American Serial Rape,1940–2010

Sex of Offender

The first commonality across samples is that all of the studies displayed here draw their conclusions from relatively small samples ranging from 16 to 108 offenders. Data on these offenders were drawn from police reports, court records, or prison interviews. Furthermore, all of the serial rape studies tabled here draw from exclusively male samples. Based on the existing literature then, it appears that serial rape is largely limited to male offenders.

Race/Ethnicity of Offender

As displayed in Table 1, two studies focused on samples that were race homogenous (Beauregard et al.’s, 2007, mostly White sample and Woodham and Labuschagne’s, 2012, Black, South African sample). Of the remaining six studies where offender characteristics were described, the samples are more diverse. The small incarcerated sample studied by Hazelwood and Warren (1990) and McCormack, Rokous, Hazelwood, and Burgess (1992) was composed of Whites (86%), Blacks (12%), and Hispanics (2%). Park, Schlesinger, Pinizzotto, and Davis’ (2008) sample of 22 serial rape offenders was 41% White, 32% Black, 18% Hispanic, and 9% Asian. Canter and Larkin (1993) and Warren et al. (1999) analyzed data from samples that were fairly evenly split between Black and White offenders (i.e., 53% Black, 47% White and 51% Black, 49% White, respectively). Finally, Kocsis and Irwin’s (1998) sample was reported to be 75% White and 25% of “Ethnic Origins.” Since these studies all analyze samples drawn from offenders formally processed by the criminal justice system in specific geographic locales (e.g., Beauregard et al.’s, 2007, Canadian sample and Woodhams and Labuschagne’s, 2012, Black, South African sample), it is not possible to arrive at any sort of summary statement about the race/ethnic membership of serial rapists. The racial/ethnic makeup of a given sample surely owes less to actual rates of offending among different groups and more to the geographic context of the study and to the relationship of race to the probability of being identified, apprehended, and incarcerated by the criminal justice system.

Age of Offender

It is difficult to summarize the age profile of offenders since investigators measure age at different stages of apprehension and adjudication. Some studies do not qualify their measurement of age of offender (e.g., Santtila, Junkkila, & Sandnabba, 2005). Other researchers appear to be entering age of offender as age at the time of the research interview (e.g., Hazelwood & Warren, 1990). A common operationalization of age is age at time of arrest (or incident) or conviction for a sex crime in the series (e.g., Kocsis & Irwin, 1997; Park, Schlesinger, Pinizzotto, & Davis, 2008). Beauregard et al. (2007) and Woodhams and Labuschagne (2012) measure offender age at the beginning of the crime series or at first known offense. Given the varieties of measurement strategies, it may not be meaningful to summarize age profiles. As displayed in Table 1, however, past researchers that provide a mean offender age mark the typical age of serial rapists at somewhere between 29 and 35 years of age.

Offender Occupation

Little is known about the occupational class status of serial rapists. Park et al. (2008) are one of the few research teams to include measures of employment status in their analysis. In their sample of offenders where occupational class was available (N = 22), 26% of the offenders were unemployed, 42% were blue-collar workers, 5% were white-collar workers, and 10.5% were students.

Attack Style

Hazelwood and Warren (1990) developed the first systematic typology of serial rapist styles. According to their typology, typical serial rapist attacks include the con, the blitz, and the surprise approach. The con approach involves using deception and persuasion to gain the victim’s trust. The con often feigns a need for assistance (e.g., asking for help from a motorist), offers assistance (e.g., offering a ride to a woman at a bus stop, knocking at the victim’s residence in the guise of a maintenance man), or poses as a trusted authority figure (e.g., police officer). The blitz approach involves the use of a sudden, direct attack that quickly subdues the victim and is more opportunistic, as it does not appear to involve careful planning. According to Hazelwood and Warren’s (1990) analysis, the blitz approach was used the least frequently of all three types and the surprise approach was the most common method used by serial rapists. The surprise approach is more premeditated than the blitz strategy and appears to involve a preselection of victims and a strategic plan to attack the victim when vulnerability is heightened (e.g., breaking into a residence when the occupant is sleeping). In a related study, Park et al. (2008) found that 79.5% of their sample employed the surprise attack, while 18.6% of their sample offenders used the blitz method and 11.6% favored the con approach.

Conceptualizing Serial Rape

According to Vronsky (2004), the “serial killer” concept was coined in the mid-1970s by Federal Bureau of Investigation (FBI) agent and behavioral profiler Robert Ressler. In its original form, the serial murder concept lacked specific definitional parameters. To Ressler, serial killers murdered a series of victims in a cyclical, patterned style. At this point, the serial killer was not defined in terms of the specific qualities of serial crime. Thus, the temporal, spatial, and stylistic characteristics of serial violence were not yet part of the definition. By the 1980s, the FBI defined serial killers as male offenders that were driven by sexual motives and who killed three or more victims (Ressler, Burgess, & Douglas, 1988). Scholarly interest in the serial killer also grew significantly in the 1980s. Researchers used a variety of definitions for serial murder. Most definitions, however, had several key components in common. Specifically, serial killers were generally conceptualized as those who killed three or more victims, driven by an intrinsic drive or urge as opposed to a profit motive, with an “emotional cooling off period” in between offenses.

Scholarly conceptualizations of serial rape vary considerably. The existing body of the literature employs a variety of definitions on what constitutes “serial rape.” And most conceptualizations of serial rape are simplistic and vague and fail to capture the unique properties of serial rape that sets it apart from other predatory offenses. Just as important as knowing what serial rape is, is knowing what it is not.

Table 2 displays a list of serial rape conceptualizations used by scholars in the field. As depicted in the table, most definitions of serial rape are rather simplistic. That is, scholars often simply define a serial rapist as one who has committed some act of sexual coercion against multiple victims. Merely defining a serial offender as one that has committed multiple acts of criminality is problematic because it fails to consider important variables related to offender motives and intent, the temporality of the crime series, offender–victim relationship, and the accomplishment of the crime itself. Serial rape is a social process that involves a complex interaction between offenders and victims across time and space. Accordingly, it would be a mistake to label one a serial offender simply because he or she has offended against multiple victims. Several of the definitions in Table 2 limit their identification of serial rapists in this manner. Lebeau (1985, 1987), Hazelwood, Reboussin, and Warren (1989), Graney and Arrigo (2001), and Santtilla, Junkkila, and Sandnabba (2005), in fact, conceptualize a serial rapist as one who has committed multiple acts of rape. These definitions fail to capture the temporal qualities of serial offenses. Serial rape researchers should follow the lead modeled by scholars of multiple homicides, who see a meaningful categorical difference between mass killers and serial murderers. Mass killers differ in kind from serial offenders because their acts of violence tend to be explosive outbursts occurring in a relatively short time frame and often—but not always—limited to one physical space. On the other hand, serial killers commit their offenses in a more measured series of attacks stretching across broader spans of time and space. Theoretically, then, scholars must give attention to time when defining serial rape. A multiple rapist, for example, is not a serial rapist if he or she commits their crimes during the course of 1 day.

OPEN IN VIEWER

Table 2.

Serial Rape/Rapist Definitions.

Source	Definition
Lebeau (1985, 1987)	An offender that commits two or more reported rapes before being apprehended by law enforcement
Hazelwood, Reboussin, and Warren (1989)	Offenders that commit rapes against 10 or more victims
Douglas, Burgess, Burgess, and Ressler (1992)	An offender who has committed attacks in three or more separate offenses separated by an emotional cooling off period
Kocsis and Irwin (1998)	Offender has committed a minimum of two offenses on victims with which they had no regular prior contact before their crime. The attacks occur at least 1 day, but no more than 1 year, apart
Graney and Arrigo (2001)	Offender has raped on two or more separate occasions against at least two different victims
Santtila, Junkkila, and Sandnabba (2005)	Serial rape involves two or more rapes committed by the same offender
Beauregard, Proulx, Rossmo, Leclerc, and Allaire (2007)	Offender committed two or more sexual assaults or other sex-related crimes (e.g., sexual homicide) involving a victim of any age and of any gender and who was a stranger to the attacker (i.e., offender had no personal relationship with victim prior to the attack)
Ottman (2008)	Serial rape involves a series of two or more rapes. Sometimes the definition of serial rape includes a cooling off period between offenses. Serial rape and serial murder may overlap and are not mutually exclusive
Current study	Three or more separate events of attempted or executed sexual coercion (with or without penetration) that occur over a period greater than 72 hr. Offenders appear to be actively seeking, hunting, or luring victims to accomplish sexual assault as their principal goal. Attacks are linked together (by police, media, or both) by some set of offense characteristics that may include the offender’s physical characteristics, a common style of attack, or characteristics of the victim.

The definitions listed in Table 2 also say little about the offender–victim relationship. The exceptions are represented by Kocsis and Irwin (1997) and Beauregard et al. (2007). Both of these research teams assert that serial rapists are those that offend against “strangers.” This is an important observation, since it highlights the importance of target selection. Serial predators, presumably, select victims due to their vulnerability, based on opportunity, and in a strategic way to maximize the probability that they will complete the act without being identified or apprehended. After all, the serial predator hopes to continue his or her crime series. While researchers have demonstrated that most sexual predators are previously known to the victim, serial rapists appear to prey more on strangers (Beauregard et al., 2007; Hazelwood & Warren, 1990) which makes it harder for them to be identified and makes a case more difficult to prosecute (Lebeau, 1987).

A useful conceptualization of serial rape should also recognize that the serial rapist’s main objective is a sexual assault. Like Hazelwood and Warren (1990) and Beauregard et al. (2007), we see the serial rapist as hunting, luring, and otherwise seeking vulnerable victims in order to carry out the principal goal of sexual assault. Thus, versatile criminals who rob victims directly or burglarize homes and then attack victims as a peripheral or opportunistic act are not counted as serial rapists. There are, however, offenders who commit sexually motivated burglaries that are principally interested in sexual assault and commit property crimes as an incidental or opportunistic act (see Pedneault, Beauregard, Harris, & Knight, 2015). Sexually motivated burglars of this kind do count as serial rapists.

With respect to the actual physical offense, we believe that the act of serial rape may include a variety of attempted or completed acts of sexual coercion. Our conceptualization is consistent with recent revisions to the FBI classification of rape as “penetration, no matter how slight, of the vagina or anus with any body part or object, or oral penetration by a sex organ of another person, without consent of the victim” (2013, p. 1). This definition is far more inclusive than any previous definitions, as it now includes a broader range of penetration and male victims as well as females.

Finally, we set our victim count parameters in line with the FBI definition of serial murder. According to the FBI: “The term ‘serial killings’ means a series of three or more killings, not less than one of which was committed within the United States, having common characteristics such as to suggest the reasonable possibility that the crimes were committed by the same actor or actors” (Behavioral Analysis Unit, National Center for the Analysis of Violent Crime, 2008, p. 8). One of the key strengths of the FBI’s definition is that it introduces case linkage as an important definitional dimension. When defining a collection of multiple attacks as a series, one must draw from all the available evidence to demonstrate that the series was committed by one actor or specific actors working together.

Taking all of these factors into consideration, our conceptualization of serial rape is as follows: Three or more separate events of attempted or executed sexual coercion (with or without penetration) that occur over a period greater than 72 hr. Offenders appear to be actively seeking, hunting, or luring victims to accomplish sexual assault as their principal goal. Attacks are linked together (by police, media, or both) by some set of offense characteristics that may include the offender’s physical characteristics, a common style of attack, or characteristics of the victim.

Rapists that kill their victims are not included in our conceptualization (to distinguish them from sexual serial killers). There are some similarities between serial killers and serial rapists. Hazelwood and Warren’s (1990) con–blitz–surprise typology is consistent with models created by serial murder scholars who designate serial killers as somewhere on the continuum between organized and disorganized offenders. On the other hand, investigators have discovered important differences between sex offenders who kill their victims and those that do not. Chan and Beauregard (2015) found that homicidal sex offenders appear to be more likely to select a victim of choice, report deviant sexual fantasies, mutilate their victim, and to admit to their offense upon apprehension. Nonhomicidal sex offenders, by comparison, appear to select victims with distinctive characteristics.

Furthermore, sexual offenders who befriend, recruit, mentor, or counsel children over an extended period of time in order to sexually victimize them are not included as we believe that they represent a significantly different category of offender in terms of style, profile, and motives. Scholars of sexual offending against children make clear distinctions between those who develop an enduring relationship with the victim before the molestation and those who hunt, target, and sexually coerce children in the same way that a rapist might attack an adult victim. For instance, Groth, Hobson, and Gary (1982) differentiated between “child molesters” and “child rapists” in their research based largely on their observation that child molesters generally use a grooming process on their victims, whereas child rapists do not as their assaults occur suddenly. Moreover, Canter, Hughes, and Kirby (1998) argue that those who use “aggressive coercion” against child victims are a distinct type. Canter et al. make a distinction between the interpersonal/intimate type (the groomer) and aggressive or opportunistic offenders who stalk and hunt children. In the case of child rapists that are criminal opportunists (and not groomers), the victimization of the child is often a one-time occurrence—the interaction between the adult offender and the child victim is not repeated because the sex act constitutes the extent and duration of their relationship. In the case of an intimate offender (the groomer), the relationship is long term and the molestation is likely to occur more than once.

Thus, our sample does include child rapists that attack victims in a manner consistent with Hazelwood and Warren’s (1990) typology. We excluded, however, rapists that were constructed in news accounts as groomers. According to Knoll (2010), grooming is “the process by which sex offenders carefully initiate and maintain sexually abusive relationships with children. Grooming is a conscious, deliberate, and carefully orchestrated approach used by the offender. The goal of grooming is to permit a sexual encounter and keep it a secret” (p. 374). The grooming process, then, is distinctly different from the blitz–con–surprise model in that it involves the cultivation and maintenance of a relationship with a child that permits a sexual encounter. This is not to say that victims of child molestation consent to their attack. As Lanning and Dietz (2014) argue, “The term compliant child victim is used to describe those children who in any way, partially or fully, cooperate in their sexual victimization without the threat or use of force or violence” (p. 2823). To “cooperate” with molestation means that the victim may believe—as a result of extended offender manipulation—that he or she is voluntarily taking part in a sexual relationship. A molester may prey on a child’s need for adult attention or ply them with gifts to facilitate closeness that he or she escalates into sexual intimacy. This is not to say that a molested child is not a victim. The grooming process, however, is different in kind from the hunting and luring and physically coercive acts perpetrated by the offenders that fall under our conceptualization of serial rape.

Having established these conceptual parameters, we narrowed our sample to those cases that fit our conceptualization of serial rape. To distinguish cases from spree or mass rapists, as in a murder series, we used victim count and time parameters. That is, guided by our definitional parameters, we identified and entered a rape series as the executed or attempted sexual coercion of three or more victims over a period greater than 72 hr. Using our conceptualization of serial rape, we identified 1,037 offenders from 1940 to 2010.

We focused on media representations of active serial rapists, the time period during which they were active, and a variety of variables related to offender characteristics and attack styles. For the purposes of the current study, four variables were used to form a basic social profile. Those measures include age, race, sex, and occupational class.

Age

This variable is defined as the age at which the offender committed their first known act of rape in a series. For example, one offender in our sample was accused of 11 different rapes in the 1950s and was subsequently arrested and convicted in 1957 when he was 18. News accounts stated that he began his series of rapes in 1955, making him 16 years old when the crimes first occurred (Chicago Tribune, 1957). In some cases, where the assailant was still at-large, we estimated age based on victim estimates of age range. When a range of ages was given (e.g., 25–30), the median age was used. Median age estimates were used in 8.5% of the cases. Age of offender was present in about 91% of the cases (n = 947).

Race/Ethnicity

The race/ethnicity of the offender was determined based on identifying language in each news account. Of the total number of cases of serial rape, only 29% had an identified race or ethnicity (n = 297). Often times, race was assigned based on victim descriptions when a perpetrator was considered at large and had not been caught. In the case of the “Horn-Rimmed Glasses” rapist from 1969 in Edison and Orioles Park, Illinois, the Chicago Tribune stated, “at least four recent victims have described the assailant to police as a white male, age 25 to 30…” (p. B24). In rare cases, race or ethnicity was not stated in the article but a police sketch was present. In these few cases, the data coders used their observations of the features depicted on the artist’s rendering to estimate race or ethnicity. If the coding team all agreed on a particular racial or ethnic category, then that category was entered. If the team could not agree, then a missing value was entered for race.

Sex of Offender

The sex of offenders was either clearly stated in news accounts or was deduced by examining the use of pronouns by either police or victims. Sex was identified in 100% of the cases. All but two of the cases (n = 1,035) were men; two of the active rapists in our sample were identified as women. In one case, the female assailant was part of a collaboration (1 of the 25 rape series collaborations in our sample). And while the rape victims in our sample were overwhelming female (96%), 4% of the cases involved males. For example, The Baytown Rapist was arrested and convicted for a series of crimes in 2006 in Baytown, Texas. According to news accounts all five of his victims were male (Los Angeles Times, 2006). Based on media accounts, however, the rape series in our data are overwhelmingly male offender–female victim attacks.

Occupational Class

Our measurement of occupational class draws on Colvin and Pauly’s (1983) class fraction typology and Vander Ven and Cullen’s (2004) occupational class coding scheme to estimate occupational class for each of our offenders. Colvin and Pauly (1983) divided the working class (i.e., those who sell their labor in the marketplace) into three discrete class fractions. Fraction I employees are those unskilled, nonunion workers found in the secondary labor market: agricultural labor, small manufacturing firms, retail and food services, household domestic work, and low-level clerical jobs (e.g., stock clerk). Fraction II laborers include skilled blue-collar workers (e.g., welders, plumbers, steelworkers, and auto assemblers), skilled clerical workers (e.g., secretaries), craftsworkers, and lower level supervisors in the trades or service industry. Fraction III status was assigned to professionals, self-employed entrepreneurs, highly trained medical personnel, technicians, managers, and advanced government workers. Offenders were assigned a class fraction status based on the description or title of their job in the news article. Occupational status was referred to in 35% of the cases (n = 362).

Some of the offenders were described as students, while others were reported as being chronically unemployed or homeless. Other offenders were described as spending much of their lives incarcerated with no known work history. Therefore, we created categories for the chronically unemployed/homeless, the chronically incarcerated, and for students. Offenders were coded as “chronically unemployed/homeless” or “chronically incarcerated,” if they were described as such in associated news accounts. Furthermore, offenders described as high school or college students were coded as “student” for the occupational class variable. This coding scheme produced six occupational class categories: chronically unemployed/homeless, Fraction I: secondary labor market, Fraction II: semiskilled laborer, Fraction III: professional, chronically incarcerated, and full-time student.

Offense Date Range

This variable is defined as the years during which media accounts reported that a serial rapist was active. The offense date range was determined by the specific years reported by police or victim accounts. If a newspaper report stated that an offender began his series in a certain year and was caught and processed in a subsequent year, we estimated that the assailant was active during all years in between. For example, one offender in our sample was convicted in 1976 for a series of rapes in Porterville, California. The Los Angeles Times stated that the offender “…pleaded guilty to raping nine women between July, 1974 and July, 1976” (1976, p. B2). This would suggest that the offender was also active during 1975. Occasionally, a rapist would be deemed a serial rapist, get caught and convicted and upon release, would commit a series of rapes again. In this instance, the offender could not have been active while incarcerated, as his attacks were temporarily suspended. For example, a serial predator was arrested and convicted in 1966 for a series of rapes in Hermosa Beach, CA, and was released in the 1970s. While he was out of prison, the offender committed another series of rapes in 1972 and 1973, making his known offense date range: 1966, 1972, and 1973 (Los Angeles Times).

Attack Style

Hazelwood and Warren (1990) identified three basic types of offender approach: the con, the blitz, and the surprise. The con approach involves the use of a theatrical ruse to temporarily gain the trust of a potential victim. One offender, for example, was convicted in 1980 for attacks on 14 women. According to news accounts, he would gain access to the victims’ homes by saying he was a plumber sent over to check on their house or pretend he needed help with his car and once inside, would attack (Leeds, 1980).

The surprise approach involves premeditation and often a preselection of victims, however, unlike the con approach, this method does not rely on personal interaction with the victim beforehand. For example, an offender convicted in Los Angeles, California, in 1978, was dubbed by police “The Pillowcase Rapist” because he would sneak in through windows and place a pillowcase over the victim so they could not identify him (Decker, 1978). The blitz approach is opportunistic and typically involves an unexpected attack on the victim. The “Halls Hill Rapist” in Arlington, Virginia, would grab women from behind after they exited buses and drag them at knifepoint behind buildings (Dougherty, 1981).

In addition to these three methods of attack, our analysis identified a previously neglected method of assault—drugging. Drugging was considered a discrete method because it employs different tactics that do not fit into the previous three categories. Drugging involves the use of medicine, sedative, or a street drug to subdue the victim (e.g., laughing gas). In the 1980s, for example, a physician admitted to raping five of his patients while they were sedated. The offender informed patients that he needed to put them under sedation to perform small surgeries: “[The offender] admitted that he had anesthetized five women patients between 1982 and 1984 and…then sexually assaulted them” (Myers, 1987, p. A2).

The drugging style differs from the other attack categories in important ways. First, in most cases (73% of drugging series), the drugger’s victims entered the attack site (e.g., a doctor’s or dentist’s office) voluntarily to seek medical treatments and without being lured or persuaded through face-to-face interaction. While the con-rapist uses rhetorical strategies to move the victim to the attack site, the medical professional, who drugs and rapes, draws in victims by simply operating a medical services office. This type of drugging rapist uses his professional establishment and access to incapacitating drugs as resources in his attacks. The other 37% of the drugging rapists used drugs or alcohol to incapacitate a victim while on a date or during a professional interaction (e.g., a photographer who dosed victims during a shoot). Thus, unlike blitzers or surprise rapists who hunt in public spaces and use brute force to overcome victims, or con-rapists that create a dramaturgical ruse to isolate victims, the drugging rapist interrupts an otherwise legitimate and noncoercive interaction with the surreptitious administration of drugs in order to render a victim unconscious and vulnerable to attack.

Offender Social Profile and Style

The social profile generated by the current study reflects media representations of over 1,000 active serial rapists and 70 years of offense tracking. Our sample is unique in that it draws from newspaper accounts of serial offenders in every stage of identification and criminal justice processing (e.g., at-large offenders, arrested, convicted, and incarcerated). Our social profile includes estimates of sex, age, race, and occupational class. With respect to sex, all but 2 of our 1,037 offenders were male. Regarding offender–victim relationships in serial rape cases, news accounts rarely provided explicit information on the known relationship prior to the attack. While it appears that the overwhelming majority of the attacks occurred between strangers—especially in the case of blitz, surprise, and con attacks—there were doctor–patient drugging cases where the offender and victim appeared to have at least some previous history of interaction through the administration of medical services.

Offender Style of Attack

Attack style patterns are displayed in Figure 1. Seventy-four percent of the cases had an identified method of attack (n = 768). In some cases, offenders used more than one attack style, in which case all styles were entered and counted. Five percent of the rapists described in news accounts used more than one attack style in their series (n = 52). The most common combination of attack styles was the use of blitz and surprise during the series. Thirty-five percent of offenders that used more than one style combined the blitz and surprise methods.

OPEN IN VIEWER

Figure 1.

Serial rapist style of attack.

As illustrated in Figure 1, 47.4% (n = 389) of our cases involved offenders that employed the surprise approach, 27.9% (n = 229) used the con approach, 22.2% (n = 182) employed the blitz style, and 2.6% (n = 21) used the drugging strategy.

Serial Rape, 1940–1963

Figure 2 depicts historical trends in media representations of serial rape over the 1940–2010 time period. The 1940–1963 period is generally characterized by consistently low levels of media-reported serial rape. According to newspaper coverage of this period, the number of active serial rapists range from a low of 0 in 1940 to a high count of 13 in 1953. The number of serial rape accounts during this period is in single digits for 20 of the 24 years included in the span of analysis. The yearly average for number of media-reported serial rapists during this era is 7.04.

OPEN IN VIEWER

Figure 2.

Historical trends in media representation of American serial rape, 1940–2010.

It might be argued that these consistently low numbers are an artifact and an expression of the limitations of using newspaper data. That is, it may be that the large newspapers included in our sample were less likely to capture rape series’ taking place in more remote or rural areas or in regions that fall outside of the radar of large city newspapers. News sharing and the broad dissemination of crime stories are likely to have been relatively lower during this period as compared to more contemporary news spreading practices. As Jenkins (1992) points out, large newspapers like The New York Times significantly expanded their regional news coverage in the 1950s. Serial murder scholars, however, suggest that large newspapers did, in fact, uncover stories that reached far beyond their local area. According to Jenkins (1992, p. 4), apparent increases in serial murder cannot be explained by changing news reporting practices alone:

Throughout the century, major serial murder cases have usually been viewed as stories of great journalistic interest, and between about 1900 and 1940 the New York Times reported extensively and enthusiastically on serial murder cases as far afield as Colorado, Iowa, Alaska, and South Dakota…Even so, the danger of overlooking cases prior to the 1960s is reduced by using Newton’s Hunting Humans, which draws on regional and local newspapers from no fewer than 25 states: papers such as the Eagle-Beacon of Wichita, Kansas; the Daily News of Anchorage, Alaska; and the Clarion Ledger of Jackson, Mississippi. Even a search as wide as this apparently failed to find a large number of midcentury cases that escaped the attention of the major metropolitan press.

Serial Rape, 1964–1979

Figure 2 illustrates a steady increase in the number of newspaper accounts of serial rapists during the 1964–1979 time period. According to newspapers accounts, the number of offenders ranges from a low of 13 in 1968 to a high of 59 in 1979. The yearly average for the number of serial rapists reported by news outlets during this 16-year period is 26.9.

The routine activities approach (Felson & Eckert, 2016) may offer a useful theoretical explanation for the rising trend in serial rape between the mid-1960s and 1970s. That is, cultural changes, policy shifts, and the increasing spatial mobility of victims and offenders may have tipped the social climate to the advantage of motivated offenders who were increasingly likely to find suitable targets in a context lacking capable guardianship.

Serial rapists, by and large, appear to exploit or manufacture pockets and moments of low guardianship in order to increase target vulnerability. The blitz rapist, for example, relies in part on the existence of solitary victims traveling in unguarded spaces. Blitzers generally mount sudden, explosive attacks on targets in open spaces. Any social trend that increases unaccompanied travel, then, will result in increasing levels of predatory crime. Similarly, surprise rapists exploit moments of low guardianship when they, for example, stealthily break into the homes of single-living women or lie in wait in unguarded public rooms (e.g., apartment laundry rooms). The social and cultural conditions that emerged in the 1960s and 1970s may have helped the serial rapist to accomplish his or her goals more efficiently and more frequently.

As Jenkins (1992) argues, the expanding cultural freedoms of young people during the 1960s and 1970s may have had the unintended effect of putting more people in harm’s way. With reference to increasing vulnerability to serial predators, Jenkins (1992) posits that “the vital changes in the 1960s might have been the greater independence of the younger generation and changes in their sexual behavior and attitudes. This greatly enhanced the opportunities for a potential offender to find himself or herself in intimate circumstances with a victim, and the increased physical mobility of these years made it less likely that a young person’s disappearance would be immediately noticed” (pp. 14–15).

Furthermore, one might argue that the 1960s and 1970s represent a period of emergent norms that encouraged people to be more accepting and trusting of strangers and those with alternative lifestyles and to reject the assistance of formal agents of social control. According to Jenkins (1992, p. 15),

Changes in mores increased a willingness to experiment with alternative belief systems and life styles and made many people less prepared to reject or suspect individuals who might appear strange or deviant. The sum total of these changes was to facilitate encounters between strangers that might have been far more difficult only a few years previously. At the same time, the political fragmentation of the late 1960s discouraged young people from invoking police assistance in what might have been seen as suspicious circumstances.

Serial Rape, 1980–1991

As depicted in Figure 2, the period from 1980 to 1991 is characterized by a brief decline in newspaper accounts of serial rapists from 1980 to 1984 followed by a steady climb from 1985 to the highest recorded level in 1991. The number of media reported cases for 1991 is 84 or twice the number of estimated in 1983. The brief decline in offenders identified by news outlets during the 1980–1984 years matches the general decline in American predatory crime identified by a host of researchers. Steffensmeier and Harer (1987, 1999), for example, observed drops in both rape and homicide during this period. According to the authors, this brief decrease in predatory crime is at least partially explained by changes in the age structure in American society. Since predatory crime peaks during the late teens and early twenties, fluctuations in the age composition of a population can impact the crime rates of that population (Steffensmeier, Streifel, & Harer, 1987). The most crime-prone age group shrank as a proportion of the population in the 1980s, which helps to explain early 1980s declines. According to Steffensmeier, Streifel, and Harer (1987), “In the United States…both the dramatic increases in the overall crime rate in the 1960s and early 1970s and the decreases in the rate in the 1980s are linked to the post–World War II ‘baby boom’ and the post-1960 ‘baby bust’” (p. 702). Steffensmeier and Harer (1987) use age-adjustment models to demonstrate that as much as 40% of the 1980–1984 decline can be explained by changes in the age structure of the population.

Figure 2 shows a steady increase in media accounts of active serial rape from 1985 to 1991. This trend follows the steady general increase in serial rape accounts from the mid-1960s onward. As previously noted, according to our estimates, the number of active serial rapists reported by the media reached a zenith in 1991. The early 1990s spike in serial rape is consistent with many other studies that show a dramatic increase in predatory crime at this time. Crime researchers have observed a dramatic spike in a variety of offenses in the early 1990s including increases in homicide and rape (see Levitt, 2004; Steffensmeier & Harer, 1999). Our data suggest that serial rape—as depicted by media sources—followed similar historical trend lines.

Serial Rape, 1992–2010

According to media accounts, the number of serial rapists declined steadily from 1992 to 2010, with a brief increase in 2003–2004. The dive in cases of serial rape identified by media sources picks up again in 2005 and plummets to single digits in 2009 and 2010 for the first time since 1963. This gradual, substantial drop in serial rape accounts reflects the broader decrease in American violent crime over this time period. According to Levitt (2004), “crime fell sharply in the United States in the1990 s, in all categories of crime and all parts of the nation. Homicide rates plunged 43% from the peak in 1991 to 2001, reaching the lowest levels in 35 years. The FBI’s violent and property crime indexes fell 34 and 29%, respectively, over that same period” (p. 163).

The dramatic drop in predatory crime (in general) and in serial rape (in particular) may be attributed to a variety of factors. Drawing from the routine activities approach, it may be argued that social trends, criminal justice practices, and the increasing use of surveillance technology may have effectively reduced the extent to which motivated offenders, suitable targets, and the lack of capable guardianship converged in time and space (see Baumer & Wolff, 2014; Felson, 1995).

The pool of motivated offenders was undoubtedly reduced by legislation and judicial practices that grew the American prison population to unprecedented levels during this period. According to Conklin (2003), “…13 to 54 percent of the decline in crime rates in the 1990s was due to growth of the prison population” (p. 95). Similarly, Levitt (2004) argues that “The 1990s was a period of enormous growth in the number of people behind bars…. By 2000, more than two million individuals were incarcerated at any point in time, roughly four times the number locked up in 1972. Of that prison population growth, more than half took place in the 1990s” (p. 177). To the extent that many would-be serial predators were among the masses incarcerated during this period, it is no wonder that serial rape experienced a downward trend.

The decline in serial rape since 1992 may also be explained by increasing and more sophisticated guardianship. The growth in police forces is one way to measure increased guardianship. Levitt (2004) reports that the number of American police officers grew by 14% in the 1990s. More importantly, advancements in law enforcement technology and investigation techniques have made it increasingly difficult for serial predators to maintain a crime series. According to Jenkins (1992), modern law enforcement officers are more skilled and experienced at discovering and tracking serial offenders: “The agency’s modern counterparts would be more familiar with patterns of serial homicide and would thus tend to speculate with greater freedom. Modern agencies also have superior recordkeeping techniques and more experience of interagency cooperation, enhancing the likelihood that a suspect could be linked to a large number of earlier offenses” (pp. 4–5). Furthermore, the forensic sciences advanced rapidly during the apparent decline in serial offending. In the late 1980s, DNA evidence entered into the court system. The Committee on Identifying the Needs of the Forensic Sciences Community (2009) suggests that over the past 20 years, science and DNA evidence have allowed for more prosecutions and convictions, as well as solving previously unsolved cases. The FBI (2008) has drawn increasingly from forensic evidence based upon a person’s DNA and a nationally maintained database which has allowed for prosecution in previously unsolved cases, and allowed for linkage between cases that were not known before. These advancements have allowed for an increasing potential for identifying offenders and, thus, have made serial offending more difficult to accomplish. Since serial rapists tend to have more forensic awareness than single-victim offenders (Corovic et al., 2012), advancing technology may have served to discourage would-be offenders from acting upon criminal desires.

Finally, the broad use of everyday surveillance and security technology in American life has undoubtedly resulted in fewer serial offenses. Guardianship, in the form of surveillance cameras on city streets and in public establishments, for example, has increased dramatically in recent decades. According to Welsh and Farrington (2009), CCTV cameras are currently being installed at an unprecedented rate in the United States and have been associated with crime decreases in some settings. Since serial rapists—especially those that hunt at night in sparsely inhabited spaces—rely on moments of reduced guardianship, the growth in surveillance technology may have had the effect of reducing their offense opportunities.