|Home | About | Journals | Submit | Contact Us | Français|
This study assesses the entrance of substance-abusing female offenders (N=1,209) into the criminal justice system through temporal patterns (using age of first victimization, drug use and arrest). Nine pathways were identified. Unexpectedly, the leading path was a sequence where drug use preceded arrest in absence of childhood victimization. However, women under a path inclusive of victimization possessed more risk factors. Findings support feminist pathway research, which states that childhood victimization is generally present in female offenders’ lives. Nevertheless, results also revealed that a drug pathway without childhood abuse proved to be as important and even more dominant among criminal justice-involved women.
There is a growing amount of literature that provides support for feminist pathway research (Belknap, 2007; Belknap & Holsinger, 2006; Bloom et al., 2003a; Burguess-Proctor, 2012; Carbone-Lopez & Miller, 2012; Chesney-Lind, 1997; Covington, 1998; Daly, 1992; Huebner, DeJong & Cobbina, 2010; Kaufman & Widom, 1999; Richie, 1996; Makarios, 2007; Salisbury & Van Voorhis, 2009; Wright, Salisbury& Van Voorhis, 2007). A recent and increasingly popular perspective applied to women is the feminist pathway perspective, which examines the causes of women’s illegal behavior by emphasizing childhood abuse as a significant risk for drug use and subsequent offending (Belknap, 2007; Belknap & Holsinger, 2006; Cernkovich, Lanctôt, & Giordano, 2008; Huebner et al., 2010; Siegel & Williams, 2003; Kaufman & Widom, 1999; Makarios, 2007; Salisbury & Van Voorhis, 2009; Wright et al., 2007). The purpose of this study is to test the hypothesis, in a sample of substance-abusing female offenders, childhood victimization played a significant precursory role in a drug use to arrest pathway.
A number of studies grounded in feminist criminology identified, in addition to early abuse, key life events that propelled women into crime. These included family dysfunction, mental health issues, domestic violence, and economic marginalization (Bloom et al., 2003a; Chesney-Lind, 1997; Covington, 1998; Daly, 1992; Richie, 1996; Salisbury & Van Voorhis, 2009). Given research findings on the predictive validity of these risk factors associated with women’s drug use and crime, what is notably absent from the literature is an examination of the timing traumatic events (or risk factors) occurred in a woman’s life (Belknap & Holsinger, 2006). Age, for example, was found to be a significant risk factor for women who had a higher risk of both alcohol and drug abuse (Danielson et al., 2009). In other words, women with an earlier age of first drug use had an increased risk of drug abuse later in life (Ibid, 2009). Also, the timing of violence exposure was determined to be a potentially relevant factor in the investigation of substance abuse among female offenders. Thus, this research attempts to address the timing of exposure to victimization, drug use and arrest by chronologically ordering the events by age of first occurrence. As opposed to clustering women by whether or not the event (i.e., victimization, drug use, arrest) happened in their lives, the age at which the event occurred allows for a descriptive view on 1) whether victimization was present at all, 2) did victimization precede the onset of deviant behaviors, and 3) whether additional pathways that may disrupt the order of the feminist pathway perspective existed. However, this research does not test whether age has detrimental effects on deviant outcomes.
The relationship between age and crime is evident through developmental theories of crime (Moffit & Caspi, 1999; Sampson & Laub, 1993; Tonry & Farrington, 1991). They assume that age is a significant factor for understanding criminal behavior. By examining crime through a life-course perspective, we assume that the same event can have different consequences depending on when in the life course the event occurred (Strauss & Ramirez, 2004). For example, several studies have established the early onset of crime leads to a greater risk of prolonged criminal behavior (Glueck & Glueck, 1950; Sampson & Laub, 1993; Wolgang, 1983). The significance of age has transcended the link between early onset of crime and later offending. Research on the timing of physical victimization suggest that children abused at earlier ages suffered from more detrimental psychological functioning effects (Carslon, Furby, Armstrong & Schlaes, 1997; Keiley et al., 2001). While children abused at later ages were found to exhibit more external misbehavior than non-abused children (Keiley et al., 2001). Although, the concept of age was important in formulating the age-based variables for this research, it is important to note that age was only used to create the pathway typologies for the women in this sample. Age was fundamental in plotting a sequential view of factors related to the onset of crime for women.
A vast majority of qualitative studies used a pathway perspective to construct additional typological categories/pathways to highlight important features or events that lead women to offend (Daly, 1992; James & Meyerding, 1977; Richie, 1996; Owen, 1998). One pathway tracked poverty, lack of education and other forms of marginalization in justice-involved women (Richie, 1996; Owen, 1998). Another proposed pathway was that of “street women” who likely ran away from abusive homes and became involved in the criminal justice system due to theft, drug abuse, drug dealing or prostitution; that is, survival strategies were, in and of themselves, criminal (Daly, 1992; Owen, 1998). Lastly, “battered women” pathway connected abusive intimate partner relationships with a multiplicity of abuse, drug abuse and criminality (Daly, 1992; Owen, 1998; Richie, 1996). The main premise was that women in the criminal justice system follow a path where childhood victimization is prevalent, and followed by drug use eventually leading to contact(s) with the criminal justice system.
Recently, several studies have attempted to fill a gap in the women’s pathways to crime literature by utilizing larger samples and quantitative methodologies. Larger samples have allowed researchers to quantify several pathways (Salisburg & Van Voorhis, 2009; Brennan et al., 2010; Brennan et al., 2012, Reisig et al., 2006). Researchers have classified paths into gendered pathways and gendered neutral pathways (risk factors associated with offending that are not specific to women) (Brennan et al., 2012; Van Voorhis, 2009). This quantitative approach to pathway research has also allowed for greater complexity to women’s paths to the criminal justice system (Brennan et al., 2012).
Quantitative and qualitative research on women’s pathways to crime revealed differing conceptualization of risk factors related to women’s drug use and offending. Additionally, Reisig and colleagues (2006) used Daly’s (1992) typological pathways to assess women’s recidivism patterns based on their classification as offenders (high, medium and low). Similar to the current study, Reisig et al. (2006) identified meaningful subgroups within their sample of female offenders that reflected grouped risk factors associated with female criminality. More recently Carbone-Lopez and Miller (2012) examined the cumulative impact of early transitions into adult roles and responsibilities on the onset of methamphetamine use. Results showed family, caretaking, motherhood, independent living, peer and romantic relationships facilitated their initiation to methamphetamine use. Carbone-Lopez and Miller (2012) state these findings highlight the importance of attending to the timing and sequencing of women’s experiences.
This paper adds to the growing literature on gendered pathways (Huebner et al., 2010; Siegel & Williams, 2003; Kaufman & Widom, 1999; Makarios, 2007; Salisbury & Van Voorhis, 2009; Wright, Salisbury & Van Voorhis, 2007) by assessing women’s paths to crime using the age of first childhood victimization. The prevailing theme across most studies using a feminist pathway perspective is that most substance-abusing women offenders were sexually or physically victimized in their early years prior to their involvement with drugs and/or crime. This paper explores whether the sequence of the feminist pathway perspective (i.e., childhood victimization→ drug use→ arrest) emerges as the dominant path in a group of approximately 1,200 substance-abusing female offenders.
Drug use is one of the factors associated with women’s path into the criminal justice system. Female offenders have a high prevalence of drug abuse (Mumola & Karberg, 2006), which has been found to be strongly correlated with repeated involvement in the criminal justice system (Bloom et al., 2003a; Brennan, Breitenbach & Dieterich, 2010; Evans et al., 2002; Farabee, Joshi& Anglin, 2001; Sommers et al., 2000; Dunlap, Johnson & Maher, 1997; Mallik-Kane & Visher, 2008; McClellan et al., 1997). National statistics reported that 60 percent of women in State prisons and 43 percent of the women in Federal prisons met the criteria for drug dependence or abuse (Mumola & Karberg, 2006). Findings from a study about substance abuse and crime showed female inmates were more likely than male inmates to experiment with drugs before any criminal justice involvement (McClellan et al., 1997).
Although drug-related offending and drug use are also significant in men’s involvement with the justice system, men’s most commonly traveled and direct path to the criminal justice system is through violent criminal acts (McClellan et al., 1997). An explanation is that women’s drug use may be tied directly to their experiences of interpersonal violence (Anderson, 2008; Bloom et al., 2003a). Some suggest that women are more likely to use drugs to cope with internalized feelings of shame (due to prior abuse), whereas men may externalize the same negative event with feelings of anger (Gilligan, 1993; McClellan et al., 1997).
Consequently, prior research has placed men at the center of women’s drug use (Bloom et al., 2004; Brown, 2002; Daly, 1992; Richie, 1996). Female offenders have reported that drug abuse was a consequence of a previous “battering incident” where drugs were used as a way to “reconnect” with men (Richie, 1996). Additionally, imprisoned women have also emphasized their entrance into drug trafficking and the criminal justice system was co-opted by their intimate partners. However, research on women drug sellers and traffickers disputes this claim and finds that men are not always at the center of women’s entrance into drug use and the drug market economy (Maher, 1997; Anderson, 2008). In fact, women have displayed a level of autonomy or agency in determining their own entrance into the drug market economy as sellers and resisted the label of victims (Maher, 1997; Sommers, Baskin & Fagan, 2000).
Drug use rarely occurs within a vacuum; that is, it is often a mediating factor to crime which extends from overlapping issues within women’s lives such as victimization (Widom & Ames, 1994; Widom, Ireland, & Glynn, 1994; Wilson & Widom, 2009). Substantial research shows victimization to be a precursor to involvement in drug use and crime across gender (DeHart, 2004; Maxfield & Widom, 1996; Widom, 1989b, Widom & Ames, 1994; Wilson & Widom, 2009). The most pervasive claim of the child victimization literature is that abused children have a greater likelihood of offending in adolescence and in adulthood (Mass et al., 2008; Widom & Maxfield, 2001; Smith & Thornberry, 1995). The “cycle of violence” research generated results showing abused and neglected children have a higher likelihood of early arrests for delinquency, adult criminality, and violent criminal behavior, as compared to non-abused children (Smith, Ireland & Thornberry, 2005; Widom & Maxfield, 2001). Analyses by gender also showed abused and neglected females were more likely to suffer from depression, and possibly hospitalization, as a consequence of early adverse childhood experiences rather than exhibiting reactions through direct outward aggression as displayed by males (Widom & Maxfield, 2001). The prevalent argument, however, is that child abuse or neglect do not directly cause crime or delinquency. Instead, women’s offending was related to victimization, (regardless of the type, e.g., sexual, physical, neglect), through depression, anxiety and substance abuse (Ireland & Widom, 1994; Widom, Ireland & Glynn, 1995; Salisbury & Van Voorhis, 2009).
While exposure to traumatic events, such as sexual or physical abuse and neglect, have been linked to substance abuse and criminal behavior (Danielson, Amstadter, Dangelmaier, Resnick, Saunders & Kilpatrick, 2009; Widom, 1995; Widom, Marmorstein & White, 2006; Wilson & Widom, 2009), previous research has not chronologically ordered the events associated with the main variables: victimization, drug use, and arrest (Belknap & Holsinger, 2006; Danielson, et al., 2009). The analysis of the temporal ordering of these events will translate to the examination of whether a feminist pathway beginning with childhood victimization is more prevalent among a group of substance-abusing female offenders (Belknap, 2007; Belknap & Holsinger, 2006).
The current study is designed to first explore the variation and distribution of the sequential patterns of women’s pathways into the criminal justice system. This research seeks to assess the distribution of common pathways leading to first arrest in a large sample of criminal justice-involved women (using age of childhood victimization and first drug use).
The study also investigates differences in the characteristics of women within the most populated pathways. Research suggests that many women who were sexually, physically abused or neglected are also likely to experience other types of adverse childhood experiences, putting them at greater risk for behavioral, physiological and psychological disorders (Anda et al, 2006; Dong et al., 2004; Felitti et al., 1998). Therefore, characteristics and risk factors will be compared across the dominant pathways to determine whether women following one path are more likely to experience other adverse experiences (e.g., familial drug use, parental incarceration).
The data for this study was derived from Project WORTH (Women’s Options for Recovery, Treatment and Health), a large-scale evaluation project of women who entered drug treatment through the criminal justice system in New York City, NY and Portland, OR. The data was collected by interviewing 1,209 women (641 in New York City and 568 in Portland) who were mandated to participate in drug treatment programs (outpatient, inpatient and jail). The information collected provided the best sample available for examining the questions at hand; however, it should be recognized that because the analyses are based on secondary data, not all items and information relevant to the questions at hand were available. Baseline interviews were administered within a two-week period from the admission of the women to treatment between 1995 and 1999. The interviews collected in-depth information about women’s drug use and interactions with the criminal justice system. The original study was not designed to identify the common pathways of women, instead, the original project was an evaluation of outpatient, residential and in-prison drug treatment programs. The main outcomes of interest were reduction in drug use and criminal activity. Moreover, the interviewers were highly sensitive to the women’s issues, skillful in building rapport and instilling confidence. The women were also assured their privacy would be protected with a Federal Certificate of Confidentiality.
The interviewers in the original study used timelines1 to help the women remember when something happened by plotting key life events. This method emphasized events that represented a phase of a person’s life (e.g. marriage, work, childbirth, birthdays, holidays) to organize their memory, which would then serve as cues for drug use, arrests and victimization (Caspi et al., 1996; Freedman et al., 1988; Roberts & Horney, 2010). Generally, Life Event Calendars (LEC), which use a timeline format, have been found to be effective in increasing the recall of interviewees (Freedman et al., 1988; Morris & Slowcum, 2010). This improvement in recall is due to the respondents’ ability to place different activities within the same time frame (Freedman et al., 1988).
The study used a purposive sample of women in the criminal justice system who had substance abuse problems. The sample was made up of women charged and convicted primarily for non-violent drug offenses. The majority were polydrug users and began using drugs during their teenage years (M= 14, SD= 5.30). Prior to the current conviction the women self- reported an average of nine arrests (SD= 13.3) and seven convictions (SD= 9.6) (the crime was not specified). When asked about the type of crime they most often committed and got in trouble for, the women cited “drug dealing” (48.5%) as their primary offense followed by “stealing” (19.9%) and “trading sex for money or drugs” (14.7%). It is important to note that these numbers are self-reported and were not corroborated by official records. The specific mandated conditions of drug treatment for the women were unknown. Women were sent to outpatient, residential and in-prison drug treatment programs (the difference in drug treatment placement was due to the severity of their drug use and/or specifics of their criminal record). The women reported on average attending 2.71 (SD= 6.45) drug treatments in the past (excluding detox).
Since drug use and crime were expected by definition in this sample, the main focus was to identify the age of first drug use, age of first arrest and then assess whether childhood victimization existed among this group. This was the basis for formulating the pathways.
Specifically, this research sought to answer the following questions:
The focus of the paper is on the pathways, which by definition in this study is a sequential relationship between three age variables: age of first arrest, age of first drug use and age of first time they were abused as children. The logic behind this approach was derived from the life course perspective (Elder, 1985) and Life Course Theory (Sampson & Laub, 1993). Both the perspective and the theory emphasize the lasting effect of particular life events that occur during childhood, adolescence, and early adulthood. In particular, the life course perspective distinctively calls for the study of pathways to crime through an “age-differentiated life span” (Elder, 1985, p.17). The current study organized victimization, drug use and first arrest by the age at which each event first occurred so that it would be clear whether childhood victimization preceded deviant behavior in most of the sample.
Age of first arrest: assessed the age at which the women first became involved with the criminal justice system. The interviewers plotted several 12 month period timelines based on the topic discussed (e.g. crime). The respondent and the interviewer began to plot events based on a 12 month period prior to their current incarceration or current admission to outpatient treatment program. The interviewer would then move backwards in time. In particular, the question regarding age of first arrest was in the section of the questionnaire that asked about illegal activities. The interviewer first asked general questions (i.e., who taught you the most about committing “crime A”? The interviewer then asked, how old were you the first time you were arrested? The interviewers were trained to give respondents additional probes to help them remember as accurately as possible. Although this only measured the first phase of involvement in the system, early age at first arrest has been considered one of the strongest predictors of long-term and serious criminal activity (Loeber & LeBlanc, 1990).
Age of first drug use: participants were asked the age at which they used an illegal substance for the first time. The drugs included under “illegal substances” were alcohol to intoxication (prior to the age of 21), inhalants (glue, amyl nitrate), marijuana or hashish (blunts, buddah), hallucinogens (LSD, mushrooms, PCP), cocaine (not crack), crack/freebase, heroin, heroin and cocaine(speedball), methadone (prescribed or street), other opiates (codeine, morphine, percodan), methamphetamine (speed, ice, crystal, bennies), other amphetamines (uppers, diet pills), Librium/Valium/mild tranquilizers, barbiturates (downers, yellow jackets), and other sedatives (quaaludes).
Age of first childhood abuse: the women were asked how old they were the first time they were sexually, physically abused and/or neglected.
Childhood Sexual, Physical Abuse and Neglect: assessed through a series of questions a measure of sexual and/or physical abuse and neglect before the age of 18. The women were asked (1) whether any member of their family2 had (a) ever fondled, touched, (b) pressured or (c) physically forced them to have sex against their will? (0=No; 1=Yes) and/or (2) deprived them of food, locked or tied them up, beat or burned them? (0=No; 1=Yes). These variables were later combined into a single dichotomous variable, childhood victimization (0=No, 1=Yes). Although, it would have been interesting to assess these groups independently, combining them into a single group increased cell counts and confidence in the accuracy of the findings. Moreover, prior research grouped early sexual and physical abuse into a “childhood victimization” category when assessing women’s future criminality (BJS, 1999; Daly, 1992; Hubbard & Pratt, 2002; Wright, Salisbury & Van Voorhis, 2007; Salisbury & Van Voorhis, 2009).
Family Risk Factors: variables included familial drug use, parental incarceration and foster care. Familial drug use was assessed through self-reports during the face-to-face interviews. This was indicated by a multiple-item dichotomous variable (0=No; 1=Yes). The question in the interview which reflected this variable was: “Thinking back to the time while you were growing up, did any of the following family members (a) mother, (b) father, (c) other adult who raised you, (d) brothers/sisters, (e) other children in house, (f) other adults in house used illegal drugs regularly -- 3 or more days a week?” A positive response to any of these questions resulted in “yes” to familial drug use. Parental incarceration was measured by asking whether either the respondent’s mother or father had ever been in jail or prison (0=No; 1=Yes). The foster care variable was measured by asking whether the respondent had ever been in foster care or was given up for adoption (0=No; 1=Yes).
Prior Mental Disorder Diagnosis: This construct was operationalized by asking: have you ever in your life been prescribed medication for any psychological or emotional difficulties? (0=No; 1=Yes). This measure is a proxy to an official diagnosis and does not account for undiagnosed mental illness, which is prevalent among women in the criminal justice system.
The analysis tested the overarching hypothesis that the main path (childhood victimization→ leads to drug use→ leads to arrest) as outlined by the feminist pathway perspective would emerge as the dominant pathway among a group of substance-abusing female offenders. Analyses were conducted using PASW Statistics 18.0 (SPSS 18). First, bivariate analyses explored the association among the primary variables. Second, all pathway sequences were determined by using the ages of first occurrence of childhood victimization, first drug use, and first arrest. Pathways without any cases were eliminated. Third, the frequency distribution of all the pathways was calculated. Finally, logistic regression was used to investigate differences in characteristics and risk factors of women in dominant pathways.
The study sample of 1,209 consisted of approximately 39 percent Black/African American, followed by 36 percent White and 18.8 percent Hispanic. There were no significant differences among the sample based on geographical location, except for race. In contrast to the New York sample, the majority of the women in Portland, OR identified themselves as White. The majority reported having less than a high school diploma (67.3%) and never being married (52.7%). In addition, 28.5 percent were unemployed 12 months prior to incarceration or treatment and were on average 33 years old at the time of the interview.
Approximately 48.2 percent of the women in the sample were abused during childhood. The average age was 9 years old (SD= 5.10) when they were first abused and the majority of the women experienced the abuse more than once (82.6%). On average the women were 14 years old (SD= 5.30, range 4–49) when they used drugs for the first time. Considering that all of the women on this sample were criminally involved, the average age of first arrest was 23 (SD=8.10, range: 10–65), a later age of onset than usually reported among women in the criminal justice system. Moreover, a mean of 8 years (SD=7.90) passed between the time the women were first abused and the time when they were first arrested.
In addition, approximately 46 percent of the women in the sample reported that marijuana was their first experience with an illegal drug followed by 33 percent who stated that alcohol was their first illegal drug. Answers about the women’s current drug problem revealed the majority of them considered crack cocaine as their main problem (36.7%) followed by 21 percent who acknowledged that heroin was their primary drug followed by 16 percent who considered methamphetamine their core drug issue.
The primary purpose of this study was to investigate the sequential distribution patterns of women’s pathways into the criminal justice system. The number of all possible pathways were calculated by using the age-based variables (i.e. first childhood abuse, first drug use and first arrest). A permutation formula for 3 sets of pathway combinations was later used to configure all the possible sequences. There were three possible sequences drawn from the sample which yielded a total of nine pathways (see Table 1).
Table 1 shows the percentage of the sample in each of the nine pathways. The sample was categorized into nine pathways (paths 0 through 8) by organizing each variable into a chronological sequence. All of the pathways are mutually exclusive from each other. That is, women were placed in only one pathway. Pathway 0, 1 and 2 display a sequence where childhood victimization is not present. In Pathway 0, first drug use and first arrest occurred concurrently (n=41; 3.5%). In Pathway 1, the most common path, first drug use occurred prior to first arrest; almost half of the sample (n=542; 45.7%) was included in this category. Contrary to Pathway 1, in Pathway 2 first arrest occurred prior to first drug use; only 3.5 percent (n=41) were in this category. Pathways 3 through 8 reflected women who reported a history of childhood abuse. Pathway 3 (childhood abuse, drug use and then arrest) was the second most common pathway in the sample (n=431; 36.3%).
Pathway 4 described women who were also abused as children but were arrested prior to their first drug use. Only 2.7 percent (n=32) of the women in the sample followed this path. Pathway 5 portrayed a third variation. In this case women used drugs first prior to child victimization and were later arrested. There were a few more women in this pathway (6.3%, n=75) than there were in Pathway 4. Furthermore, there were 1.7 percent (n=21) of women in Pathway 6, which began with women’s drug use, followed by first arrest and childhood victimization. As mentioned above, some of the women were arrested prior to the age of 18; therefore in sequences where arrests occurred prior to childhood victimization, it means that those women entered the criminal justice/juvenile system earlier in life. Lastly, there were only two women under Pathway 7 and one woman under Pathway 8. Both of these pathways began with first arrest; in Pathway 7, women endured childhood victimization followed by their first experience with drugs. In Pathway 8, women were first arrested, and then had their first experience with drugs prior to childhood victimization.
Overall, it is clear that Pathway 1 (D→A) was the most common sequence. Childhood victimization was absent from this pathway, resulting in a key finding that differs from previous literature. Pathway 3 (CV→D→A) was the second most prevalent. This pathway included childhood victimization, drug use and arrest. Furthermore, 19.5 percent of the women followed a sequence different from the two dominant pathways (pathways 4 (CV→A→D), 5 (D→CV→A), and 6 (D→A→CV)). Pathways 7 (A→CV→D) and 8 (A→D→CV) included a very limited number of women (n=2; n=1 respectively). Due to the small numbers, both of these pathways were dropped from further analyses.
A joint contingency table analysis (see Table 2) allowed for a closer look at the risk factors and characteristics of the two groups of women who were under the most dominant paths (1 and 3). Significant results were found among the common predictor variables. Overall, women in Pathway 3, which consisted of women who were first victimized in childhood, then began using drugs and were later arrested, had higher rates of familial drug use, parental incarceration, prior foster care and reported a prior mental health condition in comparison to women in Pathway 1 (the most dominant path). Race, marital status and employment status were not significant.
Specifically, 46.9 percent of women in Pathway 3 grew up in families where drug use was common in comparison to 32.5 percent of the women in Pathway 1 (χ2=20.94, p<.001). Moreover, 26.2 percent of women in Pathway 3 reported having had at least one parent in prison during their childhood, whereas 11.3 percent of the women in Pathway 1 (χ2=36.6, p<.001) recounted the same experience. The same difference was reflected in prior foster care; a higher percentage of women in Pathway 3 had been in foster care in comparison to those women in Pathway 1 (25.8% vs. 10.6% ; χ2=38.76, p<.001). Also, a higher percentage of women in Pathway 3 appeared to have been prescribed medication for a previous mental health diagnosis compared to the women in Pathway 1 (42.0% vs. 30.2%; χ2=14.3, p<.001). Lastly, women in Pathway 3 were found to have a slightly lower average age of first drug use and average age of first arrest compared to women in Pathway 1 (see Table 2).
A logistic regression model (see Table 3) was constructed to further evaluate whether the two main Pathways (1 and 3) were related to other characteristics. Specifically, this analysis sought to find whether demographic variables and risk factors of female offenders were related to Pathways 1 or 3 (Pathway 1= 0 and Pathway 3= 1). The model was statistically significant (χ2=84.2, p<.001; R2 = .112). Familial drug use, parental incarceration, foster care and a previous diagnosed mental health condition were significantly related to Pathway 3. First, experiencing family drug use in childhood increased the odds of following Pathway 3 by a factor of 1.53 (logit=.427, p<.01). Also, having a parent in prison during childhood increased the odds of following Pathway 3 by a factor of 2.12 (logit=.753, p<.001). Next, prior foster care also increased the odds of being under Pathway 3 by a factor of 2.55 (logit=.937, p<.001). Lastly, previous prescription of mental health medication increased the odds of following Pathway 3 by a factor of 1.5 (logit=.375, p<.05). Consequently, having the following characteristics, familial drug use, parental incarceration, prior foster care and prior prescription of mental health medication for diagnosis placed women at greater risk of having followed Pathway 3. It is important to note that temporal order is not established among the risk factor variables and the correlations observed are not assumed to cause (or be caused by) the victimization – substance abuse – arrest pathway.
The present research study was designed primarily to address women’s pathways into the criminal justice system. The specific aims were to: (1) identify all populated pathways among substance-abusing female offenders; (2) detect the most common sequences followed by the women; and (3) examine the relationship between other risk factors and the common pathways.
Of the nine possible pathways, only two had very few cases and did not appear to be viable pathways. Similar to previous research this study established that women come into contact with the criminal justice system through distinct etiological routes (Brennan et al., 2010; Daly, 1992; DeHart, 2004; Simpson et al., 2008). However, the present research chronologically organized these routes by using the age at which key events first occurred (childhood abuse, drug use and arrest).
One of the main assumptions of feminist pathway perspective is that in order to understand the causes of women’s delinquent behaviors, one must emphasize and focus on women’s childhood victimization (Belknap & Holsinger, 2006). Previous feminist pathway research reported high incidences of childhood abuse among female offenders (Bloom, et al., 2004; Bloom et al., 2003a; Bloom et al., 2003b; Burguess-Proctor, 2012; Chen et al., 2004; Greenfeld & Snell, 2000), and thus assumes that childhood abuse is central to women’s offending trajectory. Most of this research positions childhood sexual and physical victimization in the beginning of female offenders’ life trajectories. This is then expected to lead to drug use and inevitably to arrest/contact with the criminal justice system (Bloom et al., 2003a; Bloom & Covington, 2009; Owen, 1998; Van Voorhis & Presser, 2001). This is the most common argument regarding women’s pathways into criminal behavior—that child abuse creates a trauma in a young girl’s life, the adolescent girl begins to use drugs as a coping strategy and arrest results often as a by-product of this combination.
Thus, the findings of this research supports: (1) the assumption that childhood victimization is generally present in women offender’s lives and also confirms (2) the prevalence of this particular path as proposed by feminist pathway perspective. However, this study finds that a drug pathway, without childhood abuse, is also as important and more dominant among criminal justice-involved women (Pathway 1). The absence of the key variable, childhood victimization, was inconsistent with much of the previous literature on women’s pathways, which found that childhood abuse was central to the events that lead women (and girls) to justice involvement (Bloom et al., 2004; Bloom et al., 2003a; Bloom et al., 2003b; Daly, 1992; Gilfus, 1992). Although childhood victimization is high among this population, it is not the most common pathway. This finding is congruous with the work of Carbone-Lopez and Miller (2012) which found strong evidence for women’s pathway into drug use (methamphetamine) that did not include traumatic childhood events (i.e. sexual/physical abuse, family dysfunction). They cautioned against the overemphasis on the role of victimization and trauma on later offending.
The finding that Pathway 1 (D→A) was the most predominant path slightly changes the discussion of substance-abusing women as victims. The drug use to arrest pathway, is consistent with literature that posits women are more likely to be involved in crime if they are drug users (Brown , 2002; Chesney-Lind, 1997; Cobbina, 2009; Bloom et al., 2003; Van Voorhis, et al., 2001, 2008).The absence of victimization in this pathway could be due to: (1) simply the non-existence of the event or (2) the lack of disclosure about victim status. It is possible that not all women disclosed their victim status, but if this is the case, this is an important point to acknowledge. This has direct implications to substance-abusing women’s identity of self. This corresponds to Anderson’s symbolic resistance agency, which states that “women actively reject stigma, derogatory images and undesirable expectations by managing and creating new alternative identities…” (p.5). Nevertheless, as Katsulis and Blankenship (2008) point out, substance-abusing females can change the narrative of their identity depending on the setting. For example, women in a criminal justice system may use the victim role narrative not to disrupt an image of a repentant woman. Conversely, this identity is less useful while in a street setting, where women may not want to portray themselves as victims.
However, this study does not reject that women and girls often come in contact with the criminal justice system based on survival of abuse/victimization, since it proved to be the second most prevalent pathway among this group of women. This research highlights that there is not just one entry into the criminal justice system that accounts for the origins of women’s crime. Similar to the theorizing process and multiple studies conducted about men’s criminality, the study of women’s pathways reveal overlapping and complicated mechanisms for their criminal behavior.
Results from the third research question: Are the common pathways related to other characteristics (i.e., demographics, family drug use, parental incarceration, prior foster care, mental health) demonstrated that women who are under Pathway 3 (CV→D→A) had more risk factors than women who followed Pathway 1 (D→A). Higher percentages of the women in Pathway 3 experienced previous familial drug use, parental incarceration, prior foster care, and prior mental health diagnosis. The same variables were statistically significant in predicting the likelihood of following Pathway 3 versus Pathway 1. These findings are consistent with previous research, which affirmed that many of the gendered risk factors are associated with Pathway 3; the path often cited by feminist criminologists (Reisig et al., 2006). Furthermore, these results support previous research, which stressed that experiencing multiple forms of abuse or household dysfunction during childhood produces harmful consequences on adult mental health (Chapman et al., 2004) and other health and behavioral statuses. In fact, women with a history of familial drug use, parental incarceration and prior foster care had a higher risk of having childhood victimization in its path to the criminal justice system. Thus, it is safe to assume that this could have been the result of inadequate or inappropriate parental care.
Furthermore, according to Reisig et al. (2006), women under a path that closely mirrored women in Pathway 3 (CV→D→A), have been found to be considerably misclassified in correctional settings. These women tend to be classified as least likely to recidivate (Holtfreter & Morash, 2003; Reisig et al., 2006; Van Voorhis & Presser, 2001). Due to this low-risk misclassification, women under Pathway 3 receive less resources and treatment (Reisig et al., 2006). As per the findings of this study, this could have severe implications for the criminal justice system (i.e. recidivism) given the additional trauma and familial conditions that women under Pathway 3 endure. This misclassification perpetuates our ignorance or denial about women’s poor mental health, which therein continues to go untreated. Women who are appropriately diagnosed and provided comprehensive treatment have a greater capacity to support themselves once they are released.
Also, the results revealed that childhood abuse occurred within the context of poor family functioning. Previous research suggests that childhood abuse is often not an isolated event in children’s lives. They are also more likely to be exposed to other adverse childhood experiences such as parental divorce/separation, domestic violence, familial criminality, familial substance abuse and exposure to a household member with a mental illness (Anda et al., 2006; Chapman et al., 2004; Dong et al., 2004; Felitti et al., 1998). Therefore it is possible that the women also suffered added trauma from not feeling able to share or turn to her family members for help. This case was highlighted in Baskin and Sommers (1998) research, where many of the women reported that they felt “unprotected and alone” in the aftermath of the abuse due to the uncertainty of their familial environment. Moreover, this study’s findings support Daly’s (1992) “harmed and harming women” pathway and more recent research assessing women’s likelihood of recidivism (Reisig et al., 2006). “Harmed and harming” women, like those in Pathway 3 suffered from neglect, sexual and/or physical abuse as children, began abusing drugs as teens and later entered the criminal justice system through felony court. In addition, Reisig et al. (2006) found that women under a harmed and harming pathway also lived in unhealthy family conditions and began abusing substances in early childhood. Daly (1992) states that these women were unable to cope with difficult situations and displayed symptoms of psychological harm.
Prior childhood victimization is one of the main threads connecting feminist pathway research. Women’s narratives regarding their previous sexual, physical abuse and neglect has led to a profound understanding regarding the link between childhood abuse and women’s psychological well-being, drug use and criminality (Bloom et al., 2003; Chesney-Lind, 1997; Covington, 1998; Giordano, 2010; Owen, 1998). Moreover, feminist pathway perspective “…examines girls’ and women’s histories, allowing them, when possible, a voice in order to understand the link between childhood, adult events, traumas and the likelihood of subsequent offending” (Belknap, 2007, p. 71). It is this retrospective/life history aspect that allows for a deeper understanding of the psychological/physical processes that impact women’s and girls’ status as an offender. Giordano (2010) justified this approach by highlighting the need to understand women’s past to truly discern the reasons for their criminal lives as adults.
Since this study used a purposive sampling technique in order to identify women in the criminal justice system who had substance abuse problems, all women were criminal justice-involved, and therefore this research could not differentiate between offenders and non-offenders. Therefore, these results are not generalizable to women who were not in the criminal justice system or women in the criminal justice system who were not drug users. Similarly, due to the targeted sample, the study cannot make statements about women who were victimized as children who never used drugs or entered the criminal justice system. It is also possible that these pathways are not gender-specific, however, we cannot be certain due to the nature of the sample. Nevertheless, the large sample size increases the confidence of the results among substance-abusing female offenders. Moreover, due to the geographic location of the sample, this study cannot make assumptions about substance abusing female offenders in other cities. However, given the opposing characteristics of both cities (New York and Portland), a representative racial and ethnic sample of women in the criminal justice system was achieved.
The self-reported nature of women’s drug use fails to differentiate the magnitude of their substance abuse. The interview also directed women to disclose any childhood abuse by family, adopted family and friends of family. This may have excluded answers from women who were victimized as children by complete strangers. Also, even after the use of timelines to help the women recall past arrests (i.e. crime), drug use and victimization; the women could have had an issue recalling exact times or ages. Thus, there is a chance that the sequencing of pathways could be biased. For the most part, interviewers felt that the women were honest about their answers. Moreover, respondents were not randomly selected by the researchers. Instead, participants self-selected themselves and volunteered for the interview. There is evidence that non-random selection can yield biased results; more so when participants self-select than when others make the selection decision (Shadish, Cook & Campbell, 2002). The degree of this potential bias is unknown. Nonetheless, this sample was racially and ethnically diverse and the distributions of other demographic variables demonstrated substantial variation. Despite the above limitations the findings are clear in demonstrating that multiple pathways for women in the criminal justice system exist. More importantly, a path lacking previous victimization highlights the need to research additional avenues by which women drug users enter the system.
As with the majority of feminist pathway research, this study calls attention to the way we respond to women’s paths into the criminal justice system. Programs and services in correctional facilities and the criminal justice system as a whole should take into account the context of women’s lives with the understanding that this may explain, at least to some degree, how and why women get arrested. Even though the child victimization pathway was not the most dominant one, results from this study reveal that women in this particular path have more risk factors associated with their lives. These include familial drug use, parental incarceration, prior foster care and mental health issues. Therefore treatment programs that specifically target childhood victimization in combination with problematic family life may have a positive effect on women’s criminal and drug-using patterns of behavior. Other researchers have also called for similar programming, which emphasizes targeting familial factors when developing a gender-specific curriculum for delinquent girls (Bloom et al., 2003; Bloom et al., 2004; Kruttschnitt & Giordano, 2009; Van Voorhis, 2001). Beyond services behind bars, these types of programs need to be known and readily available to probation officers, judges and other court staff that refer girls and women to programs. This may seem like a natural process to follow after incarceration/detention but decreasing budgets often lead to less gender-specific programming in cities and states.
Intrinsically, the results of this study also have implications for the development of drug and delinquency/crime prevention programs for girls. A comprehensive curriculum that includes opportunities for the girls in these settings to exert some sort of agency, which they may be missing in other aspects of their lives, can also yield more promising results with girls who have been abused. Often researchers and/or practitioners have a sense of “what works,” but as the results of this study and the vast majority of feminist criminology research show, maintaining one assumption of how girls/women enter the criminal justice system limits the type of approaches to support this population.
Another implication based on the findings is the need for an assessment that takes into consideration the age in which “risk factors” occur in women’s lives. In other words, the assumption that age matters should be translated into assessment protocols for criminal justice and juvenile populations. Childhood victimization may have lasting effects if not addressed. Although, this study did not test how the age of victimization affected deviant behavior directly, descriptive statistics showed that girls were on average nine years old when they were victimized and were arrested later in adulthood. These results substantiate previous research which showed that the criminogenic effects of childhood victimization may not be evident until later in life (Ireland & Widom, 1994). Thus it is critical that services are provided and target the deleterious effects caused by victimization, even if trauma is not visible. Furthermore, during the assessment period, we should allow for greater verbalization of women’s own thoughts regarding 1) the reasons why they entered the system and 2) their strengths and buffers, which may have kept them out of the criminal justice system (Cobbina, 2009; DeHart, 2004).
Based on our findings, future research should sequence other key life events (e.g. running away) into the pathways. Moreover, the relevancy of perpetrators other than family members should be included in future data collection to test its long-term effects. Future research would also benefit from testing whether the average time between child abuse and first drug use would make a difference in the time it would take for the victim to enter the criminal justice system.
1An evaluation of timelines or Life Event Calendar methods suggested that it provides more accurate estimates of events closer to the year prior to the interview rather than events in the more distant past. Nevertheless, it is possible that respondents can provide accurate responses of the timing of criminal events, which occurred within a longer timeframe if they are asked to report the year in which an event occurred rather than the exact month (Morris & Slocum, 2010). Traditional survey methods ask questions in a segmented format and do not reflect the interrelatedness of life events (Belli, 1998, 2001). Instead, LECs promote sequential thinking regarding a person's past which in turn can maximize the quality of retrospective reports (Axinn, Pearce & Ghimire, 1999; Belli, 1998), In addition, Morris and Slocum (2010) found individuals with fewer arrests and arrests for more serious crimes reported the number of months they had been arrested more accurately. Respondents were also more likely to correctly identify when arrests occurred if the arrest month coincided with a period of incarceration. Incarceration may act as a landmark that facilitates the accurate placement of arrests in time because it disrupts normal routines. Lastly, significant to the current research, substance use was not found to affect the validity of self-reports (Morris & Slocum, 2010).
2These include: mother alone, father alone, stepmother alone, stepfather alone, mother and father or parent and step-parent, foster care mother, foster care father, adopted mother, adopted father, grandmother, grandfather, aunt or great-aunt, brother, sister, other relative, friend of my family.