The prevalence of men having what we have referred to as a transactional sexual relationship, i.e. where they perceive that the relationship or sex act was predicated on a provider expectation (whether fulfilled or not), was very high at 66%, and much higher than the prevalence of having had sex with someone they identified as ‘a prostitute’ (18%). Through asking about transactional relationships or sex with different types of partners (although not with wives), we have shown that almost all men who ever report these, have such relationships with their main partners, and further that there is almost no difference in the nature of the provider expectation between partner types. We acknowledge however that our question formulation required men to speculate about their sexual partner’s motives, but in so doing we measured their perceptions of expectations. This very strongly suggests that a large group of men within the South African population perceive themselves to be expected to provide for their partner in various ways within a relationship and that they would not have an opportunity for that relationship if they did not fulfil that role [11
]. Most explicit in the case of transactional sex with once off partners, but also otherwise implicit, given that what men reported were sexual relationships with other women, is the idea that by so doing men are entitled to sex with this partner i.e. on the woman’s part receiving entails obligation.
The study findings suggest that in as far as we have attempted to measure and study transactional relationships and sex, what we have succeeded in capturing may best be understood as a conservative notion of male gender roles in relationships, where men are expected to provide and women in return are sexually available. This has been well described in qualitative work from Africa (e.g [22
]). Clearly this notion prevails in all of the social groups studied, albeit with some difference in frequency, but is most dominantly seen among low income, lower educated, Black African and Coloured men. This sits uncomfortably with the notion of a category of ‘transactional sex’ or even ‘transactional relationships’ although we have used such terminology in this paper. These notions suggest a certain instrumentality that individual men may apply to some relationships or sexual acts but not others, whereas it seems we have rather tapped into a provider notion which would probably apply across all these relationships. The notion of transactional sex has predominantly been developed and utilised through research with women, some of which captures a very particular instrumentality aspired to in sexual encounters or relationships [16
]. An extreme example is the somewhat mythical ‘3C’s girl, exploiting men for cash, cars and cellphones [8
]. We would like to suggest that women’s discourse of transactional relationships and sex may appear empowered and liberal, and thus contrast with men’s perceptions of provider expectations, which are a reference to a very conservative gender norm, but there may often be a clash in reality. If conservative gender norms generate expectations that men will provide for women, then it is possible for a women to feel empowered by ‘exploiting men’ (to use the language of one of Leclerc-Madldla’s informants [8
]), whilst the ‘exploited men’ view themselves in a conservative gender role. Thus this agentic expression of a ‘modern’ feminine position comes at a particular cost in terms of complicity with the patriarchal gender order (c.f Jewkes, Morrell and Haram [12
The study findings suggest that many men perceive that they are expected to provide a considerable range of goods to their women partners. Whilst this list included some items that could be rendered in kind (such as repairs), by and large the men indicated that the realm of sexual relationships is cash economy. It is very easy to see how these social expectations put pressure on men, especially in context of unemployment, and may be strongly resented by men who have little or no money. Further this gives rise to circumstances in which men may mistrust the motivations and commitment of women sexual partners (c.f. Mganja et al [22
Whilst it was earlier suggested that transactional sex blends with sex with a woman in prostitution, it seemed that many of the men did not see it that way. The manner in which men responded to the questions indicated that they distinguished between sexual relationships and acts that revolved around providing and reciprocal sex, from sex with a woman in prostitution. This was most clearly seen in the reports of transactional sex with a once off partner where the man provided cash. To an outsider, this may appear remarkably similar to the type of transaction found in prostitution, but most men made a distinction. In fact two thirds of men who described having been in such a situation said they had not had sex with a prostitute. This may be because having sex with a woman in prostitution is largely stigmatised whereas transactional sex is regarded as ‘normal’ and acceptable (by both men and women) [10
There were racial differences, with a very much lower proportion of Black African men reporting having had sex with a woman in prostitution than Coloured and Indian men, although an important part of this difference was due to many Black African men being of lower income. Although there were significant differences in both the prevalence of transactional sex and sex with a woman in prostitution between income categories, the differences were largest when it came to sex with a woman in prostitution. This has been a relatively infrequent experience for men who were not earning, compared to those earning R500 of more per month. The prevalence of transactional sex was very high for all income categories and the proportional difference between categories was not large, suggesting that expectations of men fulfilling provider roles are relatively inelastic when related to income. The findings that very low income men have paid for sex is in keeping with observations of the sex industry internationally [28
] and findings that low income men may spend a greater proportion of their income on casual partnerships than higher income men [29
The prevalence of sex with a woman in prostitution in this study was nevertheless higher than that reported from Sweden (18% v. 13%) [30
] cited in [32
]. But in both countries it cuts across all ages and income classes. In South Africa, like elsewhere, many men who have had sex with a woman in prostitution are married or cohabiting, although the cross-sectional nature of the research prevents us from knowing their relationships status at the time when they did it.
We have shown that men who are older (from their mid-twenties) are more likely to have had transactional sex, but there was no association between age and having sex with a woman in prostitution. This suggests that men who have sex with a woman in prostitution do so for the first time when young or they will not do so at all. This finding is similar to that of research with men in the United Kingdom, where research has shown that if a man had not paid for sex by the age of 25 he was less likely to do so in the future [33
]. The majority of men in the study who reported having had a transactional relationship or sex with a woman in prostitution were single or married and the prevalence of these practices in these groups of men did not differ. Men who were cohabiting were significantly more likely than married men to have done both of these practices. In South Africa’s conservative gender context, it is possible to interpret cohabiting as reflecting an unwillingness to commit to a partner, with such men being more likely to seek sex in a range of contexts where there is no expectation of commitment, especially where it is overtly or implicitly commodified. However we recognise as well that cohabiting may be a practice of marginalised very low income men who feel excluded from mainstream gender goals of establishing stable marriage unit. We recognise that this social meaning of cohabiting is different from that found in many higher income countries.
The prevalence of having had transactional sex (or sex in a provider role) disclosed by men is much higher than that reported by women, for example by Dunkle et al [6
] (which was 21% with non-primary partners). There were some differences in the way the exposure was ascertained, but it is also possible that men are more likely to assert that involvement was due to an expectation of material gain. This may be in part a statement of provider prowess, obviously partly reflects expectations of women that sex partners will give them gifts, but it is also a commonly voiced complaint by men about women that they only are interested in men for what they can get from them. In this respect a greater likelihood of disclosing transactional sex may be associated with greater misogyny.
A strength of the study was the use of APDAs for data collection as these provided a confidential environment in which disclosure of anti-social and illegal behaviour could be enabled. Through removing the face to face component to interviews, APDAs greatly reduced the performative aspect of interview responses and so gave us more confidence that there would not have been a problem of over-reporting. Under or over-reporting, either deliberately, or through misconstruing women’s intentionality, is a recognised problem in research on transactional sex and stigmatised acts such as having sex with a woman in prostitution. It is impossible to estimate the magnitude of this in the study. Recall bias is a potential problem with any study of this nature. Self-completion of the questionnaire and the option of skipping questions resulted in some missing data on some items. We have not replaced missing values. Inevitably we had to keep this analysis within the constraints of the questions included in the questionnaire. It is unfortunate that these precluded us from obtaining other information about the men’s relationships or sexual acts, apart from this reported here. We did not retain information on the number of eligible men per household and so were not able to weight the analysis for this, but we have no reason to believe this would have made much difference to the estimates of association [35
The findings have methodological importance as men from all racial groups clearly distinguish between adoption of a provider role/transactional sex and sex with a woman in prostitution and so it is methodologically important to maintain these distinctions in research. The most common provider role/ transactional sexual relationships were reported with main partners. This indicates that questions that focus on exchange in casual or once off partner relationships will miss the largest category of commodified sex or relationships.
Transactional sex or the male provider role, in particular, supports subtle understandings of gender inequality even while it seems more acceptable than the inequalities that are manifest in the commoditized arena of prostitution. By this reading, transactional sex assumes a resourced male in a supplicant sexual relationship with a woman who is assumed to be passive, but potentially can give or withhold sex. Transactional sex thus supports patriarchal sex roles (including a passive woman). On the other hand, transactional sex can also be considered as the arena in which women are more proactive, leaving the arena of marriage and familial negotiation, to pursue a more active role in exploring their material and sexual needs [23
]. Similarly, for men with resources, it is the place where these resources are converted into heterosexualised masculinity in a way that avoids the public slur of isoka lamanyala
(a man who takes womanising too far [11
]: 165‐6]). It is precisely the possibility of these multiple readings of transactional sex that constitutes a resistance to any equation of the practice as a form of prostitution.