‘If you have children, you have responsibilities’: motherhood, sex work and HIV in southern Tanzania

Abstract

Many female sex workers begin sex work as mothers, or because they are mothers, and others seek childbearing. Motherhood may influence women’s livelihoods as sex workers and their subsequent HIV risks. We used qualitative research methods (30 indepth interviews and three focus group discussions) and employed Connell’s theory of Gender and Power to explore the intersections between motherhood, sex work, and HIV-related risk. Participants were adult women who self-reported exchanging sex for money within the past month and worked in entertainment venues in southern Tanzania. Participants had two children on average, and two-thirds had children at home. Women situated their socially stigmatised work within their respectable identities as mothers caring for their children. Being mothers affected sex workers’ negotiating power in complex manners, which led to both reported increases in HIV- related risk behaviours (accepting more clients, accepting more money for no condom, anal sex), and decreases in risk behaviours (using condoms, demanding condom use, testing for HIV). Sex workers/mothers were aware of risks at work, but with children to support, their choices were constrained. Future policies and programming should consider sex workers’ financial and practical needs as mothers, including those related to their children such as school fees and childcare.

Introduction

A recent meta-analysis showed that female sex workers in sub-Saharan Africa had 12.4 times higher odds of having HIV than women in the general population (Baral et al. 2012). Female sex workers in Tanzania also have high HIV prevalence (30% in Dar es Salaam [NACP (National AIDS Control Programme) 2012]) and incidence (13.9/100 person-years in Mbeya [Riedner et al. 2006]). Data on HIV prevalence among sex workers in the study site, Iringa and Njombe Regions, is unavailable, but the regions have the highest HIV prevalence in Tanzania (Iringa, 9%; Njombe, 14%) (TACAIDS (Tanzania Commission for AIDS), Zanzibar AIDS Commission (ZAC), National Bureau of Statistics (NBS), Office of the Chief Government Statistician (OCGS), and ICF International 2013).

Additionally, an estimated two-thirds of sex workers in sub-Saharan Africa have children (Scorgie et al. 2012). Financially supporting children is reported as a major motivation for women to participate in sex work, due to comparatively high pay (Kerriganet al. 2001; Murray and MODEMU 2002). Furthermore, some sex workers have reported seeking childbearing through sex work (Kerrigan et al. 2003; Wayal et al. 2011), perhaps as an exit strategy (Murray and MODEMU 2002; Haram 2003).

Despite evidence that sex workers are often mothers and at high HIV risk, little is known about how motherhood might influence HIV-related risk behaviours. In this paper, we employ Connell’s theory of Gender and Power to explore the intersections between motherhood, sex work and HIV-related risk behaviours among sex workers in southern Tanzania.

Gender, power, motherhood and sex work

Connell’s theory of Gender and Power (Connell 1987, 2002) posits four structures of gender – labour, power, emotional, and symbolic relations – which ideologically shape ‘women’ and ‘men’, ‘masculinity’ and ‘femininity’ (Connell 2002). These structures constrain society, but are (re)constituted through practice (Connell 1987) and are thus ‘vulnerable to major changes’ (Maharaj 1995, 53). The structures are used to analyse various institutional and socio-historical contexts (Maharaj 1995); we use the four structures to analyse motherhood and sex work.

Labour structures (the gendered divisions of work) are seen in motherhood, where the ‘ideological construct of mothering’ envisions the ‘mother absorbed in nurturing activities’ at home with her children (Arendell 2000, 1195). In Tanzania, mothers are expected to engage in childcare and household tasks, with fathers providing financial support (Kiaga 2007). Though not particular to motherhood, gendered labour divisions are stark in Tanzania, where women occupy more unskilled positions than men (NBS 2011) and employees are selected for particular tasks ‘suitable’ to their gender (Fischer 2014). Gendered labour structures also apply to sex work. Firstly, sex work is work (Delacoste and Alexander 1987), and its labour structures, particularly its criminalisation, lead to occupational health issues (Alexander 1998). The lack of economically viable employment opportunities for women has been identified as a major reason women enter sex work in sub-Saharan Africa (Scorgie et al. 2012). Furthermore, it is generally men who manage, regulate, and police sex work (Delacoste and Alexander 1987). Even in sub-Saharan Africa where women work independently (Scorgie et al. 2012), the gendered division of labour remains: it is mostly men who buy and women who sell sex.

The gendered divisions of power structures of ‘authority, control, and coercion’: (Maharaj 1995, 60) manifest themselves in women’s (mothers’) low control over household decisions in Tanzania (NBS 2011) and in laws and policies, which favour fathers in inheritance, divorce, and custody (Kiaga 2007). African feminism, however, argues for ‘motherism’ whereby motherhood is seen as empowering and powerful rather than repressive (Akujobi 2011). Power structures are seen also in some sex workers’ HIV- related vulnerability to disease and violence, including their inability to negotiate condom use (see Choi and Holroyd 2007; Pack et al. 2014). Although male clients often wield power over sex workers, much HIV prevention work addresses often disempowered women, without also addressing their clients (see Barrington et al. 2009).

The structure of emotional relations (affect, family, gender roles, and attachments) (Connell 2002), ‘describes how women’s sexuality is attached to other social concerns’ such as ‘impurity and immorality’ (Wingood and DiClemente 2000, 544). In Tanzania, this manifests itself in how motherhood is essential to women’s respectability (Haram 2003), adulthood (O’Malley 2002), and womanhood (Haws 2009). Tanzanian women are imagined to be asexual before marriage, and monogamous within it. Women with multiple partners are labeled malaya, ‘prostitutes’, while men are encouraged to have multiple partners (Wight et al. 2006). Expectations about women’s sexuality certainly are attached to debates over sex work (Delacoste and Alexander 1987) and contribute to stigma and violence against sex workers (Scorgie et al. 2012). Women who are both sex workers and mothers structure their emotional relations in various ways. For example, Indian sex workers have asserted being ‘mothers first’, subsuming their sex work to mother-child attachments (Basu and Dutta 2011). Ugandan women reported preferring more complexity and fluidity, inhabiting multiple relational identities at once: mothers, but also ‘wives, partners, friends, and workers’ (Zalwango et al. 2010, 89).

Structures of symbolic relations (socially constructed, gendered interpretations and meanings) certainly apply to motherhood and sex work. Just as one becomes a woman by practising societally expected femininities (Connell 2002), one becomes a mother by practising motherhood, not by giving birth. Thus, motherhood differs across societies, and ‘African motherhood’ is a ‘respected and mythologised’ institution (Akujobi 2011, 6). This contrasts with sex workers, who are stigmatised and blamed for HIV (Scambler and Paoli 2008); they symbolise and embody HIV risk, rather than being seen as human beings at risk. The structure becomes complex and even contradictory when women are both mothers and sex workers. Castañeda et al. (1996, 233), argued that women in Mexico faced a ‘societal schizophrenia’ as both ‘mother’ and ‘prostitute’ in a society that denied that one person could be both. Similarly, in the Dominican Republic, sex workers symbolically separated street from home relationships (Kerrigan et al. 2001).

These structures are not static and ahistorical; indeed, Connell (1987) emphasises gender’s historicity and malleability. Amadiume (1987) argues that gender has changed dramatically since pre-colonial times in Africa. Arguably, Tanzania’s history, from the slave trade, missionary movements, and colonialism, to post-colonialism, structural adjustment programmes, globalisation and the HIV epidemic (Shillington 2005), has influenced the current practice of gender. That women interviewed in this study referred to husbands/fathers as breadwinners and wives/mothers’ place as in the home speaks as much to missionary Christianity as to ‘African tradition’, if not more so (Amadiume 1987).

There is some evidence that these complexities translate into HIV-related risk behaviours as well; sex worker mothers in India were less likely to use condoms consistently (Reed et al. 2012), while other mothers reported positive healthcare-seeking behaviours for the sake of their children (Basu and Dutta 2011). How these complexities play out in sex worker mothers in southern Tanzania is the focus here.

Sex work in Tanzania

To understand the context of the participants’ experiences, we will begin with a brief overview of the nature of sex work in the area.¹ Although sex work is criminalised in Tanzania (GofURT 1981), it is widely tolerated. Many sex workers seek clients in entertainment venues such as bars, where some are employed at low wages. Sex work here is independent; there are no managers, and the worker and client alone determine price, sex acts, etc. The study sites, Iringa and Njombe Regions, are characterised by the Tanzanian-Zambian highway and multiple plantations. These bring migrants and transport workers, who comprise a significant portion of sex workers’ likely high-risk clientele. While we observed that there were no clinical services specifically for sex workers, there were limited interventions targeting sex workers (condom demonstration and distribution, limited bar-based outreach HIV testing, and an economic empowerment intervention for vulnerable women) (Beckham et al. 2013).

Methods

Between February and July 2012, we conducted a qualitative study to explore sex workers’ experiences with motherhood. Data included in-depth interviews with 30 sex workers and three focus group discussions (FGD) with 22 workers (six, eight, and eight participants). The analysis presented here was one part of a larger strategic assessment to define a comprehensive response to HIV in the area.

The second author, an experienced, university-educated, Tanzanian woman, underwent a two-week training course on qualitative research and ethics. After gaining permission from local leaders, she approached the owners/managers of entertainment venues, explained the purpose of the study, and requested approval to recruit there. Women were approached directly, informed about the study, screened for eligibility and asked if they were interested in learning more about the study, and if so, began the informed consent process. Additional women were recruited through snowball sampling.

Women were eligible to participate if they were at least 18 years old, self-reported exchanging sex for money in the past month, and worked in entertainment venues. The interviewer read and explained an informed consent form and answered any questions before obtaining oral consent. Oral consent was used to protect confidentiality. Each participant chose a time and place for the interview that was private and convenient for her (e.g., a back room at the bar or a private residence).

We used demographic and phenomenal purposive sampling (Sandelowski 1995) to maximise variability in ages, locations, venue types, numbers of children, and living situation (children did/did not live with mother). For interviews and focus group discussions, we recruited in different urban and semi-urban locations across the regions in order to maximise any differences in types of work and clientele. For the 30 in-depth interviews, we developed an interview guide with open-ended questions to direct questioning, and the interviewer probed extensively. Topics included daily life, work experiences, reproductive history, motherhood, childcare, paternity, relationships, healthcare seeking and health issues including HIV prevention and treatment. A similar guide was used for the three focus group discussions, which were conducted in Iringa, Ilula, and Makambako.

Interviews were conducted in Swahili, audio recorded, and transcribed verbatim. To protect confidentiality, we allocated pseudonyms to all participants. Within 24 hours of each interview, the interviewer wrote up a summary of major topics, themes and methodological issues. We held weekly meetings during the course of data collection to discuss findings, troubleshoot, and monitor sampling variation.

The first author read all transcripts in Swahili multiple times and added notes. Additionally, we created a matrix to facilitate comparisons across participants on key quantifiable characteristics including location, venue type, age, education, reported HIV status, current and past intimate/marital partnerships, condom use, and information about children (numbers, ages, living situation, paternity). We used Stata/SE (version 12.0, StataCorp, College Station, TX) to calculate summary statistics.

Next, the first author thematically coded the Swahili transcripts using Atlas.ti (version 7.0, Scientific Software Development GmbH, Eden Prairie, MN). She began the analysis with incident-by-incident coding to identify sections concerning motherhood, child-rearing, and children. She then conducted initial line-by-line coding on five transcripts, then used axial coding to define categories and sub-categories among these initial codes. We identified one of these categories, how motherhood and childrearing affect sex work, for further development in this paper. The first author then conducted detailed coding on sections previously coded as relevant to motherhood within all transcripts, allowing themes to emerge from the data. The first author also wrote memos on key themes throughout analysis to develop ideas (Charmaz 2006; Friese 2011). The first author translated the quotations, and the second author checked them.

The Johns Hopkins Bloomberg School of Public Health, Muhimbili University of Health and Allied Sciences, and the Tanzania National Institute for Medical Research granted ethical approval for this study.

Results

Participant demographics

Participants were recruited from five towns across the two study regions, and ranged in age from 20 to 40 years (mean 28.9). They worked in various venue types, mostly in bars (baa and grosari) that sold bottled alcohol, or small clubs (kilabu) that sold locally brewed alcohol (22/30). None were currently married but about half reported intimate partners. Three- quarters had primary school education (Table 1). Three women indicated they were living with HIV. Women had a median of two living children, and median age of children was seven years. At least three women were breastfeeding. Three women did not currently have living children. Of those with living children, two-thirds had at least one child living with them. Of these, 12 had children living alone with them, and others had help with childcare.

Table 1.

Demographic characteristics of female sex workers (n = 30).

Characteristic	Value/Type	n	%*
Age (mean = 28.9)	20–24	8	26.7
	25–29	10	33.3
	30–34	7	23.3
	35–40	5	16.7
Venue	Baa (large bar)	7	23.3
	Kilabu (bar with local brew)	14	46.7
	Grosari (small bar)	1	3.3
	Truck Stop	1	3.3
	Multiple Sites	7	23.3
Town (Region)	Iringa (Iringa)	12	40.0
	Ilula (Iringa)	9	30.0
	Mafinga (Iringa)	3	10.0
	Kilolo (Iringa)	1	3.3
	Makambako (Njombe)	5	16.7
Education	None	2	6.7
	Primary School	23	76.7
	Some Secondary School	5	16.7
Marital Status	Never Married	17	56.7
	Divorced/Separated	10	33.3
	Widowed	3	10.0

^*Percentages may not sum to 100 due to rounding.

Motherhood affected sex work and HIV risk in complex ways. We present three main themes which demonstrate these dynamics: ‘motherhood/respectability vs. sex work/ stigma’; ‘for the children’; and ‘motherhood, power and HIV risk’. Participants felt that motherhood brought social respectability, while sex work brought social stigma. As both mothers and sex workers, they situated their socially stigmatised work within their respectable identities as mothers caring for their children at any cost, including risking HIV transmission.

Motherhood/respectability vs. sex work/stigma

Participants idealised motherhood, and directly associated it with respectability (heshima) and womanhood, which contrasted sharply with the stigma they experienced in relation to sex work. Several women referred to the status and respect women gained as mothers. As Sofia (26, Ilula, no children) stated: ‘To be called Mama is respectability.’ Naiwa (28, Mafinga, 3 children away) elaborated: ‘[Motherhood] brings respect… When you are called Mama of So-and-So, already you are confident that “I am a woman who is completed.”‘

While women focused on the respectability of motherhood, they demonstrated the stigma they faced as sex workers. Although few women used the Swahili term for stigma (unyanyapaa), they expressed similar constructs through phrases such as ‘they isolate us because we are very deviant’ (Beatrice, 31, Ilula, 2 children at home), and ‘in the community we are very despised’ (FGD, Makambako, Participant 5). Pamela (23, Ilula, 1 child away) explained: ‘I cannot tell [others] the work that I do. It is shameful to be selling in a bar. It is indeed work, but it’s shameful… A person comes and acts crass, [saying], “You are just… a barmaid, a whore [malaya].”‘

As a result, women often hid their stigmatised work by emphasising their motherhood. Most maintained that they went about their work in ‘secrecy’, ‘hiding themselves’ under the guise of respectable mothers and ‘regular women’. In contrast to the miniskirts, tight clothing, and other ‘shocking attire’ they wore when seeking clients, women reported dressing ‘with respect’ and ‘like a mother’ in long skirts, loose blouses and wraparounds at health centres and other public settings. This helped them be ‘treated like any other woman’. Dressed this way, ‘They cannot know I do this work’ (Sofia, 26, Ilula, no children). In part because society stigmatised sex workers, women negotiated their social identities as respectable mothers, doing whatever was necessary to care ‘for the children’.

For the children

Participants across sites described the ‘ideal’ mother as one who stayed at home with her children, and was financially supported by their father. However, many participants had sole responsibility for their children’s material needs. As Estelle (23, Iringa, 2 children at home) expressed, ‘If you have children, you have responsibilities.’ Women saw this as extending their role as mother to encompass the role of father/breadwinner. As Mpiluka (33, Iringa, 1 child at home) explained, ‘I am the mother and the father.’ In these circumstances, many found whatever work they could, including bar work, daily wage labour, farming, selling alcohol or prepared foods, and sex work. Mwajuma (34, Iringa, 1 child at home), who sold local brew for a USD 1.25 profit per 20-litre bucket, explained how she supplemented her meagre income through sex work in order to support her child: ‘Now you have children at home … will that money be enough? Really, it will not be enough. I mean … you purposely get clients and get money another way.’

Given both the economic realities of single motherhood and the stigma against sex work, women often framed their work as for the children. Mariam (22, Ilula, 1 child away) explained: ‘They do this … because they have small children… They do this so that their children live well and do not get problems. They get money so that they fulfil their children’s needs and their needs.’

For some women, framing their work as being ‘for the children’ normalised it, and they reported feeling good since it enabled them to take care of their children. For example, Evangeline (27, Mafinga, 2 children at home) felt that society treated sex work ‘like normal work’ because it was for the children. She explained: ‘It [society] treats it like it’s normal work, because, for example, I myself, I do that work and I educate and take care of my family.’ Mpiluka (33, Iringa, 1 child at home) reported feeling good about sex work because it helped her feed her child: ‘No, I cannot feel bad, because when he eats, he gets full, you understand?’

One woman represented a negative case here; her work was not oriented around her children. Rather, ‘I have no goals other than to buy make-up, clothing, and food’ (Shida, 28, Iringa, 1 child at home, 1 away). Shida never wanted children, but bore two to avoid the stigma of childlessness. Nevertheless, she connected her sex work to feeding her child. When asked how she felt about bringing her breastfeeding infant to work with her, she replied: ‘Now how will I feel, my sister? Don’t I feel just great, because have I not gotten money for food? I have eaten and I feed the child well, she is developing great. I feel just great.’

Motherhood, power, and HIV risk

Being mothers affected sex workers’ negotiating power, compromising their power vis-á-vis clients in particular ways, while increasing it in other ways. This in turn led to both reported increases and decreases in HIV-related risks. First, motherhood constrained women’s power in certain ways. Women reported that clients’ offers were especially hard to refuse when mothers considered their children’s daily needs. Many of the women lived hand-to-mouth, relying on income from clients to supply their children’s daily meals. Estelle explained her dire situation of facing her hungry children with no money to buy them breakfast, which compromised her ability to refuse clients:

It’s like this in raising children. You will be shocked one day you wake up and truly you don’t have even 10 [shillings], and right there you look at your child… Oh! Tsk, that’s it, if a person happens to tell you ‘If you do this, I’ll give you such-and-such amount’, will you refuse? And if you look at home, you have no money? You agree just like that. (24, Iringa, 2 children at home)

Women reported that the constraints of having children led to increases in HIV-related risk behaviours, such as seeking more clients and having unprotected vaginal and anal sex. Women specifically sought clients when they needed cash for their children’s expenses, since sex work paid relatively well, e.g., TZS 400–1000 (USD 0.25–0.60) for daily wage labour vs. TZS 5000–50,000 (USD 3–30) per client. For example, Mpiluka (33, Iringa, 1 child at home) increased her client load as necessary to pay secondary school fees for her son. She explained: ‘Even at the bar I wasn’t drinking [to save money]… You find the college students are drunk, [I get] thirty [thousand shillings for sex] or whatnot. That’s how I send my child to school.’

Likewise, many mothers accepted clients’ requests for unprotected sex, since this paid up to double the price, and they had children to feed. A FGD participant described this type of situation, where she felt powerless to refuse a client when she considered her hungry children at home:

[I told him,] ‘I ask [that we use] a condom’. [He answered] ‘A condom for what? Didn’t you already take my money, so what do you want a condom for? I’ll increase it with this other [money], take this here’. If I see I have children there at home dying from hunger … it’s just necessary that I agree and be hung out to dry. (FGD, Makambako, Participant 8)

Another compromise women made for their children was agreeing to anal sex. Although many said they preferred to avoid anal sex, some agreed because it fetched higher prices, and they needed money for their children:

You can meet with one who says, ‘I don’t need, maybe, to have sex in front’. Oh! He wants to do it behind?! So doing it like that just hurts, [but] because you want money, what will you do? And you have thoughts about your children … The children have no father, so you find that a thing like this just forces you [to agree]. You give it, so that you get money. (Angel, 22, Ilula, 2 children at home)

In contrast, there were ways in which motherhood empowered women in their work. First, women could appeal to their motherhood to earn extra money from clients. For example, sometimes truckers asked women to accompany them on long trips. Mothers with children living away with relatives had more freedom to accept these multi-day trips, but women with children at home had to consider the costs of leaving them. Some capitalised on the opportunity to ask for additional money to leave with the children. Chuki explained:

Maybe he’s a driver, you will hear, ‘I’m going on a trip to Tunduma [Zambian border], can you provide me company?’ You tell him, ‘All right, but, my friend, I have a family … If we leave 10,000, 15,000, it can be enough for them’ … He gives you maybe 15,000 or 20,000, and you leave it with the family. You go with him. (38, Makambako, 2 children at home)

Sex workers also used their motherhood to leverage sympathy from the police in disputes with clients, at least in one town. Women across sites reported how clients refused to pay or reduced the payment amount after their encounter. Some women felt they could not approach police about such theft because sex work was criminalised, but women in Makambako reported that police were helpful and sided with the women in such cases. In one case, the police ordered the client to pay the worker for the services she had given, partly because she had children:

If we go to the [police] station, they speak like this [to the client], ‘This young woman had a problem that made her agree … You need to give her that money so that it can help her. It will be that she has children. She is supposed to prepare their breakfast right now. Now what will she cook that breakfast with?’ So they help us at the station. (FGD, Makambako, Participant 4)

Additionally, motherhood served as a risk-reduction strategy for some sex workers; mothers wanted to live to see their children grow. Thus, motherhood was a key motivation for women to protect themselves from acquiring HIV, or from developing or worsening their AIDS. Lusajo (35, Iringa, 2 children at home, 1 away), for example, trusted no sexual partners in ‘these times of AIDS’ because she had a family. She said: ‘There is no one whom I trust… These are the times of AIDS. You need to be there to continue to watch your children, family. Yes, isn’t it family?’

Participants were well aware of HIV transmission risks at work, and awareness of condoms as a prevention strategy was high (28/30 interviewees used condoms, though inconsistently). Mothers explicitly connected condom use to their desire to protect their health so they could raise their children. For example, Chuki (38, Makambako, 2 children at home, 2 away) stated:

I try really hard to be attentive because I am a mother, and I have a family. I am supposed to take care of my family until … my time to live has ended. I try very hard to use condoms.

Likewise, Sara explained how she appealed to her motherhood to convince a regular client to use condoms:

I cannot go with him without a condom. Truly, I told him, ‘Bwana, hey, the contraceptive is right here’. He doesn’t want it, even if I have a problem [HIV]. [I tell him] ‘Bwana, go.I cannot lose my life. I cannot leave my children behind for your sake. (39, Makambako, 2 children away)

The desire to use condoms did not necessarily translate into actual use; clients reportedly resisted this male-controlled method often. Women commonly complained about clients resisting condom use - and forcing sex without condoms.

Having children who depend on them also increased women’s desire to test for HIV. Agnes described how her child figured prominently into her decision-making about testing:

Now then, I was keeping in mind that first I have a small child. That child still depends on me … If I die today, my child will suffer, and here I am without a father and without a mother … And I went in and I tested. (28, Iringa, 1 child away)

For women like Agnes, testing was the first step to accessing treatment for HIV, enabling them to live longer and thereby raise their children to adulthood and independence.

These results demonstrate how motherhood intersects with sex work in often complex and nuanced ways. Women negotiated their stigmatised work lives within the context of their broader lives as mothers striving to care for their children, which in turn lead to both increases and decreases in HIV-related risk behaviours.

Discussion

Gender, power, motherhood and sex work

In this paper, we have examined how motherhood affects sex work and sex workers’ HIV- related risk behaviours. We will now apply Connell’s theory of Gender and Power to illuminate the gendered structures that patterned the lives of women who are both mothers and sex workers in Tanzania. We do this through exploring how the four structures of gender (labour, power, emotional relations, symbolic relations) intertwine in the lives of women.

Gendered labour structures affect women’s lives in multiple ways. First, women reported that an ‘ideal’ (married) mother stays at home, attending to children and housework. None of the women here, though, lived this ‘ideal’. Single motherhood was financially difficult for them; female-headed households in Tanzania are more likely to be poorer and less likely to have enough food (Katapa 2006). Additionally, labour options are limited for women in the study site. Fewer women than men finish secondary school (16.2% vs. 22.9% nationally), and twice as many women as men have no education (NBS 2011). These structures perpetuate inequalities in Tanzania, and make meeting basic needs more difficult for women than men. Thus, women filled certain positions gendered as ‘women’s work’, such as bar work, but these positions paid women too little to support families. Mothers especially were attracted to the high pay and flexible working hours of sex work, increasing client load as necessary to fulfil children’s financial needs.

Furthermore, the women’s situations in this study reflect the gendered division of power. Gender power figures prominently in other writing on sexuality (Pettifor et al.2012) and sex work (Choi and Holroyd 2007; Aubé-Maurice et al. 2012); what this study adds is how gender power structures sex work in particular ways for mothers. In the socially expected ‘ideal’ husband-wife relationship in this setting, women have relatively little power over household decision-making. In parallel, clients wielded considerable power over sex workers, first plying them with beer and food to demonstrate their earning power, then later sometimes denying them payment, refusing to use condoms, or threatening and using violence. This was especially true for sex workers who were mothers, who felt constrained in their decision-making because they had hungry children at home. However, women also demonstrated ‘motherism’ (Akujobi 2011), leveraging their roles and respectability as mothers to regain some power vis-á-vis clients and negotiate for more money and demand condom use.

This study also demonstrates gendered emotional relations. For women in this study, the dominant emotional attachment was with their children. This bond was so strong that they would do anything for them, including going against prevailing norms about sexuality and face stigma. This echoes how sex workers/mothers in Costa Rica justified their work to themselves and others as ‘for the children’, giving them commodities and opportunities they themselves lacked as children (Rivers-Moore 2010). In a financial sense, it was, indeed, for the children; women in this study paid for their children’s food, clothing, and schooling. Symbolically, though, ‘for the children’ was a justification to themselves and others that their work was legitimate, despite being criminalised, ‘shameful’, and ‘dangerous’. By emphasising that work was ‘for the children’, sex workers were claiming the status and respectability of motherhood and disassociating themselves from the stigma attached to their work.

Lastly, symbolic relations figure prominently in this study, especially in what it means to be a woman, a mother, a wife, and a sex worker. Sex work was associated with AIDS and death, drunkenness, shame and promiscuity; it meant scorn. Motherhood, on the other hand, was associated with womanhood, domesticity, marriage and childrearing; it meant heshima. Heshima is a complex cultural concept meaning ‘dignity, honor, and respect’ (McMahon 2005, 95), but also can be translated as fame, renown, or standing in society. Heshima involves taking one’s place in life, in a family, in society, and in history; it is also a salient gendered moral principle throughout Tanzania (O’Malley 2002; Haram 2003; McMahon 2005; Fischer 2014). In this sense, motherhood allows women to take their place, or assume their proper role in a family. Women in this study, as in northern Tanzania (Haram 2003), used heshima, which they sought through motherhood, to counteract the stigma they felt from society. Like in Costa Rica, ‘motherhood is central to sex workers’ ability to combat stigma’ (Rivers-Moore 2010, 722). Educating, feeding, and clothing their children was a way to gain the respectability that was being denied them by societal views on proper womanhood, single motherhood, divorce, and sex work. By dressing in particular ways, ‘with respect’, and ‘like a mother’, by ‘hiding’ their sex work – by practising motherhood – they desired to and were able to appear as ‘regular women’ in society. For these women, being a mother and a sex worker meant taking care of the children, at whatever cost, including social sanction and risks to health.

Motherhood and HIV risk

Our results indicate pathways though which motherhood directly increased HIV risk for sex workers, such as agreeing to unprotected vaginal and anal sex for more money. Any sex worker, mother or not, might make similar choices for a variety of reasons (or be denied the choice by a client). However, mothers in this study explicitly connected their HIV-related decision-making to their children. In a choice between facing their hungry children tomorrow or risking HIV transmission, women chose to feed their children. This pathway to HIV risk needs further research to understand how often this happens, and to what degree mothers may be at higher risk than non-mothers.

Sex workers in Kolkata, India, also reported having to choose between earning money to feed their children and safeguarding their own health (Basu and Dutta 2011). In Andhra Pradesh, India, sex workers with ‘motherhood challenges’ were significantly less likely to use condoms consistently, and more likely to accept more money for no condom during sex (Reed et al. 2012). Reed et al. (2012), suggested it was the financial and temporal burdens of motherhood that affected risks. These mechanisms were suggested by the present study, with reported increased HIV risk-taking linked to the financial burden of children.

In contrast to Reed’s analysis, the women in this study also reported ways in which motherhood decreased their HIV-related risk behaviours. Motherhood helped women negotiate for higher prices and gain sympathy from police. They also reported using condoms and testing for HIV to preserve their health for their children’s sake, and negotiating for condom use by mentioning the children’s welfare. Similarly, sex workers elsewhere reported striving to protect their health so that they can stay healthy as their children grow (Murray and MODEMU 2002; Rivers-Moore 2010; Basu and Dutta 2011). It is possible these relationships with their children translated into safer HIV-related behaviours. Mechanisms through which motherhood reduces HIV-risk behaviours should also be considered for further research to test whether, for example, desires for increased condom use (for the children) translate into actual increased use. Future research must also consider the complexities here: motherhood may both increase and decrease HIV-related risk behaviours.

Limitations

This study has limitations. The data are cross-sectional; a longitudinal approach would have allowed more time to ask follow-up clarification questions and delve more deeply into complexities. Additionally, it is possible that women made judgements about what they thought the interviewer wanted to hear, which could have influenced their answers. Women did, however, freely reveal stories about not using condoms, suggesting they felt able to reveal what they ‘really’ did, rather than just what they were ‘supposed’ to do.

Programmatic implications

This study has implications for future intervention programmes for sex workers. First, the women in this study interlaced discussions of their work lives with references to their children, emphasising their respectability as mothers. As demonstrated by research in India, they were ‘mothers first’ (Basu and Dutta 2011, 107). Standard interventions with narrow focus on HIV and condom use, as are common in Africa (Chersich et al. 2013), may be inadequate to meet women’s needs given their constraints as mothers, and may contribute to stigmatisation (Scambler and Paoli 2008). Programmes should recognise the multifaceted aspects of sex workers’ lives as workers, mothers, partners, and human beings, and how all of these aspects, roles, and identities impact each other in complex ways.

Second, many sex workers in this study supported their children independently. Thus, ways to reduce the financial burden of motherhood, such as connecting sex workers to existing financial services, could help sex workers identify ways to earn more money for their children’s basic needs, help them access savings and loans programmes or form informal savings cooperatives among fellow workers. Significantly, many of these types of programmes already exist, but sex workers face greater barriers in accessing services due to criminalisation and stigmatisation (Scorgie et al. 2012) Programmes could assist sex workers in navigating these services, and advocate for their rights to access them.

Third, sex workers in this study were especially concerned about their children’s educations, and reported increasing their risk behaviours in order to pay school fees. Sex workers in Sonagachi, India also placed high priority on their children’s educations, so advocating for their children’s educational rights became a component of that successful programme (Jana et al. 2004). Programmes could support children’s education through scholarships for the children’s sake, but also to empower women to refuse unprotected sex and unwanted encounters. Fourth, childcare responsibilities constrained the times women could work, which may have increased HIV-related risk behaviours, as suggested by Reed et al. (2012). Thus, programming should consider strategies to address these needs, such as organising 24-hour childcare, as in India (Jana et al. 2004; Basu and Dutta 2011), or encouraging cooperative childcare among workers.

Finally, the World Health Organization (WHO) recommends community mobilisation and empowerment-based interventions among sex workers (WHO 2012), and such interventions have been successful at reducing HIV risk, decreasing stigma, increasing women’s self-efficacy, and advocating for legal and human rights (Kerrigan et al. 2013). Successful empowerment-based interventions in India credit part of their success to reacting to women’s needs as mothers, creating childcare services and advocating for children’s educational rights (Jana et al. 2004; Basu and Dutta 2011). Likewise, data in this study indicate that responding to sex workers’ needs as mothers may be a key component to reducing risk.

Conclusions

Sex workers who were mothers were aware of risks they could encounter at work, but with children to care for, they were constrained in their choices. They explicitly connected their HIV-related decision-making to their children, which led to increases in risk behaviours, such as unprotected sex; their children’s hunger overcame their worries about health risks. In contrast, women appealed to their roles as mothers to leverage gendered power to achieve their ends, and sought ways to protect themselves from disease for their children’s sakes, including using condoms and testing for HIV. Given the high proportion of sex workers who are mothers in sub-Saharan Africa, programming could be greatly enhanced by recognising female sex workers’ needs as mothers with children to support, in addition to more traditional HIV prevention and treatment strategies.