�II. Achieving Safety-Related Outcomes
A primary objective of State child welfare systems is to ensure that children who have been found to be victims of abuse or neglect are protected from further abuse or neglect, whether they remain in their own homes or are placed by the State child welfare agency in a foster care setting. The Department established the following safety-related national outcomes and measures to assess State performance with regard to protecting child victims from further abuse or neglect:
Outcome 1—Reduce recurrence of child abuse and/or neglect
Outcome measure 1.1—Of all children who were victims of substantiated or indicated child abuse and/or neglect during the first 6 months of the reporting period, what percentage had another substantiated or indicated report within a 6-month period?10
Outcome 2—Reduce the incidence of child abuse and/or neglect in foster care
Outcome measure 2.1—Of all children who were in foster care during the reporting period, what percentage was the subject of substantiated or indicated maltreatment by a foster parent or facility staff member?
The Department established the following national standards for these outcome measures for use in its national child welfare monitoring system, the Child and Family Services Review (CFSR). (Appendices E and F provide additional information on the CFSR and the national standards.):
- Of all children who were victims of substantiated or indicated child abuse and/or neglect during the first 6 months of the period under review, 6.1 percent or fewer children had another substantiated or indicated report within 6 months of the first report.
- Of all children in foster care in the State during the period under review, 0.57 percent or fewer were the subject of substantiated or indicated maltreatment by a foster parent or facility staff member.
This chapter presents the following information:
- State performance in 2003 on the safety-related outcome measures.11
- The change in State performance from 2000 to 2003 on the safety-related outcome measures.
- The relationship between State performance on the safety-related outcome measures and the number and characteristics of child maltreatment victims.
For the most part, data reported in this chapter come from the Department's National Child Abuse and Neglect Data System (NCANDS). Through NCANDS, the Department collects and analyzes data on children who come into contact with public child protective services (CPS) agencies as alleged victims of child abuse or neglect. The Department prepares an annual report on NCANDS data entitled Child Maltreatment. The highlights of findings from Child Maltreatment 2003 are presented in appendix G.
The number of States included in each analysis varies because NCANDS is a voluntary system and not all States submit the relevant data to NCANDS. The primary data source for the Report to Congress safety-related outcome measures is the NCANDS Child File. The Child File is a data file that States submit to NCANDS on a periodic basis that contains detailed case information about children who are the subjects of an investigation or assessment in response to a maltreatment allegation. The Child File includes only those children for whom there has been a disposition (i.e., a "finding") pertaining to the maltreatment allegation. The Department has strongly encouraged participation in this data system and provides technical assistance to States to assist them in collecting and reporting the relevant data. Participation in the NCANDS Child File has increased considerably over the past 5 years; 45 States participated in the NCANDS Child File in 2003 compared to 29 States in 1999.
A comprehensive understanding of State performance on the safety-related outcome measures requires a consideration of State statutes, policies, and practices pertaining to child maltreatment that may affect performance. For example, with regard to maltreatment recurrence within a 6-month period (outcome measure 1.1), in several States, it is a child welfare agency policy or an agency-wide practice that when a maltreatment allegation is received on a family currently being served by the CPS system, the allegation is referred to the current caseworker and/or supervisor for follow up rather than being referred for a formal investigation or assessment. In this situation, information on the child who is the subject of the allegation is not reported to the NCANDS Child File because there is no formal investigation or assessment of the allegation and no formal disposition. Consequently, the allegation is not "counted" as a possible incident of maltreatment recurrence.
In addition, some States have implemented an "alternative response" approach to maltreatment allegations. In many of these States, maltreatment allegations that are determined to involve very low risks for child safety are not formally investigated. Instead, the allegation is referred for a family assessment and no disposition is reached regarding the validity of the maltreatment allegation. Because there is no disposition, States do not submit an NCANDS Child File for the children who are the subject of allegations that are referred for an alternative response. States that have implemented this approach may show declines in the percentage of children experiencing maltreatment recurrence and may have lower incidences of maltreatment recurrence than States that have not implemented an alternative response approach. Other examples of the factors that may affect performance on the measure of maltreatment recurrence are State definitions of child maltreatment and State requirements concerning the level of evidence necessary to substantiate a maltreatment allegation.
The policies and practices pertaining to the types of information submitted to the NCANDS Child File also may affect State performance regarding the percentage of children in foster care who are maltreated by their foster parents or a facility staff member. For example, some States submit an NCANDS Child File for all children who are found to be victims of maltreatment by a foster parent or facility staff member, even if the child is not under the placement and care responsibility of the State child welfare system (e.g., the child is in the mental health system, juvenile justice system, or in a private care facility such as a mental health or substance abuse treatment facility). Although this is likely to involve only a small number of children, it can affect the calculation of the measure. That is, for these States, the calculation of State performance on the measure of maltreatment of children in foster care would include children in the numerator (number of children maltreated by a foster parent or facility staff member) who are not in the denominator (number of children in foster care during the reporting period).
The Department recently addressed this issue by instituting a system for identifying those child victims reported to the NCANDS Child File who also are reported to AFCARS as being in foster care. However, because in 2003, this system was not yet implemented by States in a consistent manner, for some States, the percentage of children reported as maltreated while in foster care in 2003 continues to be affected by this issue.
Because of the potential effect of State statutes and policies on performance, this report presents the findings regarding State performance in 2003 on the safety-related outcomes to depict the variation that exists among States rather than to compare the performance of States. Information regarding the change in performance on the outcome measures from 2000 to 2003 is provided to demonstrate the extent and direction of change occurring within individual States.
State Performance in 2003 on the Safety-Related Outcome Measures
Table II-1 presents key findings regarding State performance in 2003 on the safety-related outcome measures. In 2003, 45 States provided sufficient data for outcome measure 1.1, compared to only 34 States in 2000; 41 States submitted sufficient data in 2003 for outcome measure 2.1, compared to only 30 States in 2000.
| Outcome Measures | Performance Variables | ||
|---|---|---|---|
| Range | Mean percent | Median percent | |
| 1.1: Percent of child victims experiencing a recurrence of child maltreatment within a 6-month period (N=45 States) | 2.1-14.4% | 7.3% | 7.1% |
| 2.1: Percent of children in foster care who were victims of maltreatment by a foster parent or facility staff member (N=41 States) | 0.02-1.53 | 0.44 | 0.39 |
* The number of States identified for each measure reflects those that provided sufficient data to NCANDS to calculate the measure. Back |
|||
Figures II-1 and II-2 depict 2003 State performance on the safety-related measures. States not included in figure II-1 are Alabama, Alaska, Georgia, North Dakota, Oregon, Wisconsin, and Puerto Rico. States not included in figure II-2 are Alabama, Alaska, Connecticut, Georgia, Louisiana, New Hampshire, North Dakota, Oregon, Tennessee, Wisconsin, and Puerto Rico.
As shown in Figure II-1, States varied in the percent of children experiencing maltreatment recurrence within a 6-month period. At the low end, seven States had recurrence rates of 4.0 percent or less (Arizona, Colorado, Delaware, Pennsylvania, South Carolina, Tennessee, and Virginia). At the high end, eight States had recurrence rates of 10 percent or more (Connecticut, Iowa, Massachusetts, New York, Rhode Island, South Dakota, Washington, and West Virginia).
Figure II-2 indicates that States also varied with regard to the percent of children in foster care who were maltreated by foster parents or facility staff members in 2003, although the percentages are quite small for this measure. In five States, less than 0.10 percent of children in foster care were reported to NCANDS as victims of maltreatment by a foster parent or facility staff member (Maine, Maryland, Nevada, Vermont, and Wyoming). In four States, more than one percent of the children in foster care were reported to be victims of maltreatment by a foster parent or facility staff member (Hawaii, Massachusetts, New Mexico, and Rhode Island).
Change in State Performance from 2000 to 2003 on the Safety-Related Outcome Measures
Change in State performance on a particular outcome measure was assessed by calculating the percent change from 2000 to 2003.12 Using a percent change calculation permits an assessment of the extent of change occurring over time by taking into account the size of the percentages being compared. For example, a change from 6 percent in 2000 to 7 percent in 2003 represents a change of only 1 percentage point; however, it reflects a +16.7 percent change. In contrast, a change from 75 percent to 80 percent represents a change of 5 percentage points, but only a +6.7 percent change. For purposes of the analyses presented in this chapter, if the percent change in performance from 2000 to 2003 was less than 5.0 in either direction (i.e., positive or negative), a determination was made that there was "no change" in performance. Because the outcome measures are assessed using percentages, small fluctuations across years are to be expected. Rather than identifying any percent change as a performance change, it was decided that a percent change greater than 5.0 would be a practical indicator of actual performance change.
Table II-2 presents key findings regarding the change in State performance from 2000 to 2003 on measures pertaining to children's safety. The analyses included only those States that provided data for the measures in both years. Because the outcomes assessed focus on the reduction of recurrence of child abuse and/or neglect (outcome 1) and the reduction of the incidence of child abuse and/or neglect in foster care (outcome 2), a negative percent change represents an improvement in performance while a positive percent change represents a performance decline.
| Outcome Measures | Change in Performance from 2000 to 2003 | ||
|---|---|---|---|
| Improved performance (number and percent of States) | Declined performance (number and percent of States) | No change in performance (number and percent of States) | |
| Measure 1.1: Percent of child victims experiencing a recurrence of child maltreatment within a 6-month period (N=34 States) | 18 (53%) | 10 (29%) | 6 (18%) |
| Measure 2.1: Percent of children in foster care who were victims of maltreatment by a foster parent or facility staff member (N=26 States) | 19 (73%) | 5 (19%) | 2 (8%) |
* The number of States for each measure (N) includes those States that submitted sufficient data to calculate the measure in both 2000 and 2003. Back |
|||
Figures II-3 and II-4 depict the extent of change of individual States in performance on these safety-related measures. States excluded from Figure II-3 are Alabama, Alaska, Colorado, District of Columbia, Georgia, Idaho, Maryland, Mississippi, Nevada, North Dakota, Ohio, Oregon, Puerto Rico, South Carolina, South Dakota, Tennessee, Virginia, and Wisconsin. States excluded from figure II-4 are Alabama, Alaska, California, Colorado, Connecticut, Delaware, District of Columbia, Georgia, Idaho, Louisiana, Maryland, Mississippi, Nevada, New Hampshire, New Mexico, North Dakota, Ohio, Oregon, Puerto Rico, South Carolina, South Dakota, Tennessee, Virginia, Washington, West Virginia, and Wisconsin.
As shown in Figure II-3, four States had more than a -25.0 percent change on the measure of maltreatment recurrence, reflecting a substantial performance improvement (Arizona, Montana, New Hampshire, and Vermont). In contrast, four States had at least a +40.0 percent change, reflecting a considerable decline in performance (Maine, Michigan, Missouri, and West Virginia).
Similar results were found for the measure of maltreatment of children in foster care. As shown in figure II-4, there was a substantial improvement in performance in four States (Arizona, New Jersey, Vermont, and Wyoming) and a substantial decline in performance in four States (Arkansas, Florida, Minnesota, and Nebraska). However, because most of the percentages for this measure are less than 1.0, differences in percentage points that are quite small can result in a fairly large percent change.
Relationships Between State Performance on Safety-Related Outcome Measures and the Number and Characteristics of Child Victims
Analyses were conducted to examine the relationships between State performance in 2003 on the safety-related outcome measures and the following:
- Child maltreatment victim rates
- Age of child victims
- Race/ethnicity of child victims
- Types of maltreatment experienced by child victims
Child maltreatment victim rates
A State's child maltreatment victim rate is the number of child victims per 1,000 children in the State's population, with "victim" defined as a child reported to NCANDS who is the subject of a substantiated or indicated maltreatment allegation.13 In 2003, child victim rates varied considerably across States, from 1.6 in Pennsylvania14 to 42.2 in Alaska, with a median of 10.6. In seven States, there were fewer than five child victims per 1,000 children in the population (Arizona, Idaho, New Hampshire, New Jersey, Pennsylvania, Virginia, and Washington). In contrast, in six States, there were more than 20 child victims per 1,000 children in the population (Alaska, District of Columbia, Florida, Massachusetts, South Dakota, and West Virginia); in two of these six States (Alaska and Florida), there were more than 35 child victims per 1,000 children in the population.
A relationship was found between State victim rates in 2003 and the percent of maltreatment recurrence within a 6-month period in 2003 (Pearson's r = 0.63). As shown in Figure II-5, States with a relatively high percent of maltreatment recurrence within a 6-month period also tended to have a relatively high child victim rate. States not included in figure II-5 are Alabama, Alaska, Georgia, North Dakota, Oregon, Puerto Rico, and Wisconsin.
An additional finding was that 23 (82 percent) of the 28 States that exhibited a change (either positive or negative) from 2000 to 2003 in performance on the measure of maltreatment recurrence also exhibited a change in the child victim rate in the same direction. That is, if the maltreatment recurrence percentage increased, the victim rate also increased and vice versa. The relationship between victim rates and maltreatment recurrence is discussed further in chapter V.
Child maltreatment victim characteristics
Table II-3 presents information pertaining to the age, type of maltreatment experienced, and race/ethnicity of child maltreatment victims. As shown in the table, the characteristics of child victims varied across States.
| Characteristics of Child Victims | Percent range across States | Median percent across States |
|---|---|---|
| Age of children at time of maltreatment report | ||
| Not yet 1 year old | 5.4 - 16.0 % | 9.7 % |
| At least 1 but not yet 6 years old | 20.2 - 37.2 | 30.9 |
| At least 6 years old, but not yet 11 years old | 24.1 - 31.5 | 27.6 |
| At least 11 years old, but not yet 16 years old | 18.3 - 37.6 | 24.7 |
| At least 16 years old | 3.2 - 12.1 | 5.3 |
| Type of maltreatment experienced* | ||
| Neglect (not medical neglect) | 3.9 - 90.4 | 64.8 |
| Physical abuse | 3.1 - 59.1 | 20.8 |
| Psychological/emotional abuse | 0 - 58.0 | 2.5 |
| Sexual abuse | 3.0 - 57.2 | 9.3 |
| Medical neglect | 0 - 12.5 | 2.0 |
| Other | 0 - 88.2 | 0.1 |
| Children's race/ethnicity | ||
| Alaska Native/American Indian | <0.1 - 46.8 | 0.3 |
| Asian | <0.1 - 11.5 | 0.5 |
| Native Hawaiian/Other Pacific Islander | <0.1 - 19.8 | 0.1 |
| Black (non-Hispanic) | 0.8 - 55.9 | 15.5 |
| Hispanic (of any race) | 0.3 - 46.6 | 4.7 |
| White (non-Hispanic) | 0.4 - 94.6 | 53.2 |
| Multiple Races | <0.1 - 26.2 | 1.2 |
* Variations in types of maltreatment experienced may be attributed to differences in State definitions of maltreatment. For example, in some States but not others, psychological or emotional abuse is included in the State definition of child abuse and neglect. Back |
||
The following relationships were found between State variation in the characteristics of child victims and State performance on the safety-related outcome measures:
- States with a relatively high percentage of children who were victims of sexual abuse tended to have a relatively low percentage of maltreatment recurrence within a 6-month period and vice versa (Pearson's r = -0.40).
- States with a relatively high percentage of children who were victims of neglect tended to have a relatively high percentage of maltreatment recurrence within a 6-month period and vice versa (Pearson's r = 0.40).
No relationships were found between State performance on the safety-related outcome measures and State variation with regard to the age of child victims. Also, no relationships were found between State performance on the safety-related outcome measures and State variation with regard to the race/ethnicity of child victims. However, in many States, the percentage of children of color in the child victim population was considerably higher than their percentage in the State's child population. The differences were particularly large for Black children and Alaska Native/American Indian children. Specific information regarding these differences is presented in table II-4.
| Relationship** | Race/Ethnicity of Children | |||
|---|---|---|---|---|
| Hispanic | White | Black | Alaska Native/ American Indian |
|
| Number and percent of States in which the percent of child victims exceeded the percent of children in the State's population | 9 (18%) | 0 | 41 (82%) | 21 (42%) |
| Number and percent of States in which the percent of child victims was about equal to the percent of children in the State's population | 27 (54%) | 40 (80%) | 9 (18%) | 10 (20%) |
| Number and percent of States in which the percent of child victims was less than the percent of children in the State's population | 14 (28%) | 10 (20%) | 0 | 19 (38%) |
| Total | 50 (100%) | 50 (100%) | 50 (100%) | 50 (100%) |
|
* Data are not available for Puerto Rico or Pennsylvania. Pennsylvania is not included because State law does not permit reporting of the race/ethnicity of child victims. Puerto Rico does not submit these data to NCANDS. Back ** The following approach was used to determine the nature of the relationship: A determination of "Exceeded" was made if the percent of child victims of a particular race/ethnicity was at least 1.3 times the percent in the State's child population. A determination of "less-than" was made if the percent of child victims of a particular race/ethnicity was less than 0.7 times the percent in the State's child population. Back |
||||
The following findings indicate that the over-representation of a particular race/ethnicity in the child victim population in 2003 was quite large in some States:
- In Connecticut, Massachusetts, Minnesota, North Dakota, and South Dakota the percent of child victims who were Hispanic was at least twice the percent of Hispanic children in the population.
- In Minnesota, Nebraska, Oregon, and Washington, the percent of child victims who were Native American was more than four times the percent of Native American children in the population.
- In Iowa, Minnesota, North Dakota, Utah, and Wyoming, the percent of Black child victims was more than three times the percent of Black children in the State population.
It has been a consistent finding in the Reports to Congress and in the research field that Black children and Alaska Native/American Indian children are over-represented in the child victim population in many States and White children are under-represented in the child victim populations in all States. The causes and implications of both the over-representation and the under-representation are not clearly understood at this time.
10 The Child File of the National Child Abuse and Neglect Data System cannot track children from one year to the next. Therefore, a 6-month time frame for maltreatment recurrence was established to ensure that the measure captured the actual incidence of maltreatment recurrence rather than estimating the incidence. In addition, multiple research findings indicate that the highest incidence of maltreatment recurrence tends to occur within 6 months of a prior incident. Therefore, the Department determined that the 6-month time frame was sufficient to assess State performance in this area. Back
11 NCANDS data were collected and reported on a calendar year basis through 2002. Beginning in 2003, they were reported on a Federal Fiscal Year basis (October 1 through September 30) to be consistent with AFCARS data. Back
12 Percent change is calculated by subtracting "old" data from "new" data, dividing by "old" data and multiplying by 100. For example, if maltreatment recurrence was 9.2 in 2000 and 7.6 in 2003, the percent change would be [(7.6-9.2)/9.2] x 100 = -17.4 percent change. Back
13 A child who is the subject of more than one substantiated or indicated maltreatment allegation may be counted more than once. Back
14 Pennsylvania's low victim rate is due to the fact that reports of child neglect (unless determined to be "severe") are not handled by the child protective services system. Therefore, the State does not report child neglect to NCANDS unless it is severe neglect. Back
| << Chapter I | Chapter III >> |
