An investigative analysis of 463 incidents of single-victim child abductions identified through Federal Law Enforcement

Aggression and Violent Behavior 30 (2016) 59–67

Contents lists available at ScienceDirect

Aggression and Violent Behavior

An investigative analysis of 463 incidents of single-victim child abductions identified through Federal Law Enforcement☆ Janet I. Warren a,⁎, James M. Wellbeloved-Stone a, Mark A. Hilts b, William H. Donaldson b, Yvonne E. Muirhead b, Sarah W. Craun b, Anna Grace Burnette a, Sara B. Millspaugh c a b c

Institute of Law, Psychiatry, and Public Policy, University of Virginia Health Systems, PO Box 800660, Charlottesville, VA 22908-0660, United States Behavioral Analysis Unit, Federal Bureau of Investigation, 1 Range Road, CIRG/NCAVC/BAU, Quantico, VA 22135, United States Department of Psychology, Rosalind Franklin University of Medicine and Science, 3333 Green Bay Road, North Chicago, IL 60064-3095, United States

a r t i c l e

i n f o

Article history: Received 3 May 2016 Received in revised form 6 July 2016 Accepted 6 July 2016 Available online 7 July 2016 Keywords: Child abduction Criminal history Victim age Victim gender Child murder

a b s t r a c t We examined the characteristics of perpetrator, victim, and crime scene for 463 child abduction incidents involving a single perpetrator and single victim based upon case material submitted to federal law enforcement. The victims were predominantly female with sexual assault being the primary motivation for the abduction of both the female and male victims. Within this sample 55% of the female child victims and 49% of the male child victims were found dead or not recovered. Offenders who were identified as being criminally versatile were found more often to abduct the youngest and oldest children, to be between the ages of 30 to 59, and a stranger to their victim. In contrast, perpetrators with prior crimes against children tended more often to be below the age of 30 years, to demonstrate more of a propensity for abducting children of minority status, and to perpetrate crimes with a lower probability of holding their victims for more than 8 h. Only 5% of offenders who abducted a female child and none of the perpetrators who abducted male child victims during or after 1994 were found to be registered on a state or federal sex offender registry. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction Child abductions are frightening events, which distress the larger community, and encapsulate some of the worst fears of parents. In the midst of the reactions to an abduction, law enforcement officers are expected to respond quickly and to manage these events effectively, ideally protecting the child victim from further harm. Prior research has indicated that almost half the children who are murdered in the course of an abduction are killed within an hour of their kidnapping, while nearly 80% are killed within 3 h (Brown, Keppel, Weis, & Skeen, 2006). These statistics underscore the importance of crime-specific research to support, inform, and expedite law enforcement's investigative efforts when abduction is the suspected or known cause of a child's disappearance. Over the past 20 years, in order to assist investigative efforts into such crimes, a body of research has developed that mines large aggregate datasets to explore possible associations between the crime scene

☆ This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. ⁎ Corresponding author: Institute of Law, Psychiatry, and Public Policy, University of Virginia Health Systems, PO Box 800660, Charlottesville, VA 22908-0660, United States. E-mail address: [email protected] (J.I. Warren).

http://dx.doi.org/10.1016/j.avb.2016.07.006 1359-1789/© 2016 Elsevier Ltd. All rights reserved.

behavior of identified offenders and their personal characteristics, these associations possibly being applicable or relevant to similar crimes that may be occurring. Rather predicting behavior based upon identity as seen in most social science research, this inversion of analytic thinking which predicts identity based upon behavior, is referred to as Criminal Investigative Analysis (CIA). CIA is an investigative adjunct that is used to inform cases in which more-direct investigative efforts have failed to bring any resolution to a crime or crime series. Specifically, CIA assists law enforcement by examining information about the victim, the apparent intent of the crime, and the circumstances and condition of the victim upon release or disposal (Morton & Lord, 2005). In so doing, CIA prioritizes suspects for further investigation and offers a framework for narrowing the suspect pool based upon the characteristics of the victim and the particular type of violent, interpersonal crime that is believed to have occurred (Morton & Lord, 2005). One aspect of CIA involves the exploration of victim selection as it might relate to the motivation for the crime, the prior relationship between the perpetrator and victim, and the characteristics of perpetrator, including age, race, past criminal history, and history of violent offending. In our current study, we examined the dynamics of child abduction, with and without murder, involving a single victim and a single perpetrator, with a specific focus on the race, gender, and age of the victim as it informs possible descriptive characteristics of the then

60

J.I. Warren et al. / Aggression and Violent Behavior 30 (2016) 59–67

unidentified offender. These variables were identified for our study as they represent what is possibly the only information that is reported to law enforcement at the time that a child goes missing. 1.1. Typologies of child abduction Previous research has categorized child abductions into typologies using different relationships, motivations, and characteristics of the crime to provide insight into both the victim and the offender, potentially offering insight into relevant investigative strategies. The U.S. Department of Justice's National Incident-Based Reporting System (NIBRS) organizes child kidnappings by relationship of the perpetrator to the victim (Finkelhor & Ormrod, 2000). An analysis of 1214 juvenile kidnapping cases in 12 states indicated that there are three primary types of kidnapping; those involving a family member, acquaintance, or stranger. Family kidnappings accounted for 49%, acquaintance kidnappings for 27%, and stranger kidnappings for 24% of all cases. Family kidnappings are typically perpetrated by a parent, most often a female, involve victims of both sexes, often under the age of six, and tend to originate in the family home. The acquaintance kidnapping category incorporates a larger proportion of juvenile offenders and adolescent female victims abducted from their homes, and is often associated with other types of violent and sexual crimes, thus increasing the chances of severe injury to the victim. Stranger kidnappings, which are perpetrated almost exclusively by males, more often involve public outdoor locations and the use of weapons, and include both school-aged and adolescent victims. Stranger kidnappings perpetrated against females tend to involve sexual motives, while those perpetrated against male victims are more likely to be based on immediate financial gain. Although abductions by family members are common, the majority of the literature focuses on abductions committed by non-family perpetrators. Finkelhor, Hammer, and Sedlak (2002), using a randomized national household survey combined with a survey of law enforcement agencies in the NISMART-2 study, estimated that over a one-year period, 58,200 children experienced some type of nonfamily abduction and 115 experienced what was defined as a stereotypical kidnapping. Nonfamily kidnappings are defined as criminal events designed to obtain ransom or permanent possession of a child, which involve a victim taken by force or threat of force and held for over an hour, or a victim under the age of 15 or otherwise mentally incompetent who voluntarily leaves with the perpetrator. Stereotypical kidnappings are characterized by the holding of the victim overnight, the transportation of the victim at least 50 miles, and the perpetration of the crime with the intent to obtain a financial gain, hold the victim indefinitely, or commit murder (Finkelhor et al., 2002). Brown et al. (2006) confirmed the potentiallyviolent relationship between the non-family perpetrators to murderabduction victims by concluding that 44.4% of all murdered abduction victims were killed by strangers, 41.9% by a friend or acquaintance, and 13.7% by family or intimate individuals. A second way of categorizing abductions derives from the perceived motivation of the offender. Erickson and Friendship (2002), using a British sample of 149 offenders, developed a motivational framework for identifying different types of child abductions. They found that the sexually-motivated group, which made up the preponderance of the cases, was the only motivational group in which the offenders were found to have a previous criminal history of child abductions. Abductions motivated by maternal desire were primarily committed by nonfamily female offenders; these women were found to have the lowest rate of prior criminal offending, and were prompted in their criminal behavior by the desire to obtain a child to represent as their own, to maintain relationships with intimate partners, or to become mothers. Boudreaux, Lord, and Dutra (1999) examined 550 cases of abducted children obtained through two databases: the Violent Criminal Apprehension Program (ViCAP) and the data maintained by the Child Abduction and Serial Killer Unit (CASKU). The authors identified five motivational themes: sex, emotion-based, profit, infant abduction

(also known as maternal desire), and unknown. Sexually-motivated abductions required physical evidence of a sexual act, and were almost exclusively perpetrated by male offenders. They were also the most common type of abduction perpetrated by men and accounted for 60% of all male-perpetrated abductions. Emotion-based abductions included crimes that occurred in response to, or as an effect of the offender's heightened emotions; for example, anger, resentment, and hatred. They were found to be associated with the abduction of infants, toddlers, and preschoolers, but less often with older youth who were reaching the age associated with puberty. Emotion-based abductions accounted for 51% of female-perpetrated abductions, and 27% of maleperpetrated abductions. Profit-based offenses were found to be more prevalent among high school male victims, and were seldom associated with any type of sexual motivation; this category accounted for 13% of male-perpetrated abductions. Infant abductions involved the abduction of a child with the intent to keep him or her as the abductor's own and were exclusively committed by female offenders—accounting for 44% of all female-perpetrated abductions. Lanning and Burgess (1995) produced another motivational typology which they derived from an examination of data obtained through the Massachusetts Treatment Center and the FBI. They propose seven motivational categories: nontraditional, ransom, profit, sexual, killing, miscellaneous criminal, and political. In describing the sexually motivated abductions, they conclude that there can be multiple motivations for a single abduction, particularly when multiple perpetrators are involved. They also comment on the efforts made by some perpetrators to camouflage, even to themselves, the sexual motivation of their crimes. The typology of child molesters developed by Lanning (1992), which identifies situational and preferential offenders, is also thought to be applicable to sexually motivated child abductions. Lanning and Burgess (1995) concluded that nonfamily child abductors who are sexually motivated manifest diverse patterns of behavior, including relationship to victim, degrees of force used in the abduction, and methods used to obtain access to a particular victim.

1.2. Demographic characteristics of victims Research concerning victim characteristics has been consistent in documenting that the majority of children abducted by non-family members are females. Finkelhor et al. (2002) attribute this overrepresentation of female victims to the sexual motivation of these crimes, and report that 65% of nonfamily abductions and 69% of stereotypical abductions involved female victims. They further underscore that the targets of many abductions are teenage girls, and contrast this finding with popular media misrepresentation of very young abduction victims. Finkelhor and Ormrod (2000) similarly found acquaintance abduction victims to be predominantly female, which they contrasted with family abductions, which were determined to be approximately equal between male and female victims. Based upon their prior research, Boudreaux et al. (1999) explored the effects of victim gender on different patterns of victimization. They found that although females were at higher risk and accounted for 70% of abduction victims, there was no significant difference in gender between identified victims when the victim was of a younger age (Boudreaux et al., 1999). The authors propose that the presence of maternal desire as a motivation for abduction explains the lack of gender specificity when neonates and infants are victimized. Boudreaux et al. (1999) also analyzed the gender of abductors with the abducted children's ages, and found that females who abducted children accounted for 91% of neonates, 54% of infants, 41% of toddlers, 15% of children in preschool, 1 % of children in elementary school, 3% of children in middle school, and 2% of children in high school. They further found that females were more likely to be abducted for sexual motivations, and that males and females were equally abducted for emotional, profit-based, and infant-obtaining reasons.

J.I. Warren et al. / Aggression and Violent Behavior 30 (2016) 59–67

1.3. Criminal history Research challenges the public perception of child abductors being criminally monotonic in their past offending, highlighting a rather broad array of past criminal offenses among those who ultimately abduct and possibly murder a child. As previously mentioned, Erickson and Friendship (2002) found different patterns of criminal offending among their four child abduction types. Notably, sexually-motivated abductors had the highest rate of prior sexual crimes, with the “other” abductors exhibiting the highest rate of more diverse patterns of criminal offending. Brown et al. (2006) studied the characteristics of child abductors who were known to have murdered their victims. Within their sample of 735 victims, 46% of the abduction-murder offenders had a substantial history of previous crimes against children. The most common prior offenses within this group included sexual assault (21%), rape and attempted rape of a child (13%), murder or attempted murder of a child (11%), and abduction of a child (7%). Beasley et al. (2009) sought to identify specific patterns of criminal behavior of fatal and non-fatal child-abductors. Within their sample, a third of all abductors were found to be criminally versatile, defined as having more than three different types of offenses in their criminal history. Few significant differences were found in the criminal histories of perpetrators who were known to have murdered their victims and those whose victims were found alive. Between both groups, the majority of offenders had not committed previous crimes involving the abduction and sexual assault of children. Of policy relevance was the finding that a very small percentage of the perpetrators were registered sex offenders, which may have been influenced by the enactment date of sex offender registration laws. The authors concluded that offenders typically abduct children against the backdrop of chronic criminal careers including property crimes and crimes against both adults and children. 1.4. Current study Using the data made available by retrospective analyses of abduction cases reviewed by federal law enforcement, we examine how victim variables such as age, gender, race, and known location of abduction is associated with perpetrator characteristics including age, gender, race, motivation for the crime, the presence and type of criminal history, along with aspects of the abduction including its fatal or non-fatal outcome. Based on a sample of both family and nonfamily abductions, we explore these relationships between perpetrator and victim in an effort to better inform the timely response of law enforcement. These analyses are informed by the CIA perspective, and aid in illuminating the interplay of motivational and interpersonal dynamics as they influence victim choice by diverse types of offenders. 2. Method 2.1. Data sources The FBI's National Center for the Analyses of Violent Crime (NCAVC) collected 617 abduction cases from the Violent Criminal Apprehension Program (ViCAP) and the Kidnapping and Abduction Statistical Application (KASA). ViCAP was established by the Department of Justice in 1985 to provide information about serial offenders believed to offend in multiple jurisdictions over the course of their criminal careers (Critical Incident Response Group, n.d.). In 2002, software was developed for ViCAP enabling law enforcement officials to enter case data directly into the national database, resulting in submission of case information by N 5000 law enforcement agencies nationwide on an ongoing and real-time basis. KASA was developed by the FBI's Criminal Justice Information Services to identify cases of kidnapping designated for

61

analysis by two of its investigative units, the Behavioral Analysis Crimes Against Children Unit, and the Violent Crimes/Fugitive Unit. The criminal history of each known perpetrator used in our study was collected from the National Crime Information Center (NCIC). The NCIC was established in 1967 to centralize criminal history records and includes 21 crime categories, including seven property crimes and 14 crimes against person (NCIC, n.d.). The presence of offense histories in multiple states was coded for, as was criminal versatility, which was defined in the current study as having offenses in two or more crime categories such as assault, drug crimes, fraud, embezzlement, arson, pornography, prostitution, driving under the influence, forcible or non-forcible sex offenses, and kidnapping (Beasley et al., 2009). Six hundred and seventeen incidents of child abduction, with and without murder, were identified as part of the current research project; of these, 463 involving a single perpetrator and single victim were examined in this first set of data analyses. Offenders were incorporated into the study if they were believed by the investigating agencies to be the person responsible for the crime (but for some reason could not be arrested), or were arrested for the attempted or completed abduction of a child. Incidents involving some type of physical contact and/ or restraint occurring in a single location, such as the victim's residence, were included in the study; incidents involving only an attempted verbal luring of the child were not included. The definition of abduction of a child in our current study follows that provided by statute (18 U.S.C. § 1201), which includes not only the coerced, unauthorized, or otherwise illegal movement of a child, but also the holding or confining of a child's movement, possibly through the use of restraints, with the intent to perpetrate a criminal offense. The data expanded upon prior research concerning child abductions that were collected by crime analysts, as well as researchers from the University of Virginia, resulting in a database of 226 variables. The initial data was collected by crime analysts, supplemented by two additional protocols developed and coded by University researchers. The supplemental protocols collected data that allowed the database to be restructured according to individual abduction incidents, enabling the examination of incidents involving different numbers of perpetrators and victims. The final database included data concerning victim characteristics, offender characteristics (including their criminal histories), the relationship between the victim and offender, location of the abduction, duration of the abduction (less than or greater than 8 h), offender's primary intended target, the causes of death in cases of abduction-murder, and the offender's perceived motivation for the abduction. The location of the abduction referred to the site of initial contact between the offender and victim, regardless of if the victim left willingly with the offender, was lured by a ruse, or physically taken by the captor. The criminal histories were categorized according to the Federal Bureau of Investigation's National Incident-Based Reporting System (NIBRS) Volume 1: Data Collection Guidelines (Federal Bureau of Investigation, FBI, 2000). The offenders' criminal histories included any offense committed prior to the instant offense or prior to the offenders' first incident of child abduction. Based on previous research (e.g. Erickson & Friendship, 2002; Boudreaux et al., 1999; Lanning & Burgess, 1995), the motivations for the abductions were classified into six categories: sexual, heightened emotion, profit, maternal desire, mental illness/depression, and other. Sexually-motivated abductions were defined as incidents in which the offender chose the victim to fulfill sexual desires, needs, or fantasies—corroborated through the offender's statement, victim's statement, the sexual abuse of the victim, and/or the disposal of the victim's body. Maternal desire was identified in instances in which the offender was motivated by the desire to obtain and raise the victim as their own and/or feigned a pregnancy in order to maintain romantic relationships. Incidents classified as heightened emotion were abductions driven by the offender's overpowering emotions at the time of the instant offense, and did not involve fulfilling the offender's sexual desires, needs, or fantasies. Profit-motivated abductions included cases that

62

J.I. Warren et al. / Aggression and Violent Behavior 30 (2016) 59–67

involved some type of material or financial gain. The category of mental illness/depression did not require the presence of a formal psychiatric diagnosis, but did require clear indications of disturbed thinking or a severe mood disorder. The data regarding victim age were analyzed according to five age categories: 0 to 1, 2 to 5, 6 to 11, 12 to 14, and 15 years and older. These age categories were used to capture differences in the motivation and gender of both perpetrator and victims, as identified in prior research on child abduction (Boudreaux et al., 1999). The relationship between the victim and the offender was categorized according to five groups; stranger, acquaintance, family, intimate partner, and unknown. The distinction between stranger and acquaintance was based upon the presence of any prior interaction, and was not based upon familiarity or the length of time that the perpetrator was in contact with the victim.

Table 1 Methods of categorizing child abductions, by perpetrator, victim, and incident. N perpetrators Single perpetrators Multiple perpetrators Number of cases Number of single perpetrator cases Number of multiple perpetrator cases Single perpetrators Single perpetrators with 1 victim only Single perpetrators with 2 or more victims Multiple perpetrators cases Multiple perpetrator cases with 1 victim only Multiple perpetrator cases with 2 or more victims Single perpetrator range of number of victims Multiple perpetrator group range of number of victims Range of number of perpetrators in group

685 534 151 617 534 83 534 463 71 83 69 14 1–13 1–4 2–9

78% 22% 87% 13% 87% 13% 83% 17%

2.2. Data exclusion The cases drawn from these databases were compared to ensure that no duplicate files were included within the study, and to ensure that the following types of abductions were excluded from the study. Incidents in which the child was not the intended target were excluded; for example, individuals attempting to steal a vehicle may not have noticed a child in the vehicle, and were later charged with child abduction in addition to theft-specific charges. Abductions that involved multiple child victims either in a single abduction or over the course of multiple abductions were also removed as they were being analyzed separately to determine the similarities between single and multiple victim incidents. Finally, incidents that lacked adequate information regarding the offender's arrest information, description of the incident, or the victim's gender and age were removed from the study. These exclusion criteria resulted in a total N for the current study of 463 incidents of child abduction involving a single perpetrator with a single victim. Data on sex offender registration was reviewed only for cases that involved abductions on or after 1994, when the Jacob Wetterling Crimes Against Children and Sexually Violent Offender Registration Act was enacted (N = 297). 2.3. Statistical analyses In testing significance, Pearson χ2 tests were conducted, and when expected cell counts were below five, Fisher's Exact Tests (FET) were conducted, using a 10,000 sample Monte Carlo. For logistic regressions, variables were examined for being predictive of the following aspects of the perpetrator's criminal history: any adult criminal history, criminal versatility, crimes against adults, crimes against children, crimes against property, and crimes against society. The predictive power of these variables was analyzed first using only variables that might be known at the outset of an abduction, and second, by examining both the characteristics of the victim, the perpetrator, and the crime scene as could be understood only after the abduction was completed and the perpetrator identified. 3. Results Table 1 provides different but interrelated ways of describing and analyzing child abductions. These include analyses that focus on the perpetrator(s), single or multiple; analyses that examine the victim(s), single or multiple, and analyses that quantify crime scene behavior across incident(s), which can also include a single incident or many. Each of these perspectives encompasses a different set of data and variables that are distinct but interrelated. The data in Table 1 indicate that the majority of incidents were made up of single perpetrators with a single victim (75%). This combination results in a total of N of 463 incidents which constitute the basis of the current study. Contrasted to this group were single perpetrators (N = 71) who were found to have up to

13 victims, and groups of multiple perpetrators (N = 83) who were found to have from one to four victims. Table 2 contains demographics of the 463 single perpetrators with a single victim included in the current analyses. The majority of perpetrators of child abduction were male, with only 6% of the perpetrators being identified as female. A significant difference was observed comparing the age categories across the two genders P = 0.048, FET. Among the male perpetrators, 75% of the men were between the ages of 18 through 39 years as contrasted to the female perpetrators; 63% of the women fell in this same age range with a larger number of the female perpetrators being under the age of 18 years. Across both males and females the majority were White/Caucasian (76%), followed by Black/African-American (16%); no significant differences in race/ ethnicity were found by gender. The majority of both the male (75%) and female perpetrators (70%) had prior criminal histories. Table 3 examines the characteristics of the child victims, illustrating that the victims were predominantly female. Gender differences were found when comparing the age of the child victims across five different age categories, χ2(4, N = 463) = 25.33, P b 0.001. More male infants were abducted and more teenage girls aged 15 to 17 years of age. However, the most frequent age range for both the male and female victims was six to 11 years of age. Difference was also found comparing child victim gender and race/ethnicity, P = 0.007, FET. A majority of the child victims were Caucasian, and a greater proportion of male child victims were African American. The location of the abduction varied across the male and female victims, P = 0.047, FET. The most frequent location of abduction for all child victims, regardless of gender, was a public outdoor setting, 46% of abductions overall; the second most common Table 2 General characteristics of perpetrators of child abduction (N = 463). Age of offender

Male (N = 433)

Female (N = 30)

17 (4%) 197 (45%)⁎ 128 (30%)⁎ 69 (16%) 17 (4%) 5 (1%)

4 (13%)⁎ 7 (23%) 12 (40%) 5 (17%) 2 (7%) 0 (0%)

Race/ethnicity Asian African-American Hispanic/Latino Native American Caucasian Other

Male (N = 430) 5 (1%) 64 (15%) 30 (7%) 2 (0%) 328 (76%) 1 (0%)

Female (N = 29) 0 (0%) 8 (28%) 1 (3%) 0 (0%) 19 (66%) 1 (3%)

Prior arrest history Yes No

Male (N = 429) 320 (75%) 109 (25%)

Female (N = 30) 21 (70%) 9 (30%)

b18 18–29 30–39 40–49 50–59 60+

⁎ P = 0.05.

J.I. Warren et al. / Aggression and Violent Behavior 30 (2016) 59–67 Table 3 General Characteristics of Victims of Child Abduction (N = 463). Age of victim

Table 4 General characteristics of single-victim incidents of child abduction (N = 463).

Male (N = 90)

Female (N = 372)

15 (17%)⁎⁎ 12 (13%) 29 (32%) 19 (21%) 15 (17%)

13 (3%) 39 (10%) 123 (33%) 93 (25%) 104 (28%)⁎⁎

Race/ethnicity of victim Asian African-American Hispanic/Latino Native American Caucasian Other Unknown

Male (N = 90) 1 (1%) 21 (23%)⁎⁎ 5 (6%) 1 (1%) 57 (63%) 2 (2%) 3 (3%)

Female (N = 373) 5 (1%) 32 (9%) 22 (6%) 1 (0%) 292 (78%) 7 (2%) 14 (4%)

Location of abduction⁎ Victim residence Other residence Public outdoor Public inside Rural outdoor Unknown Public Private Indoor Outdoor

Male (N = 90) 17 (19%) 9 (10%) 37 (41%) 6 (7%) 4 (4%) 17 (19%) 47 (64%) 26 (36%) 32 (44%) 41 (56%)

Female (N = 373) 106 (28%)⁎ 29 (8%) 176 (47%) 15 (4%) 13 (3%) 34 (9%) 204 (60%) 135 (40%) 150 (44%) 189 (56%)

0–1 2–5 6–11 12–14 15–17

63

⁎ P b 0.05. ⁎⁎ P b 0.01.

location of abduction was the victim's residence, 27% of abductions overall, with a greater proportion of females being abducted from their residence. Table 4 examines elements of the crime scene associated with the abduction of male and female child victims. There were no differences in the relationship between the victim and the perpetrator, matching race of victim and perpetrator, primary cause of death, and the presence of some form of asphyxiation when viewed across the two genders. Significant differences were found in the perceived motivation for the abduction of male and female victims, P b 0.001, FET. The majority of both the male and female victims were perpetrated for sexual reasons, but the proportion of sexually motivated abductions was more common among the female child victims (75%) when compared to the male child victims. Abductions motivated by maternal desire were the secondmost common motivational category for both boy and girl victims. When the duration of abduction was known, the largest proportion of abductions were found to not last longer than 8 h with more female children, 66%, being recovered within 8 h, as opposed to 52% of male child victims χ2(2, N = 463) = 9.03, P = 0.011, Cramér's V = 0.140. As for the outcome of the abduction, 45% of female child victims and 56% of male child victims were recovered alive, this reflecting a nonsignificant slightly higher rate of murder among the female victims χ2(1, N = 463) = 3.05, P = 0.081, Cramér's V = 0.081. There were no significant interactions found between the cause of death and the gender of the victim with the most common primary cause of death for both male (28%) and female (42%) child victims being asphyxiation. When asphyxiation was examined across the multiple causes of death, strangulation was found to be more common than suffocation or other unspecified forms of asphyxiation, including airway occlusion, with these patterns being consistent among both male and female victims. Table 5 summarizes the prior criminal history of single perpetrators of single victim abductions by age and gender of the victim. Across all female age groups, the majority of perpetrators had a prior criminal history (64% to 78%). Perpetrators who abducted females aged 15 to 17 years of age were more likely to have prior arrests for crimes against adults as compared to perpetrators who abducted female victims of younger ages χ2(4, N = 364) = 13.752, P = 0.008, Cramér's V = 0.194. The observed

Relationship to perpetrator

Female victims (N = 372)

Male victims (N = 90)

16 (4%) 19 (5%) 138 (37%) 141 (38%) 58 (16%)

4 (4%) 4 (4%) 38 (42%) 34 (38%) 10 (11%)

Matching Race Yes No

Female victims (N = 373) 299 (80%) 74 (20%)

Male victims (N = 90) 71 (79%) 19 (21%)

Motivation for crime Sexual Heightened emotion Profit Maternal desire M.I.-depression Other Unknown

Female victims (N = 373) 283 (75%)⁎⁎⁎ 7 (2%) 2 (1%) 13 (4%) 4 (1%) 2 (1%) 62 (16%)

Male victims (N = 90) 47 (52%) 5 (6%) 3 (3%) 8 (9%) 2 (2%) 3 (3%) 22 (25%)

Victim held N8 h⁎ Yes No Unknown

Female victims (N = 373) 78 (21%) 163 (44%)⁎⁎ 132 (35%)

Male victims (N = 90) 32 (36%) 35 (39%) 23 (25%)

Outcome of abduction Alive Deceased/not found

Female victims (N = 373) 169 (45%) 204 (55%)

Male victims (N = 90) 50 (56%) 40 (44%)

Primary cause of death Asphyxiation Blunt force injury Cut or stab Gunshot Drown Other

Female victims (N = 204) 85 (42%) 36 (18%) 24 (12%) 13 (6%) 4 (2%) 42 (21%)

Male victims (N = 40) 11 (28%) 10 (25%) 5 (13%) 7 (17%) 0 (0%) 7 (17%)

Murdered victims Any suffocation Any strangulation Unspecified asphyxiation

Female victims (N = 204) 12 (6%) 69 (34%) 24 (12%)

Male victims (N = 40) 1 (3%) 12 (30%) 2 (5%)

Relative Intimate Partner Acquaintance Stranger Unknown

⁎ P b 0.05. ⁎⁎ P b 0.01. ⁎⁎⁎ P b 0.001.

rates at which perpetrators with female victims had prior crimes against children was comparable to crimes against adults, but lower than the rate at which they were found to have prior crimes against property. A substantial proportion of the perpetrators with a female victim were found to be criminally versatile in prior criminal history, with this proportion ranging from a low of 38% for those who abducted female victims between the ages of 12 and 14, to 54% of those who abducted female infant victims. Of 239 perpetrators who abducted female children during or after 1994, very few were found to be registered on a sex offender registry (5%); however, a significant interaction was observed between sex offender registration status and age of the female child victim, P = 0.047, FET, with the female child victims of registered sex offenders being more likely to be six to 11 years of age. When the victim of the abduction was male, the majority of the perpetrators were also found to have a criminal history, also an interaction between the male victim's age and the perpetrator having a prior arrest history for kidnapping, P = 0.030, FET, with 33% of the perpetrators who abducted adolescents between the age of 15 and 17 years. Crimes against children were uncommon, but there was a significant interaction found, P = 0.014, FET, with 50% of perpetrators who abducted male victims between the age of 12 to 14 years having a prior arrest concerning a child. Property crimes were found to vary by age of the male child victim, χ2 (4, N = 90) = 8.04, P = 0.090, Cramér's V = 0.299, being more frequent among perpetrators who abducted infants between the ages of zero to one year.

64

J.I. Warren et al. / Aggression and Violent Behavior 30 (2016) 59–67

Table 5 Offender criminal history by victim gender and age (N = 463). Female victims (N = 373) Prior arrest history Any (N = 369) Kidnapping (N = 372) Criminal arrest categories⁎ Against adults (N = 364) Against children (N = 341) Against property (N = 372) Against society (N = 370) Criminal versatility (N = 369) Arrests in 2+ categories Interstate history (N = 372) Arrests in 2+ states Registered sex offender (N = 239)⁎ At time of abduction On probation/parole (N = 316) At time of abduction Male victims (N = 90) Prior arrest history⁎ Any (N = 89) Kidnapping (N = 90) Criminal arrest categories⁎ Against adults (N = 88) Against children (N = 84) Against property (N = 90) Against society (N = 89) Criminal versatility (N = 89) Arrests in 2+ categories Interstate history (N = 90) Arrests in 2+ states Registered sex offender (N = 58) At time of abduction On probation/parole (N = 82) At time of abduction

0–1

2–5

6–11

12–14

15–17

10 (77%) 2 (15%)

25 (64%) 3 (8%)

87 (73%) 15 (12%)

73 (78%) 15 (16%)

81 (78%) 14 (13%)

4 (31%) 4 (31%) 10 (77%) 4 (10%)

12 (32%) 9 (24%) 18 (46%) 10 (26%)

51 (43%) 26 (24%) 64 (52%) 35 (28%)

40 (44%) 16 (19%) 49 (53%) 24 (26%)

63 (61%)⁎⁎ 17 (18%) 54 (52%) 25 (25%)

7 (54%)

16 (41%)

54 (45%)

35 (38%)

53 (51%)

6 (46%)

15 (38%)

46 (38%)

44 (47%)

44 (42%)

0 (0%)

0 (0%)

8 (11%)⁎

3 (4%)

1 (2%)

2 (15%)

3 (9%)

8 (8%)

7 (9%)

8 (9%)

0–1

2–5

6–11

12–14

15–17

14 (93%) 1 (7%)

9 (75%) 0 (0%)

17 (61%) 1 (3%)

13 (68%) 2 (11%)

11 (73%) 5 (33%)⁎

6 (40%) 1 (7%) 12 (80%) 7 (47%)

4 (33%) 0 (0%) 5 (42%) 4 (33%)

5 (18%) 8 (30%) 11 (38%) 4 (14%)

8 (42%) 9 (50%)⁎ 10 (53%) 3 (17%)

7 (50%) 3 (21%) 6 (40%) 5 (33%)

8 (53%)

3 (25%)

7 (25%)

9 (47%)

6 (40%)

6 (40%)

1 (8%)

11 (38%)

8 (42%)

6 (40%)

0 (0%)

0 (0%)

0 (0%)

0 (0%)

0 (0%)

0 (0%)

3 (25%)

2 (8%)

2 (11%)

0 (0%)

⁎ P b 0.05. ⁎⁎ P b 0.01.

Table 6 examines characteristics of the abduction as they related to female child victims, grouped by age. An interaction was found across the relationship between the victim and perpetrator by female victim age, P = 0.015, FET, with acquaintance or stranger being both first and second most common relationship across all age categories. The motivation for the abduction was also significantly related to female victim age, P b 0.001, FET, with abductions being overwhelmingly motivated by a sexual interest in the child, except for abductions of infants, which were primarily motivated by a maternal desire for the female infant. Length of abduction was also found to significantly vary by female victim age, P b 0.001, FET, with infant victims predominantly being held over 8 h. The outcome of whether the female child victim was found alive or found murdered/presumed deceased was also related to child age, χ2 (4, N = 372) = 20.45, P b 0.001, Cramér's V = 0.234, with female children under the age of six being more likely to be found alive. Cause of death was also related to the age of the deceased female child victim, P = 0.012, FET, with asphyxiation being the most common cause of death, followed by blunt force injury. When any indication of asphyxiation was compared to the age of the deceased female child victim, it was found that the portion of victims experiencing suffocation tended to be younger, with rates of suffocation decreasing as the victims became older, P = 0.042, FET. Table 7 examines characteristics of the abduction as they related to male child victims, grouped by age. In contrast to the female child victims, there was only one significant relationship found between the age of the male child victim and characteristics of the abduction. Here, as with the female child victims, the motivation of the abduction was related to male child victim age, P b 0.001, FET, with the preponderance of abductions being motivated by a sexual interest in the male child victim,

except for infant male child abductions which were motivated primarily by a maternal desire for the child. The logistic regressions conducted, displayed in Table 8, indicated that predictive models could be identified each component of the perpetrator's criminal history. If the victim age was 0–1, perpetrators were significantly more likely to have either crimes against adults or property, while if the victim age was 15 to 17 the perpetrator was more likely to have crimes against adults, and if the victim age was either 0–1 or 15–17 the perpetrator was more likely to be criminally versatile. Victim race (dummy coded for minority status) indicated that perpetrators who abducted non-minority child victims were more likely to have criminal history of crimes against children. Of the offender variables, the age categories and relationship to the child victim were the only variables to result in significant predictions of offender criminal history. Notably, it appeared that there was a trend for older offenders to be more likely to have criminal versatility, crimes against adults, and crimes against children; in each of these aspects of criminal history, three different age categories were found to be significantly predictive, with generally larger odds ratio as the age categories increased. As for the relationship between the perpetrator and child victim, strangers were more likely to have criminal versatility, crimes against adults, and crimes against property, while family or intimate partners were less likely to have crimes against society. 4. Discussion There are two over-arching conceptual themes that can be identified based upon the current study. The first involves the multidimensional nature of child abduction and the various parameters according to

J.I. Warren et al. / Aggression and Violent Behavior 30 (2016) 59–67

65

Table 6 Incident characteristics for female victims by age (N = 373). Female victims (N = 373) Relationship to perpetrator (N = 371) Relative Intimate partner Acquaintance Stranger Unknown Matching race (N = 372) Yes No Motivation for crime (N = 372) Sexual Heightened emotion Profit Maternal desire M.I.-depression Other Unknown Victim held N8 h (N = 372) Yes No Unknown Outcome of abduction (N = 372) Alive Deceased/not found Primary cause of death (N = 203) Asphyxiation Blunt force injury Cut or stab Gunshot wound Drowning Other Murdered victims (N = 203) Any suffocation Any strangulation Unspecified

0–1 0 (0%) 0 (0%) 6 (46%)⁎⁎ 7 (54%)⁎⁎ 0 (0%) 11 (85%) 2 (15%)

2–5 2 (5%) 0 (0%) 20 (51%)⁎⁎ 14 (36%)⁎⁎ 3 (8%) 32 (82%) 7 (18%)

6–11 6 (5%) 1 (1%) 52 (42%)⁎⁎ 47 (38%)⁎⁎ 17 (14%) 104 (85%) 19 (15%)

12–14

15–17

5 (5%) 8 (9%) 33 (35%)⁎⁎ 31 (33%)⁎⁎ 16 (17%)

3 (3%) 10 (10%) 27 (26%)⁎⁎ 41 (40%)⁎⁎ 22 (21%)

72 (77%) 21 (23%)

80 (77%) 24 (23%)

28 (72%)⁎⁎⁎ 0 (0%) 0 (0%) 1 (3%) 1 (3%) 0 (0%) 9 (23%)

90 (73%)⁎⁎⁎ 3 (2%) 0 (0%) 2 (2%) 3 (2%) 1 (1%) 24 (20%)

75 (81%)⁎⁎⁎ 1 (1%) 0 (0%) 1 (1%) 0 (0%) 0 (0%) 16 (17%)

88 (85%)⁎⁎⁎ 2 (2%) 1 (1%) 1 (1%) 0 (0%) 0 (0%) 12 (12%)

10 (77%)⁎⁎⁎ 2 (15%) 1 (8%)

8 (21%) 20 (51%) 11 (28%)

15 (12%) 65 (53%) 43 (35%)

24 (26%) 31 (33%) 38 (41%)

21 (20%) 44 (42%) 39 (38%)

11 (73%)⁎⁎⁎ 2 (27%)

22 (56%)⁎⁎⁎ 17 (44%)

39 (32%) 84 (68%)

46 (49%) 47 (51%)

51 (49%) 53 (51%)

1 (8%) 1 (8%) 1 (8%) 8 (62%) 0 (0%) 1 (8%) 1 (8%)

1 (50%)⁎ 0 (0%) 0 (0%) 0 (0%) 0 (0%) 1 (50%)

6 (35%) 3 (18%) 1 (6%) 0 (0%) 1 (6%) 6 (35%)

39 (46%)⁎ 14 (17%) 6 (7%) 4 (5%) 2 (2%) 19 (23%)

19 (40%) 8 (17%) 7 (15%) 1 (2%) 0 (0%) 12 (26%)

20 (38%) 11 (21%) 10 (19%) 7 (13%) 1 (2%) 4 (8%)

1 (50%)⁎ 0 (0%) 1 (50%)

2 (12%)⁎ 3 (18%) 3 (18%)

7 (8%) 27 (33%) 12 (14%)

1 (2%) 20 (43%) 3 (7%)

1 (2%) 19 (37%) 5 (10%)

⁎ p b 0.05. ⁎⁎ p b 0.01. ⁎⁎⁎ p b 0.001.

which it is optimally investigated and studied. Rather than comporting with common stereotypes of a young girl being lured into a car by a stranger, we find that it is a tapestry of different interactions that involve criminal intent against a child. A child may be abducted by a single or multiple perpetrator(s), alone or with other victims, by a perpetrator who may be a man or woman, and who may have abducted a child before or be free of any past criminal history. Moreover, the victim may know the perpetrator or encounter a stranger, and he or she may be abducted for various reasons. The differences in these data across these attributes and descriptors point toward distinct patterns in offender motivation, victim choice, criminal intent, and abduction outcome, all of which are pivotal to the timely investigation of this type of crime. It also highlights the importance of research into child abduction that identifies and examines these different dimensions explicitly so that comparisons across studies can be made, and meta-analytic studies eventually conducted with the inclusion of crime behavior that is appropriately-defined for this particular type of offense. The second major conceptual conclusion mirrors to some extent the first, and underscores the complexity or intricacy with which the prior criminal offending of these individuals should be considered or examined. Common practice by law enforcement after a child is reported missing is to identify known child molesters living in the area based upon state and federal sex offender registries and to begin a broader review of adults who have had recent contact with the abducted child to see if any have a prior criminal history. This type of approach has face validity based upon the well-accepted risk axiom that past behavior best predicts future behavior and that the individual being sought is someone who has enacted a similar type of crime sometime in the past. Our study, while confirming that the majority of these perpetrators

do have a prior criminal history, suggests that this linear approach is overly simplistic and that the past criminal behavior of these offenders is more diverse and follows more-nuanced criminal trajectories than these common practices would suggest or imply. Our analyses confirm that adults who abduct a child are likely to have been previously arrested or convicted of a crime. This is true when the perpetrator is a man (75%) but to an almost equal extent true when the perpetrator is a woman (70%). This finding places child abduction in the broader arena of antisocial dispositions and behavior, and suggests that disparate acts such as stealing a baby or abducting and sexually assaulting a child, are unified in a common labyrinth of past criminal intent and conduct. However, this prior criminal behavior is unlikely to be captured by state and federal sex offender registries. Ironically, while sex offender registries were created, at least to some extent to identify suspects for future crimes against children, the data that were analyzed in the current study indicate that they have little ability to identify individuals with a propensity for abducting a child, often for sexual reasons. We found that when a female child victim was abducted, only 12 perpetrators out of 239 (5%) who abducted children after 1993 were found to be registered as a convicted sex offender, with eight of these being perpetrators who had abducted a female child victim aged six to 11 years. None of the 58 perpetrators who abducted male children after 1993 were registered sex offenders. When viewed broadly, these findings offer little support to the community notification function of the sex offender registry laws, and despite wide-based appeal, underscore the importance of other protective measures to ensure the safety of most children. Historically, the tendency to plea bargain sex offenses against children may have influenced this pattern, while the longer sentences imposed on sex offenders over the past 15 years

66

J.I. Warren et al. / Aggression and Violent Behavior 30 (2016) 59–67

Table 7 Incident characteristics for male victims by age (N = 90). Male victims (N = 90) Relationship to perpetrator (N = 90) Relative Intimate partner Acquaintance Stranger Unknown Matching race (N = 90) Yes No Motivation for crime (N = 90) Sexual Heightened emotion Profit Maternal desire M.I.-depression Other Unknown Victim held N8 h (N = 90) Yes No Unknown Outcome of abduction (N = 90) Alive Deceased/not found Primary cause of death (N = 40) Asphyxiation Blunt force injury Cut or stab Gunshot wound Drowning Other Murdered victims (N = 40) Any suffocation Any strangulation Unspecified

0–1

2–5

6–11

12–14

15–17

2 (13%) 0 (0%) 6 (40%) 7 (47%) 0 (0%)

2 (17%) 0 (0%) 4 (33%) 4 (33%) 2 (17%)

0 (0%) 0 (0%) 17 (59%) 9 (31%) 3 (10%)

0 (0%) 3 (16%) 6 (32%) 7 (37%) 3 (16%)

0 (0%) 1 (7%) 5 (33%) 7 (47%) 2 (13%)

11 (73%) 4 (27%)

9 (75%) 3 (25%)

22 (76%) 7 (24%)

15 (79%) 4 (21%)

14 (93%) 1 (7%)

0 (0%) 1 (7%) 0 (0%) 8 (53%) 0 (0%) 0 (0%) 6 (40%)

1 (8%) 1 (8%) 1 (8%) 0 (0%) 2 (17%) 1 (8%) 6 (50%)

19 (66%)⁎⁎⁎ 2 (7%) 0 (0%) 0 (0%) 0 (0%) 1 (3%) 7 (24%)

16 (84%)⁎⁎⁎ 1 (5%) 1 (5%) 0 (0%) 0 (0%) 0 (0%) 1 (5%)

11 (73%)⁎⁎⁎ 0 (0%) 1 (7%) 0 (0%) 0 (0%) 1 (7%) 2 (13%)

9 (60%) 3 (20%) 3 (20%)

5 (42%) 5 (42%) 2 (17%)

7 (24%) 12 (41%) 10 (34%)

8 (42%) 8 (42%) 3 (16%)

3 (20%) 7 (47%) 5 (33%)

11 (73%) 4 (27%)

7 (58%) 5 (42%)

12 (41%) 17 (59%)

11 (58%) 8 (42%)

9 (60%) 6 (40%)

0 (0%) 3 (75%) 0 (0%) 0 (0%) 0 (0%) 1 (25%)

1 (20%) 3 (60%) 1 (20%) 0 (0%) 0 (0%) 0 (0%)

6 (35%) 1 (6%) 2 (12%) 4 (24%) 0 (0%) 4 (24%)

2 (25%) 3 (38%) 0 (0%) 1 (13%) 0 (0%) 2 (25%)

2 (33%) 0 (0%) 2 (33%) 2 (33%) 0 (0%) 0 (0%)

0 (0%) 0 (0%) 0 (0%)

0 (0%) 0 (0%) 1 (20%)

1 (6%) 8 (47%) 1 (6%)

0 (0%) 2 (29%) 0 (0%)

0 (0%) 2 (33%) 0 (0%)

⁎ p b 0.05. ⁎⁎ p b 0.01. ⁎⁎⁎ p b 0.001.

may also be impeding their release, and concomitant registration on a sex offender registry. Whatever the reasons, this finding warrants further research to disentangle the factors that keep this particular population of sex offender from being charged, convicted, and registered in a way that is consistent with the intent of these laws. While we found little association between past criminal offending in general and characteristics of the victim, the offender, and the abduction itself, more distinct patterns began to emerge when we used multivariate analyses to examine more precise criminal variables, such as criminal versatility and specific categories of crime. For example, criminal versatility among offenders was related to age extremes of the victim, including infancy and ages 15 through 17 years, the offender being between 30 and 59, and the offender being a stranger to the child abduction victim. In contrast, a prior history of property crimes was associated with the victim being an infant, the offender being between the ages of 40 and 49, and the offender being a stranger to the child abduction victim. While these observations reflect statistical patterns and not descriptions of any particular individual, they do begin to provide some contours to the questions at hand when law enforcement is called upon to investigate a case in which little is known except that it is believed that a child of a specific age has been abducted. They also constitute a preview into further empirical study of the different criminal trajectories that tend to be associated with the abduction of a female or male child of a particular age. The majority of child abductions contained in this unique sample of cases that came to the attention of federal law enforcement were motivated by a sexual interest in the child. The proportion of child abduction cases with sexual motives did vary both by gender and across ages by gender, with sexual motivation being the most common for males

(52%), and the majority of females (75%). The two most common age categories of both the girl and boy victims abducted for sexual reasons were 12 to 14 years and 15 to 17 years. These two age categories reflect sexual development that is either pre-pubescent or has reached full sexual maturity, possibly reflecting the dual sexual interests of heterosexual pedophiles and non-paraphilic offenders. Our study documents rather profound gender differences in the risk for being abducted as a child. The great majority of the child victims were females, with the largest portion of the sample being composed of Caucasian females, the majority whom were between the ages of six and 17 years, and who were abducted from either a public outdoor location or their own residence. The perpetrator tended to be either an acquaintance of or a stranger to the child victim and of the same race. However, a sexual intent was the most common motivation found among both the female and male samples and the rates of lethality high for both groups with 55% of the female child victims and 44% of the male child victims being found dead or not found and presumed dead. These findings underscore the largely sexual motivation that appears to prompt these criminal acts and the significant threat of lethality that is associated with them regardless of the gender of the child victim. Not surprisingly, a number of differences were found between the men and women who abducted children. The majority of the perpetrators in our sample were male, with almost one-half of the male offenders falling in the age range of 18 to 29 years. In contrast, the largest proportion of the female offenders fell in the age range of 30 to 39 years. There were more female perpetrators that were African American when compared to male perpetrators, although both male and female perpetrators were predominantly Caucasian. More often

J.I. Warren et al. / Aggression and Violent Behavior 30 (2016) 59–67 Table 8 Predictive effects of victim, offender, and crime scene characteristics on reported prior offending.⁎ β

SE

P

OR

Any adult criminal history (N = 459) Offender age 30–39 Offender age 40–49

1.086 0.897

0.279 0.340

b0.001 0.008

2.96 2.45

Criminal versatility (N = 389) Victim age extremes Offender age 30–39 Offender age 40–49 Offender age 50–59 Relationship-stranger

0.489 0.626 0.873 1.523 0.563

0.230 0.249 0.312 0.562 0.228

0.033 0.012 0.005 0.007 0.013

1.63 1.87 2.39 4.59 1.76

Crimes against adults (N = 384) Victim age 0–1 Victim age 15–17 Offender sex Offender age 30–39 Offender age 40–49 Offender age 50–59 Relationship-stranger

1.043 1.111 −2.352 0.771 0.788 1.416 0.452

0.603 0.301 0.701 0.260 0.328 0.576 0.235

0.084 b0.001 0.001 0.003 0.016 0.014 0.054

2.84 3.04 0.10 2.16 2.20 4.12 1.57

Crimes against children (N = 409) Victim race Offender age 30–39 Offender age 40–49 Offender age 50–59

1.108 0.723 0.688 1.579

0.379 0.291 0.361 0.537

0.003 0.013 0.057 0.003

3.03 2.06 1.99 4.85

Crimes against property (N = 392) Victim age 0–1 Offender age 40–49 Relationship-stranger

1.230 0.672 0.419

0.500 0.319 0.224

0.014 0.035 0.061

3.42 1.96 1.52

Crimes against society (N = 391) Offender age 30–39 Offender age 40–49 Relationship-family or intimate

1.027 0.824 −1.716

0.279 0.343 0.629

b0.001 0.016 0.006

2.79 2.28 0.18

⁎ Note: list-wise case exclusion resulted in differing N sizes for each logistic regression.

the location from which a male child was abducted was unknown, possibly pointing to less supervision of the male child victims. When considering these findings, it is imperative that they be contextualized properly in the nature of the sample that was obtained and utilized. Each of these abductions came to the attention of federal law enforcement either in reference to requests for investigative assistance or after conclusion of the abduction through federal reporting requests or requirements. As a result, these incidents represent the most serious of child abduction crimes, are separate from other circumstances in which a child goes missing, and frequently reflect investigations that were disturbing enough to local law enforcement to prompt

67

a request for further assistance. The relevance of the findings can, therefore, only be generalized to crimes that fall with these same parameters. Because these cases were also coded by individuals other than the first responders, information that might be relevant to the crimes and the victims may also have been lost over the passage of time and geographical distance. Moreover, a number of statistical comparisons were conducted suggesting that some of our findings could be spurious, although the findings did tend to hold to common themes suggesting that they were reflecting real differences in the different attributes of the victim, perpetrator, and the crime. In spite of these caveats, these data do reflect trends that suggest relationships between the past and present behavior of men and women who chose to abduct children, and as such, provide some preliminary empirical guidance to law enforcement when required to respond swiftly to ambiguous circumstance particularly when the safety of a child is at stake. References Beasley, J. O., Hayne, A. S., Beyer, K., Cramer, G. L., Berson, S. B., Muirhead, Y., & Warren, J. I. (2009). Patterns of prior offending by child abductors: A comparison of fatal and nonfatal outcomes. International Journal of Law and Psychiatry, 32, 273–280. Boudreaux, M. C., Lord, W. D., & Dutra, R. L. (1999). Child abduction: Aged-based analyses of offender, victim, and offense characteristics in 550 cases of alleged child disappearance. Journal of Forensic Science, 44, 539–553. Brown, K. M., Keppel, R. D., Weis, J. G., & Skeen, M. E. (2006). Investigative case management for missing children homicide: Report II. (98-MC-CX-0001). Olympia, WA: Attorney General of Washington and U.S. Department of Justice, Office of Juvenile Justice and Delinquency Prevention. Critical Incident Response Group. (n.d.). (Retrieved from) http://www.fbi.gov/about-us/ cirg/investigations-and-operations-support/investigations-operations-support 18 U.S.C. § 1201 Erickson, M., & Friendship, C. (2002). A typology of child abduction events. Legal and Criminological Psychology, 7, 115–120. Federal Bureau of Investigation (2000). National Incident-Based Recording System: Volume 1, data collection guidelines. Clarksburg, WV: Criminal Justice Information Services Division, Federal Bureau of Investigation. Finkelhor, D., & Ormrod, R. (2000, June). Kidnapping of juveniles: Patterns from NIBRS. Juvenile Justice Bulletin (NCJ 181161)Washington, DC: U.S.: Department of Justice, Office of Juvenile Justice and Delinquency Prevention. Finkelhor, D., Hammer, H., & Sedlak, A. J. (2002). Non-family abducted children: National estimates and characteristics. National incidence studies of missing, abducted, runaway, and thrownaway children (NCJ 196467)Washington, DC: U.S.: Department of Justice, Office of Juvenile Justice and Delinquency. Lanning, K. V. (1992). Child molesters: A behavioral analysis (report no. NC-70). Washington, DC: National Center for Missing and Exploited Children. Lanning, K. V., & Burgess, A. W. (Eds.). (1995). Child molesters who abduct: Summary of the case in point series. Washington, DC: National Center for Missing and Exploited Children. Morton, R. J., & Lord, W. D. (2005). Criminal profiling. In J. Payne-James (Ed.), Encyclopedia of forensic and legal medicine (pp. 51–55). Oxford: Elsevier Ltd. National Crime Information Center. (n.d.). (Retrieved from) http://www.fbi.gov/aboutus/cjis/ncic