Ministry employed a set of “standardized assessment tools that have been developed and endorsed by leading practitioners and researchers to inform clinical judgment” (BCMCFD, 2004). The tools guiding decision-making and planning in FDR when it was first implemented in British Columbia were the Comprehensive Risk Assessment, the North Carolina Family Assessment Scales and the California Risk Assessment. Practice standard 14 supports the provision of FDR “in circumstances where identified risks of harm to a child can be managed 25  through the provision of intensive, time-limited support services rather than an investigation” (BCMCFD, 2004). 2.2 Rationale The turn of the century saw the highest numbers of children in care in British Columbia’s history which coincided with budget restraints introduced in the late 1990s. These restraints were in the aftermath of a rapid increase in volume of protection reports made to the Ministry after risk assessment was implemented in the wake of public outcry at the death of a child known to the Ministry but not in its care. The government attempted to respond to the challenge of an overburdened system mostly by finding ways to prevent removals and placements. Family Development Response was one such initiative and promised preventative services for children and families at a time when protection reports were on the rise and agency resources were scarce and needed to be allocated to maximize their effectiveness.  Gordon Campbell was elected Premier in May of 2001 and the Liberal government prioritized keeping expenditures low by reducing budget and staff and diverting children from the system whenever possible. This philosophy was carried by then Minister of Children and Family Development, Gordon Hogg. Legislation which favoured the family environment, the promotion of early childhood development and family development response were all initiatives under the same umbrella of maintaining the family unit. The Liberals divested from the risk assessment model when it became apparent that it was contributing to the rising numbers of children in care and instead concentrated its resources on family preservation initiatives (Callahan & Swift, 2006). Some of these methods included the use of supervision orders, youth agreements introduced in December 1999 and the “utilization management” program which provided resources to examine if children could return to their homes safely from placements 26  (Foster, 2007). For the fiscal year of 2000-1 for example, 300 youth were receiving youth agreements staying out of care and by 2004-5, 789 youth were on such agreements.  Removal and placement remained as options reserved for children who were classified as highest risk but the Ministry soon acknowledged that giving priority to these highest-risk cases disqualified many families from receiving services. Families with identified child protection concerns that did not meet the threshold of a substantiated protection concern typically had their cases closed without follow-up. Part of the rationale for implementing FDR in British Columbia was to provide intensive short-term services to lower-risk families to prevent repeated contacts with the Ministry before “more intrusive, disruptive and costly intervention becomes necessary” (BCMCFD, 2004). The problem with using substantiation as the grounds on which service is provided is the persistent finding that families with a substantiated child protection report represent the minority of families who come into contact with protective services (Loman & Siegel, 2008; Fallon, Trocmé & MacLaurin, 2011; Ruiz-Casares, Trocme & Fallon, 2012). As a result, many families must have contact with protective services multiple times before actually receiving any services and by that point, may already require more intrusive interventions like removal and substitute care.  FDR was also implemented as a corrective measure for the inherent problems associated with the investigative approach. The investigative approach has received criticism for having the potential to foster an adversarial relationship between the worker and families in an atmosphere of threats and anger depending on the families’ history with child protective services (Schreiber, Fuller & Paceley, 2013; Dumbrill, 2006; Harris, 2012; Kris, Slyer, Iannicelli & Lourie, 2012; Loman & Siegel, 2008). The Ministry recognized that investigations had the potential to create tension in the worker-family relationship and to alienate parents, extended family and the 27  community in the process. On a larger scale the Ministry sought to reverse its public image as antagonising towards parents and stigmatizing for children and families.  Prior to the introduction of differential response, the traditional and still-dominant mode of responding to reports of child abuse or neglect was a child protection investigation (Merkel-Holguin, Kaplan & Kwak, 2006). Modeled on a criminal investigation, the worker’s goal was to identify a victim (the child(ren)), a perpetrator (most often the parent), and to determine whether the abuse was likely to have happened given the evidence that the worker had gathered during the investigation (Loman & Siegel, 2008; Parton, 1998; Laird, 1985). The narrow focus on whether the act occurred (substantiation or finding) and who had been harmed had a tendency to create a threatening and punitive environment evoking “anger, fear and other negative emotions in caregivers” (Loman & Siegel, 2008).  MCFD states that: “FDR is not about reducing workload, saving money, minimizing risks or offloading responsibility for protecting children onto communities” (BCMCFD, 2004). The Ministry avowed that it implemented FDR to keep children safe and in their communities by offering parents a range of practical and clinical services along with referrals to community resources. The approach assumes that the worker will form a collaborative partnership with parents during the assessment through the use of strengths-based, solutions-focused interviewing. The worker will then use this relationship to guide service provision based on the needs and strengths of the family. Social workers performing FDR are allowed to have smaller caseloads because they are working closely with families and in partnership with service providers. The end goal of the FDR worker is to help the family build connections in the community to meet the child’s needs.   28  2.3 The Screening Process When a protection report is received, the screening social worker under the supervision of a clinical supervisor determines the most appropriate response within a period of five days. This determination is based on the child’s age and developmental level, severity and type of harm, risk level, previous history of abuse or neglect within the family and willingness of the parent to accept responsibility and participate in services, though this latter factor is “not the determining factor in deciding how to respond” (BCMCFD, 2004). Once the five-day assessment of the protection report is complete, a determination is made about whether the family receives an investigation response or an FDR. The screening team can also decide at this point to take no further action or refer to the community if there are no presenting concerns. This decision is informed by Section 13 of the Child, Family and Community Service Act (CFCSA), the legislation in British Columbia that determines when a child has been harmed or is at risk of being harmed. Based on the British Columbia Risk Assessment Model, protection reports assessed as priority one (life-threatening) or priority two (dangerous situations) would be recommended for investigation while reports assessed as priority three (damaging) or priority four (potentially damaging) would be recommended for FDR. With regards to the use of FDR with Aboriginal communities, the Aboriginal Services Branch in 2004 reviewed their own Aboriginal Operational and Practice Standards and determined that separate FDR standards were not necessary as their existing standards were inclusive of the new practice. 2.3.1 Appropriate Investigation Referrals The investigation remains within the repertoire of possible responses to protection reports but is reserved for cases where there is a higher level of assessed risk to the child, when parents 29  are unwilling to cooperate with MCFD intervention or in cases of serious physical or sexual abuse, or any applications for court orders under the CFCSA. The investigation involves an interview of the subject child (not necessarily with parental consent or approval), interviewing parents and others with knowledge of the circumstances surrounding the report and a medical examination if necessary. Once the investigation is complete, a determination is made about whether the alleged maltreatment can be substantiated and whether ongoing protective services such as court orders will be required. The investigative process usually takes thirty days to complete and children can be removed during or after this time. If the risk to the child increases during an FDR, the family may be re-referred for investigation.  2.3.2 Appropriate Family Development Response Referrals Referrals are considered to be appropriate for FDR when one of the following conditions are present: 1) there is an identified child protection concern according to Section 13 of the CFCSA; 2) the family is willing to work with MCFD to reduce identified risks; 3) the family has the capacity to benefit from the FDR process; 4) the initial assessment suggests that any identified protection concern can be resolved with intensive short-term intervention to build on the families’ existing strengths and resources; and 5) the family assessment and involvement of community partners to address the risks assists in a shift to the family ensuring the child’s safety.   Family Development Response is not appropriate when any of the following conditions are present: 1) the family does not acknowledge problems or risks to the child or children 2) the family is not willing to cooperate with interventions or referrals to community services 3) the family does not have the capacity to benefit from services 4) intensive short term intervention is not required as one brief contact is sufficient to resolve the child protection concern 5) the case may require longer term service in which case immediate transfer to a longer term family service 30  team is recommended 6) children are at high risk due to chronic or severe maltreatment and there are no community supports or limited capacity of the family to assess the safety of the child and 7) the family has a history of non-cooperation based on previous reports.  The team is to monitor the progress of the family and not cease involvement until sufficient progress has been made to indicate that the parents can maintain the safety of the child in the home and the file can be closed. The Child, Family and Community Service Act  in 1996 replaced the Family and Child Service Act of 1981and expanded the scope of protective services to include emotional abuse and an emphasis on preserving Aboriginal culture (Foster, 2006). The CFCSA enshrined certain principles that would guide FDR service provision and assessment. Among those principles most relevant to the implementation of FDR were those that stated that the family was the preferred environment for the care and upbringing of children and that the responsibility for the protection of children rested primarily with the parents. Following the previous principle came the stipulation that if a family was able to provide a safe and nurturing environment for a child with the support of available services, then these services ought to be provided. 2.4 Concluding Statement  One of the most common criticisms of differential/alternative response is that it is not a new approach but reflects attempts by implementing jurisdictions to correct the errors associated with a limited model of child welfare. McKenzie (2011) states that “[t]here is an argument that differential response is not really ‘new’, and that it simply reflects good child welfare practice which incorporates interventions based on family-centred practice, increased use of community-based resources, and an earlier form of intervention for some families”. The historical transformation from child welfare practice to a narrow focus on substantiation through assigned 31  culpability will be the focus on the next section as the outcome of residualism and concomitant proceduralism in a reduced model of social welfare.  2.5 Differential Response in Historical Context 2.5.1 Paternalism and Child Saving in Child Welfare  The introduction of industrial capitalism to Canada in the 1860s led to growing disparity between wealthy and poor when society became structured around the private market. This resulted in the rise of a “free” labouring class who would contribute to the production of goods and services, but shared little in the benefits of the wealth they had created (Jennissen & Lundy, 2011). Many families living in poverty who suffered from its consequences in inadequate housing and disease drew attention to concerns about the public health. Families who could not secure income from employment relied on charitable organizations to survive.  The primary concern of the child-saving movement was to rescue children who had come from homes where the parents were perceived to lack morals owing to their improvidence. Most workers were well-meaning middle- and upper-middle class Christian women who were “intent on teaching the poor the values of ‘proper white society’ and integrating them into the mainstream” (Callahan & Walmsley, 2006). Child welfare organizations acting on moral bases had a poor track record of critically analyzing how their values impacted important child welfare decisions, sometimes with disastrous results as in the case of the continuing disproportionate removal of Aboriginal children by the British Columbian government from the 1960s onward (Fournier & Cray, 1998; Johnston, 1983). These removals were the consequence of a jurisdictional lacuna wherein the province had no jurisdiction on reserves and the federal government had no jurisdictional obligation for the provision of preventive and supportive social services. Residualism’s worse effects were thereby revealed in large scale removals in the 32  absence of any prior attempt to prevent them. Social work as a profession shares a common history with the movement of charity organizations and despite the altruistic efforts and progress made by social workers and organizations to protect children and advance the social work profession, it has often operated from a moralistic and paternalistic position. 2.5.3 Protection and Segregation The period from the late 1800’s to the 1940’s coincides with what Charles and Gabor (2006) describe as the “protection-segregation” era in child welfare. Many families lived in poverty during the depression and children were seen as vulnerable individuals who needed protection from the abuses associated with surviving in an increasingly competitive adult economic world (Charles & Gabor, 2006). This was also when the first laws to protect children were enacted explicitly recognizing that childhood itself was a concept worth defending. The local Council of Women of Vancouver successfully petitioned British Columbia’s Legislative Assembly to pass the Infants Act on March 20, 1901 because there were no laws regulating how children were to be cared for at the time (Callahan &Walmsley, 2006). The Infants Act was the first legislation that gave Children’s Aid Societies (CAS) the authority to remove children and place them outside of the family. This act was soon followed by British Columbia’s first Child Protection Act, which contained provisions for the state to act as the guardian for neglected or orphaned children and for the CAS to act as caregiver if needed. Vancouver’s Children’s Aid Society was plagued from the outset by overcrowding, underfunding and public criticism, leading to the first British Columbia Child Welfare Survey conducted by Charlotte Whitton. The survey was designed to measure the public’s opinion and review social worker credentials and practice, and then make the results known to the public. The survey prompted major reforms in child welfare in the province. It also “enshrined the focus 33  on parents as the source of child maltreatment and established the two often-conflicting purposes of child welfare: supporting parents to care for their children and removing children from families where the parents failed to measure up” (Callahan & Walmsley, 2006). The survey only focused on childcare institutions and largely discounted child correctional facilities, institutions for the mentally ill or residential schools. It took a Eurocentric view of child welfare and confirmed worker and agency discrimination in child welfare services. The results of the survey reflected the discriminatory, middle class, Eurocentric opinions of the general public towards children and families who came into contact with protective services. During the protection-segregation era, children receiving residential services were categorized as disturbed, criminal offenders, unemployed, orphans or homeless and were segregated from each other as well as the general population (Ainsworth & Fulcher, 1981 as cited in Charles & Gabor, 2006). It was not un-common for such children to be permanently removed from their families of origin because preventive services aimed at enhancing family functioning to maintain children in their homes were not the preferred form of practice at the time. It is also possible that the persistent view of parents as deficient caregivers would preclude any justification to keep children in such homes. By the early twentieth century, child welfare continued to gain recognition in Canada as a necessary state institution to help children who had lost connections with their families and communities. 2.5.4 The Depression Era to the Welfare State In 1933, T.D. Patullo’s Liberal Government introduced a “New Deal” which transformed the province’s health and welfare services. Harry Cassidy became the first Director of Social Welfare in 1934 and set up a welfare field service in 1935 which led to the division between the provinces rural and urban areas for social workers. Social work practice in urban and rural 34  communities differed significantly and workers often experienced an array of diverse cases in different settings (Callahan & Walmsley, 2006). Group work was introduced for families with children who were experiencing similar challenges so that they could be mutual supports for each other and receive professional supports at the same time. Workers performed a variety of tasks including removals, adoptions, community-based development work, group work, rural social work and administered child welfare, mother’s pensions and social assistance.  The social welfare services provided to families during the depression years are an important development in the history of child welfare in British Columbia because they are the first steps towards preventive strategies to reduce the numbers of children coming into the care of the state through financial supports for parents. Welfare services for men like unemployment pension and worker’s compensation carried little social stigma and were usually associated with employment. On the other hand, welfare services for women and children like the mother’s pension, child welfare and social assistance tended to be more discretionary insofar as workers had flexibility in administrating benefits. The belief that individual workers could create meaningful change remained through the 1930s to the 1960s despite glaring structural inequalities that worsened through the depression.  Canada was able to recover from the Great Depression through the application of Keynesian economic theory (Magnusson et. al, 1984) and there were social workers aligned with a structural view of the requisites of a just society (Cassidy included). The orientation of child welfare services, however, remained largely within the personal change model. Massive expenditures were made on armaments for the Second World War which helped pull the government out of recession, create employment and stimulate economic recovery. Female employment soared temporarily and child care and other child centered services like dentistry 35  and health care were provided to support female employment in the war effort, but these were temporary and did not find their way into the permanent institutional fabric of social welfare. Many theoreticians in child welfare during the 1950s began to focus on the family for theoretical orientations, treatment methods and interventions (Brown, 1992). Foster care was receiving criticism during this time for needlessly separating children from their families and leading to high levels of “foster care drift” in many child welfare systems. Services provided to children in their own homes aimed at improving family functioning were gaining popularity as family treatment theories began the conceptual shift from family- to child-centered services. Later research and conceptualization of the importance of permanency planning also informed the attention being given to family-centered interventions.   From the end of World War II to the mid 1970’s, federal and provincial governments partnered to expand the development of social programs and services. The post-war economic boom resulted in many countries experiencing incredible prosperity and led to significant improvements in the well-being of most citizens. In Britain, the United States and Canada, social legislation was enacted to mark a new era which is sometimes referred to as “welfare capitalism” or the “welfare state” (Mullaly, 1993). Social rights were enshrined into legislation in Canada’s post-war liberal democracy in the forms of the welfare state and free collective bargaining (Carroll, 1983). The Green Book represented a new deal for Canadian society in which government, capital and labour recognized the need for public spending on services and programs to protect citizens from the hazards of a post-industrial society. Thus, “Social welfare developed in most Western industrialized nations on the assumption that a harmonious relationship could exist between capitalism and the social welfare sector” (Mullaly, 1993). 36  During the 1960s, several of the prominent Superintendents of Child Welfare in British Columbia resigned from office, most of them citing lack of staff and resources to do the job and an overburdened system that child care and placements or adoptions could not keep up with (Callahan & Walmsley, 2006). One year before Ruby McKay resigned in 1960 for such reasons, Member of the Legislative Assembly Dave Barrett echoed the concerns raised by the superintendents about the dwindling number of trained social workers and the resulting increase in caseload sizes. Funding limitations and increasing caseloads led many workers to direct their attention to the province’s overburdened child welfare system. Social workers used community development as a tool to gain government support for expanded programs and services to respond to children’s and families’ needs. Children who were seen as problematic were still being removed from their communities and placed in centralized institutions in the Greater Vancouver area but the notion that struggling families could be maintained with direct financial supports and social programs continued to gain momentum. 2.5.5 Neo-conservatism in Child Welfare The effectiveness of Keynesian economic theory was challenged in the 1970’s by the OPEC crisis, slow economic growth, the ongoing costs of the Vietnam War in the United States, government deficits and chronic unemployment as well as inflation and high interest rates (Mullaly, 1993). Ongoing economic decline in combination with decreased tax revenue led many Western nations like Canada into debt. By the early 1980s, Margaret Thatcher, Ronald Reagan and Brian Mulroney had similar responses to the debt crisis and rather than raising taxes, contended that the solution was to minimize government expenditures. Whether it was by pushing values of self-reliance, depicting government programs and services as intrusive or declaring the country ‘open for business’, all three governments began a concerted attack on the 37  welfare state and labour unions through the implementation of retrenchment policies aimed at cost reduction. Thus business interests pushed the welfare state into a legitimation crisis from which it has yet to recover. Commenting on Canada’s welfare state, Jennissen and Lundy (2011) observed that since the mid-1970s “there has been a dismantling of the welfare state, first with clawbacks and cutbacks to programs and services and, more recently, with a wholesale undermining of the fundamental principles of the welfare state, including the erosion of public funding, the elimination of national standards, the devolution of programs and services from the federal government back to the provinces and onto the municipalities, and the move toward a heavy reliance on charitable contributions.”. The 1980s was a transformative era in the history of global capitalism and introduced the adoption of neo-liberal policies in Canada which would further reduce any restraints on the power of globalizing business interests. During this time the nature of social work and the resources available to agencies providing programs and services changed drastically and in most cases, for the worse. In British Columbia, the Social Credit government was re-elected in 1975 following the defeat of the NDP and Bill Bennett led the party to victory in 1983 with his controversial budget “restraint” initiative, bringing about major reductions in the public service sector (which included child welfare). One of the most well-known components of this initiative was a 25 percent reduction of the public service (almost 90 percent of staff who were laid off worked directly with children and families) and the elimination of the family support service workers from ministry offices, despite their reported successes (Scarth & Sullivan, 2007). Four hundred public service positions were cut on July 8th, 1983. On February 20th, 1984 the government introduced a budget that would reduce funds to all ministries and announced their plan to lay off 2,000 public service employees 38  by March 31st of that year (Magnusson, Carroll, Doyle, Langer & Walker, 1984). Bennett’s restraint initiative is an example of the trend in many Western countries to choose budget reductions over increased social spending as a means of reducing the debt crisis. British Columbia’s Solidarity Movement, a coalition of social workers, concerned interest groups, scholars, not-for-profit agencies and labour unions resisted the cuts based on the view that they would harm poor children and families who were already at risk of entry into the child welfare system. Though some of the preventive family support programs were contracted out to agencies so they could continue to provide services to families, the Bennett Social Credit government of the 1980s marks a major shift in child welfare in British Columbia towards residualism. Neo-conservatives and the influence of the Fraser Institute (Mullaly, 1993) espoused ideas like rugged individualism to promote  the goals of the corporate sector in an environment where the political right dominated social and economic policy thought in British Columbia under Bennett. Rugged individualism conceptualizes the family as a self-sufficient unit and state funded programs and services as indications that parents have failed in their roles as providers for children. 2.5.6 The Move to Managerialism Bringing the welfare state in line with the neo-liberal agenda required the careful application of management practices from the business world into child welfare and turned service recipients into state liabilities under a new managerialism (Callahan & Swift, 2006). Managerial controls in child welfare often take three forms (McCroskey, Weil & Finch Jr., 1992, p. 251). First, many service activities under managerial controls experience deprofessionalization. Deprofessionalization reduces child welfare to a set of routine activities under the control of managers, usually out of a desire to centralize control in decision-making as 39  a means of ensuring consistency and standardization over professional discretion. Second, strategies are used to maximize efficiency and cost-effectiveness. Finally, there is an over-reliance on simplistic quantitative measures without proper attention to family and child outcomes to document service effectiveness. These measures are designed to increase the level of control management is able to exercise over front-line workers. Risk assessment was first introduced to British Columbia in the early 1990s and represented a shift towards managerialism in child welfare, a practice that emphasizes process and high levels of control. It organizes people into “risk classes” to find ways to categorize the numbers of people who require state support. The new managerialism returned social and economic problems to the arena of management practices instead of focusing on the needs of the community or societal problems. The professional knowledge of social workers was secondary to managerial knowledge “since managers can carry out required tasks through delegation and standardized controls” (Callahan & Swift, 2006, p. 158).  2.5.7 Residualism: A modern expression of neo-conservatism The gradual policy shift towards reduced government spending through the 1970s and 1980s led to significant cuts to many programs and services for children which had been established as part of Canada’s welfare state. As available resources and staff were clawed back, child welfare services which had been in place to support child development were reduced to protection under the philosophy of residualism. As a field that is influenced by the combination of private troubles and public problems, child welfare tends to focus on the private sphere as the main point of intervention (Wharf, 2006). The philosophical position that focuses on the private as the main source of troubles is known as residualism. From this perspective state programs and services are supposed to be provided as a last resort when parents have abused or neglected their 40  children and exhausted all of their own resources (Wharf, 2006, p. 1). This occurs because states operating from a residualist position do not see themselves as sharing in child-rearing responsibilities. One Canadian historian describes the central theme of residualism: The prejudices of the dominant belief system have regularly encouraged many Canadian child savers, whether private philanthropists, professional social workers or reform-minded citizens, to be blind to collective injuries whether they be of gender, race or class. Not surprisingly, such myopia favours techniques such as casework, but also psychologizing of all sorts that pathologize the individual. In contrast, community development approaches that require a panoramic perspective on children in distress, struggle against a normative vision of the world that resists institutional change. (Strong-Boag, 2002, p. 35)  In practice, residualism seeks to downplay the impact of any public problems on the well-being of children. If family capacity fails to keep children out of harm’s way, child protection from a residualist philosophy is presented as a reluctant but necessary evil which violates the sanctity of family life (Scarth & Sullivan, 2006). On the other hand, the institutional perspective argues that social work is a legitimate and permanent function in society for helping children and families to meet their potential via developmental and preventive services (Northen, 1982). The types of child welfare systems most likely to operate under residualism are referred to as threshold systems (Cameron, Coady & Adams, 2006). Wharf (2006) observed that: “The child-saving paradigm is philosophically consistent with residualism and with a Western/white view of families and children.” The roots of paternalism in child welfare and their modern expression of residualism justified by neo-conservative ideologies and increased managerialism find their ultimate expression in threshold systems and the narrowed response of protection investigations.     41  2.5.8 Threshold Systems Threshold systems in child welfare dominate most Western countries like the United States, the United Kingdom and Canada where individual rights and responsibilities are prioritized values. Parents are primarily accountable for the rearing of children in these systems and state intervention is only to occur at the threshold level where parents have violated the minimum standards of caring for their children. The requirement to demonstrate parental incapacity or harmful behaviour towards children according to child welfare legislation means that workers spend much of their time investigating parents’ faults and gathering evidence for the purpose of taking them to court. It is not surprising that parents who have had multiple involvements with child welfare in threshold systems tend to fear child protection social workers. Brown and Weil (1992), advocating for a shift to a more family-centered curriculum for social services on the other hand, define child welfare as “…the collective and necessary social and familial protections and provisions that should be provided all children and adolescents to assure their health, education, socialization, social membership, opportunities, and fair “life chances”.”. This definition is more inclusive of the developmental needs of children and emphasizes services and programs as a shared responsibility between parents and the state and a fundamental right of all children.  Residualism narrows the focus from child welfare to child protection and results in the unfortunate approach of a child protection investigation. The most persistent critiques of threshold child welfare systems are for the low levels of assistance provided to most families, the rapidly escalating costs of these systems and the lack of tangible evidence of benefits for children and families (Cameron, Coady & Adams, 2007; Buckley, 2000). Services provided tend to be of poor quality when child welfare systems operate from a residual position and parents are 42  stigmatized for dependence on the state when they try to access them. In other words, they become bad services for bad parents. Cameron, Coady and Adams (2007) observe that: “The price of child protection’s single-minded focus on specific conceptions of harm to children is that too many families resent and fear an unwanted and unhelpful intrusion into their lives”. Forcing parents to accept the majority of the responsibility for child maltreatment without consideration of family context or its underlying causes is an insufficient response.  2.6 Differential Response: A Post-Modern Hybrid 2.6.1 Managing Volume after Gove and Risk Assessment Differential response can best be understood as a post-modern hybrid of residualism and community partnership in the service of managing volume and reducing state intervention in the lives of children and families. One of the events leading to the implementation of FDR in British Columbia was the tragic death of five year-old Matthew Vaudreuil in 1992 which resulted in major reforms to the province’s ailing child welfare system. Minister Joy McPhail ordered an independent inquiry in 1994 chaired by Justice Thomas Gove and by November of 1995, a report was released with 118 recommendations with the statement that “care and protection of the child was to remain paramount in all future decisions” (Armitage & Murray, 2007).  The CFCSA was introduced in 1996 and listed the safety and well-being of the child as paramount considerations in the aftermath of the Gove Inquiry, adding to the already risk-averse culture within the Ministry. The definition of child maltreatment in the new act included the phrase “has been or is likely to be harmed”, leading to the implementation of a risk assessment model which pledged to fulfill the need to predict risk based on a comprehensive assessment. At the same time the Ministry was undergoing major staffing changes and reduced federal transfers for social spending. Risk assessment was hurriedly put into place with training completed by 43  1,526 staff by the end of April, 1997 during a time when many experienced workers were offered early retirement packages to save costs and 180 social workers (many of them new) were assigned to implement the new CFCSA. Workers anticipated the opportunity to use the Comprehensive Risk Assessment (CRA) to formulate plans in partnership with families with resources to support them in this task. Unfortunately, workers were not able to predict that the risk assessment could be used as a management tool to triage resources rather than expand their development by limiting resources to those who had the highest risk assessment scores. English and Pecora (1994 as cited in Callahan and Swift, 2007) observed that most jurisdictions choosing to implement risk assessment models have done so in response to tragedies leading to overwhelming numbers of protection reports to justify narrowing the focus of child welfare.  Workers and community professionals alike began to make reports with increased frequency in the aftermath of Matthew Vaudreuil’s death to avoid another tragedy. The Ministry soon became overburdened and without adequate funding to meet the swell of new incoming reports, the numbers of children in care in the province started to rise after steady declines from the previous two and a half decades.  Income assistance, a program responsible for serving many children in care, was moved to the Ministry of Human Resources, effectively severing its long-standing connection to child welfare and leaving many families without financial support. Minister Penny Priddy made the impossible promise that there would never be another Matthew Vaudreuil in British Columbia in the face of ongoing reductions and tightening eligibility requirements for income assistance and child welfare services for needy children and families. The recommendations of the Gove report did not coincide with an increase in government spending and the Ministry was forced to introduce further budget restraints to manage the burden.  44  2.6.2 Old Wine in New Bottles In 1992, June Brown and Marie Weil proposed a family-centered ecological model for social services that in many ways resembles differential response and illustrates how DR may be “old wine in new bottles”. One of the failures of residualism in child welfare has been that parents are often left bearing the responsibility and stigma associated with child maltreatment. The authors argue that when child protective service agencies are not adequately resourced to keep children safe in care or to follow a reliable permanency plan to completion, the protective effort itself can become neglectful and abusive. Writing during the time of family preservation and risk assessment, Brown argued against service “targeting” and restricting the definitions of child abuse and neglect. An appreciation of child welfare history shows that: “Past, hard-won lessons have taught that narrow definitions of family and child welfare problems and limited fragmented service responses do not serve well; and current data reveal that today’s families, young children and adolescents are facing a vast array of unmet needs, stressful circumstances, and difficult, even tragic life events” (Brown, 1992, p. 35). The values of the ecological model proposed by Brown (1992) mirror those of differential response. The belief at the core of the model is that every child has the right to the best possible chance at life and should be provided with the essentials for proper development. Belonging, safety, socialization, permanence and opportunities for personal growth and mastery are all components of a family-centered model as well as differential response. These values fit with the institutional view of child welfare that affords equal developmental opportunity for all children. The model and differential response also share the common assumption that children require nurturing environments where they are valued, where they matter and can be supported to achieve their developmental milestones. Both value the family as the ideal unit in which 45  children should be reared, both acknowledge the importance of the family as a system interacting with other systems and incorporating the extended family as an important aspect of family support and functioning. Both the proposed ecological model curriculum and differential response strongly emphasize the families’ right to self-determination, the child’s right to grow up in his or her family and assume that the family will act in the child’s best interests. In the event that the parent cannot act in the best interests of the child, both the ecological model and differential response recognize the importance of the right of the child to be protected by the community if they are at risk in their own homes (differential response systems maintain the investigative pathway if risk to the children arises).  Family development response and the model proposed by Brown and Weil are informed by the ecosystems paradigm and favor a comprehensive approach to the way family services are designed and delivered. This framework argues that behaviour patterns and problems (including child abuse and neglect) must be understood and resolved with a full understanding of how people and their environments are interdependent and in constant exchange. General systems theory assumes that proper functioning is obtained through the achievement of adaptation between person and environment. It is a dynamic concept that views families as systems that must adapt to their internal and external pressures and can influence and be influenced by them in their development (Sullivan, 1992). This model also shares differential response’s conviction that families are interdependent contributing social units and that “[t]hey therefore require fair access to resources and opportunities that are essential for their functioning and for their ability to carry out family obligations, especially child rearing” (Brown, 1992, p. 44).  The need for preventive services in child welfare has been acknowledged in both the ecological model and in differential response and in much of the research on child maltreatment. 46  Research has revealed the detrimental consequences of neglect on child development as children who are raised without adequate nurturance are at risk for physical, cognitive, emotional and social under-development. The ecological model and differential response both propose outreach services as a natural response to emerging child problems and suggest that: “Service at this level would make comprehensive efforts to prevent the progression of problems that are likely to lead to serious neglect and abuse if they go unchecked, and ultimately to family breakdown and separation.” (Brown, 1992). Jurisdictions choosing to implement differential response tend to do so when the need for preventive services has become too clear to ignore. Family Development Response emerged in British Columbia as an effort to manage the burdens of the immediate context of high call volumes and increased children in care by partnering with the community towards the goal of reduced state intervention. It was but one initiative among others designed to meet the goal at the time of family preservation and reversing the negative image of the Ministry as unnecessarily intrusive and residual. Differential response models “appeal to community organizations in supporting families, while recognizing the importance of building an alliance with families to promote family engagement” (Kyte, Trocme & Chamberland, 2013) and given the immediate context of implementation, wherein defrayng costs by keeping families out of the system was the modus operandi, a more appropriate title  may be family ‘deferred’ response. Current research and key findings will illustrate how temporary interventions like differential response may hold potential for certain types of cases but are in danger of failing to maintain positive changes if deeper systemic and environmental problems go un-addressed.     47  2.7 Current Research and Key Findings A recent meta-review of studies from 2000 to 2012 by Kyte, Trocmé and Chamberland (2012) on differential response summarizes some preliminary findings on effectiveness, potential areas for future research and some important questions to consider about program evaluation and implementation. Most studies evaluating the effectiveness of DR consist of quasi-experimental designs due to their ethical inability to randomize into investigative or DR tracks (Conley, 2007; English et al., 2000; Marshall et al., 2010; Ortiz, Shusterman, & Fluke, 2008). One exception is a study conducted by Loman and Siegel (2008) on the Minnesota Alternative Response where families screened in as appropriate for differential response were randomly assigned into DR or investigation pathways but the study does mention an exercise of discretionary power when families were assigned to control or experimental groups.  Most studies begin by trying to match a sample of DR and investigation families based on social demographic characteristics then use recurrence (re-entry, recidivism, re-entry) as an outcome measure indicating whether the child’s safety has been maintained within their respective response pathway. There is variation between studies on calculating recurrence with some studies looking at case closure at varied intervals (eg. one month, six months, 18 months) and some studies calculating recurrence during the intervention period. The dependent variable for many of the studies on differential response appears to be recurrence despite its limited validity as an accurate measure of child safety and well-being. One of the major findings in support of DR is that child safety does not appear to be compromised. Lower recurrence rates, longer time periods between subsequent reports and reduced severity of protection concerns in subsequent reports all seem to suggest that at the minimum, DR does not cause additional harm to children. The finding that differential response 48  families tend to have lower rates of placement than traditional investigative approaches was supported in three out of the six studies examined by the authors and positive outcomes were discovered on measures of family satisfaction and engagement, mostly through qualitative analysis. DR workers had a tendency to know and could report more information about the families they worked with than investigative workers and services were more quickly provided by DR workers. Workers and families both preferred the DR approach than they did investigations and the very implementation of a DR system appears to have helped improve the quality of investigations (Loman, 2004). Reviewing the short-term outcomes of studies where families had received differential response revealed a decreased number of placements and increased reunification with families. With regards to implementation, the authors noted that “[a] major limitation from previous studies is the inability to clearly identify in what circumstances referral to DR becomes appropriate” and that “[m]ost studies refer to arbitrary markers of risk and potential harm, funneling those families that fall under a ‘low-medium risk’ category into DR” (Kyte, Trocme & Chamberland, 2013). Clearer guidelines around screening and assignment may be necessary in light of the fact that many states that have implemented DR have screened between 60 and 80 percent of screened in reports to DR.  Finally the authors identified the important relationship between reports of neglect, families living in poverty and assignment to DR in that “[m]any cases traditionally classified as neglect in an investigative pathway look like straightforward poverty in DR” (Kyte, Trocme & Chamberland, 2013). The authors drew on the work of Siegel’s (2012) impact analysis which showed that families most likely to be helped by DR were poor families where the difference between child neglect and poverty was not easily distinguishable. The voluntary nature of DR pathway is questioned by the authors as they found that some jurisdictions directly referred 49  families to the traditional investigative track following a families’ refusal to participate in DR intervention. The question of whether the distinction between DR and traditional CPS investigations is based on concrete markers of a change in policy or approaches to clinical practice lingers as the challenge remains to demarcate differential response from effective clinical social work practice (Kyte, Trocme & Chamberland, 2013). Future research will need to focus on finding ways to determine whether DR is effective in reducing child maltreatment.   Fallon, Trocme and MacLaurin (2011) reviewed the Canadian Incidence Study of Reported Child Abuse and Neglect (CIS), the National Child Abuse and Neglect Data System (NCANDS) and the National Incidence Studies of Reported Child Abuse and Neglect (NIS) to argue that children and families who were identified as “at-risk” of maltreatment presented with as many household and caregiver concerns as those from substantiated investigations. According to the authors differential response represents an attempt to distinguish between protection investigations and assessments where family functioning and children’s needs are prioritized. As indicated by the previous review, the merits of differential response appear to include higher satisfaction rates with services and workers and greater collaboration and engagement in service planning and decision-making. Short-term outcomes include decreased numbers of placements and increased re-unification rates.  The authors highlight the need for more research about which families would best be served by differential response. Based on previous research the authors identified case factors more likely to be associated with the use of differential response including younger age groups, referrals from non-professional sources like parents, relatives, friends or children, referrals from social services or school staff; reports for neglect, emotional treatment or witnessing domestic violence, and a history of previous child protection involvement. The authors also indicate the 50  “lack of consensus on specific guidelines to inform decisions about differential response and decisions are often left to the discretion of workers or supervisors” (Fallon, Trocme & MacLaurin, 2011).  English et al., (2000) examined outcomes for 1,263 “low risk” referrals from Washington State’s Community-Based Alternative Response System (CBARS) and found that referrals to alternative response systems from screening were inappropriately high. CBAR referrals were divided into three groups: 1) a no service group (n=251), 2) an assessment only group (n=409) and 3) an assessment and at least one other CBAR service group (n=603). Data was collected between 1992 and 1995 with 18 months of post-service follow-up. Case characteristics and outcomes were compared between those within the CBAR group who accepted and those who refused treatment. Outcome measures were re-referral and post-service placement. The authors found no significant difference in re-referral between the CBAR group and investigative group within 18 months of case closure and no significant differences between the three groups of CBAR referrals. Significance of re-referral rates were found in cases where reports had come from PHN’s, cases of domestic violence and cases where parents were abusing substances.  Re-referral rates were similar between families who did or did not engage in assessment services and were highest for families where domestic violence was present. The authors concluded that the inhibitory effects of differential response on re-referral rates were modest and short-lived based on significantly lower rates at the six-month but not at the 18 month follow-up period. It could be the case that more intrusive approaches past the 18 month period are deferred by providing families with temporary interventions associated with differential response (eg. temporary parenting support, bus passes and food vouchers and community referrals). Latent risk factors, like exposure to domestic violence and parental mental illness/addiction; however, are 51  very complex problems and require the coordinated involvement of multiple professionals like police and mental health teams.  Differential response was not intended to be a long term intervention and the assessment phase in the literature ranges from anywhere between 60 to 90 days (OCFS, 2011). If the researchers found that re-referral rates were similar between groups at the 18-month mark and re-referral rates were statistically significant for protection reports coming from PHN’s, cases of domestic violence or cases where parents were abusing substances, differential response may not be accomplishing its objective of deflecting low-risk families away from CPS. This may also be the reason the authors found that referrals to differential response seemed inappropriately high. Seemingly low-risk reports about child neglect may be the first visible signs in the community of major problems in the home.  The only example of randomized assignment between differential response and investigation groups occurred during implementation of the Minnesota Alternative Response Project which was piloted in 20 counties in 2001 and was expanded statewide afterwards (Loman & Siegel, 2008). At minimum, each family had to be screened in as appropriate to receive a differential response, then assigned randomly with confirmed comparability analysis to either an investigation (1,035 in the control group) or differential response (2,860 in the experimental group). Data was obtained from February 2001 to December 2002. Analysis of twelve safety areas (food and nutrition, clothing, personal hygiene, shelter, hygiene, health care, supervision, abandonment, physical violence, emotional and/or sexual abuse) in treatment phase and follow-up revealed no evidence of greater declines in child safety between the two groups. In fact, they found that when all individual categories of change in child safety were considered, safety improvements totalled 47.7 percent for experimental families and 31.8 percent for control 52  families. The authors also found no difference in report recurrence of child abuse or neglect during the initial phase, indicating no decline in the safety of the child. Higher engagement and cooperation levels with families, higher levels of therapeutic services provided, lower rates of removal and foster care placement and more concrete assistance received by families were all positively associated with assignment to differential response. The authors found a small but statistically significant reduction in re-referral rates among experimental families and suggest that this difference may be attributable to the new approach as well as increased services. The authors state “…it showed that, in addition to offering more comprehensive services, positive benefits were achieved – independently – by changing the way workers approached families” (Loman & Siegel, 2008).  Based on analysis of two time periods for each family within a sample from experimental and control groups, cost analysis based on service costs (by local bookkeepers), and calculated staff time costs (from worker logs) determined that overall costs associated with families were lower under differential response and the authors concluded that “savings achieved by experimental families later more than offset investment costs incurred during the initial contact period” (Loman & Siegel, 2008).  Regarding outcomes for children and families, there were no consistent differences reported between groups on overall well-being, health, behaviour and relationships as well as academic progress of children one year following final contact with CPS. No differences were found in caregiver reports on improvements regarding their relationships with their children, methods of discipline, ability to care for their children, home or living arrangements, emotional or financial support from friends. It seems clear that the differential response families received more and better quality service and were more satisfied with their overall experience with workers, but the outcomes for the children appear to be similar. While safety does not seem to 53  have been compromised for children whose families received differential response, a lack of clear guidelines at screening differentiating families who are appropriate for differential response and families who are appropriate for investigation may be a confounding variable for these findings.   When the decision was made to refer a family into the differential response pathway, there was often considerable latitude from the screener in deciding on one pathway or another, except for cases of the most egregious harm or possible harm to children. The authors note that the proportion of families screened in for a differential response varied from 27% to 61% percent depending on which of the twenty counties the report was received in. The range in most counties (14 of 20) was between 45% and 60% assigned to differential response. This suggests that screeners may not have been making very clear distinctions between response pathways or suggests that even though families were randomly assigned into DR or investigation groups, there may have been a certain degree of unplanned randomness within the DR group, leading to a possible over-estimation of effect sizes.  Re-assignment of response pathway in the Minnesota Project between differential response and investigation was possible based on new information. Reassignment from differential response to investigation occurred in less than 5% of cases, but reassignment from investigation to differential response, on the other hand, occurred in less than 1% of cases. The authors conclude that even though the findings in the study were statistically significant, they were modest, meaning that any changes observed were unlikely to happen at random but not significant enough to indicate major changes in outcomes for children and families. The identified strengths of differential response appear to be more in changing the approach of the worker to emphasize cooperative relationships than in outcomes for children in the long-term. 54  That differential response did not compromise the safety of the children when compared to investigation is a promising finding but its limited inhibitory effects on abuse and neglect and wide latitude for screeners in placing too many families into the DR group may be problematic and may confound positive results.  Conley and Duerr Berrick (2010) examined the outcomes associated with Alameda County’s “Another Road to Safety” differential response program. The authors used survival analysis of California’s three track system (community response only, community and child welfare services response in combination, and child welfare response only) in which families screened out of protective services are referred to the differential response pathway (voluntary, home-based services). ‘Track-one’ families were those cases which did not meet the statutory definitions of abuse or neglect but where families were experiencing difficulties that could be addressed by services received within the community. The study compared 135 children in the treatment group of track one to 511 control children who were eligible for services but were denied due to program capacity. As with other studies on differential response, the authors note mixed results when it comes to re-reporting of families following case closure, placement and out-of-home care. These mixed findings remain consistent even after follow-up periods of six months or more. The study also echoes previous findings that the rate of substantiation is low enough to not have sufficient power to detect differences between the two groups.  The study found no statistically significant differences between groups on three variables: likelihood of a re-report following participation in differential response, timing of abuse and neglect reports, or subsequent investigations. Children from the differential response group were more likely than those in the control group to have had prior child maltreatment reports made about them. The authors advocate for providing families with support but based on the results of 55  their study, acknowledge insufficient evidence to support the hypothesis that differential response reduces the likelihood of future reports of maltreatment. One important trend identified by the study’s authors and noted in their literature review is the tendency for families assigned to the differential response pathway to be assessed as “high-risk” or even “very high risk” by staff despite the intention that differential response is only to be used in low to moderate risk cases. The explanation offered by the authors is that “very troubled families are regularly brought to the attention of child welfare agencies, and even these families often are turned away from services” (Conley & Duerr Berrick, 2010). This study suggests that child welfare agencies may struggle with how to assign those families where children have needs for service but may not fit into either pathway.  Overall, the study coincides with previous findings which describe a modest treatment effect of differential response on the prevention of subsequent maltreatment. Conley (2007) critically examines the ‘Another Road to Safety’ program in California to highlight questions about CPS’s approach to protecting children based on mounting evidence of the similarities in trajectory between substantiated and unsubstantiated child protection reports. Her first argument is that CPS services must be narrowed to avoid unnecessary intrusion into the lives of families with an individualized approach. This argument is supposed to address the problem of over-inclusion, or those families that have been screened-in who should not have been, and lead to overburdening of child welfare systems. As mentioned in the study on the CBARS program, there are families where investigation might not be necessary but who still require service. The concern is not serious enough to warrant immediate intervention from CPS through an investigation but ongoing case management and coordination is in the best interests of the children. As mentioned previously, these cases may be costly and often involve multiple 56  systems. They may also be the types of cases that are most appropriately managed with the integrated approach discussed previously by Brown and Weil (1992). The second argument she makes is for “broadening of child welfare services to create a system that truly promotes child well-being rather than intervenes only in desperate situation(s?)” (Conley, 2007). This time the problem is under-inclusion, or families where children need help but have not met the threshold for being included under the residual mode of investigation. Both arguments point to the same problem on a larger scale: that child welfare has been pared down to the point where the right assessment through regular contact no longer leads to the right services provided at the right time to meet needs other than those of the child’s immediate safety. In an attempt to address such concerns, the Another Road to Safety program involves consent by parents to meet on a weekly basis and receive voluntary services via worker referral. Caseloads are to be no larger than thirteen and on average, nine. Visits of approximately one hour each week ensure the development of a working relationship with clients to obtain a better understanding of the family’s strengths and needs. The use of the purposeful relationship is not a new idea (Perlman, 1979) but is given significant weight as a part of an effective assessment in differential response. Within thirty days, the worker develops a plan in partnership with the family to be regularly monitored with the use of developmental and health assessments of all children at regular intervals. The program lasts for nine months. During this time, workers have access to a discretionary needs fund for the purpose of preventing crisis should an urgent and unaddressed need arise. According to the author, the worker develops “a therapeutic relationship that is the intervention tool with the family” (Conley, 2007). Perlman (1979) identified the three components of a professional helping relationship that may have informed the formulation of differential response. As in differential response, the professional relationship is formed for a 57  mutually agreed-upon purpose, it is time limited and it is always for the client. The goal is to provide for a family’s basic needs with practical support, direct face-to-face time with clients and a plan that will lead to parenting changes and life skills that will promote the safety and well-being of the child, all within the context of the helping relationship.  Differential response, while voluntary, can lead to referrals back to child protective services if the family refuses. When families become aware that they can be re-referred to CPS if they refuse to engage with differential response, the voluntary nature of participation comes into question. The dual role of the child welfare system is to simultaneously investigate maltreatment reports and remove children who are unsafe while promoting the preservation of the family and offering families supports. There are many factors which could contribute to a family not being willing to participate including any parental resistance to regular visits (mental illness, domestic violence or substance use), chaos in the community, or agency factors such as long waitlists, limited resources or delays in providing service. Other complications such as lack of clarity around the ideal length of time to be involved with families, negative previous experiences with protective services, lack of clarity around whether paraprofessionals are effective at providing differential response with adequate training and supervision, and a lack of resources all factor into how families actually receive the intervention. Some jurisdictions employ paraprofessionals or workers who provide service without the formal title of child protective service case worker, while others rely on the child protection social worker to provide differential response. All of these factors will contribute to how services are provided and by extension, how services might be evaluated. Despite the problems associated with implementing differential response, the importance of tailoring the service to meet the developmental and safety needs of the child is of critical importance regardless of title.  58  Another study on differential response in eleven small, rural counties in Northern California is suggestive of achievable goals if families were willing to accept services, resulting in the authors advocating for greater service availability and effective strategies for engagement with families (Franke, Bagdasaryan & Furman, 2011). The researchers attempted to answer the question of whether there are differences in demographic characteristics, maltreatment type, and presenting problems that differentiate families assigned to differential response pathways when compared to families who were served by community-based organizations referred from sources outside of CPS. As in previous studies, the 90 families who participated in the study were in one of three tracks. California’s three-path model of differential response divides families into three types: the first path includes cases that do not meet statutory definitions of maltreatment and involves referral of families to Community-Based Organizations (CBO’s). The second path includes a joint response by CPS and CBOs for cases that meet statutory definitions of maltreatment but risk of maltreatment is low to moderate. Path three is reserved for traditional CPS investigation and response. Study findings indicated that counties were assigning families differentially based on expected level of risk and demographic patterns but the services received by families were less differentiated. Their literature review suggested that families assigned to differential response were typically for concerns about parents not supplying basic needs for the children, having conflicts with their children and having more than one child in the home. These were also less families who were less likely for the report to be about physical abuse. Allegations of sexual abuse almost always necessitated investigations; a finding consistent with the fact that most differential response systems prohibit sexual abuse cases being assigned to non-investigative pathways.  59  As shown with previous studies on jurisdictions where differential response has been implemented, CPS workers become more intrusive when families experience severe problems. The authors advocate for “a shift in the system of intake assessment such that attention is given to cases that under the traditional system would not have received further consideration or services” (Franke, Bagdasaryan & Furman, 2011). The authors also identify the challenge of providing services to families in rural areas due to scarce resources and limited service capacity to meet the demands of differential response implementation. Differential response requires a broad range of resources to be effective. The authors conclude that only a small proportion of families achieved their identified goals in the differential response pathways due to the disconnect between the formulated policies of the child welfare system and what can realistically be implemented on the front lines without the resources to do so. On a positive note, those families who did receive service were more likely than those who did not receive any services to accomplish the goals of their service plans.  Perhaps most relevant to this thesis is a recent (2008) study conducted by Marshall, Charles, Kendrick and Pakalniskiene on the efficacy of British Columbia’s Family Development Response in relation to lower rates of re-entry into CPS and child removal over a 20 month period in one region.  The researchers found that there was no significant difference between FDR and Investigation (INV) groups in recidivism rates into CPS. In addition, parents/guardians in the FDR pathway were no more likely to initiate an intake with CPS than were parents/guardians in the INV group. There were no differences in number of intakes, parent-initiation of intakes or length of time between intakes; however, the outcome/status of cases by the end of the study did differ significantly. While safety of the children was similar for both groups, fewer children assigned to the FDR group were removed or in MCFD care when 60  compared to the INV group. This means that FDR is meeting one of its primary objectives to decrease the number of children who come into the care as well as precluding more intrusive measures. The authors also suggest that “FDR helps the child protection system in British Columbia achieve success at matching child and family needs with an appropriate type of intervention” (Marshall et. al., 2008).  A review of the literature on differential response in North America tends to focus on rates of re-reporting following case closure, rates of removal and placement of children in foster care, and satisfaction ratings among workers and clients of differential response. Findings from most studies indicate positive albeit modest and short-lived effects of differential response in outcomes in the first two categories but positive feedback in the third from workers and clients who report high levels of satisfaction after receiving and implementing differential response services. The majority of studies report that while differential response may have a modest positive effect depending on availability of services and other factors, child safety does not appear to be compromised for families referred to the differential response pathway. Studies also have highlighted the importance of clear definitions and inclusion parameters for differential response as spelled out in legislation.  Most of the studies of differential response to this point indicate modest and temporary inhibitory effects on child abuse and neglect following differential response services. Families who are referred to differential response because they are classified as low-risk tend to be those families where neglect is the primary concern but the reasons for neglect may be outside of the purview of a short-term preventative intervention. Exposure to domestic violence, poverty, addiction or social isolation, mental illness or other factors can all complicate the delivery of differential response assessment and services. These are complex problems that may have more 61  to do with service cutbacks in different social welfare programs than they have to do with the families’ shortcomings. Waldfogel (1998) identifies a number of problems associated with child welfare systems that constitute additional impediments to the delivery of services via DR. These include over inclusion of families that should not have been referred to CPS, numbers of reports exceeding the systems capacity to respond effectively, under-inclusion of families who should be referred to CPS but are not, an authoritative service orientation, and lack of appropriate services. Families in residual threshold systems of child protection may require more than such systems are able to provide. It is precisely these types of families that require child welfare to be institutionalized. Reading between the lines of British Columbia and Canada’s child welfare history we see how differential response might be an attempt to address the failures of residualism in child welfare. Services which have been cut under retrenchment policies have narrowed what qualifies as a protection concern. Before child welfare became merely child protection, the physical, cognitive, mental, emotional and social needs of children were all thought to fall within the purview of child welfare. When children do not qualify as having a protection concern they do not receive services and when they do, these services tend to be of poor quality.  Clients and workers report greater levels of satisfaction and there does seem to be some evidence that differential response, at the very least, does not compromise the safety of children. Findings about differential response must be made tentatively for several reasons. First, most research on differential response examines secondary sources of data where randomization is not possible (except in the abovementioned study). Since differential response families are supposed to be screened in as low-risk, the finding that it does not compromise the safety of children may be an artifact of the previously assessed risk level which may not have been high to begin with. 62  This is just one example of possible confounding factors on research on differential response. Second, differential response is a relatively new paradigm in child welfare but the ideas which inform it have existed in child welfare since the 1960s and earlier. If there is anything that makes differential response more effective than a traditional investigation it should be remembered that many of the clinical skills and approaches contained in differential response have existed in child welfare practice since before DR existed. The collaborative relationship, integrated and community-based practice, effective planning and case management skills emphasizing permanency for children are all elements of effective child welfare practice and can be used in either DR or investigative response.                63  3 Organizational Theory  3.1 From Child Welfare to Child Protection The previous chapter put the shift from child welfare to a residual approach in child protection in historical context. The purpose of this chapter will be to apply organizational theory to understand why MCFD implemented FDR beginning with a look at three common critiques of threshold systems. The critiques will introduce how organizations like MCFD respond to external pressures and demands which will then be related to the internal systems adaptations designed to meet the demands and resources of the ecological context. The demands and opportunities of the ecological context with its consequent resource implications will set up the bifurcation of MCFD’s response into FDR and INV and draw into consideration the contextual factors that take precedence in influencing decision-making regarding pathway designation and ultimately the research questions drawn here. 3.1.1 Critiques of Threshold Systems  Threshold systems require that families meet a minimum level, or threshold of dysfunction to qualify for inclusion into the child welfare system. Cameron, Freymond, Cornfield and Palmer (2007) summarize three common critiques of threshold systems. First, the number of child abuse reports has increased dramatically within threshold systems leaving child protection agencies inundated and perpetually under-resourced. This is partly because every child maltreatment report in such systems requires a formal investigation and partly because policies guiding practice tend to draw more families towards investigation. Child protection agencies in threshold systems manage this volume by relying on managerial procedures to ration services to families in order to maximize effectiveness and efficiency.  64  A second criticism of threshold systems is that they formally investigate too many families whose problems stem from insufficient resources or difficulties with childcare responsibilities. Threshold systems draw resources away from intervening when children are truly at risk of harm because so much time is devoted to investigating families who might benefit from less coercive measures. A third criticism of threshold systems has to do with their dual mandates of providing assistance for families while simultaneously exercising legal control over them. Control tends to be emphasized in threshold systems and the negative tension between the two functions has led critics to argue for a separation of investigation and support.  The lack of quality services in threshold systems and the negative valuations of those who receive them stem partly from the child-saving movement’s drive to rescue children from “immoral” parents and communities. It is evidenced in the fact that the disadvantaged represent the majority of those who come into contact with child protective services. First Nations children are in care disproportionately to the general population and it is often underprivileged families from which children are removed. Single mothers are also disproportionately represented in child protection statistics.  British Columbia has had the highest rate of child poverty of any province in Canada for several years, and the highest poverty rate both for single mothers and for children living in two-parent families, as well as the greatest income inequality (First Call BC, 2013). The neo-conservative philosophy used to justify the retrenchment of services towards a residualist system has reduced services to the point where rationalization is required to meet the expectations of external pressures and demands in threshold systems.   65  3.1.2 Rationalization of Resources Organizations must find ways to ration valuable resources if they are expected to perform their designated functions as effectively and efficiently as possible. When the task of caring is delegated away from primary groups (extended family, friends or neighbours) to formal organizations, there is pressure to routinize care for efficiencies’ sake (Hasenfeld, 1992). Hasenfeld (2010) explains that organizations providing human services will inevitably face pressure to reduce numbers of service recipients where feasible, meet quotas for service, curb costs and comply with accountability measures based primarily on quantity. Many Western child welfare jurisdictions tried to meet this goal by using risk assessment to prioritize reports.  Ideally, risk assessment was supposed to be used as a comprehensive tool to help in the difficult but necessary task of prioritizing reports and rationing resources to children and families most in need but Parton (1998) discussed how risk assessment had a tendency to dehumanize families by “dissolv[ing] the notion of a subject or a concrete individual and put[ting] in its place a concern with a combination of risk factors”. Workers in CPS agencies may have experienced tension when the essential focus of policy and practice shifted from client relationships to monitoring and managing abstract factors determined to produce risk for children, forcing them to choose between adherence to standardization and individualized services that might take them beyond the limits of the regular work-day.  A reliance on standardized procedures impinges on the worker’s ability to exercise professional judgment, seeks to maximize the level of managerial control and assumes universal agreement between workers on what constitutes child abuse that can be responded to accordingly. Workers become ‘passive agents’ in administering service to families in environments that meet agency requirements with little flexibility (Stevenson, 1997). In these 66  environments child protection social workers must find ways to prioritize highest risk cases and triage services without compromising the goal of nurturing the development of children and fostering positive community relations (Buckley, 2000; Smith & Donovan, 2003; Parton, 1998). Inevitably, there will be many families who have not met the criteria of highest risk and remain under-served, damaging CPS’s relationship with the community and leaving a significant population of children neglected. Organizational theory will be used to explain how differential response represents a systemic adaptation in child welfare to address the demands of its ecological context. 3.2 Defining Organizational Theory Organizational theory is a subcategory of general systems theory and assumes that the unit of analysis (in this case MCFD) is in constant interaction with its environment. Much like a biological organism, a human service organization (HSO) self-regulates in response to external and internal stimuli. Adaptation is said to occur when stimuli is processed through external and internal feedback loops to meet the systems’ compulsion to achieve balance or homeostasis (Bertalanffy, 1973). Differential response from an organizational theory perspective is an example of a systemic adaptation to manage the demands associated with existing in an ecosystem that has a compromised relationship with its community as a consequence of residualism which sets the threshold for service at the point where harm to a child has either occurred or is thought likely to. Put another way, FDR can be understood as one of the Ministry’s attempts to address its damaged relationship with the community as a consequence of narrowed service provision while also rationing scarce resources.    67  3.2.1 Responding to External Demands and Pressures Organizational ecology assumes that the environment, through which resource availability and distribution is achieved, is a key influence on an organization’s service delivery system. Institutional theory adds to this assumption to the fundamental contention that socially constructed appropriate organizational practices and behaviors are supported by a broad range of external actors and forces (Tucker, Baum & Singh, 1992). The theory contends that those organizations that best reflect the expectations and demands of the institutional environment in their service systems and operation will garner the legitimacy, support and resources necessary to survive and grow. The expectations and demands of the institutional environment are themselves the outcomes of negotiations among interest groups which can include political and civic interests, social movements, professional associations and other concerned stakeholders. The state, in particular, plays an important role in controlling the allocation of programs and resources and in jurisdictions like British Columbia where child protection is the delegated responsibility of the provincial government, this control is complete. Tucker, Baum and Singh (1992) consider executive management styles along two continua: centralized/decentralized authority and an internal/external orientation. A centralized-external pattern of executive management is “geared toward outside environments so as to ensure legitimation of the organizational domain and resource recruitment” (Tucker et al., 1992, p. 102). The power and authority to make decisions is held by the Director (in the case of MCFD, the Minister) and administrative and professional staff such as line workers and team leaders are generally not involved in suggesting innovations and strategic options suited to match the dynamic environment. This pattern is described by the authors as appropriate for “organizations that operate under conditions of economic uncertainty, shortage of resources and 68  continuous budget cutbacks” (Tucker, Baum & Singh, 1992, p. 108). The executive directs their efforts towards the task environment to respond efficiently and centralizes management in an effort to maximize effectiveness with limited resources. Chapter two demonstrated how CPS agencies like MCFD have been historically susceptible to the changes and demands of the environment and would therefore be more likely to have an external orientation. They often respond to crises by centralizing control which leads to standardization and formalization of response. In child welfare where the environment is turbulent and leadership is constantly changing, it appears that “executives who concentrate on internal problems and maintenance of the organization without planning for the future have no choice but to adopt a style of management geared toward crisis management – solving immediate problems and removing obstacles as they arise” (Tucker, Baum & Singh, 1992, p. 105).  One needs only to examine the recent history of child welfare in British Columbia for examples of how CPS agencies like MCFD are pressured to respond to demands of their ecological contexts. Cutbacks to child welfare through the 1980s were justified by moral and political ideologies emphasizing family privacy and limited government intrusion, and the resultant policies and practices left CPS agencies like MCFD unable to respond to the needs of many children and families in the community. As a consequence, the volume of intakes began to swell through the 1980s and into the 1990s until the structural weaknesses of British Columbia’s threshold child welfare system were exposed to the public in 1994 with the tragic and highly publicized death of Matthew Vaudreuil, a child who was known to the Ministry but not in its care. Risk assessment was implemented in 1997 shortly after the 1995 Gove Report as the Ministry’s attempt to address the flaws of the system but it did not reverse the trend of financial restraint.   69  3.2.2 Internal Systems Adaptations The example above illustrates how environments in which child welfare agencies exist influence how services are provided to clients. Evaluations about the state’s role in the lives of parents with children, contemporary political ideologies about the role and function of the family in society and agency funding priorities informed by competing interests from different community stakeholders all have the potential to influence agency decisions about who receives services and what services look like. Decisions like these are necessarily moral and human services, by extension, are also moral.  Human services organizations adapt to the demands and resource capacities of their ecological contexts and are also constrained by prevailing morals and sometimes conflicting values. Child protection, like the work of other HSOs, is necessarily moral and depends on its institutional environment for legitimation of its activities and resource demands (Hasenfeld, 1992, p. 9). This is usually accomplished by “adopt[ing] and uphold[ing] moral systems that are supported by significant interest groups and organizations” (Hasenfeld, 1992, p. 10). CPS agencies usually conform to the prevailing ideologies of the day about children and families. These dominant ideologies in turn play a crucial role in the shaping and delivery of service technologies because “organizations are likely to adopt technologies that are sanctioned by the institutional environment” (Hasenfeld, 1992, p. 13). Service technologies will reflect practice ideologies (ie. what is ‘good’ for clients) and efforts will be directed towards ensuring that clients live up to expectations. Client compliance with technological requirements signals which types of behaviours are considered appropriate, thereby reinforcing social control and giving homage to the source of their legitimacy. The organization directs resources to those activities and initiatives that will feed its legitimacy and when the demand for resources inevitably 70  outstrips their supply in human service organizations, Hasenfeld (1992, p. 6) argues that the rationalizing of resources to clients becomes a moral act because it conveys an evaluation of the client’s social worth, reifying notions of deserving and undeserving poor discussed in chapter two. Demands for productivity might conflict with demands for attention to human needs in these scenarios and the former usually prevails. One example well-documented in the literature is the fluctuation between dual mandates in child protection. Hayes and Spratt (2014) explain how child welfare systems have come to oscillate between the dual mandates of immediate protection (child-saving) and preventative services (family preservation). First, the need to expand the definition of child maltreatment was informed by research demonstrating that only a small proportion of children actually being maltreated were being reached on the narrow ground of child maltreatment definitions while familial and community factors known to predict poor outcomes for children went unaddressed. Second, a tendency to prioritize the most serious cases of abuse and neglect based on media-led public scrutiny has resulted in child welfare systems being judged as effective to the extent they are able to identify and act on such cases. The resulting net effect has been:  …system oscillation as the policy to widen the protection net lurches back to worst case prioritization, often in reaction to child death inquiries. In seeking to correct such system oscillation, the tendency in Anglophone countries has been to seek an elusive fulcrum to balance systems with narrow concerns to provide immediate protection on one side of the scale and wider efforts to provide preventative service to meet a broader range of needs on the other (Hayes & Spratt, 2014).    In attempting to address this problem, MCFD may have implemented a range of preventative initiatives like FDR to uphold the sanctity of family preservation without sacrificing the need to act in the best interests of children enshrined in the CFCSA. Litwak and Hylton (1962) describe the phenomenon of inter-organizational relations to explain that when societies 71  view two conflicting values as equally desirable, like freedom and personal safety for example, they will separate the manifestations of such values into separate organizations like newspaper outlets and police forces respectively. In the case of child welfare in British Columbia there has not yet been a clear separation between child saving and family preservation which placed agencies like MCFD in a bind of how to preserve both values under public scrutiny. The tension between family preservation and child-saving, as mentioned previously, led to debate about whether the functions should be separated in child welfare.  In 2001, the Liberal government expected greater efficiency from the Ministry after implementing risk assessment but did not provide additional resources to support this goal. High performance expectations created pressure for MCFD to ‘do more with less’ and to find ways to make better use of existing resources. FDR and the initiatives that were implemented in parallel with it were designed to engage extended family and community members like neighbours and friends who have an existing relationship with the child. Litwak and Meyer (1966) refer to these types of resources as primary groups and explain how most successful bureaucratic organizations make efforts to establish close ties with them out of recognition of the benefits they have to offer.  The primary group is flexible and can quickly adapt to non-uniform events, a trait particularly useful when it comes to providing emergency care for children. CPS provides technical expertise like knowledge about the legislation and effects of physical discipline on children or developmental milestones and rates at which children develop at the price of slow communication and adjustment within its bureaucratic structure. Mediation, out-of-court conferences and family development response emerge within roughly the same time frame and were initiatives designed to increase cooperation between the family it’s immediate environment and the Ministry within the families’ immediate environment (eg. extended family and 72  individuals who have a positive relationship with the child were encouraged to become part of planning and providing care for children). The benefit of engaging the primary group is that some types of child protection concerns can be quickly and efficiently addressed (ie., a parent who falls asleep with a lit cigarette in their hand) with minimal CPS involvement but the optimal level of primary group engagement requires negotiation. If a bureaucratic organization like MCFD and its primary groups become too close they may be drawn into conflict about how child welfare is best achieved (eg. what constitutes a protection concern or how to address such a concern), yet if there is too much distance between them they may be unable to coordinate their efforts. Optimal balance theory, as suggested by Litwak and Meyer (1966) suggests that the optimal level of coordination between organizations and primary groups is one that maximizes the contributions while minimizing conflicts between the two. The Ministry’s implementation of FDR from this theory may be an attempt to rebuild its relationship with primary groups which is essential for its proper functioning. Child welfare agencies like MCFD recognized that primary groups needed to be harnessed as a resource and FDR represents one attempt at the system level to re-engage the primary group in addressing protection concerns. The nature of this relationship is closely related to the institutional environment in which the child welfare organization exists.  3.2.3 Resource Implications Demanding Solutions Organizational systems respond to changes in the opportunities and demands of their ecological contexts, including their political environments (pendulum shifts) and these changes often have resource implications that demand solutions. Child welfare, in particular, is one type of human service organization that has been historically vulnerable to such changes (Merkel-Holguin, Kaplan & Kwak, 2006; Parton, 1998). In times of crisis, 73  changes in management, political or ideological shifts, there can be intended and unintended effects that will impact service delivery. According to systems theorists, innovations emerge when complex organizations are seeking to find order by adapting to their environments. The time period in which FDR was implemented was a time of crisis due to high intake volumes and increasing numbers of children in care along with legislation which emphasized family preservation and image management in response to a highly publicized child death. These events could be conceptualized as external feedback and MCFD’s attempts to respond to such loops and in most cases senior management have taken the primary role in making changes.  3.3 Bifurcation into FDR and INV  The bifurcation of protection reports into FDR and INV was MCFD’s attempt to adapt to its institutional and organizational environment following several important events in British Columbia’s child welfare history. MCFD sought to re-establish positive relationships with communities and address the problems associated with becoming a threshold system. It was faced with pressure to preserve the family, yet had witnessed the consequences of what could happen if only the highest priority children and families were serviced. The following section will explain how implementation of a major change like FDR can affect a CPS agency and how case designation can be influenced by organizational and contextual factors as well as case/clinical characteristics. 3.3.1 Organizational and Contextual Factors The bifurcation of FDR and INV was a major systemic change within MCFD. Unfortunately, major re-organizational changes tend to be expensive, require enormous amounts of energy and often result in little functional change on the front lines of service 74  despite the disruption that they cause (Glor, 2007). Glor argues that newly structured components of an agency “typically settle back into behaviour patterns that are similar to those that existed previously and that were sometimes integral to the problems the re-organization was meant to solve” and concludes that for this reason most large organizational change projects fail to achieve their explicit goals. MCFD invested heavily into FDR implementation by incorporating FDR into the practice standards and transformation charter, recruiting offices and line staff for piloting and training them in the new approach. In order for there to be meaningful change, it is more important to first address any problematic patterns of functioning within the organizational context. Change is more likely to be achieved when individuals are intrinsically motivated, exist in a non-hierarchal culture and the challenge is minor, but this is seldom the case for CPS. Change is least likely to occur when motivation is extrinsic, the organizational culture is top-down and major challenges are imposed. Hierarchy is described as a barrier to implementation of system changes because “[m]anagers in many organizations in effect deliberately reduce system variety and reactivity, build barriers to heterogeneous interaction and retard the self-organization of change, in favor of control” (Glor, 2007, p. 12). This slowing of adaptation is done by placing a great deal of power and authority at the top of an organization, by restricting any unauthorized activities and motivations and by giving employees so much work that they have no time to interact with their environment.  If an organization does not have the adaptive capacity to affect instrumental change, it will introduce linearity and power-based mechanistic change without choice. Organizations acting linearly will not be able to adapt to face regular crises because they 75  lack the capacity. This occurs because there is no intrinsic motivation or bottom up skills or flexibility to encourage the emergence of adaptation. In linear systems there is an over-reliance on judgment and wisdom from the executive, yet no evidence exists in the organizational literature to suggest that better decisions are made at the executive level than at the operational level. Self-organized emergent change is also not likely in such systems. The organization that will be flexible and most likely to change is one where employees are intrinsically motivated. Adaptive change is more likely when the culture is one of bottom-up style of leadership and many ideas are promoted, encouraged and supported by management (Glor, 2007). Regular communication with outside organizations and reduced levels of challenge via de-centralization are also elements of promoting meaningful change.  Major changes in CPS are usually the result of crises which have drawn the attention of the media and the demand of the public for an immediate response. Change therefore tends to be reactive in nature and management typically answers by tightening authority, eligibility requirements and control. Social workers on the front lines of child protection in particular are influenced by major changes in the organizational and environmental context and are responsible for how such changes are manifested in service to clients. It is reasonable to assume that in such environments workers making decisions about case designation will be influenced by a myriad of factors outside of case characteristics.  3.3.2 Influences on Case Designation Intake screeners are responsible for initial case designation between INV and FDR pathways but complications can arise in the process. For reports that contain concerns of possible maltreatment but do not meet official criteria for inclusion for follow-up by a worker, for 76  example, “a space is created for ideological and pragmatic factors to influence decision making” (Buckley, 2000). These ideologies were discussed in the previous section and are likely to reflect the dominant morals and values of the institutional environment. One of the major findings of a study by English et al. (2000) was how the referrals to alternative response in their sample were inappropriately high. Other studies have documented how cases assigned to the FDR pathway resembled cases of neglect associated with poverty. Buckley (2000) questioned why seventy-six percent of referrals in her study based on neglect concerns were filtered out of the system very quickly without offering service despite existing knowledge about the detrimental effects it can have on children.  She explained that, among other factors including poverty and reluctance to be intrusive, workers seemed to be caught in a dilemma where reliance on a forensic approach to protection could offer little in situations not amenable to accessible solutions and “seemed to be caught up in a linear, one-dimensional approach to the problems which were presented to them” (Buckley, 2000). In a similar vein, other researchers have suggested that “’child protection’ does little to improve the lot of many children who come into its net, and in fact performs poorly in relation to ‘neglect’ cases” (Thorpe, 1994, p.196). Thorpe (1994) goes on to say that the available knowledge base and technology constructed around child abuse does not represent the reality for the majority of children who come into contact with the system and highlights the current focus of assessment on ‘children in need of protection’ rather than ‘children in need’.  3.3.3 Organizational Theory and Contextual Factors as Explanatory Variables The implementation of FDR from an organizational theory perspective could represent an internal systemic adaptation by MCFD in response to external demands to serve as a ‘catch-all’ to moderate the detrimental consequences of neglect for children but the reality is that neglect cases are the majority in child protection. Threshold systems that have clawed back services 77  inevitably create a class of children who have been systemically neglected by CPS itself. This neglect then causes a rift between primary groups in the community and child protective services which initiatives like differential response have attempted to address. Unfortunately, these types of approaches are not able to address the underlying problems with a system that has moved away from child welfare towards child protection. Screeners in this context might be motivated to code the majority of reports into differential response with a cursory knowledge of what is happening within the families’ own system because this class matches the profile for FDR assignment at the screening level based on standard practice protocols. This could also be why differential response could more accurately be defined as deferred response because a short-term (usually 60 to 90 day) FDR intervention cannot possibly address the deeper problems faced by the family that have to do with structural inequalities that cannot be abated with a limited response. Bednar (2003) observed that the perceived value of child welfare in society is influenced by many factors including political climate or perceived organizational efficiency in responding to reports and other external or internal pressures. The purpose of this evaluative research will be to use organizational theory to determine the objectives of FDR in BC and how these objectives might have been met.  3.4 Research Questions The purpose of this research will be to answer the following questions:  1. Is the goal of coding a certain proportion of intakes as Family Development Response consistent with the available research on abuse/maltreatment rates? 2. What is the proportional designation between family development response and a traditional investigative approach at decision point two? 78  3. What is the actual rate of re-designation of intakes coded from family development response to investigation and under what conditions does re-designation from Family Development Response to Investigation occur? Attempts will be made to isolate the different ways that intakes are re-coded. What are the objectives (explicit or implicit) of Family Development Response in British Columbia? Does it meet those objectives?                   79  4 Methods  4.1 Sampling 4.1.1 Intake Data  The sample in this study consisted of child protection intake reports from January, 2007 to March, 2012 collected from the Ministry of Children and Family Development’s (MCFD) computerized information management system (MIS). March 2012 was chosen as the cut-off date because a new Integrated Case Management (ICM) system was introduced in April 2012 and the two systems were not compatible for data collection or analysis (the new ICM was windows based while the former MIS was based on the MS-DOS operating system). To qualify to be included in the sample the report had to contain at least one identified s. 13 protection concern according to the CFCSA and assigned to either the family development response (FDR) or investigation (INV) pathway.  A population sample of protection reports was taken from the Lower Mainland of Greater Vancouver for overall descriptive analysis (N = 33,880) and from this sample four offices were chosen from the Langley/Delta Region of Surrey, British Columbia for more detailed analysis (n=1961). Two of the teams studied (one intake and one family service) were geographically co-located and were also a pilot site for FDR during its implementation while the other two teams (also geographically co-located with one intake and one family service) were not pilot sites. Thus four teams in total made up the sub-sample of the study to compare differences in the use of FDR between pilot sites and non-pilot sites. Changes in designation between FDR and investigation were also examined. For the purposes of this study, new calls received within one calendar year of the last intake being closed on a families’ Family Service file (each family has their own unique FS number) served as a 80  proxy for re-coding after initial designation. The proxy was used in an attempt to answer the third research question (what is the actual rate of re-designation of intakes coded from family development response to investigation and under what conditions does re-designation from Family Development Response to Investigation occur?). A region where FDR service and assessment was completed by the protection social worker ‘in-house’ was compared to a region where parts of the service or assessment were ‘contracted out’ to see if there was any association between contracting out service and FDR pathway assignment. Finally, three qualitative interviews were conducted to complement the quantitative data and give a fuller understanding about the context of FDR implementation and any challenges that arose during implementation.  When a protection report is made to the Ministry of Children and Family Development, the call is received by a screener who makes a decision about whether there is a presenting protection concern according to s. 13 of the CFCSA; this is referred to by the Ministry as decision point one. If there appears to be at least one s. 13 concern, the screener then consults with the Team Leader about pathway assignment to either FDR or INV at decision point two, provided the Ministry has decided against taking no further action, offering a youth response or referring back to the community. Once the decision has been made about pathway designation, a worker is assigned and conducts either an FDR assessment or an investigation. After the assessment or investigation is complete the assigned worker at the intake level must make a decision to close the file if concerns are not substantiated (INV) or if the family is not interested in services or supports (FDR). If the worker in consultation with the Team Leader makes the decision that the family requires ongoing involvement then the decision is made to transfer the case to a Family Service team. At any point re-assignment to a different pathway is allowed depending on new information obtained during the investigation or assessment. 81  4.1.2 Staff Interviews Interviews were conducted to obtain qualitative information about how FDR was implemented in the province. The purpose of the interviews was to obtain information about challenges with making coding decisions, training and give context to complement the quantitative data, especially when it came to policy changes, practice guidelines, contracting out services and outcomes (see Appendix A for interview questions). All three participants signed a consent form to participate in the study and were informed of their right to cease participation at any time. All interviews were audio recorded on a device to MP3 format which were then stored on a secure flash drive and transcribed from MP3 format. Participants were chosen based on their specialized professional knowledge and asked a set of pre-approved questions having to do with a range of topics about implementation including coding decisions, training curricula, policy and/or practice changes over time, implementation challenges and evolution of FDR and outcome monitoring. This was done primarily to track whether there were changes in FDR since it was initially implemented and to obtain a better understanding of how MCFD monitored its own performance since implementation.   Three semi-structured interviews were conducted with professionals at three levels: one child protection line worker, one clinical supervising team leader and one manager. All three interviewees were working in the Ministry when FDR was implemented and the line worker and team leader were both part of the pilot project. The manager interviewed was chosen based on her key role in FDRs implementation in the province. Participation was voluntary and participants were chosen based on their professional expertise and knowledge of FDR during MCFDs implementation.   82  4.2 Data Analysis The first research question was answered by observing how trends in the most recent iteration of the Canadian Incidence Study align with MCFD’s recent objective to code in favor of FDR. To examine the proportion of INV and FDR designations per year for the larger sample and for the smaller four-team sample, frequency tables of designation and caller type, and cross-tabulations of designation by caller type and by year were constructed. Chi-square tests of independence were carried out to test whether designation was independent of caller type, and thus to assess any changing trends in pathway assignment in order to answer the second research question. The researchers hypothesized that proportional designation would remain constant over the four year time period if case characteristics were the deciding factor in pathway assignment; any significant difference between years would suggest that outside factors influenced case designation. Data analysis for this study also focused on how coding decisions were made in the four offices (ie. whether or not the initial designation was informed by an accurate and timely assessment of vulnerability, strengths and needs) as well as the specific mechanisms used and the circumstances under which re-coding took place.  Crosstabs and chi-squares tests of independence carried out to test whether there were differences in the rates of the two designations for the total sample and for each year along the characteristics mentioned above. Re-coding mechanisms were examined to see how offices used MIS to record changes in pathway designation after initial assignment. Built into the MIS system was a coding designation specifically for cases where coding had changed from FDR to INV (IFD) so attention was paid in the overall sample to how it was used and whether there was any mechanism to code in the reverse (INV after FDR). Intakes were examined for re-designation from FDR to INV (or vice versa) after initial designation to explore whether there was any 83  crossover at time two. Chi-square tests of independence were carried out to test whether the rate of change in designation was the same across years. Finally, the way that offices are coded (each office has its own unique three letter code with the first letter indicating geographic region) allowed comparison of proportional designation between a region that contracted out the FDR function (Langley/Delta) and a region that performed the FDR function in its entirety ‘in-house’ (Vancouver) to see if contracting FDR made a difference in its use. Once again, chi-square tests of independence examined whether type of designation was the same for the two locations, for the total sample, and for each year.   84  5 Results  5.1 Preliminary Discussion 5.2 Comparing Canadian Incidence Study Data and Research Question One The most recent iteration of the Canadian Incidence Study of Reported Child Abuse and Neglect (CIS-2008) and its American counterpart, the fourth National Incident Study of Child Abuse and Neglect (NIS-2010) were used to evaluate whether the goal of coding a certain proportion of intakes as FDR is consistent with the available national data on maltreatment rates. When the two studies were examined in chapter one for trends the following observations were made. First, there appeared to be more protection reports made and substantiated about neglect. Second, domestic violence was flagged as an important factor in many of the child protection reports received in both Canada and the United States. Third, concerns were identified with a range of social issues contributing to neglect from mental illness, isolation and substance use, to poverty-related concerns indicated by receipt of social assistance. There continues to be a strong correlation between socioeconomic status and child protection reports. Fourth, child protection involvement was racialized as Aboriginal children were disproportionately represented in Canadian foster placements while Black children in the United States were reported as more likely to have experienced maltreatment than their White or Hispanic counterparts. Fifth, there appeared to be a high proportion of protection reports that did not receive a response.  One practical implication drawn from the trends observed in the CIS-2008 and the NIS 2010 was that many of the reports where the risk level was not severe enough did not receive follow up. Chapter two described how ongoing cuts to social spending in general beginning in the early 1980s resulted in an unfortunate protection approach to child welfare. The consequences of residualism are borne out in North American threshold systems. The NIS-2010 85  and CIS-2008 data showed that many children in need of help did not receive it because the reported concerns did not meet the threshold for ongoing CPS involvement. The major problem presented by these findings was how to help children in situations where risk factors for neglect and/or abuse were present but the threshold had not been met to merit protective intervention. Another way to view the problem is how to help children not screened in for service when residualism narrowed the scope of service provision.  By definition, FDR is a secondary preventive paradigm intended to address low-risk protection concerns not deemed to require an urgent response. The detrimental consequences of neglect on children are not meant to be minimized when classifying cases as low-risk. On the contrary, the rationale behind differential response is to provide intervention to families in need before they hit the high-risk threshold. Given the recent trends in child protection reports from the CIS-2008 and the NIS-2010 there may be grounds to suggest that the goal of coding a certain proportion of intakes as FDR is consistent with the available research on abuse and maltreatment rates, but this statement must be made with caution.  5.3 Overview of Larger Sample and Research Question Two  An overall sample of child protection reports (intakes screened-in because they contained at least one identified section 13 concern under the CFCSA) was drawn from the Lower Mainland (N = 33,880) for the period January 2007 to March 2012. The data were provided by MCFD in the form of a password-protected Microsoft Excel spreadsheet. Of this sample, 8,678 intakes were coded as family development response (FDR), 25,195 intakes were coded investigation (INV) and seven intakes were coded as ‘investigation following FDR’ (IFD) to answer the second research question (what is the proportional designation between family development response and investigation at decision point two?) (See table 5.1). Investigation was 86  clearly the preferred response based on the overall proportions (74.4%) despite the fact that by 2008, FDR had been a part of MCFDs protection repertoire for approximately four years. Table 5.1  Designations (Frequency and Percent), from January 2007 to March 2012 Designation Frequency  Percent Family Development Response (FDR) 8,678 25.6 Investigation (INV) 25,195 74.4 Investigation Following FDR 7 0.0 Total 33,880 100.0  Table 5.2 below shows the different types of callers by frequency and percent. Police and schools represented the most common reporters to MCFD, followed closely by community professionals. The least frequent sources of protection reports were from SPMH workers and Financial Assistance Workers followed by Residential Caregivers.           87  Table 5.2  Type of Caller (Frequency and Percent), from January 2007 to March 2012 (Ordered from most frequent to least frequent) Caller Type Frequency  Percent Police 8,009 23.6 School 6,795 20.1 Community Professional 4,791 14.1 Parent 2,539 7.5 Health Professionals 2,160 6.4 Relative 2,125 6.3 Friend/Neighbour 2,074 6.1 Concerned Citizen 1,660 4.9 CF & CS Worker 1,546 4.6 Subject Child 546 1.6 Anonymous 398 1.2 Ex-Spouse 326 1.0 Probation Officer 251 0.7 Preschool/Daycare 242 0.7 In-home Support Provider 197 0.6 Residential Caregiver 98 0.3 Financial Assistance Worker 97 0.3 SPMH Worker 25 0.1 Not Coded 1 0.0 Total 33,880 100.0  Table 5.3 shows the proportion of FDR or INV designated intakes, cross tabulated by caller type. The seven IFD designated intakes are not included here, so the total is 33,873 rather than 33,880. Investigation was the most common response to all caller types but FDR was used more frequently for calls from the police, schools or the subject child. Investigations were used most commonly for reports from a residential caregiver, an anonymous caller or a CF and CS worker. FDR is supposed to be used when the level of risk is screened as lower priority. Concerns about children’s’ exposure to domestic violence via police reports, low-level cases of neglect from school teachers and complaints from subject children may all fit the lower-risk 88  criteria. There is also the possibility that police, schools and subject children are all able to provide more comprehensive social and historical information on a child or family and an FDR designation can be made with greater assurance than if the report came from other sources. If there was a province-wide directive in 2010 to use FDR except in the most urgent cases and police, schools and subject children might have been able to provide more comprehensive information on the family context, there is a chance that screeners may have reviewed reports for information suggesting that risk was low to code in favor of FDR.                               89  Table 5.3  Caller and FDR or INV designation, from January 2007 to March 2012 (Sorted by total records per caller type) Caller FDR INV Total Police 2,435 (30.4%) 5,574 (69.6%) 8,009 School 2,066 (30.4%) 4,727 (69.6%) 6,793 Community Professional 1,151 (24.0%) 3,639 (76.0%) 4,790 Parent 668 (26.3%) 1,871 (73.7%) 2,539 Health Professionals 518 (24.0%) 1,640 (76.0%) 2,158 Relative 388 (18.3%) 1,737 (81.7%) 2,125 Friend/Neighbour 423 (20.4%) 1,651 (79.6%) 2,074 Concerned Citizen 306 (18.4%) 1,354 (81.6%) 1,660 CF & CS Worker 233 (15.1%) 1,312 (84.9%) 1,545 Subject Child 164 (30.0%) 382 (70.0%) 546 Anonymous 56 (14.1%) 342 (85.9%) 398 Ex-Spouse 89 (27.3%) 237 (72.7%) 326 Probation Officer 49 (19.5%) 202 (80.5%) 251 Preschool/Daycare 63 (26.0%) 179 (74.0%) 242 In-Home Support Service Provider 36 (18.3%) 161 (81.7%) 197 Residential Caregiver 9 (9.2%) 89 (90.8%) 98 Financial Assistance Worker 18 (18.6%) 79 (81.4%) 97 SPMH Worker 6 (25.0%) 18 (75.0%) 24 Not Coded 0 (0.0%) 1 (100.0%) 1 Total 8,678 (25.6%) 25,195 (74.4%) 33,873 The data in Table 5.3 are also presented as a segmented bar chart (Figure 5.1) to provide a graphical comparison.       90  Figure 5.1  Designation by Caller Type  An examination of the bar chart and standardized residuals from the cross tabulation in Table 5.3 (standardized residuals are not shown here) suggests large differences in the rate of FDR versus INV calls depending on caller types. To assess this formally, caller types were classified into “mandatory” reporters vs. “voluntary” reporters. The mandatory group comprised: Police, School, Community Professional, Health Professionals, CF & CS Worker, Probation Officer, Preschool/Daycare, In-Home Support Service Provider, Residential Caregiver, Financial Assistance Worker, and SPMH Worker. The voluntary group comprised: Parent, Relative, Friend/Neighbour, Concerned Citizen, Subject Child, Anonymous, and Ex-Spouse. A chi-square test of independence showed a statistically significant association between type of designation and type of reporter. Mandatory reporters had significantly higher rate of 0% 20% 40% 60% 80% 100%Not CodedSPMH WorkerResidential CaregiverFinancial Assistance WorkerIn-Home Support Service ProviderProbation OfficerAnonymousPreschool/DaycareEx-SpouseSubject ChildCF & CS WorkerConcerned CitizenRelativeFriend/NeighbourHealth ProfessionalsParentCommunity ProfessionalSchoolPoliceFDRINV91  FDR designations (27.7%) than voluntary reporters (21.7%) (X2 (1) = 111.49,  p < .001). See Table 5.3b.  Table 5.3b  Mandatory vs. Voluntary Caller and FDR or INV designation, from January 2007 to March 2012  Designation FDR  INV Total Mandatory 6,584 (27.2%) 17,620 (72.8%) 24,204 Voluntary 2,094 (21.7%) 7,575 (78.3%) 9,669 Total 8,678 25,195 33,873 (*) (*) Note. Seven IFD cases are not included here so the total is reduced from 33,880. A chi-square test of independence showed statistically significant differences in the use of FDR over time (see Table 5.4) (X2(4) = 3950, p < .001). An examination of standardized residuals (not included in the table) confirms that the rate of FDR in each year was different from the overall rate of FDR in the sample. There was a modest but gradual increase in the proportion of intakes coded as FDR in the Lower Mainland from 2008 to 2010 (the average percentage of all intakes coded as FDR was 14.5%) but by 2011 the percentage of FDR intakes reached almost forty-five per-cent. Then in 2012, the number of intakes in the Lower Mainland coded as FDR surpassed investigations for the first time since its implementation and became the majority response in the Lower Mainland to child protection reports. This is an important finding because it supports the hypothesis from chapter three that case characteristics were not the only determining factors for pathway designation. If screening teams were making coding decisions based primarily on case characteristics it would be reasonable to expect that the proportions of INV and FDR would not have changed so dramatically in such a short period of time.   92  Table 5.4  Intake Designation Type, by Year  Year FDR  INV Total (*) 2008 357 (13.0%) 2,392 (87.0%) 2,749 2009 1,310 (13.5%) 8,382 (86.5%) 9,692 2010 1,624 (17.0%) 7,952 (83.0%) 9,576 2011 4,254 (44.8%) 5,238 (55.2%) 9,492 2012 1,124 (51.7%) 1,051 (48.3%) 2,175 (*) Note. Seven IFD cases (3 in 2008, 1 in 2009, and 3 in 2011) are not included here.  Table 5.5 shows the nature of the call based on five available categories built into MCFD’s MIS system: request for family support service, request for SPMH child services, protection report, request for special needs daycare and request for youth services. Protection reports were the overwhelming majority of calls screened in for service by MCFD that would receive an FDR or INV designation (95.7%). The remaining 4.3% of calls were those where the initial purpose of the calls were to request service (either Family Support Service or Youth Services) but were screened as containing at least one s. 13 protection concern and later given an FDR or INV designation at decision point two.  Table 5.5  Nature of Request (Frequency and Percent), from January 2007 to March 2012 Designation Frequency  Percent Protection Report 32,413 95.7 Request for Family Support Service 1,104 3.3 Request for Youth Services 359 1.1 Request for SPMH Child Services 2 0.0 Request for Special Needs Daycare 2 0.0 Total 33,880 100.0  5.4 Re-coding After Initial Designation and Research Question Three Investigation Following FDR (or IFD) represented the only code built into the MIS system to track re-designation following initial pathway assignment but was used only seven 93  times in the overall sample by one office (there was no coding designation to represent FDR after an initial coding designation of INV). For the purposes of this study, new calls received within one calendar year of the last intake being closed on a families’ Family Service file (each family has their own unique FS number) served as a proxy for re-coding after initial designation. The proxy was used in an attempt to answer the third research question (what is the actual rate of re-designation of intakes coded from family development response to investigation and under what conditions does re-designation from Family Development Response to Investigation occur?) The seven IFD designated intakes were not counted as part of the proxy.  Figure 5.2. Chart to track sample sizes leading to re-designations  Total intakes = 33,880 Follow-up:  11,556 (34.1%) No Follow-up: 22,324 (65.9%) Follow-up Within One Year: 8,891 (76.9% of FU) Follow-up > 1 yr: 2,665 (23.1% of FU)  Change in Designation: 1,565 (17.6% of FUs <1 yr) No Change: 7,326 (82.4% of FUs < 1 yr)   INV to FDR: 859 (55.0% of changes) FDR to INV: 706 (45.0% of changes)    The chart shows that 34.1% (11,556 of 33,880) of intakes in the sample were repeaters on the same FS file. Of the repeaters, 76.9% (8,891 of 11,556) were received within one calendar year of the last intake being closed. Of the repeaters within one calendar year, (1,565 of 8,891) the designation of the new call changed from the previous designation (FDR to INV or vice versa). Thus, using a proxy to answer the third research question, re-designation between FDR and INV in occurred 4.6% of all intakes, a small but not unimportant rate in view of the sample size. The large majority (82.4%) of new calls received within one calendar year kept the same 94  designation as the last closed intake. Of the 1,565 intakes where the designation changed, 55.0% (859) were INV to FDR and 45.0% (706) were FDR to INV. Follow-up records after closing dates were examined to explore changes over time with respect to the number of multiple records on the same FS file, the number of repeaters within one year, and the number with a change in status (see Table 5.6). There is a large and steady increase in the number of repeating intakes on the same FS files, from 8.7% in 2008, to almost 47% in 2012. Of the follow-ups there is a large decline in the proportion within one year. In 2008, 93.8% of repeaters were within one year, but by 2012 the percentage had dropped to 60.4%. And, of the follow-ups within one year, the proportion that had a change in status increased from 8.9% in 2008 to 27.1% in 2012. For each variable, a chi-square test of independence was carried out, to test whether the proportions were independent of year. There were large effect sizes observed for each test: rate of follow-up (X2(5) = 1951.37, p <.001), rate of follow-up within one year (X2(5) = 781.27, p< .001), and follow-up change in status (X2(5) = 557.73, p < .001). Note that data from 2012 could only be collected up to March 31. An examination of standardized residuals shows differences in each pair of years except between 2011 and 2012. The information contained in these tables sought to answer questions about whether multiple records increased with time and what percentage of FS files had multiple records. It is clear that the number of multiple records climbs over the years substantially and that a significant number of intakes received by MCFD are for multiple intakes on the same FS files.     95  Table 5.6 Intakes with Follow-up calls, Follow-up calls within One Year, Change in Status, by Year Year Total Follow up Within One Year Change   Percent of “Total” Percent of “Follow-up” Percent of “Within One Year” 2007 or Before 189 36 (19.0%) 34 (94.4%) 5 (14.7%) 2008 2,754 240 (8.7%) 225 (93.8%) 20 (8.9%) 2009 9,703 2,326 (24.0%) 2,135 (91.8%) 211 (9.9%) 2010 9,564 3,724 (38.9%) 3,016 (81.0%) 384 (12.7%) 2011 9,496 4,216 (44.4%) 2,869 (68.1%) 779 (27.2%) 2012 2,174 1,014 (46.6%) 612 (60.4%) 166 (27.1%) Total 33,880 11,556 (34.1%) 8,891 (76.9%) 1,565 (17.6%) A proxy (intakes that were closed as FDR but re-opened within one calendar year with an INV designation) was used to examine if family development response (FDR) intakes that were re-designated to investigation (INV) had any child placements or court involvement. Seven hundred and six intakes in the total sample (N = 33,880) met our proxy criteria. From this sample of 706 we drew a smaller random sample of twenty intakes to peruse for court involvement and found that two of the intakes concluded with Voluntary Care Agreements (VCA) and one of the files concluded with a removal on a subsequent intake. Therefore, out of the random subsample of twenty intakes drawn from the 706 intakes that were re-designated within one calendar year from FDR to INV, three (15%) resulted in child placements or court involvement. 5.5 Overview of the Four Team Sub-sample The four team sub-sample in this study consisted of two teams that were a co-located pilot site for FDR (one intake team and one family service team) and two teams in the same 96  region that were also co-located but not an FDR pilot site (also one intake and one family service team). These four teams were chosen to compare the use of FDR between offices that were pilot sites and those that were not pilot sites. The four-team sub-sample (n=1,971) was drawn from the larger overall sample (N=33,880) and contained 1,250 (63.4% of the sub-sample) intakes from the pilot sites (GEB/GEC) and 721(36.6% of the sub-sample) intakes from the non-pilot sites (GCB/GFB). As with the larger sample, investigation is the dominant response to child protection reports but to a lesser degree than in the larger sample. The sub-sample mirrored the pattern of the larger sample in the progressive increased proportion of intakes being coded as FDR and the significant increase in 2011. It also reflects FDR having surpassed investigation as the majority response to child protection reports in 2012. A chi-square test of independence showed a statistically significant difference in the use of FDR between pilot and non-pilot sites with pilot sites more likely to use FDR than non-pilot sites (X2(1) = 133.68, p < .001). Pilot sites used FDR almost as much as investigation (47.8% versus 52.2%) while non-pilot sites used FDR only 21.5% of the time (See Table 5.7).    Table 5.7  Pilot Versus Non-Pilot Sites and FDR use from January 2007 to March 2012  Designation Pilot (GEB/GEC)  Non-Pilot (GCB/GFB) Total FDR 594 (47.8%) 155 (21.5%) 749 INV 649 (52.2%) 566 (78.5%) 1,215 Total 1,243 721 1,964 (*) (*) Note. Seven IFD cases are not included here so the total is reduced from 1,971. This offers tentative support to the finding that pilot site regions were more likely to use FDR and also supports the hypothesis that case characteristics were not the only determining factors of coding decisions. 97  Table 5.8 shows the use of FDR in pilot versus non-pilot sites by year. There were no FDR intakes to compare prior to 2007 in the sample between pilot and non-pilot sites. In 2008 the use of FDR at pilot sites is 28.8%, rose to 43.5% in 2009, and remained above 50% until 2012. In contrast, the use of FDR at the non-pilot site started in 2008 at 7.9%, dropped to 5.1% in 2009, rose to 9.7% in 2010 and increased to 45.8% in 2011 and again to 51.6% in 2012. The pilot sites show an approximate increase of 10% by year in the use of FDR but the dramatic increase between 2010 and 2011 in the non-pilot FDR sites raises questions of a significant incident in or around 2010 which led to the increased use of FDR at non-pilot sites from approximately 10% to 50%. Chi-square tests of independence were done for each year, to compare the proportion of FDR intakes in the Pilot and Non-Pilot sites. Differences were statistically significant for 2008, 2009 and 2010. Table 5.8  Pilot Versus Non-Pilot Sites and FDR use by Year  Year Pilot (GEB/GEC) Non-Pilot (GCB/GFB)   FDR INV Total FDR INV Total X2; p-value (*) 2008 44 (28.8%) 109 (71.2%) 153 5 (7.9%) 58 (92.1%) 63 11.03 p = .001 2009 135 (43.5%) 175 (56.5%) 310 10 (5.1%) 185 (94.9%) 195 86.33; p <.001 2010 183 (54.8%) 151 (45.2%) 334 20 (9.7%) 186 (90.3%) 206 110.38; p <.001 2011 193 (51.6%) 181 (48.4%) 374 87 (45.8%) 103 (54.2%) 190 1.70; p = .19 2012 39 (59.1%) 27 (40.9%) 66 33 (51.6%) 31 (48.4%) 64 0.75; p = .39 (*) Chi-square tests compare proportions of FDR in Pilot and Non-pilot sites for each year. There are several interesting changes with time when comparing pilot sites to non-pilot sites. FDR use in the pilot sites is much more frequent than in the non-pilot sites but the difference appears to lessen between 2010 and 2011 and even more in 2012. In the table we can 98  observe a stark difference in the use of FDR in the non-pilot offices between 2010 and 2011 where its use jumps from 9.7% to 45.8%. This further supports the hypothesis that case characteristics were not the only determining factor in pathway designation and that organizational and other factors may have played a role in screening decisions on the front lines of practice.   5.6 Comparison between Regions  All MCFD child protection offices have a three-letter code which is representative of their geographical catchment area. All offices in the Vancouver/Richmond area have a code that begins with ‘R’ while all offices in the Langley/Delta area have a code that begins with ‘G’. This coding system is convenient for this study since it permits comparisons between regions. In Vancouver/Richmond the FDR function was performed by child protection workers in stand-alone teams (teams that only performed the FDR function where workers performed the assessment and provided services) while Langley/Delta was more flexible in the use of FDR and contracted out aspects of it to agencies (the assessment and service provision phases were either performed by the protection worker or a contracted agency or some combination of the two).  Just over half (53.8% or 18,211 intakes) of all child protection reports in the total sample came from the Langley/Delta and Vancouver/Richmond regions combined. Of these, 11,028 (60.6% of intakes in the two regions) came from Vancouver/Richmond and 7,183 (39.4% of intakes in the two regions) came from Langley/Delta.  Table 5.9 shows the proportions of FDR and INV in the two offices. FDR was used more frequently in Vancouver/Richmond (33.9%) than in Langley/Delta (29.9%). A chi-square test of independence shows that this difference is statistically significant (X2(1) = 31.95, p < .001). Both Vancouver/Richmond and Langley/Delta regions contained FDR pilot sites but 99  Vancouver/Richmond was the only region to have stand-alone FDR teams where both assessment and service were provided by fully delegated child protection social workers. Table 5.9  Proportions of FDR and INV between Two Regions, January 2007 to March 2012  Designation Vancouver/Richmond  Langley/Delta Total FDR 3,737 (33.9%) 2,144 (29.9%) 5,881 INV 7,291 (66.1%) 5,032 (70.1%) 12,323 Column Totals 11,028 (60.6%) 7,176 (39.4%) 18,204  These findings offer tentative support to the notion that FDR was more commonly used in regions where there were stand-alone child protection teams with protection workers performing FDR from assessment to service provision. It also supports the hypothesis that case characteristics alone were not the sole factors in pathway designation. Table 5.9 shows the same operation broken down by year. Table 5.10  Contracted and In-house and FDR use by Year Year Vancouver/Richmond Langley/Delta   FDR INV Total FDR INV Total X2; p-value (*)  2007 & prior 5 (13.2%) 33 (86.8%)38 38 1 ( 2.1%) 47 (97.9%) 48 4.01; p = .045 2008 141 (17.0%) 688 (83.0%) 829 126 (18.5%) 554 (81.5%) 680 0.59; p = .44 2009 540 (17.9%) 2,482 (82.1%) 3,022 412 (19.8%) 1,669 (80.2%) 2,081 3.02; p = .082 2010 662 (20.1%) 2,634 (79.9%) 3,296 485 (24.9%) 1,462 (75.1%) 1,947 16.67; p < .001 2011 1,909 (60.9%) 1,225 (39.1%) 3,134 868 (44.5%) 1,081 (55.5%) 1,949 130.04; p < .001 2012 480 (67.7%) 229 (32.3%) 709 252 (53.5) 219 (46.5%) 471 24.22; p < .001 (*) Chi-square tests to compare proportions of FDR in Vancouver/Richmond and Langley/Delta sites for each year  100   Langley used FDR more frequently in 2010 but Vancouver overtook Langley in 2011. Chi-square tests of independence were done for each year, to compare the proportion of FDR intakes in the two regions. Differences were statistically significant for 2010, 2011 and 2012. What significant event or process accounted for the increase in intakes coded as FDR between 2010 and 2011 and how did it influence screeners making decisions on the front lines of practice? Due to time limitations of this study it was not possible to analyze differences by case characteristics and will instead look for differences in organizational circumstances to explain differences. 5.7 Staff Interviews and Research Question Four  Three interviews were conducted with staff who were at different positions within the Ministry during FDR implementation: a line worker at one of the FDR pilot sites (given the name Dorothy), a team leader who supervised workers doing FDR (give the name Colin) and a manager responsible for FDR rollout across the province (given the name Andrea). Questions were asked about training and early implementation, decision-making challenges, policy changes, contracting out FDR and outcome monitoring. These interviews were conducted for clarification purposes and do not necessarily reflect the experiences of all MCFD personnel (see Appendix A for interview questions).  5.7.1 Training and Early Implementation Andrea, one of the managers responsible for FDR implementation at the Ministry of Children and Family Development, explained how training was initially ad hoc but individualized by region, followed by provincial core curriculum training and changes to the risk assessment tool to reflect policy and legislation changes. As part of the implementation process, Colin recalled that MCFD used FDR consultants to work closely with pilot site teams and 101  provided extra provincial and regional training for team leaders providing worker supervision. Colin was a proponent of FDR from the beginning but recalled his struggles during implementation with workers who did not buy in. Despite the strategic use of strength-based, scaling and miracle questions during supervision in the hopes that these workers would eventually adopt them in practice, there were line workers who ended up leaving FDR teams. Andrea explained how there were “struggles over the years in between solution-focused and strength-based, and child protection and trying to help people embrace both is a challenge”. She went on to describe the need to reconcile the tension between professional clinical social work skills and an investigative substantiation approach, regardless of name. Dorothy offered her views on where FDR fits in child protection. Dorothy described herself as having already bought into the concept of FDR as a different way to engage with families and believed that a culture developed at the pilot site of “why wouldn’t we” in reference to the use of FDR over investigation. One of the areas she identified as a necessary but absent part of FDR training was addressing workers’ comfort with risk stating: “…the piece where the training lacked is really working with people’s values with being comfortable with risk, comfortable with risk in situations when it’s much easier to remove a child, put them in a foster home, rather than leave them in the family home and hope that the family is following the plan to keep their children safe”. Dorothy stressed the importance of clinical support and supervision from team leaders in these situations. “Really re-iterating that you are not doing this alone, you’re part of a care team. You’re not carrying this on your shoulders by yourself. Families make decisions, you’re opening doors, you don’t have a crystal ball”. She describes her screening experience at the pilot site:  …because it was an early implementation, a pilot project, we would essentially cherry pick the families – we would cherry pick the families 102  that we thought would be likely to have a positive response, that could possibly have a good outcome within 90 days – they would have little to no previous involvement, they were typically coded for neglect, or domestic violence, there was some physical but not serious physical violence, not sexual abuse.   Dorothy’s explanation on the types of cases screened towards FDR at early implementation is important because it coincides with the overall findings of this study. FDR was the more frequently assigned pathway if the report came from police, the school or the subject child (see table 5.3). It is reasonable to expect reports about domestic violence exposure and low level neglect to come from police and schools respectively. It is also likely that these reporters have access to historical information. The screener could then obtain a better understanding of whether the family was likely to respond positively to an FDR approach. Dorothy was an advocate of FDR but admitted that practice guidelines were not clear during the early phases of implementation, leaving teams to interpret the available policy to the best of their abilities.  5.7.2 Decision-Making Challenges  Colin expressed struggles on his team when making coding decisions, saying: “I think people had difficulty between provincial expectations and regional guidelines, trying to figure out the difference between the two”. Inconsistency between regional and provincial standards produced confusion among teams trying to implement FDR to the point where some workers left to go to other teams. Colin explained how intake workers who were used to investigation were confused about when to code reports as FDR, especially “…afterwards, when we became a pilot, there was a presumption for FDR. So we began to look at the screening process much more closely….” He said the team began to interpret the existing policy to find ways to differentiate between FDR and INV and whenever possible, code as FDR. Files where court involvement was 103  likely were screened towards investigation along with files where the parent or caregiver experienced serious mental illness. On the other hand, files rated as three or four according to the established risk assessment model were screened as FDR. A practical implication identified by Dorothy and Colin was confusion about when to see children and uncertainty about whether risk was being missed in the FDR pathway.  When asked about the circumstances in which coding might change from FDR to INV Colin said “…if we start the service, they don’t return calls, it’s a section 13 that is significant, we’d switch to investigate”. Initial engagement, cooperation with service providers and a willingness to receive services started a family on the FDR pathway but once the family ceased cooperation, the social worker began an investigation. Dorothy identified that FDR was a good fit for families and workers in terms of their values and believed that FDR should always be the starting point on what she called the “least intrusive ladder”. When asked about what conditions would tip a screening decision towards an investigation from FDR, Dorothy said: Certainly what we are looking at is the severity of the abuse and previous involvements and what I mean by previous involvement is their willingness to work with us, but just because someone didn’t work well last time that’s still not a tipping factor to an investigation, we’d still try again, try it a little bit differently. So it’s always FDR, very, very few files that are coded investigation right off the top, unless it meets that criteria or we had to do an immediate safety planning for the child or we weren’t able to meet with the parent, or advise the parent, so therefore it has to be an investigation. Those would be the only circumstances – that’s just how we do it.   Both Colin and Dorothy believed that FDR allowed for more professional discretion in interacting with families and that FDR on the whole was more respectful. 5.7.3 Policy Changes  According to Andrea there were a number of policy changes since the first mention of FDR in MCFD policy in 2003. Accompanying the original 2003 policy  104  … there were quite a number of different practice guidelines developed, because at that time in the Ministry the regions were able to develop guidelines for implementation that best suited their community, where there were geographic struggles. So there were a variety of ways that people approached implementation across the province.   When asked if any policy changes occurred specifically within the time frames of this study (2008 to 2012) Andrea said there were no policy changes and that what was written in 2003 was still being used in 2008 with April, 2012 as the most recent policy released relating to FDR. When asked specifically about circumstances which might have required changes in policy Andrea said, There has been a lot provincially. Every region took a unique approach to FDR and as long as they stayed within the parameters of the policy… Underpinning all of these approaches was more family engagement skills, really very basic social work skills, starting with the families.  According to Colin, the BC Government from Victoria implemented a presumption in 2010 to code protection calls in favour of FDR based on perceived successes at provincial FDR pilot sites. The presumption originated from Victoria to the local CSMs to the team leaders and was ultimately implemented by line social workers. Colin said that both he and the CSM at the time in Langley/Delta were enthusiastic about using FDR and one worker was brought onto the team who was “committed as well to FDR and really enjoyed it and was a champion of it”. The presumption meant that screeners were encouraged to examine information within protection reports to justify coding as FDR when possible and according to Colin he was directed to “use it unless I tell you not to”. Colin believed that the presumption directive contributed to the rapid increase in the use of FDR since 2010 and that it was supposed to be a practice shift “…from the Deputy, around engagement and using strength-based practices and such”.  Andrea confirmed that there was a move towards  105  really wanting to send the message that strength-based practice is really important in child protection cases. They’re required to be managed by delegated child protection workers. They’re not soft cases, and if we can really support the families at the intake level, in a way that will prevent them, and inoculate them, from getting secondary calls on them or being transferred to FS cases, not just moving to investigate, rather than just sitting there.   Colin also gave an example of how policy guidelines changed with time. When FDR was first implemented it was not clear if all children needed to be interviewed at once or only the subject child needed to be interviewed. In Colin’s region there was controversy because only the subject child of the report was being interviewed. He explained that this was because “…since we were using the old 3s and 4s, in the old risk assessment language, especially the 4s, the likelihoods, did you need to interview all the kids, then eventually policy changed and it became more clear that you did interview all the kids…”.  Andrea drew on her experience reviewing files across the province to explain how challenging it was to achieve consistency in worker supervision as well as eligibility of families for FDR designation. She went on to explain how the skills and worldview of the supervisor as well as the community in which the family lived or the team was practicing could all influence FDR eligibility. Even historical context of the community could influence comfort levels with FDR. She offered an example of how communities that have experienced more critical incidents and internal or external reviews could have an impact on worker’s comfort levels and how they assess and respond to reports. She stated that “When FDR started, the eligibility was more in a framework than it was consensus based. It was assessment-based and really relied on the judgment of the professionals involved by asking a series of practice questions and those kinds of things”. She believed that FDR was an iterative process that gradually expanded, not just in 2010. 106  5.7.4 Contracting out Family Development Response  Andrea was not aware of FDR being contracted out where she was involved but did say that Maple Ridge used co-located contractors who worked in collaboration with the delegated worker. She added that there was no consistent approach to FDR across the province so it is possible that while there was a regional decision about contracting there was no provincial decision on the matter. This statement was supported by Colin who explained how the decision to contract FDR was made as a regional decision within the Fraser region. The delegated social worker was responsible for the assessment, then the actual services were provided by an outside agency. Colin said: How I wanted it for our office was that the SW would be there for the initial intake meeting, the goals, and the client would be clear on the two roles between the SP and SW and if there was a mid-point meeting the SW would be involved and the SW would also be spoken with before it would be closed, ideally a closing meeting with the SW – ideally – however that didn’t always happen – just logistically and as it happened, I think how I would have liked to see FDR roll out in the region is that social workers would be doing more of it. That’s what I would have liked to have seen. Doing the actual service.  5.7.5 Outcome Monitoring   Andrea explained that one of the original goals of FDR implementation was “…to reduce families being reported over and over under the old family assessment model… to inoculate them, or connect them, not just do the same things over and over. So recidivism”. She also explained how a reduction in the number of removals was an unanticipated outcome that MCFD began to monitor. Colin recalled that under FDR there were less removals overall and better job satisfaction. Dorothy and Colin both highlighted better relationships with clients and, by extension, the community as an additional perceived outcome of FDR use.  107  5.8 Overall Results: Investigation was the preferred response in the overall sample (74.3%) but the proportion of FDR designated intakes reached 45 percent in 2011 and surpassed investigation (51.7%) by 2012. The three most common caller types in descending order were police, schools and community professionals. FDR was used more often for calls from the police, schools or the subject child while investigations was used more often for reports from a residential caregiver, an anonymous caller or a CF and CS worker. In approximately 34 percent of cases the same FS file had repeating intakes with 26 percent of them occurring within one year. A small percentage of intakes experienced a change in designation within one year (4.6%) and of those 1,565 intakes where the designation changed, 859 were investigation to FDR and 706 were FDR to investigation.  Four teams were chosen to compare the use of FDR between offices that were pilot sites and those that were not pilot sites. In 2008 the use of FDR at pilot sites was 28.8 percent, rose to 43.5 percent in 2009, and remained above 50 percent until 2012. In contrast, the use of FDR at the non-pilot sites started in 2008 at 7.9 percent, dropped to 5.1 percent in 2009, rose to 9.7 percent in 2010 and increased to 45.8 percent in 2011 and again to 51.6 percent in 2012. The pilot sites show a steady increase of approximately 10 percent per year in the use of FDR but the drastic increase between 2010 and 2011 in the non-pilot FDR sites raises questions of a significant incident in or around 2010 which led to the increased use of FDR at non-pilot sites from 10 to 50 percent.  Finally, a regional comparison was made between a region where the entire FDR assessment and service was completed by a delegated child protection social worker (Vancouver/Richmond) and a region where aspects of FDR (most often the services) were 108  contracted to an outside agency (Langley/Delta). FDR was used more frequently in Vancouver/Richmond (33.9%) than in Langley/Delta (29.9%).  Both regions contained FDR pilot sites but Vancouver/Richmond was the only region to have stand-alone FDR teams where both assessment and service were provided by fully delegated child protection social workers.  Three interviews were conducted to obtain a fuller understanding of the data. A line worker, a team leader and a manager were all interviewed because they had different roles and were able to provide insights at different levels of FDR implementation. All three interviewees were proponents of FDR and believed in the strengths-based, solutions focussed approach but recalled resistance from some line workers who preferred to use investigations. This friction did lead to some workers leaving teams after FDR was implemented. The line worker described how FDR was used for low level neglect or exposure to domestic violence cases or cases where cooperation from parents was likely to lead to success. The team leader and manager highlighted the confusion that came from regional versus provincial expectations of how FDR was to be implemented and described differences in implementation from the use of different assessment tools to contracting out services in certain regions. As a result, consistency in FDR implementation was a challenge for the Ministry.   Despite the challenges in early implementation it appears that Victoria gave teams the direction in 2010 to code the majority of intakes as FDR unless workers were told not to. According to the team leader this was done based on perceived successes at pilot sites and reduced removals and recidivism rates (a trend also observed by the manager). The team leader and manager explained how FDR resulted in fewer removals and lower recidivism rates and the line worker offered examples of how engaging families with FDR provided a more positive image of the Ministry to the community. The manager stated that one of the goals of FDR was to 109  reduce the number of repeating calls coming in for the same family but this assertion was not supported by the data. The data and interviews offer tentative support to the notion that case characteristics were not the only determinants of pathway designation. It appears as though the Ministry made a concerted effort to change its practice and directed staff to code more files as FDR whenever possible; the effects of this change appear most pronounced between 2010 and 2011. Unfortunately, there was not sufficient time in this study to peruse individual files to further probe for information about this practice change so the following chapter will explore organizational factors which may have contributed to this shift.                 110  6 Discussion  6.1 Results 6.1.1 Review of Research Questions The purpose of this research was to evaluate the fidelity of implementation of Family Development Response (FDR) in British Columbia, specifically, if FDR accurately distinguishes between levels of need and risk. Designation and re-designation in a sample of 33,880 intakes was examined between January, 2008 and March, 2012 (a different computer data management system was implemented in April). Four questions were developed to give structure to the research. This final chapter will begin with a discussion about the rationale behind the methods used to answer each question followed by interpretation of the findings. The chapter will then move to whether the expectations of the study were met and conclude with implications and directions for further research. i. Is the goal of coding a certain proportion of intakes as FDR consistent with the available research on abuse/maltreatment rates? Chapter two explored differential response’s origins and suggested that it might be ‘old wine in new bottles’ based on existing concepts like family preservation and family-centered clinical practice, the ecosystems perspective and an emphasis on parent and community engagement. The emergence of FDR in British Columbia was driven by a need to manage the consequences of residualism like high call volumes, resource and staffing limitations and costs associated with increased numbers of children entering into care. Given the context in which FDR was implemented in BC where defraying the costs by keeping families out of the system was the modus operandi, a more appropriate title might have been family ‘deferred’ response. The notion that the expenses of having children enter into state care can be offset by offering families front-111  loaded services to maintain children in the home is not a new one but might have served an important purpose within the immediate BC context, trying to bring preservation and investment into prevention back into the child welfare picture. ii. What is the proportional designation between family development response and a traditional investigative approach at decision point two? Investigation was the dominant overall response to protection reports in the sample data but there was a marked increase in the use of FDR between 2010 and 2011 and FDR eventually overtook INV for the first time in MCFD history as the primary response to child protection reports in British Columbia in 2012. The increase in the use of FDR in the sample between 2010 and 2011 was not gradual so efforts were made to explain where this change might have come from.  One of the expert informants explained how a practice directive originating from Victoria was disseminated to teams to code reports in presumption of FDR based on perceived successes at the provincial FDR pilot sites. These perceived successes were in the form of lower rates of removals and lower recidivism rates as disclosed by the manager. If case characteristics were the major determinants of pathway designation it might be reasonable to expect that the rate of FDR intakes in the province would not have risen so dramatically in such a short period of time. The increased use of FDR in the data coincides with the ‘presumption to code in favor of FDR’ directive of 2010 by Victoria but the timing is up for question. FDR was first mentioned in provincial policy documents in February of 2003 and had been part of the province’s response repertoire for seven years prior to the presumption being implemented. What would have happened to lead the Ministry to believe that FDR should be used as the default response to reports?  112  Chapter three explored how the bifurcation of protection reports into FDR and INV might have been the result of MCFD’s attempt to adapt to its institutional and organizational environment following several important events in British Columbia’s child welfare history. MCFD sought to re-establish positive relationships with communities and address the problems associated with becoming a threshold system. It was faced with pressure to preserve the family yet witnessed the consequences of what could happen if only the highest priority children and families were served.   iii. What is the actual rate of re-designation of intakes coded from family development response to investigation and under what conditions does re-designation from Family Development Response to Investigation occur? Attempts will be made to isolate the different ways that intakes are re-coded. According to the sample data there were 1,565 cases in which there was a pathway re-designation between FDR and INV with 859 going from INV to FDR and 706 going from FDR to INV. This result had to be obtained by the use of a proxy (repeating intakes on the same FS file within one year that had a different designation after the last intake closed) because the ‘Investigation Following FDR’ (IFD) designation was used only seven times in the sample and there was no inherent coding mechanism built into MIS to track files where the coding changed from INV to FDR. It was not possible within the scope of this study to examine individual case characteristics to determine conditions under which re-designation took place so organizational factors were examined instead.  iv. What are the objectives (explicit or implicit) of Family Development Response in British Columbia? Does it meet those objectives?  113  The expert informants interviewed for chapter five described their experiences with implementing FDR on three different levels (on the front line, supervising FDR workers and province-wide implementation). While the coding criteria for FDR was clearly articulated in 2003, increased discretion between regions along with managerial fiat over time resulted in a blurring of distinctions at the early decision making stages. The manager said that this was a general shift in direction over time as FDR became the default designation to protection reports wherever possible. This marks a more general shift in approach that was more about shifting practice and less about differentiating clients, a finding that was supported in the results section. If this was the case, the objective of “rehabilitating” the Ministry’s image might have been achieved but raises the question of whether the FDR/INV dichotomy is necessary. Insofar as re-designation (while statistically significant) occurred in fairly small numbers and reasonably equivalent proportions in the sample, there is some evidence to suggest that the differentiation between FDR and INV was accurate and to the extent that high and lower risk reports could be differentiated, successful. More important indicators of success will include whether the number of children coming into care from those streams remains different and whether families actually got help from referred services whether mandated or voluntary. If the families designated FDR were referred to services and got them, it could be determined if this may be attributed in part to a more voluntary approach introduced by a social worker trained in engagement and assessment.  The application of clinical social work skills is not isolated to one response type as much as it may be strengthened by the worker’s openness and managerial support in the exercise of professional judgment. Families might respond more positively to a social worker who has refined skills in engagement and assessing safety, and the development of a robust, well-trained front line social work staff might be a more worthwhile priority than re-imagining existing 114  models of social work theory and practice. An adequately-staffed, well-trained, well-resourced systemic answer to the problem of child abuse and neglect may be a more suitable benchmark on the way to reducing overall rates of child maltreatment and family disruption.  6.2 Comparisons with Existing Research This study evaluated the fidelity of implementation of FDR in British Columbia with respect to coding decisions by child protection social workers in the Ministry of Children and Family Development. The contribution of this research to the body of knowledge of FDR in child welfare has to do with screening and if FDR accurately distinguishes between levels of need and risk. Support for the accuracy of designation inheres in the relative small proportion of re-designation but other distinguishing outcomes remain to be studied. This research suggests that case characteristics were not the only determinants of pathway designation and that organizational factors might have played a significant role in how child protection reports were responded to.   6.3 Study limitations There are several limitations to this study. First, time limitations prevented detailed examination or perusal of individual files for case characteristics which may have differentiated INV from FDR intakes. Profiles might have been compiled to obtain a fuller understanding of how designation decisions were made by screeners but at this time organizational factors were used to explain differences. The sample size was large enough to make several comparisons like pilot sites and non-pilot sites or between geographic regions but as the interviewees reported, there were regional differences in implementation so conclusions must be made tentatively. Andrea was quite clear that regions implemented FDR based on the needs of their individualized communities so it is reasonable to expect that there were variations between offices and regions 115  regarding how coding decisions were made based on tools used, supervision and a possible myriad of other factors. In other words, the challenge of implementation consistency might preclude conclusions about any comparisons made between regions or even teams. 6.4 Broader Implications and Future Research I would have liked to take a representative sample of the data and peruse individual physical FS files in-depth for case characteristics distinguishing FDR and INV intakes and the contextual factors surrounding individual instances of pathway re-designation. I would have also liked to look at specific organizational factors like the province’s budget between 2008 and 2012 and how many child protection social work staff were on the Ministries’ payroll in this time frame to see whether these were significant correlates of FDR or INV pathway designation or re-designation. A detailed analysis of contextual factors contributing to differences in case designation based on a random sample was unfortunately not within the scope of this thesis.   Opportunities to compare FDR to INV intakes on a wide range of variables open the door to many research possibilities and questions. Number of children on the file, length of opening of intake, past history on the family, age of the children, nature and severity of concern are but a few of the variables that can be examined when comparing between INV and FDR intakes. One example of a research question might be whether FDR is appropriate for high risk cases. This could be assessed by looking backwards at case designation and the existing risk rating used by MCFD’s priority system and comparing it to outcomes in the INV or FDR groups like future number of intakes, protection court orders and child mortalities. It would even be possible to examine what case-specific factors contributed to such outcomes like contracted parenting support services, family conferences, and community-based supports. For this study only re-designation, comparison between pilot and non-pilot sites and comparison between regions 116  where FDR was performed in-house versus contracted out were compared but the possibilities seem nearly limitless.   Since this research was completed, the default to code protection all incoming protection reports as family development response has been institutionalized in the administrative recording apparatus that the Ministry employs. I take this presumption to code protection reports in favour of FDR as further credence to the argument that administrative fiat trumps case characteristics in the determination of pathway. It is clear that the use of FDR had a drastic increase in 2011 which coincided with a directive to code in its favour but this also means that case characteristics are not the primary determinant of pathway assignment.                117  References Androsoff, A. (2013). Pacifist “terrorists” in the “peaceable kingdom”: Cultural conflict in twentieth-century Canada. Journal for the Study of Radicalism, 7(1), 1 – 35.  Armitage, A. & Murray, E. (2007). Thomas gove: A commission of inquiry puts children first and proposes community governance and integration of services. In L. T. Foster & B. Wharf (Eds.), People, politics and child welfare in British Columbia. (139 – 157).Vancouver: UBC Press. Bala, N. (2004). Child welfare law in Canada: An introduction. In Bala, N., Zapf, M. K., Williams, R. J., Vogl, R. & Hornick, J. P. (Eds.). (2004). Canadian child welfare law: Children, families and the state (2nd ed.). Toronto, ON: Thompson Educational Publishing, Inc.  Bednar, S. (2003). Elements of satisfying organizational climates in child welfare agencies. Families in Society, 84(1), 7 – 12. Berg, I. K. & Kelly, S. (2000). Building solutions in child protective services. New York, NY: Norton & Company, Inc.. Bertalanffy, L. V. (1973). General systems theory; Foundations, development, applications. New York, NY: G. Braziller.  Brown, J. (1992). Family-centered services: Planning, policies, and professional standards. In J. Brown & M. Weil (Eds.), Family Practice: A Curriculum Plan for Social Services (pp. 23 – 42). 118  Buckley, H. (2000). Child protection: An unreflective practice. Social Work Education, 19(3), 253 – 263. Buckley, H. & O’Nolan, C. (2014). Child death reviews: Developing CLEAR recommendations. Child Abuse Review, 23, 89 – 103. Callahan, M. & Walmsley, C. (2007). Rethinking child welfare reform in British Columbia, 1900-1960. In L. T. Foster & B. Wharf (Eds.), People, Politics and Child Welfare in British Columbia (11 – 33). Callahan, M. & Swift, K. (2007). Great expectations and unintended consequences: Risk assessment in child welfare in British Columbia. In L. T. Foster & B. Wharf (Eds.), People, Politics and Child Welfare in British Columbia (158 - 183). Cameron, G., Coady, N. & Adams, G. R. (2007). Finding a fit: Family realities and service responses. In G. Cameron, N. Coady & G. R. Adams (Eds.), Moving toward positive systems of child and family welfare. (vii – xxii). Ontario: Wilfred Laurier University Press. Cameron, G., Freymond, N., Cornfield, D. & Palmer, S. (2007). Positive possibilities for child and family welfare: Expanding the anglo-american child protection paradigm. In G. Cameron, N. Coady & G. R. Adams (Eds.), Moving toward positive systems of child and family welfare. (vii – xxii). Ontario: Wilfred Laurier University Press. Charles, G. & Gabor, P. (2006). An historical perspective on residential services for troubled and troubling youth in Canada revisited. Relational Child and Youth Care Practice. 19(4), 17—26. Retrieved from http://www.cyc-net.org/reference/refs-history-canada.html 119  Conley, A. (2007). Differential response: A critical examination of a secondary prevention model. Children and Youth Services Review, 29(2007), 1454 – 1468. Conley, A. & Berrick, J. D. (2010). Community-based child abuse prevention: Outcomes associated with a differential response program in California. Child Maltreatment, 15(4), 282 – 292. Douglas, E. M. (2013). Child welfare workers who experience the death of a child client. Administration in Social Work, 37(1), 59 – 72.  Dumbrill, G. C. (2006). Parental experience of child protection intervention: A qualitative study. Child Abuse & Neglect,30, 27–37http://dx.doi.org/10.1016/j.chiabu.2005.08.012. English, D. J., Wingard, T., Marshall, D., Orme, M. & Orme, A. (2000). Alternative responses to child protective services: Emerging issues and concerns. Child Abuse & Neglect, 24(3), 375 – 388. Fallon, B., Trocme, N. & MacLaurin, B. (2011). Should child protection services respond differently to maltreatment, risk of maltreatment, and risk of harm? Child Abuse & Neglect, 35(2011), 236 – 239. First Call BC Child and Youth Advocacy Coalition (2013). British Columbia 2013 child poverty report card. Vancouver, BC: First Call BC Child and Youth Advocacy Coalition retrieved July 2, 2013 from http://worstincanada.org/wp-content/uploads/2013/11/First_Call_Report_Card_2013_web_FINAL.pdf 120  Foster, L. T. (2007). Trends in child welfare: What do the data show? In L. T. Foster & B. Wharf (Eds.), People, politics and child welfare in British Columbia. (34 – 65).Vancouver: UBC Press. Franke, T., Bagdasaryan, S. & Furman, W. (2011). Differential response in rural counties: Path differentiation, service receipt, and case disposition. Journal of Public Child Welfare, 5(1), 1 – 22. Fournier, S. & Cray, E. (1998). Killing the indian in the child: Four centuries of church-run schools. In S. Fournier & E. Crey (Eds.), Stolen from our embrace: The abduction of first nations children and the restoration of aboriginal communities. (47 – 80). Vancouver: Douglas & McIntyre Ltd. Glor, E. D. (2007). Assessing organizational capacity to adapt. Emergence: Complexity and Organization, 9(3), 33 – 46. Harris, N. (2012). Assessment: When does it help and when does it hinder? Parents' experiences of the assessment process. Child & Family Social Work, 17, 180–191 http://dx.doi.org/10.1111/j.1365-2206.2012.00836.x. Hasenfeld, Y. (1992). The nature of human service organizations. In Y. Hasenfeld (Ed.), Human Services as Complex Organizations. (3 – 23). Newbury Park, CA: Sage Publications. Hasenfeld, Y. (2010). Organizational responses to social policy: The case of welfare reform. Administration in Social Work, 34(2), 148 – 167. 121  Hayers, D. & Spratt, T. (2014). Child welfare as child protection then and now: What social workers did and continue to do. British Journal of Social Work, 44(3), 615 – 635.  Hughes, R. C., Rycus, J. S., Saunders-Adams, S. M., Hughes, L. K. & Hughes, K. N. (2013). Issues in differential response. Research on Social Work Practice, 23(5), 493 – 520. Ivanova, I. (2013). Reality check on government spending: Is the overspending fear mongering justified. Retrieved from https://www.policyalternatives.ca/publications/commentary/reality-check-government-spending-overspending-fear-mongering-justified. Jennissen, T. & Lundy, C. (2011). One hundred years of social work: A history of the profession in English Canada. Waterloo, ON: Wilfred Laurier University Press. Johnston, P. (1983). Native children and the child welfare system. Toronto: Canadian Council on Social Development in association with James Lorimer.  Kanani, K., Regehr, C. & Bernstein, M. B. (2002). Liability considerations in child welfare: Lessons from Canada. Child Abuse & Neglect, 26(10), 1029-1043. Kriz, K., Slyter, E., Iannicelli, A. & Lourie, J. (2012). Fear management: How child protection workers engage with non-citizen immigrant families. Children and Youth Services Review, 34, 316 – 323. Kyte, A., Trocme, N. & Chamberland, C. (2013). Evaluating where we’re at with differential response. Child Abuse & Neglect, 37(2-3), 125 – 132. 122  Laird, J. (1985). Working with the family in child welfare. In J. Laird & A Hartman (Eds.), A handbook of child welfare: context, knowledge, and practice (pp. 360 – 396) New York: Macmillan.  Litwak, E. & Meyer, H. J. (1966). A balance theory of coordination between bureaucratic organizations and community primary groups. Administrative Science Quarterly, 11(1), 31 – 58. Loman, L. A. & Siegel, G. L. (2008). Alternative response in Minnesota: Findings of the program evaluation. Protecting Children, 20(2/3), 78 – 92. Loman, L. A., & Siegel, G. L. (2004). Minnesota alternative response evaluation: Final report. Saint Louis, MO: Institute of Applied Research. Litwak, E. & Hylton, L. F. (1962). Interorganizational analysis: A hypothesis on coordinating agencies. Administrative Science Quarterly, 6(4), 395 – 420. Magnusson, W., Carroll, W. K., Doyle, C., Langer, M. & Walker, R. B. J. (1984). The new reality: The politics of restraint in British Columbia. Vancouver, BC: New Star Books. Marshall, S. K., Charles, G., Kendrick, K. & Pakalniskiene, V. (2008). Comparing differential responses within child protective services: A longitudinal examination. Child Welfare, 89(3), 57 – 77. McCroskey, J. & Meezan, W. Family-Centered Services: Approaches and effectiveness. The Future of Children, 8(1), 54 – 71.  123  McCroskey, J., Weil, M. & Finch Jr., W. A. (1992). Reclaiming leadership in family and children’s services. In J. Brown & M. Weil (Eds.), Family practice: A curriculum plan for social services. (251 - 264). Washington, DC: Child Welfare League of America. McKenzie, B. (2011). Differential response in child welfare: A new early intervention model? In K. Kufeldt & B. McKenzie (Eds.), Child Welfare: Connecting Research, Policy and Practice (pp. 101 – 115). Ontario: Wilfred Laurier Press. Merkel-Holguin, L., Kaplan, C. & Kwak, A. (2006). The national study on differential response in child welfare. Denver, CO: American Humane. Mullaly, R. (1993). Structural Social Work. Toronto: McClelland & Stewart Inc. British Columbia Ministry of Children and Family Development. (2004). Family development response or investigation: determining the most appropriate response to child protection reports. Retrieved June 6, 2014, from http://www.mcf.gov.bc.ca/child_protection/pdf/fdr_guide_internet.pdf National Quality Improvement Center on Differential Response in Child Protective Services. (November 2011). Differential response in child protective services: A literature review. Retrieved July 2, 2013, from http://www.ucdenver.edu/academics/colleges/medicalschool/departments/pediatrics/subs/can/DR/qicdr/General%20Resources/QIC-DR_Lit_Review%20version%20%202.pdf National Quality Improvement Centre on Differential Response in Child Protective Services. (June 2009). Differential response: Early implementation and fidelity. Retrieved July 2, 2013, from 124  http://www.ucdenver.edu/academics/colleges/medicalschool/departments/pediatrics/subs/can/DR/qicdr/Documents/Year%201%20Site%20Reports/CrossSiteMay%202012.pdf Northen, H. (1982). Clinical social work. New York: Columbia University Press.  Ortiz, M. J., Shusterman, G. R., & Fluke, J. D. (2008). Outcomes for children with allegations of neglect who receive alternative response and traditional investigations: Findings from NCANDS. Protecting Children, 23(1/2), 57–70. Parton, N. (1998). Risk, advanced liberalism and child welfare: The need to rediscover uncertainty and ambiguity. British Journal of Social Work, 28, 5 – 27. Perlman, H. H. (1979). Relationship: The heart of helping people. The University of Chicago Press: Chicago, IL.  Robinson, N. D. (1985). Supplemental services for families and children. In J. Laird & A Hartman (Eds.), A handbook of child welfare: context, knowledge, and practice (pp. 397 - 416) New York: Macmillan.  Ruiz-Casares, M., Trocme, N. & Fallon, B. (2012). Supervisory neglect and risk of harm. Evidence from the Canadian child welfare system. Child Abuse & Neglect, 36(2012), 471 – 480. Sawyer, R., & Lohrbach, S. (2005). Differential response in child protection: Selecting a pathway. Protecting Children, 20(2/3), 44–53. 125  Scarth, S. & Sullivan, R. (2007). Child welfare in the 1980s: A time of turbulence and change. In L. T. Foster & B. Wharf (Eds.), People, politics and child welfare in British Columbia. (83 – 96).Vancouver: UBC Press. Schrieber, J. C., Fuller, T & Paceley, M. S. (2013). Engagement in child protective services: Parent perceptions of worker skills. Child and Youth Services Review, 35(4), 707 – 715. Sedlak, A. J., Mettenburg, J., Basena, M., Petta, I., McPherson, K., Greene, A., & Li, S. (2010). Fourth national incidence study of child abuse and neglect (NIS-4): Report to congress. Washington, DC: US Department of Health and Human Services, Administration for Children and Families. Shusterman, G. R., Hollinshead, D., Fluke, J. D. & Yuan, Y. T. (2005). Alternative responses to child maltreatment: Findings from NCANDS. Protecting Children, 20(2/3), 32 – 42. Smith, B. D. & Donovan, S. E. F. (2003). Child Welfare Practice in Organizational and Institutional Context. Social Service Review, 77(4), 541 – 563. Stanbury, W. (1975). Success and failure: Indians in urban society. Vancouver: UBC Press.  Stevenson, O. (1997). Child welfare in the UK: the exercise of professional judgments by social workers, paper presented to the BASPCAN Congress, July 1997, Edinburgh. Strong-Boag, V. (2002). Getting to now: Children in distress in Canada’s past. In B. Wharf (Ed.), Community work approaches to child welfare (pp. 29-46). Peterborough, ON: Broadview Press. Sullivan, R. (1992). Human behaviour in the social environment. In J. Brown & M. Weil (Eds.), Family Practice: A Curriculum Plan for Social Services (pp. 91 - 113). 126  Trocme, N., Fallon, B., MacLaurin, B., Sinha, V., Black, T., Fast, E., et al. (2008). Canadian Incidence Study of Reported Child Abuse and Neglect – 2008: Major findings. Ottawa, ON: Public Health Agency of Canada, Injury and Child Maltreatment Section.  Tucker, D. J., Baum, J. A. C. & Singh, J. V. (1992). The institutional ecology of human service organizations. In Y. Hasenfeld (Ed.), Human Services as Complex Organizations. (47 – 72). Newbury Park, CA: Sage Publications. Waldfogel, J. (1998). Rethinking the paradigm for child protection. The future of children: Protecting children from abuse and neglect, 8(1), 104 – 119. Wallace, B., Klein, S. & Reitsma-Street, M. (2006). A call for accountability at the ministry of employment and income assistance. Retrieved July 2, 2013 from https://www.policyalternatives.ca/publications/commentary/call-accountability-ministry-employment-and-income-assistance. Wharf, B. (2006). Introduction: People, politics, and child welfare. In L. T. Foster & B. Wharf (Eds.), People, Politics and Child Welfare in British Columbia (pp. 1 – 9). Wiltse, K. T. (1985). Foster care: an overview. In J. Laird & A Hartman (Eds.), A handbook of child welfare: context, knowledge, and practice (pp. 565 – 584) New York: Macmillan.  Ziefert, M. (1985). Homemaker and day care services. In J. Laird & A Hartman (Eds.), A handbook of child welfare: context, knowledge, and practice (pp. 417 – 438) New York: Macmillan.    127  Appendix A Interview Guide  Interview Questions for Line Workers 1. When receiving section 13 reports, were there times when you had difficulties making decisions to code as Family Development Response or investigation based on policy criteria? Have there been other challenges? 2. Do you feel that the training you received prepared you to provide Family Development Response? Interview Questions for Team Leaders 1. When receiving section 13 reports, were there times when you had difficulties making decisions to code as Family Development Response or investigation based on policy criteria? Have there been other challenges? 2. How has Family Development Response practice evolved over time in the Langley/Delta service area and what influenced the changes? Where did these changes come from and when did they occur? 3. How was the decision made to provide Family Development Response as a contracted service by non-delegated workers and when was the decision made? 4. Was there anything that happened in 2010 or around that time that may account for an increase in the use of Family Development Response in the region? If so, what was it? Interview Questions for Managers 1. Have there been any policy changes since Family Development Response was first implemented? If so, what were the changes and how were they put into place? If you could provide specific dates and times, this would be very helpful (particularly between the years of 2008 and 2012). 2. Are there circumstances in which implementing Family Development Response required policy or practice changes? What necessitated those changes and how were those needs identified and addressed? 3. Have there been any challenges in achieving consistency for the application of Family Development Response eligibility criteria? Has eligibility criteria changed since initial implementation and if so, when did this change occur? 4. How was the decision made to provide Family Development Response as a contracted service by non-delegated workers and when was the decision made? 5. Was there anything that happened in 2010 or around that time that may account for an increase in the use of Family Development Response in the region? If so, what was it? 6. Has training for Family Development Response changed since it was first implemented? 7. Were there any criteria for selecting workers to perform the Family Development Response function? 8. What are the outcomes that are monitored to assess Family Development Response performance? How are they monitored?  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items