The influence of cash transfers on engagement in transactional sex and partner choice among adolescent girls and young women in Northwest Tanzania

Abstract Cash transfers are theorised to reduce adolescent girls and young women’s economic vulnerability and engagement in transactional sex; however, the processes involved remain unclear. We conducted longitudinal cross-sectional, and narrative timeline in-depth interviews with young women aged 15–23 years in northern Tanzania enrolled in a cash transfer intervention. We analysed data using a thematic approach guided by the transactional sex framework. We summarised data and compared findings to the domains of the framework. Participants reported relationships in the ‘sex and material expression of love’ domain characterised by emotional intimacy and loose ties between material support and sex, and the ‘sex for basic needs’ domain characterised by limited emotional intimacy, economic vulnerability, and a clear exchange between material support and sex. Some participants expressed that cash transfers provided business capital and savings which enabled them to decrease ‘sex for basic needs’. Cash transfers influenced transactional sex engagement by altering partner selection criteria such as from an emphasis on what men could provide to a focus on relationship stability. Findings suggest that cash transfers have the potential to reduce young women’s transactional relationships motivated by economic vulnerability.


Introduction
HIV incidence, adolescent girls and young women aged 15-24 years remain at increased risk. Adolescent girls account for 25% of new infections in sub-Saharan Africa and 80% of new infections among 15-19-year-olds (UNAIDS 201815-19-year-olds (UNAIDS , 2019b. In Tanzania, recent estimates suggest that HIV prevalence among adolescent girls aged 15-19 was 1.0% compared to 0.4% in boys of the same age group. This disparity widens with age whereby young women ages 20-24 years have an HIV prevalence 3.4%, compared to 0.9% among similarly aged young men (Ministry of Health, Community Development, Gender, Elderly and Children, and Tanzania and Ministry of Health, Zanzibar 2018).
HIV vulnerability among young women is related to unique economic and social vulnerabilities, which include limited livelihood options (Fox, Senbet, and Simbanegavi 2016), gender inequality (Jewkes et al. 2010), gender-based violence (Li et al. 2014), lack of educational opportunities (Pettifor et al. 2008) and school dropout (Stoner et al. 2017). Transactional sex can be a coping strategy for those from poorer backgrounds which increases their risk of HIV (Maganja et al. 2007;Nnko and Pool 1997;Wamoyi et al. 2011). Transactional sex may be defined as 'non-marital, non-commercial sexual relationships motivated by the implicit assumption that sex will be exchanged for material benefit or status' (Stoebenau et al. 2016). This definition builds on a wealth of previous research seeking to distinguish transactional sex from commercial sex work Dunkle et al. 2004). Unlike sex work, transactional sex refers to sexual relationships or behaviours predicated on actual or perceived receipt of goods, cash or services in exchange for sex . Recent evidence suggests that young women who engage in transactional sex are at an increased risk of HIV Kilburn et al. 2018).
Cash transfers have been hypothesised to address young women's social and economic vulnerabilities, in turn reducing behaviours that put them at risk for HIV. Cash transfers are a structural intervention in which small infusions of cash are provided either unconditionally or conditionally to reduce poverty and increase health enhancing behaviours (Browne 2013). Evidence on their efficacy as an HIV prevention strategy is mixed. Both conditional and unconditional cash transfers have been found to reduce HIV prevalence among young women (Baird et al. 2012) and the incidence of sexually transmitted infections (STIs) among young adults (de Walque et al. 2012). Government cash transfer programmes have shown delays in sexual debut among young women (Handa et al. 2014). For transactional sex specifically, one study demonstrated a reduced incidence of transactional sex (Cluver et al. 2013), whereas two others found no impact (Handa et al. 2014;Pettifor et al. 2016) among young women. Additionally, little is known about the mechanisms through which cash transfers induce sexual behaviour change in this population. Some evidence points to cash transfers exposing young to safer sexual networks (Baird et al. 2012) and behaviours (Handa et al. 2017) by keeping them in school. However, in Sub-Saharan Africa many young women may not attend secondary school warranting the need to study the potential protective pathways of cash transfers for this population.
The transactional sex framework has been used to characterise three main domains of the behaviour (Stoebenau et al. 2016). The first domain, 'sex for basic needs', describes inequitable allocation of resources that positions women as vulnerable and financially dependent on men. This equates to sexual relationships in which poor young women engage in transactional sex to satisfy subsistence needs. The second, 'sex for improved social status', reflects women's agency in using sex to obtain goods and achieve a modern lifestyle. In this domain, transactional sex is used to acquire consumer and luxury desires. The third, 'sex and material expressions of love', reflects relationships in which love, gifts and money are intertwined and men are expected to provide in partnerships. This study, which was informed by the transactional sex framework (Stoebenau et al. 2016), aimed to describe the transactional sexual relationships of young women enrolled in a cash transfer intervention, and examine how cash transfers influenced their relationships.

Methods
The study was conducted with participants from the Sauti Project, a USAID funded programme implemented in fourteen regions in Tanzania to support provision of combined biomedical, behavioural and structural interventions to address sexual and reproductive health needs, including HIV. The cash transfer programme aimed to reduce transactional sex, relationships with older partners, concurrent partnerships and inconsistent condom use.
The intervention assessed under this study included three main components: (1) a cash transfer of TZS 70,000 (USD 31) every three months over 18 months; (2) WORTHþ, a financial education programme which entailed savings and loan groups and entrepreneurship skills training; and (3) 10 h of behaviour change and communication (BCC) group peer-led sessions which covered topics such as HIV and STI prevention, gender based violence, family planning, negotiation skills, self-efficacy/agency skills, condom use skills and health seeking. Eligible participants were female, aged 15-23 years old, out of school (not currently enrolled and have attended less than 10 days of school in the last 3 months), resident in study intervention villages, and willing and able to give voluntary, informed consent/assent to all study procedures. All participants were assessed for vulnerability at enrolment using a 20-item index with five domains: (1) sexual behaviour; (2) sexual and gender violence; (3) impoverishment; (4) isolation; and (5) schooling and literacy (Han 2017). Participants under 18 years required the consent of a guardian unless they met criteria for emancipated/mature minors (i.e. minors who were married, pregnant, had given birth, or were heads of household). We purposively selected programme participants aiming for a balance in age and an overrepresentation of unmarried young women. The final sample had a similar proportion of adolescents (15-19 years old) (47%) and young women (20-24 years old) (53%). Most had a primary education or less (54%) and were unmarried (65%).

Data collection
We conducted interviews with programme participants in the Shinyanga region of Northwest Tanzania. Recruitment was facilitated by community-based organisations implementing Sauti Project activities in Shinyanga municipality (urban), Kahama town (mining), and Bulungwa village (rural). We conducted longitudinal interviews (n ¼ 20) in Shinyanga, cross sectional in-depth interviews (n ¼ 40) and narrative timeline interviews (n ¼ 20) with participants in all three areas. Baseline interviews for longitudinal participants were conducted in June 2017, while follow-up interviews for longitudinal participants, cross-sectional interviews and narrative timeline interviews were conducted in June 2018. Longitudinal and cross-sectional interviews broadly assessed how participants spent their cash transfers, to whom they disclosed, how these people reacted to the programme and any advice they provided, and how this cash may have impacted their relationships.
We used a narrative timeline approach to examine participants' sexual relationships before and during the cash transfer intervention. Narrative timeline interviews are a participatory qualitative research method where participants draw a timeline depicting relationships over time. Timelines are used as a visual tool to elicit details about a relationship (Goldenberg et al. 2016). The methodology has been used sexual health research to examine multiple partnerships (Goldenberg et al. 2016). Interviews covered topics such as the emotional, economic and sexual dynamics of relationships participants had within the past two years. Additionally, participants were asked about their experience in the cash transfer programme and how it influenced their relationships.
All interviews were conducted in Kiswahili by trained female research assistants. Individual interviews were conducted in private locations in the community, including at community-based organisation offices and near the homes of some participants. All interviews were audio recorded. Participants were given TZS 5,000 to cover transport costs to the interview.

Analysis
Interviews were transcribed verbatim in Kiswahili and translated into English. Data were coded and managed using Dedoose online qualitative software. We used thematic analysis to guide data analysis. A codebook using pre-determined codes from the interview guide was developed by the research team. Five researchers each coded one transcript and convened to discuss discrepancies and refine codes and definitions. After reaching consensus and finalising the codebook, researchers coded remaining interviews. Twentypercent of interviews were quality checked for coding consistency by analysts.
Coded data were exported from Dedoose into Microsoft Word for analysis. We created matrices summarising individual participant responses to major themes to allow for comparison across interview type. To improve the confirmability and dependability of the data, analyses were conducted in a collaborative and iterative fashion. Matrices were reviewed and discussed by the study team to agree upon interpretation of the findings. Data reported here specifically focus on participant's sexual relationships and the influence of the intervention on sexual behaviour.

Ethical review
We obtained IRB approval from The University of North Carolina and the National Institute for Medical Research in Tanzania. Informed consent was obtained prior to any interview.

Economic support, emotionality and sex in relationships
Economic support was an important component of nearly all of the relationships described by participants in the study. Participants reported that one of the primary ways they benefitted from their partners was that they 'took care of their needs' or 'helped when they had a money problem'. Participants spoke positively of partners who provided generously, without expectations, and in a timely manner. Conversely, they lamented men with whom they engaged in sexual relationships on the promise of providing material goods but who did not provide or made them beg when they asked for support.
The degree to which sex was directly linked to financial provision varied across relationships. There was a spectrum ranging from relationships in which provision and sex were loosely tied, to those where they were closely tied. Some of factors that determined where a relationship was placed on the spectrum were the emotional connection between partners and young women's socioeconomic status.

Loose transactional relationship between economic support and sex
In relationships characterised by strong emotional connections, the link between sex and economic support was less pronounced. While participants often received gifts and money from their partners, there were additional benefits that young women described, such as love, attraction, open communication, support of their business goals and the prospect of marriage. Many participants viewed the provision of money and gifts from the partners as a sign of love or care. The level and consistency of support signified a partner's intentions, particularly their intention for marriage. Participants had sex with these types of partners after they were given gifts or money, but to them this was not viewed as an exchange but as something they did voluntarily due to their affection for the partner.
You don't have to have sex so he gives you money if you truly love him. So if he says he doesn't have money I let him be. (Bahati,21 years old,unmarried) One participant emphasised how when there is love in a relationship and the prospect of having children together, then money is not necessary when deciding to have sex. This was reiterated by another participant who talked about how after commitment was established, there was no longer an 'exchange' of sex and gifts. The expectation that each gift or sexual act had to be immediately reciprocated was lifted.
Interviewer: Ahhh, so when he gives you money later you give him sex? Not all participants viewed having sex after receiving a gift as exchange. Instead, many described this as a sign of 'thanks' or to make their partner feel 'happy'. They emphasised that sex was a normal aspect of courtship, as was giving gifts. Both were expected in the relationship but are not dependent on each other. In these relationships, sex came before or far after receiving gifts. Some women were able to say no to sex with these partners and they would still provide.
Interviewer: Okay, when it happens that he has given you money how long does it take before you guys have sex gain?
Grace: Even two weeks may pass.
Interviewer: Okay, has it ever happened that he's given you money then you left without having sex with him?
Grace: Plenty of times. (Grace,20 years old,unmarried) This was echoed by another participant who stated: There were a few cases where participants felt there were misaligned expectations about sex and material support between them and their partners. Participants described relationships in which they had a strong emotional connection with their partner; however, male partners explicitly expected that material support provided would be reciprocated with sex. For most, this was a distressing experience and made them feel they were being used by their partner. I mean this is not a business, my body is not for business … For example if you tell him something, that 'help me with something'. He helps you maybe that money he assists you then it becomes like a disturbance. That like, because he has helped you … he tells you in the evening come and get it, when you go he asks you to go to a room. He asks you to give him then he will give you. It's not good I feel bad because give I give you means he does not love you. Not real love. It means it's a business being done. (Farida,21 years old,unmarried) Similarly, Winnie, a 16-year-old unmarried girl, expressed how the request for sex made her question her partner's feelings for her, 'He does not love you; he only wants to use you because he gave you money'. These perspectives indicated that participants interpreted a partner's explicit expectation of sex and/or withholding money as a sign that the partner did not really love them or that his intentions for being in the relationship was not based on love or intimacy but on the desire for sex.
Clear transactional relationship between economic support and sex In contrast, some participants described relationships which they initiated or maintained primarily due to a partner's financial support. These relationships had a strong transactional dimension where sex was conditional on the receipt of gifts or money. Unlike relationships with stronger emotional ties, Participants described few or no benefits outside of the money or gifts they received. The majority of these relationships were short (lasting a few months or less) and overlapped with other relationships. However, there were some accounts of longer-term primary relationships focussed on economic support.
Participants reported often being in environments where men often propositioned them for casual relationships by offering them financial support. Women's decisions to accept these relationships were strongly influenced by what partners gave them.
Interviewer: Mmh and what pulled you to this person to this teacher what pulled you to him? Saida: What pulled me is his money only. Because at first, he would give me lots of money. (Saida,24 years old,unmarried) If gifts or money were not given eventually, then this provided grounds to terminate the relationship. Neema, an unmarried young woman, described how she did not find any value in being with a partner unless he could give her money to meet financial challenges. Another participant reported that she left her husband whom she had a child with because he was no longer providing her with enough money.
Interviewer: Mhh, how was the decision process of your separation was done?
Zhara: Aaa I just decided because if someone is not supporting you there is no need to have him. (Zhara,19 years old,unmarried) Participants of varying economic means engaged in transactional relationships. The poorest initiated transactional sexual relationships in response to economic shocks as well as more generalised poverty in their households. These were unmarried participants with limited resources to support themselves, as well as married women whose husbands were struggling to provide. Participants described personal and peer experiences of seeking out sexual partners due to events such as: illness in the family, food insecurity, and educational needs. Others sought out partners to acquire basic needs such sanitary pads, soap and other small items. There were a few accounts of parents asking participants to seek money from male partners. Amani described how she sought out a partner when her mother was ill: There is a time I had a problem my mother was sick and my brother was at school. There was a certain boy he works in a bank. He seduced me. I never had a way to refuse him. I accepted and he really helped me. I borrowed money from him and told him I had a problem. My mother was ill he helped until mother was well. (19 years old, married) Another woman described a friend of hers who engaged in transactional sex to acquire resources for school.
She was a student and she did not have exercise books … .and she doesn't live with her parent. She has just rented a room … She ran out of exercise books and her school uniform too … So she went and asked her boyfriend for money to buy exercise books and school uniform … So he told her that he will give her the money but she will have to sleep with him first. (Nia, 21 years old, unmarried) A few participants described transactional relationships that were initiated to acquire luxury items or non-essentials. One participant described how there are young women who seek out partners to get phones: It's possible that some of them [women] would think that maybe they don't have a phone or maybe they won't be getting credit so they decide to be with someone so that when they run out of credit they can ask him for credit. (Amani, 19 years old, unmarried) Some viewed transactional relationships which they perceived were not for the sole purpose of meeting basic needs with disdain. They said women who engaged in this behaviour were 'selling' their bodies. I remember I was living with her here as my neighbour. She was just after money. You give her money; she gives you her body. (Fatima,21 years old,married) This same point was reiterated by Johanna who expressed negative feelings about the behaviour. She recalled an argument she had had with a friend who was only dating a man for money: I had a friend of mine we were about in misunderstanding with her because of that behaviour [exchanging sex for money or gifts] … . My friend says, 'I am just coming here at your home simply because I want to meet with my boyfriend as you know'. And I said, 'Why don't you tell him to go to your home?' and she says, 'He doesn't have money. I'm just taking his money then [the relationship is] over'. So, I dislike such deeds. (18 years old, unmarried) Young women who obtained luxury items from transactional partners were viewed as having put themselves at risk for disease for a certain lifestyle.
Someone will want to put on a trouser and sneakers and go to the saloon with a smart phone; if you tell her that is not how life is supposed to be and that there are diseases she does not listen. (Aisha,19 years old,married) A few participants reported refraining from entering into relationships for money because they feared acquiring HIV or becoming pregnant. Zainab, a married young woman discussed a time where she was propositioned by a man: I felt bad because you can get money then make love but at the end of the day you get infected [with HIV], pregnancy and the man goes away. So you are the one left with all the responsibilities. (20 years old, married) As this participant and a few others expressed, young women perceived that they faced worse consequences for engaging in transactional relationships compared to the men who were providers in transactional relationships.

Process through which the cash transfer programme influenced transactional sex
Some young women, primarily unmarried ones, discussed how the cash transfer intervention influenced the types of sexual relationships they initiated after enrolment in the programme. Participants believed that the cash transfers had helped women develop businesses which provided them with a steady source of income. They felt like this reduced their need to seek out transactional relationships where there was a strong connection between material support and sex.
Mary: It has changed. For example, there is a true love and there is a lover where it is not as sincere. When we say a true love, it is someone who truly loves you. The other one loves you for that time only. Several women discussed how a partner's ability to provide was no longer the primary criteria they used to select a partner. Afiya described changes in partner traits she had observed in peers enrolled in the cash transfer programme: Interviewer: For the girls may be at that time when they were not in the cash transfer what things were they looking for from the men … ?
Afiya: Most of the time they were looking for a man who had money.
Interviewer: … And for now, the young girls who are in the cash program when a man come what are the qualities that they are looking for?
Afiya: For now, the first quality he should be free from HIV, second he should respect himself, third his should have future dreams. As for the small need we are able to meet them. We are only looking for those. (Afiya,18 years old,unmarried) Participants also noted the importance of having partners who supported their ambitions and who had some of their own. This was especially important for those who had used cash transfer funds to start or supplement a business. I want that if I have my business then so does he. When I have a problem, he helps me and I would help him when he has a problem. (Reheema,19 years old,unmarried) The shift away from relationships primarily motivated by economics was most pronounced among the poorest participants who were engaging in transactional sex to secure basic subsistence needs. Young women described how they felt like they no longer had to accept just any offer from a man or maintain relationships they were in only for financial support. Salama, a young married woman who previously relied on extramarital relationships to obtain food for her family reported that she no longer has partners outside of her husband: Agatha, an unmarried woman who previously engaged in transactional relationships to take care of her child expressed how this had changed over the course of the cash transfer: Because personally that money [cash transfer] used to help me together with my child and it's the reason why I stopped having men who were not adding any value in my life. (19 years old, unmarried) There was also some indication that there had been a reduction in participants seeking out partners to acquire non-subsistence items. Many young women reported that the cash transfers helped them take care of their personal needs. One young woman described how she used her cash transfer funds to satisfy her needs such as clothing, shoes, hair plaiting and sanitary pads. She later noted how she was able to purchase these items herself which reduced her desire to seek out partners to fulfil these needs: For instance, maybe you can be approached by someone who is rich you see. Now you can be approached by someone that is able and you really know that they will provide you with everything, but you can find that you are busy with your business and you can meet your needs … So if I didn't have any work to do then I would have accepted him so that he would help me and fulfil my small needs. (Neema,21 years old,unmarried) Some unmarried participants reported that they, or their peers, experienced a shift from relationships where they were seeking material support to more stable long-term relationships which could lead to marriage. Participants felt like the cash transfer programme enabled them to reject partners who were looking for casual relationships. They expressed how men had previously cheated them due to their lack of money: Amani: Right now some receive money from the cash transfer programme and have businesses and in the past if you could not have money you could be forced to accept any person since they have money. But now, no one can play with you.
Interviewer: Mmh and play with you. So right now what do they look at from men? Maybe when a man has that thing, they now accept him?
Amani: Maybe if they just accept we marry each other. (Amani,20 years old,married) This point was reiterated by Salome, an adolescent girl who discussed how she was now in a position where she did not have to accept the consequences of pregnancy and abandonment that transactional partners came with: Salome: Because I was very young I knew nothing, but now am grown up I even have my own business Interviewer: So what things will you consider if you need a lover for now, what qualities will you look for? Salome: Maybe if he wants to get married. But to just give me some gifts and say 'let's have sex'in the end he will get you pregnant and leaves you. (17 years old, unmarried) Most participants talked about how the cash transfers made them select only partners who were interested in marriage. Mary discussed how she got married during the course of the intervention: Interviewer: Okay. Did the cash transfer programme help in choosing a husband when your husband came to court you?
Mary: The cash transfer programme also helped.
Interviewer: In choosing him? Mary: Yes.
Mary: Because I saw I love him and he loves me so we could reach our goal.
Interviewer: So this project helped you in choosing your husband?
Mary: It helped because I did not have any rush to be married so after I saw he has the intention of marrying me I agreed to be with him after already joining the cash transfer program. (19 years old, married)

Discussion
This study described young women's sexual relationships and the ways in which a cash transfer intervention affected their decisions about engaging in partnerships. We found that relationships varied in the degree to which economic support was tied to sex, and that this was shaped by emotional connections and participant's socioeconomic status. Young women acknowledged that the cash intervention enabled them to be more selective of partners and shifted their focus to more stable relationships.
The types of transactional relationships young women reported most closely resemble the domains of 'sex and material expressions of love' and 'sex for basic needs' outlined in the transactional sex conceptual framework (Stoebenau et al. 2016). Participants described relationships involving emotionality, love and intimacy, which fell into the category 'sex and material expressions of love'. Participants did not feel that sex and material support were being exchanged because they loved these partners, and sex and gifts were viewed as a part of courtship. In other studies in sub-Saharan Africa, young women report that while material provision is integral to their primary committed partnerships, it is not the motivating factor for entering or sustaining the relationship (Maganja et al. 2007). Young women report having lower material expectations from partners they love as compared to other partners (Hawkins, Price, and Mussa 2009;Kaufman and Stavrou 2004). Relationships which aligned with the 'sex for basic needs' domain, were exhibited by participants with limited economic means engaging in transactional sex due to specific economic shocks (e.g. family illness, need to pay school fees) and more systemic poverty (e.g. food insecurity). These relationships were primarily economically motivated and casual, with women feeling less of an emotional connection to these partners.
While there was limited discussion of young women themselves engaging in transactional relationships within the domain of 'sex for improved social status', some participants expressed stigmatising attitudes towards those engaged in these types of relationships. Young women spoke negatively of peers who received luxury items from partners or those who were with men for money if there was not a justified need, although it was not explicitly clear when it was considered acceptable to be with a man solely for financial support. There was a sense that these types of transactional relationships were immoral. Some participants perceived this behaviour to be sex work, and expressed that women engaging in these types of relationships were 'selling their bodies'. These relationships were also frowned upon by participants because of the associated consequences of unplanned pregnancy and STIs.
Our findings suggest that cash transfer interventions may influence some of the types of transactional relationships young women engage in, and that this influence may vary by demographic background. A common thread across participant's responses was that access to cash provided young women with the agency to reject relationships previously accepted due to limited financial options. These were relationships that they viewed as less desirable and did not seem to hold a potential for a serious commitment. The partner traits that participants said they were now looking for were indicative of a change to safer sexual relationships. Participants also mentioned that partners should be HIV negative.
Additionally, participants mentioned that the intervention helped shift the focus to more stable relationships that could lead to marriage, indicating a potential shift to relationships in the 'sex and material expressions of love domain'. This was expressed as both a willingness to only accept partners who were open to the potential for marriage, or as actively delaying marriage for others. This shift may have both positive and negative consequences. Delaying marriage past adolescence (ages 15-19), is generally protective against HIV for girls in sub-Saharan Africa (Clark, Bruce, and Dude 2006). However, some participants described only accepting partners who wanted to get married which could mean that they marry earlier. For the youngest girls (ages 15-17) this could potentially be risky, but for young women (ages 20-24) initiating a relationship that could later lead to marriage may be more appropriate for their age. Cash transfer interventions for young women should provide age-appropriate separate messages about marriage for girls and young women to ensure they do not encourage girls to marry early.
These findings highlight the potential impact of combined cash transfer interventions for HIV prevention. Notably, cash transfer interventions such as the one evaluated in the study are often time limited. Young women will need long term economic solutions to ensure that they do not engage in transactional relationships due to economic necessity. Furthermore, while the WORTH þ programme included messaging about transactional sex for young women themselves, it is critical to also target the men who solicit transactional sex from this population (Jewkes, Morrell et al. 2012). Participants in this study described how often they were approached by men for transactional sex. Findings from the Stepping Stones trial in South Africa found that targeting men with participatory learning approaches to address sexual health reduced men's engagement in transactional sex (as a provider) with non-primary partners (Jewkes et al. 2008).

Limitations and strengths
Several limitations must be considered in the interpretation of study findings. A primary message within the WORTH þ curriculum, was that young women should develop businesses in order to reduce their economic dependence on men. Consequently, participants may have provided socially desirable responses about how the programme affected their behaviour. Furthermore, because the study involved a complex intervention which combined a cash transfer, financial literacy and sexual and reproductive health education, it is hard to disentangle which component had the most influence on sexual relationships. Our research was also conducted in a geographically limited area and findings may not be transferable to cash transfer participants in other settings.
Despite these limitations the study has several strengths. We combined several qualitative methods in the analysis which increased rigour by allowing us to triangulate findings. Longitudinal interviews also allowed us to understand whether the impact of cash transfers was maintained over time. Data from this source were complemented by narrative timeline interviews conducted at a single point in time, but which examined relationships starting one year before the cash transfers up until a year or more into receiving cash transfers. Additionally, the study is one of few qualitative studies to explore the impact of cash transfers on transactional sex.

Conclusion
This study found that young women engage in different types of transactional relationships, and that cash transfers seem mainly to influence relationships motivated by basic needs and to a limited extent those motivated by improved social status. The process by which cash transfers influenced transactional relationships in this study was by changing the criteria young women used to select partners and moving them towards more stable relationships with the possibility of marriage. Interventions that provide sustained economic support and address gender norms concerning transactional sex seem likely to have the biggest impact on reducing transactional sex, and ultimately preventing HIV within this population.