3312 Northside Drive Suite D-250 Macon, Georgia 31210 478-757-2661 478-757-2666 (fax) www.oca.georgia.gov Office of the Child Advocate Tom C. Rawlings, Child Advocate 55 Park Place, NE Suite 410 Atlanta, Georgia 30303 404-656-4200 404-656-5200 (fax) www.gacfr.oca.georgia.gov OCA Analysis: Bryan Moreno and DFCS Performance Measures Introduction and Executive Summary Pursuant to its statutory mandate,1 the Office of Child Advocate regularly monitors and reviews the policies and practices of Georgia's Division of Family and Children Services (DFCS) and makes recommendations to improve the quality and effectiveness of that agency's child protective services work. With that objective, OCA reviewed and analyzed the Forsyth County case of Bryan Moreno, an autistic, nonverbal six-year old who police allege was beaten to death by his mother's boyfriend in July 2009. Reports of possible abuse to the child had been made to Forsyth DFCS three separate times during the year prior to Bryan's death, and on none of those occasions did the agency determine that Bryan was in any danger. While individual child protective services staff involved in Bryan's cases may have made errors, OCA's analysis concludes that larger systemic issues are involved here. Specifically, OCA is concerned that the county leadership may have interpreted DFCS' performance measures in a way that placed too high a priority on achieving certain caseload reduction objectives and time deadlines, to the exclusion of a focus on providing quality social work services. We encourage DFCS to review its performance measures and management practices to ensure that counties place the highest priority on customer service and child safety. Facts of the Moreno Case Although six-year-old Bryan Moreno had autism and could not tell teachers what happened, officials at his Forsyth County school could see the hand-shaped reddish and purple bruise on his upper thigh. Afraid for his safety, they called the local Division of Family and Children Services to make a report on September 2, 2008. The agency responded quickly, but the mother said she believed any bruises came from the child roughhousing with his brother and cousins. Because Bryan became agitated, the case manager was not able to see the bruises for herself or document them with photos. The case manager spoke with Eder Acosta, the boyfriend of Bryan's mother, who told the interviewer that he did not live in the home and that Bryan's mother did a good job of caring for the children. After visits to the family and collateral contacts were made, the case was closed as "unsubstantiated." On November 7, 2008 one day after the first investigation was closed in the SHINES2 system another call came in concerning Bryan. Again, mandated reporters at Brian's school were concerned about injuries she had suffered. This time, the complaint was that the child had scratches on his face. While Brian's mother had claimed a cat made them, school officials with medical training expressed concern to DFCS that the scratches were too severe to have been made by a cat. Rather than investigating again, the agency "screened out" the complaint because, according to its files, the case was "previously investigated and closed" and the prior investigator had at some point seen the child playing with kittens. 1 See OCGA 15-11-173. 2 SHINES is the state's automated child welfare information system. Moreno Analysis Page 2 of 9 12/7/2009 The agency again had the opportunity to intervene with the family when it received another complaint from a mandated reporter on January 30, 2009. This time, Bryan's older brother, who lives with his father, said he did not want to go to visit his mother. Mother and her boyfriend, Eder Acosta, fight and hit each other with their fists, he said, and Eder had punched holes in the wall. The child also reported that his brother Bryan lived in that home, and that Eder had "one time . . . hit Bryan on the leg very hard with his fist because Bryan wouldn't go to the bathroom." The agency's intake report notes that the child may have been referring to the case noted above that was investigated in September 2008. When the DFCS case manager interviewed Jose, he told her his mother and Mr. Acosta do fight but that he did not know whether Mr. Acosta hit Bryan or the other child in the home. The case manager then interviewed Mr. Acosta and Bryan's mother, apparently allowing Mr. Acosta to serve as a translator for Bryan's mother, who did not speak English. According to the DFCS documentation, the agency decided that after doing some routine follow-up work the case would be closed and the allegations unsubstantiated. It was closed in the DFCS database on March 9, 2009. On July 16, 2009, according to police reports, Eder Acosta took Bryan's mother to work and then returned to the trailer where he beat Bryan to death. The child arrived at the hospital emergency room in full cardiac arrest. When interviewed in private with a translator after Bryan's death, the mother acknowledged that Mr. Acosta was violent and that she did not tell DFCS about her situation because Mr. Acosta told her the agency would have her deported. This case could be analyzed by focusing in hindsight on the various individual actions and failures to act that prevented the Department from recognizing the danger in which Bryan Moreno lived. In the September incident, one could chastise a case manager for taking the mother's word that Bryan had received his bruises while roughhousing and for failing to fully examine Bryan as required by DFCS policy.3 One could also fault the agency's workers for failing to take seriously the November allegation, made by a mandated reporter with medical expertise, that Bryan's scratches "could not have come from a cat." And, significantly, one could fault county investigative staff for failing to treat seriously the January 2009 statement of Bryan's brother that may have explained the source of the hand-shaped bruises on Bryan the previous September. To focus on the individual actions of these line workers within a hierarchical agency, however, would be to ignore evidence of systemic issues. After all, numerous case managers and supervisors dealt with the Moreno family over the course of those six months, and the agency's failure to intervene more strongly in the family cannot be attributed to any single worker or supervisor.4 Since undertaking the investigation of this child's death, the Office of the Child Advocate has attempted to determine what, if any, systemic issues policies, practices, attitudes, and pressures may have contributed to the tragedy of Bryan Moreno. Based on all the evidence, OCA has concluded that certain current agency practices and policies may have been interpreted by this county DFCS in ways that did not promote thorough investigations of child maltreatment reports. G-Force and the Use of Performance Measures Over the past five years, the Division of Family and Children Services has concentrated on reducing the numbers of children in foster care, the number of investigations, and the caseloads of its case managers. Some of this reduction was natural and necessary following a "spike" in foster care, investigations, and ongoing family preservation or family services caseloads. That spike resulted in large part from policies in effect for a 3 DFCS policy 2104.11 4 It appears 8 different workers and supervisors were involved with different aspects of the three reports to DFCS prior to Bryan's death. Moreno Analysis Page 3 of 9 12/7/2009 period of time that required DFCS workers to thoroughly investigate most reports and encouraged the removal of children if certain risk factors existed. Possibly because of these policies, the number of "open" child protective services cases open at any one time spiked to over 32,000 in early 2004, and the number of children in foster care rose dramatically, to over 14,500. Georgia was removing children at the incredibly high rate of 5 children per 1,000 in the population.5 Realizing that such surges could overwhelm the system, the state reversed those policies and began reducing those removal and foster care numbers.6 Over the past four years, DFCS leadership has managed caseloads by focusing on practice and performance measures. Those performance measures and outcomes are highlighted in regular "G-Force" meetings that agency leadership holds with regional and county-level management. Several of these practices have already been noted with concern by OCA in earlier reports. They include a significant increase in the use of "Diversion," a dramatic reduction in the number of investigations and family preservation cases, and the inappropriate use of safety resources and temporary guardianships in lieu of foster care.7 Diversion Diversion, now referred to by the agency as "family support," is a procedure in which each county or region may decide not to investigate certain reports that, pursuant to a local protocol, are not considered serious enough for intervention. Those protocols have varied widely among counties and regions, but matters that may be diverted have generally included such problems as prenatal drug use by a mother; domestic violence between parents in which the child is not physically harmed; medical neglect of a child; and child-on-child sexual abuse. In some areas, responding to a Diversion meant having a face-to-face interview with the child and family and ensuring that services were provided. In many areas of the state, however, Diversion could be completed with a phone call and a family referral to other agencies that might be able to help.8 Agency leadership has suggested that Diversion is simply Georgia's version of a national model of social work known as "Alternative Response" or "Differential Response." OCA and DFCS' Child Protective Services Citizen Review Panel, however, expressed concerns that Diversion practice was insufficient because it varied widely from county to county, made no provision for an appropriate family assessment, and had no requirement that anyone determine whether the family received or complied with the referral to services. In response, DFCS in July 2008 commissioned a report from the Carl Vinson Institute of Government. That report, released this summer, concluded that DFCS should create a consistent, statewide alternative response policy that provides for a uniform assessment of each family's needs, collaboration with the family to obtain the necessary outside assistance, and follow-up to ensure the family has obtained the necessary assistance or help.9 DFCS has contracted with a consultant from North Carolina to help it create such a policy and implement appropriate practice. Since its current Diversion practice was implemented beginning in 2004, the agency has applied the response to over 100,000 complaints of child abuse or neglect. The agency has promoted its use through the G-Force performance management tool. For a period of time until mid-2008, the agency regularly measured each region's rate of diverting complaints to a "desired pattern" in the shape of a pyramid. The pyramid suggested, as Figure 1 shows, that it was desirable to divert a particular percentage of complaints. In early 2008, as Figure 5 Statistics courtesy Andrew Barclay, Fostering Court Improvement. 6 See "Differential Response/Family Support Services: Policy Analysis and Recommendations" (Carl Vinson Institute of Government, University of Georgia, 2009), pp. 1-2 (hereinafter "CVIOG report") 7 See "Reducing the Foster Care Rolls: Are We Using the Right Tools?" (Office of the Child Advocate, July 2009), available online at http://oca.georgia.gov/vgn/images/portal/cit_1210/62/38/145957858Safety%20Resource%20Report%20FINAL.pdf 8 CVIOG report, supra, pp. 5-6. 9 Id. Moreno Analysis Page 4 of 9 12/7/2009 1 shows, the message to the regional and county management was that 45% of all complaints to the agency should receive a "Diversion" response. While it abandoned the pyramid, the agency continues to measure the use of Diversion as part of its G-Force meetings, and it is clear that leadership has been successful in encouraging the use of Diversion. Currently, as Figure 2 shows, the agency now responds to 53% of all complaints by making a Diversion referral. This push to divert cases has occurred despite the agency's lack of a consistent statewide policy that assures only appropriate cases are Diverted. Figure 1 Reducing Investigations With the use of Diversion, according to the Carl Vinson Institute's analysis, the number of families being investigated for abuse or neglect has dropped 42% to its lowest level since 2002.10 DFCS consistently measures this decline and within Figure 2 its G-Force meeting uses graphs to track the increase or decrease in investigations and other "open" or active cases. The message appears to be that improvement is achieved by increasing the number of diversions and decreasing the number of investigations, or guarding the "front door."11 In its June 2008 presentation, for example, leadership pointed out that "Our front door policy has changed since mid-2004 when diversions was instituted. Since then the number of investigations and active cases have decreased by 63%." Currently, according to DFCS statistics from its latest G-Force presentation, the agency has open at any given time only about 11,000 cases. Of these, approximately 60% are open "family preservation" cases and 40% are investigations. As figure 3 (next page) demonstrates, the agency currently carries only half the open caseload of investigations and family preservation cases it carried in July 2002. It does so with a child protective and foster care services staff that has increased significantly since that time, allowing the agency to further lower caseload ratios.12 Reducing Foster Care Caseloads 10 CVIOG report, p. 3. 11 As the agency noted in its October 2008 G-Force presentation, "Regions developed Wildly Important Goals related to front door practices and have seen improvements in their outcomes." (emphasis added). 12 According to a historical summary posted at http://childwelfare.net/activities/legislative2001/ga_4_kids_agenda_item.html, he agency has approximately 1,500 employees handling child protective services investigations and foster care in 2000. The agency has been allocated at least 500 additional positions since that time. OCA is currently working to verify the current and historical staffing numbers. Moreno Analysis Page 5 of 9 12/7/2009 Used appropriately, "Family Preservation Services" is a wonderful tool for keeping families together even after an incident of abuse or neglect.13 Following its 2004 "spike" in the use of foster care, the agency realized that removing children from their homes was not always the most appropriate response and can be terribly traumatic for the child. The agency began focusing on reducing the use of foster care, and the G-Force performance management meetings reflect an increased focus on measuring foster care caseloads. According to current AFCARS data, Georgia had approximately 8,900 children in foster care on March 31, 2009. That figure represents a decline of over 19% from the previous year, when Georgia ended March Figure 3 2008 with approximately 10,900 children in care. Foster care usage has declined precipitously from its high in 2004, when the state had approximately 14,500 in foster care custody. The foster care numbers, like other performance measures, are consistently tracked and communicated to regional and county managers through the G-Force meetings, with the message being that foster care usage should continue to decline. See Figure 4. The number of children in the state's foster homes is now 1,000 lower than it has been at any point in the last decade even though the state's child population has grown by 20% during that same decade.14 To make equivalent comparisons, the state's "removal rate" of children to foster care now stands at 2 per 1,000 children in the population. This rate is 60% less than the removal rate during the "spike" of 2004 and 43% less than the agency's 3.5 per 1,000 foster care removal rate in 1999 and 2000. See Appendix A. By way of comparison, Missouri has a current removal rate of 2.8 per 1,000; Tennessee's is 3.1, and Florida's is 3.2. Therefore, under current practice a child in Georgia is 38% less likely to be removed to foster care over an allegation of abuse or neglect than a child in Florida, and a Georgia child today is 43% less likely to be removed to foster care than a Georgia child was in 1999 or 2000. Figure 4 13 Family Preservation Services are defined by DFCS policy 2107 as services to "help families help themselves by preserving and strengthening a child's own family and promoting a family's self sufficiency, self determination and independence." Such services may include providing assistance with housing, outpatient drug treatment, parenting classes, and other social services designed to keep the family unit intact. 14 Statistics derived from DFCS PSDS reports and other statistical databases by Andrew Barclay, Fostering Court Improvement. The foster care numbers and removal rates from 1999 to 2008 are displayed in Appendix A. Moreno Analysis Page 6 of 9 12/7/2009 Reducing Caseloads Through Timely Case Management Another goal of the G-Force meetings is to impress upon county and regional leadership the need to reduce the length of time spent on each case and, thereby, to reduce caseloads. In its July 2008 presentation to field staff, agency leadership emphasized the need to complete all investigations within 30 days, and regional management is measured based on the percentage of its investigations that are completed in a timely manner. Impact of G-Force on Local Practice OCA believes the G-Force performance measurement tools have been extremely effective in reducing the number of investigations, reducing caseloads, increasing the timely completion of casework, and reducing the foster care rolls. The success of the focus on these performance measures is clear from the statistics: In the past two years, the number of child protective services cases open at any given time has been cut in half. During that same time, the number of children in foster care declined by 40%, and the agency is currently less likely to remove a child to foster care than at any time in the past decade. The number of cases of substantiated child abuse or neglect that are "closed" without further involvement by DFCS increased by 22%. Over half of all reports of child abuse and neglect are now met with a "diversion" response, meaning that the incidents which may or may not have involved child maltreatment are handled by referring the family to another agency's services. Our agency's recent work -- and specifically our work connected with our inquiry into Bryan's case -- leads us to question whether this success has been indicative of improvement in the child protection system. In an earlier report, OCA detailed how the focus on keeping down caseload numbers has contributed to a significant and problematic rise in the number of children placed without legal permanency in "safety resources" or temporary probate court guardianships. In that report, OCA expressed concern that the focus on reducing these particular numbers was having the perhaps unintended consequence of encouraging county staff and leadership to find other means of disposing of cases. We detailed incidences in which some children were placed in "safety resources" even though their equally-at-risk siblings were in foster care and headed toward adoption. We found other examples of the agency pushing recalcitrant parents to give temporary guardianship of their children to relatives so the agency could close its case. Juvenile court judges also report they are seeing fewer cases brought to their attention. In Gwinnett County, for example, juvenile court abuse and neglect filings decreased 44% from fiscal year 2007 through fiscal year 2009. It appears DFCS leadership in Bryan Moreno's home county of Forsyth focused strongly on meeting these performance measures. A May 2008 internal investigation of the county's DFCS by DHR's Office of the Inspector General found at least 16 cases of 138 that, by the agency's own admission, were improperly assigned a Diversion response. That investigation also found 7 of 73 screenouts "contained information that appeared to need further investigation." According to the Inspector General's report, "it was noted that the questionable cases contained allegations related to drug abuse, neglect, domestic violence, one case involving CPS history of sexual abuse in another state, verbal and emotional abuse, and cases with substantiated history."15 From interviews with current and former staff at Forsyth County DFCS, OCA has determined management placed strong pressure on line case managers and supervisors to investigate and close cases within very strict 15 OIG report, in possession of OCA. OCA is also very concerned that when we spoke with former DFCS employees who participated in that OIG investigation, several felt they had been threatened, fired or retaliated against for sharing their concerns with the Inspector General. The Inspector General's report also indicates that during the investigation, employees voluntarily raised concerns regarding the work environment and allegations of misconduct. The report indicates that these allegations would be referred to the DHR Office of Human Resources. It is unclear what, if any, action was taken following that referral. Moreno Analysis Page 7 of 9 12/7/2009 time frames, possibly with insufficient consideration for the quality of the investigation. Standard statewide agency policy, for example, suggests that cases be completed within 30 days but that a waiver can be requested if there are reasonable grounds for not completing the investigation within that time period.16 Current and former case managers, however, report that in Forsyth County, line workers were pushed to complete investigations and diversions within 21 calendar days. Emails and other communications reviewed by OCA demonstrate that management keeps close track of cases over 21 days and requires workers to supply "anticipated closure dates" on such cases. If a case has been open more than 30 days, supervisors may expect to have to explain during a "cadence call" to the "state office" why the case remains open. Documents reviewed by OCA demonstrate that Forsyth DFCS leadership placed a high priority on meeting numerical targets for reducing the length of time such investigations and diversions are open. Focusing on the Right Work? There is great value in ensuring that the agency intervenes in a family's life only to the extent necessary; that it completes its work with a family as quickly as possible; and that its case managers not be overwhelmed with work to the extent they cannot give each family the necessary attention. OCA is concerned, however, that these particular performance measures do not adequately take into account that different cases have different levels of complexity. Investigating a case alleging lack of adequate housing may be quite different from investigating a suspicious child death. Some parents, witnesses, and children cooperate readily, while others avoid the case manager, refuse drug screens, or are so transient that they are hard for the case manager to track down. Yet Forsyth County DFCS' push for all cases to be completed within a stated deadline does not appear to take such differences into consideration. Likewise, to focus on increasing the number of diversions and reducing the length of time spent on a family's case while, at the same time focusing on reducing the foster care rolls, is of concern. If the agency focused on increasing the number of family preservation cases as well as the quality and quantity of services, a reduction in the foster care rolls might be seen as a sign of success. Instead, the performance measures can be interpreted as having the objective of reducing the entire spectrum of services to families, limiting their access to "the front door" and pushing them through the system as quickly as possible. No evidence can prove or disprove the effect of these policies and practices on the life and death of Bryan Moreno. It is evident, however, that the county's focus on these particular performance measures encouraged workers to prioritize lowering caseloads rather than improving the quality of investigations and family interventions. OCA requests leadership consider the following alternative scenario: When Bryan was reported to DFCS by a mandated reporter at the school, the intake worker well remembered the caution she had received from her director following the last G-Force meeting. "The State Office is really focused on how responsive we are to reports from school counselors and mandated reporters," he said. "You know they're doing regular surveys to determine how responsive and service-oriented each office is considered by our partners in addressing child abuse." Upon hearing about Bryan's case, the case manager's first call was to the school. She spoke with all Bryan's teachers, counselors, and the school nurse. When she visited Bryan's mother and was unable to see the bruises because he was so agitated, she called the school and asked if the nurse could view Bryan the next day and take photos. 16 DFCS Policy 2104.28. According to Region 2 case managers, supervisors, and regional staff, the timeframe for completing investigations is 45 days, not 30. Moreno Analysis Page 8 of 9 12/7/2009 The case manager also knew that quality of investigations was a top priority for the state office and that each region's investigations were analyzed on the number of contacts made in each case and the quality of the case manager's interaction with other child protection partners doctors, law enforcement, and counselors. So although she heard Mr. Acosta say that Bryan's mother was a good parent, she was curious. She talked with many people in the community and found out that Mr. Acosta himself had a history of violence. So she encouraged and assisted Bryan's mother to go speak with a Spanish-speaking family counselor. She also sought the advice of Bryan's special education teachers so she could figure out how best to investigate abuse of a nonverbal child with his special needs. Given the language barriers and the difficulties of dealing with a nonverbal victim, the case manager asked for and received permission to extend her investigation beyond 30 days. It was given without objection by her Director, who reminded her that qualitative outcomes were more important than recommended deadlines. Thus, when the school again reported Bryan for the scratches on his face in early November, the case manager sought out medical advice and politely challenged, through a translator, the mother's assertion that a cat scratched the child. By this time, the mother had begun to trust the case manager and confessed that Mr. Acosta was hurting Bryan and had threatened to hurt her if she told anyone. To ensure Bryan's needs would be met and he and his mother protected, the case worker asked DFCS' attorney to file a deprivation petition in juvenile court to request a protective order for Bryan. The case manager also worked quickly to get Bryan and his mother placed in a domestic violence shelter. The old saying goes that "you get what you measure." In the current situation, it appears that county staff may have read into the G-Force performance measurements a mandate that they focus on achieving reductions, whether the statistic reduced be caseloads, investigations, time to complete investigations, or foster care use. While leadership may very well desire to put qualifiers in that message i.e., reducing unnecessary foster care use or reducing unnecessary investigations, it appears that message is not getting through, at least in this particular county. These performance measures run the danger of effectively encouraging the agency's employees to provide less service to families in need, and for that reason these measurements should be modified or downplayed and the agency should rely more on other measurements of customer service levels. OCA agrees that the agency's employees should be subjected to appropriate performance measures. OCA would recommend the increased use of both qualitative and quantitative performance measures that encourage good work. Some of the quantitative measures, such as statistics measuring the recurrence of maltreatment, should be retained. To those should be added measures such as the amounts and types of services provided, the satisfaction level of professionals and consumers with DFCS' work, or the number of family assessments completed within a given time frame. In fact, it appears the agency is using some very good performance measures when it comes to children who are actually in foster care. The G-Force presentations have provided managers with good information and performance measures for improving the well-being of children in care, ensuring more appropriate permanency, improving handling of teens in care, and reducing the time between termination of parental rights and adoption. By adopting similar qualitative measures to determine the levels of service in intake, investigations, and family preservation programs, the agency can refocus staff on the core mission of providing an adequate level of service to the state's children and families in need of help across the continuum from intake and investigations through foster care. Moreno Analysis Page 9 of 9 12/7/2009 CY 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 SFY 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 FFY 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 APPENDIX A Removal Rates and Foster Care Population, 1999-200817 Child Population 2,144,922 2,193,546 2,242,171 2,290,782 2,339,409 2,388,027 2,436,646 2,485,270 2,533,885 2,582,509 Child Population 2,120,616 2,169,234 2,217,852 2,266,482 2,315,091 2,363,719 2,412,334 2,460,958 2,509,585 2,558,195 Child Pop 2,132,773 2,181,380 2,230,010 2,278,621 2,327,244 2,375,868 2,424,494 2,473,111 2,521,734 2,570,352 Children Removed to Foster Care During the Year Count Annual Rate per 1K 7,528 3.5 7,631 3.5 8,802 3.9 9,781 4.3 11,043 4.7 11,869 5.0 10,926 4.5 9,437 3.8 8,718 3.4 6,481 2.5 Children Removed to Foster Care During the Year Count Annual Rate per 1K 7,612 3.5 8,256 3.7 8,850 3.9 10,283 4.4 12,340 5.2 10,906 4.5 10,671 4.3 9,398 3.7 6,918 2.7 Children Removed to Foster Care During the Year Count Annual Rate per 1K 7,659 3.6 7,340 3.4 8,604 3.9 9,367 4.1 10,787 4.6 11,951 5.0 11,068 4.6 9,880 4.0 9,076 3.6 6,636 2.6 Child Population (on last day of year) 2,165,183 2,213,801 2,262,432 2,311,041 2,359,663 2,408,283 2,456,906 2,505,526 2,554,148 2,602,764 Child Population (on last day of year) 2,140,874 2,189,488 2,238,110 2,286,738 2,335,358 2,383,978 2,432,596 2,481,218 2,529,831 2,578,458 Child Pop 2,153,033 2,201,650 2,250,264 2,298,886 2,347,505 2,396,136 2,444,753 2,493,373 2,541,998 2,590,611 Children In Foster Care on Last Day of Year Count Rate per 1K 10,180 4.7 10,812 4.9 12,031 5.4 13,108 5.7 13,606 5.8 13,839 5.8 13,436 5.5 12,323 5.0 11,386 4.5 9,422 3.6 Children In Foster Care on Last Day of Year Count Rate per 1K 10,318 4.9 10,607 4.9 11,638 5.2 12,530 5.5 13,183 5.7 14,511 6.1 13,667 5.7 13,299 5.4 12,700 5.1 10,401 4.1 Children In Foster Care on Last Day of Year Count Rate per 1K 10,707 5.0 10,635 4.9 12,052 5.4 13,279 5.8 13,711 5.9 14,275 6.0 13,878 5.7 12,899 5.2 12,120 4.8 9,980 3.9 17 Courtesy Andrew Barclay, Fostering Court Improvement