Abstract
For young adult sex workers, the risk of arrest and incarceration are dramatically influenced by the venue of sex exchange and individual and neighborhood characteristics. Using a unique venue-based survey sample of young adults in Detroit who are exchanging sex, multivariable logistic regression models were used to identify associations with arrest and incarceration. Criminal justice involvement was normative, and risk was increased by working on the street venue, using drugs, lacking stable housing, juvenile arrest or incarceration, dropping out of school before age 18, and neighborhood characteristics. Several promising points of intervention could reduce criminal justice involvement for young adults exchanging sex.
Keywords: Criminal justice involvement, Sex work, Drug use, Neighborhoods
1. Introduction
For young adults aged 18–30 involved in sex exchange, frequent engagement with the criminal justice system is common, and interactions between sex workers and police occur regularly. These interactions are most common for street-based sex workers, who sell sex in public spaces, but also occur across the broad range of venues in which sex or simulated sexual activity (such as stripping, erotic dancing) is exchanged for money, drugs, shelter, or other needed items. Police interactions can result in a summons or desk-appearance ticket, or being taken into police custody (Thukral & Ditmore, 2003; Thukral, Ditmore, & Murphy, 2005). Up to three-quarters of individuals exchanging sex have been arrested previously, and almost half were initially involved in the justice system before the age of 18 (Cohan et al., 2006; Raphael & Shapiro, 2002). Although estimates of the incarceration rate for sex workers vary, up to one-third have served sentences in either a jail or a prison, many more than once (Hankel, Heil, Dewey, & Martinez, 2015). For the purposes of the analyses and discussion presented here, “sex workers” will be used as an umbrella term including individuals who exchange any type of sex, or simulated sexual activity, such as erotic dancing or stripping.
Although criminal justice involvement is common for sex workers, differential experiences of arrest and incarceration can be conceptualized using a social ecological framework. Originally developed to describe the effects of community and environment on child development, this framework emphasizes the ways in which individual, sex exchange venue, neighborhood, and policy/policing environment factors all shape an individual’s risk for criminal justice involvement (Bronfenbrenner, 1977; Sallis, Owen, & Fisher, 2008).
As a result of decades of targeted economic and criminal justice policies, young, unemployed men of color bear a disproportionate burden of incarceration (Spohn & Holleran, 2000; B. Western, 2006). Young adults without a high school education and those with incarcerated parents are also significantly more likely to be involved in the criminal justice system (Harper & McLanahan, 2004; Bruce Western & Petit, 2010). Additionally, drug use results in more frequent and more serious criminal justice involvement and the relationship between sex work and drug use is complicated. Addiction may preclude many legal forms of employment, leaving sex work as one of few viable options for generating income. Additionally, some sex workers may use drugs as a means of coping with the trauma and conditions of that work (Cobbina & Oselin, 2011; Sallmann, 2010).
Chances for criminal justice involvement are not evenly distributed across venues for sexual commerce (e.g., the street or strip clubs), and sex workers are not evenly distributed by race, gender, and class across venues (Cohan et al., 2006; Murphy & Venkatesh, 2006). Sex workers who are racial and ethnic minorities, those who use drugs, and those who are among the youngest and oldest are more likely to be exchanging on the street rather than in brick-and-mortar venues (Cohan et al., 2006; Cunningham & Kendall, 2011; Murphy & Venkatesh, 2006; Thukral & Ditmore, 2003; Thukral et al., 2005). Individuals involved in street-based sexual commerce have the highest risk of arrest and incarceration due to punitive policies that target the selling of sex in outdoor public spaces and thereby increase the penalties for street-based sex exchange. Only those with sufficient resources can transition to private, indoor spaces where sexual services are marketed through online ads and websites reviewing escorts and call girls, and where the risk of arrest and incarceration is significantly lower (Bernstein, 2004; Cunningham & Kendall, 2011).
Neighborhood characteristics further determine patterns of policing, arrest, and incarceration for sex workers. Residential segregation and concentrated disadvantage additionally place young adults who are poor or are people of color at risk for more intensive policing and more punitive sentences, and neighborhoods known for drug trade are often under heightened surveillance (Ousey & Lee, 2008; Rodriguez, 2011).
In this study, the primary objective was to examine the risk and protective factors for arrest and incarceration among a diverse sample of young adults exchanging sex in Detroit. This was a secondary analysis of data from the Detroit Youth Passages project gathered to help understand how sex exchange contributes to economic and social vulnerability. The collateral consequences of criminal justice involvement are wide-ranging, including exclusion from state and federal health care and welfare programs, limited employment opportunities, and stigma from family, friends, and community (Pinard, 2006). The secondary objective of this study was, in collaboration with our community partners, to identify potential points of intervention to decrease criminal justice involvement among this group.
2. Materials and Methods
2.1. Study Design
The Detroit Youth Passages project is a collaboration between three community-based organizations (CBOs) in Detroit (Alternatives for Girls, Detroit Hispanic Development Corporation, and Ruth Ellis Center) and the University of Michigan School of Public Health. The project included an ethnographic study using participant observation, semi-structured and life-history interviews with young people identified through partner organizations, a photovoice project, and a venue-based, interviewer-administered survey that was designed to broadly explore the ways in which engaging in sex exchange contributed to sexual vulnerabilities inherent in the contexts navigated by young adults of diverse genders, sexualities, and social positions (Graham et al., 2013; Lopez et al., 2012; Padilla, 2012). The work described here represents a secondary analysis of the survey data. The project was specifically targeted towards young adults who were uniquely situated with respect to gender and sexual identity and socioeconomic status, including cisgender and transgender women and gay and bisexual men with histories of economic and residential instability.
2.2. Sampling
A venue-based sample was chosen principally because it operationalized the concept of sexual geographies that, in part, framed the project. Venue-based sampling helped identify these relatively isolated, small, distinct target populations of interest (Kendall et al., 2008; MacKellar et al., 2007). Due to the challenges of recruitment in venues heavily regulated by managers, and with the goal of ensuring some degree of representation across a wide range of characteristics, snowball sampling was also employed within the venues, and participants were encouraged to refer contacts to the study. Prior descriptions of participant recruitment strategies support that venue-based and snowball sampling frameworks may allow for increased sampling of lower socioeconomic status groups and others that are missed by traditional study samples (Kendall et al., 2008).
2.3. Recruitment
The study team identified venues that would allow survey sampling of cisgender and transgender young adults who were representative of the racial and socio-economic diversity in the city and also were engaged in sex exchange at the time of the survey. The study team included former sex workers and outreach workers with experience working with our target groups in Detroit and were instrumental in defining the venues and subsequently categorizing establishments and locations in the city. The three initially identified venues were the street, after-hours parties and social clubs, and strip clubs. The strip clubs were further divided into high-end and low-end strip clubs, with high-end clubs having more security and also more visible security, more space between the performers and clients, and a lower client-to-performer ratio. Low-end strip clubs had less security and less visible security, more contact between performers and clients, and a higher client-to-performer ratio. Consistent with the eligibility criteria for services at our partner CBOs, we defined “young adults” as people 18–30 years old.
After identifying the venues, select members of the research team made preliminary visits to each one. They recorded extensive field notes that detailed the time and space dynamics of the flow of customers and workers, the availability of space for interviews, and the demographics of the potential participants. During these initial visits, the research team members approached the owners or proprietors of the venue (excepting the street) and described the broad scope of the study. On selected dates and times that were representative of the venue, research team members conducted a convenience sample of interviews. They attempted to recruit participants at slow times, when the $30 incentive for a 30-minute interview would augment, rather than cut into, earnings (Snow et al., 2013). Interviews were generally conducted in a quiet or private space, where participants felt most comfortable. They were also offered the opportunity to have a private meeting outside of the venue if they preferred. Interviewers recruited participants on the street venue from blocks that members of the study team had identified as having a high density of street-based sex work. These respondents were offered the option of conducting the interview in a car or restaurant at the time of recruitment, or to set up a private meeting at another time. The details of the street-based recruitment process are detailed elsewhere (Snow et al., 2013).
Trained interviewers used paper surveys to conduct interviews with 278 respondents from May, 2012 to August, 2012. Trained research assistants then entered the survey responses into a secure database for data cleaning and analysis. The data entry was double checked for coding errors, and then the entire dataset was reviewed by a third individual for consistency.
2.4. Measures
Several measures of criminal justice involvement were used as the dependent variables in this study. Interviewers asked respondents whether, before or after the age of 18, they had experienced arrest, incarceration in a juvenile detention center (before age 18 only), adult jail, adult prison, or immigration detention center, and whether they had been deported. Only two respondents indicated that they had been held in a detention center or deported, so no further analyses were conducted using these variables. Arrest and jail experiences before and after age 18 were selected as the outcomes of interest because they were the most common in this population. These responses were used first as dependent variables, and then juvenile arrest and jail experience were used as independent variables in models of adult criminal justice involvement.
Other independent variables from our social ecological model of risk for criminal justice involvement were operationalized as demographic measures, patterns of sexual exchange, types of drug use, neighborhood characteristics, and stress about police. Race was assessed with a single question asking respondents to select all applicable racial and ethnic identities. A single variable was then created with four categories: 1) White/non-Latino/a; 2) Latino/a; 3) Black/African American; and 4) Mixed race or other. Individuals were assigned to a single category with the attempt to place the individual in the category that an outsider, such as a police officer, might place them. Individuals who indicated that they were Black/African American were placed in that category (irrespective of other racial or ethnic identities). Then, individuals who indicated that they were Latino/a were placed in that category, but only if they had not indicated that they were also Black/African American. Individuals who selected White/non-Latino only were placed in that category, and individuals who selected that they were of mixed race or “other” race, but did not indicate that they were any of the other racial or ethnic categories were grouped together. Sex and gender identity were measured with two questions, the first asking for sex assigned at birth, and the second regarding transgender identity.
Age was measured in years and coded continuously. The highest level of education attained was categorized as “Some high school,” “High school diploma or GED,” “Some college,” and “Bachelor’s degree or more.” Respondents were asked whether they had access to reliable transportation, with response options of “Never,” “Rarely,” “Most of the time,” and “All of the time.”
Current patterns of sexual exchange were measured in several ways. Interviewers coded the venue where a participant exchanged sex based on the recruitment location: on the street; in a low-end strip club; in a high-end strip club; or in an after-hours party/social club. They asked respondents whether they had exchanged sex or simulated sexual activity, for cash or money, gang membership, drugs or alcohol, food, clothes, a cell phone, somewhere to stay, items for friends, children, or family, items in jail, transportation, home or car repair, anything else they could not afford, or help to stay in the country. Very few people indicated exchanging for gang membership, home or car repair, immigration help, or items in jail, so these were not included in the analysis. Interviewers also asked respondents about the age at which they first exchanged sex as well as the age at which they began regularly engaging in sex exchange. In addition, they reported the frequency of sex exchange as the number of exchanges in the past month, which was then categorized as “none,” “less than the 50th percentile (1–6 exchanges),” “between the 50th and 90th percentile (7–30 exchanges),” or “above the 90th percentile (>30 exchanges).” Interviewers assessed whether respondents worked for a pimp, madam, or other managerial figure with the question, “Does anyone oversee your [sex exchange] work?”
The research team constructed a neighborhood attribute scale by summing the responses to a series of ten statements about neighborhood resources. Interviewers asked respondents to agree, disagree, or neither agree nor disagree with statements about neighborhood attributes (e.g., “In my neighborhood it is easy to get to know people,” “There is enough street lighting in my neighborhood,” and, “Most buildings are in good condition in my neighborhood”). Based on prior definitions of residential instability, respondents were considered to experience residential instability if they lived in a shelter, hotel, or transitional housing at the time, were homeless, or had lived in three or more places in the past year (Cotton & Schwartz-Barcott, 2016). Housing insecurity was measured with a single question, “Are you ever worried about not having a place to stay?”
Respondents indicated whether they currently or had ever used: alcohol; marijuana; crack (free base form of cocaine that can be smoked); cocaine (other than crack); heroin; methamphetamine (“meth”), speed, bennies, uppers or other stimulants; ecstasy, LSD, PCP, mushrooms or other hallucinogens; paint thinners, glues, gas, or other inhalants; non-prescribed use of prescription drugs (e.g., Oxycontin, Xanax, hormones). Prior studies of young adult drug use divide alcohol and marijuana use, which are more common, from “hard drugs,” including heroin, crack, and cocaine (Faggiano, Minozzi, Versino, & Buscemi, 2014; Grant et al., 2016). In addition to individual variables for each drug, two composite measures of drug use were constructed: daily use of any substance, including alcohol and marijuana, and daily use of any substance, excepting alcohol and marijuana, both of which were legal for some portion of the study sample in the setting of a widely contested state-wide legalization of medical marijuana (Brush, 2013).
The respondents’ stress levels about “seeing police” were measured as a single item with a three point scale from “not stressful,” to “very stressful.” This item was only included on the questionnaires used in the street venue and the high- and low-end strip clubs. A relative lack of police presence in the social clubs/after-hours was noted in the preliminary visits and corroborated by the study team members who were familiar with the local scene. Responses for the social clubs/after-hours clubs were coded as missing.
2.5. Statistical analysis
Descriptive statistics for involvement in the criminal justice system were calculated across venues and demographic variables. Given the significant variation by racial and gender identity across the venues, the decision was made to include venue in all analyses, even for what would conventionally be calculated as bivariate relationships. Bivariate relationships adjusted for venue were calculated using logistic regression on each of the measures of criminal justice involvement with the measures of demographics, patterns of sexual exchange, drug use, neighborhood characteristics, and stress about police as the independent variables. Next, multivariable logistic regression models were constructed including venue, racial identity, and gender identity a priori, as well as all variables that showed statistically significant relationships in the bivariate analyses. The study was not adequately powered to test interactions between different variables from our conceptual framework.
3. Results
3.1. Description of the sample
Table 1 shows the demographic make-up of the sample (n = 278). Respondents in the two strip club venues were nearly entirely cisgender women. Except in the high-end strip clubs, each venue had majority Black/African American respondents; the high end strip club respondents were more evenly divided between White/Caucasian and Black/African American respondents. The average age of the sample was 24, and this was similar across venues. Nearly two-thirds of the sample had completed high school or a GED, and seventy percent had work other than sex exchange.
Table 1.
Demographic description of the study sample by venue
| Street | Low-End Strip Clubs | High-End Strip Clubs | Social/Afterhours Clubs | Total* | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Mean | (Range) | Mean | (Range) | Mean | (Range) | Mean | (Range) | Mean | (Range) | |
| Age | 24.8 | (18–30) | 23.6 | (18–30) | 23.9 | (19–29) | 23.5 | (18–30) | 24 | (18–30) |
| N | (%) | N | (%) | N | (%) | N | (%) | N | (%) | |
| Gender Identity | ||||||||||
| Transgender Woman | 35 | (34.0) | 2 | (4.8) | 0 | ( − ) | 36 | (34.6) | 73 | (26.3) |
| Cisgender Man | 32 | (31.1) | 9 | (21.4) | 0 | ( − ) | 33 | (31.7) | 74 | (26.6) |
| Cisgender Woman | 36 | (35.0) | 31 | (73.8) | 29 | (100.0) | 35 | (33.7) | 131 | (47.1) |
| Racial Identity | ||||||||||
| Black | 67 | (65.0) | 33 | (78.6) | 11 | (37.9) | 55 | (52.9) | 166 | (59.7) |
| White | 19 | (18.4) | 1 | (2.4) | 12 | (41.4) | 22 | (21.2) | 54 | (19.4) |
| Hispanic | 6 | (5.8) | 1 | (2.4) | 1 | (3.4) | 11 | (10.6) | 19 | (6.8) |
| Mixed/Asian/Other | 11 | (10.7) | 7 | (16.7) | 5 | (17.2) | 16 | (15.4) | 39 | (14.0) |
| Highest level of education achieved | ||||||||||
| Some high school | 30 | (29.4) | 7 | (16.7) | 6 | (20.7) | 16 | (15.4) | 59 | (21.3) |
| High school GED | 53 | (52.0) | 11 | (26.2) | 10 | (34.5) | 38 | (36.5) | 112 | (40.4) |
| Some college | 16 | (15.7) | 21 | (50.0) | 12 | (41.4) | 37 | (35.6) | 86 | (31.0) |
| Bachelors degree or more | 3 | (2.9) | 3 | (7.1) | 1 | (3.4) | 13 | (12.5) | 20 | (7.2) |
| Has other work (apart from sex exchange) | ||||||||||
| No | 60 | (58.3) | 1 | (2.4) | 1 | (3.4) | 19 | (18.3) | 81 | (29.2) |
| Yes | 43 | (41.7) | 40 | (97.6) | 28 | (96.6) | 85 | (81.7) | 196 | (70.8) |
| Worried about not having a place to stay | ||||||||||
| No | 43 | (46.2) | 28 | (68.3) | 21 | (72.4) | 61 | (67.8) | 153 | (60.5) |
| Yes | 50 | (53.8) | 13 | (31.7) | 8 | (27.6) | 29 | (32.2) | 100 | (39.5) |
| Access to reliable transportation | ||||||||||
| Never | 10 | (9.8) | 2 | (4.8) | 1 | (3.4) | 9 | (8.7) | 22 | (7.9) |
| Some of the time | 60 | (58.8) | 8 | (19.0) | 7 | (24.1) | 31 | (29.8) | 106 | (38.3) |
| All of the time | 32 | (31.4) | 32 | (76.2) | 21 | (72.4) | 64 | (61.5) | 149 | (53.8) |
| Total* | 103 | 42 | 29 | 104 | 278 | |||||
Abbreviations: N – number; % - Percentage
Individual blocks may not sum to totals if there was missing data
The majority of the sample had experienced involvement with the criminal justice system, as demonstrated in Table 2. Across all venues, 30–50% of respondents had been arrested before the age of 18, and approximately 25% of respondents had been incarcerated in either a juvenile or adult facility before age 18. After age 18, rates of arrest ranged from 40% in the after-hours/social club venue to just over 60% on the street. A slightly smaller proportion of respondents had been incarcerated in an adult jail facility. Only the street venue included a substantial percentage of respondents (15%) who had been incarcerated in an adult prison facility.
Table 2.
Involvement in the criminal justice system by venue
| Street | Low-End Strip Clubs | High-End Strip Clubs | Social/ Afterhours Clubs | Total* | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| N | (%) | N | (%) | N | (%) | N | (%) | N | (%) | |
| Criminal justice involvement BEFORE age 18 | ||||||||||
| Arrested | 37 | 37.4) | 13 | (33.3) | 14 | (50.0) | 30 | (29.4) | 94 | (35.1) |
| Juvenile detention | 27 | 27.0) | 9 | (22.0) | 9 | (32.1) | 20 | (20.0) | 65 | (24.2) |
| Adult jail facility | 11 | (11.5) | 4 | (9.8) | 3 | (10.3) | 4 | (4.0) | 22 | (8.3) |
| Adult prison facility | 3 | (3.1) | 1 | (2.6) | 1 | (3.4) | 3 | (3.0) | 8 | (3.0) |
| Criminal justice involvement AFTER age 18 | ||||||||||
| Arrested | 60 | (61.9) | 22 | (55.0) | 15 | (55.6) | 41 | (40.6) | 138 | (52.1) |
| Adult jail facility | 56 | (57.7) | 18 | (42.9) | 13 | (46.4) | 37 | (36.3) | 124 | (46.1) |
| Adult prison facility | 15 | (15.2) | 1 | (2.4) | 0 | ( − ) | 5 | (5.0) | 21 | (7.8) |
| Total* | 103 | 42 | 29 | 104 | 278 | |||||
Abbreviations: N – number; % - Percentage
Individual blocks may not sum to totals if there was missing data
Patterns of sexual exchange are shown in Table 3. Approximately 140 respondents reported exchanging oral, vaginal, or anal sex, and over one-third of the respondents did not answer these questions (data not shown). The average number of exchanges in the past month was 15.8, with the most exchanges taking place in the street venue. The average age of first exchange was around 18, although some respondents reported exchanging sex as minors. Most respondents had exchanged for cash, and a substantial minority had exchanged for either food, clothing, phone, drugs or alcohol, a place to stay, items for friends, children, or family, transportation, or something else that was not listed. The proportion of respondents who endorsed someone managing or overseeing their sex exchange work ranged from none in the after-hours/social club venue to half in the high-end strip club venue.
Table 3.
Patterns of sexual exchange by venue
| Street | Low End Strip Clubs | High End Strip Clubs | Social/Afterhours Clubs | Total* | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Mean | (Range) | Mean | (Range) | Mean | (Range) | Mean | (Range) | Mean | (Range) | |
| Number of exchanges in the past month | 26.4 | (0–208) | 12.3 | (0–30) | 11.8 | (0–24) | 7.1 | (0–55) | 15.8 | (0–208) |
| Age at first exchange | 18.3 | (11–30) | 19.3 | (6–27) | 17.9 | (15–29) | 17.9 | (6–30) | 18.5 | (6–30) |
| N | (%) | N | (%) | N | (%) | N | (%) | N | (%) | |
| Items exchanged | ||||||||||
| Cash | 97 | (94.2) | 41 | (97.6) | 28 | (96.6) | 85 | (81.7) | 251 | (90.3) |
| Food, clothing, phone | 41 | (39.8) | 11 | (26.2) | 10 | (34.5) | 27 | (26.0) | 89 | (32.0) |
| Drugs or alcohol | 51 | (49.5) | 12 | (28.6) | 9 | (31.0) | 27 | (26.0) | 99 | (35.6) |
| Place to stay | 41 | (39.8) | 10 | (23.8) | 2 | (6.9) | 21 | (20.2) | 74 | (26.6) |
| Items for friends, children, or family | 31 | (30.1) | 8 | (19.0) | 5 | (17.2) | 9 | (8.7) | 53 | (19.1) |
| Transportation | 30 | (29.1) | 7 | (16.7) | 4 | (13.8) | 16 | (15.4) | 57 | (20.5) |
| Something else | 53 | (51.5) | 13 | (31.0) | 10 | (34.5) | 40 | (38.5) | 116 | (41.7) |
| Anyone oversees your work? | ||||||||||
| No | 77 | (93.9) | 28 | (82.4) | 12 | (50.0) | 19 | (100.0) | 136 | (85.5) |
| Yes | 5 | (6.1) | 6 | (17.6) | 12 | (50.0) | 0 | ( − ) | 23 | (14.5) |
| Total* | 103 | (100.0) | 42 | (100.0) | 29 | (100.0) | 104 | (100.0) | 278 | (100.0) |
Abbreviations: N – number; % - Percentage
Individual blocks may not sum to totals if there was missing data
Drug use among respondents is shown in Table 4. The most commonly used substances were alcohol and marijuana, with nearly three-quarters of respondents in the low-end strip clubs, high-end strip clubs, and after-hours/social clubs endorsing current alcohol use, compared with slightly less than half of respondents in the street venue. Across all venues, approximately 9% of respondents reported daily use of drugs other than alcohol and marijuana, with the highest proportion being in the street venue (13%). Of the small number of respondents to a question about drug and alcohol use in the context of sex exchange, the majority answered that they prefer to be “a little drunk/high.”
Table 4.
Drug use by venue
| Street | Low-End Strip Club | High-End Strip Clubs | Social/Afterhours Clubs | Total* | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| N | (%) | N | (%) | N | (%) | N | (%) | N | (%) | |
| Current drug use | ||||||||||
| Alcohol | 49 | (48.0) | 30 | (71.4) | 22 | (75.9) | 78 | (75.7) | 179 | (64.9) |
| Marijuana | 54 | (52.9) | 27 | (64.3) | 13 | (44.8) | 52 | (50.0) | 146 | (52.7) |
| Crack | 13 | (12.7) | 0 | ( − ) | 1 | (3.6) | 2 | (1.9) | 16 | (5.8) |
| Cocaine | 11 | (11.1) | 0 | ( − ) | 3 | (10.3) | 8 | (7.7) | 22 | (8.1) |
| Heroin | 8 | (7.9) | 1 | (2.4) | 1 | (3.6) | 1 | (1.0) | 11 | (4.0) |
| Meth, speed, other uppers | 2 | (2.0) | 1 | (2.4) | 1 | (3.6) | 4 | (3.8) | 8 | (2.9) |
| Ecstasy, LSD, PCP, mushroom, other hallucinogen | 6 | (5.9) | 2 | (4.8) | 3 | (10.3) | 11 | (10.6) | 22 | (7.9) |
| Inhalents | 0 | ( − ) | 0 | ( − ) | 2 | (7.1) | 0 | ( − ) | 2 | (0.7) |
| Non-medical prescription drugs | 5 | (4.9) | 5 | (11.9) | 3 | (10.3) | 12 | (11.7) | 25 | (9.1) |
| Daily use of drugs other than alcohol or marijuana | 0 | |||||||||
| No | 89 | (86.4) | 40 | (95.2) | 28 | (96.6) | 96 | (92.3) | 253 | (91.0) |
| Yes | 14 | (13.6) | 2 | (4.8) | 1 | (3.4) | 8 | (7.7) | 25 | (9.0) |
| During sex exchange, prefer to be… | ||||||||||
| Straight/Sober | 32 | (41.0) | 11 | (32.3) | 11 | (47.8) | 6 | (33.3) | 60 | (39.2) |
| A little drunk/high | 35 | (44.8) | 20 | (58.8) | 9 | (39.1) | 6 | (33.3) | 70 | (47.5) |
| Really drunk/high | 11 | (14.1) | 3 | (8.8) | 3 | (13.0) | 6 | (33.3) | 23 | (15.0) |
| Total* | 103 | 42 | 29 | 104 | 278 | |||||
Abbreviations: N – number; % - Percentage
Individual blocks may not sum to totals if there was missing data
Table 5 shows the distribution of residential instability, neighborhood attributes, access to reliable transportation, and stress about police. The average number of residences in the past year was 3.5 (range 0–50) and this varied from an average of 1.9 in the low-end strip club venue to 5.3 on the street venue. Only 20% of respondents said that their neighborhood was dangerous during the day (ranging from 3% in the high-end strip clubs to 32% on the street), but this increased to just over 40% at night (range 21%−56%) (data not shown). On the street venue, half of respondents experienced residential instability, which was nearly twice as frequent as in the other venues, and about the same proportions endorsed being worried about having a place to stay. Nearly half of the entire sample had inconsistent access to reliable transportation. Stress about seeing the police was highest on the street venue, with 55% rating it “very stressful.” The majority of respondents in other venues rated seeing the police as “not stressful.”
Table 5.
Residential instability by venue
| Street | Low-End Strip Clubs | High-End Strip Clubs | Social/Afterhours Clubs | Total* | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Mean | (Range) | Mean | (Range) | Mean | (Range) | Mean | (Range) | Mean | (Range) | |
| Number of residences in the past year | 5.3 | (1–50) | 1.9 | (1–7) | 2.7 | (1–20) | 2.6 | (0–50) | 3.5 | (0–50) |
| Number of positive neighborhood attributes | 4 | (0–8) | 3.8 | (1–8) | 3.5 | (2–7) | 4.3 | (0–9) | 4 | (0–9) |
| N | (%) | N | (%) | N | (%) | N | (%) | N | (%) | |
| Residential instability | 52 | (51.0) | 8 | (19.0) | 9 | (31.0) | 26 | (25.2) | 95 | (34.4) |
| Ever worried about not having a place to stay | 50 | (53.8) | 13 | (31.7) | 8 | (27.6) | 29 | (32.2) | 100 | (39.5) |
| Access to reliable transportation | ||||||||||
| Never | 10 | (9.8) | 2 | (4.8) | 1 | (3.4) | 9 | (8.7) | 22 | (7.9) |
| Some of the time | 60 | (58.8) | 8 | (19.0) | 7 | (24.1) | 31 | (29.8) | 106 | (38.3) |
| All of the time | 32 | (31.4) | 32 | (76.2) | 21 | (72.4) | 64 | (61.5) | 149 | (53.8) |
| Stress of seeing police ǂ | ||||||||||
| Not stressful | 20 | (24.7) | 24 | (72.7) | 18 | (75.0) | - | - | 62 | (44.9) |
| Somewhat stressful | 16 | (19.8) | 4 | (12.1) | 3 | (12.5) | - | - | 23 | (16.7) |
| Very stressful | 45 | (55.6) | 5 | (15.2) | 3 | (12.5) | - | - | 53 | (38.4) |
| Total* | 103 | 42 | 29 | 104 | 278 | |||||
Abbreviations: N – number; % - Percentage
Individual blocks may not sum to totals if there was missing data
Not measured in the social/afterhours club venue
3.2. Arrest before age 18
Significant bivariate relationships with arrest before the age of 18, adjusted for venue, were observed for working in a high-end strip club, having less than a high school diploma or GED, being aged 10–14 at the time of first sexual exchange, less frequent sexual exchange than the 50th percentile, and being very stressed about seeing police (Table 6, model 1). Once these variables were placed simultaneously into a multivariable model, only working on the street venue and having less than a high school diploma or GED remained significant (Table 6).
Table 6.
Bivariate and multivariable models of arrest before age 18
| Arrest BEFORE age 18 | (1) | (2) | (3) | (4) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | ||||||||
| Venue | |||||||||||||||
| Street | 1.432 | [0.80 - | 2.58] | 1.25 | [0.67 - | 2.31] | 1.10 | [0.55 - | 2.22] | 0.15 | * | [0.03 - | 0.70] | ||
| Low-end Strip Club | 1.200 | [0.54 - | 2.65] | 1.21 | [0.54 - | 2.71] | 0.94 | [0.37 - | 2.40] | 0.25 | [0.06 - | 1.12] | |||
| High-end Strip Club | 2.400 | [1.02 - | 5.64] | 2.33 | [0.98 - | 5.57] | 3.76 | * | [1.28 - | 11.06] | 1.00 | [1.00 - | 1.00] | ||
| Level of education | |||||||||||||||
| Some high school | 2.645 | [1.36 - | 5.15] | 2.65 | ** | [1.36 - | 5.15] | 2.07 | [0.97 - | 4.43] | 3.75 | * | [1.19 - | 11.77] | |
| Any post-high school education | 1.007 | [0.54 - | 1.86] | 1.01 | [0.54 - | 1.86] | 1.06 | [0.52 - | 2.15] | 1.43 | [0.40 - | 5.08] | |||
| Age at first exchange | |||||||||||||||
| <9 | 1 | [1.00 - | 1.00] | 1.00 | [1.00 - | 1.00] | 1.00 | [1.00 - | 1.00] | ||||||
| 10–14 | 3.923 | [1.15 - | 13.34] | 3.71 | * | [1.03 - | 13.37] | 1.00 | [1.00 - | 1.00] | |||||
| 15–18 | 1.239 | [0.68 - | 2.27] | 1.49 | [0.77 - | 2.89] | 2.03 | [0.73 - | 5.69] | ||||||
| >25 | 0.628 | [0.18 - | 2.25] | 0.96 | [0.24 - | 3.76] | 0.78 | [0.14 - | 4.41] | ||||||
| Number of exchanges per month | |||||||||||||||
| 1–6 | 0.213 | [0.05 - | 1.03] | 0.30 | [0.06 - | 1.52] | 2.03 | [0.21 - | 19.30] | ||||||
| 7–30 | 1.631 | [0.79 - | 3.35] | 1.60 | [0.72 - | 3.53] | 1.39 | [0.44 - | 4.45] | ||||||
| >30 | 0.958 | [0.49 - | 1.86] | 1.09 | [0.52 - | 2.28] | 3.22 | [0.88 - | 11.83] | ||||||
| Stressfulness of encounters with police | |||||||||||||||
| Somewhat stressful | 2.335 | [0.82 - | 6.67] | 1.72 | [0.40 - | 7.45] | |||||||||
| Very stressful | 2.645 | [1.05 - | 6.66] | 2.42 | [0.70 - | 8.37] | |||||||||
| N | 268 | 223 | 100 | ||||||||||||
Exponentiated coefficients; 95% confidence intervals in brackets
p<0.05
p<0.01
p<0.001
3.3. Jail before age 18
Having less than a high school diploma or GED, exchanging sex for the first time at age 10–14, exchanging for transportation, being worried about a place to stay, and ever using crack were significantly associated with going to an adult jail before the age of 18 in bivariate analyses (Table 7, model 1). Only having less than a high school diploma or GED and ever using crack remained significant in the multivariable models, as shown in Table 7, models 2–4.
Table 7.
Bivariate and multivariable models of going to an adult jail before age 18
| Jail BEFORE age 18 | (1) | (2) | (3) | (4) | (5) | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | ||||||||||
| Venue (Ref = Afterhours/Social Club) | |||||||||||||||||||
| Street | 3.12 | [0.95 – 10.12] | 2.63 | [0.77 – 9.04] | 1.74 | [0.46 – 6.54] | 2.10 | [0.48 – 9.11] | 1.87 | [0.42 – 8.43] | |||||||||
| Low-end Strip Club | 2.60 | [0.62 – 10.92] | 2.81 | [0.64 – 12.24] | 2.49 | [0.54 – 11.55] | 2.53 | [0.42 – 15.20] | 2.83 | [0.46 – 17.35] | |||||||||
| High-end Strip Club | 2.77 | [0.58 – 13.16] | 2.86 | [0.57 – 14.24] | 3.09 | [0.59 – 16.09] | 4.14 | [0.69 – 24.70] | 4.08 | [0.66 – 25.22] | |||||||||
| Level of education (Ref = High school/GED) | |||||||||||||||||||
| Some high school | 2.97 | [1.05 – 8.36] | 2.97 | * | [1.05 – 8.36] | 3.86 | * | [1.23 – 12.08] | 3.38 | * | [1.03 – 11.05] | 3.73 | * | [1.09 – 12.77] | |||||
| Some college | 0.55 | [0.13 – 2.28] | 0.55 | [0.13 – 2.28] | 0.53 | [0.11 – 2.43] | 0.48 | [0.08 – 2.80] | 0.55 | [0.09 – 3.38] | |||||||||
| Bachelors degree or more | 2.18 | [0.39 – 12.07] | 2.18 | [0.39 – 12.07] | 2.39 | [0.38 – 15.14] | 2.59 | [0.37 – 18.31] | 3.65 | [0.49 – 27.37] | |||||||||
| Age at first exchange (Ref = 18–25) | |||||||||||||||||||
| <9 | 1 | [1.00 – 1.00] | 1.00 | [1.00 – 1.00] | 1.00 | [1.00 – 1.00] | 1.00 | [1.00 – 1.00] | |||||||||||
| 10–14 | 4.82 | [0.95 – 24.46] | 3.96 | [0.71 – 21.93] | 4.12 | [0.54 – 31.61] | 5.40 | [0.69 – 42.33] | |||||||||||
| 15–18 | 1.48 | [0.48 – 4.54] | 1.31 | [0.41 – 4.22] | 1.52 | [0.43 – 5.41] | 1.83 | [0.48 – 6.98] | |||||||||||
| >25 | 1.79 | [0.30 – 10.51] | 2.19 | [0.31 – 15.26] | 1.85 | [0.23 – 14.76] | 2.09 | [0.25 – 17.41] | |||||||||||
| Exchange for transportation | 2.87 | [1.13 – 7.264] | 3.13 | * | [1.08 – 9.10] | 3.14 | * | [1.01 – 9.81] | 2.86 | [0.88 – 9.30] | |||||||||
| Worried about a place to stay | 2.89 | [1.08 – 7.78] | 2.21 | [0.75 – 6.55] | 1.97 | [0.65 – 6.00] | |||||||||||||
| Ever used crack | 3.45 | [1.19 – 10.02] | 3.88 | [0.99 – 15.20] | |||||||||||||||
| N | 265 | 237 | 216 | 213 | |||||||||||||||
Abbreviations: OR – odds ratio; CI – confidence interval; Ref – reference
p<0.05
p<0.01
p<0.001
3.4. Arrest after age 18
Significant bivariate positive relationships with arrest after age 18 are shown as model 1 in Table 8. Once multivariable models were constructed, significant positive associations with arrest after age 18 were with working on the street venue, increasing age, having more positive neighborhood attributes, worry about a place to stay, and daily use of drugs other than alcohol and marijuana (Table 8, models 2–7). When neighborhood attributes, worry about housing, and drug use were included in the same model, only drug use remained a significant independent variable, and worry about housing neared significance (Table 8, model 8).
Table 8.
Bivariate and multivariable models of arrest after age 18
| Arrest AFTER age 18 | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | ||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | |||||||||||||||||
| Venue (Ref = Afterhours/Social Club) | ||||||||||||||||||||||||||||||||
| Street | 2.37 | [1.34 – 4.20] | 2.03 | * | [1.11 – 3.71] | 1.99 | * | [1.06 – 3.73] | 2.14 | * | [1.12 – 4.06] | 1.89 | [0.94 – 3.81] | 1.70 | [0.32 – 9.10] | 1.92 | [1.00 – 3.69] | 1.91 | [0.93 – 3.95] | |||||||||||||
| Low-end Strip Club | 1.79 | [0.85 – 3.74] | 1.58 | [0.71 – 3.53] | 1.25 | [0.52 – 2.98] | 1.41 | [0.57 – 3.45] | 1.48 | [0.58 – 3.74] | 2.12 | [0.51 – 8.87] | 1.62 | [0.66 – 3.99] | 1.87 | [0.71 – 4.92] | ||||||||||||||||
| High-end Strip Club | 1.83 | [0.78 – 4.31] | 1.36 | [0.52 – 3.56] | 0.56 | [0.18 – 1.69] | 0.67 | [0.21 – 2.07] | 0.63 | [0.20 – 2.02] | 1.00 | [1.00 – 1.00] | 0.87 | [0.28 – 2.73] | 0.94 | [0.28 – 3.11] | ||||||||||||||||
| Gender (Ref = Cisgender Women) | ||||||||||||||||||||||||||||||||
| Transgender Women | 0.82 | [0.43 – 1.60] | 0.88 | [0.44 – 1.76] | 0.84 | [0.41 – 1.72] | 0.82 | [0.39 – 1.69] | 1.04 | [0.47 – 2.32] | 1.26 | [0.25 – 6.35] | 1.16 | [0.54 – 2.47] | 1.28 | [0.55 – 3.00] | ||||||||||||||||
| Cisgender Men | 0.52 | [0.27 – 0.99] | 0.52 | [0.26 – 1.02] | 0.57 | [0.27 – 1.18] | 0.51 | [0.24 – 1.09] | 0.53 | [0.24 – 1.17] | 0.50 | [0.10 – 2.38] | 0.74 | [0.34 – 1.59] | 0.62 | [0.27 – 1.44] | ||||||||||||||||
| Age | 1.16 | [1.08 – 1.25] | 1.17 | *** | [1.08 – 1.26] | 1.18 | *** | [1.09 – 1.28] | 1.21 | *** | [1.11 – 1.32] | 1.21 | *** | [1.10 – 1.33] | 1.23 | ** | [1.06 – 1.42] | 1.19 | *** | [1.09 – 1.29] | 1.23 | *** | [1.12 – 1.36] | |||||||||
| Number of exchanges per month (Ref = 1–6) | ||||||||||||||||||||||||||||||||
| 0 | 1.02 | [0.35 – 2.99] | 1.00 | [0.32 – 3.11] | 0.94 | [0.30 – 2.96] | 0.99 | [0.30 – 3.25] | 1.43 | [0.15 – 13.43] | 0.91 | [0.28 – 2.89] | 0.86 | [0.25 – 2.90] | ||||||||||||||||||
| 7–30 | 2.93 | [1.41 – 6.12] | 2.43* | [1.11 – 5.31] | 2.16 | [0.97 – 4.80] | 1.89 | [0.81 – 4.39] | 1.62 | [0.50 – 5.31] | 1.94 | [0.86 – 4.36] | 1.54 | [0.64 – 3.67] | ||||||||||||||||||
| >30 | 1.45 | [0.76 – 2.75] | 0.92 | [0.45 – 1.87] | 0.81 | [0.39 – 1.68] | 0.92 | [0.43 – 2.00] | 6.71 | * | [1.01 – 44.56] | 0.81 | [0.39 – 1.69] | 0.76 | [0.34 – .68] | |||||||||||||||||
| Additional positive neighborhood traits | 1.13 | [1.00 – 1.27] | 1.21 | ** | [1.05 – 1.39] | 1.12 | [0.96 – 1.31] | |||||||||||||||||||||||||
| Worried about a place to stay | 2.10 | [1.21 – 3.65] | 2.55 | ** | [1.36 – 4.78] | 1.89 | [0.97 – 3.66] | |||||||||||||||||||||||||
| Level of stress about police (Ref = Not stressful) | ||||||||||||||||||||||||||||||||
| Somewhat stressful | 2.19 | [0.74 – 6.53] | 2.21 | [0.55 – 8.85] | ||||||||||||||||||||||||||||
| Very stressful | 3.20 | [1.25 – 8.16] | 2.62 | [0.85 – 8.07] | ||||||||||||||||||||||||||||
| Daily use of drugs other than alcohol or marijuana | 26.60 | [3.51 – 201.7] | 23.21 | ** | [2.90 – 185.53] | 11.30 | * | [1.33 – 95.79] | ||||||||||||||||||||||||
| N | 261 | 244 | 244 | 221 | 111 | 244 | 221 | |||||||||||||||||||||||||
Abbreviations: OR – odds ratio; CI – confidence interval; Ref – reference
p<0.05
p<0.01
p<0.001
In order to more clearly interpret the effect of neighborhood attributes, we conducted several post-hoc analyses. First, the overall scale was divided into three parts: physical resources (street lights, garbage collection, parks, and buildings in good condition), social resources (easy to get to know people, residents care for property, like the neighborhood, and identify with people in the neighborhood), and dangerousness (dangerous during the day and at night). Each of these smaller subscales (increasing physical resources, increasing social resources, and increasing dangerousness of the neighborhood) was found to increase the risk of adult arrest and incarceration (data not shown).
3.5. Jail after age 18
There were many significant bivariate relationships with jail after age 18, shown in Table 9, model 1. In the multivariable models, only age, arrest prior to age 18, daily drug use, and stress about seeing police remained strongly associated with jail after age 18 (Table 9, models 2–7). The variables indicating ever using a single drug lost significance once demographic variables were controlled for, and these were not included in the final models (data not shown). Of note, because stress about police was only asked in a subset of the venues (street, low- and high-end strip clubs), any many respondents did not answer that question, the sample size was significantly reduced when this was included in the models. When stress about police and daily drug use were included together in the multivariable models, the reduced sample size resulted in perfect correlation with incarceration by daily drug use, so an odds ratio could not be calculated (Table 9, model 7).
Table 9.
Bivariate and multivariable models of going to an adult jail after age 18
| Jail AFTER age 18 | (1) | (2) | (3) | (4) | (5) | (6) | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | ||||||||||||
| Venue (Ref = Afterhours/Social Club) | |||||||||||||||||||||||
| Street | 2.40 | [1.36 – 4.24] | 1.61 | [0.81 – 3.22] | 1.71 | [0.81 – 3.58] | 1.35 | [0.59 – 3.13] | 1.55 | [0.58 – 4.13] | 1.53 | [0.57 – 4.08] | |||||||||||
| Low-end Strip Club | 1.33 | [0.63 – 2.74] | 2.02 | [0.80 – 5.11] | 2.18 | [0.78 – 6.08] | 1.80 | [0.59 – 5.48] | 1.70 | [0.51 – 5.64] | 1.70 | [0.51 – 5.70] | |||||||||||
| High-end Strip Club | 1.52 | [0.65 – 3.55] | 1.02 | [0.35 – 2.92] | 1.12 | [0.36 – 3.48] | 0.72 | [0.19 – 2.77] | 0.86 | [0.20 – 3.72] | 0.96 | [0.22 – 4.18] | |||||||||||
| Gender (Ref = Cisgender Women) | |||||||||||||||||||||||
| Transgender Women | 0.72 | [0.37 – 1.40] | 0.90 | [0.39 – 2.06] | 1.24 | [0.51 – 3.03] | 1.32 | [0.52 – 3.40] | 1.12 | [0.38 – 3.27] | 1.21 | [0.40 – 3.63] | |||||||||||
| Cisgender Men | 0.53 | [0.28 – 1.01] | 0.60 | [0.28 – 1.32] | 0.78 | [0.34 – 1.82] | 0.75 | [0.29 – 1.95] | 0.42 | [0.14 – 1.23] | 0.47 | [0.16 – 1.39] | |||||||||||
| Race/Ethnicity (Ref = White/Caucasian) | |||||||||||||||||||||||
| Black/African American | 0.40 | [0.20 – 0.81] | 0.29 | ** | [0.12 – 0.67] | 0.29 | ** | [0.12 – 0.74] | 0.40 | [0.14 – 1.14] | 0.60 | [0.18 – 1.98] | 0.67 | [0.20 – 2.27] | |||||||||
| Latino/a | 2.16 | [0.65 – 7.19] | 1.70 | [0.44 – 6.57] | 1.78 | [0.40 – 7.88] | 3.49 | [0.69 – 17.73] | 6.26 | [0.90 – 43.35] | 6.61 | [0.92 – 47.50] | |||||||||||
| Mixed/Asian/Other | 0.44 | [0.18 – 1.08] | 0.27 | * | [0.10 – 0.77] | 0.21 | ** | [0.07 – 0.68] | 0.36 | [0.10 – 1.32] | 0.28 | [0.06 – 1.34] | 0.31 | [0.06 – 1.52] | |||||||||
| Age | 1.17 | [1.09 – 1.26] | 1.25 | *** | [1.14 – 1.36] | 1.29 | *** | [1.16 – 1.42] | 1.26 | *** | [1.13 – 1.41] | 1.35 | *** | [1.18 – 1.54] | 1.34 | *** | [1.17 – 1.53] | ||||||
| Level of education (Ref = High school/GED) | |||||||||||||||||||||||
| Some high school | 2.24 | [1.14 – 4.42] | 3.24 | ** | [1.50 – 7.01] | 2.25 | [0.94 – 5.39] | 1.57 | [0.59 – 4.16] | 1.40 | [0.45 – 4.32] | 1.45 | [0.46 – 4.56] | ||||||||||
| Some college | 0.59 | [0.32 – 1.10] | 0.46 | * | [0.23 – 0.91] | 0.36 | ** | [0.17 – 0.77] | 0.39 | * | [0.17 – 0.91] | 0.59 | [0.24 – 1.47] | 0.53 | [0.21 – 1.35] | ||||||||
| Bachelor’s degree or more | 1.28 | [0.47 – 3.50] | 0.49 | [0.16 – 1.52] | 0.51 | [0.15 – 1.70] | 1.00 | [0.27 – 3.73] | 0.81 | [0.19 – 3.49] | 0.82 | [0.19 – 3.57] | |||||||||||
| Arrested before age 18 | 3.37 | [1.94 – 5.85] | 4.38 | *** | [2.13 – 9.01] | 4.65 | *** | [2.09 – 10.37] | 4.29 | ** | [1.72 – 10.66] | 4.20 | ** | [1.67 – 10.55] | |||||||||
| In jail before age 18 | 4.15 | [1.45 – 11.87] | 1.47 | [0.35 – 6.17] | 1.61 | [0.36 – 7.13] | 1.55 | [0.27 – 8.88] | 1.41 | [0.24 – 8.17] | |||||||||||||
| Number of exchanges per month (Ref = 1–6) | |||||||||||||||||||||||
| 0 | 1.06 | [0.33 – 3.44] | 0.93 | [0.21 – 4.05] | 1.02 | [0.21 – 4.83] | 0.87 | [0.18 – 4.13] | |||||||||||||||
| 7–30 | 4.06 | [1.92 – 8.62] | 2.74 | * | [1.03 – 7.26] | 1.40 | [0.47 – 4.18] | 1.42 | [0.47 – 4.30] | ||||||||||||||
| >30 | 2.35 | [1.19 – 4.63] | 1.83 | [0.74 – 4.48] | 1.57 | [0.57 – 4.32] | 1.55 | [0.56 – 4.31] | |||||||||||||||
| Exchange for shelter | 2.57 | [1.43 – 4.61] | 2.48 | * | [1.05 – 5.85] | 2.22 | [0.82 – 6.01] | 2.28 | [0.83 – 6.21] | ||||||||||||||
| Has other work | 0.37 | [0.20 – 0.70] | 0.64 | [0.26 – 1.61] | 1.60 | [0.51 – 5.05] | 1.76 | [0.55 – 5.64] | |||||||||||||||
| Additional residence in the past year | 1.10 | [1.02 – 1.20] | 1.09 | [0.96 −1.24] | 1.09 | [0.95 – 1.26] | |||||||||||||||||
| Worried about a place to stay | 2.67 | [1.55 – 4.60] | 2.09 | [0.85 – 5.17] | 1.89 | [0.76 – 4.73] | |||||||||||||||||
| Each additional positive neighborhood trait | 1.20 | [1.06 – 1.36] | 1.15 | [0.95 – 1.41] | 1.13 | [0.93 – 1.38] | |||||||||||||||||
| Have reliable transportation (Ref = Always) | |||||||||||||||||||||||
| Never | 4.10 | [1.48 – 11.38] | 1.45 | [0.20 – 10.46] | 0.65 | [0.06 – 7.64] | |||||||||||||||||
| Sometimes | 1.45 | [0.83 – 2.54] | 1.64 | [0.67 – 4.05] | 1.54 | [0.62 – 3.82] | |||||||||||||||||
| Daily use of drugs | 33.51 | [4.42 – 254.0] | 9.31 | [0.61 – 141.34] | |||||||||||||||||||
| Level of stress about police (Ref = Not stressful) | |||||||||||||||||||||||
| Somewhat stressful | 2.74 | [0.96 – 7.85] | |||||||||||||||||||||
| Very stressful | 6.80 | [2.58 – 17.92] | |||||||||||||||||||||
| N | 265 | 251 | 232 | 208 | 208 | ||||||||||||||||||
Abbreviations: OR – odds ratio; CI – confidence interval; Ref – reference
p<0.05
p<0.01
p<0.001
4. Discussion
In this diverse population of young adults exchanging sex, the majority had been either arrested or incarcerated prior to participating in the study. Those variables that were highly associated with arrest and incarceration were those that placed respondents in highly policed areas, increasing their exposure to law enforcement. This included heightened exposure on the street venue, the placement of drug users into spaces targeted for policing of the drug trade, and the increased surveillance of individuals who had been previously arrested or incarcerated. Older respondents exchanging sex had spent more years under the scrutiny of law enforcement, and they were more likely to have been both arrested and incarcerated. Factors that also constrained sex work to more visible areas, such as lacking stable housing or dropping out of school before age 18, were also significantly related to criminal justice involvement. Consistent with a social ecological framework, these results support a geography of risk, wherein the interaction between individual behavior and the places where an individual lives and works largely determines the risk of criminal justice involvement (Mason, Cheung, & Walker, 2004; Shannon et al., 2008).
Individuals who exchange sex and cannot meet basic needs, as evidenced by exchanging for transportation and shelter, are often highly visible to law enforcement, and as a result are also at higher risk for criminal justice involvement. Those with the fewest resources end up exchanging on the street and concentrated in neighborhoods and city blocks known for street-based sex work, which are targets for law enforcement (Ousey & Lee, 2008; Thukral & Ditmore, 2003). The inability to meet basic needs is a consistent indicator of criminal justice involvement across studies, which may reflect a cyclic relationship (Cohan et al., 2006). Sex workers without resources are more likely to be arrested and incarcerated due to increased exposure to law enforcement, and sex workers who have been arrested and incarcerated may rely on sex work to meet basic needs.
Exposure to increased scrutiny because of a juvenile arrest or incarceration was significantly associated with adult criminal justice involvement. There are several potential explanations for this relationship. Firstly, individuals with a criminal record may be under increased formal (i.e., parole or probation) or informal (i.e., being known to local law enforcement officers) surveillance. Secondly, arrest and incarceration may reduce access to basic resources through the collateral consequences of a criminal record that can include exclusion from state and federal health care and welfare programs and limited employment opportunities (Pager, 2003; Pinard, 2006; Ramaswamy & Freudenberg, 2012). Thirdly, the association between juvenile arrest or incarceration and adult criminal justice involvement may be, in part, due to the fact that the most significant factor associated with arrest and incarceration prior to age 18 in this sample was having dropped out of high school. This is consistent with studies of the general population, and suggests that even in this group of highly vulnerable young adults, completing high school is highly protective (Pettit & Western, 2004). Even when other risk factors, such as drug use and venue of exchange, were included in the analyses, education remained a significant protective factor.
Consistent with other studies that have identified drug use as one of the most common reasons that sex workers become involved in the criminal justice involvement, it was strongly associated with arrest and incarceration in this study (Hankel et al., 2015; Lorvick, Comfort, Kral, & Lambdin, 2017). Although sustaining an addiction may result in illegal activity or increased sex work to pay for drugs, it certainly brings sex workers into physical spaces that are highly policed (Hankel et al., 2015). This increased intensity of policing is largely a result of policies rooted in the political rhetoric of the “war on drugs,” which disproportionately focus policing in poor, minority, and urban neighborhoods perceived to have high rates of drug crime (Alexander, 2012). For example, tough-on-crime political rhetoric emphasized incarceration as the most appropriate response to the crack “epidemic” in urban areas, rather than drug treatment or providing jobs for urban residents (B. Western, 2006). This shifted punishments toward increased jail and prison sentences for minor drug and street crimes, leading to higher rates of arrest and incarceration among drug users, including sex workers.
In contrast to a tremendous amount of previous work, and our own conceptual model in which the interactions between race, neighborhood, and criminal justice policy place racial and ethnic minority-identified people at disproportionate risk for criminal justice involvement, race was not significantly associated with arrest or incarceration in the adjusted analyses. Indeed, in the unadjusted analyses, respondents who identified as Black were significantly less likely to be incarcerated in an adult jail compared with their White counterparts, although this effect was attenuated and non-significant once the models were completely adjusted. As a research group, there was some discussion of a particular group of 15–20 cisgender women exchanging sex on the street; all of them identified as White and were daily users of heroin or cocaine and, we conjectured, might skew our findings in such a small overall sample (Snow et al., 2013). We ultimately concluded, however, that our results were likely a result of focusing on some of the most marginalized populations in Detroit, where the effect of race/ethnicity may be less pronounced because of the high levels of concentrated disadvantage across racial/ethnic categories.
Individuals reporting neighborhoods with more positive attributes were more likely to be involved in the criminal justice system. This was consistent even when the positive attributes were broken down into measures of access to neighborhood physical resources or social resources, and whether the neighborhood was dangerous; individuals reporting that their neighborhoods were better resourced, more socially connected, and more dangerous were all more likely to be arrested and incarcerated. These findings may reflect differential exposure to law enforcement in the neighborhood where an individual works, compared with where an individual lives. Individuals living in more resourced neighborhoods may be going to less resourced areas in order to exchange sex, or may stand out more to law enforcement in their neighborhoods. Alternatively, high degrees of social connection may reflect the residential segregation and concentrated disadvantage that make the neighborhood a “hot spot” for law enforcement. These findings suggest that sex workers living or working in segregated urban neighborhoods may not benefit as much from the physical and social resources they do have access to because those neighborhoods are nonetheless heavily policed (Ousey & Lee, 2008; Rodriguez, 2011).
In keeping with the social ecological model of risk for criminal justice involvement, the program and policy interventions supported by our findings span several levels. At the individual level, service providers working with young adults who exchange sex should screen for criminal justice involvement and legal needs and make referrals to legal aid societies and other community resources for free or low cost legal help. These organizations can help to navigate the system and to determine whether young adults are eligible to expunge charges from their criminal record.
Our findings also support interventions at the family and school to keep children and young adults in school and to provide access to resources to meet basic needs. The most successful programs to do this are early life interventions aimed at strengthening families and providing comprehensive education, family, and health service; these have been shown to decrease rates of school dropout and juvenile arrests (Reynolds, Temple, Robertson, & Mann, 2001). Across childhood and young adulthood, programs that achieve this include Nurse Family Partnerships, where trained nurse home visitors follow a detailed protocol that provides child care training and social skills development for the mother during the prenatal period and over the first two years of the child’s life, preschool programs that also work with families to develop parenting and life skills, and school-based programs emphasizing shared decision-making and problem-solving involving teachers, parents, students, community members and administrators (Greenwood, 2008).
Although structural interventions are often difficult to design and implement, they hold the potential to divert young adults away from the criminal justice system and offer a promising approach to changing the nature and intensity of policing of sex workers in urban neighborhoods. One example are drug treatment courts, and despite some mixed results, there is promise in diverting young adults away from the criminal justice system and into drug treatment programs in order to decrease not only recidivism but also relapse to substance use through these types of programs (Gallagher et al., 2015; Jewell, Rose, Bush, & Bartz, 2016). Demonstrating the potential for even more sweeping changes from structural interventions, Biradavolu and colleagues describe and community in India in which a non-governmental organization and community-based organization of sex workers created new standards for acceptable police behavior, set up a network to monitor compliance, and created a rapid reaction team to punish non-compliance through confrontation, publicity and legal action (Biradavolu, Burris, George, Jena, & Blankenship, 2009).
The respondents in this study were a small, non-random sample from a highly specific population of street-, strip club-, and after-hours party-based sex workers, limiting the generalizability of the findings. Due to concerns about confidentiality that limited pre-enrollment data collection, we cannot say how those who refused participation differ from those who responded to the survey. Our analysis of subsets was also limited by the diversity of the group and the small sample size within each venue. In an effort to minimize the effects of these sampling limitations, all analyses were stratified by venue. Online sex work is also increasingly common, and not captured in our survey (Thukral & Ditmore, 2003; Thukral et al., 2005). However, Cunningham and others find that only 13% of online sex workers have transitioned from working on the street to working online, and that those remaining in the street venue are those with the fewest resources, validating the continued study of sex workers in off-line venues (2011). Access to this difficult-to-reach and vulnerable urban population in Detroit was also an important strength of this research. We were able to identify the needs and potential points of intervention in this unique sample of high-risk young adults who are often overlooked.
The relatively small sample and complex relationships between the variables in this cross-sectional analysis also resulted in large effect sizes, large confidence intervals, and significant confounding. These less stable point estimates may be related to the interactions and confounding between these variables that we lacked sufficient power to test. We have presented models as we developed them, sequentially, so that the reader can see how each additional variable affected the estimates for the others, and how the sample size was affected by each subsequent addition. For example, when the stress about police variable was added to the multivariable models, the sample size was halved and many of the relationships lost significance and confidence intervals were increased substantially. This can be seen in Table 6 and Table 8, and stress about police was not included in Table 9 because the sample size was so small as to render the analysis meaningless. Although a longitudinal study, rather than a cross-sectional one, would have allowed for a stronger causal argument, with measurement of exposures at a time before the outcomes occurred, to date these types of studies have not been performed in this often hidden, highly marginalized community. As researchers, we are optimistic that whether future studies bear out a causal relationship between sex exchange and criminal justice involvement, or the reverse is found to be true, this early work will motivate interventions aimed at decreasing both among young adults.
Among the diverse young adults in this study, involvement in the justice system was a normative experience. The majority had been either arrested or incarcerated prior to involvement in the study, likely in large part a result of intense policing in poor, urban neighborhoods and particularly among people of color. In this setting, there was a consistent relationship between age and both arrest and incarceration, as it might be seen in the development of age-related morbidity. Other variables that additionally heightened exposure to law enforcement were also associated with heightened risk of arrest and incarceration, such as working on the street, using drugs, and living in neighborhoods with neighbors like them. These results suggest that over time, more of this sample of young adults living and working in highly policed neighborhoods and already with significant criminal justice experience, would find themselves either under community supervision as a probationer, inmate, or parolee, or with a criminal record. The true cost of policing and incarcerating young adults exchanging sex are not only those related to housing and feeding them during the period of incarceration or monitoring them after release, but also the long-term health and economic effects of involvement in the justice system (Massoglia, 2008; B. Western, 2006).
Figure 1.
Social-ecological model of risk for criminal justice involvement for sex workers
Highlights.
The majority of young adults who exchange sex have been involved in the criminal justice system.
Working on the street, using drugs and lacking stable housing increased risk for arrest and incarceration.
Completing high school was highly protective against criminal justice involvement.
Acknowledgements
This work was supported by the Ford Foundation Youth, Sexuality, Health and Rights Initiative. The first author was supported the University of Michigan Medical Scientist Training Program (NIGMS T32GM07863) during portions of this work. Special acknowledgement is due to Allison Brenner for her thoughtful comments on many drafts of this manuscript.
References
- Alexander M (2012). The new Jim Crow: Mass incarceration in the age of colorblindness: The New Press. [Google Scholar]
- Bernstein E (2004). The Transformation of Sexual Commerce and Urban Space in San Francisco. Footnotes, 32(1). [Google Scholar]
- Biradavolu MR, Burris S, George A, Jena A, & Blankenship KM (2009). Can sex workers regulate police? Learning from an HIV prevention project for sex workers in southern India. Social Science & Medicine, 68(8), 1541–1547. doi: 10.1016/j.socscimed.2009.01.040 [DOI] [PubMed] [Google Scholar]
- Bronfenbrenner U (1977). Toward an Experimental Ecology of Human Development. American Psychologist, 32(7), 513–531. [Google Scholar]
- Brush M (2013, February 12, 2013). TIMELINE: A short history of Michigan’s medical marijuana law. Retrieved from http://michiganradio.org/post/timeline-short-history-michigans-medical-marijuana-law [Google Scholar]
- Cobbina JE, & Oselin SS (2011). It’s Not Only for the Money: An Analysis of Adolescent versus Adult Entry into Street Prostitution. Sociological Inquiry, 81(3), 310–332. doi: 10.1111/j.1475-682X.2011.00375.x [DOI] [PubMed] [Google Scholar]
- Cohan D, Lutnik A, Davidson P, Cloniger C, Herlyn A, Breyer J, … Klausner JD (2006). Sex worker health: San Francisco style. Sexually Transmitted Infections, 82(5), 418–422. doi: 10.1136/sti.2006.020628 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cotton BP, & Schwartz-Barcott D (2016). Residential Instability Among Low-Income Families: A Concept Analysis. Archives of Psychiatric Nursing, 30(2), 257–261. doi: 10.1016/j.apnu.2015.11.006 [DOI] [PubMed] [Google Scholar]
- Cunningham S, & Kendall TD (2011). Prostitution 2.0: The changing face of sex work. Journal of Urban Economics, 69(3), 273–287. doi: 10.1016/j.jue.2010.12.001 [DOI] [Google Scholar]
- Faggiano F, Minozzi S, Versino E, & Buscemi D (2014). Universal school-based prevention for illicit drug use. The Cochrane database of systematic reviews(12), CD003020. doi: 10.1002/14651858.CD003020.pub3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gallagher JR, Nordberg A, Deranek MS, Ivory E, Carlton J, & Miller JW (2015). Predicting Termination from Drug Court and Comparing Recidivism Patterns: Treating Substance Use Disorders in Criminal Justice Settings. Alcoholism Treatment Quarterly, 33(1), 28–43. doi: 10.1080/07347324.2015.982451 [DOI] [Google Scholar]
- Graham LFL, Reyes AM, Lopez W, Gracey A, Snow RC, & Padilla MB (2013). Addressing Economic Devastation and Built Environment Degradation to Prevent Violence: A Photovoice Project of Detroit Youth Passages. Community Literacy Journal, 8(1), 41–52. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&db=eue&AN=92032469&site=ehost-live [Google Scholar]
- Grant BF, Saha TD, Ruan WJ, Goldstein RB, Chou SP, Jung J, … Hasin DS (2016). Epidemiology of DSM-5 Drug Use Disorder: Results From the National Epidemiologic Survey on Alcohol and Related Conditions-III. JAMA psychiatry, 73(1), 39–47. doi: 10.1001/jamapsychiatry.2015.2132 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Greenwood P (2008). Prevention and Intervention Programs for Juvenile Offenders. The Future of Children, 18(2), 185–210. [DOI] [PubMed] [Google Scholar]
- Hankel J, Heil M, Dewey S, & Martinez N (2015). Characteristics of women seeking services at a transitional housing facility for women leaving street-based sex work: Implications for social service providers. Journal of Social Service Research, 42(1), 41–56. doi: 10.1080/01488376.2015.1077188 [DOI] [Google Scholar]
- Harper C, & McLanahan S (2004). Father Absence and Youth Incarceration. Journal of Research on Adolescence, 14(3), 69–397. [Google Scholar]
- Jewell JD, Rose P, Bush R, & Bartz K (2016). The Long Term Effectiveness of Drug Treatment Court on Reducing Recidivism and Predictors of Voluntary Withdrawal. International Journal of Mental Health and Addiction, 15(1), 28–39. doi: 10.1007/s11469-016-9652-8 [DOI] [Google Scholar]
- Kendall C, Kerr LRFS, Gondim RC, Werneck GL, Macena RHM, Pontes MK, … McFarland W (2008). An Empirical Comparison of Respondent-driven Sampling, Time Location Sampling, and Snowball Sampling for Behavioral Surveillance in Men Who Have Sex with Men, Fortaleza, Brazil. AIDS and Behavior, 12(S1), 97–104. doi: 10.1007/s10461-008-9390-4 [DOI] [PubMed] [Google Scholar]
- Lopez WD, Graham LF, Reardon C, Reyes AM, Reyes A, & Padilla M (2012). “No jobs, more crime. More jobs, less crime”: structural factors affecting the health of Latino men in Detroit. Journal of Men’s Health, 9(4), 255–260. doi: 10.1016/j.jomh.2012.03.007 [DOI] [Google Scholar]
- Lorvick J, Comfort M, Kral AH, & Lambdin BH (2017). Exploring lifetime accumulation of criminal justice involvement and associated health and social outcomes in a community-based sample of women who use drugs. Journal of Urban Health, 95(4), 584–593. doi: 10.1007/s11524-017-0204-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- MacKellar DA, Gallagher KM, Finlayson T, Sanchez T, Lansky A, & Sullivan PS (2007). Surveillance of HIV Risk and Prevention Behaviors of Men Who Have Sex with Men—A National Application of Venue-Based, Time-Space Sampling. Public Health Reports, 122(S1), 39–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mason M, Cheung I, & Walker L (2004). Substance Use, Social Networks, and the Geography of Urban Adolescents. Substance use & misuse, 39(10–12), 1751–1777. doi: 10.1081/ja-200033222 [DOI] [PubMed] [Google Scholar]
- Massoglia M (2008). Incarceration as Exposure: The Prison, Infectious Disease, and Other Stress-Related Illnesses. Journal of Health and Social Behavior, 49(1), 56–71. [DOI] [PubMed] [Google Scholar]
- Murphy AK, & Venkatesh SA (2006). Vice Careers: The Changing Contours of Sex Work in New York City. Qualitative Sociology, 29(2), 129–154. doi: 10.1007/s11133-006-9012-2 [DOI] [Google Scholar]
- Ousey GC, & Lee MR (2008). Racial Disparity in Formal Social Control: An Investigation of Alternative Explanations of Arrest Rate Inequality. Journal of Research in Crime and Delinquency, 454(3), 322–355. [Google Scholar]
- Padilla M (2012). Spatial stigma, sexuality, and neoliberal decline in Detroit, Michigan. Scholar & Feminist Online, 11(1–2). [Google Scholar]
- Pager D (2003). The Mark of a Criminal Record. American Journal of Sociology, 108(5), 937–975. doi: 10.1086/374403 [DOI] [Google Scholar]
- Pettit B, & Western B (2004). Mass Imprisonment and the Life Course: Race and Class Inequality in U.S. Incarceration. American Sociological Review, 69(2), 151–169. [Google Scholar]
- Pinard M (2006). An Integrated Perspective on the Collateral Consequences of Criminal Convictions and Reentry Issues Faced by Formerly Incarcerated Individuals. Boston University Law Review, 86, 623–690. [Google Scholar]
- Ramaswamy M, & Freudenberg N (2012). The Cycle of Social Exclusion for Urban, Young Men of Color in the United States: What is the Role of Incarceration? Journal of poverty, 16(2), 119–146. doi: 10.1080/10875549.2011.639862 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Raphael J, & Shapiro DL (2002). Sisters Speak Out: The Lives and Needs of Prostituted Women in Chicago. Retrieved from Chicago: [Google Scholar]
- Reynolds AJ, Temple JA, Robertson DL, & Mann EA (2001). Long-term Effects of an Early Childhood Interention on Educational Achievement and Juvenile Arrest: A 15-Year Follow-up of Low-Income Children in Public Schools. Journal of the American Medical Association, 285(18), 2339–2346. [DOI] [PubMed] [Google Scholar]
- Rodriguez N (2011). Concentrated Disadvantage and the Incarceration of Youth: Examining How Context Affects Juvenile Justice. Journal of Research in Crime and Delinquency, 50(2), 189–215. doi: 10.1177/0022427811425538 [DOI] [Google Scholar]
- Sallis JF, Owen N, & Fisher EB (2008). Ecological Models of Health Behavior. In Glanz K, Rimer BK, & Viswanath K (Eds.), Health Behavior and Health Education: Theory, Research, and Practice (4th ed.). San Francisco: John Wiley & Sons, Inc. [Google Scholar]
- Sallmann J (2010). “Going Hand-in-Hand”: Connections Between Women’s Prostitution and Substance Use. Journal of Social Work Practice in the Addictions, 10(2), 115–138. doi: 10.1080/15332561003730155 [DOI] [Google Scholar]
- Shannon K, Rusch M, Shoveller J, Alexson D, Gibson K, & Tyndall MW (2008). Mapping violence and policing as an environmental–structural barrier to health service and syringe availability among substance-using women in street-level sex work. International Journal of Drug Policy, 19(2), 140–147. doi: 10.1016/j.drugpo.2007.11.024 [DOI] [PubMed] [Google Scholar]
- Snow RC, Williams A, Collins C, Moorman J, Rangel T, Barick A, … Reyes AM (2013). Paying to Listen: Notes from a Survey of Sexual Commerce. Community Literacy Journal, 8(1), 53–69. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&db=eue&AN=92032470&site=ehost-live [Google Scholar]
- Spohn C, & Holleran D (2000). The imprisonment penalty paid by young, unemployed Black and Hispanic male offenders. Criminology, 38(1), 281–306. [Google Scholar]
- Thukral J, & Ditmore M (2003). Revolving Door: An Analysis of Street-Based Prostitution in New York City. Retrieved from
- Thukral J, Ditmore M, & Murphy A (2005). Behind Closed Doors: An Analysis of Indoor Sex Work in New York City. Retrieved from
- Western B (2006). Punishment and Inequality in America. New York: Russell Sage Foundation. [Google Scholar]
- Western B, & Petit B (2010). Incarceration & social inequality. Dædalus(Summer), 8–19. [DOI] [PubMed]