A Predictive Theory of Intentions to Exit Street-Level Prostitution

Entering Street-Level Prostitution

The literature describes a combination of “push” and “pull” factors that prompt women to engage in prostitution (Mayhew & Mossman, 2007). Poverty, a push factor, is a common reason some women enter street-level prostitution (Farley & Kelly, 2000; Sanders, 2007; Williamson & Folaron, 2003). Other push factors include an unstable home life, childhood physical and sexual abuse, and neglect (Farley et al., 2003; McClanahan, McClelland, Abram, & Teplin, 1999; Nadon, Koverola, & Schludermann, 1998; Norton-Hawk, 2001; Silbert & Pines, 1982; Ward & Roe-Sepowitz, 2009; Weitzer, 2009; Williamson & Folaron, 2003) or drug addiction (Brawn & Roe-Sepowitz, 2008; Edlund & Korn, 2002; Young, Boyd, & Hubbell, 2000). Pull factors include one’s glamorization of the lifestyle (Williamson & Folaron, 2003), a sense of excitement from dangerous and risky behavior (Dalla, 2006a), feeling in control or empowered (Mayhew & Mossman, 2007), encouragement from others (Kennedy, Klein, Bristowe, Cooper, & Yuille, 2007), and the desire for economic independence (Edlund & Korn, 2002; Weitzer, 2009).

Regardless of one’s motivations for entering prostitution, undeniable are the damaging effects street-level prostitution has on women’s emotional and physical health including disease, violence, rape, and murder (Farley et al., 2003; Farley & Kelly, 2000; Pauw & Brener, 1997; Williamson & Folaron, 2003). It is important to differentiate between street-level prostitution and indoor prostitution (e.g., escorts, brothels) because street prostitution has more risks than other kinds of prostitution (Sanders, O’Neill, & Pitcher, 2009; Weitzer, 1991, 2005, 2009; West, 2000). Street prostitution is a dangerous way of life characterized by coercion, exploitation, and violence. Several studies indicated that 41 to 68% of street-level prostituted women experienced threats with weapons, and physical and sexual assaults by clients while prostituting (Farley & Kelly, 2000; Pearl, 1987; Raphael & Shapiro, 2004; Surratt, Inciardi, Kurtz, & Kiley, 2004). When compared to nonprostitutes of similar demographics, prostituted women are nearly 18 times more likely to be murdered (Potterat et al., 2004). Due to the nature of the work, these women are at an increased risk of contracting human immunodeficiency virus (HIV) and sexually transmitted infections (STIs) than women who do not prostitute (Farley & Kelly, 2000). Many prostituted women are addicted to substances and use them to cope with their lifestyle (Allen, Flaherty, & Ely, 2010; Farley & Kelly, 2000; Hedin & Månsson, 2004; Logan, & Leukefeld, 2000). In addition, most street-level prostituted women experience some form of mental health problems such as depression or bipolar disorder (Davis, 2000; Farley & Kelly, 2000; Weitzer, 2009; Williamson & Folaron, 2003; Young et al., 2000), and in a study of 854 prostituted persons in nine countries, 68% met the criteria for posttraumatic stress disorder (Farley et al., 2003).

Barriers to Exiting

While entrance histories are unique, some of the factors that pushed women into street-level prostitution (e.g., economic necessity, drug addiction, coercion from others) are factors that continue to make it difficult for them to exit. These barriers to exiting prostitution are well described in the literature as a constellation of individual, relational, and structural factors (Baker, Dalla, & Williamson, 2010; Månsson & Hedin, 1999; Sanders, 2007). Poverty and a lack of resources are persistent needs of street-level prostituted women and act as barriers to exiting (Farley & Kelly, 2000; Mayhew & Mossman, 2007). Some women perform “survival sex” where they trade sexual services for basic necessities such as food or a place to stay (Dalla, 2006a). Addiction (Allen et al., 2010; Benson & Matthews, 1995; Farley & Kelly, 2000; Mayhew & Mossman, 2007), severe mental illness (Farley & Barkan, 1998), and interpersonal stigma from selling oneself (Mayhew & Mossman, 2007) prevents some prostituted women from seeking formal services or help from family. Coercion, deception, and violence from pimps and fellow prostitutes serve to normalize and reinforce prostitution (Kennedy et al., 2007; Nixon, Tutty, Downe, Gorkoff, & Ursel, 2002); these associates provide an underground network of social support and encourage continued involvement in street-level prostitution (Mayhew & Mossman, 2007). Women attempting to exit who lack education and job skills or have criminal histories are often unable to gain legitimate employment, thus compelling them to resort to prostitution (Edlund & Korn, 2002; Sanders, 2007).

The Effectiveness of Formal Prostitution Exiting Services

Prostitution exiting programs offer formal services for women who are trying to escape prostitution, but only a few states in the United States offer such programs (Farley, 2010). Typically, these programs provide substance abuse treatment and counseling for psychological traumas common among prostituted women. Best practice exiting interventions are holistic and include a combination of counseling, housing, employment training, and education (Arnold, Stewart, & McNeece, 2000; Mayhew & Mossman, 2007). As an added benefit, exiting programs reduce environmental barriers to exiting (e.g., separation from violent pimps and johns, no need to engage in survival sex). With these services, commonsense would suggest the “push and pull” factors should no longer be an issue. It is curious, however, that despite ameliorating many barriers to exiting, reentry is routine among women in exiting programs. For instance, in Davis’s (2000) study, only 73 women completed treatment and left prostitution, while 291 dropped out of the program. Only a quarter of Saphira and Herbet’s (2004) study of 47 prostituted women were able to exit, though two-thirds made several attempts. In Dalla’s (2006a) study, 13 of 18 women returned to prostitution. In a study of brothel sex workers in Thailand, 60% (n = 42) quit and reentered sex work at least once (Manopaiboon et al., 2003). While 16 of these women successfully left prostitution, 23% said they would return to prostitution if their situation made it necessary. Finally, in a Canadian study of 201 sex workers, 7 in 10 had exited sex work at least once, over half exited three or more times, and women who did exit made an average of 5.6 exiting attempts (Benoit & Millar 2001).

The pervasiveness of the “enter, exit, reenter” cycle among street-level prostituted women suggests there may be a fundamental difference between women who can permanently exit prostitution and those who cannot. Researchers have begun studying exiting more closely, proposing several models that describe the exiting process. Much of this research is based on in-depth interviews between prostituted women and scholars in the field. These qualitative examinations are necessary and appropriate given the complex nature of the population and the need for sensitive, nonexploitative research practices (Shaver, 2005). These methods provide rich data and much insight into the lived experiences of prostituted women but have only been able to describe the stages street-level prostituted women naturally experience as they transition out of prostitution and into a conventional lifestyle. Research now calls for more sophisticated, testable theories that can better explain and predict an exit from prostitution, while incorporating existing models (Baker et al., 2010; Dalla, 2006b). This article proposes a theory examining attitudes, norms, and self-efficacy beliefs toward prostitution; notions of agency; and the impact of societal context that may contribute to the exiting process. This theory may help predict an exit from street-level prostitution.

Williamson and Folaron’s “Phases of the Lifestyle”

Several researchers have conceptualized the exiting process among street-level prostituted women. Williamson and Folaron (2003) described a five-stage entry-to-exit process where one becomes enticed by the prostitution lifestyle, then learns the prostitution culture, and begins to survive and thrive in prostitution or “trust the game” (p. 279). Prostituted women exit once they become disillusioned with the lifestyle after violence, drug addiction, arrests, and trauma have taken their toll. This last stage is described as “taking stock and getting out” (p. 283). While their description was not solely about exiting, it highlights the accumulation of negative events and attitude changes that are necessary precursors to exiting: “It is the sum total of daily hassles, acute traumas, and chronic conditions that precipitate a woman’s decision to exit prostitution” (p. 283).

Månsson and Hedin’s “Breakaway Model”

Månsson and Hedin’s (1999) prostitution “breakaway” model, based on role exit theory (see Fuchs Ebaugh, 1988; Oselin, 2010; Vanwesenbeeck, 1994), describes how most women experience a turning point such as a traumatic, eye-opening event that prompts their break away from prostitution. After this event, they contemplate leaving, try to locate exiting resources, and then either quickly or gradually leave prostitution. A complex interplay of structural, relational, and individual factors also impact a woman’s ability to exit prostitution. Structural factors included access to welfare or resources for securing conventional employment; relational factors included seeking or repairing relationships with family, children, and men as well as accessing formal support from social workers; individual factors were beliefs in one’s ability to change. After exiting, a woman faced continual challenges as old lifestyle habits conflict with her newly assumed role. As women incorporated these more conventional lifestyles, they began to heal from the traumas suffered while prostituting. Hedin and Månsson (2004) found the ability of an ex-prostituted woman to change social networks was integral to her continued success away from prostitution. In sum, they concluded exiting prostitution required “the restructuring of everyday life and the occurrence of critical incidents” (p. 72), believing in one’s abilities to make positive change, and building a network of supports.

Sanders’s “Typology of Transitions”

In another discussion of exiting, Sanders (2007) identified four pathways or transitions out of prostitution, many of which are rooted in interpersonal or environmental factors. Some street-level prostitutes, Sanders described, experience reactionary transitions out of prostitution after experiencing a significant life-changing event, such as a death, rape, pregnancy or incarceration. Others experience a gradual transition out as they begin to access formal support services and put elements of a conventional lifestyle into place. A natural progression out of prostitution, Sanders observed, begins with an intrinsic desire for a new and safer lifestyle. The fourth transition, yo-yoing was the common drifting in and out of prostitution, treatment centers, and the criminal justice system. Sanders added that the positive impact of cognitive transformations (e.g., willingness to change, cutting off old ties and developing new ones) and structural disadvantages such as a lack of affordable housing, treatment for substance abuse and employment alternatives, affected the exiting process.

Baker and Colleagues’ “Integrative Model of Exiting”

More recently, a theory combining earlier exiting models with nonprostitution specific theories such as Prochaska, DiClemente, and Norcross’s (1992) Stages of Change was proposed by Baker et al. (2010). Their integrative model describes the exiting process in six stages from immersion into the lifestyle, to coming to a visceral and conscious awareness about leaving prostitution, then engaging in deliberate acts to exit prostitution (e.g., seeking formal and informal services), which would ultimately lead to an initial exit. This initial exit, they caution, often results in reentry. A “final exit” is the last stage and occurs after numerous exiting attempts are made, resulting in a complete change in identity, habits, and social networks. Baker and colleagues’ research highlighted cognitive and behavioral changes necessary in a successful exit, as well as the alleviation of structural, relational, and individual barriers women face when exiting prostitution.

Limitations of the Current Exiting Models

Part of a theory’s usefulness is determined by empirical testing (Renzetti, 2006), and while there is much inherent value in these four models of exiting, a more defined and operationalized theory ¹ is needed (Baker et al., 2010; Dalla, 2006b). To be testable, a theory should lend itself to refutable hypotheses and contain well-operationalized variables used to measure differences across groups and for replication. Ambiguity in definitions across research studies may result in vague theory building and other methodological difficulties (Herzberger, 1990; Lee & Renzetti, 1990). As with most research on vulnerable populations, the four models described here possess similar challenges. A more rigorous theoretical model that draws from earlier models, and one that builds on the cognitive change processes described in the literature, is needed to deepen our understanding of the prostitution exiting process. The theory that follows is an attempt to synthesize existing literature into more focused and measurable variables to empirically test.

Attitudes

As identified in the integrative model, attitude consists of the outcomes or results of a targeted behavior and a person’s evaluation of those outcomes. According to the model, the target person evaluates how likely and how significant an outcome might be; this is simply his/her attitude toward the consequences of performing (or not performing) the behavior (Ajzen & Fishbein, 1980). For example, two outcomes of condom use are the prevention of pregnancy and the prevention of STIs. If the targeted person is a postmenopausal woman in a monogamous relationship, for example, she may feel that pregnancy or contracting an STI is an unlikely outcome. Thus, her attitude toward condom use to prevent pregnancy or STIs might be negative or at least neutral, so she may not intend to use them. In contrast, a college-aged, nonmonogamous man may want to avoid STIs, but pregnancy prevention may be a lesser concern. Therefore, his attitude toward condom use to prevent STIs may be positive, but neutral toward pregnancy prevention. Regardless, his intention to use condoms based on his evaluations of the outcomes of using (or not using) condoms is likely greater than the monogamous, postmenopausal woman.

In the context of prostitution, attitudes might encompass a woman’s evaluation of the likelihood of being arrested, being physically assaulted, or making a lot of money should she stay in prostitution. If a prostituted woman does not think she will be arrested, and does not care about being arrested, her attitude toward prostitution may be less negative than a woman with a criminal record who wants to avoid additional arrests. Thus she may continue to prostitute because she does not associate negative attitudes toward prostitution. Conversely, a prostituted woman may believe she can make a lot of money from prostitution, but may not think the consequences of making money are worth it. Her attitude toward prostitution is negative even though the outcome of money is likely; her intention may be to not prostitute unless she desperately needed money.

Norms

Norms are defined by the behavioral prediction model as the perception of significant others’ expectations, whether these significant others are performing the target behavior and the target person’s motivation to comply; it represents the normative expectations of others in a person’s environment (Ajzen & Fishbein, 1980) and incorporates an element of social pressure (Fishbein, 2000). The expectations from specific and significant others to perform (or not perform) a behavior creates subjective norms within the targeted person. It is important to clarify that significant others’ versus nonsignificant others’ opinions are integral to the construct because it is the opinion of significant others that produce pressure and a behavioral proscription for the target person (Fishbein, 2000). In the condom use example, if the monogamous, postmenopausal woman’s husband does not use condoms and does not think she should either, she may be motivated to comply with him and thus will not use condoms. Conversely, if her doctor told her to use a condom, she may not be as motivated to comply as she was with her husband. The college-aged man, on the other hand, may have a partner who insists he uses a condom, and because he wants to please his partner, he may use a condom.

Norms, in the context of prostitution, would include the beliefs held by a prostituted woman’s significant others, which could be family members, pimps or associates, and her willingness to comply with the opinions of these others. Their performance (or nonperformance) of the behavior also influences her subjective norms. A pimp obviously wants a prostituted woman to engage in prostitution, and she would likely submit to his requests. Similarly, associates of hers who also prostitute might reinforce the behavior. A family member who she cares about, on the other hand, may not want her to prostitute and if she is motivated to comply, she may not prostitute.

Self-Efficacy

The last variable underlying intention in the predictive model is self-efficacy. Self-efficacy is a person’s belief that he/she can perform the target behavior under specific barriers or obstacles. Self-efficacy requires a sense of competence in one’s ability to perform the behavior. In other words, the more the targeted person believes he/she has the skills and abilities to perform the behavior under difficult circumstances, the stronger his/her self-efficacy will be with respect to performing the target behavior (Fishbein, 2000). Under this definition, it is possible a person who does not perceive him/herself as competent cannot perform the target behavior, which is different from a person who does not intend to act in such a manner; the self-efficacy construct captures this distinction. The partner of the nonmonogamous man, for instance, may not feel competent to ask her mate to use a condom, although she may intend for him to do so.

Self-efficacy beliefs of a prostituted woman might encompass how certain she can or cannot resist prostitution under duress, such as if she needed money or drugs. Under these difficult situations, she might be tempted to engage in prostitution, especially if she does not have alternative means to make money. In contrast, a prostituted woman with higher efficacy beliefs may perceive herself as capable of resisting prostitution under these circumstances.

Skills and Environmental Constraints

Two additional constructs in the integrative model of behavioral prediction are skills and environmental constraints. These two elements are separate from intentions, as a person can intend to perform a behavior, but may not have the necessary skills or may experience environmental constraints preventing performance of the behavior (Fishbein et al., 2001). A lack of skills or the presence of environmental constraints would obviously impact a person’s ability to perform a given behavior (Fishbein, 2000; Fishbein & Yzer, 2003). Skills, in the context of prostitution, may include life skills, coping skills learned in therapy, or basic employment skills prostituted women need to live a conventional lifestyle. Environmental constraints, like homelessness or unemployment, for example, may inhibit a street-level prostituted woman’s ability to exit versus a woman who is living in a shelter and has legitimate employment.

Finally, the integrative model accounts for distal variables such as demographics, culture, personality traits and other individual characteristics. The model suggests these variables indirectly influence behavior because there is not a single, stable relationship between these individual characteristics and a given behavior (Fishbein, 2000). For example, men and women may have similar attitudes, norms and self-efficacy about some behaviors, but may differ with respect to other behaviors. In other words, the influence of external variables shifts depending on the behavior and the context. Within prostitution, age, length of time involved in prostitution and severity of addiction may be distal variables that indirectly influence one’s ability to exit.

Synthesis with Existing Models of Exiting

The integrative model of behavioral prediction is well suited to measure and predict intentions to exit prostitution. It captures the cognitive, relational and environmental variables that help or hinder an exit described in existing models. For example, Williamson and Folaron (2003) suggested that a sum of daily hassles must occur for a person to shift from glamorizing the life to being disillusioned with it; this is captured by the attitudes construct of the predictive model. Both Månsson and Hedin (1999) and Sanders (2007) described relationship changes that have some bearing on a person’s decision and ability to exit, which is expressed in the norms construct. Månsson and Hedin (1999) also described how one’s internal drives and abilities to achieve positive results were part of the breakaway process, paralleling notions of self-efficacy in the predictive model. Finally, existing prostitution-specific exiting models, as well as the predictive model, place emphasis on skills (e.g., job or life skills) and the reduction of environmental constraints (e.g., obtaining housing) necessary for an exit. Baker et al. (2010) stressed the planning process prostituted women undergo when exiting; this is similar to the skills construct in the predictive model. Also, Sanders (2007) identified securing housing and removing oneself from street life as factors that contributed to a successful exit, which coincide with the predictive model’s discussion of environmental constraints. These similarities provide support for the development and examination of a predictive theory of intentions to exit street-level prostitution.

While there is an obvious overlap between existing models and the integrative model of behavioral prediction, one distinctive difference is the emphasis on latent psychosocial variables underlying behavior. The attitude, norm, and self-efficacy constructs, which have been unaddressed in prostitution exiting research, may uncover reasons why a disproportionate number of women cannot successfully exit prostitution. For instance, exiting programs address environmental constraints by providing women who want to leave prostitution with shelter, basic necessities, and a separation from deviant associates (e.g., pimps, johns), as well as provide skills learned in counseling or vocational programs. Yet, even when provided these resources, women continue to reenter prostitution, suggesting that some women are more ready or have stronger desires to exit than others. Thus, it appears that current programming efforts are ineffective because they ignore intentions and only provide skills or reduce some environmental barriers to exiting. Consequently, differences in attitudes, norms, and self-efficacy beliefs toward prostitution may explain why some women are able to exit street-level prostitution. The integrative model of behavioral prediction offers a testable theory to examine differences in intentions, attitudes, norms, and self-efficacy. If these variables differ among women, then it is possible to predict an exit from street-level prostitution.

Agency and Societal Context as Enhancements to the Model

An assumption the predictive model makes about behavior is that an exit must be under a person’s volitional control. One can argue, however, that a woman’s decision to exit or continue street-level prostitution is not likely under her control, especially since research has indicated that 42% to 80% of street-level prostituted women are, at some time, in pimp-controlled prostitution (Giobbe, 1993; Norton-Hawk, 2004; Williamson & Cluse-Tolar, 2002). Issues of power, and physical and emotional control, are inherent in the pimp-prostitute dynamic. Some women in pimp-controlled prostitution have difficulty leaving a pimp, and therefore prostitution, because they feel love, trust, or a sense of loyalty to their pimp, despite the threatening, abusive, and coercive tactics pimps use to keep them prostituting (Dalla, 2006a; Williamson & Cluse-Tolar, 2002). Other women rely on pimps to provide them protection, shelter, or other necessities (Dalla, 2002b), which they could lose if they stopped prostituting. Dalla (2006a) suggests that control from nonpimp partners and severe drug addiction can also affect one’s ability to exit prostitution (Ratner, 1993). Therefore, one enhancement to the model would be to include agency as a moderator of intention. Agency is the capacity of an individual to act independently and to make his/her own free choices. Agency is a uniquely important feature of the prostitution lifestyle, as the degree to which a prostituted woman can make free choices will affect her intention and ability to exit prostitution. For instance, a woman under a pimp’s control may want and intend to exit, but her pimp may force her to engage in prostitution through the use of violence. In some cases, a prostituted woman may also wish to stay under the protection of her pimp. In addition, the fear of violent retaliation keeps some women from even attempting to exit (Nixon et al., 2002), despite the strongest intentions to leave. While the agency construct may appear similar to normative beliefs, its focus on one’s capacity to act independently makes it a distinctive variable. Additionally, Ajzen’s (2002) notion of perceived behavioral control, defined as the “perceived ease or difficulty of performing a behavior” (p. 665), also attempts to deal with a lack of volitional control. Agency, however, is conceptually unique because it taps into power and control dynamics caused by external entities (i.e., a person or force vs. one’s own sense of control to perform a behavior).

Another enhancement to the predictive model is to incorporate the influence of societal context. As Baker et al. (2010) and Månsson and Hedin (1999) have suggested, structural forces (e.g., ability to find a job, other material conditions) affect whether a woman is capable of exiting prostitution. For the street-level prostituted woman, her ability to support herself is key to living a conventional life away from prostitution. While employment and equitable wages may be a function of one’s skills and qualifications, it is also influenced by wider societal influences. For instance, in an economic downturn fewer jobs are available, making legitimate work difficult to find for an ex-prostituted woman despite her intentions. Welfare assistance also makes transitioning out of prostitution easier (Månsson & Hedin, 1999), and policy changes reducing the amount of public assistance available have real consequences on one’s ability to exit. Changes in criminal sanctioning of prostitution may have similar effects. Societal context may seem analogous to environmental constraints, but differs because societal forces occur at the institutional level, while environmental constraints can be addressed at the individual level.

Impact on the Criminal Justice Process

The predictive theory of intentions to exit street-level prostitution can benefit criminal justice entities working with prostituted women. Accurately predicting an exit from street-level prostitution could be used as supplemental documentation to seek reduction in sentencing for women arrested for prostitution. One barrier to exiting prostitution that is difficult to address is a criminal record; proponents of arrest feel it has a deterrent effect, and incarceration may act as a turning point that prompts some women to exit (Sanders, 2007). However, having felony charges makes finding legitimate employment very difficult and may trap women in prostitution (Norton-Hawk, 2001; Sanders et al., 2009). As an indicator of future behavior, this model could be used as evidence to reduce or avoid criminal charges among women arrested for prostitution, circumventing the difficulties a criminal record has on finding legitimate employment. As part of a customized probation plan, it can also be used to place a prostituted woman in an appropriate intervention or as a pre/post test to measure changes in attitudes, norms, and self-efficacy beliefs.

Addressing Child Custody Issues

It is not uncommon for prostituted women to lose custody of their children because of prostitution and/or drug addition (Arnold et al., 2000; Dalla, 2002). Losing custody of their children was like “hitting rock bottom” for some prostituted women, and being reunited with them was a motivation to leave prostitution (Dalla, 2002, 2006b; Hedin & Månsson, 2004). Predictions made by this theory could be used as part of a reunification assessment for ex-prostituted women seeking to regain custody of their children.

Limitations

As addressed by Fishbein and Ajzen (1975), application of this theory must be situated within the context of the population and the behavior. The theory presented here was primarily based on literature involving street prostituted women in Western culture and may not be generalizable to juveniles, male prostitutes, those engaged in indoor prostitution,or prostitution occurring other countries where social mores likely affect attitudes and behaviors differently than what has been presented. This limitation should not deter others from applying a similar model to these populations. Indeed, the flexibility to adapt a theory to a variety of populations and behaviors, while making it infinitely more complex, ultimately allows us to capture nuances and contextual factors that are important to understanding a phenomenon as complex as human behavior (Renzetti, 2006).