[ Main Page of Report | Contents of Report ]
Successful service delivery requires a good understanding of one's client population and interventions tailored to those clients' needs. With respect to services that prevent or ameliorate child maltreatment, agencies serve children in the context of their families. But in practice, services are most often delivered to mothers. Men are noticeably absent in most social service agencies' waiting rooms, parenting classes, and family group decision meetings. Yet more than 40 percent of child maltreatment perpetrators were men, according to national data for 2002 (U.S. Department of Health and Human Services, 2004). The intent of the analyses presented below was to look more closely at male perpetrators of child maltreatment in order to better understand their characteristics and patterns of maltreatment. By understanding more clearly who these men are and whether they are similar to or different from traditional child welfare clients (i.e. mothers), we may better design prevention and intervention approaches that meet their needs and protect their children and others in their care.
This paper used a unique multi-State data set of nearly 200,000 child maltreatment perpetrators to better understand the characteristics of male perpetrators, their maltreatment patterns, and the outcomes associated with their maltreatments. Furthermore, the influence of a mother coperpetrator on the circumstances surrounding the child maltreatment or the outcomes was also investigated.
Until the past decade, child welfare research primarily focused on mothers of maltreated children rather than fathers. For example, a 1990 review of major social work journals published in the preceding 27-year period (814 issues) elicited only 21 articles focusing on fathers (Greif & Bailey, 1990).
Much of the recent research investigating the role of fathers and other males as perpetrators of child maltreatment has focused on circumstances related to specific types of maltreatment and relatively small samples or clinical populations. For example, Phelan (1995), Manion, McIntyre, Firestone, Ligezinska, Ensome, & Wells (1996), Manion, Firestone, Cloutier, Ligezinska, McIntyre, & Ensom (1998), and Rudin, Zalewski, & Bodmer-Turner (1995) focused on sexual abuse; Bagley & Pritchard (2000), and Klevens, Bayon, & Sierra (2000) focused on physical abuse; and Brewster, Nelson, Hymel, Colby, Lucas, McCanne, et al. (1998), Smithey (1998), and Adinkrah (2003) focused on fatalities.
Some research shows that when we take issues of severity into consideration, fathers or father surrogates (cohabiting husbands or boyfriends who are not biologically related to the child) are responsible for more severe physical abuse and fatalities than women perpetrators (Brewster et al., 1998; Klevens et al., 2000; U.S. Advisory Board on Child Abuse and Neglect, 1995). Daly & Wilson (1999) have argued that parents are less likely than surrogate parents to physically abuse or seriously injure their biological offspring due to their greater investment in the genetic continuity of their family. In a longitudinal analysis of a cohort of children at risk for child maltreatment, Radhakrishna, Bou-Saada, Hunter, Catellier, & Kotch (2001) demonstrated that the presence of a father surrogate in the home increased the risk of a maltreatment report to more than twice that of families with both biological parents in the home.
The research on male perpetrators is complicated by inconsistent or imprecise definitions of perpetrators and classifications of male perpetrators. Persons found by a child welfare agency to be responsible for the maltreatment of a child may have committed an act of maltreatment, allowed an act of maltreatment to occur, or contributed to the risk of maltreatment. Furthermore, there is no consistent typology for male perpetrators. Some studies of male perpetrators do not distinguish between biological fathers and stepfathers in their analyses (Smith & Saunders, 1995), or they do not consistently group fathers in the various types of father relationships. Manion et al. (1996) cited Canadian national data indicating the percentage of sexual abuse cases that were incestuous (18%), involved strangers (18%), or involved other friends and acquaintances (18%). Bagley & Pritchard (2000), in their examination of a cohort of convicted child sexual abuse perpetrators, distinguished among biological relatives, consisting of birth fathers and other biologically related males such as uncles and grandfathers (18%), nonbiological relatives consisting of stepfathers and cohabiters (10%), and extrafamilial offenders, who had no family link with their victim (72%). Dubowitz et al. (2001) distinguished between fathers and other father figures, which included stepfathers, mother's boyfriends, and other male relatives, but recognized that these relationships were based on the child's perception, and were not based on other parental or adult information.
Recent research has begun to examine the specific benefits of fathers to child well-being. The economic and social stressors to which a single mother is exposed put her at risk for maltreating her children. Child Trends (2002) reported that children are more likely to be abused by mothers in single-parent families than in two-parent families. Children raised in two-parent families have been shown to have better school performance, superior peer relationships, and fewer behavioral problems than children living in single parent families (Lamb, 2001). Child Trends (2002) documented a large number of indicators to better understand the role that fathers play in the lives of their children. These indicators demonstrated that a considerable percentage of fathers were highly engaged in a variety of roles with their children, including play, discipline, and primary caregiving.
Dubowitz, Black, Cox, Kerr, Litrownik, Radhakrishna, et al. (2001) concluded that the presence of a father or father figure, regardless of whether he lived in the same home, was associated with better cognitive development and greater perceived competence among 6-year-olds when coupled with the child's perception of father support. Researchers have shown that even nonresidential fathers can contribute to positive child outcomes when they maintain an active involvement in their children's lives (Lamb, 2001; Federal Interagency Forum on Child and Family Statistics, 2003). Results of a community survey by Nobes & Smith (2002) that found that children in two-parent families were punished significantly more frequently than children in single mother households, raising the possibility of an association with maltreatment. The general direction of current research, however, has increased interest in supporting and strengthening the relationship of fathers and their children.
The objective of this study was to better understand the characteristics of male perpetrators of child maltreatment as well as how they and their patterns of maltreatment vary among subgroups of men and compare to those of women. This research takes advantage of a large data set of child maltreatment investigations reported by States to the Federal Government through the National Child Abuse and Neglect Data System (NCANDS). The study classifies subgroups of male perpetrators according to relationship to the victim, and examines how investigation outcomes differ depending on whether the male perpetrator was found to have acted alone or in conjunction with the victim's mother.
The key research questions for this study were the following:
[ Go to Contents ]
Case-level data from the 2002 National Child Abuse and Neglect Data System (NCANDS) were used as the basis for the analysis. Only cases of substantiated or indicated maltreatment were included in the data set. These data are submitted on a voluntary basis in a common record format to the Federal Government by State CPS agencies. The submissions are a rich source of information about children who are the subjects of child maltreatment investigations, including data about the investigations, child demographics, and types of maltreatment, perpetrators, and services. Each year the U.S. Department of Health and Human Services prepares an annual report that analyzes the NCANDS data. For calendar year 2002, Child File data submissions from 42 States containing case-specific information on close to 3 million reported children were submitted to NCANDS. Data from 18 States, including 202,376 unique perpetrators, were used to create the data set for this research.(1) These States were chosen for the completeness and validity of their data on the relationship of the perpetrators to the victims. The population in these 18 States was comparable to the national population on a range of demographic characteristics including age, race distribution, and poverty level. (See table 1.) The findings in this study are not necessarily representative of all reporting States or the entire nation as there may be other differences that are not reported or observed. For some analyses, fewer States were included if some States did not provide valid data on the variables of interest.
|Demographic Indicators||All States||18 States|
|% of U.S. Population||100%||35%|
|Population under 18 years: Total||26%||26%|
|Families||Percent of total population living in family settings||80%||81%|
|Married-couple family, with own children under 18 years old (percent of families)||35%||36%|
|Families with female householder, no husband present,
with own children under 18 years old
(percent of families)
|Average Poverty Rate||12.2%||12.0%|
|Race||White (not Hispanic), percent of total population||69%||73%|
|Black or African American (not Hispanic), percent of total population||12%||11%|
|American Indian and Alaska Native (not Hispanic), percent of total population||1%||1%|
|Asian / Pacific Islander (not Hispanic), percent of total population||4%||2%|
|Other or Multiple Race (not Hispanic), percent of total population||2%||1%|
|Hispanic or Latino, percent of total population||13%||12%|
|Sex||Male, percent of total population||49%||49%|
|Female, percent of total population||51%||51%|
|Child Abuse and Neglect||Rate of children investigated per 1,000 children in population||43.8||39.4|
|Rate of victims of child maltreatment per 1,000 children in population||12.3||10.9|
Consistent with the research questions, the analysis of the data depended on developing a data set of unique perpetrators. Despite the relative simplicity of obtaining unduplicated perpetrator data since encrypted unique identifiers are presented in the data set, the design for this research depended on examining complex event and relationship data associated with each perpetrator. For instance, a perpetrator may have victimized more than one child, in one or more report events. In addition, the perpetrator may have been related to one child as a biological parent and to another as a stepparent. Similarly, the perpetrator may have been associated with different or similar types of maltreatment with each child, and may have acted alone or in concert with other perpetrators each time or only sometimes.
Data on all reports, children, and maltreatments were merged and recoded to represent the categories of reports, children, and maltreatments associated with each unique perpetrator.
Figure 1 provides a graphical overview of the stages of data construction.
Construction of the Unique Perpetrator Data Set
Stage 1 refers to the NCANDS case-level Child File. The NCANDS Child File establishes a record for each child in a report alleging child maltreatment. This record entity is called a report-child pair. Data on up to three perpetrators and the types of maltreatment associated with each can be included in each report-child pair record. Data on perpetrators are collected only if the child has been found to be a victim of maltreatment.
Stage 2 refers to creating records for each perpetrator in each report-child pair record. In other words, if a report-child pair record had three perpetrators, then three report-child-perpetrator triad records were created.
Stage 3 refers to aggregating data for each perpetrator to create a new file with all unduplicated male perpetrators. During Stage 3, several additional variables were derived:
Stage 4 included additional screening of data to exclude certain categories of missing data.
The final data set was composed of 192,321 unduplicated perpetrators based on data from 18 States.
While the advantages of this large, multi-State data set are many, some limitations to these data should be noted as well.
The designation of the individuals in this data set as perpetrators was made by CPS representatives during the disposition of a report or reports of child maltreatment. Inevitably, some individuals with similar circumstances or relationships may have been inconsistently recorded as perpetrators across agencies, counties, or States. For example, a boyfriend living with a mother who was found to be neglecting her children may have been included as a coperpetrator in one CPS agency, but not in a different agency. While efforts were made to reduce these inconsistencies in the data set, some undoubtedly remain.
Furthermore, the NCANDS data did not include whether the perpetrators were living in the same home as the victims. Similarly, the relationship between coperpetrators was not captured. A child may have been maltreated by both his biological mother and biological father, but this data set does not indicate whether these parents were married to each other or whether they lived in the same household.
It is also important to note that some categories of male perpetrators that are discussed are much smaller than others. For example, adoptive fathers account for only 1 percent of all male perpetrators, but are discussed along with other groups that are much larger. The implications of findings related to these small groups should be considered carefully.
[ Go to Contents ]
The findings are discussed in terms of the overall research questions. First, a general description of male perpetrators is presented. Second, the categories of primary interest for this research biological fathers and father surrogates including adoptive fathers, stepfathers, and mothers' boyfriends are compared with male perpetrators who were not parents. Finally, comparisons of fathers, father surrogates, and male nonparents acting alone, with the mother, and with others are described. Supporting data tables for the analyses presented in this section are included in appendix A.
|Nearly one-half of all perpetrators were male,
and of these, one-half were biological fathers.
Of the 192,321 unique perpetrators in the data set, 89,028 (46%) were male and 103,293 (54%) were female. Figure 2 shows male perpetrators by their relationship to their victims. More than one-half of all male perpetrators (51%) were biological fathers. The second largest group was male nonparents (26%), who included male relatives (12%), male nonrelatives (13%), and those with a combination of nonparental relationships (1%). Boyfriends, stepfathers, combination fathers, and adoptive fathers accounted for 10 percent, 8 percent, 5 percent, and 1 percent of all male perpetrators, respectively. Among female perpetrators, 86 percent were biological mothers, 10 percent were nonparents, and the remaining 4 percent were stepmothers, adoptive mothers, fathers' girlfriends, or combination mothers.
(n = 89,028)
Age of Male Perpetrators
|Male perpetrators were older than femal perpetrators.|
Male perpetrators were 5 years older, on average, than female perpetrators. The average age of male perpetrators was 38.6 years, while among females the average age was 33.4 years. (See figure 3.) This coincides with the typical age disparity between married partners. According to U.S. Census data, among married couples with children, husbands are on average 2.4 years older than their wives.(3)
Age of Male and Female Perpetrators(4)
(n = 192,321)
Age distributions were also examined by male perpetrator relationship. (See table 2.) The average age for biological fathers was 37.7 years, for stepfathers 38.9 years, for adoptive fathers 47.0 years, for mothers' boyfriends 39.3 years, for combination fathers 34.8 years, and for male nonparents 40.8 years.
|Age of Perpetrator||Biological Father||Adoptive Father||Stepfather||Mother's Boyfriend||Combination Father||Nonparent||Total|
|Age 20 or Younger||1,034||0||50||303||89||7,396||8,872|
|Age 61 or Older||318||37||75||46||26||1,156||1,658|
|Male and female perpetrators
did not differ in terms of race.
Male and female perpetrators did not differ in terms of race. More than one-half of male perpetrators were White (58%), as were female perpetrators (57%). Sixteen percent of male perpetrators were African American; 21 percent of female perpetrators were African American. Among both male and female perpetrators, 13 percent were Hispanic. When the race of male perpetrators was examined for the different father categories, all had similar racial distributions. Among all father categories, the largest racial group was White and the second largest group was African American.(5)
Caregiver Status of Perpetrator
|Female perpetrators were more
likely to be caregivers to their
victims than male perpetrators.
In 10 of the 18 States, data were provided on whether the perpetrator was in a caregiver role with the victim at the time of the maltreatment. If the perpetrator was in a caregiving role with any of the associated victims, he or she was considered to be a caretaker. Female perpetrators were much more likely to be in a caregiving role than were male perpetrators; 87 percent of females were caregivers compared with 60 percent of males.
Biological or legal fathers were more likely to be designated as being in a caregiving relationship to the maltreated child than were nonparent males.(6) Perpetrators in most of the father categories were almost as likely as female perpetrators to be in caregiver roles. The highest percentages of caregivers were among combination fathers (86%), stepfathers (82%), biological fathers (81%), and adoptive fathers (72%). Among mothers' boyfriends, only 43 percent were caregivers. Among male nonparents, the percentage of those in a caregiving role was 15 percent.
These analyses address the second key research question are male perpetrators associated with any specific patterns of maltreatment? The basic characteristics of the child victims associated with male perpetrators in each of the six male perpetrator categories (biological father, stepfather, adoptive father, mother's boyfriend, combination father, and nonparent), as well as the type of maltreatment and the total number of child victims, are examined.
Number of Child Victims
|Both male and female
perpetrators were most likely
to be associated with only one
The majority of both male and female perpetrators were associated with only one child victim 67 percent of males and 61 percent of females. Biological fathers were most likely of all the male perpetrators to be associated with more than one child.(7) Among biological fathers, 35 percent were associated with more than one child. Among mothers' boyfriends, 31 percent were associated with more than one child. Among stepfathers, 22 percent were associated with more than one child. Among adoptive fathers, only 15 percent were associated with more than one child. Among nonparents, 18 percent were associated with more than one child. All combination fathers were associated with more than one child, because by definition they have more than one relationship with more than one child.
Age of Child Victims
|Male perpetrators were
associated more often
with older victims.
When compared with female perpetrators, male perpetrators were involved with more children older than age 8 and fewer children younger than age 1. Twenty-one percent of females were associated with child victims under age 1, while only 11 percent of male perpetrators were associated with infant victims; and 29 percent of male perpetrators were associated with victims between age 12 and 15, compared with 22 percent of females.
The different groups of male perpetrators varied in their associations with children in each age group. (See figure 4.) Biological fathers, combination fathers, and mothers' boyfriends were most frequently associated with infants and children under age 3. Stepfathers, adoptive fathers, and nonparents were most frequently associated with older children and adolescents.
Distribution of Perpetrators by Age of Victims(8,9)
(n = 89,028)
Sex of Child Victims
|Stepfathers, adoptive fathers, and
nonparents maltreated girls more
often than other male perpetrators.
Perpetrators were categorized as having been associated with girls, boys, or both boys and girls. Male perpetrators were associated more often only with female victims, while female perpetrators were associated with male and female victims in almost equal numbers. However, this pattern does not hold true for all categories of male perpetrators. More than one-half of the perpetrators who were stepfathers or adoptive fathers maltreated girls. The majority of combination fathers were associated with both male and female victims; this is related to the fact that all these perpetrators were associated with multiple victims. Among biological fathers and mothers' boyfriends, the proportions associated with boys and with girls were more evenly distributed. Nonparents had the highest proportion of perpetrators (65%) who maltreated girls. (See figure 5.)
Male Perpetrators by Sex of Victim(10)
(n = 89,028)
Perpetrators Acting Alone or with a Partner
|More than one-half of male perpetrators
of child maltreatment always acted alone.
Some perpetrators acted alone, and some acted with another person or people. (See appendix A, table A-1.) Sometimes, the other person was the mother of the victim, sometimes he or she was a different person. The majority (59%) of male perpetrators always acted alone; 34 percent acted at least once with the mother of the victim; and 6 percent acted at least once with other(s) but never acted with the mother.
Among biological fathers, 55 percent always acted alone, while 41 percent were associated with the mother at least once, and 4 percent acted with another person. A slightly larger percentage of stepfathers (64%) and adoptive fathers (60%) always acted alone. Among nonparent male perpetrators, 73 percent always acted alone, only 12 percent acted at least once with the mother, and 14 percent acted with another person or people. Thus more than one-half of these male perpetrators were perpetrators of maltreatment without the involvement of either the victim's mother or another person as a perpetrator.(11)
Type of Maltreatment
|Male perpetrators physically or sexually abused
their victims more and neglected their victims less
than did female perpetrators.
Male and female perpetrators had distinct patterns of maltreatment. More than one-third of male perpetrators (36%) were associated with neglect; 66 percent of female perpetrators were associated with neglect. (See figure 6.) A quarter of males (26%) were associated with sexual abuse, while 2 percent of female perpetrators were associated with sexual abuse. The percentages associated with physical abuse were comparable, although slightly higher for males (22%) than for females (18%).
Maltreatment by Male and Female Perpetrators(12)
(n = 192,321)
Within male perpetrator categories there were also distinct patterns in the type of maltreatment. (See table 3.) One-half of biological fathers were associated with neglect. Adoptive fathers and stepfathers had distributions similar to each other, approximately one-quarter to one-third of each group was associated with one of the main types of maltreatment physical abuse, neglect, or sexual abuse.
|Type of Maltreatment||Biological Father||Adoptive Father||Stepfather||Mother's Boyfriend||Combination Father||Nonparent||Total|
|Physical Abuse Only||11,920||157||2,331||2,629||388||2,544||19,969|
|Neglect or Medical Neglect Only||22,580||111||1,345||2,662||2,596||2,819||32,113|
|Sexual Abuse Only||3,226||114||2,064||1,712||339||15,579||23,034|
|Other or Emotional Abuse Only||2,971||15||286||498||124||379||4,273|
For both adoptive fathers and stepfathers, physical abuse was the most frequent type. Nonparental perpetrators were primarily associated with sexual abuse (68%). Biological fathers had the smallest percentage of sexual abuse cases (7%) compared to between 20 percent and 30 percent for boyfriends, adoptive fathers, and stepfathers. The pattern for combination fathers was similar to biological fathers, although with higher proportions being associated with neglect (56%), and with multiple types of maltreatment (25%).
Of great interest is the examination of risk factors associated with perpetrators. The analyses of these data are complicated both by the data collection process as well as the data submission process in NCANDS. Information on such risk factors is largely dependent upon the assessments that are conducted by the CPS agency during investigations and are therefore likely to be underreported. Additionally, since risk factors are specifically associated with each child in a report rather than each perpetrator, the risk factors discussed below would be associated with all perpetrators in a specific report-child pair record.
|Mothers' boyfriends who were caregivers were
associated with alcohol and drug use more than
were other types of male perpetrators.
They also had a higher incidence of
domestic violence between adults.
Of the States that identified whether the perpetrator was a caregiver or not, seven States were also able to indicate whether the caregivers in the report had drug or alcohol risk factors.(13) Because the data on these risk factors were associated with caregivers, only perpetrators who were caregivers were included in these analyses. Sixteen percent of all male perpetrators who were caregivers were associated with these risk factors, compared with 20 percent of all female perpetrators who were caregivers. Boyfriends who were caregivers were more often associated with the risk factors of alcohol, drug, or drug and alcohol (25%) than were the other types of male perpetrators. (See figure 7.) Biological fathers and stepfathers, who were caregivers, had the next highest percentages of involvement with drugs and alcohol (18% and 14% respectively).
Drug and Alcohol Abuse Among Male Caregiver Perpetrators(14)
(n = 24,085)
Eight States provided data on the presence of domestic violence in the families of maltreatment victims. Combination fathers who were caregivers were most often associated with the risk factors of family violence (35%), followed by boyfriends and stepfathers (23% and 20%, respectively). (See figure 8.) Biological fathers who were caregivers had the next highest percentages of involvement with family violence (18%).
Family Violence Among Male Caregiver Perpetrators(15)
(n = 14,410), Data from 8 States
These analyses address the third key research question are there different outcomes associated with different types of male perpetrators? Perpetrators were compared in terms of services and recidivism.
|Postinvestigation services were provided
more often with female perpetrators
than with male perpetrators.
Nearly one-half of all male perpetrators were associated with at least one investigation after which additional services were provided and recorded by the child welfare agency (47%). A larger percentage of female perpetrators (57%) received services. Approximately 16 percent of male perpetrators were associated with at least one victim who was placed in foster care as a result of the investigation compared to 24 percent of female perpetrators.
The proportions of male perpetrators who were associated with in-home services were quite comparable for all categories except combination fathers and nonparents. Approximately one-third of most categories of male perpetrators were associated with post-investigation in-home services. Among combination fathers 43 percent had in-home services provided and among nonparents only 24 percent had in-home services provided. The percentages of perpetrators associated with at least one child placed in foster care ranged from 10 percent (nonparents) to 29 percent (adoptive fathers). (See figure 9.) Approximately one-half of all perpetrators who were biological fathers, stepfathers, or mothers' boyfriends did not receive any services. The percentage that did not receive any services was lower among adoptive fathers (40%) and higher among nonparents (66%).
Services Received by Categories of Male Perpetrators(16)
(n = 89,028)
|Male perpetrators had a lower rate of recidivism
than female perpetrators.
Stepfathers and adoptive fathers rate
of recidivism was lower than
most other male perpetrators.
Survival analysis techniques were used to estimate the proportion of perpetrators who had a second finding of having maltreated a child within 12 months of the first finding in the reporting period.(17) Overall, male perpetrators had a lower recidivist rate, 6 percent at 6 months and 9 percent at 12 months, compared with female perpetrators (8% and 12% respectively).
Within 12 months, it was projected that 7 percent of biological fathers were associated with a second maltreated child, while approximately 4 percent of adoptive and stepfathers, 7 percent of boyfriends, and 8 percent of nonparents experienced an additional report of maltreatment in the same year. (See figure 10.) In addition, combination fathers experienced a 38 percent recidivism rate after 12 months. This is due to the fact that these perpetrators were necessarily associated with multiple children and frequently multiple reports.(18)
Recidivism of Categories of Male Perpetrators(19)
(n = 89,028)
Among male perpetrators who were associated with postinvestigation services (either foster care or in-home services), the proportion recidivating in 1 year was twice the proportion of males who were not associated with services 12 percent compared to 6 percent. This finding is consistent with findings on female perpetrators in that 20 percent of females associated with services recidivate within 12 months compared with 8 percent who were not associated with services. Perpetrators receiving services may have an intrinsically higher risk of recidivism that cannot be fully addressed by services, or the increased recidivism may possibly be due to increased attention and surveillance by CPS. This finding regarding services and perpetrator recidivism is also consistent with studies of child maltreatment recurrence (Fluke, Yuan, & Edwards et al,. 1999; DePanfilis & Zuravin, 1999a; U.S. Department of Health and Human Services, 2004).
Perpetrators could either act alone or in concert with others. As noted above, male perpetrators were classified by whether they always acted alone or they were responsible for the maltreatment along with the victim's mother (biological mother, stepmother or adoptive mother) in at least one victimization event. Male perpetrators who acted only with a person or persons other than the victim's mother were not included in these analyses.(20) Of a total of 79,031 male perpetrators, 65 percent only acted alone, and 35 percent were associated with at least one victim's mother on at least one occasion. To simplify the comparisons for these analyses, a father surrogate category was created to combine adoptive fathers, stepfathers, and mothers' boyfriends. The following categories were compared: biological father acting alone; biological father acting with mother; biological father acting with other person or people; father surrogate acting alone; father surrogate acting with mother; father surrogate acting with other person or people; male nonparent acting alone; male nonparent acting with mother; and male nonparent acting with other person or people. Fifty-five percent of biological fathers always acted alone; and 58 percent of father surrogates (adoptive fathers, stepfathers, and mothers' boyfriends) and 73 percent of male nonparents acted alone.
Age of Child Victims
|Biological fathers acting with mothers
were associated with maltreatment of infants
more often than were those acting alone.
Male perpetrators showed different patterns in the age of their child victims. (See table 4.) Among biological fathers who acted with mothers, 25 percent were associated with children under age 1, but among those who acted alone only 9 percent were associated with infant victims. Among other groups of male perpetrators, very few were associated with the youngest victims regardless of whether they acted with the mother or alone. Biological fathers were associated with fewer preteen (age 8-11) and teenage (age 12-15) victims than were the other male perpetrators, regardless of whether they acted alone or with the mother.
|Age of Youngest Child||Acting Alone||Acting with Mother||Total|
|Biological Father||Father Surrogate||Nonparent||Total||Biological Father||Father Surrogate||Nonparent||Total|
|Younger than Age 1||2,320||229||199||2,748||4,720||433||215||5,368||8,116|
|Age 16 or Older||2,600||1,016||1,736||5,352||1,260||553||202||2,015||7,367|
Sex of Child Victims
Male perpetrators acting alone were consistently associated with maltreating girls more often compared with male perpetrators acting with mothers. Among biological fathers acting alone, 42 percent had maltreated girls; 55 percent of father surrogates acting alone and 68 percent of male nonparents acting alone were associated with girls. (See figure 11.) In contrast to all categories of male perpetrators acting alone, for the categories of male perpetrators acting with the victim's mother the percentage associated with girls was lower, and a higher percentage acted with both boys and girls.
Actions and Categories of Male Perpetrators by Sex of Victims(21)
(n = 79,031)
|Male perpetrators were more likely to maltreat girls
when acting alone than when acting with mothers.
Male perpetrators acting alone followed a very different pattern of maltreatment from those acting with the victim's mother. In all instances, male perpetrators acting alone were more likely to be associated with sexual abuse than if they acted with mothers. Also, biological father and father surrogate perpetrators were more likely to be associated with physical abuse and less likely to be associated with neglect if they acted alone. (See table 5.)
|Type of Maltreatment||Acting Alone||Acting with Mother||Total|
|Biological Father||Father Surrogate||Nonparent||Total||Biological Father||Father Surrogate||Nonparent||Total|
|Physical Abuse Only||9,982||3,854||1,947||15,783||1,686||1,172||294||3,152||18,935|
|Neglect or Medical Neglect Only||8,587||1,079||1,027||10,693||13,201||2,944||1,032||17,177||27,870|
|Sexual Abuse Only||2,698||3,198||13,055||18,951||340||575||700||1,615||20,566|
|Other or Emotional Abuse Only||2,001||484||234||2,718||874||297||64||1,235||3,953|
Male perpetrators who acted alone were less likely to receive services than males who acted with the victims' mothers. (See figure 12.) In all categories, male perpetrators who acted alone were less likely to have a child placed in foster care or to receive in-home services. This finding may be consistent with the types of maltreatment associated with male perpetrators acting alone; in sexual abuse cases, the perpetrator may be removed from the household so that the child is protected without the need for foster care.
Services Received by Perpetrators Acting Alone or in Concert with Mothers(22)
(n = 79,031)
|Among male perpetrators,
recidivism was highest among
biological fathers acting with mothers
and lowest among father surrogates acting alone.
Recidivism was projected to be highest among biological fathers acting with mothers (10%), and lowest among father surrogates acting alone (4%). (See figure 13.) Five percent of biological fathers alone were recidivists compared with 10 percent of biological fathers acting with mothers. Similarly, among father surrogates, 4 percent of those acting alone were again perpetrators within 1 year, compared with 8 percent of those acting with mothers. Among nonparents, the percent of recidivists was approximately the same for both groups, although slightly higher for those acting alone (7%) than for those acting with the mother (6%).
Recidivism of Perpetrators Acting Alone or in Concert with Others(23)
[ Go to Contents ]
The distribution and characteristics of male perpetrators are among the least studied aspects of child maltreatment. Nonetheless, generalizations regarding this group are common. The use of NCANDS data to explore information about this group of perpetrators has revealed a clearer picture from a multi-State perspective.
This study confirms earlier findings (U.S. Department of Health and Human Services, 2004) that females outnumbered male perpetrators among cases investigated by CPS, by a margin of 10 percent. Males were slightly older than females, but otherwise no other basic demographic differences between male and female perpetrators were observed.
More than one-half of the male perpetrators were biological fathers, nearly one-quarter occupied some other parental role (adoptive fathers, stepfather, mothers' boyfriends), and an additional one-quarter were in nonparental relationships (including relatives, foster parents, daycare providers, or friends) to their victims. With respect to basic demographic descriptors, male perpetrators with different relationships to their victims varied to only a limited degree in their age and race.
In contrast, the categories of male perpetrators varied by the characteristics and experiences of their victims. Biological fathers were associated with the youngest victims compared with other male perpetrators, and among these groups, stepfathers and adoptive fathers were associated with relatively higher percentages of preteen or teenage victims. Further, nonparent perpetrators had the largest proportion of female victims, and more than one-half of the stepfather and adoptive father perpetrators had exclusively female victims. In contrast, fewer than one-half of the perpetrators who were biological fathers and mothers' boyfriends were associated with only female victims.
The study also confirmed NCANDS findings (U.S. Department of Health and Human Services, 2004) that patterns of sexual abuse were more common among male perpetrators, whereas neglect was more common among female perpetrators. About one-fifth of perpetrators, both male and female, physically abused their victims. The maltreatment pattern for biological fathers was similar to the overall female pattern that is, the majority was associated only with neglect, and less than 10 percent were sexually abusive, although about one-quarter were physically abusive. In contrast, about one-quarter of the stepfather and adoptive fathers sexually abused their victims and roughly one-third physically abused their victims. Interestingly, maltreatment patterns for mothers' boyfriends appeared to fall between the biological fathers and the combined group of stepfathers and adoptive fathers, such that the proportion of sexual abuse was less than it was among stepfathers and adoptive fathers, but greater than it was among biological fathers, with the pattern being reversed for neglect. Finally, more than two-thirds of the nonparent perpetrators sexually abused their victims.
To explore this further, the categories of male perpetrators were broken down by those who acted alone and those who acted in concert with a victim's mother. One of the most important findings is that a majority of male perpetrators in this study were found to have always acted alone. For these analyses, the stepfather, adoptive father, and mother's boyfriend categories were combined as father surrogates. When these categories were arrayed with respect to the victim characteristics, and particularly their maltreatment experience, a continuum emerges from the contrasts. Male perpetrators acting alone were more likely to have committed sexual abuse or physical abuse, to have abused girls, to have abused older children, and not to have received services than similar male perpetrators who acted with mothers. Male perpetrators who acted with mothers were more likely to be associated with neglect than any other type of maltreatment. In other words, the overall perpetration pattern for males acting with mothers has more similarities to the overall pattern for females when compared to males acting alone.
Both within the group of perpetrators acting alone and with the mother, the biological fathers' victims were younger and proportionally fewer were sexually abused. Surrogate fathers victimized a greater proportion of older female children and were associated more often with sexual abuse compared with biological fathers. However, when compared with biological and nonparent perpetrators, the surrogate fathers fell in between in terms of age, gender, and sexual abuse of their victims. Taken together, the data appear to support a two-dimensional continuum of male perpetration patterns based on the relationship to the child and whether the perpetrator acted alone.
Finally, intervention outcomes appear to be tied into the continuum as well. With the exception of mothers' boyfriends, more than one-half of male perpetrators in all father categories received some services. However, the rate of foster care service provision was three times as frequent among male perpetrators acting with mothers as it was among males acting alone, across categories. On the other hand, the proportion associated with the provision of in-home services was relatively consistent among male perpetrators, regardless of involvement of the mother. Only among nonparents was in-home service notably higher among those perpetrators involved with the mother than those who were acting alone.
Although recidivism rates were low, biological fathers were more likely to be perpetrators of maltreatment again than were most other male perpetrators. The lowest levels of recidivism were among mothers' boyfriends and stepfathers. The lack of permanence in the relationship between a boyfriend and the mother may be responsible for this; such a perpetrator may be excluded from the household before recidivism can occur. Among stepfathers, lower recidivism rates may also result from such exclusion from the household, either on the part of the mother or CPS, or it may reflect the successful implementation of intervention services.
A pattern of decreasing recidivism can also be tied to the perpetrator continuum, so that those males acting with mothers were more likely to recidivate than were males acting alone. Also consistent with the continuum, among perpetrators acting with mothers, biological fathers acting with mothers were most likely to recidivate followed by father surrogates and then nonparents. The father surrogate perpetrators acting alone were least likely to recidivate. The type of maltreatment or level of access to the child was not considered for these analyses.
These findings may have important policy implications. The first implication is that if prevention and treatment interventions for child maltreatment are targeted only to women, a large proportion of perpetrators will not benefit from these efforts. Secondly, in-home services, in their most narrow sense, may be missing the opportunity to involve men who maltreat children but who are not living in the home. As many as one-third more biological fathers act alone than act with a mother in maltreating their children, suggesting a degree of disconnectedness in family relations that may be significant.
The second implication is that perpetrators who are not biological fathers were more commonly associated with physical abuse and sexual abuse, older children, and female children. Similarly, when acting alone, biological fathers and father surrogates were more often perpetrators of physical and sexual abuse, but when acting with the mother were more often associated with neglect. These perpetrator category distinctions echo broader distinctions in the NCANDS data and add to the notion that the dynamics associated with neglect are quite dissimilar from physical and sexual abuse. But again, this study highlights that a relatively large proportion of male perpetrators are involved in neglect when both parents are present, a finding not emphasized in prior research. Additionally, the relatively large proportion of stepfathers and adoptive fathers associated with sexual abuse, as well as with older, female children, suggests the need for prevention efforts in blended and adoptive families.
The third area of findings with potential policy implications is that male perpetrators have many different relationships to their victims. The findings and the literature suggest that interventions that strengthen the role of fathers to prevent further child maltreatment and improve child well-being are a complex undertaking. This study provides insights into this complexity by identifying clear subgroups of perpetrators. Because of the distinct differences among these male perpetrators with different relationships to their victims, interventions of all types may need to be more highly differentiated. The classification scheme presented could be refined and combined with more information about victims, and ultimately families, resulting in a data driven classification scheme of CPS populations for whom specific and targeted interventions may be designed. For example, it seems discouraging that biological father perpetrators who acted with the victims' mother were more likely to recidivate than were other male perpetrators. However, as the data show, these are also situations in which neglect is more frequently of concern and children are younger, both important risk factors for recurrence (U.S. Department of Health and Human Services, 2004). This is precisely the set of children and families who must be served more effectively to reduce maltreatment recurrence. The data clearly imply that fathers of these families must be included in the service plans.
Further exploration of information on male perpetrators is certainly warranted. For example, multivariate analyses have not been conducted; these might highlight even more clearly how the variables relate to each other and the relative importance of some variables in refining the classification scheme. In addition, obtaining a clearer picture of how the various categories of perpetrators fit within households would provide insights into the service and recidivism outcomes. The NCANDS data remain an important resource in developing this understanding as they provide a comprehensive view of the range of child maltreatment circumstances among CPS populations.
[ Go to Contents ]
1. These States were: Colorado, Delaware, Iowa, Idaho, Illinois, Indiana, Kentucky, Massachusetts, Minnesota, Missouri, Mississippi, Montana, New Mexico, Ohio, Oklahoma, Texas, Utah, and Virginia.
2. Table A-1 provides supporting data for this figure.
3. Current Population Survey 2004, March Supplement.
4. Table A-2 provides supporting data for this figure.
5. Table A-3 provides supporting data on race of male perpetrators.
6. Table A-4 provides supporting data on male perpetrators as caregivers.
7. Table A-5 provides supporting data on the number of child victims associated with male perpetrators in each category.
8. Table A-6 provides supporting data for Figure 4.
9. Percentages add to more than 100 percent due to perpetrators being counted more than once if they maltreated more than one child. Since all combination fathers maltreated more than one child, the percentages for this group are more than 200 percent.
10. Table A-7 provides supporting data for Figure 5.
11. Table A-2 provides supporting data on the number of perpetrators in each category who acted alone, with the mother, or with another person.
12. Table A-8 provides supporting data for Figure 6.
13. These risk factors were coded as caregiver alcohol abuse and caregiver drug abuse in the NCANDS data set. They were recoded as alcohol abuse, drug abuse, and drug and alcohol abuse.
14. Table A-9 provides supporting data for Figure 7.
15. Table A-10 provides supporting data for Figure 8.
16. Table A-11 provides supporting data for Figure 9.
17. Survival analysis is routinely used to provide an unbiased estimate of the likelihood of occurrence of a certain events in this case, subsequent maltreatment by a perpetrator. Survival analysis controls for the situations in which the subjects can no longer be observed. Survival estimates are especially useful in this instance because not all perpetrators can be tracked for an equal period of time.
18. Combination fathers were omitted from Figure 10 due to inconsistency of scale and because they are defined, in part, by multiple events and may not represent a comparable category. Nonparents also include some complex relationships, with more than one event.
19. Table A-12 provides supporting data for Figure10.
20. For these analyses, 5,594 perpetrators who acted only with a person other than the victim's mother were excluded.
21. Table A-13 provides supporting data for Figure 11.
22. Table A-14 provides supporting data for Figure 12.
23. Table A-15 provides supporting data for Figure 13.
Top of Page |
Main Page of Report | Contents of Report
Human Services Policy (HSP)
Assistant Secretary for Planning and Evaluation ASPE)
U.S. Department of Health and Human Services (HHS)