Female Stalkers and Their Victims

J. Reid Meloy, PhD, and Cynthia Boyd, PhD


Dr. Meloy is Associate Clinical Professor, Department of Psychiatry, University of California, San Diego, and is a graduate of California School of Professional Psychology, San Diego, CA. This study was supported by a grant from Forensis, Inc. Address correspondence to: J. Reid Meloy, PhD, P.O. Box 90699, San Diego, CA 92169. E-mail: jrmeloy@san.rr.com

Demographic, clinical, and forensic data were gathered in an archival study of 82 female stalkers from the United States, Canada, and Australia. Female stalkers were predominantly single, heterosexual, educated individuals in their mid 30s who had pursued their victims for more than a year. Major mental disorder and personality disorder were suggested, especially borderline personality disorder. They usually threatened violence, and if they did threaten, were more likely to be violent. Frequency of interpersonal violence was 25 percent, but there was limited use of weapons, and injuries were minor. Stalking victims were most likely to be slightly older male acquaintances; but if the victim was a prior sexual intimate of the female stalker, her risk of being violent toward him exceeded 50 percent. Unlike male stalkers who often pursue their victims to restore intimacy, these female stalkers often pursued their victims to establish intimacy. Common emotions and motivations included anger, obsessional thoughts, rage at abandonment, loneliness, dependency, jealousy, and perceived betrayal. Results are interpreted from a clinical and risk management perspective.
J Am Acad Psychiatry Law 31:211–19, 2003

Female criminality is rarely studied and little understood. Although the crime of stalking is receiving a growing amount of research attention,1 the 15 to 20 percent of stalkers who are women are usually subsumed by the larger proportion of male stalkers in all research designs.(2–6) Gender differences among stalkers have been studied only once,(7) in an Australian community forensic mental health clinic. Purcell et al.(7) found that male stalkers in that study outnumbered females by a ratio of four to one. Similarities were more frequent than differences in most demographic, clinical, and forensic variables. The females were significantly less likely to have a history of criminal offenses, violent criminal offenses, or substance abuse diagnoses. They were significantly less likely than men to stalk a stranger, but more likely to pursue a prior professional contact, motivated by “a desire to establish a close and loving intimacy with the victim” (Ref. 7, p 2058). The threat and assault rates showed no significant differences, but females were less likely to threaten and then assault. Although the frequency of assault was somewhat higher for the males (36.7%), one of five female stalkers (22.5%) attacked the object of pursuit.
The crime of stalking, typically defined as “the willful, malicious, and repeated following and harassing of another person that threatens his or her safety” (Ref. 2, p 258), is increasingly recognized as a serious social and criminal problem, affecting large proportions of the populations in which it has been studied. Lifetime risks of being the victim of stalking have been measured in the United States, Australia, and Great Britain, and range from 8 to 15 percent for women and 2 to 4 percent for men.(8) If there is one stalker for each victim, then a large random telephone survey of stalking victims in the United States published in 1997 would indicate that 179,135 women had stalked another person in the previous 12 months.(9)

The purpose of this study was to assemble a sample of cases of stalking by females and study their various demographic, clinical, and forensic characteristics. Given the relatively small number of women charged with or convicted of stalking in any one jurisdiction, we attempted to gather cases from as many different collaborators in as many different geographical locations as possible, to increase the external validity of our findings.

The study design was an archival survey of mental health and law enforcement professionals. A static group comparison was used to test certain hypotheses within the sample. Research collaborators who were professionals with expertise and experience in stalking cases were sought from the United States, Canada, and Australia between June 2000 and March 2001. These professionals were known to the first author, in some cases through their research publications. Stalking cases were included based on the following selection criterion: an adult female (18 or more years of age) who had engaged in stalking behavior as defined by the California Penal Code: “. . .any person who willfully, maliciously, and repeatedly follows or harasses another person and who makes a credible threat with the intent to place that person in reasonable fear for his or her safety, or the safety of his or her family.”(10) Specific charges or convictions were not necessary for inclusion, but the behavior had to have occurred on more than one occasion.
The survey instrument was developed by the authors based primarily on published research studying male perpetrators. It consisted of 140 categorical questions distributed across 11 sections: (1) perpetrator demographics; (2) psychiatric and psychological information; (3) use of drugs or alcohol; (4) criminal history; (5) patterns of pursuit; (6) emotions and motivations for stalking; (7) threats; (8) violence; (9) escalation; (10) victim characteristics; and (11) victim response. The survey instrument is available from the authors. Because of the archival nature of the study, the subject’s consent was unnecessary. Subjects were not identified by name. Data were analyzed on computer with SPSS 9.0 for Windows. Frequencies and percentages were calculated based on the sufficiency of the data to determine the presence or absence of each variable. Inferential hypotheses were tested using chi-square analysis. Statistical significance was set at .05.

Thirty-nine individuals responded to requests for participation as our research collaborators. Thirtythree (84.6%) completed and returned the survey instrument: 20 mental health professionals, 8 law enforcement professionals, 1 private security professional, 3 stalking victims, and 1 relative of a stalking victim. Data used to complete the research instrument were gathered from the case files of the research collaborators. Eighty-two cases of stalking by females were entered into our database from the United States (n=62), Canada (n=10), and Australia (n= 10).

Perpetrator Demographics
The subjects’ ages ranged between 18 and 58 years (mean, 37; SD, 9.50). Most were white (n=61; 77%), and the remainder were African American (n= 6; 8%), Asian (n= 5; 6%), Hispanic (n= 4; 5%), and other (n= 3; 4%). (When the total number does not reach 82, data for the particular variable were missing.) Eighty-eight percent (n= 68) had achieved at least a high school degree. Thirty-eight percent had a college or graduate degree. Although reliable IQ data were available on only 24 subjects, 96 percent had an estimated average to superior IQ (n= 23). Eighty percent of the subjects were heterosexual (n=60), 8 percent lesbian (n= 6), and 12 percent bisexual (n= 9). Fifty-eight percent were single (n= 45), 13 percent were married (n= 10), and 21 percent were divorced (n=16). Sixty-seven percent were childless (n= 55), and 33 percent had one or more children (n= 27). Seven percent had three to five children (n= 6). Forty-five percent (n=18) had a history of sexual abuse, 30 percent (n= 12) had experienced physical abuse, and 8 percent (n= 3) had suffered emotional trauma. However, data on these abuse variables were missing for half of the subjects.

Psychiatric and Psychological Information
Only a subsample of subjects could be reliably evaluated for Axis I (n= 24) and Axis II (n= 22) diagnoses. Axis I diagnoses, in descending order of frequency, included delusional disorder (n= 5), major depression (n= 4), schizophrenia (n= 3), dysthymia (n= 2), bipolar disorder (n= 2), adjustment disorder (n= 2), schizoaffective disorder (n= 2), and psychotic disorder not otherwise specified (NOS) (n= 2). Axis II diagnoses, in descending order of frequency, included borderline personality disorder (n= 10), narcissistic personality disorder (n= 3), dependent personality disorder (n= 3), personality disorder NOS (n= 2), and obsessivecompulsive personality disorder (n= 1). One female had a diagnosis of more than one personality disorder.
Half of the females (49%) who were psychiatrically evaluated at the time they engaged in stalking (n= 39) were judged to be psychotic. The most prominent psychotic symptoms were delusions (n= 14). One-third (n= 27) of the women had at least one inpatient psychiatric admission by history. In the year prior to the stalking, 38 percent of the women (n= 25) had suffered at least one major personal loss, usually a relationship. Seventeen percent (n= 11) reported multiple losses, such as a relationship, finances, child custody, and a home.

Use of Drugs or Alcohol
Three-fourths of the women (n= 49) denied a substance abuse history. One-third of the women, however, reportedly used the following substances while they were stalking, in descending order of frequency: alcohol (n= 17), cannabis (n= 10), amphetamines (n= 4), opiates (n= 1), and sedativehypnotic drugs (n= 1). There was no reported use of cocaine, hallucinogens, phencyclidine (PCP), or steroids.

Criminal History
Twenty-eight (37%) of the female stalkers had an adult criminal history prior to stalking, but only three (7%) had an available juvenile history of crime. Sixteen percent (n= 13) had one to four prior arrests for stalking other individuals.

Patterns of Pursuit
The female stalkers were most often prior acquaintances of the victim (n= 40, 50%). Twenty-seven percent (n= 22) were prior sexual intimates of the victim, 21 percent (n= 17) were strangers, and 2 percent (n= 2) were family members. Various means were used to pursue their victims in these contexts and are listed in Table 1. Multiple approaches were used by most of the stalkers. Slightly less than half of the subjects (n= 40, 49%) actually followed their victims.

Emotions and Motivations
Table 2 presents the emotions and motivations for the female stalkers, as reported by the research collaborators from police reports, victims’ reports, witness testimony, psychotherapists’ reports, and selfreports. The categories were derived from the extant research. More than one emotion or motivation was reported in most cases.

A threat was defined as a “written or oral communication that implicitly or explicitly states a wish or intent to damage, injure, or kill the target” (Ref. 11, p 90). Sixty-five percent (n= 50) of the female stalkers threatened their victims. Forty-one women threatened orally, 23 wrote threats, and 34 threatened multiple times. Nineteen women made homicidal threats.
Threats were divided into two categories: expressive and instrumental.(11) Expressive threats are primarily used to regulate affect in the threatener and are often inferred through the ventilation of anger or frustration (an example from our study: “I’ll wreck his business!!!”). Instrumental threats are used to control or influence the behavior of the victim through an aversive outcome (an example from our study: “Don’t ever call me crazy. I don’t want to hurt you like I did the others”). Forty women made expressive threats, and 42 made instrumental threats. Most women made both expressive and instrumental threats. The threats were followed by personal violence toward the victim in 15 (30%) cases, which is the true-positive rate. The false-positive rate would therefore be 70 percent. There was, however, a falsenegative rate of 15 percent (n= 3), which refers to the female stalkers who became personally violent who did not communicate a threat beforehand. There was no significant relationship between prior sexual intimacy and the presence of a threat (X2 [1, n = 77] = 1.16, p = .281).

Violence Toward Person and Property
Violence was defined as “an intentional act of aggression directed toward a specific human being or property which physically injures or damages, or is likely to physically injure or damage, the person or property” (Ref. 12, p 6). Twenty-five percent (n= 20) of the female stalkers were physically violent toward the victims.
Other aspects of violent behavior were reported for only a small subsample of women and are represented in Table 3. In most cases when data were available, the injury to the victim did not require medical care. When a weapon was used, it was usually a knife, gun, or automobile. Other weapons included a brick, rock, purse, book, garden hoe, and portable electric fan. There were 83 victims in this study, and 3 were killed by their female stalkers. In most cases the violence was affective (emotional, reactive) rather than predatory (unemotional, planned).(13–14) All of the homicides were predatory when analyzed according to forensic criteria.(14)
There was a significant relationship between prior sexual intimacy and violence toward the victim (X2 [1, n= 77] = 10.62, p < .05). The strength of this relationship was moderate (Ø =.37, p<.05). Fiftyfive percent of the prior sexual intimates were violent; 15 percent of the remainder of the sample were violent. Acquaintance status and violence toward the victim was also significantly related (X2 [1, n=79]= 4.02, p < .05). The strength of this relationship was mild (Ø = .23, p < .05). There was no significant relationship between stranger status and violence. The presence of a threat and subsequent acts of violence toward the victim was positively and significantly related (X2 [1, n= 76] = 4.30, p < .05). The strength of this relationship was mild (Ø = .24, p < .05).

The female stalkers had contact with their victims daily (n= 37, 52%), weekly (n= 24, 34%), or monthly (n= 10, 14%). The duration of stalking lasted less than 1 year in 23 (32%) cases ; from 1 to 5 years in 38 (54%) cases; from 6 to 10 years in 9 (13%) cases; and more than a decade in 1 case. Escalation was defined by an increase in frequency of contact during the course of stalking (n= 51, 66%) and an increase in intrusiveness (n= 55, 73%). For example, the number of letters sent would increase over a set period of time (frequency), or letter writing as a method of pursuit would be supplanted by physical following (intrusiveness). Although the stalking activity of the majority of female stalkers escalated according to our definition, we did not measure the magnitude of the intrusiveness or the rate of increase in frequency.

Victims’ Characteristics
The victims of the female stalkers were mostly males (n= 52, 67%). Ninety percent (n= 72) were white, five percent (n= 4) were Hispanic, two percent (n= 2) were African American, and one percent (n= 1) was Asian. Their ages ranged from 16 to 68 years (mean, 41.30; SD, 11.08). Ninety-four percent (n= 74) were heterosexual. Forty-six percent (n= 36) were married, 17 percent (n= 13) were divorced, and 27 percent (n= 22) were single. Half of the victims had no dependents (n= 43, 52%). Victims’ Response to Stalking Eleven of the victims (six females and five males) had been stalked in the past, usually by a stranger (n= 7). Only two were stalked previously by a prior sexual intimate and three by an acquaintance. Onethird of the victims initiated contact with the female stalkers after the stalking began (n= 22), which subsequently increased the stalking behavior in 68 percent of these cases (n= 15). It reportedly decreased stalking behavior in one case and had no effect in six cases. A complete list of victims’ responses is listed in Table 4.
Ninety-seven percent of the victims were reportedly aware that they were ultimately responsible for their own personal safety, and 90 percent reported positive treatment by law enforcement.

This study has several limitations. We were not able to evaluate most of the female subjects clinically, although 20 mental health professionals provided clinical data on a subsample. We were dependent on the observation and interpretation of data from other professionals without standard measures of reliability.
Our clinical data are particularly weak, given the absence of structured interviews and psychological testing for most subjects.
As with all archival studies, we had missing data and were therefore precluded from making statistical predictions. As with all survey studies, we had no control or comparison group. Given the small number of female stalkers available anywhere for study, our 82 subjects were selected nonrandomly, limiting the generalizability of our findings. We attempted to compensate for this limitation by drawing subjects from many different geographical locations in three different countries. There were also no medical evaluations of our subjects, an important limitation given the presence of organic and/or neurological problems in two studies of a small sample of male stalkers.(24,33) Despite these shortcomings, we think our findings establish some important groundwork for further research.
Female stalkers appear to be similar to male stalkers in age and relationship status: most appear to be single women in their mid-30s. Purcell et al.(7) found the median age of their sample to be 35, and male stalkers are consistently found to be in their mid 30s and unmarried.(2–6) A history of failed sexual pair bonds appears to mark both male and female stalkers. The crime of stalking, moreover, is labor intensive, and it is a testament to either their desperation or efficiency, or both, that one-third of the female stalkers were probably raising children at the same time. Once again, education and IQ appear to be higher among female stalkers than female criminals in general, a finding that has been replicated in male samples. (6) Female stalkers appear to be smart and educated, which may translate into a criminal sophistication that, despite their other disabilities, may contribute to the successful pursuit of their male victims. Although the physical and sexual abuse data are limited, such abuse is disturbingly prominent and similar to findings in other samples of female criminals. (15) It also suggests a traumagenic rather than constitutional basis for the high frequency of borderline personality disorder among the female stalkers(16) and raises the probability of post-traumatic stress disorder (PTSD).
Although our psychiatric data were also limited, when evaluations were available, most of the women had both Axis I and Axis II diagnoses. As in males, Axis I conditions varied among the women and clustered around depression, psychosis, or both. The absence of any anxiety disorders, particularly PTSD, was unexpected and inconsistent with the prevalence of borderline personality disorder. One in five women with a verifiable Axis I diagnosis had delusional disorder, a rare condition even among psychiatric patients, but one that may occur more frequently among female than male stalkers. Delusional disorder has been shown to be more frequent in females than in males in civil mental health settings.(17) Axis II disorders appear to follow the pattern among men, with Cluster B personality disorders most likely. Borderline personality disorder was expected and found for three reasons: first, it is most often diagnosed in women; second, it is defined in DSM-IV by its intensity, instability, and a fear of abandonment; and third, it correlates with an underlying “preoccupied” attachment pathology in the domestic violence and stalking literature concerning men.(18) Object relations theory suggests that individuals with preoccupied attachments hold a positive emotional valence toward others (idealization) and a negative emotional valence toward themselves (devaluation). (19) Their internal representations are part object related and preoedipal.(20) They are likely organized at a borderline level of personality.(21)
Suggestively absent among the female stalkers is antisocial personality disorder (ASPD). This finding is similar to the low rate of ASPD (and psychopathy by inference) among male stalkers—usually less than 10 percent.(6) Typical rates of ASPD among male offenders are 50 to 75 percent.(22) These findings are likely a product of gender differences for the ASPD diagnosis and the absence of a dismissive attachment pathology: an internal working model in which others are devalued and the self is idealized, a narcissistic stance from which others are used, exploited, and discarded, rather than desperately sought.(23) A significant personal loss for many of the women in our sample during the year before the onset of stalking may have further precipitated such pursuits, or at least exacerbated the underlying, insecure attachment style.(18) Kienlen et al.(24) found that 80 percent of a small sample of imprisoned male stalkers had had a significant personal loss (relationship or employment) in the 7 months preceding the onset of stalking. Most of the subjects in Kienlen et al. also suffered a change or loss of a primary caretaker during childhood.
Prior research in male stalkers indicates that most knew their victims, and the most common prior relationship was sexual intimacy (45–50%).(2–6) In our study the most common prior relationship was “acquaintance,” while prior sexual intimacy had occurred in only one of four cases. The female stalker was a complete stranger to the victim in one of five cases. Purcell et al.(7) had almost exactly the same finding for sexual intimacy in their sample of 40 female stalkers, 10 of whom are included in our study. Other studies of male stalking victims are consistent with these findings. Hall (25) reported that two-thirds of male victims in her study were stalked by prior acquaintances, not intimates, and Tjaden and Thoennes (9) found that males tend to be stalked by female acquaintances and strangers, not by prior sexual intimates. These results, which are a significant departure from those showing that victims of most male stalkers are prior sexual intimates, provide suggestive support for the theory of Purcell et al.(7) that female stalkers are motivated to establish intimacy with their victims, whereas male stalkers are attempting to maintain intimacy with their former partners.
Pursuit behavior of the females was multiple and varied, as was true of male stalkers.(4,6,8) It appears, however, that the most frequently prohibited act in stalking laws, following the victim, is engaged in less often by females than by males. Our finding of 49 percent following their victims is virtually identical with the 50 percent who followed their victims in Purcell et al.,(7) which, in turn, was significantly less than their male comparative sample. Among male stalkers, 48 to 80 percent follow their victims.(2–6) Perhaps unwanted following is more overtly aggressive, therefore more likely to be a male stalking behavior.
The women, moreover, appear to be creatively aggressive in more covert ways: they intrude on the victim’s associates, vandalize property, use surveillance, break and enter, and steal the victim’s possessions. All of these patterns do not risk a direct physical confrontation with the male victim, at least for the moment.
What motivates female stalkers? Anger and hostility were reported in two-thirds of our sample, as it was in 55 percent of a previous sample of male stalkers. (6) The commonality of obsession, perhaps a cognitive corollary of anger in these women, gives further credence to the prominent role of preoccupied thoughts among all stalkers.(1,4,8) Rage caused by abandonment in half the sample is consistent with a diagnosis of borderline personality disorder and may reactivate early feelings and fears, whether real or imagined, of parental neglect. Such rage toward the victim is likely to be based on transference perceptions and feelings, perhaps psychotic ones, since at least half of the abandonment-related rage was reported in our subjects who had no prior intimate relation with their victims.
Other feelings common to the ennui of human experience—loneliness, dependency, jealousy, a desire to retaliate, and a need for power and control— are present among female stalkers, but may need a diagnosable psychiatric condition to be acted out in aberrant and unwanted behavior. Such conditions may defeat one lesson to be learned in the course of maturation: internal states can be felt without being acted on. As in other studies of stalking,(6) explicit sexual motivations appear to be unusual.
Although the fertile ground for stalking by males—social isolation, loneliness, and social incompetence— is also suggested by our data for females, the role of pathological narcissism, specifically entitlement and grandiosity in the form of narcissistic linking fantasies, seems less.(11) The women seem more intent on forming an attachment, regardless of the negative consequences, to assuage feelings of loneliness, dependency, and anger, rather than to restore a narcissistically idealized relationship, which often characterizes male stalkers.
Female stalkers threaten their victims at about the same rate as males (50–75%).(26) We also found a greater likelihood of violence if a threat was communicated by the female stalker to her victim. The strength of this relationship was mild, however—a finding similar in male stalkers.(5,12) This finding underscores an important risk management axiom: threats in private stalking cases increase risk, but because they are so common they should not be used to predict violence. An analysis of verbalized threats should be secondary to an analysis of behavior that may pose a danger to the victim.(27)
We discovered that threats issued by female and male stalkers are acted on at a similar rate. Truepositive rates for males are 25 to 35 percent.(28) The true-positive rate for this sample of female stalkers for personal violence was 30 percent. The false-negative rate (no threat followed by violence) is low, as expected and within the range of that of male stalkers (13–23%; Ref. 28) of private parties, usually prior sexual intimates. If a female stalker does not threaten, there appears to be a one in seven chance that she will attack the victim.
The frequency of interpersonal violence among male stalkers is 25 to 40 percent.(28) When prior sexual intimates are the victims, violence frequency substantially exceeds 50 percent.(5,12) Female stalkers are no different. When prior sexual intimates were victimized, the frequency of violence substantially increased to 55 percent. Purcell et al.(7) reported an assault rate of 22.5 percent for all of their female stalkers, almost identical to our overall finding for interpersonal violence.
The good news is that the violence was not serious in most cases, did not require medical care for the injuries sustained, and did not involve a weapon (see Table 3). Even when a weapon was used, our limited data suggest that it did not cause injury in two of three cases. This finding is similar to that of Meloy et al.(12) who reported no injury when a weapon was used in a sample of male stalkers (n= 59). Their finding suggested that weapons were used to frighten, intimidate, or control, rather than to injure the victim. Female stalkers may have a similar motivation when they brandish a knife or a gun.
Violence research involving study of male stalkers has also confirmed a strong and significant relationship between risk of violence and prior sexual intimacy with the victim.(12) One study found that it predicted violence correctly 97 percent of the time.(12) We tested this hypothesis with female stalkers by using nonparametric comparisons and found the same, albeit weaker, relationship. Interpersonal violence had a moderate association with prior sexual intimates, a mild association with acquaintances, and no association with strangers. It appears that risk of violence weakens as the interpersonal relationship between parties is more distant. The more actual detachment in the relationship between female stalker and victim, the more idealized the relationship may remain, with room for the victim to maneuver psychologically to avoid direct anger and hostility. For example, third parties may provide a convenient target for the displacement of aggression(29) in the face ofactual rejection. Psychotic diagnoses are also significantly more prevalent among male stalkers who pursue strangers(6,24) (we had insufficient data to test this in our female sample), and female stalkers who are delusional, perhaps erotomanically so, could more easily maintain their irrational beliefs in the midst of actual rejection. Whatever the basis for continuous idealization of the victim, distance and delusion could also diminish the intensity of emotions of anger, jealousy, or fear of abandonment, the fuel for affective violence.(11)
The estimated homicide rate among male stalkers is 0.25 percent.(30) We had two subjects who intentionally killed three victims in this study, but our sample is too small to construct a valid homicide rate for female stalkers. Both women lived in San Diego County, California; were prior sexual intimates of the victims; and committed acts of predatory violence that were planned, purposeful, and emotionless. In the first case a 26-year-old woman threatened and intruded on her former boyfriend for more than a year, rented an apartment near him, and shot him to death with a .357 magnum revolver in front of his apartment. Psychological testing indicated an average IQ. Her evaluating psychologist and psychiatrist concurred that she had a borderline personality disorder with antisocial features.(31) In the second case, a 41-year-old woman engaged her former husband in five years of vandalism, threats, telephone calls, trespassing, property damage, and financial and child custody disputes. She then shot and killed him and his new wife with a .38 caliber revolver in the early morning hours as they lay sleeping in their bed.(31) At trial, her diagnosis was major depression and borderline personality disorder with narcissistic and histrionic features. One of us corresponded with this individual during the course of this study, and she wrote of her victim, “[He] was a sick bastard with a lot of power with the crooked judges who ran those courts in the 1980s.” She had been in prison for more than a decade.
All violent offenders of both genders tend to target same-gender victims 70 to 75 percent of the time.(15) This was not the case in our study. The violent female stalkers targeted males 67 percent of the time. Our findings suggest that stalking, whether violent or not, is a crime that usually occurs between sexes, regardless of the gender of the perpetrator, and samegender stalking is the exception to the rule.(8) The abnormal desire to establish a heterosexual pair bond with an unwilling person may subsume the violent patterns seen among other kinds of female criminals.
Most female stalkers pursue their victims for one to five years. Although we had no data to determine the mean length of the sample’s pursuits, this range falls within the expected duration of most stalking cases, which is 1.8 years.(9) In Mullen et al.(8) a large sample of Australian stalkers, both men and women, had durations of stalking ranging from 8 to 38 months. Their female stalkers had a median duration of stalking of 22 months.(7)
Victims of female stalkers, on average, were men at least a decade older than the female victims of male stalkers. Tjaden and Thoennes(9) reported an average age of 28 in their random probability telephone survey in the United States of victims of stalking. Female stalkers and their victims are almost chronological peers, the women averaging 37 to 38 years and the men averaging 40 to 41 years of age. There is an evolutionary symmetry to all these data, even though the behavior is maladaptive; most men pursue younger women for sexual pair bonding, and most women prefer older men for sexual pair bonding.(32) Women are more likely than men, however, to engage in same-sex stalking. One-third of the victims in our study were women, while Purcell et al.(7) found that almost half of the victims in their study were women. The reason for these findings remains elusive.
The victims responded to their female stalkers with a range of coping behavior (see Table 4) that reflect attempts to avoid the stalker, legally limit her ability to approach the victim, and actively help law enforcement build a criminal case against her. Our data present an unduly optimistic picture, however, because most victims of stalking, whether male or female, do not report the incidents to the police.(9) In our study, the active victim response is a product of selection bias, since virtually all our subjects were provided to us by law enforcement or mental health.
We do not know how well these strategies worked, but we do know that direct confrontation of the female stalker by her victim usually did not. Among the 22 victims who initiated contact with the female stalker after it began, the stalking increased 68 percent of the time. This is the first empirical support for the assertion1,(26) that any type of initiated contact, regardless of the words or affect exchanged, is an intermittent positive reinforcement and is likely to increase the frequency of subsequent stalking behavior.

We are deeply grateful to the following individuals for their collaboration in our study: Stacy Baradino, Kimberly Bistis, Paul Bristow, Susan Crabtree, Joe Davis, Dennis Donovan, Gary Farkas, Susan Fisher, Larry Froley, James Johnson, Kristine Kienlen, Steve Lang, Suzanne MacKaye, Wayne Maxey, Dave McMullen, Paul Mullen, Doreen Orion, Vanessa Payne, Linda Pounds, Mike Proctor, James Reavis, Jill Ricke, Diana Riehm, Ann Rose, Bernie Schell, Glenn Sheil, Christine Taylor, Ann Troy, Judy Tyler, Kerry Wells, and Rick Wildman.

1. Meloy JR (editor): The Psychology of Stalking: Clinical and Forensic Perspectives. San Diego, CA: Academic Press, 1998
2. Meloy JR, Gothard S: A demographic and clinical comparison of obsessional followers and offenders with mental disorders. Am J Psychiatry 152:258–63, 1995
3. Harmon R, Rosner R, Owens H: Sex and violence in a forensic population of obsessional harassers. Psychol Public Policy Law 4:236–49, 1998
4. Zona M, Sharma K, Lane J: A comparative study of erotomanic and obsessional subjects in a forensic sample. J Forensic Sci 38: 894–903, 1993
5. Palarea R, Zona M, Lane J, et al: The dangerous nature of intimate relationship stalking: threats, violence and associated risk factors. Behav Sci Law 17:269–83, 1999
6. Meloy JR, Rivers L, Siegel L, et al: A replication study of obsessional followers and offenders with mental disorders. J Forensic Sci 45:189–94, 2000
7. Purcell R, Pathe M, Mullen P: A study of women who stalk. Am J Psychiatry 158:2056–60, 2001
8. Mullen P, Pathe M, Purcell R: Stalkers and Their Victims. London: Cambridge University Press, 2000
9. Tjaden P, Thoennes N: Stalking in America: Findings from the National Violence Against Women Survey. Denver, CO: Center for Policy Research, 1997
10. California Penal Code, § 646.9 (West 2002)
11. Meloy JR: Stalking: an old behavior, a new crime. Psychiatr Clin North Am 22:85–99, 1999
12. Meloy JR, Davis B, Lovette J: Violence risk factors among stalkers. J Threat Assess 1:1–16, 2001
13. Meloy JR: The Psychopathic Mind: Origins, Dynamics and Treatment. Northvale, NJ: Aronson, 1988
14. Meloy JR: Predatory violence during mass murder. J Forensic Sci 42:326–9, 1997
15. Greenfield L, Snell T: Women Offenders (Bureau of Justice Statistics Special Report NCJ 175688). Washington, DC: U.S. Department of Justice, 1999
16. Herman J, van der Kolk B: Traumatic antecedents of borderline personality disorder, in Psychological Trauma. Edited by van der Kolk B. Washington, DC: American Psychiatric Press, 1987, pp 111–26
17. Rudden M, Sweeney J, Frances A: Diagnosis and clinical course of erotomanic and other delusional patients. Am J Psychiatry 147: 625–8, 1990
18. Dutton D: Intimate abusiveness. Clin Psychol 2:207–24, 1995
19. Bartholomew K: Avoidance of intimacy: an attachment perspective. J Soc Pers Rel 7:147–78, 1990
20. Meloy JR: Pathologies of attachment, violence, and criminality, in Comprehensive Handbook of Psychology (vol 11). Edited by Goldstein A. New York: John Wiley, 2002, pp 509–26
21. Kernberg O: Severe Personality Disorders. New Haven: Yale University Press, 1984
22. Hare RD: Manual for the Psychopathy Checklist, Revised. Toronto: Multihealth Systems, 1991
23. Meloy JR: The Mark of Cain: Psychoanalytic Insight and the Psychopath. Hillsdale, NJ: The Analytic Press, 2001
24. Kienlen K, Birmingham D, Solberg K, et al: A comparative study of psychotic and non-psychotic stalking. J Am Acad Psychiatry Law 25:317–34, 1997
25. Hall D: The victims of stalking, in The Psychology of Stalking: Clinical and Forensic Perspectives. Edited by Meloy JR. San Diego, CA: Academic Press, 1998, pp 115–37
26. Meloy JR: Stalking (obsessional following): a review of some preliminary studies. Aggress Violent Behav 1:147–62, 1996
27. Fein R, Vossekuil B: Assassination in the United States: an operational study of recent assassins, attackers, and near-lethal approachers. J Forensic Sci 44:321–33, 1999
28. Meloy JR: Communicated threats and violence toward public and private targets: discerning differences among those who stalk and attack. J Forensic Sci 46:1211–13, 2001
29. Meloy JR: Erotomania, triangulation, and homicide. J Forensic Sci 44:421–4, 1999
30. Meloy JR: Stalking and violence, in Stalking and Psychosexual Obsession. Edited by Boon J, Sheridan L. London: John Wiley, 2002, pp 105–24
31. Meloy JR: Violent Attachments. Northvale, NJ: Aronson, 1992
32. Buss DM: Sex differences in human mate preferences: evolutionary hypotheses tested in 37 cultures. Behav Brain Sci 12:1–49, 1989
33. Schwartz-Watts D, Morgan D: Violent versus nonviolent stalkers. J Am Acad Psychiatry Law 26:241–5, 1998