Is It Time to Revisit the Deﬁnition of Serial Homicide? New Evidence and Theory

: Historically, serial homicide has been deﬁned in various ways by experts. Recently, there have been renewed efforts to arrive at a consensus deﬁnition, yet these efforts have not yet been resolved. At the heart of the controversy appears to be the prioritization of either qualitative deﬁni-tional features, such as offenders’ intentions and motives, or more observable quantitative features, speciﬁcally a minimum threshold of completed murders. The present technical note brieﬂy summarizes this controversy before considering new empirical and theoretical research developments. These developments support a deﬁnition that includes a three-victim minimum threshold of forensically linked murderers by the same person(s), occurring in separate events over time, wherein a primary motive is often personal gratiﬁcation (leisure experience).


Introduction
Separate literature reviews have shown that the term serial homicide (or serial murder) has never been consistently defined, with definitions often varying regarding a minimum victim threshold, length of time between murders ("cooling off" period), relationship between offenders to victims, and primary motivation [1][2][3][4]. At least some of these differences may be due to legal and practical approaches to multiple homicide compared to research driven perspectives [1][2][3][4]. For example, the Federal Bureau of Investigation (FBI) in 2008, apparently in an effort to increase practical utility and provide potential assistance for local law enforcement, changed its definition of serial murder to "the unlawful killing of two or more victims by the same offender(s) in separate events", and then offering differing potential motives-anger, criminal enterprise, financial gain, ideology, power/thrill, psychosis, and sexually driven-as general categories to help guide investigations [5]. However, many researchers continue to struggle with how to operationally define serial homicide-some simply avoid operational definitions, other researchers rely on definitions put forth by others (including the FBI 2008 definition), and still others propose their own particular definition for their research purposes [1].
Two decades ago, Skrapec, after reviewing the available literature at the time, defined serial homicide as "three or more forensically linked murders committed as discrete events by the same person(s) over an extended period of time and where the primary motivation is personal gratification" [2] (p. 22). In their more recent review, Adjorlolo and Chan observed that many of the qualifiers (i.e., "cooling off" period, relationship of victims to offenders, specific intentions) in various serial homicide definitions have been shown to be insignificant [1]. They follow Skrapec in defining serial homicide as (a) having forensically linked murders by the same person(s) in discrete events over time, and (b) the primary motive is personal gratification, yet they lower the victim threshold from three completed murders to two, which is consistent with that of the FBI.

Definitional Prioritization: Offender Intention or Minimum Number of Victims?
In the third edition of their Crime Classification Manual, Douglas and colleagues, in presenting their definition of serial murder, followed the FBI 2008 definition in reducing the victim threshold from three victims (beginning in their first edition) to only two, while emphasizing offender intent and motivation, rather than numbers of victims [6]. However, these authors also continue to cite the three-victim threshold definition created by the United States Congress in 1998 for serial killing, in which such murders have common characteristics and at least one killing occurring in the United States [6]. Attempts to understand the rare phenomenon of serial homicide remain difficult given that multiple definitions are available and are commonly used, and the purposes of definitions differ (legal vs. academic). Proponents of a two-victim minimum threshold [1,6,7] argue that psychological and situational attributes, especially strong intentions and inclinations to murder multiple people in separate events if not constrained, are of primary definitional importance. In other words, an offender with two (or perhaps fewer) victims, but with a mindset and plan to kill more, would likely kill more victims if he or she was not apprehended. The two-victim threshold generally has been favored in legal settings [8]. It has also been noted that regarding serial homicide, in particular, the details of cases matter, in which understanding such details can generate valuable new information [7].

New Empirical Evidence and Theoretical Insights on Serial Homicide
The present author agrees that details of serial homicide matter and, as with understanding any complex behavior, diverse research methods are warranted depending on specific questions. Both qualitative and quantitative studies, which have different strengths and limitations, should be welcome, along with research from multiple relevant disciplines. That being said, both recent empirical research and new theoretical insights appear valuable in helping to shape and clarify basic definitional parameters of serial homicide.

Empirical Research on Minimum Victim Threshold
A recent large study utilized multinomial regression analysis in an attempt to identify potential behavioral differences in male serial homicide offenders (N = 2275) who had committed two murders compared to those with three to seven victims and also offenders with eight or more murders [9]. The findings showed that offenders with only two victims differ significantly from their more prolific counterparts regarding both motive and crime scene behaviors, leading investigators to conclude that three should be the minimum number of victims in defining serial homicide [9]. A similar study of both male and female serial homicide offenders (N = 1258) found that two-victim offenders differed from their more prolific counterparts on important variables related to victim demographics (sex and race) and prevalence of rape at or near the time of murder [10]. These two large empirical studies found clear differences between serial homicide offenders (minimum of three victims) from serial homicide offenders with two victims but perhaps a similar psychological mindset [11]. Together, these studies clearly call for a three-victim minimum threshold, rather than two.

Insights from General Behavior Theory
In their thorough review on serial homicide definitions, Adjorlolo and Chan refer to the robust Theory of Planned Behavior (TPB) and its predecessor, the Theory of Reasoned Action (TRA), to support a need for experts to consider behavioral intention, not just behavior, in definitions of serial homicide [1]. These theories, particularly TPB, have been empirically tested across numerous diverse behaviors [12,13]. However, it is certainly possible to interpret these theories differently in conceptualizing serial homicide definitions. First, TPB was created by simply adding the construct "perceived behavioral control", which is akin to self-efficacy, to TRA, which then accounts for significantly more variance [12,13]. Self-efficacy (or a similar construct) seems to be essential for theories attempting to explain behavioral changes. More importantly, while behavioral intention is correlated with actual behavior, such correlation is far from perfect. In TPB, "attitudes toward the behavior", "subjective norms", and perceived behavioral control all contribute to behavioral intention, while perceived behavioral control additionally has a direct pathway to actual behavior [12,13]. The Integrated Behavioral Model (IBM), a more recent extension of TPB, adds the constructs "knowledge and skills", "salience of behavior", "environmental constraints", and "habit" to predict the likelihood of continued behavior [12,13].
Besides having predictive value regarding a wide range of human behaviors, IBM constructs have relevance to serial homicide behavior (as separate from serial homicide intention). For example, having the knowledge and skills to commit multiple murders and/or fewer environmental constraints, would be significant in differentiating offenders with two victims compared to those with more than two. Similarly, more prolific offenders would seem to vary regarding killing as an established habit and/or killing as a salient behavior compared to those with only two victims.

New Research on Leisure as a Primary Motive
Experts have known for a long time that many serial homicide offenders enjoy killing, and based on prior literature reviews, scholars have included the general motive of personal gratification as a key element of serial homicide definitions [1,2]. Skrapec notes the German word "Lustmörd" refers to an offender's desire to kill and the pleasure derived from it [2]. She adds that the personal meaning of killing and its intrinsic motivation are important features of serial homicide [2]. Personal meaning, intrinsic motivation, and relative choice are important attributes of leisure experience [14], and a recent analysis of hundreds of serial homicide cases found that for such offenders, their killing behavior functions as a particular form of leisure-as a game, simple fun and enjoyment, thrills and intense sensations, and/or as a form of personal celebration [15]. These findings are remarkably consistent with those of Fridel and Fox [9], which showed that offenders with three or more kills are significantly much more likely to kill primarily for personal enjoyment, compared to counterparts with only two victims.
While committing serial murder often appears to require considerable effort, skill, and planning that are consistent with more serious leisure activities, some offenders kill more casually and spontaneously as a form of simple sensation seeking and active entertainment [16]. Additionally, serial homicide offenders must negotiate intrapersonal, interpersonal, and structural constraints to desired leisure in order to commit murder successfully-the ability to negotiate constraints directly impacts the killing success of offenders (number of victims) [14]. Leisure research suggests that important characteristics of engaging in desired leisure, such as being able to successfully negotiate particular constraints, having sufficient knowledge and skills to engage in the desired behavior, and enjoying the activity sufficiently to pursue it repeatedly, are consistent with the theoretical constructs of IBM. Killing for personal enjoyment or gratification (i.e., leisure experience) is a salient component in most serial homicide cases with three or more victims, although other motives certainly are present in some cases.

Summary and Conclusions
While an agreed-upon definition of serial homicide has been elusive, recent scholarship, both empirical and theoretical, brings new insights that can clarify important points of contention-specifically, whether to prioritize behavioral intention for offenders who have committed two murders, or to set a minimum victim threshold higher than two. Of course, coming to an agreement on defining serial homicide, considering the differences between practical and academic purposes, is not necessary, though it would be extremely valuable in facilitating communication between experts in different settings [8,9]. Large empirical studies [9,10] suggest that there are important differences between offenders with two victims compared to those with greater numbers (minimally three victims). Furthermore, recent theoretical developments also support a victim threshold greater than two. Intention differs from behavior, and the likelihood of repeated behavior, both as leisure and otherwise, depends on specific constructs that are directly relevant to serial killing. While some factors implicating success in repeated behavior, such as serial homicide, are external to the individual (notably, structural and environmental constraints), others are internal, thus varying among potential and actual serial homicide offenders.
Although serial homicide offenders seem to continue to kill if left to their own volition, it is also important to remember that the notion of such offenders being unable to stop killing is a common myth [3,17]. To the contrary, serial homicide offenders derive sufficient personal gratification from killing, as a leisure experience, that they strongly desire continued engagement with such homicidal pursuits.
In light of new empirical and theoretical scholarship, it is impressive how both Adjorlolo and Chan [1] and, especially, Skrapec [2], over two decades ago, arrived at surprisingly accurate definitions of serial homicide, given the recent research. Current scholarship continues to largely support the definition proposed by Skrapec in 2001, that serial murder is "three or more forensically linked murders committed as discrete events by the same person(s) over an extended period of time and where the primary motive is personal gratification" [2]. Both definitions emphasize the importance of linking separate crime scenes by physical evidence, including cases where no suspect has been identified, rather than accepting offenders' claims for murders that have not been solved. However, research has failed to support a significant "cooling off" period or time segments between murders [18,19], thus the time dimension outlined by Adjorlolo and Chan, that murders are committed "over a period of time", appears to be more precise in that regard [1].
In short, new research supports the three-victim minimum favored by Skrapec [2], the time dimension proposed by Adjorlolo and Chan [1], and the killing as motivated by personal enjoyment (as leisure experience) outlined in both reviews [1,2]. Furthermore, although aspiring or hopeful serial homicide offenders [11] appear to be distinct from actual serial homicide offenders [9,10], there remains much that researchers and professionals can continue to learn from this group. To conclude, it is strongly recommended that, based on new research, experts follow the conceptualization refined most recently by Adjorlolo and Chan [1], but with an updated three-victim (fatal) threshold rather than two.