Introduction: contraception and normative tensions
1The social value attributed to high fertility is one of the main obstacles to contraceptive use in Southern countries (Lesthaeghe, 1989; Cleland et al., 2006). Fear of side-effects of modern contraceptive methods is often identified as a major factor holding back their use (Bankole et al., 2006). In addition, some writers have suggested that this fear relates more generally to the rejection of Western reproductive norms which value the fertility reduction that these products would help to establish (Otoide et al., 2001; Guillaume and Desgrées du Loû, 2002). Contraceptive provision that fails to match women’s expectations (because of providers’ reluctance to supply contraception to certain women, inadequate distribution points, or price) is also often cited as an explanation for limited use of hormonal contraception (Brown et al., 1995; Banza, 2001; Katz and Nare, 2002; Sinai et al., 2006).
2As Feyisetan and Casterline (2000) and Watkins (2000) have pointed out, many of these studies limit their analyses to the rational actor paradigm; that is, they attempt to identify factors linked to non-use of contraception while presupposing that women systematically wish to use it. In addition, these analyses focus on the question of contraceptive use, without considering unplanned pregnancies among women who are using a contraceptive method. Even when there is easy access to contraception, a significant proportion of pregnancies – between 30% and 50% – are unplanned, and most of these are terminated (Glasier et al., 2006).
3These observations demonstrate the limits of approaches based on “unmet need for contraception”, which has been a central concept of family planning policies implemented in recent years (Cleland et al., 2006).
4Focusing on four African countries (Burkina Faso, Ghana, Morocco, Senegal), this article aims to characterize the range of situations that typically lead to unplanned pregnancy, whether no contraception was used, or the woman faced problems in contraceptive use.
5Our central hypothesis is that contraceptive use is at the intersection of social rationales which, in a given social context, relate to reproductive norms (Lesthaeghe, 1989; Bledsoe, 2002; Johnson-Hanks, 2006), sexual norms (Cole and Thomas, 2009), gender relations (Andro and Hertrich, 2002; Smith, 2010) and relations with contraceptive providers (doctors, family planning centres). There is often tension between these norms, and it is when such tensions arise that unprotected or poorly protected sexual intercourse occurs. It is these norms and phases of tension, specific to each society, that are analysed here.
6Comparing the practices of women in countries with differing levels of access to contraception, notably for unmarried women, we can move beyond conventional supply-based analyses to consider the specific issues likely to be faced by women whose sexual lives are not socially legitimate.
7In terms of birth control policies in Africa, Morocco and Ghana are two pioneering countries which adopted population policies from the late 1960s onwards; Senegal and Burkina Faso, in contrast, did not do so until the 1990s.
8There is one immediately obvious difference between these four countries: Moroccan family planning programmes are effective and broadly spread across all social groups (with a small difference in prevalence by level of education or place of residence), but they are restrictive. They exclusively target women in a union, thus demonstrating the very strong stigmatization of sex outside marriage and the total refusal to acknowledge that unmarried people engage in it – especially women, who cannot openly express their sexuality. In the three sub-Saharan African countries in our study, birth control policies also give priority to the marriage-sex-reproduction triptych. However, although non-marital sexual activity is still perceived as socially unacceptable, it is on the increase in those countries, driven in particular by the rise in women’s age at marriage (Bledsoe and Cohen, 1993). In other words, although it is condemned socially and morally, there is a relative degree of tolerance towards young people’s sexuality in Accra and, to a lesser extent, in Ouagadougou and Dakar – though less so for women than for men. There is no such tolerance for women in Morocco. In the three West African countries – in contrast to Morocco – there is no denial of reality and young people are taken into account in family planning programmes, even if access to contraception often remains problematic (Blanc and Way, 1998).
9The institutional and normative context is reflected in surveys of contraceptive practices. In Morocco, the only available data on contraception concern women in a union, whereas for the other countries they are available for all sexually active women who have had intercourse in the month preceding the survey. In the city of Rabat,  where contraception is easily accessible for married women at health centres and pharmacies, more than half of women living in a union (53%) use a modern contraceptive method (mostly the pill) and 14% use a natural method. In the other three capitals, where contraception is also available from health centres (in particular, mother-and-child welfare centres) and from pharmacies, but is less easily accessible than in Rabat, contraceptive prevalence is lower among sexually active women. In Ouagadougou, 48% of sexually active women use a modern contraceptive method (22% condoms; 10% the pill; 6% injectables), while 10% employ a traditional method; in Accra, 35% of women use a modern method (11% condoms; 7% the pill; 7% injectables; 4% female sterilization) and 12% a traditional method; in Dakar, only 19% of women use a modern method (10% the pill; 6% injectables) and 2% a traditional method. Condoms are very widely used, particularly in Burkina and in Ghana thanks to HIV programmes.
10Note that access to legal abortion in these countries is highly restricted. In Senegal, abortion is permitted only to save the mother’s life; likewise in Morocco, where it is also allowed if the mother’s health is threatened; in Burkina and in Ghana, the law also covers situations involving rape. In all cases, access is subject to specific conditions and authorizations that make the right to abortion more theoretical than real.
I – Methodology 
11In order to identify obstacles to contraceptive use and to analyse situations in which it is problematic, we favoured a qualitative approach designed to reveal the social tensions women are likely to face in observing the various norms pertaining to reproduction, sexuality and gender.
12In 2006-2007, we conducted semi-structured, in-depth interviews with 50 women in the capital of each of the four countries under study (Rabat, Ouagadougou, Dakar and Accra). The team defined the sample quotas by age (18-35), relationship status and social situation (individual respondents had to be, or have been, sexually active) (Table 1). Respondents were recruited using the “snowball” sampling method (Combessie, 2007), on the basis of information gathered from family planning centres or pharmacies. We also made use of our interviewers’ personal social networks.
Social and demographic characteristics of women interviewed in the ECAF survey (numbers)
Social and demographic characteristics of women interviewed in the ECAF survey (numbers)
13The guidelines for the semi-structured, biographical interviews were drawn up in collaboration with teams in each country to identify the best way of addressing questions about sexuality. The same guidelines were used in each country. The interviewers (both male and female) invited the respondents to give their views on whether or not, in future, people may be free to choose their own spouse without any family intervention. Only after that did interviewers move on to discuss the respondents’ personal experience. By going through their life histories, follow-up questions could be used to clarify the circumstances of their sexual debut, their views on contraception, their contraceptive use, and any history of abortion.
14All the interviews were handled as a single corpus. All the researchers involved in the analysis presented here read the transcripts several times and discussed them collectively. The interviews were examined in two different ways, starting with a thematic analysis to identify themes that recurred from one narrative to another and any variations on those themes, followed by an analysis of these themes according to the respondents’ characteristics. The interview was then examined as a whole: the order in which the themes arose revealed how each individual history might explain the unfolding of the process analysed here, namely attitudes to contraception and contraceptive practices.
15Thematic analysis was performed using NVivo7 software. We first identified the themes which emerged from repeated reading of the interviews. Then each interview was coded according to the respondent’s socio-demographic attributes and any themes defined as relevant for our analysis. Where a recorded comment touched on several themes, it was classified under several thematic groups. The software gives the position of particular comments within the interview so that the context in which they emerged during the discourse could be taken into account in the analysis. The material we had collected could thus be cross-matched, linked, grouped or differentiated on the basis of different attributes and themes.
16In this kind of qualitative analysis, the construction of a set of themes is never fixed or final; NVivo is a flexible tool that allows new signifiers to emerge throughout the research, so our thematic categories could be revised and adapted as the analysis progressed.
17Two summary documents were produced for each interview. The “portrait” offered a theoretical-oriented summary of the interview that enabled us to clarify the respondents position with respect to prevailing family and gender models. The “life course grid”, in contrast, used a summary table to outline the major stages in individual trajectories, covering family life, changes in place of residence, education, working life, sexual life, births and contraception (Battagliola et al., 1992). In collaboration with the teams from each country, these documents were compiled and written up as each interview was being read.
18Analysis of the interviews allowed us to identify the different social norms that organize social and family models in the four capital cities we studied – norms whose observance may create a problematic relationship with contraception. They relate first and foremost to (i) the reproductive norm whereby women must have a child immediately after marriage and then space later births; adherence to this norm means that the woman’s fecundity must be maintained until the desired number of children has been reached. Contraceptive attitudes and practices also depend on sexual norms that are dominated by a gendered double standard, which (ii) prohibits sex outside marriage, especially premarital sex for young women, and also (iii) endorses the primacy of male sexual pleasure.
19These different norms will first be presented through several examples to show how they contribute to the construction of a problematic relationship with contraception. Secondly, we will demonstrate how they interact with one another and create tensions, producing situations that, in turn, lead to contraceptive failure. Lastly, an example of a detailed life history will be presented to show how contraceptive “moments of vulnerability” in a woman’s life are generated by these tensions between norms – tensions which develop differently according to the interpersonal and social contexts in which she finds herself over her lifetime.
II – Social norms and contraceptive attitudes and practices
Motherhood and social status
20The reproductive norm, which represents socially valued conditions for having a child, encourages couples to have a baby soon after marriage, enabling the woman to provide rapid proof of her fecundity. A woman who departs from this model may find herself in a awkward situation if her partner or her family circle expect her to become a mother quickly, even though she might prefer to defer motherhood in order to complete her education or embark on a career. The attitude of contraceptive providers may also make it harder to start using a contraceptive method, as Cocotte’s situation illustrates. Cocotte is a young Senegalese woman, aged 24, with higher education, who wanted to postpone her first birth until after she had found her first job; however, she did not succeed in doing so and became pregnant while still an intern in a ministerial office.
… I was a virgin when I got married, so… after, I went, just the week before I got married, I went to the doctor because I didn’t want to get pregnant and I was at the… time, when you are really fertile, when you can get pregnant, so I went to see the doctor, you know, I asked his advice because I wanted to take pills, I wanted… not to have a baby very quickly after getting married, so, he… strongly advised me not to take anything and to have a child, then… do what I wanted to do and wait however many years I wanted before having a second one… And that’s what I did, because, after the honeymoon, I got pregnant, but in the end, as I didn’t want to at all… I didn’t want it, I cried and cried, perhaps it was the stress or something like that, I had a miscarriage.
22The social advantages of motherhood explain the attention paid by all the women we interviewed to preserving their fecundity before forming a union, and afterwards until the desired number of children had been reached. This defined a very particular attitude to hormonal contraception (the pill, injections or implants). Many women were reluctant to use these methods, which they believed put their future fecundity at risk. Some refused them for this reason, preferring other methods, while some who did end up using them were tempted to minimize the risks they ran. They felt they needed to take regular “breaks” or even, for the pill, to take it only sporadically in order to limit the dose of hormones they ingested. For example, after six years of taking various pills, Soumaya, a 31-year-old Moroccan housewife with one child, decided to take a break from her contraception in order to “have a bit of a rest”, thinking that she was still protected. She became pregnant in the first month of this “break” and decided to have an abortion. It was easy for our correspondents to rationalize this method of managing hormonal contraception to “minimize risk” because many providers share the same fears and do not give women the information they need to manage these breaks effectively. Lydia, an uneducated 35-year-old Ghanaian housewife, was married to an alcoholic with secondary education who routinely beat her and brought very little money into the home. She thought that the four children she had brought into the world were more than enough. She became pregnant nonetheless, since she was poorly informed as to how to use the method she had chosen.
So, after my fourth child, I decided not to have any more children, so that I could look after the ones I already have properly. But I’ve noticed that if you live with a man and refuse to give in to him, his reaction is going to be: “What’s up?” At that time I used the five-year injectable. They told me that after five years I could take a break from it for a while. Then, when I took a break, I got pregnant again in less than three months! You see? – Less than three months! It’s become a heavy burden on me.
24This adherence to the reproductive norm and the same fear of impairing one’s future fecundity is observed in all four capital cities. Similarly, in all four countries, women who had fewer social resources were more inclined to want more children. The differences between countries lay in the reproductive norm – that is, the specific conditions that give childbearing its social value (age, number of children, social and family situation, etc.).
The social prohibition of non-marital sexuality
25Another norm that leads to a problematic relationship with contraception is the social prohibition of non-marital sexuality for young women – premarital sexuality in particular. If women who are sexually active outside marriage become pregnant, they run the risk of revealing a form of sexual behaviour that is socially proscribed, and this should encourage them to use methods for avoiding pregnancy. But equally, using contraception confirms the existence of a sexual life that they are not supposed to have. This absence of a socially recognized right to sexuality makes access to and use of contraception difficult. A number of young women reported that they had begun their sexual life without any contraception, since the issue had never been discussed in the family and only extremely briefly at school. For example, Codou, an unmarried 18-year-old Senegalese secondary school student, became pregnant on the very first occasion she had intercourse with a chosen partner, when she was clearly ignorant of the risk:
Interviewer: So it was only one night… just once, or more?
Codou: Yes, it was just the once, because it astonished me, I asked myself how someone could just have intercourse once and be pregnant, but afterwards, when I looked at everything I’d been taught, I did know that it was… It was possible, I knew that it was possible and also when I felt sick and that kind of thing, I went to the hospital for advice and they told me I was pregnant, but…
27This social prohibition of non-marital sexuality makes access to contraception especially difficult. Kardiata, a 24-year-old Senegalese woman, married with one child, with higher education and not in employment, told us that she had dreaded going to the pharmacy to ask for the pill, even though she had already had an unplanned child outside marriage:
Kardiata: … I couldn’t imagine having to go to the pharmacy to buy the pill, because that, that was just too much for me (laughter)!
Interviewer: In what sense, ‘too much’?
Kardiata: Ah, you – you’re not married and you’re coming here asking for the pill!
I: So it was only because of that that you didn’t take it?
Kardiata: Yes, it was only because of that, it was just those kinds of constraints. You know, even your parents probably see it and say “yes, so you’re going on doing that even though you’ve got a child”, you understand, you can’t be certain.
29It was the same reluctance to reveal premarital sexual activity that led Monique, a 22-year-old single Senegalese woman in her second year of studying for a vocational certificate, to stop taking the pill because the medical centre required her partner to be present when they prescribed it; he did not want to go because some of the staff there knew him and that made him feel awkward.
30The strength of this prohibition on premarital sexuality varies across countries: it is less strict in Ghana and in Burkina Faso, very explicit in Senegal and particularly marked in Morocco. It is reflected in the requirement for a woman to be a virgin on marriage. For this reason, some young people opt for non-penetrative intercourse, but even that is not totally risk-free, since some couples end up engaging in penetration that was not initially planned.
31For example, Wissal, an 18-year-old Moroccan school student, single when the events in question took place, told us how she came to have sexual intercourse for the first time:
Wissal: It was before the month of Ramadan, we started at a friend’s birthday party, we had our first kiss, he kissed me on my girlfriend’s birthday and that was all. After that, I felt it was good, I loved all that, we started seeing one another, after a year I started going to his house (his parents’ house), I went to his parents’ house in the normal way, I sometimes stayed to eat with them, bit by bit I started sleeping with him (externally, between her legs, sexual activity avoiding penetration) until the day when I felt something, I called out, like that, and I found some blood, and I didn’t say anything, I cried and I kept quiet.
The primacy of male sexual pleasure
33Finally, another factor likely to interfere with effective contraceptive use arises from the primacy of male sexual pleasure.
34Sofia, a 26-year-old Moroccan woman with higher education, working in management at the Ministry of Justice and married to a man who was also university educated, with whom she had two children, described this very graphically. She used the rhythm method, but became pregnant because her husband had not observed the periods of abstinence.
I have a regular cycle… When my period ends, I always count, I always do that… But even so, I have slept with my husband, I’ve sometimes been stupid and slept with him. I just wanted to do my duty towards him, when he said to me “Let’s go to bed together”, I couldn’t say no, even if we were fighting, I slept with him…
36Her account illustrates the obligation that society places on women to respond to their husband’s sexual expectations – something we found in all four countries studied.
37The primacy of male pleasure often also leads to unprotected intercourse because the partner does not want to use a condom. Generally, the man gives a lessening of sexual pleasure as the reason for his refusal. Even if the woman reminds her partner about the risk of pregnancy, it can be pushed into the background, as the situation of Madeleine from Burkina Faso illustrates. This single student nurse related an episode that occurred with her lover, a married man:
Well, we went out together for some time, we used a condom every time, and one day he invited me to his office… And I knew that if I went there, he would get me to have intercourse. I didn’t hesitate, I stopped at a pharmacy, I bought some condoms, I put them in my bag because every time, it was me who was in the habit of buying the condoms, he didn’t buy them. And when I’d bought the condoms, I went to his place, we chatted a bit, well… Things overtook us, me too. Just as we were about to actually do it, I said to him “No, please wait, I’ve got condoms in my bag, you should put one on”. He said “No, I can’t wait any longer, I feel overwhelmed, I can’t wait…”. And I don’t know when, how it happened, but he did penetrate me like that. And when we’d finished, we got up, I immediately went mad.
39This primacy of male sexual pleasure has even more normative force when the woman is in a situation of socioeconomic and/or emotional dependence on her partner.
III – Where contraceptive failure meets normative injunctions
40For women, unprotected or poorly protected sexual intercourse most often occurs in situations where several of the normative pressures that have just been outlined arise together, whether their effects are cumulative or opposing.
41The example of Mawa illustrates this situation: a single 21-year-old student from Burkina Faso, she found herself unintentionally pregnant. She had decided to take the pill because she wanted to be protected against pregnancy. But she stopped taking it after three weeks because of side-effects: she “felt odd”, to the point where she could no longer manage to keep up in class even though it was an exam year for her. She was eating a lot, felt nauseous and had migraines which might have aroused her mother’s suspicions – something she absolutely wanted to avoid. Apart from its side-effects, the pill did not really correspond to her needs. In fact, in her opinion, she did not have intercourse frequently enough to justify its use.
I don’t have regular intercourse with my, my partner. Perhaps twice a month. Perhaps even only once a month. Especially now that he has problems, it’s not something he thinks about, and all that. And, well, it’s really irregular, our sexual relationship. … I don’t take the pill because I said to myself “What’s the point in taking the pill, and then a month goes by and you haven’t had sex?” All you’re doing is punishing yourself physically.
43Here, normative pressures had a cumulative effect. Even though Mawa dreaded pregnancy (prohibition on premarital sex), the fear of her sexual practices being discovered by her mother and the fear of taking a hormone that she judged to be dangerous for her fecundity (getting married and having a child in order to acquire social status) both prevented her from using the pill effectively.
44This fear that harmful effects of hormonal contraception could compromise their future fecundity led a certain number of the women we interviewed – especially those who had not yet had children – to prefer condoms, especially because these also protected them against STIs. But the example of Nicole, a 20-year-old single secondary school student, shows that in Burkina Faso, as elsewhere, it is not always easy to convince one’s partner to use a condom. Nicole was very aware of the risks linked to sexual activity and was keen to protect herself against them; but this did not suit her partner, who was concerned above all to satisfy his libido with the fewest possible constraints. As a couple, their relationship was punctuated by disagreements and tensions, notably about the use of condoms and her partner’s refusal to take an HIV test. She had a fairly good knowledge of contraceptive methods, notably the pill, whose main drawback, in her opinion, was the risk of forgetting to take it. She did not exclude the possibility of using it later, after having her first child or when she was married. In the meantime, there were occasions when she had unprotected intercourse, and she explained, referring to a recent episode:
He says “With a condom, I don’t feel you, I can’t come…”. But if I take too many pills and then if after that, when I try to have a baby, I don’t get pregnant, my husband is going to say that he has married a woman who can’t have children.
46Here, normative pressures interact in an opposing way. On the one hand, protecting her future fecundity prevented Nicole – like Mawa – from using a form of hormonal contraception that she could control, but on the other hand, even though she wanted to protect herself against pregnancy, she could not use the method that suited her most – condoms – because of the priority given to male pleasure.
47Naima’s story illustrates another configuration of the normative tensions to which women are subject – and which can lead to contraceptive failures. This 33-year-old Moroccan, a practising Muslim and a manager in public administration, had lived in Rabat since starting university. When we met her, she had been married for six years and had two children. Her husband was a geological engineer, and the couple was in the process of buying their own home. She had met him at the age of 26 and they had gone out together for a year before marrying for love. During that time, she had “external” sexual relations with the man whom she had very quickly come to view as her future husband. After marriage, she did not use contraception, and became pregnant even though she did not really wish to have a child at that particular time. After that birth, she decided to take the pill. She stopped taking it from time to time, in order to “have a rest”, because the pill “got on her nerves” and gave her brown patches on her face. She took the view that “it’s against our religion for a poor woman to be stuffing herself with the pill for four years”. Yet her husband refused to use condoms, considering this method appropriate only for intercourse with prostitutes; he also considered that contraception was women’s business. So she used the rhythm method and quickly became pregnant. She had an abortion; deciding to do so alone, after her husband said “it’s your problem, they’re your sins”. It was an event that upset her, since she considered that she had “killed a soul”.
48Two years later, while she was still using the rhythm method, she became pregnant again. Two children fitted the couple’s ideal model of the family, and even though Naima said that she did not really want this pregnancy, she decided to carry it to term. The following year, again “resting from the pill”, she became pregnant again and had a second abortion. At the age of 33, Naima was still having difficulties with contraception, since she could not tolerate the pill but did not want to use an IUD: to her, it was a “foreign body”, even though her gynaecologist recommended it and advised her against taking the pill at age 40.
49Naima was now very disappointed in her marriage and no longer felt any sexual desire. Her husband did the shopping and was, in her view, fairly understanding: he “took care of the children himself” while she went for English or music lessons, and he let her go abroad; however, she had to take total responsibility for the children’s upbringing. She considered that, despite progress in some areas, there was still a high level of inequality between men and women in the society where she was living; she said that she suffered from this situation more acutely because she was educated. She felt that she had had children because of social pressure and in order to please her husband, even though she had dreamt of going abroad to study.
50Thus her life history revealed how her adherence to the dominant social model – in this case, in Moroccan society today, the two-child family – combined with respect for reproductive norms and the primacy of male pleasure, led this woman to have several unplanned pregnancies, two of which she terminated.
IV – Contraceptive failure: the social and interpersonal context
51Women’s ability to distance themselves from norms that may hinder effective contraceptive use depend not only on their social resources but also on the interpersonal contexts of their relationships. These contexts may also change over the course of a given relationship or over phases of the life cycle; this is illustrated by the story of Janta, aged 25, a young married woman from Burkina Faso, mother of two children, who sold soap for a living. An event-history interpretation of her interview reveals that a woman’s attitudes to contraception and her ability to manage its use may vary over the course of her sexual and emotional life. Difficulties in using contraception can arise in very varied situations and for different reasons.
52Janta was born in Ouahigouya (a medium-sized town north of Ouagadougou) into a low-income, polygamous Muslim family (her father was a driver, her mother a rural community worker). The family included 22 children, born of three wives. The fourth of her own mother’s 11 children, Janta was particularly loved by the wives of her “little daddy” (her own father’s younger brother, who worked in a bakery) and she was adopted by these women at the age of five – a common practice in Burkina Faso. At her own request, Janta left school while still at the primary stage; she returned to her own parents’ home at age 15, when her uncle’s wife left him. She then met her first partner, who was 25 and a labourer in a building firm. She embarked on a chaotic relationship with him, and many times threatened to break it off if he insisted on having sexual intercourse. Several months later, during a New Year party, he ended up forcing her to have sex. The words she used to recount this experience show how the legitimacy of male desire is internalized and illustrate the complexity of the continuum between unwanted intercourse and rape.
Interviewer: When did your first sexual intercourse take place?
Janta: (Long silence)
I: Come on!
Janta: (Long silence) I wasn’t forced into it, but it wasn’t easy (laughs). When he started, I didn’t really agree, but he ended up forcing me so as to make me agree.
I: How did it start?
Janta: I didn’t know that we were going to have sex. That day, we went out, on the 31st (of December), we just went out to sit around, dance, have a drink, and then when we got into the courtyard, we sat down for a chat. We were together in the house and we were talking and his big brothers got up to go out and left us alone in the house. That’s how it was. (Long silence) He was talking and I said no, I had to go back (home). He said no, that because I was at his house, it meant I couldn’t go back. (Long silence) He was talking and I refused and that was that, he used force then and that was that, I had to agree.
54In a context like this, Janta was not ready to consider the risks of pregnancy, since she had not been thinking of having intercourse at that particular time. In addition, she reported not knowing at the time of any method for avoiding pregnancy.
55She very soon discovered that she was pregnant. Her family forced her to go and live with her partner, with a view to marriage. This union would make amends for her violation of the social prohibition on sex and childbearing outside marriage. After the birth of the baby, life in this very big family (a polygamous family with about 100 members) went badly for Janta, notably with her mother-in-law, who tried to control her son’s household. Despite all this, after the delivery, she decided to go and sleep in her mother-in-law’s house in order to avoid having sex with her partner. She practised post-partum abstinence until her son was 17 months old, and then continued to breast-feed him until he was three, which she thought would protect her.
56During this extended period of breast-feeding, she limited intercourse to once a month, two weeks after her period. She did not want to take “tablets” (the pill) offered to her by the hospital, because she was frightened of side-effects. She was also afraid that she would not be capable of taking it regularly. This rhythm method also enabled her to avoid “tainting the milk” by becoming pregnant when she was still feeding her child. Her decision was accepted by her partner who, she said, behaved considerately, as he did not want to make his future wife unhappy.
57After the birth of her son, Janta started a small business selling fried yams, which, she said, allowed her to earn more money, while her partner had just lost his job. Arguments with her mother-in-law, who was opposed to their marriage, made her life unbearable – and things got worse when the mother-in-law started to take her earnings from her. When her child was weaned at the age of three, she decided to take her savings and run away, and she left to live with an aunt in Ouagadougou. There she met the man who later became her husband, a 32-year-old hay seller. She went out with him for a year without having sexual relations. Her aunt was fearful of the reactions of Janta’s parents and of the child’s father if she did not prevent her niece from marrying this man. She asked her to go back to her parents’ village to ask their permission to get married. Janta, knowing that her parents would probably oblige her to go back to the child’s father, then decided, with the agreement of her suitor, to force her family’s hand by becoming pregnant.
58Once she was pregnant, she went back to the village to return her first child to his father (as a child belongs to the father, she could not keep him in her new home). She then found herself under various family pressures. Her mother, who did not know about the pregnancy, wanted her to go back to her child’s father. Her brother supported her in her desire to remain with her new partner. Her sisters and her aunt, whom she took into confidence about her pregnancy, pressed her to have an abortion in order to conceal her extra-conjugal affair. After five months of discussion, and as the pregnancy was becoming increasingly obvious, her mother finally allowed her to make her own choice and Janta went back to Ouagadougou without her first child.
59She then moved in with her boyfriend and, after her baby was born, the couple married. They resumed sexual relations nine months after the birth. They both wanted to delay the arrival of a second child, but her husband was resolutely opposed to modern contraceptive methods. Janta, who feared another pregnancy and suffered from painful periods, took advantage of the child’s medical visits to ask for advice on contraception. She used injectable contraception without her husband’s knowledge, but suffered from side-effects (amenorrhoea, dizzy spells, bouts of nausea, etc.). Her husband discovered that she had deceived him and was furious with her. This led her to believe that her ailments were divine punishment for having disobeyed her husband, and she feared that this would prevent her from having any more children. For this reason, she did not return for a second injection, but started using the rhythm method again. A little while later, they made a joint decision to have a second child, and she was pregnant at the time of our interview.
60We have chosen to tell Janta’s story because it illustrates how far attitudes and practices regarding contraception vary, not only according to relationship contexts but also over the course of the same relationship. Throughout her life, Janta seems to have conformed to the demands of her social environment, which requires that a woman – even an independent one – remain under the authority of her family and then her spouse; yet the interpersonal and social contexts in which she found herself played a part in creating her specific contraceptive attitudes and practices. Her forced intercourse occurred in a context of submission to the man’s desires, where having sexual intercourse was socially unimaginable and where she had no social resources. Nevertheless, she later managed, with the same partner, to impose her own sexual and contraceptive choices, within the constrained framework of the valued family model. As she was financially autonomous – all the more so because her partner had no job – she even managed to escape a family setting that made her unhappy. Later, again, it was her internalized acceptance of male domination that explained her remorse about her secret use of injectable contraception with her second partner – even though it was her own choice – and her decision not to continue with this form of contraception. Her narrative of her everyday life shows that she played a part in decision-making in her current household – notably because she achieved a degree of financial independence through her soap-selling business – but still within the framework of a family model characterized by the unequal social positions of women and men.
V – Discussion
61Our qualitative research has brought to light the social factors that explain why contraceptive failures or unprotected intercourse occur, but, as always in research of this type, we cannot quantify the respective weight of these different factors. However, the social and demographic diversity of our samples in the four capital cities has enabled us to describe in detail the range of situations and to capture the way in which – in the urban environment – the difficulties with contraception defined here as “unprotected or poorly protected intercourse” arise, individually and socially, for women who do not wish to become pregnant.
62Like any analytical framework, the one we constructed from the women’s responses has its limits. In particular, to the extent that we did not collect narratives about contraceptive practices from both partners in a relationship, our female respondents can provide only an imperfect record of the importance of the interpersonal dimension, whether this involves disagreements between the two partners or, in contrast, clear support.  However, by placing contraceptive issues at the very centre of affective, social and sexual life stories we were able to confirm the importance of the interpersonal dimension, as well as the dynamic nature of contraceptive attitudes and practices at different stages of these trajectories.
63Certain types of contraceptive failure have not been highlighted in our analysis – for example, those linked to lack of information, whether resulting from inaccurate knowledge (duration of efficacy of hormonal contraception, fecund period, etc.) or from forgetting to take the pill. Such failures do not seem to be specific to the countries we studied, since they essentially reflect the difficult demands of using contraception throughout one’s reproductive life, and this difficulty is observed in other contexts, including those where contraceptive prevalence is very high (Moreau et al., 2006). Similarly, we did not explore issues of access to contraception in detail, as only married women in Morocco found such access relatively easy, even though – as we have seen – the question of supply is not simply one of accessibility but also of providers’ attitudes (see article by Mayhew et al. in this issue). 
Contraception, a central aspect of the life course
64Our analytical framework admits another interpretation of the factors involved in contraceptive failure and, in particular, enables us to deconstruct certain categories which are often cited as explaining unplanned pregnancies, i.e. those relating to contraceptive supply or to observance of the reproductive norm (Seiber et al., 2007).
65By gaining an understanding of the actual practices that are central to women’s social, sexual and emotional life histories, we have shown that unplanned pregnancies reflect women’s difficulties in coping with simultaneous normative pressures arising from gender, reproductive and sexual norms. These difficulties seem to vary a great deal, not only across life-cycle phases but also over the course of a single relationship, highlighting the limits of approaches on a strictly individual level and based on rational behaviour. Our results show that, with the exception of the very few women who know nothing about the risks of pregnancy, the great majority of women adhere to the idea of controlling their fertility through contraception. However, at certain times in their lives, this does not necessarily lead to regular – and therefore effective – contraceptive use, because, apart from the normative ambivalence sometimes exhibited by professional providers, the dominant social and family model may, in certain circumstances, conflict with the use of any contraception, hormonal contraceptives in particular.
66From this point of view, issues of access to and use of contraception, some of which appear to represent particular obstacles to effective contraceptive practice, differ across the various life-cycle phases. There is a very clear dividing line between women whose sexuality is not socially accepted and the others. Women who have a sexual life outside a conjugal union, notably before marriage, seem particularly likely to encounter difficulties with contraception, and this issue of relationship status far outweighs that of not knowing how to use contraception. They find it difficult to envisage using contraceptive services that primarily target married women, and are forced to resort to methods that are less effective and require their partner’s participation. Yet the man does not always feel that he is involved, and he may even be opposed to the use of certain methods which would reduce his sexual pleasure, as shown in the accounts given by some women we interviewed.
67In Southern societies, where childless or sterile women cannot gain real social recognition (Journet, 1990), women must maintain their ability to acquire status through childbearing within marriage. Marriage marks an obligatory watershed in the lives of women and men; there is often a strong expectation that couples will have a child immediately after marriage, and sexuality outside this recognized customary or legal framework is socially stigmatized (Bankole et al., 2006). As we have seen, the use of contraception immediately after marriage and before the birth of a child is problematic.
68Beyond the specific nature of the different life-cycle phases, the need to take account of the type of relationship is an important factor in interpreting women’s narratives. This perspective reveals how gender relations construct specific contraceptive attitudes and practices (Bawah et al., 1999; Greene and Biddlecom, 2000; Andro and Desgrées du Loû, 2009). Negotiation between partners around contraceptive use is inextricably linked with the power relations that exist within a couple (Andro and Hertrich, 2002; Kaggwa et al., 2008; Kulczycki, 2008; Tolley et al., 2005; see also Adjamagbo and Koné’s article on unwanted pregnancies in Dakar, and Rossier et al. on young people in Ouagadougou in this issue). The primacy of male sexual pleasure, which is found in all societies (Tabet, 1985; Héritier, 1996), is also a manifestation of these power relations. It influences the decision to use a form of contraception, be it periodic abstinence or condoms. The latter, used almost exclusively outside marriage, have taken on particular significance because of the HIV pandemic. It is mostly with the intention of maintaining the quality of male sexual pleasure – often in order not to threaten the relationship – that a condom may not in the end be used, even if the woman wants to protect herself against unplanned pregnancy. Some women are themselves reluctant to use condoms as their own pleasure is diminished; this can sometimes lead to unprotected intercourse, even though the women concerned are aware of the risk of pregnancy. This kind of relationship, in which the power relations between partners are considerably more egalitarian, remains the exception, however.
69The social rationale of male domination that leads to the primacy of male sexual pleasure is taken to its extreme in cases where intercourse is forced upon the woman. Such encounters, frequently mentioned by our respondents, are, by definition, unanticipated and therefore unprotected if the woman is not already using contraception.
70Finally, our research also reveals the importance of analysing contraceptive issues in relation to the “phases of a relationship” (Kulczycki, 2008). In that regard, the example of Janta was an eloquent one, since this young woman found she had difficulty with contraception in several different situations in turn, sometimes with the same partner. The great majority of the women we interviewed, like Janta, experienced numerous different contraceptive situations over the course of their lives. Here, our results mirror those established in other national contexts or for other health issues such as HIV infection (Bajos and Marquet, 2000): it is not a question of “at-risk people”, but rather, of situations that encourage risk-taking.
71The similarity of our research protocols used in the four countries to explore the contraceptive difficulties faced by women, whatever the context of their everyday lives, enabled us to highlight the factors – notably in terms of the strength of sexual and family norms – which are specific to national contexts, and those which transcend them. Our analysis of the interviews recorded in the four countries operates at both the micro sociological level, through the respondents’ personal narratives, and the macro sociological level, through consideration of the social, legal, political and normative contexts (Gagnon and Simon, 1973). The aim is to distinguish the specific features of the social models in each country that reflect linkages between elements as diverse as the legal and court system relating to the family, inequalities of access to the healthcare system, the education system or the labour market, the mode of organization of the family, the impact of religion, the level of recognition of individual autonomy and the ways men and women relate to one another materially and symbolically.
72Analysis of our interviews shows that the factors associated with difficulties in contraceptive use are relatively similar in all the countries and include, in particular, the prime importance of maintaining fecundity, the stigmatization of sexuality outside marriage and the primacy of male sexual pleasure. However, what distinguishes one country from another is the degree of rigidity of these norms and the extent to which gendered “double standards” apply. The example of the stigmatization of premarital sex illustrates the influence of specific national features. It is particularly marked in Morocco, where all sex outside marriage is forbidden. There are no data on the subject in population surveys (no modules dealing with sexual debut as distinct from first union formation, for instance) and family planning programmes are targeted exclusively at married women. Young Moroccan women avoid penetrative sex and therefore do not feel concerned about pregnancy risk or contraceptive techniques. For them, preserving their virginity takes the place of contraception – though the risk of unplanned penetration still exists (see the article by Bakass and Ferrand in this issue). In Senegal, practices that avoid penetration before marriage are sometimes mentioned, but more rarely than in Morocco. As in Burkina Faso, premarital sex is tolerated for women and accepted for men. Childbearing outside marriage is still not socially accepted, but we frequently encountered pregnancies described as “unplanned” and used strategically to create a situation that would lead to marriage or to the acceptance of a union in the family circle. In Ghana, by contrast, premarital sex is more accepted and does not appear to be an obstacle to young women’s contraceptive use.
73Finally, it was with regard to the characteristics of gender and family models that difficulties with contraception turned out to be nationally specific. A contrast between Morocco and the three other countries emerged from our analyses: Morocco has the highest recorded contraceptive prevalence, in a context where long-standing policies promote widespread use of contraception and are well rooted in the public health sector. Moroccan women – and indeed, Moroccan men – seem to adhere fully to the two-child model of marital fertility. For women in the other three countries, this was a model adopted by default, currently valued for health and economic reasons, but with some ambivalence: it is often perceived as a “Western-style” model imposed from outside, except – as Johnson-Hanks showed in her study in Cameroon (2006) – for women who have sizeable social capital.
74While contraceptive prevalence differs across our three sub-Saharan African countries, the use of medical contraceptive methods (the pill, implants, IUD, sterilization) is fairly similar. The absence of marked cross-national differences between women’s attitudes to contraception and contraceptive use may reflect similar changes in family and sexual models. But this trend could also be a superficial one, concealing the persistence of a patriarchal type of domination (by father and by husband) which ultimately leaves women with little real autonomy for managing their reproductive, sexual and emotional lives.
75In all, our results show that an approach based on difficulties with contraception – that is, unprotected or poorly protected intercourse involving women who do not want to become pregnant and most of whom accept the idea of using contraception – identifies the range of situations where women may face an unplanned pregnancy and in some cases decide to terminate it. In many such cases, emergency contraception is used (Teixeira et al., 2012). Consideration of the normative tensions to which women are subject reveals difficulties with contraception other than those linked to healthcare access or simple observance of the reproductive norm. In any case, the question of effective contraceptive use goes far beyond the issue of “unmet needs”, a concept widely used in the field of family planning to assess the effectiveness of programmes and which focuses solely on non-use of contraception, most often applying the rational actor paradigm.
The composition of the ECAF team is presented in the introduction to this group of articles.
Calculations based on the results of the following surveys: Enquête sur la population et la santé familiale (EPSF), Morocco, 2003-2004; Enquête démographique et de santé au Sénégal, 2005; Enquête démographique et de santé du Burkina Faso, 2003; Ghana - Demographic and Health Survey, 2003.
Our research methodology is set out in detail in the introduction to this issue.
Although the research design did involve interviews with men (25 per country), these were not the partners of our female respondents, so their responses cannot be matched.
Apart from their technical function, these family planning providers often seem – from the women’s point of view – to be social actors who also promote the dominant sexual and family model, reminding women of their duty to bear children within marriage or requiring the presence of the (future) spouse when a single woman requests contraception.