The Context of Economic Insecurity and Its Relation to Violence and Risk Factors for HIV Among Female Sex Workers in Andhra Pradesh, India

SYNOPSIS

Objective

We examined the context of economic insecurity and debt among female sex workers (FSWs), how this context varies among FSWs, and its association with experiences of violence and sexual risk factors for human immunodeficiency virus (HIV).

Methods

We recruited FSWs aged ≥18 years (n=673) through respondent-driven sampling for a survey on HIV risk in this region. Using logistic regression models (adjusted for partner status, education, financial support, and literacy), we assessed the relation between debt and sexual and physical victimization as well as sexual risk. We also conducted qualitative interviews with a subsample of the survey participants and examined these for related themes.

Results

In adjusted logistic regression models, FSWs who reported debt were more likely to report the following: recent physical violence (adjusted odds ratio [AOR] = 2.4, 95% confidence interval [CI] 1.5, 3.9), unprotected sex with occasional clients in the past week (AOR=2.3, 95% CI 1.2, 4.3), anal sex with clients in the past 30 days (AOR=2.0, 95% CI 1.1, 3.9), and at least one sexually transmitted infection symptom in the past six months (AOR=1.6, 95% CI 1.1, 2.4). FSWs with debt were more likely to report current husbands or other male partners, and less likely to report condom use with these partners, further increasing their sexual risk. Qualitative data elaborated on these findings.

Conclusion

Findings indicate the violence- and HIV-related vulnerability of FSWs who report debt and further highlight how male partners may contribute to the debt and economic insecurity of FSWs.

During the past decade, the HIV pandemic has brought increased attention to female sex workers (FSWs) in India and elsewhere. Dire socioeconomic conditions are often noted as a primary reason for women's involvement in sex work.^1,2 It is likely that such economic conditions also create an urgency that contributes to unprotected sex with clients and associated risk for human immunodeficiency virus (HIV). For example, FSWs often report making more money for sex trades when no condom is used;³ it is reasonable to suggest that those facing burdensome economic circumstances may be more likely to engage in such risky sex trades. However, despite the likely pervasiveness of economic hardship among this population and its significant role in initiation into sex work, very little research has specifically investigated the impact of economic insecurity on vulnerability to HIV among FSWs.

FSWs in India and elsewhere are highly vulnerable to various forms of violence (perpetrated by police, clients, and others).^4,5 Economic insecurity and debt may contribute to this violence; for example, FSWs who need to obtain funds to pay debts may be more likely to work in riskier contexts that increase their vulnerability to violence. While addressing violence is an important issue by itself, it is also a critical factor for understanding HIV risk among FSWs, as numerous studies have documented the contribution of violence to HIV risk in this population.^4,6–11

In this article, we examine issues related to economic conditions among FSWs in Andhra Pradesh, India. Specifically, we focus on debt as a specific form of economic insecurity that is likely to play a role in both the initiation and continuation of sex work, and to influence the risks associated with it. We draw from both qualitative and quantitative data to examine whether FSWs are burdened with and impacted by debt, and whether reported economic insecurity, as measured by debt, is associated with experiences of violence and HIV risk.

HIV is of critical concern in India, where 2.5 million people are estimated to be living with HIV.¹² The southern Indian state of Andhra Pradesh is among the states with the highest rates of HIV and where FSWs are largely affected.¹³ In previous research in this region, we found that consistent condom use among this population was associated with economic vulnerability, as measured by degree of economic independence.¹⁴ Beyond this finding, however, there is little systematic analysis of the impact of economic insecurity on violence or HIV-related risk among FSWs. We expanded on this work by focusing on (1) some of the ways that debt may influence involvement in sex trade, (2) characteristics of FSWs that are associated with debt, and (3) the association between debt and experiences of violence and sexual risk.

METHODS

Procedure and sample

Data for this analysis were collected as part of Project Parivartan, which analyzes the implementation and impact of a community mobilization intervention targeted to reduce HIV risk among FSWs in Rajahmundry, a city in the East Godavari District of Andhra Pradesh, India. Quantitative data are from a cross-sectional survey administered from March 30 to May 30, 2007. Participants were females aged ≥18 years who reported having sex in exchange for money in the year prior to the survey and who were capable of providing informed consent.

Project staff recruited participants via respondent-driven sampling (RDS). Specifically, five initial participants, or “seeds,” were recruited from the target population and asked to distribute up to three coupons to members in their social networks who met study inclusion criteria. Subsequent participants, who could only screen into the study if they had a coupon, were similarly provided with three coupons to recruit social network members who met the inclusion criteria. Given the high number of waves (i.e., each seed recruited women who recruited more women, and this continued for multiple waves), we believe that each round represented adequate coverage of the population of FSWs in this region.^15,16 (Our previous work describes this recruitment process in more detail.¹⁷)

Study staff developed surveys in English, had them translated into Telugu (the official language of Andhra Pradesh), and then had them translated back into English to confirm the correct translation. A total of 673 FSWs gave informed consent and completed the survey. No participants were missing data on relevant study variables; therefore, our analyses included all 673 respondents.

Qualitative data came from semi-structured life history (LH) interviews conducted by study ethnographers at two points in time. They first conducted one-on-one interviews from April 1 to June 29, 2006, with 23 participants who had completed a previous wave of the survey component (which also occurred during these months in 2006). Specifically, survey participants who reported that they were interested were invited to return at a later time to complete the LH interview. From January to February 2007, the study staff conducted a second round of five LH interviews with additional FSWs who had been involved in the community mobilization intervention but who had not previously participated in an LH interview. All interviews (n=28) were taped and transcribed; interviews were first translated into English before being transcribed.

Study staff and trained female Telugu-speaking research assistants obtained informed consent and administered the surveys and LH interviews in a Project Parivartan field office. Survey and interview completion time for participants ranged from 90 to 120 minutes. At the conclusion of both the survey and interview, participants were compensated for their time and costs associated with transportation in an amount determined to be appropriate with local norms. Participants were also provided with the option of being referred and accompanied to a local FSW intervention if distress was detected. Survey participants were given an additional incentive payment for each coupon that resulted in the successful recruitment of new study participants. The research protocol was approved by the Yale University Human Investigation Committee, the Duke University Health Systems Institutional Review Board, and the VHS-YRG Care Medical Centre Institutional Review Board in Chennai, Tamil Nadu, India.

Measures

The LH interviews addressed such topics as family background, childhood, and marriage; livelihood issues including food, housing, and debt; sex work trade issues including experiences with law enforcement; condom use; health care; and leisure-time activities. The following sections describe how we measured the variables used in the quantitative analysis.

Demographic variables.

We measured age continuously and grouped respondents into five age categories (ages 18–24, 25–29, 30–34, 35–39, and ≥40 years). Education was measured as primary education, secondary education (high school), or higher level, and literacy was measured by asking participants if they could read or write (yes/no). We calculated years as a sex worker by subtracting the age of entry into sex work from the respondent's current age. We also assessed respondents' marital status (never married, currently married, or not married but have a male partner), number of children (any children and in what age group [<5 or 5–18 years of age]), and financial support (solely support self or report support from others). In addition, FSWs were asked to report the venue in which they worked (brothel, street, lodge or hotel, home, highway, or other).

Debt.

Participants were asked whether they were currently in debt (yes/no) and from whom they had taken loans (moneylenders, friends, or other sex workers) in the six months prior to the survey.

Violence.

We considered two forms of violence: sexual and physical. We measured sexual violence by asking participants whether anyone forced them to have vaginal, anal, or oral sex against their will in the past six months (yes/no). Participants were grouped as experiencing physical violence if they responded positively to questions about whether they had experienced either (1) being beaten (e.g., hit, slapped, pushed, kicked, punched, choked, or burned) or (2) being threatened with a knife, gun, or other weapon, or having had a weapon used against them.

Indicators of HIV risk.

We measured whether participants received more money for sex without a condom by asking them whether this had happened in the past six months (yes/no). We measured anal sex trade by asking participants whether they had had anal sex with clients in the past 30 days. We measured consistent condom use by asking participants how often they had used condoms with regular and occasional clients in the past seven days; participants who reported “always” using condoms for each type of client and who reported using condoms the “last” time with these clients were categorized as consistent condom users. Participants were grouped as experiencing sexually transmitted infection (STI) symptoms in the six months prior to the survey if they reported experiencing any of the following symptoms: lower abdominal pain not related to diarrhea or menses, foul smelling vaginal discharge, burning while urinating, genital ulcers/sores, swelling in groin area, or itching.

Data analyses

For the qualitative analyses, ethnographers coded the interviews according to themes. We identified codes related to economic insecurity and debt, and conducted detailed readings of these sections, discussed relevant themes, and developed consensus on the patterns and themes reported.

For the quantitative analyses, we assessed sample characteristics to identify factors that differed between women reporting debt and those reporting no debt (using Chi-square contingency tables). We used logistic regression models to analyze debt in relation to sexual and physical victimization, condom use with clients, anal sex trades with clients, and reported STI symptoms. Sample characteristics associated with debt in bivariate analyses at p<0.05 were included in all adjusted models. For logistic regression findings, crude and adjusted odds ratios (AORs) are presented with associated 95% confidence intervals (CIs). Exploratory analyses further examined condom use among male partner types (husband, boyfriends or other male partners, and regular or occasional clients). We conducted all analyses using SAS® version 9.1.¹⁸

RESULTS

The context and types of economic debt

The majority (80%) of survey participants reported current debt; among those who reported taking loans in the six months prior to the survey (63%, n=426), the majority of women (74%, n=316) reported borrowing from moneylenders. Few borrowed from friends (15%, n=62) or other sex workers (16%, n=68). During LH interviews, women indicated that debt was often the result of seasonal variation in sex work and could be necessary for purchasing basic necessities, such as food and housing for themselves and their children, as these quotes demonstrate:

The money is not sufficient at home and when we go for the business, no one is calling us. We had no food for a week last month. There was a continuous rain for five days and we could not go for work.… What can we do, we have to pay 10 rupees every week as interest if we take [a loan].

When did we live life with money? There was not a day when I could go to sleep without the mental tension of paying the debts and never had the feeling of earning and living within the means without loans to repay.

In the rainy season, we don't have business and we have to starve … I did not have any money or rice at home. I … was already in debt.…

LH interviews also highlighted that debt often incurred further debt, as women took out new loans to pay off the old ones. Thus, managing debt appeared to be a complex process. While quantitative data did not specifically ask women whether they took loans to pay for other debt, these data showed that, on average, women reported taking from 1.7 different sources during the six months prior to the survey:

I took from four places 2,000 each and have to pay every week.… I have pressure from these finance ones. … If we don't pay … they calculate compound interest.… To pay we have to take loan again.

If we cannot pay for two months then we will have to pay 80 rupees and then 100 rupees at the end of two months.… Where can we get so much money? That way the interest is going on increasing. For the 6,000 I have taken, the interest has gone up to 50,000.

LH interviews suggest that debt specifically and economic insecurity more generally could also be reasons to initiate sex work. Quantitative data also support the finding that sex work is often initiated to provide economic security; 84% (n=564) of women reported initiating sex work to provide for themselves and their families. More specifically, during interviews respondents frequently connected their economic hardship and initiation of sex work to the departure or abandonment of their husbands and subsequent economic insecurity:

I had to [initiate sex work] for my children. My parents were not there and my in-laws did not help. I had loans to clear.…

I have no man. If a man were there, we can borrow money and he will clear it. Whether drinking or stealing or prostitution, everything is to satisfy the hunger of the stomach. If man is perfect, no woman will do things like this. I am doing this because I have no man.

It is after marriage and after children were born and he left me that I started coming out for work. I do it secretly for the sake of children, to provide food to them.

While women reported that the lack or loss of a male partner left them financially insecure and led to their involvement in sex work, our survey data also demonstrate that some relationships with male partners (relationships that women reported having while working as sex workers) also contributed to their debt and economic insecurity. Indeed, very few other demographic characteristics of FSWs were significantly associated with debt (Table 1). In terms of partnership status, women were more likely to report debt if they were currently married (p=0.03), had a male partner (p=0.01), or received financial support from others (i.e., if they were not the sole providers for their households) (p=0.0008) (Table 2). LH interviews help to situate these findings. Women reported that current male partners not only spent money women made from sex work, but also often accrued debt. While both LH interviews and survey data suggest that some women initiate sex work as a result of abandonment by husbands (as discussed in the previous section), these data also suggest that women may initiate sex work as a result of debts incurred by husbands or male partners, and that FSWs who have husbands or non-client partners may be at risk for greater economic insecurity as a result of having these male partners. Some examples of this effect include the following statements:

Table 1.

Characteristics of a sample of female sex workers (n=673) in Andhra Pradesh, India (total and by reported debt), 2007

^ap<0.05, Chi-square analysis

^bNumbers may exceed 100%; categories are not mutually exclusive.

Table 2.

Differences in reported debt by relationship factors and financial support among a sample of female sex workers (n=673) in Andhra Pradesh, India, 2007

A man told me that he would provide for my children and take care of the money I earned and put it in the bank: he started living with me as a [male partner]. He lived with me for six years and took all the money I earned in the six years and said I did not give him anything. It broke my heart and I stopped having lovers.

My husband was a drunkard. He raised loans everywhere.

In the recent times … [I struggle with my husband's] frequent drinking, beating, and raising loans and my paying them off. Recently I bought a box and started saving and my husband came and carried it away.

[My husband] borrowed 1,000 rupees for drinking alcohol … [We now have a] debt [altogether] of 20,000 rupees. I am dying with these debts.

Debt, violence, and HIV

Findings from logistic regression models (adjusted for partner status, education, financial support, and literacy) highlight that FSWs who reported debt were more likely to report the following: recent physical violence (AOR=2.4, 95% CI 1.5, 3.9), unprotected sex with occasional clients in the past week (AOR=2.3, 95% CI 1.2, 4.3), anal sex with clients in the past 30 days (AOR=2.0, 95% CI 1.1, 3.9), and at least one STI symptom in the past six months (AOR=1.6, 95% CI 1.1, 2.4) (Table 3). While qualitative interview responses did not specifically link debt to increased sexual risks for HIV, women often provided scenarios of how debt influenced the context—and urgency—of their work. In situations of economic duress, for example when women are struggling to afford basic necessities such as food for their families, it is likely that not using a condom may be more negotiable during these times.

Table 3.

Debt and its relation to violence and other HIV sexual risk factors among a sample of female sex workers (n=673) in Andhra Pradesh, India, 2007

^aIndented percent and n represent the proportion of participants who reported each HIV risk factor by whether participants reported debt. For example, a total of 12.2% (n=82) of the study sample reported sexual assault; among those with debt, 12.9% reported sexual assault, and among those with no debt, 9.1% reported sexual assault.

^bAll logistic and linear regression models were adjusted for relationship status (whether married and/or have a temporary partner, financial support, education, and literacy).

^cp<0.01

^dp<0.05

HIV = human immunodeficiency virus

OR = odds ratio

CI = confidence interval

STI = sexually transmitted infection

In the rainy season we don't have business and we have to starve … I did not have any money or rice at home. I … was already in debts and no one would give me credit then. I went to the highway in the rain with an umbrella and a veil, by walk, as I had no money for auto also. A driver stopped, seeing me, and asked where I would take him if he wanted to sleep with me … we went to the shrubs and in the rain. I spread the blanket on the ground but it was raining when he was doing it on the top and the rainwater under the blanket was flowing up from the pressure. There were such hard times and I had to do business under such circumstances also. I bought two kilos of rice … and 250 grams of oil and some vegetables and went home. We have such difficulties during rainy season.

DISCUSSION

A very large percentage (80%) of FSWs in this south India-based sample reported having some type of debt at the time of the survey, thus underscoring their high levels of economic insecurity. Our work provides one of the few analyses of this experience.

Interestingly, a greater proportion of FSWs who reported higher levels of literacy and education (particularly high school education vs. those with no high school education) also reported being in debt. While this finding was unexpected, it may indicate that women with higher education and literacy levels are more able to navigate resources to obtain loans and, thus, reported more debt. More research is needed to better understand which groups of FSWs are more able/likely to take on debt.

Also noteworthy, FSWs who reported being married currently, having a male partner, or receiving financial support from others (non-lenders) were more likely to report debt. This finding may partially reflect local standards where it is easier for men to obtain loans. However, our qualitative data suggest that male partners may aggravate women's financial burdens—buying alcohol or otherwise spending money on themselves (rather than serving as resources to help obtain loans for the needs of women and their children). Further, while FSWs in our sample were earning money through sex work, they may have less control over their earnings when married or living with a male partner due to gendered power dynamics that disfavor women within their intimate relationships. Previous research similarly indicated that male partners, largely due to these gendered norms, have more decision-making power regarding how money is spent; thus, money is more likely to be allocated to nonessential expenses (e.g., alcohol).¹⁹ Similarly, our findings suggest that women earning money in the sex trade, but not in control of spending that money, have greater challenges in supporting their families and ultimately have increased debt and associated HIV-related risk. Our finding that women who reported debt were more likely to experience violence is also consistent with previous work documenting that risk for violence relates to low levels of autonomy regarding household decision-making.²⁰ Future research must seek to clarify these complex issues.

Our analysis also indicated that risk for physical violence among FSWs was associated with the reporting of debt above and beyond relationship status. While underlying reasons for the heightened vulnerability to such violence need to be further explored, this finding may indicate that FSWs with debt are more likely to engage in riskier sex trades to make more money. For instance, women who face economic hardship may be more likely to agree to travel somewhere with a client (e.g., to a location where there is more than one client) to make more money; however, such instances of sex work may increase their vulnerability to violence. While a small proportion of FSWs in our sample reported working in brothels (13%), prior work with FSWs in multiple global contexts, including India,^21,22 has highlighted how debt bondage to employers or brothel managers can place women at risk for violence from brothel managers and male clients. Experiences of violence also render FSWs at risk for unprotected sex and other HIV risk behaviors.^4,6,7

Furthermore, consistent with an existing study among FSWs in Thailand that found a link between debt to an employer and HIV infection,²³ we found that FSWs who reported debt were more likely to indicate greater exposure to sources of HIV risk via unprotected sex with occasional clients and anal sex (riskier type of sex trade), or STI symptoms (having an STI can facilitate acquisition of HIV and also signals greater sexual health risk, including sexual risk for HIV²⁴). Having debt may attenuate the ability of FSWs to demand or negotiate condom use with clients. More work is needed, however, to understand why debt was found to be associated with increased unprotected sex among only certain types of clients (occasional clients but not with regular clients); notably, debt was not associated with reporting more engagements with either occasional or regular clients (data not shown).

Given that associations between violence victimization and unprotected sex have been well-documented in prior work among FSWs, and that the current study linked debt with experiences of violence as well as sexual risk behaviors, current study findings emphasize the inter-relation among debt, HIV sexual risk factors, and violence. Finally, given that women's reports of debt were associated with being married currently or having a male partner, women in debt may have increased risk not only as a result of increased unprotected sex with clients, but also as a result of their reported high rates of unprotected sex with their husbands or male partners. More work is needed to better understand how male partners contribute to women's debt, as well as to identify other factors associated with women's debt that are ultimately heightening women's risk for HIV and violence (e.g., substance use may play a role but has not been well-investigated in this geographic region).

Limitations

The present findings must be considered along with study limitations. The cross-sectional design did not establish the temporality of these associations, and did not allow prospective follow-up of FSWs, a challenge in studies with hard-to-reach populations. Additionally, the items used for analyses relied on self-reported responses; stigma can often result in underreporting of sensitive issues or socially undesirable behaviors.^25,26 However, such underreporting would decrease power to detect an association between debt and violence or other self-reported outcomes, and the current study found various strong associations among these factors.

Future assessment of violence experienced among FSWs would benefit by exploring varying effects by perpetrator type; however, our study did not have a sufficient sample size to investigate this issue further. Our assessment of self-reported STI symptoms could be improved in future studies with the use of biological markers to detect such STI infections. Additionally, it is not known whether the sample, recruited by RDS, is truly representative of the underlying population of FSWs in this geographic location. The sampling strategy could have introduced bias if selection of participants was based on some factor relevant to our study outcomes; while we cannot guarantee that we have reached all groups or acquired a sample truly representative of the underlying population, previous studies have found RDS to be the best method to date for sampling this population.^27–29 Also, our study findings focused primarily on associations between variables (rather than prevalence estimates) and included qualitative data as well to help support these findings. Selection of LH interview participants was based on participation in the survey and, therefore, participants were not chosen randomly. However, participants were selected based on their willingness to talk, rather than any variable thought to be associated with study outcomes. Current study findings are most applicable to populations of FSWs working in Rajahmundry, Andhra Pradesh, and may not be generalizable to larger populations of sex workers from this or another Indian state.

CONCLUSION

Our study findings highlight the violence and HIV-related vulnerability of FSWs who report debt, and FSWs who report debt represent the majority (80%) of our sample. Our research further illustrates how male partners may be contributing to FSWs' debt and economic insecurity (as well as associated risk for HIV and violence), rather than providing needed economic resources for women. More research is needed to determine types of partnerships women have with non-clients and how FSWs might best ensure that their own economic stability is not threatened by maintaining such relationships, and, in turn, furthering their risk for HIV and violence.

Our findings also have implications for programmatic strategies related to reducing FSWs' high HIV-risk behaviors with clients. Programs that increase economic opportunities for women have often been theorized as a way to decrease initiation into sex work, but our findings suggest that such programs may also reduce risks for HIV and violence among women already working as sex workers. Such findings are important for HIV prevention among this population and emphasize the need for future interventions to consider challenges within women's life contexts when implementing strategies to promote safer sex trade engagements.