Confirmatory Factor Analysis and Construct Validity of the Internalized Sex Work Stigma Scale among a Cohort of Cisgender Female Sex Workers in Baltimore, Maryland, United States

Abstract

Internalized sex work stigma among cisgender female sex workers (FSW) is produced within contexts of social marginalization and associated with a range of ill-effects, including psychological distress, and lower rates of healthcare-seeking. This study seeks to uncover latent domains of the new Internalized Sex Work Stigma Scale (ISWSS) using data from 367 FSW in Baltimore, Maryland, USA. The sample was 56% white with high substance use (82% smoked crack cocaine, 58% injected any drug). The average ISWSS score was 34.8 (s.d.=5.8, possible range: 12–48) and internal consistency was high (0.82). Confirmatory factor analysis revealed four sub-scales: worthlessness, guilt and shame, stigma acceptance, and sex work illegitimacy. Internal consistency of sub-scales was high (0.69–0.90); the scale also demonstrated construct validity with depression and agency. In bivariate logistic regressions, higher ISWSS, worthlessness, shame and guilt, and acceptance scores predicted higher odds of rushing client negotiations due to police. In unadjusted multinomial regressions, feeling respected by police predicted lower ISWSS, worthlessness, guilt and shame, acceptance, and illegitimacy scores. Identified factors are congruent with existing literature about how FSW manage sex work-specific stigma. Understanding unique dimensions and impacts of internalized sex work stigma can inform interventions and policy to reduce morbidities experienced by FSW.

Introduction

Stigma is considered a fundamental cause of social and health inequities that drives population-level morbidity and mortality, and refers to the “co-occurrence of labelling, stereotyping, separation, status loss, and discrimination in a context in which power is exercised” (Hatzenbuehler, Phelan, & Link, 2013). Stigma can be both enacted and perceived or ‘felt’ in social norms, values and expectations, with the result that feelings of shame, deviance from society, and self-blame may be internalized by targets of stigma (Scambler & Paoli, 2008a). Power is an essential component of stigma, whereby stigma results in social and economic exclusion, often resulting in stigmatized persons experiencing exploitation and less access to power and resources (Link & Phelan, 2001). Though sociologist Erving Goffman originally conceptualized stigma as a largely individual and interpersonal phenomenon, stigma has evolved to be understood as operating on several different socioecological levels (Goffman, 1963).

Female sex workers (FSW) face extensive stigma and can be assigned derogatory labels by society, perceived as criminals or “disease vectors,” or viewed as commodities rather than individual persons by clients (Vanwesenbeeck, De Graaf, Van Zessen, Straver, & Visser, 1993). In a recent review of what the authors term “prostitution stigma,” Benoit et al. argue that sex work-specific stigma is a fundamental cause of inequity for sex workers (Benoit, Jansson, Smith, & Flagg, 2018). FSW in locations where sex work is criminalized or illegal often experience structural violence that is institutionalized and exerted by systems rather than individuals and constrains ability to attain optimal health and life opportunities (Farmer, Nizeye, Stulac, & Keshavjee, 2006; Galtung, 1969). Laws, regulations and policies that criminalize sex work in order to maintain the “social order” are an example of structural violence as they reinforce stigmatizing views of sex workers and reproduce the social norm that sex work is an illegitimate source of income (Benoit, et al., 2018). Similarly, the justice and health care systems often ignore FSWs’ reports of harassment or violence perpetrated against them, and/or shame or discriminate against FSW based on perceptions about their choice to engage in “risky” behaviors (Benoit, et al., 2018; Vanwesenbeeck, 2001). Taken together, these experiences often lead to a denial of basic legal rights to sex workers when arrested. In countries where sex work is legal or decriminalized, such as New Zealand, research has found that decriminalization was an important first step in improved experiences of stigma among FSW and public attitudes about sex workers (Howard, 2019; Jonsson & Jakobsson, 2017; Platt et al., 2018). However, harassment of street-based or other “public” faces of sex work still remains (Armstrong, 2019). Armstrong argues that continued stigma against street-based sex workers in decriminalized settings is a heightened version of the sexism that women face every day with respect to sexual or verbal harassment and a way to punish women who do not conform to social norms of sexuality (Armstrong, 2019).

At the interpersonal level, FSW often experience public humiliation, harassment, or rejection by family and friends; FSW may even be stigmatized by other FSW who seek to differentiate themselves from “worse” sex workers (Ryan, Nambiar, & Ferguson, 2019; Scambler & Paoli, 2008b). In countries where sex work is legal or decriminalized, some street-based FSW still felt “othered” or experienced harassment when picking up clients on the street, showing lingering cultural stigmas that exist even when the structural violence of the criminal justice system is removed (Abel & Fitzgerald, 2010; Armstrong, 2016). Pervasive stigmatizing cultural views and power imbalances can produce negative self-image and feelings of shame that signify internalized stigma, which in turn results in “frequently disruptive” and “disabling” fear that may drive FSW to withdraw from society to avoid disclosing sex work history and resulting judgement (Scambler & Paoli, 2008a). It is important to note that not all FSW internalize sex work stigma or feel shame for selling sex. FSW have rejected internalizing shame through their ability to mobilize and organize to advocate for themselves in political or social spheres, though the criminalization of sex work can make this option difficult or impossible (Abel & Fitzgerald, 2010; Armstrong, 2019). The positive reframing of sex work as legitimate work and unworthy of stigma, and seeing paid sex as altruistic, the action of consenting adults, and authentic connections has shown to help FSW and male sex workers (MSW) resist internalized sex work stigma (Jiao & Bungay, 2019; Smith, Grov, Seal, and McCall, 2013; Koken, 2004; Bernstein, 2007; Shih 2004).

In general, FSWs’ ability to manage dual social roles can help manage shame, i.e. FSW tend to feel less shameful about selling sex when they can construct a public identity when selling sex and a private identity in other contexts (Abel & Fitzgerald, 2010; Shih, 2004). Indoor-based MSW and FSW may also prevent internalized stigma by differentiating themselves from street-based sex workers, though as noted this can create a stigmatizing hierarchy within sex work (Benoit, et al., 2018; Koken, Bimbi, Parsons, & Halkitis, 2004; Morrison & Whitehead, 2005). In criminalized settings, however, some FSW may not be able to create a public role related to their work, as such, but rather a hidden role due to fear of arrest. However, research has found that FSW are less able than MSW to construct these dual roles, as are street-based sex workers and those who sell sex for drugs (Abel & Fitzgerald, 2010). The limited body of literature shows high levels of stigma, including internalized stigma, among FSW in several countries where sex work is illegal, with no studies focused on FSW in the U.S. Among a sample of Chinese FSW who inject drugs, up to 93% agreed or strongly agreed with any one of the nine items on a measure of internalized sex work stigma (Gu et al., 2014). In a cohort of HIV-positive Zambian and South African FSW, 91% reported experiencing at least one stigmatizing event, 44% reported feeling ashamed of selling sex, and 62% reported feeling as though they lost respect from others (Hargreaves, Busza, Mushati, Fearon, & Cowan, 2017). FSW in this sample reported perceiving sex work-related stigma more often than HIV-related stigma (Hargreaves, et al., 2017). In India, Liu et al. measured stigma and sources of income among 151 FSW and found that increased income from non-sex work sources was associated with decreased internalized stigma scores (Liu et al., 2011). The internalization of stigma may lead to acceptance of human rights violations, such as gender-based violence and egregious policing, as deserved; create barriers to service access and feelings of disempowerment; and contribute to disparities in depressive symptoms between FSW and non-FSW (Benoit, et al., 2018; Ryan, et al., 2019).

Though the literature suggests a significant burden of stigma among FSW in other settings, stigma occurs and manifests differently in different contexts, where culture and power dynamics may operate in unique ways (Parker & Aggleton, 2003). For example, high HIV prevalence among sex workers in South Africa (40–72%) may contribute to sex work stigma in unique ways compared to the United States (17%) or India where HIV prevalence is lower (15%) (Paz-Bailey, Noble, Salo, & Tregear, 2016; Ramesh et al., 2010; UCSF, 2015). Brothel-based sex work—more common in India or South Africa than the United States— may confer some collective identity that protects against stigma but has also been found to foster competition among FSW to earn money for brothel madams, heightening social isolation and consequently internalized sex work stigma (Ghose, Swendeman, & George, 2011; Goldenberg, Duff, & Krusi, 2015; Stadler & Delany, 2006). Limited prior work of sex work-related stigma has been conducted in the US, and other studies have not used a standardized measure of internalized stigma to enhance comparability (Fitzgerald-Husek et al., 2017).

The present sample is a cohort of FSW in Baltimore, Maryland, United States. With few exceptions, sex work is illegal throughout the United States and is fully criminalized in Baltimore, Maryland. However, FSW continue to find clients in exotic dance clubs, through websites, and on the street, one of the most widely utilized methods by FSW in Baltimore (Decker et al., 2017; Sherman, Lilleston, & Reuben, 2011; Sherman et al., 2019). Areas of frequent street-based sex work in Baltimore tend to overlap with places with high concentration of drug activity and exist within primarily residential areas, bringing increased policing activity to these locations (Allen et al., 2018; Sherman et al., 2015). Studies have found high substance use among street-based FSW in Baltimore and the dual criminalization of sex work and drug use also increases vulnerability to arrest and police harassment (Footer, Silberzahn, Tormohlen, & Sherman, 2016; Park et al., 2019; Sherman, et al., 2015).

This manuscript describes a confirmatory factor analysis of a novel internalized sex work stigma scale. It also explores construct validity with depression and sense of agency, which we hypothesize will be correlated with higher and lower internalized stigma, respectively. Correlates between internalized stigma sub-scales and measures of interpersonal and structural violence are assessed. We hypothesize that a) police and client perpetrated violence would be associated with higher internalized sex work stigma scores, and b) that higher internalized sex work stigma scores would be associated with altered sex work practices due to police presence (e.g. rushing condom negotiation with clients).

Method

Procedure

EMERALD is a longitudinal study to assess a structural community-level intervention on HIV and STI risks among FSW in Baltimore, Maryland. The intervention consists of a drop-in center serving physical and mental health and basic needs of women-identified guests, as well as outreach services. A detailed description of the study protocol and intervention are forthcoming (Clouse et al., 2019). Recruitment for the current study was informed by a previously completed 12-month longitudinal cohort study of FSWs in Baltimore City that employed targeted sampling (Allen, et al., 2018). Briefly, the prior study paired a series of geospatial analyses of indicators of potential sex work (e.g., arrest data, 911 calls for prostitution) with data derived from an ethnographic study conducted with the Baltimore City Police Department to explore police attitudes toward sex work (inclusive of 300 hours of ride-alongs) to identify locations of potential street-based sex work activity throughout Baltimore City. The current study used information from this FSW cohort study regarding the geotemporal distribution of sex work in Baltimore City along with updated geospatial analyses of potential indicators of sex work to develop an updated sampling frame for FSW recruitment. Collectively, these analyses resulted in the identification of 10 locations in Baltimore City with likely street-based sex work activity. Of these, 6 reflected the intervention area in West Baltimore and 4 comprised the comparison areas in East and South Baltimore.

Eligibility criteria for EMERALD included: 1) aged 18 or older; 2) cisgender (i.e. identifies as a woman and was assigned female at birth; not transgender) woman; 3) sold or traded oral, vaginal, or anal sex “for money or things like food, drugs, or favors in the past 3 months;” 4) picked up clients 3 or more times in the past three months; and 5) willing to provide contact information for follow up visits. Trained study staff recruited participants using a mobile van from September 2017 to February 2019. Eligible participants provided written informed consent and completed a 50-minute Audio Computer-Assisted Self-Interview (ACASI) survey. Participants were given a tablet computer on which to complete the survey along with headphones to listen to all questions and responses read aloud. The survey consisted of demographics, sex work history, drug use, and psychosocial measures. Participants completed the survey in private spaces on the mobile van with study staff available if they had questions. Trained study staff conducted rapid HIV testing using OraQuick and women were given swabs and instructions for self-administered gonorrhea and chlamydia tests that were tested by the Baltimore City Health Department. The study was approved by the Johns Hopkins Bloomberg School of Public Health Institutional Review Board.

Measures

Internalized Sex Work Stigma Scale (ISWSS)

The ISWSS was adapted from a scale used by Carrasco and colleagues to measure sex work-related internalized stigma with HIV-positive FSW in the Dominican Republic, which has also been used in a study of FSW in Iringa, Tanzania (Carrasco et al., 2018; Kerrigan et al., 2017). Items in the original measure were asked twice, once referring to HIV positive status and again referring to sex worker identity. One item about avoiding friendships rather than disclosing HIV status was not asked about sex worker identity; we added this item to the ISWSS in order to understand a social dimension of internalized stigma. Relatedly, we also added an additional item informed by content in a stigma scale developed for FSW in India by Liu and colleagues (Liu, et al., 2011). The Liu scale focused exclusively on two subscales of perceived sex work-related stigma from family and the community and was not specific to internalized stigma (alpha = 0.91), however we included an item about feeling comfortable disclosing sex worker identity to others based on the importance of perceptions of stigma from family and friends found in Liu and colleagues but missing in the Carrasco stigma scale. Based on our prior work in the Baltimore sex work field, the items from these scales were deemed to be culturally relevant to the Baltimore context (Sherman et al., 2017; Sherman, et al., 2015; Sherman, et al., 2011). However, while the scale used by Carrasco et al. in the Dominican Republic includes only negatively worded items, the ISWSS expanded to include five positively worded items presenting sex work as normative (e.g. “you like selling sex”) prior to its use in Tanzania to reflect FSW attitudes expressed in other work about positive aspects of sex work for variability of experiences. The current 12-item ISWSS includes items about internalized reactions to perceptions of family or friends’ attitudes about sex work (e.g. “you feel comfortable telling others you are a sex worker”), how women view themselves (e.g. “working as a sex worker makes you feel like a bad person”) and perceptions of broader social views of sex work (e.g. “you see selling sex as work, just like any other job”). Responses are on a 4-point Likert scale (totally disagree, disagree, agree, totally agree).

Interpersonal and Structural Violence Correlates

Client-perpetrated physical and sexual violence in the past 6 months was assessed using a modified version of the Revised Conflict Tactic Scale (Straus & Douglas, 2004). Women were asked the degree to which they agreed with the statement: “I feel I am treated with respect by police officers” with responses on a four-point Likert scale ranging from strongly agree to strongly disagree. Answers were dichotomized to agree vs. disagree. Changes in sex work-related practices due to police in the area (rushed negotiations with clients, avoided carrying condoms) were also assessed.

Construct Validity Measures

Depression and agency were chosen as construct validity measures because of their association with stigma in prior literature (Earnshaw, Smith, Cunningham, & Copenhaver, 2015; Hong et al., 2010; Kalichman et al., 2009; Logie, James, Tharao, & Loutfy, 2013; Shih, 2004). Agency was measured using a modified 5-item version of the Pearlin Mastery Scale, which assesses “the extent to which one regards one’s life-chances as being under one’s own control in contrast to being fatalistically ruled” (potential range 5–25; Cronbach’s α=0.80) (Pearlin & Schooler, 1978). Depression was measured using the Patient Health Questionnaire-9 with responses on a 4-point Likert scale from “not at all” to “nearly every day” (potential range 1–27) (Kroenke, Spitzer, & Williams, 2001). Internal consistency was alpha=0.90.

Data Analysis

Exploratory Factor Analysis

Using data from the ISWSS at six-month follow-up in the SAPPHIRE cohort of street-based cisgender FSW in Baltimore who were recruited from October 2016 through August 2017 (Sherman, et al., 2019), an exploratory factor analysis (EFA) was first conducted (n=126). The number of factors extracted from the EFA was iteratively determined by the number of eigenvalues >1.0, the scree plot, uniqueness of items, and rotated factor loadings. Polychoric correlation was used, as was oblique (promax) rotation given correlation between factors. Given the small sample size of the dataset used for the EFA, we reproduced the EFA with the EMERALD dataset before conducting the final confirmatory factor analysis (CFA) to ensure a proper factor structure.

Confirmatory Factor Analysis

A CFA with polychoric correlation and maximum likelihood estimation was then conducted using the four-factor structure with baseline data from the EMERALD study. Participants who were missing responses to any scale question (including responding “don’t know” or “refused”) were excluded from analyses (n=18). To assess missing patterns, we created a binary indicator variable of full ISWSS completion. We then conducted chi-square and t-tests to compare completers and non-completers on key variables including: age, race, homelessness, time in sex work, motivations for starting sex work, and drug use variables. We did the same with the construct validity measures of agency and depression and outcome variables. There were no significant differences between women who completed all ISWSS questions and women who did not at p<0.05 level.

Model parameters and global fit statistics (Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), and the Root Mean Square Error of Approximation (RMSEA) and comparative fit statistics (Likelihood Ratio Test (LRT), Akaike Information Criteria (AIC), Bayesian Information Criteria (BIC) were used to determine the final model (Bentler, 1990; Hu & Bentler, 1999).

Construct Validity and Correlates

Positively worded items were reverse coded so higher scores for all items could be interpreted as indicating a higher degree of internalized stigma. Sub-scale scores were produced using the predict command in Stata which uses factor loadings in the score calculation. A test of internal consistency produced alphas for the full scale and sub-scales. To test construct validity, pairwise correlations were calculated between the full scale, sub-scales, and measure of depression and agency. Before using the continuous scores in models, approximate linearity was tested on ISWSS and sub-scale scores by creating a 10-level categorical variable for each score and adding it to logistic models; results showed that linearity was not a valid assumption. Therefore, we created three-level variables (high, medium, low) of each the full scale and sub-scale scores using tertiles. Unadjusted multinomial regressions were used to determine correlates of each score and interpersonal and structural violence variables conceptualized as predictors (e.g., client-perpetrated violence and treatment by police). Bivariate logistic regressions were used to determine correlates of each score and structural violence variables conceptualized as outcomes (e.g., rushing negotiations and avoiding carrying condoms due to police). As these analyses are exploratory, trends in significance at p<.10 were indicated. All analyses were conducted in Stata SE 15.1 (College Station, TX).

Results

Exploratory Factor Analysis

Two-thirds (67%) of the sample used for EFA was between 18–39 years old, 74% non-Hispanic white, and 52% had less than a high school education. Over half (53%) had sold sex for 10 years or longer and 20% started selling sex under 18 years old; the most common motivations for selling sex were to get drugs (77%), for basic necessities like food or shelter (29%), and to support family members (18%). The sample was also characterized by high recent drug use including injecting drugs (62%) and crack cocaine smoking (78%). Eigenvalues and a scree plot suggested a three-factor solution, however rotated factor loadings showed cross-loading and high uniqueness values (>0.90) for two items. A four-factor model was then extracted which resolved cross-loading and showed lower, though still relatively high, uniqueness values (0.79 and 0.84). A four-factor model was deemed optimal: feeling worthless because of sex work, feeling shameful or guilty about engaging in sex work, acceptance of sex work stigma, and seeing sex work as illegitimate work. The two unique items were related to sex work illegitimacy and were retained as their own factor due to the importance of perceived legitimacy of stigmatization on how an individual responds to stigma (Shih, 2004). The EFA conducted with EMERALD data reproduced a four-factor solution; the illegitimacy items showed less extreme uniqueness values (0.57 and 0.56) and clear loadings on to their own factor. Factor loadings on the remaining items ranged from 0.57 to 0.89 and uniqueness values ranged from 0.26 to 0.50.

EMERALD Descriptive Characteristics

The final CFA sample was n=367. More than half (53%) of participants were between 18–39 years old, 46% had less than a high school education, and 44% were non-White (Table 2). Fifty-nine percent of participants had sold sex for more than 10 years and 22% started selling sex before age 18. Most participants (80%) began selling sex, at least in part, to get drugs; prevalence of recent (past 6 months) drug use was high, including smoking crack cocaine (82%), injecting heroin (56%), and injecting any drug (58%).

Table 2.

Descriptive characteristics of sample of female sex workers in Baltimore, Maryland (n=367)

Variable	Total (N=367)
Age
18–29	86 (23.4)
30–39	146 (39.8)
40–49	94 (25.6)
50+	41 (11.2)
Education
Less than HS graduation	167 (45.5)
HS graduate/GED	92 (25.1)
Some college+	108 (29.4)
Race
White	206 (56.1)
Black	133 (36.2)
Other	28 (7.6)
Ever diagnosed with mental illness	208 (57.0)
Arrested, ever	300 (81.7)
Arrested, past 6 months*	99 (27.1)
Arrested for solicitation or prostitution, ever	188 (51.3)
Time in Sex Work
<10 years	150 (40.9)
>10 years	217 (59.1)
Age Started Sex Work*
<18 years old	80 (21.9)
18–30 years	201 (55.1)
> 30 years old	84 (23.0)
Motivation for starting sex work (select all that apply)±
No other job opportunities	69 (18.9)
To support children or family	99 (27.1)
To get drugs	291 (79.5)
For basic necessities	124 (33.9)
For pleasure	18 (4.9)
Condomless vaginal or anal sex with clients, past week	164 (44.7)
Client physical or sexual violence, past 6 months	134 (36.5)
Rushed negotiations with clients, past 6 months	168 (45.9)
Avoided carrying condoms because of police, past 6 months	57 (15.6)
Smoked crack, past 6 months	300 (81.7)
Injected heroin, past 6 months	206 (56.1)
Injected any drug, past 6 months	214 (58.3)
Smoke crack cocaine daily, past 6 months	246 (67.0)
Daily heroin injection, past 6 months	165 (45.0)

Footnotes:

^*n=2 missing;

^±n=1 missing

Confirmatory Factor Analysis and Construct Validity

The four factors extracted include: worthlessness (“WORTH”), guilt & shame (“SHAME”), stigma acceptance (“ACCEPT”), and perceived sex work illegitimacy (“LEGIT”) (Figure 1); the constituent items of each factor appear in Table 1. The sex work illegitimacy factor has only two items, which is generally not preferable as its own factor; however, factors with two items are permissible when items are highly correlated with only each other and uncorrelated with other items (Tabachnick, Fidell, & Ullman, 2007). The two items comprising the sex work illegitimacy factor showed relatively high correlation with only each other (r=0.61) compared to other items (all r’s <0.33) in the EMERALD sample. Further, the two items express important dimensions of internalized stigma among FSW as it relates to the perceived legitimacy of sex work (Benoit, et al., 2018; Ryan, et al., 2019). Global fit statistics were within recommended values (RMSEA=0.06, CFI=0.97, TLI=0.96). All factor loadings were statistically significant (p<.001).

Figure 1.

Model of confirmatory factor analysis of the Internalized Sex Work Stigma Scale.

Table 1.

Frequencies of Internalized Sex Work Stigma Scale among female sex workers in Baltimore, Maryland (n=367)

Question	Subscale*	Totally disagree	Disagree	Agree	Totally agree
Working as a sex worker makes you feel like a bad person.	W	40 (10.9)	66 (18.0)	160 (43.6)	101 (27.5)
You feel like you’re not as good as others because you are a sex worker.	W	37 (10.1)	78 (21.3)	156 (42.5)	96 (26.2)
People’s attitudes about sex work make you feel worse about yourself.	W	32 (8.7)	81 (22.1)	166 (45.2)	88 (23.9)
You feel completely worthless because you are a sex worker.	W	42 (11.4)	62 (16.9)	165 (45.0)	98 (26.7)
You feel guilty because you are a sex worker.	G	70 (19.1)	65 (17.7)	109 (29.7)	123 (33.5)
You feel ashamed of sex work.	G	50 (13.6)	49 (13.4)	116 (31.6)	152 (41.4)
It’s easier to avoid friendships than worry about telling others you are a sex worker.	G	45 (12.3)	64 (17.4)	146 (39.8)	112 (30.5)
You like your job as a sex worker.	A	182 (49.6)	133 (36.2)	39 (10.6)	13 (3.5)
You feel okay about being a sex worker.	A	172 (46.9)	137 (37.3)	52 (14.2)	6 (1.6)
You feel comfortable telling others that you are a sex worker.	A	194 (52.9)	124 (33.8)	38 (10.4)	11 (3)
You deserve respect as a sex worker.	I	53 (14.4)	73 (19.9)	171 (46.6)	70 (19.1)
You see sex work as work, just like any other job.	I	51 (13.9)	98 (26.7)	155 (42.2)	63 (17.2)

Note: Underlined words in each question correspond to items referenced in Figure 1.

^*W = worthlessness; G = guilt and shame; A = sex work acceptance; I = sex work illegitimacy

The full ISWSS and sub-scales showed good internal consistency and construct validity (Table 3). Internal consistency was high: alpha=0.82 for the full ISWSS and ranged from alpha=0.69 to 0.90 for sub-scales. The ISWSS was significantly correlated with worthlessness (r(365)=0.84, p<0.001), guilt and shame (r(365)=0.75, p<0.001), and stigma acceptance subscales (r(365)=0.54, p<0.001), though it was not correlated with sex work illegitimacy. Sub-scales were all significantly correlated with each other though not always in the expected direction: sex work illegitimacy was significantly negatively correlated with worthlessness (r(365)=−0.30, p<0.001) and guilt and shame (r(365)=−0.53, p<0.001) subscales. Correlations between the full ISWSS, sub-scales, and agency and depression showed good construct validity. As expected, depression was significantly correlated with the ISWSS (r(359)=0.44, p<.001), worthlessness (r(359)=0.51, p<.001), guilt/shame (r(359)=0.38, p<.001), and sex work illegitimacy (r(359)=−0.17, p=0.001), though it was not correlated with stigma acceptance. Agency was significantly correlated with the ISWSS (r(364)=−0.21, p<0.001), worthlessness (r(364)=−0.33, p<0.001), guilt/shame (r(364)=−0.23, p<0.001) and was positively associated with sex work illegitimacy (r(364)=0.18, p<0.001),

Table 3.

Internal consistency, correlations and construct validity of the Internalized Sex Work Stigma Scale and subscales in a sample of female sex workers in Baltimore, Maryland (n=367)

	Alpha	ISWSS+	Worthlessness	Guilt and Shame	Stigma acceptance	Sex Work Illegitimacy
ISWSS+	0.82
Worthlessness	0.90	0.84***
Guilt and Shame	0.82	0.75***	0.70***
Stigma Acceptance	0.83	0.54***	0.16**	0.11*
Sex Work Illegitimacy	0.69	0.03	−0.30***	−0.53***	0.36***
Construct Validity
Depression	0.90	0.44***	0.51***	0.38***	0.10	−0.17**
Agency	0.80	−0.21***	−0.33***	−0.23***	0.10	0.18***

Footnotes:

^***p<.001,

^**p≤.01,

^*p≤.05;

⁺ISWSS = Internalized Sex Work Stigma Scale

Correlations Between Scale and Violence Measures

Unadjusted multinomial regressions also showed interpersonal and structural violence predictors of ISWSS and sub-scale scores (Table 4). Compared to a low score, client-perpetrated violence predicted greater risk of high worthlessness score (odds ratio [OR]=2.00, 95% confidence interval [CI]=1.18–3.39, p=0.01). Compared to low scores, being treated with respect by police predicted lower risk of medium (OR=0.57, 95% CI=0.35–0.95, p=0.03) or high (OR=0.42, 95% CI=0.25–0.71, p=0.001) ISWSS score, high (OR=0.46, 95% CI=0.27–0.77, p=0.003) worthlessness score, medium (OR=0.63, 95% CI=0.38–1.04, p=0.07) and high (OR=0.51, 95% CI=0.30–0.85, p=0.01) guilt and shame score, and medium (OR=0.59, 95% CI=0.35–0.98, p=0.04) and high (OR=0.51, 95% CI=0.31–0.85, p=0.01) stigma acceptance score.

Table 4.

Structural violence predictors of Internalized Sex Work Stigma Scale and sub-scales in a sample of female sex workers in Baltimore Maryland (n=367)

	Internalized Sex Work Stigma Scale (ref = low)		Worthlessness (ref= low)		Guilt and Shame (ref= low)		Stigma Acceptance (ref= low)		Sex Work Illegitimacy (ref= low)
	Medium	High	Medium	High	Medium	High	Medium	High	Medium	High
Client physical or sexual violence+ (ref = no)	1.20 (0.71, 2.00)	1.15 (0.68, 1.95)	1.47 (0.86, 2.52)	2.00 (1.18, 3.39) **	1.21 (0.72, 2.04)	1.25 (0.74, 2.11)	1.10 (0.65, 1.84)	1.02 (0.60, 1.71)	0.72 (0.43, 1.20)	0.74 (0.44, 1.24)
Treated with respect by police officers (ref = disagree)	0.57 (0.35, 0.95)**	0.42 (0.25, 0.71)***	0.73 (0.44, 1.21)	0.46 (0.27, 0.77)***	0.63 (0.38, 1.04)*	0.51 (0.30, 0.85)***	0.59 (0.35, 0.98)**	0.51 (0.31, 0.85)***	1.43 (0.86, 2.39)	1.08 (0.64, 1.81)

Footnotes:

⁺past 6 months;

^***p≤.01,

^**p<.05,

^*p<.10

Bivariate logistic regressions showed ISWSS and sub-scale score predictors of changes in sex work practices due to police (Table 5). Compared to low scores, predictors of FSW rushing client negotiations due to police included: medium (OR=1.79, 95% CI= 1.07–2.97, p=0.03) or high (OR=2.85, 95% CI= 1.69–4.81, p<0.001) ISWSS scores, medium (OR=1.78, 95% CI=1.06–2.97, p=0.03) or high (OR=2.43, 95% CI=1.45–4.07, p =0.001) worthlessness scores, high (OR=2.56, 95% CI=1.52–4.28, p<0.001) guilt and shame score, high (OR=2.31, 95% CI=1.39–3.85, p=0.001) stigma acceptance score, and medium (OR=1.59, 95% CI=0.96–2.63, p=0.07) sex work illegitimacy score. Avoiding carrying condoms due to police was predicted by a medium (OR=0.52, 95% CI=0.25–1.09, p=0.08) compared to a low sex work illegitimacy score.

Table 5.

Interpersonal and structural violence as outcomes of Internalized Sex Work Stigma Scale and sub-scales in a sample of female sex workers in Baltimore Maryland (n=367)

	Rush negotiations with clients due to police+	Avoided carrying condoms due to police+
Internalized Sex Work Stigma Scale (ref = low)
medium	1.79 (1.07, 2.97) **	0.87 (0.43, 1.76)
high	2.85 (1.69, 4.81) ***	1.16 (0.59, 2.29)
Worthlessness (ref= low)
medium	1.78 (1.06, 2.97) **	0.73 (0.35, 1.50)
high	2.43 (1.45, 4.07) ***	1.13 (0.58, 2.20)
Guilt and Shame (ref= low)
medium	1.53 (0.91, 2.55)	0.94 (0.45, 1.96)
high	2.56 (1.52, 4.28) ***	1.54 (0.78, 3.04)
Stigma Acceptance (ref= low)
medium	1.03 (0.61, 1.71)	0.63 (0.31, 1.30)
high	2.31 (1.39, 3.85) ***	0.99 (0.51, 1.92)
Sex Work Illegitimacy (ref= low)
medium	1.59 (0.96, 2.63) *	0.52 (0.25, 1.09) *
high	1.08 (0.65, 1.80)	0.90 (0.47, 1.74)

Footnotes:

⁺past 6 months;

^***p≤.01,

^**p<.05,

^*p<.10

Discussion

The current study is one of the first to validate a measure of internalized stigma among FSW in the U.S. We found evidence of a four-factor scale of internalized sex work stigma; the resulting factors introduce complexity and nuance to this construct including worthlessness and guilt and shame, and the internal manifestations of stigma that may result from larger cultural and social stereotypes of FSW as criminals and sex work as illegitimate.

Importantly, these results are congruent with adaptive and maladaptive ways that FSW may respond to and manage sex work stigma, as described in prior literature. First is internalization, as described above, which is captured in the negative feelings about self in both the worthlessness and guilt and shame factors (Benoit, et al., 2018; Sallmann, 2010). Other stigma management strategies include reframing sex work as work and emphasizing the positive aspects of the work and resistance to internalizing stigmatizing labels and attitudes, possibly involving collective action (Benoit, et al., 2018; Benoit et al., 2019; Sallmann, 2010). Although we reverse-coded the items that comprise the sex work acceptance and sex work illegitimacy factors, the original items demonstrate the extent to which FSW have internally resisted sex work stigma and have reframed the personal and social value and legitimacy of sex work (i.e. feeling okay with being a sex worker). Benoit et al. also write about information management as a technique to mitigate stigmatization (Benoit, et al., 2018; Benoit, et al., 2019). This appears to be a common theme between many factors: restricting peers’ knowledge of sex work involvement by avoiding friendships or disclosing sex work to others can avoid potential negative reactions that heighten internalized stigma.

Sub-scales of the ISWSS describe dimensions of symbolic and structural stigma as reflected in internalized sex work stigma. Symbolic stigma is often expressed through implicit and explicit communication and has previously manifested in labeling FSW as “dirty,” blaming FSW for their drug use or violence victimization, and stereotyping FSW as deceitful (Ryan, et al., 2019). The worthlessness scale describes these dimensions of symbolic stigma, such as feeling “bad” or “worthless” for selling sex, or feeling worse about oneself because of the attitudes of others. Symbolic stigma against FSW often reflects social and cultural norms about women’s sexuality; internalized stigma felt by FSW may reflect their integration of larger cultural stereotypes and devaluation regarding FSW (Ryan, et al., 2019). The stigma acceptance and guilt and shame scales similarly show the internalized aspects of symbolic stigma: acceptance of sex work stigma can lead to guilt and shame about selling sex, both of which, to some extent, reflect FSWs’ own belief in the labeling and stereotypes about themselves.

Correlations between ISWSS sub-scales and depression and agency provide evidence of construct validity that fits with findings from the literature (Earnshaw, et al., 2015; Hong, et al., 2010; Kalichman, et al., 2009; Logie, et al., 2013; Shih, 2004). As hypothesized, depression was positively correlated with the full ISWSS, worthlessness and guilt and shame sub-scales and agency was negatively associated with the full ISWSS, worthlessness, and guilt and shame. However, sex work illegitimacy was marginally negatively correlated with depression and positively associated with agency. It is important to remember that sex work illegitimacy is comprised of only two items; further work should be done to explore this specific dimension of internalized stigma and help elucidate these unexpected findings. Future research should also validate the ISWSS against additional constructs such as anxiety, self-esteem, or self-concept to assess construct validity and internalized stigma (Stevelink, Wu, Voorend, & van Brakel, 2012).

The ISWSS is a novel and robust scale that provides a nuanced understanding of many dimensions of internalized sex work stigma. Key correlates of sub-scales show associations between interpersonal and structural violence and unique aspects of internalized stigma. For example, experiencing physical or sexual violence from a client was associated with greater worthlessness scores. Prior literature has found a consistent link between violence victimization and poor psychosocial outcomes in the general population and FSW more specifically (Campbell, Dworkin, & Cabral, 2009; Rossler et al., 2010); experiencing violence from clients could drive greater feelings of worthlessness specifically related to sex work. Additionally, FSW are also less likely to be taken seriously by law enforcement, if they report violence at all (Krüsi, Kerr, Taylor, Rhodes, & Shannon, 2016; Krüsi et al., 2014). FSW may feel greater feelings of worthlessness if they experience violence but feel unsupported in seeking recourse or even blamed for the violence.

Structural violence toward FSW is correlated with ISWSS worthlessness, guilt and shame, and stigma acceptance sub-scales measures. It is impossible to establish causality with the current analysis, which is cross-sectional, even more so given the complexity of internalized stigma and dynamic ongoing police interactions. FSW may feel ashamed of selling sex because they are viewed suspiciously or harassed by police and must subsequently rush negotiations with clients, though it is possible that FSW rush negotiations with clients because they feel guilt and shame about selling sex and do not want to raise police attention. This reflects larger HIV-related stigma literature that demonstrates the ways in which perceived (or felt-normative) and enacted (acts of mistreatment) stigma are associated with internalized stigma (Turan et al., 2017). It is critical to understand that internalized stigma is produced in larger social and structural contexts of exclusion to avoid pathologizing sex workers (Pantelic, Sprague, & Stangl, 2019).

Understanding the dimensions of internalized sex work-related stigma is important to design more effective interventions and policies. Structural violence is expressed through institutions, policies, and systems (Galtung, 1969). The sex work illegitimacy sub-scale raises issues of internalized stigma that results from the criminalization of sex work, which positions sex work as a job that is distinct and lesser than other licit jobs and not worthy of respect. Sex work stigma literature has mainly reported on contexts in which selling sex in any form (United States, China, South Africa) or profiting from sex work (India) is illegal, making an examination of the internalized aspects of sex work illegitimacy relevant and necessary (Footer, et al., 2016; Open Society Foundation, 2012). Reinforcing structural and symbolic stigmas can lead to anticipating stigmatizing experiences from others and subsequently internalizing larger cultural and social beliefs, making each stigma factor distinct but interrelated. For example, a FSW believing her job is illegitimate and not worthy of respect can lead to feelings of worthlessness and guilt or shame for engaging in the work. This interaction between stigma dimensions is an important area for future research, as it implies synergy between stigma dimensions that have been reported in sexual stigma research but are underexplored in sex work stigma research (Logie, James, Tharao, & Loutfy, 2011).

Understanding stigma dimensions and their health effects can also inform more actionable interventions (Logie, et al., 2011). On an individual level, interventions that help sex workers cognitively reframe stigmatizing experiences and assert agency in interactions with clients or mangers have shown promising results and can be useful interventions in samples where worthlessness or guilt and shame sub-scales are most salient (Benoit, et al., 2019; Koken, et al., 2004; Morrison & Whitehead, 2005). However, many of these interventions have focused on MSW and their results, though promising, may manifest differently in light of sexist views of women and sexuality (Shih 2004; Sanders 2017). Education and training on sex work from a harm reduction perspective (i.e. not pressuring exiting sex work) with health care providers and other service providers has also been shown to reduce stigma in practice-based settings (Bodkin, Delahunty-Pike, & O’Shea, 2015; Sanders, 2018). Indoor MSW in Canada reported the importance of setting boundaries with clients and maintaining a sense of control over client choice, as well as peer connections with other sex workers, in resisting internalized sex work stigma (Jiao & Bungay, 2019). Without a favourable political or social context, individual-level stigma reduction interventions have limited usefulness or may even open up sex workers to violence (Weitzer, 2018). Therefore, decriminalizing sex work is crucial for reducing the entrenched social stigma of sex work as illegitimate work that cannot be disclosed for fear of arrest as well as structural violence by police (Howard, 2019; Platt, et al., 2018; Weitzer, 2018). Immordino et al. found that an association between increasing legality of sex work and more positive attitudes about sex work on country-wide levels (Immordino & Russo, 2015). However, decriminalization of sex work is not a panacea and sex work stigma and related harassment can still linger for sub-groups of FSW such as those who sell sex on the street or use drugs, even in decriminalized settings (Abel & Fitzgerald, 2010; Armstrong, 2016). Empowering sex worker advocacy organizations and collective action can also provide bottom-up social change through legislation or media portrayals of sex work, for example (Weitzer 2018). Though this can be difficult in jurisdictions where sex work is criminalized, support for sex worker rights and health from organizations such as the National Organization of Women, the World Health Organization, and Amnesty International have broadened a coalition fighting back against stigma (Weitzer, 2018).

There are several limitations of this analysis. First, the sex work illegitimacy sub-scale consists of only two items, which is acceptable but not ideal. Future research should explore FSW feelings of sex work illegitimacy and its relationship to internalized sex work stigma. Second, we used computer tablets with ACASI for data collection. Though ACASI allows women to complete surveys about sensitive information in private, it also may introduce opportunities for women to answer “refuse” or “don’t know” to questions that may otherwise be probed by an interviewer. Relatedly, we coded any “don’t know” or “refused” answers as missing and therefore 18 women were dropped from our sample, though a comparison of participants with missing and non-missing data on key variables did not reveal any significant differences.

This cohort was recruited in an urban environment in the United States and may not be generalizable to other contexts. Further, our sample is comprised of a cohort of FSW that do not reflect all sex worker experiences in the United States or beyond. The sample is highly structurally vulnerable, has high past engagement with police and the criminal justice system, and a majority of the sample entered sex work to support their drug use; very few entered sex work for pleasure or enjoyment. Though street-based sex work was not an eligibility criterion, a very small percentage of women sold sex in off-street locations. Research has shown that levels of stigma tend to be higher on average for street-based sex workers compared to sex workers in off-street locations, even in settings where sex work is decriminalized or legal (Hubbard, 2013). Therefore, it is important to note that this study is not intended to be generalizable to all sex workers. Future research should validate this scale in other locations, other sex work contexts such as rural settings, and in other countries where the legality of sex work differs from the United States. Additionally, the intersection of sex work stigma with other marginalized identities (such as race, gender identity and sexual orientation) could be explored in the future.

The ISWSS and its constituent factors can provide researchers a tool with which to quantify and better understand the role that internalized sex work stigma plays in the lives of FSW. Future research can discover heterogeneity in the manifestations of internalized sex work stigma and its role in health outcomes, thereby providing more specific evidence for future interventions.