This sample was compiled through the use of Internet archive and search engine searches of the key words American serial killers, female serial killers, serial killers, serial murder, serialists, serial murderers, serial homicide, female murderers, and multiple murders. Information was collected from media sources such as publicly available police documentation, court documents, encyclopedias, and books. These sources were reached through Internet archives, such as Google news archive, Internet search engines, public libraries, the Library of Congress, The California Digital Newspaper Collection, the library resources of the Illinois School of Professional Psychology at Argosy University, and websites containing collections of media articles about serial murder, such as Murderpedia.org and The Unknown History of Misandry. The authenticity of the information of the latter sites was examined through random checks and verbatim comparisons to the texts in the original media archives. The data were collected and coded by a single collector, the researcher. The data were analyzed utilizing SurveyMonkey and SPSS.
The initial size of the sample was 101 subjects, meeting the original criteria to include females who had murdered at least two individuals on at least two separate occasions, and at least some of the murders were committed in the United States. During the data collection, eight subjects were removed from the sample because of lack of sufficient or reliable information. Thus, the final sample size became 93 (n = 93) subjects. The sample included one female child, who was excluded when analyzing variables that appear irrelevant because of age. The goal of providing a minimum of 10 information sources per subject was met for 74% (n = 69) of the subjects.
The highest number of subjects, 40% (n = 37), were born between the years 1851 and 1900, 23% (n = 21) were born between 1901 and 1950, 18% (n = 17) between 1951 and 2000, 4% (n = 4) between 1801 and 1850, one subject was born before 1800. No information about the year of birth was available for 14% (n = 13) of the women in the sample. The highest number of female serial killers born in the same year was three, and the years were 1884 and 1886.
As it relates to the racial background of the subjects, 94% (n = 88) appeared to be Caucasian, and 5% (n = 5) were African American. There was no representation of other races in the sample.
Tracking marital status quickly proved irrelevant, as it appeared to change frequently because of multiple marriages, spouse deaths, and divorces. An attempt was made to collect data concerning marital status at the time of apprehension, but that proved to be inadequate as well as subjects sometimes were apprehended repeatedly. Results concerning the number of marriages for the subjects in the sample are displayed in Figure 1. The women in the sample were married 2.3 times on average.
Figure 1. Number of marriages per subject.
Note. Answered = 93, skipped = 0.
Most of the subjects had biological or adopted children. Because of the frequent changes in marital status, stepchildren were not considered unless they were adopted by the subject. The child subject was excluded from this variable. The women in the subsample had 2.9 children on average. Results concerning the number of children per subject are displayed in Figure 2.
Figure 2. Number of children per subject.
Note. Answered = 92, skipped = 0.
Subjects’ occupation was explored through the narratives presented in the media. As most were not professionally trained, they sometimes engaged in more than one type of job. Of the women, 17% (n = 16) ran their own businesses or had a farm, 16% (n = 15) worked housekeeping jobs, 15% (n = 14) were caregivers, nurse’s aides, or babysitters. Nurses with formal training or licensing represented 11% (n = 10) of the sample. Consistent with the spirit of the times and culture, 14% (n = 13) of the women were homemakers or socialites. Unfortunately, data were absent for 18% (n = 17) of the subjects.
Subjects’ narratives were explored for the presence of difficulties in their families of origin as well as the families they later created themselves. Unfortunately, information about the subjects’ families of origin was quite scarce. Of the narratives, 50% (n = 47) did not contain data about the issues subjects encountered in their families of origin. Of the women in the sample, 13% (n = 12) experienced some form of parental abandonment, usually in the form of being raised by relatives, foster, or adoptive parents. Of the families of origin, 11% (n = 10) were described as prominent or respectable in their local communities, which may have presented pressure on the children in the family. Of the women, 10% (n = 9) experienced financial hardship in their families of origin, and 9% (n = 8) experienced exploitation as children, usually in the form of being forced to leave school and work or care for an ill family member. A more extreme form of child exploitation involved torture or starvation. Also, 9% (n = 8) of the subjects experienced immigration with their families of origin. Another 9% (n = 8) experienced the death of a parent during their childhood. Of the total, 5% (n = 5) reported having experienced physical and sexual abuse, respectively. Criminal behavior was present in 3% (n = 3) of the families of origin of the serial killers in the sample. While this number must be accepted with caution considering the limited availability of information on the topic, it may point to the lack of multigenerational patterns in the families of American female serial killers.
Information about the issues subjects experienced in their own families seemed to appear more frequently in the media, yet, far from abundantly, as 19% (n = 18) returned no data. Of the narratives, 20% (n = 19) contained references to relational issues with no known domestic violence. Of the sample, 19% (n = 18) reported spousal abandonment, separation, or divorce in their families and 16% (n = 15) reported infidelity on the side of their spouses or their own sides. Somewhat consistent with the state of medicine at the time when most of the women in the sample lived, 14% (n = 13) experienced a loss of a child (not to homicide) or separation of their children because of circumstances or forced by the courts. An equal number of subjects, 13% (n = 12) experienced domestic violence or financial hardship. Also, an equal number of women, 11% (n = 10), experienced alcoholism or death of a spouse (not inflicted by the subjects).
Information about the birth order of the subjects in the study was rarely present in the media sources; such data were absent from 65% (n = 60) of the narratives. An additional category of “adopted/raised in foster care” was added to the traditional Adlerian categories of oldest, middle, youngest, and only child to reflect the distinction of families with adopted or foster children. Results regarding birth order are presented in Figure 3.
Figure 3. Birth order of the subject.
Note. Answered = 92, skipped = 1.
The highest number of subjects, 32% (30), committed their first murder between the ages of 21 and 30, 29% (n = 27) of the women in the sample committed their first murder between 31 and 40 years of age, 12% (n = 11) between 41 and 50, 6% (n = 6) before they were 20, 3% (n = 3) were older than 51. No data were available about the age during the first murder for 17% (n = 16) of the subjects. The average age of the subjects during their first murder was 32 years. One subject appeared to be an outlier because of being a child (five years old) and was excluded.
The average number of individuals the subjects in the sample murdered was 5.5. The subjects known as baby farmers, whose number of victims was excessive or no clear number was available, appeared to be outliers and were excluded from the analysis of this variable. If multiple numbers of murders were present with equal reliability of information through various sources, the most conservative number was used.
The method of murder 74% (n = 69) of the subjects utilized included poison, medication overdose, fraudulent medical treatment (excluding surgery), and starvation. 16% (n = 15) used strangulation, suffocation, or smothering when killing their victims, 11% (n = 10) of the women shot their victims, and 6% (n = 6) of subjects stabbed, dismembered, or killed their victims through unlawful surgeries.
Results regarding the relationship of the subjects to their victims are presented in Figure 4. The victims included spouses or romantic partners of the subjects, their own children (biological children only), individuals in their care, patients, or clients, family members or relatives (excluding spouse, children, stepchildren/adoptive children, parents, and siblings), friends, neighbors, co-workers, and employers, step children and adopted children, parents, siblings, strangers, and spouse/partner of love interest.
Figure 4. Subject’s relationship to her victims.
Note. Answered = 93, skipped = 0.
Results regarding subjects’ first victims (per relationship) are presented in Figure 5. A stranger was never the first victim.
Figure 5. First victim per subject.
Note. Answered = 93, skipped = 0.
Murder was not the only criminal activity for most of the subjects. Apart from the completed murders, 44% (n = 44) of the subjects made additional attempts at murder, 24% (n = 22) engaged in fraud or forgery, 13% (n = 12) were involved in theft, robbery, or burglary, and 9% (n = 8) were arsonists. Additionally, 4% (n =4) managed to escape from prison.
Financial gain and material benefits were part of the subjects’ rationale for their murders in 65% (n = 60) of the cases. In 21% (n = 19) of the narratives, rationale involved eliminating a situational stressor, reducing situational anxiety, or reducing situational helplessness. This pattern of problem-solving through murder continued into marriage issues. In 10% (n = 9) of the cases, rationale involved solving relational issues in the marriage, marital dissatisfaction, and marital boredom. Another 10% (n = 9) included jealousy in their rationale for murder, and 7% (n = 6) involved eliminating a romantic partner to engage with a new one or eliminating the spouse of a love interest.
Information about the events or stressors in the life of each subject before the first murder also proved to be insufficient and was missing from 36% (n = 33) of the narratives. Financial hardship was the leading stressor (19%, n = 18) that subjects experienced before they murdered for the first time. Of the subjects, 15% (n = 14) experienced marital problems (e.g., dissatisfaction, boredom, conflict) at that time of their lives. Also, 15% (n = 14) had entered a new romantic relationship (sometimes based on infidelity) or a new marriage prior to their first murder, and 12% (n = 11) had relational issues with a family member (excluding spouse). Of the sample, 10% (n = 9) experienced infidelity, their own or their partner’s. Another 10% (n = 9) had multiple romantic relationships or multiple marriages over a brief period of time (usually several years). Of the subjects, 8% (n = 7) engaged in other criminal activity or their criminal behavior was exposed. Also, 8% (n = 7) experienced job loss or felt dissatisfied with their jobs.
Information about these women’s physical and mental health issues was very difficult to find in the media resources. Information about the mental health issues subjects experienced was absent for 52% (n = 47) of the sample and about health issues for 79% (n = 73) of the sample. Furthermore, available information about mental health issues seemed questionable, as it rarely involved a formal diagnosis or the diagnosis varied from one expert to another. Thus, the results concerning mental health issues should be interpreted with caution. Of the women in the sample, 14% (n = 13) attempted suicide on at least one occasion, 11% (n = 10) had known substance abuse issues (excluding alcohol), another 11% (n = 10) were declared insane, and 8% (n = 7) had issues with alcohol abuse. As it relates to their health, 3% (n = 3) had epilepsy, 3% (n = 3) struggled with obesity, 2% (n = 2) had a hearing impairment, 2% (n = 2) had joint issues or arthritis, 2% (n = 2) had heart disease, and another 2% (n = 2) had cancer.
Media also failed to show an interest in the hobbies and interests of these women. There were no available data for 59% (n = 54) of the sample. Of the subjects, 16% (n = 15) were very involved in their churches and religious activities; 8% (n = 7) had a known obsession with death, funerals, autopsies, or poison; 5% (n = 5) played a musical instrument or sang, and 4% (n = 4) were fond of gardening.
The following descriptions of the women in the sample were provided by their neighbors, friends, families, media, court officials, or anyone who came into contact with them. The descriptions were grouped together by similarity of meaning. Of the women, 26% (n = 24) were viewed as friendly, kind, caring, loving, personable, and devoted mothers and wives. In addition, 6% (n = 24) were described as not showing emotion, remaining calm (especially in difficult moments such as verdict reading or funerals); 23% (n = 21) were viewed as beautiful or attractive; 16% (n = 15) were described as having no conscience, cruel, cold-hearted, and vicious; 13% (p = 12) were seen as unattractive; 12% (n = 11) were seen as manipulative, calculating, and “twisted but not stupid.” Another 11% (n = 10) were viewed as strange or odd. Intriguingly, in two cases of heavier set women, their exact alleged weight was mentioned repeatedly in media articles.
In conclusion, more information is needed to develop a better understanding of who these women were, how they were viewed by others, what their stressors were, and what motivated them to murder repeatedly. Future research in the area should consider exploring the length of time between the first murder and the time they were convicted, or at least tried, as some of the subjects were tried in their 60s and 70s.
Considering the scarce research in the area of female serial murder, this study intended to contribute to the area of knowledge through examining the women in the sample from a systemic perspective. The initial hypothesis stated that there were common features among the subjects in the sample. Availability and reliability of information were major contributing factors to this researcher’s ability to arrive at a definite answer about the veracity of the hypothesis.
However, several conclusions emerged after an examination of the results. Both larger and smaller societal systems tended to support these women through jobs, marriages, and the way others viewed them.
Furthermore, the societal and legal systems faced cognitive dissonance in the way they approached these female serial killers. Some of the women in the sample continued to kill for many years and were tried in old age. Additionally, the descriptions provided in the results were given usually with the knowledge of murder allegations. In other words, people knew that these women likely murdered, and still described them as “caring,” “compassionate,” “nice person,” etc. The legal system appeared consistently reluctant to assess these women as criminals. The court acquitted Agnes Orner for the murder of her second husband deeming it to have been self-defense. However, her next husband met his death in a similar way 13 years later. One judge even stated, “If you were a man, I would hang you” (Prescott Evening Courier, 10 March, 1926, p. 1) during the trial of Elsie Bible Malinsky, who allegedly murdered four individuals, among them, women whose husbands she wanted to marry. Then, three years later, she also murdered one of her husbands because she was dissatisfied with the marriage.
Arguably, one of the best indicators of the way the women in the sample were viewed by society were the nicknames they were given by media and their communities. These monikers often contained historical references, such as Borgia, after Lucrezia Borgia, or legendary characters, such as the Damsel of Death, the Angel of Death, The Black Widow, Lady Bluebeard, or the Good Samaritan (aka Poison Woman).
Some are linked to the method of murder, such as the Antifreeze Killer, Arsenic Anna, Belle the Butcher, or Morphine Queen, the Starvation Doctor, the Trunk Murderess, the Strychnine Saint (aka Mercy Mary), the Derby Poisoner, the Soup Killer, or the Potato Soup Serial Killer.
In other cases, nicknames reflected appearance, character, profession, geographic, or age references, namely Jolly Jane (the Jolly Psychopath), the Giggling Grandma (aka the Jolly Widow, aka Arsenic Nannie), Death Row Granny, the Panhandle Black Widow, Deadly Babysitter, Tiger Lady (aka Dowager of Tehachapi and Carbon Copy Killer), Squirrelly Shirley (a grade school nickname), the Pampered Killer, Murderess of the Century, the Woman in Black, the Wholesale Murderess, the Spoiled Child Serial Killer, Cat Woman, Gypsy Queen, the Serial Killer Mom, Lonely Hearts Black Widow, Vance Avenue Alma, Baby Maniac (Baby Borgia), Illinois Serial Killer Grannie, and Mother Slayer.
It is obvious that with the exception of a few, the nicknames do not reflect mass fear as some of the notorious male serial killers’ monikers do, as indicated by Hickey (2010). The author added that female serial killers’ criminal abilities appear to be underestimated. Kelleher and Kelleher’s (1998) results indicated that female serialists are quite successful and manage to maintain longer killing careers than males serialists do. The narratives of the subjects of the present study suggested that it was not unusual for female serial killers to be tried at an older age. Nancy Doss was in her early 20s during her first murder but was tried as a “grandma” as her nickname indicated (the Giggling Grandma). Marie Noe also killed for the first time in her early 20s but managed to murder 8 of her own children over the course of 19 years before she was tried. Similarly, Lulu Johnson murdered her first victim at 18 years of age but was 63 during her trial. Illinois Serial Killer Grannie Lillie Winter was 51 years old at the time of her first murder, but she was 76 at the time of her trial. Alternatively, Agnes Orner was reportedly tried five times.
The results of the present study suggest that the women in the sample were often described as attractive. While, overall, the women appeared to blend into their communities and remain somewhat invisible, their appearance seemed to set them apart in the eyes of the people who came in contact with them. Does the fascination with serial killers impact the way society viewed them or were they truly beautiful women?
Marital status of the women in the sample was of interest initially, and an attempt was made to collect such data. However, examining the subjects’ narratives further demonstrated that this demographic often changed several times during their lifetimes. An attempt to collect data on marital status at the time of apprehension was made. This approach proved to be irrelevant as well because some subjects were apprehended more than once. Based on the results related to the number of marriages, it is reasonable to conclude that marriage was present in the lives of the subjects. This finding is consistent with Schurman-Kauflin’s (2000) conclusion that women who kill endorsed increased anxiety and depression related to being alone, which respectively resulted in a feeling of powerlessness. Additionally, the multiple marriages of these women arguably suggest that they possess very good social skills, considering homicide was often a part of their past.
It has often been reported that the long-term consequences of one’s romantic decision making have a significant impact upon life satisfaction (Joel, MacDonald, & Plaks, 2013). Current research indicates that unsuccessful romantic decisions are one of the most commonly mentioned life regrets (Joel et al., 2013; Morrison & Roese, 2011). The women in the sample engaged in multiple marriages and often experienced marital issues. Some of their romantic relational decision making seemed very strategic and methodical. Clara Carl killed her first husband for financial gain, then met someone else who she thought was wealthy. Soon, she realized he was not wealthy, but he convinced her not to divorce by making her the beneficiary of his life insurance policy. Emma Hepperman met her last of seven husbands through an ad for a housekeeper. Elsie Bible Malinsky’s first husband divorced her after suspecting that she tried to poison him. She poisoned the wife of her then future husband, so she could marry him, but later divorced him after she felt dissatisfied with the marriage. She then poisoned the wife of her future third husband, after getting a job as the housekeeper in the family. Her third husband reportedly heard this story in the courtroom during her trial. Grace Sims was known for having affairs with married men and then blackmailing them for money. Other serialists appear more opportunistic, such as Lizzie Halliday, who was reportedly married seven times. During one marriage, she ran away with a neighbor while married to Halliday and stole horses. Her partner deserted her, and she got arrested. In general, it appears that the women in the sample made their romantic choices in a manner consistent with interdependence theory, which assumed that people strive to maximize rewards and minimize costs in their relationships and relationship rewards and costs can be assessed against measurable standards (Joel et al., 2013).
Furthermore, in her book, All the Single Ladies, journalist Rebecca Traister spoke of marriage as an organizing principle. It appears female serial killers submit to this principle as they continue to renew marital status. Considering the United States is a country where singles outnumber married individuals (Li, 2014), it is unclear whether this trend will continue in the future among the population of female serialists. Yet, media seem to portray female serial killers as victims who are manipulated by a dominant male (Bonn, 2014). A review of the narratives of the women in the sample was unable to support this view.
In their study on the variation in the sex ratio of females to males who kill a spouse, Gauthier and Bankston (2004) explained that if women sought an escape from an abusive or troubled marriage, choosing the divorce resolution to the problem, if it is accessible, may present financial problems during property settlement. The authors predicted that when treatment of property is controlled, easier access for women to divorce would reduce the rate of female offenders. At the same time, the rate of male killers would increase because, with easier access to divorce for their spouses, men may perceive the need to use more coercive problem resolution that is more likely to result in lethal aggression (Gauthier & Bankston, 2004).
Although, most of the women in the study were not well educated and did not have professional training, examining their occupations suggested that they were flexible, adaptive, and took the initiative. Their criminal behavior and questionable moral judgment aside, the women in the study changed occupations, started new businesses, and reinvented themselves. Dorothea Puente was a prostitute at an early age; then she became a nurse’s aide and eventually ran a boarding house. During various times in her life, Judy Buenoano worked as a nurse’s aide, cocktail server, and owned a nail salon. Belle Gunness owned a store, then a farm. Some women managed to turn housekeeping jobs into marriages (i.e., Ellen Etheridge, Elsie Bible Malinsky, and Lydia Sherman). Others practiced jobs for which they held no proper licensure, such as Linda Burfield Hazzard, who practiced medicine without a proper license. She managed to open a sanitarium offering treatment based on fasting and even published a book about the starvation method of healing, namely Fasting for the Cure of Disease.
Results demonstrated that the women in the sample were often described as showing no emotion. Others labeled them as “emotionless,” “as cool as a cucumber,” “her eyes were dead,” having a “stolid demeanor,” a “bland indifferent beauty,” or “with as much indifference as a cornered opossum.” While these descriptions typically follow media interest related to alleged involvement in murder, they often depicted the subjects’ reactions in emotionally-charged situations such as court proceedings, delivery of the verdict, funerals of family members or loved ones. It should be noted that the described situations typically did not present with demands for expressing empathy for others, but rather, an expression of the subject’s own emotional experience. The results arguably indicate not only a pattern of emotion processing or expression in the subjects but also, the public’s expectation of a different reaction in such circumstances. Not surprisingly, these results are consistent with the findings of Dutton (2012) in his exploration of psychopathy as well as the conclusion of Frei et al. (2006) that female serial killers exhibited psychopathic traits. Dutton (2012) pointed out that ability to stay calm under pressure is one of the indicators of psychopathy. Love and Holder (2014) indicated that shallow affect, impulsiveness, and lack of empathy are some of the characteristics of psychopathy as a personal construct. In their study exploring the relationship between psychopathy, personality, and subjective well-being, they found “psychopathy was associated with elevated levels of depression and negative affect, and low levels of life satisfaction, happiness, and positive affect” (p. 112). While the present study lacks data addressing the particular constructs described by Love and Holder, it is reasonable to hypothesize that the women in the sample experienced negative affect, considering the life problems they encountered and the solution approaches they utilized.
Impulsiveness is often described in the literature as a characteristic consistent with psychopathy (Love & Holder, 2014). The role of impulsiveness in the cases of female serial killers’ murderous behavior arguably remains unclear, as they most frequently utilize poison as their murder method, which requires time and effort to obtain and respectively allows time for the individual to reconsider his or her decision and change behavior. Additionally, their victims often involved relatives and friends, to whom they were frequently exposed. Thus, these murders do not appear to be a result of an opportunistic impulsive choice. Impulsiveness, however, seems to be present in the cases involving a short-term situational stressor, such as a crying baby, where the individual made the extreme choice to murder to achieve relief from the effects of the stressor or the cases of nurses seeking attention and recognition through harming patients and then taking on the role of the rescuer. These women appear to have been seeking situational gain, and their rationale and decision-making process varied from the women who sought financial gain through inheritance or insurance, which suggests a longer premeditated process.
Among the very few known cases reflecting the presence of a multigenerational pattern is that of Elizabeth Potegian, whose mother allegedly killed her six husbands. Potegian murdered her husband and a stepchild, stating “I’ll make ashes of them.” She was an Armenian immigrant, who had moved to the United States with her parents. She married a wealthy widower who had children from a previous marriage. She stated that 40 days after his death she had decided to poison the children. After she had been apprehended, Potegian attempted to blame her mother for the murders. Her mother committed suicide, and upon hearing about it, Potegian attempted suicide in her prison cell. This narrative reveals a strong bond between mother and daughter, and arguably suggests more of a learned behavioral pattern, through the strength of the relationship and modeling, than a possible genetic link. Blair, Peschardt, Budhani, Mitchell, and Pine (2006) examined the development of psychopathy and proposed that there was a genetic contribution to the component of emotional dysfunction linked to psychopathy. Consistently, the present study could not identify a multigenerational pattern of criminal behavior in the narratives of the women in the sample, although it essentially explored three generations of families through family-of-origin and own family issues.
The present study indicated that some of the women in the sample were active in their churches. The concept of religiosity is complex and contains multiple aspects as previously explored by researchers (Holdcroft, 2006). Unfortunately, information pertaining to the ideological dimension of religiosity (Faulkner & De Jong, 1966) concerning the women in the sample was rather limited. Therefore, it is reasonable to focus the discussion of this finding on the practical church involvement instead. Margie Velma Barfield was described as a “Bible-loving grandmother,” and she taught Sunday school at her church. Amy Archer-Gilligan was described as “deeply religious.” Janie Lou Gibbs reportedly gave some of the insurance funds she obtained through the murder of her husband to her church. Terri Rachals was not only an avid churchgoer, but also sang in the church choir, and so did Bobbie Sue Terrell. At the same time, others may have used religion as a way to gain compassion from the public. Karla Fay Tucker claimed she found God in her prison cell after reading the Bible and became a Christian. The World Council of Churches as well as a number of political figures and reportedly Pope John Paul II gave their support to Tucker and appealed the court.
Faulkner and De Jong (1966) examined the interrelationships among scales for five dimensions of religiosity: ideological, intellectual, ritualistic, experiential, and consequential. Exploring the items involved in each scale, it is reasonable to hypothesize that the church involvement described in the women’s narratives is best reflected in the ritualistic and the consequential scales. The researchers found diversity in the degree of relationships supporting the thesis that religious involvement is characterized by these dimensions, some of which appear to be more closely related. The lowest correlations were related to the consequential dimension.
Additionally, Green and Elliott (2010) examined the effects of religiosity on well-being (i.e., reflective of job satisfaction, financial status, and marital happiness) and health. The study found that individuals who described themselves as religious also tended to report happiness, psychological well-being, and better physical health. Although there are no relevant data in the present study to be compared properly to the findings of the Green and Elliott study, one may argue that the female serialists in the sample did not view themselves as happy considering their reported life stressors, life-changing events such as numerous marriages and divorces, and marital distress.
It is also important to consider the role church plays in communities as a dimension of the social system in which the women functioned.
Examining the evolutionary roots of antisocial behavior in its severe form and its prevalence as it relates to sex differences while utilizing a life-history approach, Eme (2016) found that there was a sizable sex difference (in favor of males) in fast strategies including risky and aggressive behaviors for evolutionary reasons. Environments involving violence, punitive parenting, witnessing death, challenging economic conditions, and family composition changes tended to be conducive of fast life-history strategies focused on immediate gain and disregarding future goals, as the environment did not foster an orientation to the future (Eme, 2016). Many of the narratives of the women in the sample contained themes consistent with the environments described above. They encountered financial hardship, traumatic experiences during childhood, lack of stability in family composition in their families of origin and their own families. Additionally, considering larger system influences, the majority of these women lived during times that were historically harsh to women. While the women in the study were not the only ones subjected to such systemic conditions, one can hypothesize that their living environments were not promoting safety and well-being, but would rather prove fast life-history strategies related to risky behaviors as adaptive. The narratives serve as evidence that criminal behavior helped these women solve various problems temporarily. They solved financial hardship through fraud, arson, and murder; eliminated relational problems through murder; or relieved situational stress through misdemeanor or murder, essentially making violence instrumental.
However, society still seems to struggle to accept women as being capable of murder. In an article for The New Yorker, Patrick Radden Keefe (2013) explored the famous narratives of Waneta Hoyt and Amy Bishop. The story of the Hoyts who had allegedly lost their children to sudden infant death syndrome served as an example in medical literature for over 20 years. Amy Bishop killed her own brother, punched someone at a restaurant, and was accused of mailing a bomb to a supervisor. None of these behaviors seemed alarming enough (even in combination) to the societal system. Furthermore, per the author, between 1976 and 2013, only 12 women have been executed in the United States, compared to the staggering number of men (1308). Considering a smaller system context, namely the nuclear family, the spouses of both Hoyt and Bishop did not appear alarmed by their wives’ behavioral tendencies either. Reportedly, even after Hoyt confessed to smothering five of her children, her husband still asserted that she was innocent and her confession was a result of coercion.
Results suggested that the subjects faced problematic situations of different degrees of demand prior to the first murder. These situations involved financial hardship, domestic violence, marital conflict, relational difficulties, even crying babies. These results are consistent with Schurman-Kauflin’s (2000) conclusions that serial killers tend to live with increased levels of anxiety originating from their poor capacity to handle life stress. Thus, murder appears to be a problem-solving method, being instrumental in nature. Additionally, Baron-Cohen (2012) suggested that psychopaths tend to possess a distinct willingness to satisfy all of their desires, which may be another factor contributing to the repeated decisions to solve problems through murder. While cannot causality cannot be implied when discussing the events in these women’s lives that occurred before the first murder, one can explore sequencing. Arguably, there appear to be patterns in the sequencing of events involving a stressor and a murder. Unfortunately, the limited availability of information prevented further exploration of the sequencing of events involving subsequent murders. However, Osborne and Salfati (2014) attempted to remove the clinical aspect from the cooling-off periods and, instead, examine patterns related to geography, social involvement, and victim selection during the intervals between murders. Although all of the subjects in the study were male, their findings should be reported, considering the scarcity of such data. The researchers found “the offender’s level of social involvement may have a stronger influence on time interval length than individual aspects of the offender’s hunting behavior” (p. 13). In other words, the cooling-off periods would be prolonged if the offender was actively involved in his community (Osborne & Salfati, 2014). Considering the continual exposure female serial killers often have to their victims because of existing relationships, it would be beneficial if future research determines whether social involvement also prolongs the cooling-off periods of female serialists.
In her study of traits and thrill of serial killers, Simon (2015) discussed an intriguing pattern among female serialists (32 were taken into custody) in Hungary dating to the beginning of the 20th century. The women used serial murder to solve problems such as perceived inadequacy of husbands, unwanted children, very old individuals, and mentally ill or handicapped relatives.
Somewhat consistent with the Hungarian women mentioned above, many of the subjects’ narratives in the present study revealed a pattern of low tolerance to stress. Marie Noe’s experience was described as follows, “Killing the babies was easy,” Marie Noe said. Here is how she killed 31-day old Richard Alan: “He couldn’t tell me what was bothering him. He just kept crying. I put him on his belly instead of his back in his bassinet, and there was a pillow under his face. Then, I took my hand and pressed his face down into the pillow until he stopped moving” (Racher, 1999).
In her book, The Female Brain, Dr. Louann Brizendine proposed that female aggression is based on the goal of placing the individual in the center of her relationships (Brizendine, 2006). Her finding provided one possible explanation for the relational component in the victimology of female serial killers. Using Dr. Brizendine’s proposition, it is reasonable to hypothesize that murdering friends and family members is an attempt to solve relational problems while intending to achieve proficiency over the individual’s relational system.
In general, the popular concept that serial killings are committed by irrational or deranged individuals (Hale, 1993) contradicts the idea of murder as a purposeful act in a problem-solving process. Hale (1993) and Simons (2001) contended that the behavior of the killer is logical and meaningful because it seems to be a response to a perceived wrong. Hale (1993) illustrated this idea with the cases of several serialists (all male) who suffered emotional trauma such as maternal deprivation, humiliation, or physical abuse. Yet, he also questioned why these individuals did not go back to the original source of their humiliation (i.e. mother or father) and explained it through learning theory, originally proposed by Dollard and Miller in 1939. According to Hale (1993), these serial killers never reached a successful resolution to their instigated goal response. Instead, the individuals encountered frustration. Arguably, female serial killers do not seem to follow the principle of avoidance of the original source of humiliation. Their behavior appears to be directed at the perceived source of their frustration (i.e., crying children or a husband who somehow does not meet expectations). At the same time, some of the women in the study have reportedly experienced early instances of humiliation, which may have been internalized, such as suffering physical and emotional abuse during childhood, being subjected to unreasonable demands of physical labor, or being taken out of school. According to Hale (1993), in such cases, the individual suppresses aggressiveness, which leads to anxiety or frustration. Later, the same aggressiveness would be directed toward less threatening subjects. Unfortunately, the control remains in possession of the original source of humiliation, as he or she has blocked the basic drive in the now serial killer.
Although, the term affluenza gained popularity with the case of Ethan Couch, intending to reflect his inability to fully understand consequences of behavior because of having led a life of privilege; long before the case of Ethan Couch, who killed four people while drinking and driving, there were the cases of Nora Edwards and Elizabeth Wharton (aka Ellen Wharton). Nora Edwards murdered four people (i.e., two husbands, a stepson, and her own daughter), threatened the life of a physician, and attempted to kill one of her husbands. A physician described her as having been spoiled as a youth and having never progressed from that stage. She was even referred to as the Spoiled Child Serial Killer. Elizabeth Wharton was a socialite, described as “a child of wealth and petted darling of society” (Wheeling Daily Intelligencer, 1871, August 28, p. 3). She came from a wealthy family and married an army officer. Others described her as an exemplary and devoted wife and highly regarded in church. Wharton’s murders became a media event, as one of her victims was a well-known army general and a Wharton family friend.
Underwood (2003) stated that in infanticide, as the age of the victim increases, the level of lethal violence increases as well. He also proposed that mothers are more likely to engage in more passive forms of murder, such as asphyxiation or drowning, as well as to kill younger children. Alternatively, fathers and stepfathers tended to use more active violence, such as beating, or shaking, and to murder older children. The women in the present study tended to utilize poison as their method of murder. Arguably, this method is viewed as rather passive. However, the level of cruelty associated with it is disputably high, particularly because their victims experience prolonged dying and the perpetrators are exposed to the process of dying, as their victims are usually family members or people they know. Keeney and Heide (1994) argued that this slow death could be considered a form of torture.
Up to 70% of parents who engaged in infanticide endorsed experiencing explicit thoughts or fantasies involving aggression toward their children during stressful times and perceived themselves as inadequate related to their own parenting abilities (Levitzky & Cooper, 2000). These findings are consistent with the narratives of some of the women in the present study. Marybeth Tinning killed nine of her children by age five. She allegedly thought she was not a good mother and explained that she did not intend to kill her daughter, just wanted her to stop crying. Waneta Hoyt murdered five of her children because they were crying; she wanted to silence them and did not know what else to do for them. Similarly, Debra Sue Tuggle murdered her fiancé’s daughter because she was crying while Debra was watching TV. She also murdered three of her own children.
The present study attempted to fill the gaps in existing research related to the common characteristics of American female serial killers from a systemic perspective. It is safe to conclude that the larger system (i.e. communities) did not reject the women in the study, just the opposite, the system supported them through marriages, jobs, and community involvement. The women struggled to tolerate stress in adaptive ways and seemed to default to extreme measures, such as murder or other crime, to solve their problems. The violence in which they engaged was instrumental in nature rather than explosive. Most of the women in the study married and had children. A multigenerational pattern of serial homicide in this sample could not be identified.
The high number of sources of information used per subject was a strength of the study; yet, the nature of these information sources (mainly media) must be noted as a weakness. Similarly, considering birth order as a possible common feature among the women in the sample should be considered a strength. However, the lack of sufficient data concerning this variable was a challenge.
Another strength of the study was the long period it covered. Future research may choose to focus on identifying consistent and inconsistent patterns in female serialists over time, thus, uncovering larger system influences on gender role, expectations based on gender, as well as violence expression. Additionally, the present study attempted to consider others’ perceptions of the women in the study. Although such descriptions may be prejudiced because of the time when they were obtained, they still reflect how these women were viewed as well as the way a female murderer was approached by society.
In their study of German serial killers, Harbort and Mokros (2001) hypothesized that their killing experience might be impacting both their social situation and their psychological state. Thus, the authors proposed “future studies on the characteristics of serial murderers could use the time span that elapsed between the first homicide and the apprehension of the offenders as a covariant measure” (p. 329). The present study must note the same issue because of the lack of information representing a reliable, definite timeline of the subjects’ lives. Unfortunately, this appears to be a methodological issue related to the sources of information available to researchers.
Additionally, when discussing their finding concerning the intelligence of German serialists, Harbort and Mokros (2001) mentioned a possible systematic bias based on differences between apprehended (i.e., their sample) and unapprehended serial murderers. While the German sample was exhaustive of the entire population of convicted serialists, the sample of the present study cannot state the same. Thus, the likelihood of a systematic bias based on the availability of information for certain subjects and lack of such for others, as well as failing to account for serialists who have not been apprehended, must be noted as a weakness of the present study.
Although the focus of this study did not involve the forensic realm, it is reasonable to conclude that if a potential serialist is suspected to be female, this individual is likely to be found within the frame of a community, employed, well connected to others, and romantically involved. This individual is likely to have experienced difficulties in childhood and may be presently burdened by multiple stressors.
Future research may benefit from a partnership with law enforcement agencies to obtain reliable information as well as data related to periods between murders, which is scarce in media. Additionally, cross-cultural patterns among female serial killers active in different countries should be explored to determine the effects of gender. Examining gender-based relationships within the population of female serial killers may explain a possible pattern of objectification of others based on their role in the offender’s perception of stressful situations as well as solutions and stress relief.