National Institute of Justice Centers for Disease Control and Prevention

U.S. Department of Justice Office of Justice Programs National Institute of Justice

National Institute of Justice Centers for Disease Control and Prevention R e s e a r c h i n B r i e f April 1998 Stalking in America: Findings From the National Violence Against Women Survey Partners in Research on Violence Against Women by Patricia Tjaden and Nancy Thoennes

We are pleased to publish this first in Unprecedented interest in stalking over The data are from the National a series of joint reports on violence Violence Against Women (NVAW) against women. the past decade has produced media accounts of stalking victims,1 passage of Survey, a nationally representative When the National Institute of Justice antistalking laws in all 50 States and the telephone survey of 8,000 U.S. (NIJ) and the National Center for 2 women and 8,000 U.S. men (see Injury Prevention and Control (NCIPC) District of Columbia, and development launched our partnership to study of a model antistalking code.3 Despite “Survey Methodology and Demo- violence against women, there were graphic Description of the Sample,” no national data on stalking and its this interest, research on stalking has impact. This first-ever national survey been limited to studies of small, unrep- page 15). The survey, which asked has already made a major contribu- resentative, or clinical samples of known detailed questions about respondents’ tion. We eagerly await findings from 4 experiences with violence, including other aspects of the study and stalkers; law journal reviews of the anticipate reports on the incidence constitutionality and effectiveness of stalking, was sponsored jointly by the and prevalence of violence, partner 5 National Institute of Justice and the violence, and rape. specific antistalking statutes; and case studies of individual stalkers.6 Thus, Centers for Disease Control and This NIJ/NCIPC partnership to learn empirical data have been lacking on Prevention through a grant to the more about violence against women Center for Policy Research. is based on a research agenda such fundamental questions about developed by the National Academy stalking as: of Sciences, which was mandated by What is stalking? the Violence Against Women Act • How much stalking is there in the (Title IV of the Crime Act of 1994) and supported by NIJ and CDC. United States? Stalking generally refers to harassing • Who stalks whom? or threatening behavior that an indi- Joint research initiatives stimulate a vidual engages in repeatedly, such as rich cross-fertilization of ideas and • How often do stalkers overtly threaten bring interdisciplinary perspectives to following a person, appearing at a their victims? our knowledge base. The interests of person’s home or place of business, both NCIPC and NIJ in this area are • How often is stalking reported to the grounded in their common focus on making harassing phone calls, leaving social policy. NIJ brings a criminal police? written messages or objects, or vandal- justice perspective; NCIPC examines the issue from a public health and • What are the psychological and social izing a person’s property. These prevention perspective. consequences of stalking? actions may or may not be accompa- nied by a credible threat of serious We hope the result will be a deeper and broader understanding of the This Research in Brief presents data from harm, and they may or may not be implications of this violence and the the first-ever national study on stalking precursors to an assault or murder.7 effective policy response. and addresses these and related questions. Since the data show stalking to be much Legal definitions of stalking vary widely Jeremy Travis, Director National Institute of Justice more prevalent than previously thought from State to State. Though most States and include other findings of broad public define stalking as the willful, malicious, Mark Rosenberg, Director concern, they have significance for legisla- and repeated following and harassing of National Center for Injury Prevention and Control, Centers tors, policymakers, intervention planners, another person, some States include in for Disease Control and Prevention and researchers as well as the public their definition such activities as lying- health and criminal justice communities. in-wait, surveillance, nonconsensual Research in Brief

Issues and Findings

Discussed in this Brief: Results of a • Although stalking is a gender-neutral • About half of all stalking victims report nationally representative telephone survey of crime, most (78 percent) stalking victims are their stalking to the police. About a quarter of 8,000 women and 8,000 men about their female and most (87 percent) stalking stalking cases reported to the police result in experiences with stalking, cosponsored by perpetrators are male. suspects being arrested. While there is some the National Institute of Justice and the evidence that antistalking laws have increased Centers for Disease Control and Prevention • Adults between 18 and 29 years old are reports to the police, more research is needed and conducted by the Center for Policy the primary targets of stalking, comprising 52 to determine antistalking laws’ full effect on Research. The survey provides the first percent of all victims. reports to the police. national data on stalking in the United • Most stalking cases involve perpetrators States. • About 12 percent of all stalking cases and victims who know each other; 23 percent result in criminal prosecution, and about a Key issues: This study provides empirical of all female victims and 36 percent of all quarter of female stalking victims and about a data on the prevalence and characteristics of male victims are stalked by strangers. tenth of male stalking victims obtain restrain- stalking in the general population: How ing orders against their stalkers. Of all victims • Women are significantly more likely than much stalking is there in the United States? with restraining orders, 69 percent of the men (59 percent and 30 percent, respectively) Who stalks whom? How often do stalkers women and 81 percent of the men said their to be stalked by intimate partners, about half overtly threaten their victims? How often is stalkers violated the order. More research is of whom stalk their partners while the stalking reported to the police? What are the needed on the effectiveness of formal and relationship is intact. Since most stalking psychological and social consequences of informal justice system interventions in cases involve victims and perpetrators who stalking? Also considered in this report is the stalking cases. know each other, future research should key issue of how to define stalking. focus on intimate and acquaintance stalking, • Thirty percent of female stalking victims Key findings and policy implications: rather than “celebrity” stalking. and 20 percent of male stalking victims seek Analysis of survey data produced the psychological counseling as a result of their • There is a strong link between stalking following results: victimization. Stalking victims are significantly and other forms of violence in intimate more likely than nonstalking victims to live in relationships: 81 percent of women who • Stalking is more prevalent than previously fear for their personal safety and to carry were stalked by a current or former husband thought: 8 percent of women and 2 percent something to defend themselves. To better or cohabiting partner were also physically of men in the United States have been stalked meet the needs of stalking victims, the mental assaulted by that partner and 31 percent at some time in their life; an estimated health community should receive comprehen- were also sexually assaulted by that partner. It 1,006,970 women and 370,990 men are sive training on appropriate treatment of is imperative, therefore, that America’s stalked annually. Given these findings, stalking victims. stalking should be treated as a legitimate criminal justice community receive compre- criminal justice and public health concern. hensive training on the special safety needs of • The average stalking case lasts 1.8 years. victims of intimate partner stalking. Since nearly a fifth of all stalking victims move • American Indian/Alaska Native women to new locations to escape their stalkers, it is • Less than half of all stalking victims are are significantly more likely to report being important that address confidentiality directly threatened by their stalkers, although stalked than women of other racial or ethnic programs be made available to stalking victims. backgrounds. More research is needed to the victims, by definition, experience a high establish the degree of variance and level of fear. Thus, “credible threat” require- Target audience: Criminal justice and public determine how much of the variance may be ments should be eliminated from the health researchers and practitioners. Legisla- explained by demographic, social, and definition of stalking in all State stalking tors, policymakers, and intervention planners environmental factors. statutes. at all levels of government. communication, telephone harass- make a credible threat of violence developed by the National Institute ment, and vandalism.8 While most against the victim; others include in of Justice.11 The survey defines States require that the alleged stalker their requirements threats against the stalking as “a course of conduct engage in a course of conduct show- victim’s immediate family; and still directed at a specific person that ing that the crime was not an isolated others require only that the alleged involves repeated visual or physical event, some States specify how many stalker’s course of conduct constitute proximity, nonconsensual communi- acts (usually two or more) must occur an implied threat.10 cation, or verbal, written or implied before the conduct can be considered threats, or a combination thereof, stalking.9 State stalking laws also The definition of stalking used in the that would cause a reasonable person vary in their threat and fear require- NVAW Survey closely resembles the fear,” with repeated meaning on two ments. Most stalking laws require that definition of stalking used in the or more occasions. The model the perpetrator, to qualify as a stalker, model antistalking code for States antistalking code does not require

2 Research in Brief stalkers to make a credible threat of Exhibit 1. Percentage and Estimated Number of Men and Women Stalked violence against victims, but it does in Lifetime require victims to feel a high level of fear (“fear of bodily harm”). Simi- Persons Stalked in Lifetime larly, the definition of stalking used Group Percentagea Estimated Numberb in the NVAW Survey does not require stalkers to make a credible Men (N = 8,000) 2.2 2,040,460 threat against victims, but it does Women (N = 8,000) 8.1 8,156,460 require victims to feel a high level of a. Differences between men and women are significant at ≤ .001. fear (see “Survey Screening Ques- b. Based on estimates of men and women aged 18 years and older, U.S. Bureau of the Census, Current tions,” page 17). Population Survey, 1995.

How much stalking is there? Exhibit 2. Percentage and Estimated Number of Men and Women Stalked In the NVAW Survey, stalking in Previous 12 Months victimization was measured in terms Persons Stalked in Previous 12 Months of lifetime prevalence and annual prevalence. Lifetime prevalence Group Percentagea Estimated Numberb refers to the percentage of persons Men (N = 8,000) 0.4 370,990 within a demographic group (e.g., Women (N = 8,000) 1.0 1,006,970 male or female) who were stalked sometime in their lifetime. Annual a. Differences between men and women are significant at ≤ .001. prevalence refers to the percentage of b. Based on estimates of men and women aged 18 years and older, U.S. Bureau of the Census, Current Population Survey, 1995. persons within a demographic group who were stalked sometime in the 12 months preceding the survey. percent of male victims were stalked years comprise nearly half (47 by three different persons. Using a definition of stalking that percent) the adult population from requires victims to feel a high level which the sample was drawn, a large The survey also found that 1 percent proportion of men and women in the of fear, the NVAW Survey found that of all women surveyed and 0.4 8 percent of women and 2 percent of survey sample were at risk of being percent of all men surveyed were stalked in the 12 months preceding men in the United States have been stalked during the 12 months pre- stalked at some time in their life.12 the interview. As the proportion of ceding the survey. These findings the U.S. population aged 18–39 equate to an estimated 1,006,970 Based on U.S. Census estimates of years declines, so should the number women and an estimated 370,990 the number of women and men in the of persons stalked annually. How- men who are stalked annually in the country, one out of every 12 U.S. ever, the lifetime estimates of stalk- United States (see exhibit 2). women (8.2 million) has been stalked ing victimization should remain at some time in her life, and one out The average annual estimates of relatively constant. of every 45 U.S. men (2 million) has stalking victimization generated by been stalked at some time in his life Another reason annual estimates of the survey are relatively high com- stalking victimization are relatively (see exhibit 1).13 pared to the average lifetime esti- high compared to lifetime rates is that mates. Two factors account for this Ninety percent of the stalking victims stalking, by definition, involves finding. The first has to do with the identified by the survey were stalked repeated and ongoing victimization. age of the population most at risk of by just one person during their life. Thus, some men and women are being stalked. The survey found that Nine percent of female victims and 8 stalked for months or years on end. 74 percent of stalking victims are percent of male victims were stalked Because some men and women are between 18 and 39 years old. Since by two different persons, and 1 stalked from one year to the next, the men and women between 18 and 39 percent of female victims and 2 average annual estimates of stalking

3 Research in Brief

victimization cannot be added to Exhibit 3. Percentage of Men and Women Victimized in Previous 12 produce an estimate of the total Months, by Type of Violence number of men and women who will be stalked in two, three, or more years. Persons Victimized in Previous 12 Months (%) Thus, average annual rates of stalking Type of Men Women victimization will appear higher than Violence (N = 8,000) (N = 8,000) expected when compared to lifetime a rates of stalking victimization. Rape <0.1 0.3 Physical assault 3.4 1.9 Comparison with previous Stalking 0.4 1.0 stalking estimates Any of the above 3.9 3.0 Prior to this study, information on a. Based on five or fewer cases. stalking prevalence was limited to guesses provided by mental health Stalking prevalence based on on African-American, American professionals based on their work broader definition. If a less Indian/Alaska Native, Asian/Pacific with known stalkers. The most stringent definition of stalking is Islander, and mixed race women are frequently cited “guesstimates” of used—one requiring victims to feel combined, there is no difference in stalking prevalence were made by only somewhat frightened or a little stalking prevalence between white forensic psychiatrist Park Dietz, who frightened by their assailant’s behav- women and minority women: 8.2 in 1992 reported that 5 percent of ior—stalking prevalence rates rise percent of white women (see exhibit U.S. women are stalked at some time dramatically. Specifically, the 4) and 8.2 percent of nonwhite women in their life and approximately lifetime stalking prevalence rate (not shown) reported ever being 200,000 U.S. women are stalked stalked in their lifetime. However, a 14 increases from 8 percent to 12 each year. Thus, the NVAW percent for women and from 2 percent comparison of stalking prevalence Survey’s estimate that 8 percent of to 4 percent for men; and the annual across specific racial and ethnic U.S. women have been stalked at stalking prevalence rate increases groupings shows that American some time in their life is 1.6 times from 1 percent to 6 percent for women Indian/Alaska Native women report greater than Dietz’s guesstimate, and and from 0.4 percent to 1.5 percent significantly more stalking victimiza- the survey’s estimate that 1,006,970 for men. Based on these higher tion than women of other racial and U.S. women are stalked annually is prevalence rates, an estimated 12.1 ethnic backgrounds (see exhibit 4). five times greater than Deitz’s million U.S. women and 3.7 million This finding should be viewed with guesstimate. U.S. men are stalked at some time in caution, however, given the small How prevalent is stalking com- their life; and 6 million women and number of American Indian/Alaska pared to other forms of violence 1.4 million men are stalked annually Native women in the sample. This against women in the United in the United States. These results finding also underscores the need for States? The NVAW Survey found show how stalking prevalence varies specificity when comparing preva- that 0.3 percent of all women sur- with the level of fear included in the lence rates among women of different veyed experienced a completed or definition. A higher standard of fear racial or ethnic backgrounds. produces lower prevalence rates, and attempted rape in the 12 months Since information on violence against preceding the survey, and 1.9 a lower standard of fear produces higher prevalence rates. American Indian and Alaska Native percent experienced a physical women is limited, it is difficult to assault in the 12 months preceding explain why they report more stalk- the survey (see exhibit 3). Thus, in a Stalking risk for racial and ing victimization. A previous study 1-year period, women are three times ethnic minorities found that the overall homicide rates more likely to be stalked than raped, for Native Americans were about two Information from the NVAW Survey but they are two times more likely to times greater than U.S. national presents a complex picture of stalk- be physically assaulted than stalked. rates.15 Thus, there is some evidence ing, race, and ethnicity. When data that Native Americans are at signifi-

4 Research in Brief

cantly greater risk of violence—fatal Exhibit 4. Percentage of Men and Women Stalked in Lifetime, by Race and and nonfatal—than other Americans. Ethnicity of Victim How much of the variance in stalking prevalence may be explained by Persons Stalked in Lifetime (%) demographic, social, and environ- Asian/ American mental factors remains unclear and African- Pacific Indian/ Mixed requires further study. Moreover, Group Total White American Islander Alaska Native Race there may be significant differences Men (N=7,759) (N=6,424) (N=659) (N=165) (N=105) (N=406) in stalking prevalence among women 2.3 2.1 2.4 1.8a 4.8 3.9 of diverse American Indian tribes and b Alaska Native communities that Women (N=7,850) (N=6,452) (N=780) (N=133) (N=88) (N=397) 8.2 8.2 6.5 4.5 17.0 10.6 cannot be determined from the survey, since data on all Native a. Based on five or fewer cases. ≤ Americans were combined. b. Differences between racial and ethnic groups are significant at .05. There is some evidence that Asian and Pacific Islander women are at signifi- Exhibit 5. Percentage of Men and Women Stalked in Lifetime, by Hispanic/ cantly less risk of being stalked than Non-Hispanic Origin of Victim women of other racial and ethnic backgrounds (see exhibit 4). Again, Persons Stalked in Lifetime (%) however, given the small number of Group Total Hispanica Non-Hispanic Asian/Pacific Islander women in the Men (N=7,916) (N=581) (N=7,335) sample, this finding must be viewed 2.2 3.3 2.1 with caution. It has been suggested Women (N=7,945) (N=628) (N=7,317) that traditional Asian values emphasiz- 8.1 7.6 8.2 ing close family ties and harmony may discourage Asian women from disclos- a. Persons of Hispanic origin may be of any race. ing physical and emotional abuse by intimate partners.16 Thus, the smaller stalking prevalence rate found among The survey found no significant comparison, 94 percent of the stalkers Asian/Pacific Islander women may be, difference in stalking prevalence identified by female victims and 60 at least in part, an artifact of among men and women of Hispanic percent of the stalkers identified by underreporting. There may also be a and non-Hispanic origin (see exhibit male victims were male. Overall, 87 significant difference in stalking 5). Since previous studies compar- percent of the stalkers identified by prevalence between Asian women and ing the prevalence of violence among victims were male. Hispanic and non-Hispanic women Pacific Islander women that cannot be Young adults are also the primary determined from the survey, since data have produced contradictory conclu- 17 targets of stalkers. Fifty-two per- on these two groups were combined. sions, these findings neither confirm nor contradict earlier cent of the stalking victims were The survey found no significant differ- findings. 18–29 years old and 22 percent ence in stalking prevalence among men were 30–39 years old when the stalking started (see exhibit 6). On of different racial and ethnic back- Who stalks whom? grounds. This finding must also be average, victims were 28 years old viewed with caution given the sample’s Though stalking is a gender-neutral when the stalking started. small number of male victims falling crime, women are the primary victims The survey confirms previous reports into specific racial and ethnic group- of stalking and men are the primary that most victims know their stalker.18 ings. A larger sample of male stalking perpetrators. Seventy-eight percent of Only 23 percent of female stalking victims is needed to produce more the stalking victims identified by the victims and 36 percent of male reliable information on the relative risk survey were women, and 22 percent stalking victims were stalked by of stalking among men of different were men. Thus, four out of five strangers. The survey also shows that racial and ethnic backgrounds. stalking victims are women. By

5 Research in Brief

Exhibit 6. Victim’s Age When First current or former dates or boyfriends. Exhibit 8. Point in Intimate Stalkeda Overall, 59 percent of female victims, Relationship When Stalking compared with 30 percent of male of Womena Occurs victims, were stalked by some type of intimate partner (see exhibit 7).

18–29 Years It has been reported previously that Before 52% when women are stalked by intimate Relationship <18 Years Ends 12% partners, the stalking typically occurs After 21% Relationship after the woman attempts to leave the Ends ≥40 Years relationship.19 To test this assump- 43% Both Before 15% tion, the NVAW Survey asked women 30–39 Years and After 22% who had been stalked by former Relationship Ends husbands or partners when in the 36% relationship the stalking occurred. a. N=797 male and female victims. Percentages Twenty-one percent of these victims do not total 100 due to rounding. said the stalking occurred before the a. N=263 female victims. relationship ended, 43 percent said it women tend to be stalked by intimate occurred after the relationship ended, partners, defined as current or former and 36 percent said it occurred both (see exhibit 7), 90 percent of whom spouses, current or former cohabi- before and after the relationship are male. It is unclear from the tants (of the same or opposite sex), or ended (see exhibit 8). Thus, contrary survey data why men are stalked by current or former boyfriends or to popular opinion, women are often male strangers and male acquaintan- girlfriends. Thirty-eight percent of stalked by intimate partners while the ces. There is some evidence that female stalking victims were stalked relationship is still intact. homosexual men are at greater risk of by current or former husbands, 10 being stalked than heterosexual men: percent by current or former cohabit- The survey found that men tend to be Stalking prevalence was significantly ing partners, and 14 percent by stalked by strangers and acquaintances greater among men who had ever lived with a man as a couple compared with men who had never lived Exhibit 7. Relationship Between Victim and Offender with a man as a couple (see exhibit 9). Thus, in some stalking cases involv- 40 38 ing male victims and stranger or 36 34 acquaintance perpetrators, the perpetrator may be motivated by 30 hatred toward homosexuals, while in a others the perpetrator may be moti- 23 vated by sexual attraction. It is also 20 19 possible that some men are stalked by

14 male strangers and male acquaintan- 13 ces in the context of inter- or intra-

Percentage of Cases Percentage 10 10 10 9 group gang rivalries. Clearly, more research is needed to determine 4 2 under what circumstances men are 0 stalked by male strangers and male Spouse/ Cohabiting Date/ Relative Other Acquaintanceb Strangerb Ex-spouseb Partner/ Former Date Than Spouse acquaintances. Ex-partner Male victims (N=179) Female victims (N=650) Although men tend to be stalked by strangers and acquaintances, women a. Percentages exceed 100% because some victims had more than one stalker. are at significantly greater risk of b. Differences between males and females are significant at ≤.05.

6 Research in Brief

being stalked by strangers and Exhibit 9. Percentage of Men Stalked in Lifetime, by Whether They Ever acquaintances than men. A com- Cohabited with a Man parison of stalking prevalence among women and men by victim-offender Cohabitation Experience relationship shows that 1.8 percent Men Stalked/ Cohabited with a Man Never Cohabited with a Man of all U.S. women, compared with 0.8 Not Stalked in (N=65) (N=7,935) percent of all U.S. men, have been Lifetimea %% stalked by strangers; and 1.6 percent Stalked 7.7b 2.2 of all U.S. women, compared with 0.8 percent of all U.S. men, have been Not stalked 92.3 97.8 stalked by acquaintances (see a. Differences between men who “cohabited” and “never cohabited” are significant at < .01. exhibit 10). b. Based on five or fewer cases.

How do stalkers harass and Exhibit 10. Percentage of Men and Women Stalked in Lifetime, by Victim- terrorize? Offender Relationship

When asked to describe specific Persons Stalked in Lifetime (%) activities their stalkers engaged in to Victim-Offender Men Women harass and terrorize them, women Relationship (N=8,000) (N=8,000) were significantly more likely than a men to report that their stalkers Intimate 0.6 4.8 followed them, spied on them, or Relative 0.1b 0.3 stood outside their home or place of Acquaintancea 0.8 1.6

work or recreation (see exhibit 11). a Women were also significantly more Stranger 0.8 1.8 likely to report that their stalkers a. Differences between men and women are significant at ≤ .05. made unsolicited phone calls. About b. Based on five or fewer cases. equal percentages of female and male victims reported that their Exhibit 11. Stalking Activities Engaged in by Stalkers stalkers sent them unwanted letters or items, vandalized their property, or killed or threatened to kill a Followed, spied on, 82 a family pet (see exhibit 11). stood outside home, etc. 72

Made unwanted 61 b How often do stalkers phone calls 42 threaten overtly? Sent or left unwanted 33 letters, items 27 Many State antistalking laws include 29 in their definition of stalking a Vandalized property requirement that stalkers make an 30

overt threat of violence against their Killed or threatened 9 20 victim. Survey findings suggest that to kill family pet 6 this requirement may be ill-advised. 0 20 40 60 80 100 By definition, stalking victims in this c survey were either very frightened of Percentage of Cases their assailant’s behavior or feared Female Victims (N=625) Male Victims (N=168) their assailant would seriously harm or kill them or someone close to a. Differences between males and females are significant at ≤.05. ≤ them. Despite the high level of fear b. Differences between males and females are significant at .001. c. Percentages exceed 100% because the question had multiple responses. required, the survey found that less

7 Research in Brief

than half the victims—both male and the women who were stalked by a Exhibit 12. Percentage of Victims female—were directly threatened by current or former husband or cohabit- Who Were Overtly Threatened their stalker (see exhibit 12). This ing partner were also sexually as- by Their Stalkers finding shows that stalkers do not saulted by the same partner. By always threaten their victim verbally comparison, 20 percent of the women 50 45 or in writing; more often they engage who were ever married or ever lived 43 in a course of conduct that, taken in with a man were physically assaulted 40 context, causes a reasonable person by a current or former husband or to feel fearful. The model anti- partner, and 5 percent of women who 30 stalking code reflects this reality by were ever married or ever lived with a

not including in its definition of man were sexually assaulted by a Percentage 20 stalking a requirement that the current or former husband or partner. stalker make a credible threat of Thus, husbands or partners who stalk 10 violence against the victim.21 their partners are four times more likely than husbands or partners in the 0 Male Victims Female Victims Why stalkers stalk their general population to physically (N=179) (N=651) victims assault their partners, and they are six times more likely than husbands and To generate information on motiva- partners in the general population to abusive behavior they experienced at tions for stalking, the survey asked sexually assault their partners. the hands of their current or former spouses or cohabiting partners. The victims why they thought they had The survey also provides compelling been stalked. Since stalking occurs survey found that ex-husbands who evidence of the link between stalking in a variety of situations and between stalked (either before or after the and controlling and emotionally relationship ended) were signifi- people who have various relation- abusive behavior in intimate rela- ships, it is not surprising that re- cantly more likely than ex-husbands tionships. To provide a context for sponses to this question varied. who did not stalk to engage in violence occurring between intimate emotionally abusive and controlling Based on victims’ perceptions of why partners, respondents to the survey they were stalked, it appears that behavior toward their wife (see were asked a series of questions much stalking is motivated by exhibit 14 for details). about controlling and emotionally stalkers’ desire to control, or instill fear in, their victim (see exhibit 13). The survey results dispel the myth Exhibit 13. Victims' Perceptions of Why They Were Stalkeda that most stalkers are psychotic or delusional. Only 7 percent of the victims said they were stalked Stalker wanted to control victim 21 because their stalkers were mentally Stalker wanted to keep victim 20 ill or abusing drugs or alcohol. in the relationship Stalker wanted to scare victim 16 Relationship between stalking Victim not sure why and other forms of violence 12 Stalker was mentally ill or 7 The National Violence Against Women abusing drugs or alcohol Survey provides compelling evidence Stalker wanted or liked the attention 5 of the link between stalking and other Stalker wanted to catch victim forms of violence in intimate relation- doing something 1 ships. Eighty-one percent of the 0 51015 20 25 women who were stalked by a current Percentage of Cases or former husband or cohabiting partner were also physically assaulted a. N=624 male and female victims. by the same partner, and 31 percent of

8 Research in Brief

How often is stalking Exhibit 14. Percentage of Ex-Husbands Who Engaged in Emotionally reported to police? Abusive or Controlling Behavior, by Whether They Stalkeda

Fifty-five percent of female victims Ex-Husbands Who Ex-Husbands Who and 48 percent of male victims said Types of Emotionally Abusive/ Stalked (%) Did Not Stalk (%) their stalking was reported to the Controlling Behaviorb (N=166) (N=2,645) police. In most of these cases, the Had a hard time seeing things victims made the report (see exhibit from her point of view 87.7 57.8 15). The percentage of women Was jealous or possessive 83.7 46.3 reporting stalking is identical to the percentage of female victims report- Tried to provoke arguments 90.3 45.3 ing lone-offender violent crimes to Tried to limit her contact with family and friends 77.1 32.3 police during 1987–89, as measured Insisted on knowing where she was at all times 80.7 34.4 by the National Crime Victimization Survey.22 Made her feel inadequate 85.5 40.9 Shouted or swore at her 88.0 44.5 Police responses to stalking cases involving male victims and female Frightened her 92.2 33.1 victims were virtually identical, with Prevented her from knowing about or having two notable exceptions: Police were access to family income 59.6 20.8 significantly more likely to arrest or Prevented her from working outside the home 30.7 13.0 detain a suspect in cases involving Insisted on changing residences even when female victims, and they were signifi- she didn’t need or want to 33.9 11.9 cantly more likely to refer female a. Based on responses for first ex-husbands only. victims to services (see exhibit 15). b. Differences between ex-husbands who stalked and ex-husbands who did not stalk are significant at ≤ .001. There is some evidence that stalking reports to the police by victims have increased since passage of anti- Exhibit 15. Percentage and Characteristics of Stalking Cases Reported to stalking laws. According to informa- the Police, by Sex of Victim tion from the survey, stalking cases occurring before 1990—the year Stalking Victims (%) California passed the Nation’s first Reported to Police/Response Male Female Total antistalking law—were significantly less likely to be reported to the police Was case reported to the police? (N=178) (N=641) (N=819) than stalking cases occurring after Yes 47.7 54.6 53.1 1995, the year all 50 States and No 52.3 45.4 46.9 the District of Columbia had laws Who reported the case?a (N=84) (N=350) (N=434) proscribing stalking. There was no Victim 75.0 84.0 82.3 significant difference, however, in the Other 25.0 16.0 17.7 number of arrests made in stalking Police Responsea,b (N=84) (N=350) (N=434) cases that occurred before 1990 and Took report 66.7 68.6 68.0 those that occurred after 1995. Arrested or detained perpetratorc 16.7 25.1 23.5 When asked why they chose not to Referred to prosecutor or court 19.0 24.3 23.3 report their stalking to the police, Referred to victim servicesc 8.3 15.1 13.8 victims were most likely to state that Gave advice on self-protective measures 29.8 34.0 33.2 their stalking was not a police Did nothing 16.7 19.4 18.9 matter, they thought the police would a. Based on responses from victims whose stalking was reported to the police. not be able to do anything, or they b. Percentages exceed 100 percent because of multiple responses. c. Differences between males and females are significant at ≤ .05.

9 Research in Brief

feared reprisals from their stalkers Exhibit 16. Victims' Reasons for Not Reporting Stalking to Policea (see exhibit 16).

Overall, stalking victims gave police a Not a police matter 20 50/50 approval rating (see exhibit 17). Respondents who said their stalkers Police couldn't do anything 17 were arrested were significantly more Afraid of reprisal from stalker 16 likely to be satisfied with the way the police handled their case than Handled it myself 12 respondents who said their stalkers Reported to someone else 7 were not arrested (76 percent versus Police wouldn't believe me 7 42 percent). Private, personal matter 6 Victims who thought the police “should have done more” in their case Didn't want police, courts involved 5 were asked to describe what specific Incident too minor 4 actions they thought the police should 0 5 10 15 20 have taken. Forty-two percent thought Percentage of Victims the police should have put their assailant in jail, 20 percent said the a. N=348 male and female victims. police should have taken their situation more seriously, and 16 percent Exhibit 17. Victims' Satisfaction With the Policea said the police should have done more to protect them (see exhibit 18). 60 54 50 51 How often are stalkers 50 criminally prosecuted? Overall, 13 percent of female victims 40 and 9 percent of male victims reported that their stalkers were crimi- 30 nally prosecuted (see exhibit 19). These figures increase to 24 percent 20 and 19 percent, respectively, when of Victims Percentage only those cases with police reports 10 are considered. The stalkers were charged with a wide variety of crimes, 0 Victim was satisfied Victim thought situation Victim thought police including stalking, harassment, with actions taken improved after did everything menacing or threatening, vandalism, by police police report they could trespassing, breaking and entering, robbery, disorderly conduct, intimida- a. N = 435 male and female victims. tion, and simple and aggravated assault. Survey participants reported that about half the stalkers (54 Obtaining protective or more likely than male victims (28 percent) who had criminal charges restraining orders against percent and 10 percent) to obtain a filed against them were convicted of a stalkers protective or restraining order crime. Of those convicted, nearly against their stalker (see exhibit 20). two-thirds (63 percent) were believed Results from the survey also indicate This finding is expected since to have been sent to jail or prison. that female victims were significantly women are significantly more likely

10 Research in Brief

than men to be stalked by intimate Exhibit 18. Victims' View of Other Actions Police Should Have Takena partners who have a history of being violent toward them. Of those who obtained restraining orders, 69 Arrested perpetrator 42 percent of the women and 81 percent Taken complaint more seriously 20 of the men said their stalker violated the order. Provided protection 16

Given perpetrator a warning 14 What are psychological and Been more supportive social consequences of 13 stalking? Followed through with investigation 12 Made perpetrator leave The survey produced strong confirma- 5 tion of the negative mental health 0 10 20 30 40 50 impact of stalking. About a third of Percentage of Victims the women (30 percent) and a fifth of the men (20 percent) said they sought a. N=201 male and female victims who thought police should have done more. psychological counseling as a result of their stalking victimization. In Exhibit 19. Percentage and Outcomes of Criminal Prosecutions in Stalking addition, stalking victims were Cases, by Sex of Victim significantly more likely than nonstalking victims to be very con- Stalking Victims (%) cerned about their personal safety and Outcome Male Female Total about being stalked, to carry some- Was perpetrator prosecuted? (N=178) (N=645) (N=823) thing on their person to defend themselves, and to think personal Yes 9.0 13.1 12.1 safety for men and women had gotten No 91.0 86.9 87.9 worse in recent years (see exhibit 21). Was perpetrator convicted?a (N=15) (N=72) (N=87) Over a quarter (26 percent) of the Yes 60.0 52.8 54.0 stalking victims said their victimiza- No 40.0 47.2 46.0 tion caused them to lose time from Was perpetrator sentenced to work. While the survey did not jail or prison?b (N=9) (N=37) (N=46) query victims about why they lost Yes 77.8 59.5 63.0 time from work, it can be assumed No 22.2c 40.5 37.0 they missed work for a variety of a. Based on responses from victims whose perpetrator was prosecuted. reasons—to attend court hearings, to b. Based on responses from victims whose perpetrator was convicted. meet with a psychologist or other c. Based on five or fewer sample cases. mental health professional, to avoid contact with their assailant, and to consult with an attorney. When Stalking victims were asked whether When and why does stalking asked how many days of work they they took any measures (other than stop? lost, 7 percent of these victims said reporting their victimization to the they never returned to work. On police or obtaining a protective At the time of the interview, 92 percent average, however, victims who lost order) to protect themselves from of the victims were no longer being time from and returned to work their stalker. Fifty-six percent of the stalked. Based on information pro- missed 11 days. women and 51 percent of the men vided by these victims, about two- reported taking some type of self- thirds of all stalking cases last a year protective measure (see exhibit 22). or less, about a quarter last 2–5 years,

11 Research in Brief

Exhibit 20. Percentage and Outcomes of Protective Orders in Stalking ing had ceased; 19 percent said the Cases, by Sex of Victim stalking stopped because they (the victims) moved away (see exhibit Stalking Victims (%) 24). These findings suggest that Outcome Male Female Total address confidentiality programs may be an effective means of com- Did victim obtain a protective or restraining order?a (N=175) (N=597) (N=772) bating stalking. These programs encourage victims who face contin- Yes 9.7 28.0 23.8 ued pursuit and unusual safety risks No 90.3 72.0 76.2 to develop a personal safety plan that Was the order violated?a,b (N=16) (N=166) (N=182) includes relocating as far from their Yes 81.3 68.7 69.8 assailant as possible and securing a No 18.7 31.3 30.2 confidential mailing address that provides mail forwarding service but ≤ a. Differences between males and females are significant at .05. does not reveal their new location.23 b. Based on responses from victims who obtained a restraining order. Some stalking cases are resolved when the perpetrator gets a new love interest. Exhibit 21. Fear for Personal Safety Among Victims and Nonvictims Eighteen percent of the victims said of Stalking the stalking stopped because their assailant got a new spouse, partner, or 70 68 boyfriend/girlfriend. 62 60 It has been reported previously that informal law enforcement interven- 50 45 tions, such as detective contacts, can 42 be an effective means of deterring 40 stalkers, particularly in cases where 30 29 30 the victim and the suspect had some 24 prior relationship and where the

Percentage of Cases Percentage 20 stalker is not suffering from mental illness.24 Findings from the NVAW 10 10 Survey provide some support for this theory. Victims were more likely to 0 credit informal, rather than formal, Thinks personal safety Very concerned Very concerned Carries something for men and women about personal about being to defend selfb justice system interventions for the has gotten worsea safetyb stalkedb cessation of their stalking. For Victims (N=826) Nonvictims (N=15,053) example, 15 percent of victims said their stalking stopped after their a. Differences between victims and nonvictims are significant at ≤.01. assailants received a warning from b. Differences between victims and nonvictims are significant at ≤.001. the police. By comparison, only 9 percent of victims said their stalking ceased because their stalker was and about a tenth last more than 5 involving nonintimates (2.2 years and arrested, 1 percent said their stalk- years (see exhibit 23). On average, 1.1 years, respectively). ing stopped because their stalker stalking cases last 1.8 years. However, was convicted of a crime, and less stalking cases involving intimates or Victims who were no longer being than 1 percent said the stalking former intimates last, on average, stalked at the time of interview were stopped because they obtained a significantly longer than stalking cases asked why they thought their stalk- restraining order against their

12 Research in Brief

Exhibit 22. Self-Protective Measures Undertaken by Stalking Victimsa Exhibit 23. Distribution of Cases by Number of Years Stalking Lasteda

Took "extra" precautions 22

Enlisted help of family and friends 18 1 Year Got a gun 17 16% Changed address 11 2–5 Moved out of town 11 < 1 Year Years 52% Avoided perpetrator 7 23%

Talked to an attorney 5 5+ Years Varied driving habits 5 9% Moved to a shelter 4

Stopped going to work, school, out 4 a. N=759 cases. Got public records sealed 1

1 Hired a private investigator stalker. The fact that so few victims 0 5 10 15 20 25 credited formal justice system Percentage of Victims interventions is not surprising given the paucity of arrests, criminal a. N=440 male and female victims who took self-protective measures. prosecutions, and restraining orders in stalking cases.

Exhibit 24. Victims' Perception of Why Stalking Stoppeda Policy implications

Victim moved 19 Prior to this study, empirical data on the prevalence and characteristics of Stalker got new love interest 18 stalking in the general population Police warned stalker 15 were virtually nonexistent. There-

Victim talked to stalker 10 fore, information provided in this report can help inform policy and Stalker was arrested 9 interventions directed at stalking. Stalker moved 7 Based on findings from the National Violence Against Women Survey, the Stalker got help 6 Center for Policy Research offers the Victim got new love interest 4 following conclusions:

Stalker died 4 1. Stalking should be treated as a significant social problem. The It just stopped 3 survey found that stalking is much b Stalker was convicted of a crime 1 more prevalent than previously 05 10 15 20 thought, affecting an estimated 1.4 Percentage of Cases million adults per year in the United States. Since this figure does not a. N=665 cases. include cases involving victims b. Based on 5 or fewer cases. under the age of 18, nor victims who are homeless or living in homes

13 Research in Brief without telephones, the estimate is focus on stalking occurring between of their victimization. In addition, probably an undercount of the true intimates and acquaintances. stalking victims are significantly number of persons stalked each year. more likely than nonstalking victims Given the scope of the stalking 4. The Nation’s criminal justice to be very fearful for their personal problem revealed by this survey, it is community should receive com- safety, to carry something on their imperative that stalking be treated as prehensive training on the par- person to protect themselves, and to a legitimate criminal justice problem ticular safety needs of stalking think personal safety for men and and public health concern. victims. The survey produced women has declined in recent years. dramatic confirmation of the link To better meet the needs of stalking 2. Credible threat requirements between stalking and physical victims, mental health professionals should be eliminated from violence in intimate relationships. need additional information about antistalking statutes. Some State Fully 81 percent of the women who the characteristics of stalking, the statutes include in their definition of were stalked by an intimate partner mental health impact of stalking, and stalking a requirement that stalkers (either before or after the relation- the mental health needs of stalking make a credible threat of violence ship ended) were also physically victims. against their victims. Since stalking assaulted by that partner, and 31 is often a “crime of deeds” rather percent were also sexually assaulted 7. Stalking intervention strate- than a “crime of words,” this re- by that partner. To help law enforce- gies should include address quirement makes it more difficult to ment officers, prosecutors, and confidentiality programs. Survey prosecute stalkers. Findings from defense attorneys make appropriate data indicate that about a fifth of all the survey show that stalkers often case processing and management stalking victims move to a new do not threaten their victims verbally decisions, they must be made aware location to escape their stalker. or in writing but instead engage in a of the very real safety risks faced by Given these findings, it is important course of conduct that, taken in these stalking victims. that address confidentiality programs context, causes a reasonable person be made available to stalking vic- to feel fearful. Despite being very 5. More research must be tims. These programs encourage frightened or fearing bodily harm or conducted on the effectiveness victims who face continued pursuit death, less than half of the stalking of formal and informal law and unusual safety risks to develop a victims identified by the survey were enforcement interventions. The personal safety plan that includes directly threatened by their stalkers. survey found that 70 percent of all relocating as far from their offender This finding supports the view of restraining orders obtained against as possible and securing a confiden- many stalking experts that language stalkers were violated. The survey tial mailing address that provides which might be construed as requir- also found that stalking victims were mail forwarding service without ing an actual verbal or written threat more likely to credit the cessation of revealing their new location. Be- should be eliminated from all State their stalking to informal police cause these measures focus on the antistalking statutes. interventions, such as police warn- behavior of the victim rather than the ings, than to formal justice system perpetrator, they may be perceived 3. Research on stalking should interventions, such as arrests or as unfair and unjust; but they may be move beyond “celebrity stalking” restraining orders. More research is the most effective way some victims and focus on acquaintance and needed to determine under what can elude their stalkers. intimate partner stalking. Prior to situations various law enforcement this study, most stalking research interventions are most effective. focused on celebrity or political stalking. Findings from the survey 6. The mental health commu- show that the vast majority of stalking nity should receive comprehen- cases involve people who know each sive training on the appropriate other, with fully half of all stalking treatment of stalking victims. cases arising in the context of current The survey found that about a or former intimate relationships. quarter of all stalking victims seek Therefore, future research should psychological counseling as a result

14 Research in Brief

Survey Methodology and Demographic Description of the Sample

The National Violence Against Women Survey was conducted during Table 1. Comparison of Demographic Characteristics of Men and Women in November 1995–May 1996 by National Violence Against Women Survey (NVAWS) and U.S. Population interviewers at Schulman, Ronca, Men (%)a Women (%)a Bucuvalas, Inc. (SRBI), a national survey research organization in New Demographic U.S. U.S. b b York City, under the direction of Dr. Characteristics NVAWS Population NVAWS Population John Boyle. Survey design and data Age (N=7,920) (N=92,748,000) (N=7,856) (N=100,679,000) analysis were conducted by the 18–24 11.4 13.0 9.8 11.9 authors of this report. 25–29 10.4 10.2 9.6 9.4 The sample was drawn as a national, 30–39 25.4 23.8 24.6 21.9 random-digit-dialing (RDD) sample of 40–49 24.0 20.0 22.5 18.9 telephone households in the United 50–59 13.5 13.0 14.4 12.9 States. The sample was stratified by 60–69 8.8 10.1 9.9 10.7 U.S. Census region to control for differential response rates by region. 70–79 5.2 7.0 6.8 8.9 Within regional strata, a simple 80+ 1.5 2.9 2.5 5.5 random sample of working, residen- Race/Ethnicity (N=7,353) (N=93,282,000) (N=7,453) (N=101,117,000) tial, “hundreds banks” phone White 87.4 84.8 86.6 83.7 numbers was drawn. A hundreds African-American 9.0 10.9 10.5 12.0 bank is the first eight digits of any 10- digit telephone number (e.g., 301– American Indian/Alaska 1.4 0.7 1.2 0.7 608–38XX). A randomly generated Native two-digit number was appended to Asian/Pacific Islander 2.2 3.5 1.8 3.6 each randomly sampled hundreds Hispanic Originc (N=7,916) (N=93,282,000) (N=7,945) (N=101,117,000) bank to produce the full 10-digit, Hispanic 7.3 9.4 7.9 8.5 random-digit number. The random- Non-Hispanic 92.7 90.6 92.1 91.5 digit numbers were called by SRBI interviewers from their central Marital Status (N=7,928) (N=92,007,000) (N=7,921) (N=99,588,000) telephone interviewing facility. Never married 21.2 26.8 15.5 19.4 Nonworking and nonresidential Currently married 66.8 62.7 62.7 59.2 numbers were screened out. Once a Divorced, separated 10.2 8.3 13.3 10.3 residential household was reached, eligible adults in each household were Widowed 1.9 2.5 8.6 11.1 identified. In households with Educationd (N=7,010) (N=79,463,000) (N=7,068) (N=86,975,000) multiple eligibles, the most-recent- Less than high school 9.4 18.3 10.7 18.4 birthday method was used to High school and 29.3 31.9 34.6 35.7 systematically select the designated equivalent respondent. The household participa- Any college 48.3 40.4 45.7 39.7 tion rate was 72 percent for females Advanced degree 13.0 9.4 9.0 6.2 and 69 percent for males.25 Of the eligible respondents who started the a. Percentages may not total 100 due to rounding. interview, 97 percent of the women b. Based on U.S. Bureau of the Census estimates, Current Population Survey, 1995. and 98 percent of the men followed c. Persons of Hispanic origin may be of any race. through to completion. d. For persons aged 25 years and older.

15 Research in Brief

Survey Methodology and Demographic Description of the Sample (Continued)

The unweighted sample of the National Table 2. Comparison of Demographic Characteristics of Men and Women Violence Against Women Survey, when in Weighted and Unweighted National Violence Against Women (NVAW) compared with the Census Bureau’s Survey Sample 1995 Current Population Survey of adult NVAW Survey Sample men and adult women, was remarkably similar to the general population from Demographic Men (%)a Women (%)a which it was drawn (see Table 1, page Characteristics Weightedb Unweighted Weightedb Unweighted 15). Weighting was considered to Age (N=7,920) (N=7,920) (N=7,856) (N=7,856) correct for possible biases introduced by 18–24 11.2 11.4 9.6 9.8 the fact that some households had 25–29 10.5 10.4 9.8 9.6 multiple phone lines. Since such weighting had a negligible effect on the 30–39 25.7 25.4 24.6 24.6 demographic composition of the sample 40–49 23.6 24.0 22.1 22.5 (see Table 2), weights were not used in 50–59 13.3 13.5 14.4 14.4 this data analysis. 60–69 8.9 8.8 10.0 9.9 70–79 5.3 5.2 6.9 6.8 Precision of sample estimates. The results presented in this report were 80+ 1.5 1.5 2.5 2.5 tested to determine whether observed Race/Ethnicity (N=7,353) (N=7,353) (N=7,453) (N=7,453) differences between groups (e.g., men/ White 87.4 87.4 86.6 86.6 women) were statistically significant. African-American 9.0 9.0 10.5 10.5 Only comparisons that passed a hypothesis test at the 95 percent American Indian/Alaska 2.2 1.4 1.8 0.7 ≤ Native confidence level (p .05 ) were considered statistically significant and were Asian/Pacific Islander 1.4 2.2 1.2 1.8 discussed in this report. Hispanic Originc (N=7,916) (N=7,916) (N=7,945) (N=7,945) Hispanic 7.4 7.3 8.0 7.9 By its nature, a telephone survey is limited to the population living in Non-Hispanic 92.6 92.7 92.0 92.1 households with telephones. Thus, the a. Percentages may not total 100 due to rounding. survey does not reflect the experiences of b. Weighted for number of telephone lines per household. men and women living in phoneless c. Persons of Hispanic origin may be of any race. households, group facilities or institu- tions, or on the streets. The absence of A total of 8,000 women and 8,000 men Sample weighting. A completed sample interviews with phoneless households 18 years and older were interviewed using in a social survey will be subject to certain results in an underrepresentation of a computer-assisted interviewing system. selection biases that may introduce certain demographic characteristics Only female interviewers were used to sampling errors, in addition to sampling typical of households that lack telephone interview women. For male respondents, variability, into sample estimates. These service (e.g., poor, headed by a single approximately half of the interviews were potential sources of sample bias may be adult, located in a rural or inner-city area, conducted by female interviewers and addressed by sample weighting. Unless renters). However, since approximately half by male interviewers. A Spanish there is a considerable bias in the achieved 94 percent of the American population language translation was administered by sample, however, many researchers prefer live in households with telephones, this bilingual interviewers for Spanish- to leave the achieved sample unweighted to underrepresentation is small. speaking respondents. A technical report avoid the complexities of statistical tests describing the survey methods in more with weighted samples. detail is forthcoming.*

*To obtain copies of the technical report, call or write to the Center for Policy Research, 1570 Emerson St., Denver CO 80218, 303–837–1555.

16 Research in Brief

4. See, for example: Dietz, Park, and Martell, Daniell, “Threatening and Survey Screening Questions Otherwise Inappropriate Letters to Members of the United States Congress,” Journal of Forensic Because much confusion exists about • Left unwanted items for you to find? Sciences, 36(5), 1991; Holmes, what it means to be stalked, the National • Tried to communicate in other ways Ronald, “Stalking in America: Types Violence Against Women Survey did not use against your will? the word “stalking” in its screening and Methods of Criminal Stalkers,” questions. Including the word would have • Vandalized your property or Journal of Contemporary Criminal assumed that victimized persons knew how destroyed something you loved? Justice, 9(4), December 1993; Zona, to define stalking and perceived what Respondents who answered yes to one M.A., et al., “Comparative Study of happened to them as stalking. Instead, the or more of these questions were asked Erotomanic and Obsessional Sub- survey used the following behaviorally- whether anyone had ever done any of jects in a Forensic Sample,” Journal specific questions to screen respondents for these things to them on more than one of Forensic Sciences, 38(4), July stalking victimization: occasion. Because stalking involves 1993; Rudden, M., et al., “Diagnosis repeated behaviors, only respondents and Clinical Course of Erotomania Not including bill collectors, telephone who said yes were considered possible solicitors, or other sales people, and Other Delusional Patients,” stalking victims. Respondents who American Journal of Psychiatry, has anyone, male or female, ever... reported being victimized on more than 147(5):625–628, 1990. one occasion were subsequently asked • Followed or spied on you? how frightened their assailant’s 5. See, for example: Bernstein, • Sent you unsolicited letters or written behavior made them feel and whether Susan E., “Living Under Siege: Do correspondence? they feared their assailant would Stalking Laws Protect Domestic • Made unsolicited phone calls to you? seriously harm them or someone close to them. Only respondents who were Violence Victims?”, Cardoza Law • Stood outside your home, school, or very frightened or feared bodily harm Review, 15(1993):525–529; Boychuk, workplace? were counted as stalking victims. Katherine M., “Are Stalking Laws • Showed up at places you were even Unconstitutionally Vague or Over- though he or she had no business broad?” Northwestern University Law being there? Review, 88(2):769–802, 1994; Guy, Robert A., Jr., “Nature and Constitu- tionality of Stalking Laws,” Notes Experts Say Measures are Vague,” Vanderbilt Law Review, 46(4):991– USA Today, July 21, 1992; Tharp, 1029, 1993; Gilligan, Mattlaw, 1. See, for example: Ellement, John, Mike, “In the Mind of a Stalker,” “Stalking the Stalker: Developing “Police Arrest Boston Man, 18, for U.S. News and World Report, Febru- New Laws to Thwart Those Who Violating State Stalking Law,” Boston ary 17, 1992. Terrorize Others,” Georgia Law Globe, May 28, 1992; Sullivan, Review, 27(1992):285–342; Harmon, Kristin N., “Woman’s Case Illustrates 2. Hunzeker, Donna, “Stalking Brenda K., “Illinois’ Newly Need for State Stalking Law, Some Laws,” State Legislative Report, Amended Stalking Law: Are All the Say,” Houston Chronicle, April 19, Denver, Col.: National Conference of Problems Solved?” Southern Illinois 1992; Meyer, Josh, “Man Held in State Legislatures, 17(19):1–6, University Law Journal, Stalking of Pop Singer Janet Jack- October 1992. 19(1994):165–198; Lingg, Richard son,” Los Angeles Times, June 25, A., “Stopping Stalkers: A Critical 3. National Criminal Justice Asso- 1992; Lardner, George, “The Stalk- Examination of Anti-Stalking Legis- ciation, Project to Develop a Model ing of Kristin: The Law Made It Easy lation,” Saint John’s Law Review, Anti-Stalking Code for States, for My Daughter’s Killer,” Washing- 67(2):347–381, 1993; McAnaney, Washington, D.C.: U.S. Department ton Post, November 22, 1992; Kathleen G., et al., “From Impu- of Justice, National Institute of Puente, Maria, “Legislators Tackling dence to Crime: Anti-Stalking Justice, October 1993. the Terror of Stalking: But Some Laws,” Notre Dame Law Review,

17 Research in Brief

68(1993):819–909; Morin, K.S., 11. National Criminal Justice 16. National Research Council, “The Phenomenon of Stalking: Do Association, Project to Develop a Understanding Violence Against Existing State Statutes Provide Model Anti-Stalking Code for States Women, Washington D.C.: National Adequate Protection?” San Diego (note 3). Academy Press, 1996:40–41. Justice Journal, 1(1):123–162, 1993; Sohn, Ellen, “Antistalking Statutes: 12. The findings of the survey, as in 17. Sorenson, S.B., Stein, J.A., Do They Actually Protect Victims?” any sample survey, are subject to Siegel, J.M., Golding, J.M., and Criminal Law Bulletin, 30(3):203– sample fluctuations or sampling Burnam, M.A., “The Prevalence of 241, 1994; Strikis, Silvija, “Stopping error. Using the sampling methods Adult Sexual Assault: The Los Stalking,” Note, Georgetown Law described in this report (see “Survey Angeles Epidemiologic Catchment Journal, 81(1993):2772–2813; Methodology”), the maximum Area Project,” American Journal of Thomas, Kenneth R., “How to Stop sampling error at the 95 confidence Epidemiology, 126:154–1164, 1987; the Stalker: State Anti-Stalking level for a sample of 8,000 is plus or Sorenson, S.B., and Tells, C.A., Laws,” Criminal Law Bulletin, minus 1.1 percentage points if the “Self-Reports of Spousal Violence in 29(2):124–136, 1992; Walker, Julie response distribution on a categori- a Mexican American and a Non- Miles, “Anti-Stalking Legislation: cal variable is a 50/50 split. Hispanic White Population, Violence and Victims, 6(1991):3–16. Does It Protect the Victim Without 13. According to U.S. Bureau of the Violating the Rights of the Ac- Census estimates, there were 18. A survey of 90 Florida law cused?” Denver University of Law 100,697,000 women and 92,748,000 enforcement agencies reported that Review, 71(2):273–302, 1993. men aged 18 years and older resid- in most stalking cases the victim 6. Lardner, George, The Stalking of ing in the United States in 1995. knew the offender. See Tucker, J.T., Kristin: A Father Investigates the “The Effectiveness of Florida Stalk- 14. While testimony provided at a ing Statutes Section 784,048,” Murder of His Daughter, New York: September 29, 1992, Senate Judi- Atlantic Monthly Press, 1995; Orion, Florida Law Review, 45(4):609–707, ciary Committee Hearing on S.B. 1993. Doreen, I Know You Really Love Me: 2922 (Violence Against Women) is A Psychiatrist’s Journal of Erotoma- generally cited as the source for 19. See National Institute of Justice, nia, Stalking, and Obsessive Love, these estimates, the figures first Domestic Violence, Stalking, and New York: Macmillan, 1997. appeared in a USA Today article on Antistalking Legislation: 1 (note 9). stalking (see Puente in note 1). The 7. Thomas, “How to Stop the 20. Ibid. Stalker: State Anti-Stalking Laws” statistics contained in the article (note 5). were attributed to “guesses” pro- 21. See National Criminal Justice vided by Dr. Park Dietz, a Los Association, Project to Develop a 8. Hunzeker, “Stalking Laws” (note Angeles-based forensic psychiatrist, Model Anti-Stalking Code for States 2). presumably on the basis of his (note 3). research on a nonrepresentative 9. National Institute of Justice, sample of known celebrity stalkers 22. Bachman, Ronet, Violence Domestic Violence, Stalking, and (see Dietz in note 4). Against Women: A National Crime Antistalking Legislation: An Annual Victimization Survey Report, Wash- Report to Congress under the Violence 15. Wallace, L.J.D., Calhoun, A.D., ington, D.C.: U.S. Department of Against Women Act, Washington, Powell, K.E., O’Neill, J., and James, Justice, Bureau of Justice Statistics, D.C.: U.S. Department of Justice, S.P., Homicide and Suicide Among January 1993. National Institute of Justice, April Native Americans, 1979–1992, 1996. Violence Surveillance Summary 23. See, for example, the Address Series, No. 2, Atlanta, Ga.: Centers Confidentiality Program, Post Office 10. Ibid. for Disease Control and Prevention, Box 69, Olympia, Washington, National Center for Injury Prevention 98507-0069, (360) 753-2971. and Control, 1996.

18 Research in Brief

24. Williams, W.L., Lane, J.C.D., Patricia Tjaden, Ph.D., and Nancy document are solely those of the and Zona, M.A., “Stalking: Success- Thoennes, Ph.D., are with the ful Intervention Strategies,” The authors and do not necessarily Center for Policy Research. Police Chief, (February 1996):24–26; reflect the views of the funding and Zona, M.A., Sharma, K.K., and agencies. The authors thank Lois Lane, J.C., “A Comparative Study of This research was supported by Mock at the National Institute of Erotomanic and Obsessional Sub- grant number 93-IJ-CX-0012, Justice and Linda Saltzman at the jects in a Forensic Sample,” Journal awarded to the the Denver-based Centers for Disease Control and of Forensic Sciences, 38(4):894–903, Center for Policy Research by the Prevention for their advice and July 1993. National Institute of Justice (NIJ) support in completing this project. and sponsored jointly by NIJ and The authors also thank anonymous 25. The participation rate for the the Centers for Disease Control reviewers who provided helpful survey was determined using the and Prevention. The opinions and comments on an earlier draft of industry standard advocated by the conclusions expressed in this this report. Council of Applied Survey Research Organizations, which calculates the rate as the number of completed interviews, including those that are The National Institute of Justice is a com- screened out as ineligible, divided ponent of the Office of Justice Programs, by the total number of completed which also includes the Bureau of Justice interviews, screened-out interviews, Assistance, the Bureau of Justice Statistics, refusals, and terminated interviews. the Office of Juvenile Justice and Delin- quency Prevention, and the Office for Using this standard, the participation Victims of Crime. rate for women was (8000 + 4829) ÷ (8000 + 4829 + 4608 + 352) = .72 NCJ 169592 and the participation rate for men was (8005 + 8828) ÷ (8005 + 8828+ 7552 + 65) = .69.

19 Quick Access to Information About Violence Against Women

For news about NIJ and CDC’s most recent publications and activities related to violence against women and family violence, go to the World Wide Web pages:

NIJ’s address is: http://www.ojp.usdoj.gov/nij. Click on “Programs” for a description of the agency’s violence against women and family violence program.

CDC’s National Center for Injury Prevention and Control’s address is: http://www.cdc.gov/ncipc. Click on “Violence Prevention” for the broad range of violence activity undertaken by the National Center for Injury Prevention and Control. The direct address to CDC’s Family and Intimate Violence Prevention Team is: http://www.cdc.gov/ncipc/dvp/fivpt.

And check out the “What’s New” section on each agency’s home page.

U.S. Department of Justice Office of Justice Programs BULK RATE National Institute of Justice U.S. POSTAGE PAID DOJ/NIJ Permit No. G–91 Washington, DC 20531

Official Business Penalty for Private Use $300