When mothers kill their children

When mothers kill their children

When Mothers Kill Their Children RALPH A. WEISHEIT Illinois State University Of 460 female homicide offenders admitted to a state prison for women ...

628KB Sizes 0 Downloads 48 Views

When Mothers Kill Their Children


Illinois State University

Of 460 female homicide offenders admitted to a state prison for women in 19401983, 39 were institutionalized for killing their children. Among these offenders, several interesting changes over time were noted. First, women in the earlier time periods were more likely to be married at the time of the offense. Second, the racial mix of offenders changed from predominantly white to predominantly black. Third, offenders in the 1940-1966 period were often motivated by a strong sensitivity to social rules while those in the early 1980s often acted with an insensitivity to socially accepted rules for the treatment of children. Regarding race, education, and marital status, child-killers differed considerably from other types of homicide offenders in the 1940-1966 period but were not distinguishable from others in the 1981-1983 period. That is, child-killers are becoming more similar over time to other types of female homicide offenders.

INTRODUCTION Although homicide is an often-studied form of criminal activity, relatively little attention has been directed to the study of the female homicide offender. The available research does suggest, however, that the female homicide offender differs from other female offenders and from male homicide offenders. Compared to other female offenders, the homicide offender tends to be older, more inclined to alcohol abuse, and less likely to have a prior criminal record, or to come from a family with a prior record (Ward, Jackson, and Ward, 1969). Compared to male offenders, the female homicide offender is more likely to chose an intimate (usually their mate) as a victim, to act alone, to kill in self-defense, and to act as the result of victim provocation (Wilbanks, 1982, 1983). Wilbanks (1983) also noted that men were more than thirteen times more likely than women to commit a homicide. Other research (Totman, 1978) depicts the female The Social Sclence Journal, Volume 23, Number 4, pages 439-448. Copyright 9 1986 by JAI Press, Inc. All rights of reproductlon in any form reserved. ISSN: 0035-7634.



homicide offender as unemployed (or marginally employed), socially isolated, and the subject of physical a n d / o r emotional abuse at the hands of intimates. Totman characterizes the act of homicide by many females as a symbolic action in which the offender actively attempts to seize control of her own life situation. This seems particularly true when the victim is the offender's mate. The killing of a mate was observed to actually bring the offender closer to her surviving family members and to instill in her a new sense of self-control, of power over her own destiny (Totman, 1978). While the killing of one's mate as a problem-solving technique cannot be condoned, it is often understandble, particularly when the offender has been a long-term sufferer of abuse at the hands of her mate/victim. While mates are the most frequent victims of the female homicide offender, another frequent victim is the offender's child. Studies reporting the percent of the victims who are children ranges from 8.5 (Wilbanks, 1983) to 28 (Totman, 1978), with two studies reporting that child-killers made up about 20 percent of their sample (Cole et al., 1968; Ward et al., 1969). Unlike the killing of a sadistic or brutal male companion, the killing of a child is particularly difficult to comprehend of justify. Totman (1978) notes that even among the female killers in her study, the child-killers held a particularly low status, both in the eyes of other female killers and in their own eyes. Unlike the woman who kills her mate, the child-killer is not the recipient of abuse from the victim but is usually involved in a long-term pattern of abuse aimed at the child. Other researchers have suggested the killing of a child is not an overtly instrumental act analogous to the killing of a mate. That is, there is not usually the express purpose of killing or even seriously harming the victim. These authors suggest the child is either a convenient target for frustration (Totman, 1978), or the mothers use excessively harsh punishment with little awareness of the physical damage inflicted on the child (Ward et al., 1969). The fact that female homicide in general is linked to the role of women in the larger society has important implications for female homicide involving the killing of young children. To the extent that women's roles have changed substantially over time, we might expect the nature of her involvement in the killing of children also would change. The degree to which women's roles have actually changed, however, is still a matter of debate. While some have characterized changes in the social roles of women as "tremendous" (Thorton and Freedman, 1979:841), others have suggested that what changes have occurred have largely bypassed low-income and minority women (Chafe, 1977), the very group most likely to become involved in homicide. Further, greater participation in the social world by women With children may actually increase tension and frustration if they take on new responsibilities (such as a full-time job) while still carrying the full burden of the traditional roles of homemaker and mother. The study of women who kill their children is particularly important for several reasons. First, because it is an act that may be closely tied to the role of women in society, information about these kinds of murders may be particularly informative about female criminality in general. Second, Totman (1978) has suggested that the killing of a child may be one of the few kinds of homicide for which prevention efforts might show some success. The death of a child is often the end of a long-term pattern of abuse in which there are often many opportunities for community and family members to recognize the problem, and the mother herself may often recognize the need for outside help. Third, adjustment to incarceration may be particularly difficult for these women. They are less


When Mothers Kill Their Children

likely to find other female inmates supportive, and even staff may have difficulty relating to such individuals. The lack of basic information about women who kill their children places the prison administrator at a decided disadvantage in dealing with their adjustment problems. The purpose of the present study is to contribute to basic knowledge about women who kill their children and to determine the extent to which there have been changes in the characteristics of these offenders over time.

THE DATA The current study is based on data from the prison files of 39 women incarcerated for the killing of their children during the years 1940-1966 and 1981-1983 in the state of lllinois. ~ These data are part of a larger study of 460 female homicide offenders. Thus, the child-killers represent 8.5 percent of the total sample, a figure that corresponds to that in the Wilbanks' (1983) study based on data from police and coroners. Although other data sources might have been used, both practical and methodological considerations led to the selection of institutional records as sources of data. First, the technique is analogous to that used in the cross-sectional research conducted by Ward et al. (1969) and Totman (1978), thus allowing the present data to be compared with theirs. Second, compared with other crimes, dispositions for the crime of homicide should be least influenced by such institutional factors as prison overcrowding. Further, extra-legal factors (such as social class of the offender) should least influence the disposition of the most serious offenses, such as homicide. Third, while institutional records are notorious for omitting important information about offenders, this should be least problematic for the most serious offenses. Further, the relatively long institutional stay of these offenders suggests that institutional personnel would be interested in having as much information about these offenders and the circumstances surrounding the event as possible. Fourth, the institution that houses these offenders is one of the only facilities in the state for women offenders. This largely precludes transferring female inmates to other institutions for administrative purposes. Thus, major shifts in the characteristics of this sample cannot be the result of changing administrative decisions regarding the manner in which serious offenders are distributed among major correctional facilities, as might be true for male offenders. Finally, the offense of homicide is one in which institutionalization is highly likely. Ward et al. (1969) suggest that approximately 96 percent of all women convicted for homicide are imprisoned. Similarly, Wilbanks (1983) found that only 2 of the 17 women convicted of homicide were given probation; the remainder were incarcerated for some period of time. While there are several advantages to studying an institutionalized population, there are also several disadvantages, particularly when the subjects are those whose victims were their own children. More than most homicide offender groups, women who kill their children can be expected to be viewed as mentally unfit for trial or for ordinary punishment. Totman (1978) notes that most of the child-killers themselves feltthey should have been placed



in a mental health facility rather than a prison. Thus, an unusually high proportion of these offenders might be filtered out of the criminal justice system (and sent to mental health agencies) before they reach prison. The extent to which this is a problem is not known but is a worthy subject for future research. Given that even the offenders themselves find the killing of a child an unconscionable act, it may be difficult to obtain detailed and accurate descriptions of the circumstances surrounding the event from these offenders. Several authors have noted, for example, the frequency with which such offenders seem to have no recollection of the act. While the above points may produce some distortions in the data, the extent to which these conditions exist is unknown and it is unclear to what extent they bias the results. The data include relatively detailed record information on 39 female homicide offenders charged with the killing of their children. In addition to basic demographic information about the offender, data were collected concerning the circumstances surrounding the event and a descriptive narrative was generated from each case. These narratives are generally taken from the descriptions of the offender and the event as provided by the State Attorney's office, with only minor editing to protect the identity of the offender and others explicitly named in the narrative. In the sections that follow, the child-killers will be studied as a distinct group to identify changes over time in the characteristics of the offender and in circumstances surrounding the event. Next, the child-killer will be compared with other female homicide offenders.

CHANGES OVER TIME There were too few cases to examine yearly changes in characteristics of these offenders and it was thus necessary to collapse data across years. While the process of collapsing categories may distort the findings, the small number of cases left few alternatives and previous work with the larger data set suggested that were changes over time could be observed, they were between the 1940-1966 group and the 1981-1983 group. These two groups contained 25 and 14 offenders, respectively. 2 An examination of background characteristics of the offenders suggested that there were generally few differences between past and present offenders who killed their children. Women from the two time periods were not significantly different in regard to age, number of children, number of children in their custody at the time of the offense, years of education, employment status, or prior offenses (Table !). The average childkiller was 26 years old, had been married one time, had three children but only had custody of two at the time of the offense, had 10 years of education, was unemployed (76 percent), and had one prior offense. 3 There were also no clear patterns of change in the extent to which child-killing was an urban phenomenon. Approximately 70 percent of all offenders, including child-killers, were from the largest urban county in Illinois, which includes the Chicago area. None of the other 101 lllinois counties was represented by more than 3 percent of female homicide offenders. Similarly, the social class of offenders showed little variation over time, with nearly all of the women coming from the lowest income categories. Noticeably absent was any apparent influence of World War I1 on the nature or extent of infanticide. Although there were six cases admitted during these war years, only two were even indirectly linked to the war. In the first case, a married women with five young

When Mothers Kill Their Children


Table 1. Comparisonof Child-Killers with Other Female Homicide Offenders. Child-Killers Characteristic* Mean age Number of children Number of children in custody Years of education Percent black Percent married Percent employed full-time Number of casesb

Other Homicide

1940-1966 1980-.1983



26.0 2.7 2.0 9.7 28.0 52.0 25.0

24.9 2.9 2.6 10.0 71.4 28.6 21.4

33.4 1.5 .7 7.8 78.7 32.6 31.0

27.9 1.5 .7 10.6 70.9 29.1 20.8





Notes: "At the time of the offense. bFuIlcount not including missingcases for any one variable.

children was charged with murder after her six-month-old child died of neglect and starvation. The child was covered with large ulcers and the filthy diaper was filled with maggots. The mother spent much of her time in bars in the company of other men, leaving her children unattended and without food. A second war-related case is described more fully below. In both cases it would be more accurate to describe the mother as free from social control than to describe her as "liberated." The apparent absence of influence by the war is consistent with the findings of Steffensmeier, Rosenthal, and Shehan (1980), who found that observable increases during this time were for sex-related offenses and were not related to more egalitarian sex roles. There were two areas in which child-killers differed over time. First, those in the 1940-1966 time period were more likely to be married (13 of 25, or 52 percent) than were child-killers in the 1981-1983 period (4 of 14, or 29 percent). Second, there was a dramatic change in the racial composition of the offender population. While nearly three-fourths of the child-killers were white in the 1940-1966 period--72 percent (18 of 25)--in the 1980-1983 period only 29 percent (4 of 14) were white. This pattern persists regardless of the way in which year is categorized and is particularly interesting in light of the absence of change in other factors, such as employment status. Indeed, it is difficult to explain the apparent shift from a predominantly white to a predominantly black offender population. It might be argued, for example, that the increased percentage of blacks over time is due to changes in the formal processing of offenders by the criminal justice system. However, this explanation is difficult to reconcile with the fact that such changes were observed only for child-killers. Other types of female homicide offenders showed no comparable changes over time, and in fact showed a slight reduction in the percentage of blacks (see Table 1). As will be seen below, this shift coincides with some rather interesting changes in the circumstances of the killing. Having briefly considered characteristics of the offenders themselves, our attention next turned to characteristics of the offense. In this regard, too, there were surprisingly few changes over time. Consistent with the findings of Totman (1978), these offenders primarily committed their acts with the direct use of their hands--through beating, strangulation, or suffocation of the child. Only 5 of the 39 cases involved the use of a gun



or knife and only I case involved the use of poison. 4 Regardless of the time period examined, most of the acts (35 of 39) took place in the home with the mother as sole offender. One circumstance that did change over time was the apparent motivation for the offense. Unfortunately, for most offenders there was no direct way to measure motivation for the offense--many women themselves later described their act as senseless. For a number of cases, however, the circumstances surrounding the event suggested the importance of social regulation, and may reflect changes in this over time. Many of the child-killers could be characterized as either overregulated or underregulated by their social environment. The overregulated child-killer commits her act precisely because she is so sensitive (perhaps overly sensitive) to her social environment. The clearest cases of the overregulated child-killers are those who killed their children at birth, many of whom were attempting to avoid the shame/stigma of having an illegitimate child. Killing at birth accounted for 5 of the 8 child killings between 1940 and 1946 but for none of the 22 cases from 1960 forward, even though legal abortions were not available until 1972. The following scenario, taken from the files, illustrates the overregulatcd offender: Case 264 (1945): The defendant was married to a man who was serving overseas. She lived at home with her mother, father, younger sister, and a twenty-month-old baby by her husband. The defendant and her mother frequented taverns where they both kept company with men she met in the taverns. On the evening in question, the defendant and her mother were at a tavern until approximately 1:30 in the morning when the defendant went home alone. She became ill during the night and the next morning she gave birth to a child while in her father's bedroom. After the child was born, the defendant picked up the child from the floor, severed its navel cord with her fingers while holding the baby in her arms. Several times she went to the door with the intention of calling her mother but instead of calling her mother she finally choked the child by tying the cord from her maternity dress tightly about the baby's neck. After this she tied another part of the dress tightly around its neck and shoved the baby's head in the sleeve of a dress and threw it under the bed. It appears that the baby was fathered by someone other than her husband and the defendant was extremely ashamed. As the above case shows, these offenders were not necessarily models of proper behavior in other aspects of their life. The child, however, apparently represented a tangible sign of their misconduct that could not be denied or simply explained away. Further, for these offenders there was often a clear attempt to hide the offense from others, in stark contrast to the underregulated offender to whom our attention now turns. The tmderregulated offender acts with an insensitivity to the boundary separating proper from improper behavior in regard to the treatment of their children. Some killed while administering excessively harsh punishment, while others killed as a result of neglect. In many of these cases the offender denies any intent to seriously harm the child. The following scenario illustrates the underregulated offender: Case ! 16 (1981): The subject's 5-year-old daughter was supposedly beaten with an extension c o r d - - o v e r 40 bruises were on the body. The child died of a heart attack

When Mothers Kill Their Children


as a result of this serious beating. Subject states she did "whip" her daughter with a belt at approximately 3:00 p.m. but the child was all right. She ate at approximately 5:00 p.m. Shortly thereafter, subject states she placed the child in the bathtub, about five minutes later she returned to the b a t h r o o m and found the child unconscious in the tub. Subject states she has only been "disciplining" her daughter due to her "bothering" a hot stove. In contrast to the overregulated child-killer, the underregulated will often call the authorities herself, or seek medical attention for the child. Some claim no recollection of the act but there is a general denial of intent to kill the child and in some cases even indignation that authorities would interfere with or question her child-rearing practices. Although both types of offender can be found in any given time period, the overregulated was most frequently observed in the 1940s and 1950s, while the underrcgulated were most often observed in the 1960s and 1980s. These two general types of child-killers are suggested by the data but further research is clearly needed in which a larger number of cases and more detailed information about the events are available. It is unclear, for example, the extent to which changes represent actual changes in the offenders themselves, or changes in such things as abortion practices. The shift from predominantly overregulated to predominantly underregulated offenders is indirectly supported by studies of child-killers conducted at various points in time. Morton's 1934 study of 126 female homicide offenders describes the child-killers as follows: it is interesting to note that each of the cases give as the reason for their action the desire to save their victim from what was considered a worse fate, and this is the underlying principle in practically all infant murders that have come under my observation (p. 74) This "appeal to higher loyalties" represents yet another variation o f the overregulated offender. These offenders act (or overreact) out of an acute sensitivity to social regulation. Although killing the child is seen as wrong, the behavior is justifiable because the act was motivated by a higher rule, such as preventing the child from living under the stigma of being illegitimate or of being a failure. In their view what they did was right. By contrast, contemporary descriptions of child-killers are more consistent with the image of the underregulated offender. Excerpts from two studies illustrate this well: Cases that involve the death of minor children through repeated physical abuse do not fall into the "premeditated"category because it is not clear that the women intended to kill the child: rather their expressed intent was to influence physical harm as "punishment.'(Ward et al., 1969) The children seemed rather to be nonentities or irritants. They seemed to be symbols of frustration, boredom, confinement and failure. Subjects reacted to and against their children as evidence ofall that was wrong with their world. (Totman, 1978:79) Most offenders who could be classified as overregulated are those who were charged with the killing of newborns, although there were a few exceptions. For example, one woman shot and killed her 13-year-old daughter and her 18-year-old son as they were



sleeping, believing that it was better to kill them than to allow them to grow up to be failures, as she believed they were destined to do. It is also interesting to note the link between race and the killing of newborns. Of the 8 women charged with killing newborns only 2 were black. The reasons for this are unclear, perhaps a category of offender was eliminated by the legalization of abortion. This, too, is an observation meriting further research. It is possible that the observed changes result more from changes in social awareness than from actual changes in behavior over time. Phol (1977), for example, argues that child abuse was not "discovered" until the 1960s, at which time a wave of special legislation and attendant publicity was observed. It should be noted, however, that the special legislation and publicity focused primarily on the beating of children, which at one time was justified as being in the child's own best interests. There is no reason to believe that the killing of a child was any more acceptable in the 1940s and 1950s than it is today. In fact, the absence of females imprisoned for the killing of newborns from 1960 forward is surprising in light of the lengthy discussion by Piers (1978) regarding the continued practice of killing newborns in this country. While we can only speculate why our observations differ from those of Piers, it is possible that the killing of newborns had become increasingly defined as a medical problem (e.g., the result of postpartum depression), and therefore best handled outside of the criminal justice system. Further research is needed to examine changes in the process by which child-killers are selected for handling by the legal system. The current sample allows us to only examine the outcome of the process rather than the process itself.

IS THE CHiLD-KILLERUNIQUE? Having established that in regard to race and motivation, female child-killers may have changed over time, our attention now turns to a comparison of child-killers with other types of female homicide offenders. The child-killers tended to be younger (26 years versus 33 years), more likely to have children (3 children versus 2 children), and to have children in their custody at the time of the offense. There were other interesting changes distinguishing child-killers from other female homicide offenders (see Table l). For example, child-killers in the 1940-1966 period had m o r e education than other types of female homicide offenders, but in the 1981-1983 period they had about the same amount of education. These differences are almost entirely due to changes in the educational level of other offenders and to the absence of change in child-killers. The child-killers also differed from others in regard to race. Only 28 percent of the child-killers were black in the 1940-1966 period, compared to 79 percent black among others. However, in the 1981-1983 period, blacks were equally represented among child-killers and among others. A final respect in which there were changes over time between child-killers and others concerns their marital status. In the 1940-1966 period, child-killers were more likely than others to have been married at the time of the offense but in the 1981-1983 period about the same proportion of childkillers and other types were married. These data suggest that where changes have taken place, there is a tendency for child-killers to more closely resemble other types of female homicide offenders.

When Mothers Kill Their Children


DISCUSSION The data suggest that the types of women who kill their children have changed. Childkillers in the 198 i-1983 period were more likely to have been black and less likely to have been married than were child-killers in the 1940-1966 time period. More important, child-killers in the early 1980s were more characteristically underregulated by social rules while those in the 1940-1966 period were more characteristically overregulated. That is, compared to her predecessor, the child-killer of the early 1980s was insensitive to the social rules regarding the discipline and treatment of children. There were several areas, however, in which there were few changes from 1940 to the present. Child-killers changed little in regard to age, number of children, number of children in their custody, education, employment status or prior offenses. Child-killers tended to be younger than other types of female homicide offenders and, at the same time, they had more children than others. C o m p a r i n g child-killers with others over time suggests that in three areas--level of education, race, and marital s t a t u s - - t h e child-killer of the early 1980s was more similar to other female homicide offenders than was the child-killer of the past. Overall, the child-killer is more difficult to distinguish as a unique offender type than was true several decades ago. The findings are suggestive but clearly highlight the need for further research. The number of child-killers in this sample is too small to allow for an elaborate analysis of the problem. In particular, it seems likely that race may interact with other variables, such as education and marital status. More important, further research is needed to examine motivational factors in child-killing, especially as they relate to women's roles and changing expectations for others. It is also clear that further research is needed utilizing data other than that provided in institutional records. It is important to determine the extent to which observed changes represent changes in the offender population as opposed to changes in the official processing of offenders. It is also important to determine the extent to which changes over time reflect changes in the extent to which mental health professionals have been utilized in lieu of formal processing through the criminal justice system.

ACKNOWLEDGMENTS I wish to express my sincere appreciation to the members of the Illinois Department of Corrections who were most cooperative in providing access to the necessary documents. I would als o like to acknowledge the helpful comments of an anonymous reviewer.

NOTES I. Complete data prior to 1940were not available, and data for 1967through 1980could not be collected with a certainty that all cases were included because of several technical limitations on the data, including:(1) an experiment with mixed-sex facilities, which involved sending female offenders (several of whom were charged with homicide) to male facilities; (2) a change in the system of assigning numbers to inmates that replaced the sequential numbering system used earlier; and (3) an end to the practice of keeping detailed admission logs that indicated the date of admission and the primary offensecharged. The practice of sequentially numbering new admissions made it relativelysimple to be certain that every file was accounted for. Sequential numbering, combined with an intake log, simplified the process of determining which files were to be examined for the 1940--1966period. When these two practices were ended, it became



difficult to determine whether every file for a given year had been examined. The experiment in mixed-sex prisons created problems in that the files of female offenders sent to male institutions were stored with those of males. Similarly, such females were usually admitted directly to male institutions in which the practice of keeping detailed admission logs was also being phased out. These problems, combined with changes in the numbering system, made it difficult to know whether all female homicide offenders has been accounted for in the 1967 through 1980 period. Thus, rather than examining the records of what may have been a highly uncharacteristic sample of female homicide offenders, it was decided to exclude cases in the 1967-1980 time period entirely. 2. It is well known that the manner in which categories are collapsed may substantially influence the results. Due to the small number of cases there were limitations on the number of categories that might be used. As a check on our conclusions based on the treatment of year as a dichotomous variable, the data were also split into three groups: 1940-1951, 1952-1961, and 1981-1983 and the analysis given below were run with the same results as when the time periods were dichotomized. Exceptions to this general pattern are noted in the text. Further, when changes over time were examined for interval-level variables, year was correlated with the appropriate variable. Here, too, the results were generally the same as when time periods were dichotomized. For the sake of parsimony, then, the data presented here are grouped into only two time periods, 1940-1966 and 1981-1983. 3. Despite its obvious importance, little will be said about prior offenses. This indicator is probably very unreliable over time. The most current records include rap sheets from the FBI and from the state of Illinois. As one moves back further in time, however, the prior record is less based on these relatively comprehensive data files and more on the word of the offender upon admission. An examination of the prison records suggests that the scope and the consistency of data provided to the Uniform Crime Reports were rather limited during the forties, fifties, and much of the sixties in Illinois. Thus, comparisons over time are problematic. 4. Even this case is not one in which the sinister female slips the unsuspecting victim poison, well hidden in food or drink, in the hopes of'getting away with murder." Quoting from the official record: Case 62: From the time [the victim] was 3 weeks old until she died on [date], the offender fed her 2 bottles one-half full of beer per day. As a result of her feeding the baby alcohol every day, [the child] died of acute alcohol intoxication in association with fatty liver and asperation of the gastric contents.

REFERENCES Chafe, W.H. Women and Equalio': Changing Patterns in American Culture (New York: Oxford University Press, 1977). Cole, K.E., Fisher, G., and Cole, SS. "Women Who Kill," Archives of General Psychiatry, 19(1968): !-8. Morion, J. "Female llomicides," Journal of Mental Science, 80(1934):65-74. Phol, S.J. "The 'Discovery' of Child Abuse," Social Problems, 24(1977):310-323. Piers, M.W. Infanticide (New York: Norton Press, 1978). Steffensmeier, D.J., Rosenthal, A.S., and Shehan, C. "World War II and Its Effects on the Sex Differential in Arrests: An Empirical Test of the Sex-Role 'Equality and Crime Proposition," The Sociological Quarterly, 21(1978): 403--416. Thornton, A., and Freedman, D.S. "Changes in Sex-Role Attitudes of Women, 1962-1977: Evidence from a Panel Study," American Sociological Review, 44(1979):832-842. Totman, J. The Murderess: A Psychosocial Stud)' ofCrhninal Ilomicide (San Francisco, R. and E. Research Associates, 1978). Ward, D.A., Jackson, M., and Ward, R.E. "Crimes of Violence by Women," in Crimes of Violence, edited by D. Mulvihill (Washington, DC: U.S. Government Printing, 1969), pp. 843--909. Wilbanks, W. "Murdered Women and Women Who Murder:. A Critique of the Literature," in Judge, La~9"er, Victim. Thief: Women, Gender Roles. and Criminal Justice, edited by N.|I. Rafter and E.A. Stanko (Boston: Northeastern University Press, 1982), pp. 151-180. "The Female Homicide Offender in Dade County, Florida," Criminal Justice Review, 8(1983):9-14.