Origins and development

6The exact historical and geographical origin of female genital mutilation is unknown. The hypothesis that the practice originated in the Middle East and the Arabian peninsula and was then carried across the African continent by Arab traders is not shared by all specialists (Erlich, 1986; Hosken, 1982). What does seem to be accepted is that FGM is an age-old practice, possibly dating as far back as Ancient Egypt, [3] which may have originated in what is now Sudan and Egypt. The archaeological community is divided over whether marks found on Egyptian mummies are evidence of excision (Knight, 2001). The first reference to excision, recorded on papyrus, dates from the second century BCE in Egypt (Couchard, 2003). Later sources include accounts of travellers like the Ancient Greek geographer Strabo, who, after travelling to Egypt (around 25 BCE), described the operation as a customary practice (Hosken, 1982).

7According to Mackie (1996), female genital mutilation spread from the western shore of the Red Sea (in what is now Egypt) to neighbouring regions of Africa to the south and west. He also establishes a link between infibulation, [4] the most invasive form of FGM, which is mainly practised in eastern Africa (Eritrea, Djibouti, Somalia, Egypt and Sudan), and the slave trade, particularly during the period of Islamic expansion in Africa. This extreme form of FGM, whose name is derived from the Latin fibula (a brooch or pin), may also have been practised on female slaves in Ancient Rome to prevent sexual intercourse and avoid pregnancies, which would have rendered slaves unfit for work (Hosken, 1982). Despite the uncertainty about its origin, the evidence suggests that FGM existed long before the emergence and expansion of Islam in Africa, even if religious justifications were subsequently used to legitimize it. This is supported by the fact that FGM is practised in communities of Christians (Copts, Catholics and Protestants), Jews and animists. Unlike male circumcision, which, in Judaism and Islam, is the sign of a covenant between God, Abraham and his descendants, there is no commandment on excision in the books of the main monotheistic religions [5] (Couchard, 2003; Thiam, 1978).

8Female genital mutilation (clitoridectomy [6] and infibulation) has also been historically practised in the Western world, even if not imposed on whole communities. Chastity belts, [7] a form of mechanical infibulation as opposed to the scarring practised on female slaves in Ancient Rome, were used in the Middle Ages (Hosken, 1982). In the nineteenth century, the pathologization of certain sexual practices, particularly female masturbation (Laqueur, 2005), led to the practice of surgical clitoridectomy, believed to cure the ills and deviant behaviour of women who lacked sexual restraint. This type of surgery, mainly practised in Europe in a context of repressive medicalization of sexuality, was first performed by a British doctor, Isaac Baker Brown, who believed it to be an effective cure for female masturbation and hysteria [8] (Sindzingre, 1979). Although Baker Brown was expelled from the medical profession in 1867, in the United States the practice persisted into the 1960s (Cutner, 1985).

9More recently, gender reassignment surgery performed on intersex newborn babies has been called genital mutilation by campaigners for the rights of those concerned (Löwy, 2003). This type of surgery, first performed in the 1950s, [9] is still practised in some countries, including France (Lee et al., 2006).

A rite of passage or a marker of unequal gender relations?

10Excision was first described in the anthropological literature, giving rise to functionalist and culturalist analyses linked to a psychoanalytical approach (Sindzingre, 1979). Female genital mutilation was mainly seen as a rite of passage, according to the three-phase interpretive model (separation of the individual from the group, marginalization then reintegration) established in the early twentieth century by the ethnographer Arnold Van Gennep (1909). Under this type of approach, which has been applied to various regions of Africa, excision is considered equivalent to male circumcision and is often referred to as “female circumcision” to emphasize the analogy between the two practices, which are described as markers of gender, age and sometimes ethnicity (Cartry, 1968; Chéron, 1933; Colleyn, 1975; Droz, 2000; Muller, 1993). These studies provide detailed documentation of initiation ceremonies, and a degree of justification, by emphasizing the mythical aspects of the rituals. [10]

11These approaches were challenged in the 1970s, when the feminist campaign against excision was at its most vigorous. The equivalence between circumcision and excision was strongly contested, along with their common theoretical framework, and female genital mutilation was reframed within the broader issue of gender relations. Nicole Sindzingre was the first in the 1970s to argue against the idea of excision as a rite of initiation. She highlighted the asymmetry in practice between male circumcision and female excision ceremonies described in the anthropological literature. Firstly, in terms of its impact on physical integrity, female genital mutilation is not the equivalent of male circumcision. Furthermore, while male circumcision ceremonies are described as collective rituals with high social value, excision is usually presented as a “shortened” rite, [11] conducted within the family circle and centred on the individual (Sindzingre, 1977, 1979). However, it is primarily through the justifications for the practice – a concern to eliminate sexual ambiguity or original androgyny, a requirement of “purification” as a pre-requisite for marriage and childbirth, and a wish to curb sexual urges in order to ensure a girl’s virginity and a wife’s fidelity – that excision ties in with the question of the representation of femininity and gender relations more broadly.

12Within the variety of discourses on female genital mutilation, it is possible to identify a common logic that not only links the practice to a concern for biological reproduction (through marriage and procreation) but also to a concern for social reproduction, since this sexual marking also marks the social roles of each gender. In many societies, the clitoris represents the “male part” with which the female sex is endowed at birth, a representation that is also found in myths of original androgyny or bisexuality [12] (Couchard, 2003). Removing the clitoris thus provides a necessary means to make women’s bodies completely feminine (and exclusively devoted to procreation), but also to place them in a subordinate position within the male order by conferring on men the exclusive exercise of male authority, symbolized by the clitoris, the equivalent of the penis [13] (Fainzang, 1985). Taking up Pierre Bourdieu’s analysis (1982) of rites of institution, a term he preferred to rites of passage, excision can be seen as a ritual practice to legitimize the difference between the sexes that underpins unequal power relations: excision is designed to “de-virilize” the woman in order to reduce her power, whereas circumcision “re-virilizes” the man in order to increase his authority (Fainzang, 1985). This paradigm, which denounces FGM as violence against women and incorporates the practice into the construction of unequal gender relations, has not been totally effective in deculturalizing the practice [14] (Boni, 2009). FGM has since been analysed not only in terms of the imposition of patriarchal social norms but also from the perspective of the right to physical integrity and sexual freedom (Mbow, 1999).

The stages in the international campaign

15The United Nations Commission on Human Rights first discussed the traditional practice of FGM in 1952. In 1958 the UN Economic and Social Council explicitly raised the issue of FGM and the harm it causes as a problem for the international community (Resolution 680 B II (XXVI) of the Economic and Social Council: Ritual Operations, 1958). At that time, the practice was approached primarily from a culturalist viewpoint. The World Health Organization refused to become involved, at the time considering FGM as a social and cultural practice rather than a health issue and therefore outside its competence (United Nations, 1959).

16In 1977 the NGO Working Group on Traditional Practices was set up, opening up a discussion of the consequences of FGM on the health of women and girls. The previous anthropological approach to the practice had effectively rendered the harmful effects of FGM invisible (Thiam, 1978). In 1979, the WHO took a stance on the issue for the first time by inventorying the medical consequences of FGM. The WHO’s Regional Office for the Eastern Mediterranean in Khartoum convened a seminar on “traditional practices affecting the health of women and children”, attended by NGOs and doctors, at which Fran Hosken presented her report on genital and sexual mutilation of women (WHO, 1979).

17At the World Conference for the United Nations Decade for Women, held in Copenhagen in 1980, there was a tense confrontation between the European and African delegations. The majority of the latter were still calling for the practice to be recognized as a rite of passage to adulthood on a par with the circumcision of boys (Sow, 1997). However, by the global conference on women in Nairobi in 1985, positions had changed and a broader consensus began to emerge, with recognition that the practice was harmful. International agencies became increasingly involved from that date onwards. The Working Group on Traditional Practices Affecting the Health of Women and Children submitted its first report to the UN Commission on Human Rights in 1986 (E/CN.4/1986/42). In the 1990s efforts to ban FGM became more structured. In 1990, the Inter-African Committee on Traditional Practices, set up by feminist organizations, adopted the term “mutilation”, following UNICEF’s lead.

18The UN General Assembly adopted the Declaration on the Elimination of Violence against Women in 1993, which refers explicitly to female genital mutilation. In 1994, the United Nations Sub-Commission on Prevention of Discrimination and Protection of Minorities adopted the first Plan of Action for the Elimination of Harmful Traditional Practices affecting the Health of Women and Children. The United Nations’ abolitionist stance was reiterated at the International Conference on Population and Development in Cairo in 1994 and Fourth World Conference on Women in Beijing in 1995.

19Under the new policy framework, the World Health Organization sponsored the first joint statement with UNICEF and UNFPA in 1997, officializing their support for programmes to prevent and eliminate the practice of FGM and undertaking to support the action of governments in that direction (WHO, 1997). Knowledge of and mobilization on the issue prompted the WHO to draft the first typology of FGM in 1997, jointly with UNICEF and UNFPA (WHO, 1997) (see Section I.3).

20International legal instruments could not have been developed and adopted without the campaigns in the countries concerned. Since 1984, the role of the Inter-African Committee on Traditional Practices has been fundamental. The 1981 Protocol to the African Charter on Human and Peoples’ Rights on the Rights of Women in Africa, known as the “Maputo Protocol”, is a legal instrument, adopted by consensus in 2003 by the heads of state of the African Union. Article 5 of the protocol explicitly prohibits and condemns FGM and other harmful practices. It calls on the signatory states to take measures to develop public awareness, to pass legislation backed by sanctions to prohibit FGM, to support victims of harmful practices and to protect women who are at risk (zero tolerance to FGM). In 2008, an inter-agency statement led by the WHO, UNICEF and UNFPA set forth the international position on eradicating female genital mutilation (WHO, 2008).

From controversy to globalization of the issue

21The gradual emergence of a consensus around an international policy on FGM has been hampered by the competing discourses of various international bodies. Elizabeth Boyle (2005) pointed out that, within the United Nations itself, the recognition of the universal rights of women and the right to bodily integrity has long competed with the principles of sovereign autonomy and respect for traditions and family transmission. In the end, the former principles took precedence in the elaboration of the international doctrine on FGM.

22The doctrine is underpinned by two legal principles: the right to health and human rights. Some authors have described the “uneasy alliance” between human rights and the right to health in discussions of FGM (Gruenbaum, 2001; Hernlund and Shell-Duncan, 2007). It was through emphasis on the health effects of FGM that the practice came to be seen not in terms of a ritual of socialization but as a grave violation of the physical integrity of the women subjected to it, thus providing grounds for analysis from a human-rights perspective (Abusharaf, 2006). However, the health approach has also proved counter-productive, because opponents cite a lack of medical evidence (Obermeyer, 1999) and because of the medicalization of FGM procedures (Section VI).

23Moreover, the motives behind the efforts of international feminist movements to ban the practice have long come under suspicion. The international campaign has too often portrayed African women as enduring the custom without resisting it, even though it endangers the lives of their daughters. This reductionist representation has led to the international campaign being perceived as racialist and post-colonial, taking the form of a crusade by feminists from the North that has overshadowed the initiatives emanating from the societies concerned (Boddy, 2007; La Barbera, 2009).

24Perceptions of the practice have nonetheless changed considerably since the turn of the twenty-first century. FGM, perceived as an exclusively African problem in the twentieth century, has now become a global issue, for two main reasons. Firstly, recent studies show that FGM is also traditionally practised in other regions of the world, where the prevalence of the phenomenon was previously unknown, and in some countries of the Middle East [15] and Asia, [16] particularly Indonesia (UNICEF, 2015). Secondly, the globalization of migration flows and the settlement in Northern countries of families from regions where FGM is traditionally practised have led destination countries to consider the practice as a domestic public health issue (Bell, 2005; Johnsdotter and Essen, 2010). This new dimension of FGM has raised questions about the possible existence of the practice in yet more countries and about the implications of perpetuating or abandoning the practice in the context of migration.

Legislative developments

25In the countries of origin and of destination, legislation on FGM has gradually moved in line with the intensified international and regional efforts to ban the practice since the 1990s (Toubia, 1993). In Dakar, in 2005, the African Parliamentary Conference adopted a resolution calling on states to enact laws to ban FGM. [17]

26Of the 30 countries with the highest prevalence of FGM, 25 have passed decrees or laws on the practice in recent decades. In the vast majority of countries, laws have been passed since the late 1990s; [18] in 15 countries, they were introduced in the 2000s and 2010s (Appendix Table A.2). The scope of this legislation varies considerably across countries [19] and the divergence between international standards and local social norms makes it difficult to enforce (Boyle et al., 2002).

27The introduction of a legislative framework in the countries of origin has been simultaneous with similar developments in the countries of immigration. The first destination countries to criminalize FGM, in the late 1970s and early 1980s, were France (1979), Sweden (1982) and the United Kingdom (1985). The United States, Canada, Australia and Norway passed legislation in the 1990s, and the other European countries in the 2000s (Boyle, 2005). Some European countries have specific laws on FGM, while others (such as France) have included FGM in their legislation on child abuse and mutilation (European Institute for Gender Equality, 2013). Almost all of the laws include a principle of extra-territoriality, which makes it possible to protect girls who habitually reside outside their country of origin; young girls are often at higher risk of undergoing FGM during temporary stays in their parents’ home country. These legislative provisions have led to prosecutions in six European countries, although for many years France was the only country to have taken cases of FGM [20] to court (Boyle, 2005; Leye et al., 2007). In France, FGM has been a criminal offence since 1979 (Articles 222.08, 222.09 and 222.10 of the Criminal Code); in 2006, the statute of limitations was extended to allow victims to life a complaint up to 20 years after their majority at age 18.

The current WHO classification

29In 1997, the WHO proposed its first classification based on four types of practice according to the anatomical extent of the cutting (Table 1) (WHO, 1996, 2008). After the typology was released in 1997, experts pointed out some limitations, namely that the proposed categories over-simplified the diversity of actual practices. The classification was revised in 2007, based on the conclusions of a group of experts commissioned by the WHO. The categories in the 1997 classification were amended slightly and subdivisions were created to cover the wide range of procedures more fully. The inter-agency statement, published jointly by eight UN agencies in 2008, indicates that FGM encompasses a range of practices that, while they all violate the integrity of the female genitalia, are nonetheless extremely varied (WHO, 2008).

30Since 2008, the WHO has therefore recommended that female genital mutilation be classified into four main types, defined on the basis of the procedure performed at the time of the mutilation: Type I, often described as clitoridectomy (partial or total removal of the clitoral hood and clitoral glans); Type II, often called excision (removal of the inner labia and the clitoris); Type III, often called infibulation (narrowing of the vaginal orifice by stitching the outer labia over the opening, with or without removal of the clitoris); and Type IV which includes the other less common types (incising, cauterization, scarring). Types I, II and III can be further divided into sub-types (Table 1). The most common forms of mutilation are Types I and II. In West Africa, the most common form of FGM is Type II, whereas the rarer Type III is mainly found in eastern Africa (UNICEF, 2013) (Section III.2).

The limitations of the classification

31Until the 2000s, specific modules on FGM in socio-demographic surveys (Section II.1) explicitly asked women about the type of FGM they had undergone by inviting them to choose from one of the three main types defined by the WHO (excision, clitoridectomy and infibulation). [21] The quality of the information gathered was questionable, however. Several studies comparing the data collected from respondents with the data from clinical examinations revealed considerable discrepancies, particularly in the regions where Type III FGM (infibulation) is traditionally practised, and where the women often reported having undergone Type I or II (Elmusharaf et al., 2006b). In practice, the women do not always know which procedure they have undergone and are unable to give a precise answer. Moreover, the terms used by respondents to describe the types of mutilation performed on them vary across contexts and do not always conform to the WHO’s physiological descriptions (Yoder et al., 2004). Starting in the 2000s, the question on FGM was reworded in the Demographic and Health Surveys (DHS) and the Multiple Indicator Cluster Surveys (MICS), primarily in order to map prevalence of the most invasive procedure, i.e. Type III. [22] The first clinical studies performed in the 1990s showed that Type III FGM was associated with more serious health risks, particularly obstetric complications (Obermeyer, 1999, 2003; WHO Study Group on Female Genital Mutilation and Obstetric Outcome, 2006). Although the WHO classification appears to be unsuitable for surveys based on self-reporting (Section II.3), it is still useful for clinical studies (Yoder et al., 2004).

32The classification developed by the WHO in 1997 was revised in 2007 because the categories initially proposed were too reductionist and failed to capture the diversity of procedures (Table 1). The typology is constructed on the basis of two factors: the extent of tissue removal and the type of procedure performed at the time of the mutilation (cutting and/or stitching). It involves assessing the amount of tissue removed by the FGM practitioner, which varies by region, ethnic group or age when the FGM was performed; and reporting whether the vulva was stitched or not. The hypothesis of a causal link between the extent of tissue removal and the severity of consequences is central to the WHO typology. It is not always verified, [23] however, and the severity of consequences (particularly psychological and sexual) can vary with socio-demographic characteristics (age and marital status). Moreover, the typology does not consider the social and health environment in which the women concerned are now living. Among migrant women, the quality of obstetric healthcare at the time of childbirth in the country of immigration can minimize the consequences of FGM; the situation is very different in countries where little perinatal care is available (Andro et al., 2014; Essén et al., 2005; Zenner et al., 2013).

In the countries of origin

36Data in the countries of origin come from two main sources: Demographic and Health Surveys (DHS) [26] and Multiple Indicator Cluster Surveys (MICS) organized by UNICEF. [27] The first module specifically on FGM was introduced in the individual questionnaire for women in the DHS conducted in North Sudan in 1989-1990, then extended to the DHS conducted in all of the African countries concerned by the practice (Côte d’Ivoire, 1994; Egypt, 1995; Eritrea, 1995; Mali, 1995-1996; Central African Republic, 1994-1995). The FGM module is now included in the DHS in 25 countries (Yoder and Wang, 2013). Since the 2000s, the MICS have also been used to gather data on FGM in 17 countries, including seven [28] for which no data had previously existed (UNICEF, 2013). The Indonesian survey conducted on 30,000 households in 2013 was not a DHS- or MICS-type survey, and the questions on FGM only concerned girls aged under 12 (UNICEF, 2015). In the 30 countries where FGM is concentrated (all in Africa, except for Iraq, Yemen and Indonesia), 89 nationally representative surveys are available, covering a 25-year period (1989-2014). For almost seven in ten of these countries, the data from at least three surveys are available (Appendix Table A.1).

37The module on FGM in the DHS questionnaires is standardized, although there are some variants in different countries and some changes since the first version in the 1990s (Yoder et al., 2004; Yoder and Wang, 2013). The module, administered to female survey respondents aged 15-49, is introduced by a filter question on knowledge of FGM. The module consists of three sets of questions (Appendix Table A.1):

• The respondent’s own FGM status: cut or not, type of cutting, circumstances of cutting (age when cut and person who performed the procedure);
• The FGM status of the respondent’s daughter(s) (aged under 15): [29] cut or not, type of cutting, circumstances of cutting (same questions as for the mother) and intention for the future (asked of women who had at least one daughter aged under 15 who had not been cut at the time of the survey);
• Perceptions and attitudes: benefits of cutting/not cutting, reasons for the practice, attitude to continuing or abandoning the practice and perception of its impact on health.

38In the 2000s, the questions on perceptions and attitudes were also included in the individual questionnaire administered to male survey respondents. Since 2010, the DHS and MICS have used a similar questionnaire. Some questions were removed (the health impact of FGM, respondents’ intentions for their daughters), while other questions (cut or not, type and circumstances of cutting) were extended to include all daughters aged under 15 living with their mother (Yoder and Wang, 2013).

39The data were analysed to measure the extent of the practice by calculating the percentages of women and girls who have undergone FGM in each country. These indicators are considered to be prevalence rates in the epidemiological sense. The prevalence of a condition at a point in time t is the number of cases (individuals) with the condition (here, having undergone FGM) relative to the total population (here, the total number of women). This measure, based on representative samples, is then extrapolated to estimate the total number of women and girls who have undergone FGM (Yoder and Khan, 2008; Yoder et al., 2013). Furthermore, matching the data on FGM against the women’s socio-demographic characteristics gives an indication of the characteristics and determinants of the practice, which vary from one country to another, or within the same country, by ethnicity, educational level, income, etc. The variations in prevalence by age group and the existence of data collected at different dates [30] make it possible to assess trends in the phenomenon over time. Lastly, information about perceptions and attitudes (collected from men and women) gives an idea of the rationales underpinning the abandonment or perpetuation of FGM in these countries. Since the late 1990s, the results of the DHS and MICS have been presented in several reports that give a detailed overview of the practice in the most affected countries (Carr, 1997; UNICEF, 2005, 2013; Yoder et al., 2004, 2013; Yoder and Khan, 2008; Yoder and Wang, 2013).

In countries of immigration

40In Europe and North America, FGM concerns only a specific part of the population, namely women who originate from at-risk countries. FGM is not a social norm in these regions; on the contrary, it is a deviant, clandestine practice, which is prohibited and has been against the law for several decades. In countries of immigration, there are no nationally representative surveys with a DHS-type module on FGM. In the 2010s, two socio-demographic surveys explicitly on FGM were conducted in two European countries: Italy (Farina and Ortensi, 2014b; Ortensi et al., 2015) and France (Andro et al., 2009). The target populations were migrant women (and daughters of migrants in the French survey) and the surveys were conducted in a sexual and reproductive health framework. The Italian survey was performed in a single region, Lombardy, on a representative sample of 2,011 migrant women and girls aged 15-49; the French survey was conducted in five regions [31] on a sample of 2,882 migrant women aged 18 and over. The target population of both surveys (women having undergone or at risk of undergoing FGM) is small and hard to reach. Applying survey protocols designed to overcome these problems (Marpsat and Razafindratsima, 2012), the women were surveyed at health centres (family planning centres, motherand-baby centres, gynaecological appointments in hospitals, etc.). They were selected using time-location sampling (TLS), [32] combined with respondent-driven sampling for the Italian survey. [33] In both surveys, questions about the FGM status of the women surveyed and their daughters were asked using the module on FGM from the DHS. The French survey was also designed as a case-control study to measure the effects of FGM on the health of the women concerned (in the same way as clinical surveys). It also contained a module on reconstructive surgery, [34] which is available in France; a reconstructive surgery programme was set up by a French urologist in the 1990s, covered by French public health insurance since 2004 (Section VI.3).

Uncertainty linked to self-reporting

43FGM status recorded by socio-demographic surveys is based on self-reporting by the women surveyed. It is assumed that the women are aware of their condition, and are able to answer the questions without fear. The first assumption, that cut women have an accurate awareness of their status, is not always verified. Several studies, which compare women’s self-reports with the findings of clinical examinations by healthcare practitioners, reveal discrepancies between the two: while one study, conducted in the Gambia, found a difference of only 3% between the two types of data (Morison et al., 2001), studies conducted in Tanzania and Nigeria found a larger divergence (Klouman et al., 2005; Snow et al., 2002). Researchers attribute these differences to two main factors: firstly, some women, who underwent FGM at very young ages, are not fully aware of their status, and secondly, some more superficial types of FGM do not necessarily cause a visible alteration of the external genitalia and are not diagnosed by clinical examination.

44The same observations have been made in migrant populations, particularly in the French survey, which included respondents’ self-reports and diagnoses by healthcare practitioners (with the women’s prior consent): among the respondents for whom both types of data are available (60% of the sample), the match was around 90%. More than half of the difference could be attributed to the clinician’s failure to establish a diagnosis (the clinician answered, “Don’t know”). In countries of immigration, such diagnostic failures are linked to a lack of medical training in identifying FGM (Andro et al., 2009). Interviews have also revealed that it is fairly common for women to discover their FGM status only when they become sexually active, and in some cases only when they give birth (Andro et al., 2010).

Under-reporting linked to the legislative context

45Another under-reporting bias may be linked to legislative changes in certain countries (Section I.2). A longitudinal study conducted in northern Ghana in 1995 and 2000 assessed the consistency of women’s self-reports over time: 15% of the women surveyed on both dates gave different answers, with the majority of that group reporting having undergone FGM in the first survey in 1995, and of not having undergone FGM in the second survey in 2000. The researchers attributed the difference to a legislative effect: the first law criminalizing FGM in Ghana was introduced in 1994 and was followed by awareness-raising campaigns (Jackson et al., 2003). [37] A reluctance to report having undergone FGM in an environment where the practice has been banned has also been observed in other African countries where data have been collected on different dates. [38] From the mid-2000s, when international and African bodies intensified their campaign against FGM, [39] several DHS surveys recorded unexplained decreases in the prevalence of FGM in some age groups, which did not seem to reflect real declines but were probably the result of under-reporting by the women surveyed (UNICEF, 2013).

46Such under-reporting is even more likely in migration contexts, particularly in surveys of migrants’ descendants with origins in an at-risk country. In France, for example, where the practice of FGM has no social legitimacy in the mainstream population and where the legislation is particularly strict (Section I.2), it is difficult for women born or raised in France to report having undergone FGM and even more difficult for them to report FGM performed on their daughters. It is therefore important to consider the contexts in which the questions on FGM are asked, in order to adapt the survey protocols accordingly and to increase the number of data sources (Askew, 2005).

Direct estimates on the basis of socio-demographic surveys

49In 1997, an initial estimate (Table 2) established on the basis of Demographic and Health Surveys suggested that there were 30 million women and girls with FGM in seven countries (Carr, 1997). Ten years later, aggregated data from 27 African countries led to an estimate of 92 million (Yoder and Khan, 2008). In 2013, the estimated number in Africa and the Middle East was 125 million (UNICEF, 2013; Table 2). In February 2016, UNICEF published a new estimate of the affected population of women and girls around the world, updating estimates established in 2013 (on the basis of population growth rates) and adding the population of women and girls with FGM in Indonesia on the basis of data collected in 2013 from girls below age 12. The UNICEF estimate increased from 125 million to 200 million. This large difference is linked notably to the demographic weight of Indonesia (255 million inhabitants in 2015), where an estimated one in two girls or women have undergone FGM (UNICEF, 2016).

50The calculation method used in the most recent and precisely documented estimates (2008 and 2013) is based, first, on the proportion of women with FGM in each country [42] as calculated on the basis of DHS and MICS survey data, and second, on the numbers of women in each country as indicated by the US Census Bureau.

51The sample populations of the DHS and MICS surveys include only women aged 15-49. An initial direct estimate can be established by applying the prevalence rates provided by demographic surveys to total numbers of women aged 15-49, breaking down the rates into five-year age groups as prevalence can vary across age groups (Section IV.2). For women aged 50 or above and for girls aged 10-14 (for whom prevalence data are lacking), the rates for the closest known age group (respectively, 45-49 years and 15-19 years) are applied (Appendix figure A.1).

Indirect estimates in the absence of survey data

52In countries of immigration, direct estimates are impossible for two reasons: the first is the lack of representative surveys at the national level comparable to the Demographic and Health Surveys (DHS) which include a module on FGM for the whole female population residing in these countries (Section II.1). [43] The second resides in the difficulty of identifying the relevant population, notably in countries with no population register. This population consists of immigrant women (born abroad) from countries where FGM is traditionally practiced, and women born in countries of immigration to at least one parent from one of these countries. For the first group, depending on the country, public statistical data by country of origin is not always available (notably due to the small numbers of relevant individuals), and some may also have a residency status that makes identification very difficult (undocumented individuals, refugees, asylum seekers). Women in the second category can only be identified using knowledge of their parents’ country of birth, a question that is rarely asked in large national surveys (Simon, 2012).

53There is thus no clearly defined, homogeneous methodology for estimating prevalence in the various countries of immigration. The European Parliament resolution of 24 March 2009 on Combating Female Genital Mutilation in the EU (European Parliament, 2009) estimated that there were around 500,000 women with FGM living in the EU, and that 180,000 girls were at risk of FGM each year. The methodology used to arrive at these figures was not specified (Leye et al., 2014). While there is currently no way to calculate an overall estimate (like those established for the countries of origin), estimates produced using indirect methods – based on the extrapolation of observed prevalence in countries of origin – are available for a number of countries (Table 3). [44]

Table 3

Estimates of the total number of women with FGM in countries of immigration^(a)

Reference Number of women and girls with FGM Country Types of data used Andro and Lesclingand, 2007 53,000 (a) France Survey combined with census (Étude de l’histoire familiale [family history survey]) and DHS-MICS Ministero delle Pari Opportunita, 2009 35,000 Italy Population register, residence permit data, and DHS-MICS Hänselmann et al., 2011 24,000 Germany Population census and DHS-MICS Dubourg and Richard, 2011 13,000 Belgium National population register, register on refugees and asylum seekers, birth records, and DHS-MICS PRB, 2013 507,000 United States Census and DHS-MICS Exterkate, 2013 29,000 Netherlands Census, register of female asylum seekers, DHS-MICS Macfarlane and Dorkenoo, 2014 137,000 England and Wales Population census, birth register, and DHS-MICS

Estimates of the total number of women with FGM in countries of immigration^(a)

(a) The estimate was for adult women only.
Note: For complete country-by-country documentation on European countries, see the website of the European Institute for Gender Equality (EIGE): http://eige.europa.eu/gender-based-violence/literature-and-legislation

54This indirect estimation method consists of applying the observed prevalence in countries of origin to the populations of women and girls from at-risk countries (Appendix figure A.2). Its details vary depending on the public statistical data that are available for each country (Leye et al., 2014). In Europe, estimates were established beginning in 2005, notably in the western European countries with the largest populations of immigrants and descendants of immigrants from at-risk countries (Belgium, France, Germany, Italy, and the United Kingdom). In the late 2000s, on the initiative of the European Institute for Gender Equality (EIGE), a group of European experts undertook a major review of existing work on FGM in Europe, and notably of estimates produced in individual countries (European Institute for Gender Equality, 2013). Generally, the first step consists in identifying the reference population, defined as all women and girls who come from the 30 countries where the practice of FGM exists and is documented, or with at least one parent from one of those countries (Section II.1), by means of different sources (population census, population registers, general population surveys, registers of refugees or asylum seekers, etc.). The prevalence rates provided by the DHS/MICS surveys are then applied to this reference population (Appendix figure A.2). Depending on the variables available in a given country of immigration, these rates may be broken down by age, level of education, and age at arrival in the country (Leye et al., 2014).

55These indirect estimates are subject to a number of limitations and biases. The identification of the relevant population depends on the data available from censuses, the existence of a population register, and ease of access to registers of asylum seekers and births. The heterogeneity of sources makes it difficult to use a common methodology in different countries. Moreover, depending on the history of migration flows to each country, the presence of a second, or even a third generation also implies locally specific definitions of the “at-risk population”. For migrant women, the definition is largely shared, namely all women born in one of the 30 countries where the practice is identified and prevalence has been measured using DHS and MICS surveys. For subsequent generations, the definition of women with “origins” in at-risk countries (those born in countries of immigration, but with parentage in an at-risk country) can vary: for example, having one or both parents born in an at-risk country. However, as mentioned above, information on parents’ country of birth is rarely available (Simon, 2012).

56Other limitations or biases of these indirect estimates are linked to the method of extrapolation, i.e. the application of prevalences measured in these countries of origin to the population identified as at-risk in countries of immigration. As we will see in Section III.2, the practice of FGM varies with ethnicity (or geographic origin), level of education, place of residence (urban/ rural), income, and age (in the countries where the practice is decreasing over the generations), among other factors. While it is generally possible, when calculating estimates in countries of immigration, to apply observed prevalence rates from countries of origin by age and level of education (variables that are also available from surveys in countries of immigration), it is rarely possible to do so on the basis of ethnic origin using public statistical data in the North. And yet prevalence can vary widely by ethnic group within a given country of origin: in Senegal, while the national prevalence of FGM is 26%, it is practically non-existent among the Wolof (1%) and Serer (2%), but very widespread among the Poular (55%), Diola (52%), Soninke (65%), and Mandingo (82%) ethnic groups (DHS-MICS Senegal, 2010-2011). The application of a mean national prevalence by country of origin can thus lead to substantial under- or overestimation, depending on the migrants’ ethnic origin. [45]

57Moreover, it is difficult for these indirect estimates to factor in the effects of migration itself. Other studies have demonstrated that the migrant population is not socio-demographically representative of the population that remains in the country of origin (Massey, 1998), and also that migration can have an effect on the actual practice of FGM, notably among girls who migrated in early childhood and who had not undergone FGM at the time. Furthermore, protection against FGM has become an admissible reason for seeking asylum in several European countries. Since 2009, the United Nations High Commission for Refugees has recognized that a woman’s or girl’s fear of being subjected to FGM constitutes one of the five grounds for recognition as a refugee (“membership in a particular social group”). [46] However, according to a recent UNHCR study, the number of women claiming asylum on the basis of a risk of mutilation remains quite low (UNHCR, 2013). [47]

58Other recent studies have refined the methodology for estimating numbers of women with FGM by taking into account the largest possible set of sociodemographic variables in order to better characterize the migrant population (Ortensi et al., 2015). They also apply different hypotheses depending on age at arrival in the country of immigration, assuming, for example, that girls who arrive before the age of 15 years are not subject to the same risks as those who arrive after this age, who were more exposed to these risks in their country of origin (Andro and Lesclingand, 2007; Exterkate, 2013).

59And finally, the method of extrapolation is particularly difficult to apply to the first-generation (and even second-generation) descendants of immigrants. In addition to selection effects, it may be assumed that immersion and socialization in the destination society lead to the progressive abandonment of FGM (Section IV.1). But quantitative data on the abandonment or perpetuation of this practice in the context of migration are generally lacking, aside from Italian and French sociodemographic surveys (Andro and Lesclingand, 2008; Farina and Ortensi, 2014b). In the absence of such data, the application of prevalence rates observed in countries of origin to the daughters of migrants is a highly approximate solution at best.

Prevalence rates by country and region

60The prevalence of FGM varies widely across the 30 countries (almost all in Africa and the Middle East) where it is most common (Figure 1). They can be grouped into four broad categories by prevalence rate: (1) countries where the practice is nearly universal, with prevalence of 80% or higher; (2) countries where the majority of women undergo genital mutilation, but prevalence is more moderate (50-79%); (3) countries where only a portion of the population (25-49%) is concerned by this practice; and (4) countries where FGM is a minority practice, with prevalence below 25%. In Africa, the practice extends through a wide central band running across the continent from west to east, with prevalence particularly high in a large portion of west Africa (Mali, Guinea, Sierra Leone, Burkina Faso, and Mauritania) and the easternmost part of east Africa (Somalia, Djibouti, Eritrea, Egypt, and Sudan). FGM is not practiced in the Maghreb, southern Africa, or a large portion of central Africa (Figure 1).

Figure 1

Prevalence of FGM in Africa

Prevalence of FGM in Africa

61These national prevalence rates are not the best way to approach this very long-standing practice, historically more common in some societies than in others. Even in countries with very high national prevalence, FGM is absent or rare in certain populations. [48] Within-country contrasts by region of residence and ethnicity, two variables that are often correlated, are particularly striking. Large geographical differences exist in all cases, including in countries with very high national prevalence (Figure 2).

Figure 2

Regional variations in the prevalence of FGM in Mali, Senegal, and Tanzania^(*)

Regional variations in the prevalence of FGM in Mali, Senegal, and Tanzania^(*)

(*) National prevalence.
Note: There is a more recent DHS for Mali (conducted in 2012-2013), but whose sample did not cover all the regions of northern Mali (Tombouctou, Kidal and Gao) because of political unrest in 2012 (Mali DHS-V, 2012-2013).

62This is the case, for example, in Mali, where FGM is virtually non-existent in the (sparsely inhabited) north of the country, home notably to Songhai and Tamasheq populations, among whom genital mutilation is rare or not practiced (Mali DHS-IV, 2006). In Senegal, levels are highest in the east and the south, in regions neighbouring Mali and Guinea, where 9 in 10 women are mutilated. In Tanzania, national prevalence (15%) is relatively low, but the practice is common in a few regions in the northeast (Figure 2).

63As mentioned above (Section II.3), as women do not know precisely what form of FGM they were subjected to, the questions in the most recent DHS and MICS surveys attempt to distinguish just two types of mutilation: excision with or without removal of tissue, and infibulation (Figure 3).

Figure 3

Geographical distribution of different forms of FGM in Africa circa 2010^(a),(b)

Geographical distribution of different forms of FGM in Africa circa 2010^(a),(b)

(a) With or without removal of tissue (types I and II).
(b) Type III mutilation.

64Self-reported data on the type of mutilation are available in survey data from 22 countries. [49] In six countries, [50] above 5% of women reported not knowing what type of mutilation they had undergone, with the proportion reaching 19% in Mauritania and 26% in Mali.

65In most countries, the form of mutilation most often reported is cutting with or without removal of tissue: in 15 countries, more than two thirds of women surveyed reported this type of mutilation (Figure 3A). The most invasive type of mutilation, infibulation, is localized in eastern Africa, in Somalia, Djibouti, and Eritrea, where 77%, 62% and 35% of women, respectively, reported having undergone this type of FGM. It is much rarer in other regions, where it generally represents less than 10% of cases (Figure 3B). Women’s responses in these surveys indicate that, overall, the distribution of types of FGM practiced is stable over the generations. [51] In certain countries where the most invasive form of FGM is predominant, as in Djibouti, results suggest that the practice of infibulation on girls is decreasing. Note, however, that this proportion is not definitive, as some girls may undergo it at a later age (Carillon and Petit, 2009). Finally, several studies show that in regions where mutilation is more often carried out by health professionals, as in Nigeria and Kenya, the least invasive forms seem to be favoured (Orubuloye et al., 2001; Njue and Askew, 2004).

Associated factors: education, place of residence, economic status, and religion

66DHS and MICS survey data can be used to examine and highlight possible relationships between FGM status and a number of individual sociodemographic variables, such as level of education, place of residence, economic status, and religion.

67Women’s schooling is associated with a decline in FGM [52] in practically all countries (albeit to varying degrees): the risk for the most educated women is lower than that for women with no formal education. In some countries, the risk of FGM is three to five times greater for the least educated women than for women with higher levels of education, notably in Egypt, Sierra Leone, Mauritania, and Liberia (Figure 4).

Figure 4

Relative risk of FGM by level of education

Relative risk of FGM by level of education

Note: Relative risk of FGM for women with primary education or less versus women with secondary education or more, by national prevalence. Relative risk is the ratio of the probabilities of these two groups. A relative risk of 1 indicates that the risk of FGM is identical in both cases. A value greater than 1 indicates that the least educated women have a greater risk of FGM than the most educated women.

68Level of education cannot be interpreted as a directly causal explanatory factor, as women do not control genital cutting (as we will see, it occurs before schooling), but it can serve as a proxy to measure the influence of family background. Investment in schooling, and notably girls’ schooling, may correlate with greater openness to arguments against this practice and an understanding of its negative consequences. The influence of education is confirmed by the proportion of girls with FGM by mother’s level of education: in countries with high, medium, and low prevalence, the proportion of girls who undergo FGM decreases as their mother’s level of education increases (UNICEF, 2013).

69While education seems to be an important factor in trends over time, survey results also indicate that other factors, such as place of residence and economic status, also play a role. The risk of mutilation is almost always higher in rural than in urban areas [53] (Figure 5).

Figure 5

Relative risks of FGM in Africa by place of residence

Relative risks of FGM in Africa by place of residence

Note: Relative risk of FGM among women residing in rural versus urban areas, by national prevalence.
Relative risk is the ratio of the probabilities for women in these two groups of being (rather than not being) mutilated. A relative risk of 1 indicates that the risk of FGM is identical in both cases. A value greater than 1 indicates that women in rural areas have a greater risk of mutilation than women in urban areas.

70While relative risk is lower overall for this factor than for level of education, the countries where differentials in levels of education are highest [54] are also those where women in rural areas are most disproportionately at risk of mutilation (Figure 5). [55] Note, however, that women’s place of residence at the time of the survey is not a truly accurate indicator of women’s geographical origin. Because levels of rural-urban migration in Africa are high (Temin et al., 2013), [56] a non-negligible proportion of women who were living in an urban area at the time of the surveys were originally from rural areas. In spite of this limitation, which is inherent to this variable, it is also possible that the greater ethnic and social diversity found in cities, and thus the opportunity to have contact with communities that do not practice FGM, may affect individual expectations and practices over time. This hypothesis is supported by data from certain countries on the relationship between women respondents’ place of residence and their daughters’ risk of mutilation: [57] In Kenya, for example, the daughters of women surveyed in rural areas are four times more likely to have undergone FGM than those of women living in urban areas. In Burkina Faso, Mauritania, and Senegal, the relative risk is 2, while elsewhere it is close to 1 (UNICEF, 2013).

71Data on socioeconomic status and FGM (Figure 6) show that the risk is most often higher in very poor households than in rich households, except in the cases of Nigeria, Mali, and the Gambia, where inequality is low and regional (and ethnic) differences are greater. In contrast, relative risk is particularly high in Mauritania, Guinea, and Egypt.

Figure 6

Relative risk of FGM by household economic status

Relative risk of FGM by household economic status

Note: Relative risk of FGM among women living in the poorest households (lowest quintile) versus women living in the richest households (highest quintile). Relative risk is the ratio of the probabilities for the women in these two groups of having undergone FGM. A relative risk of 1 indicates that the risk of FGM is identical in both cases. A value greater than 1 indicates that poorer women have a greater risk of mutilation than wealthier women.

72While wealth is linked to other social characteristics (in particular, place of residence and/or household level of education), it remains clearly associated with decreased risk of FGM in certain countries.

73Regarding religion, data from the DHS and MICS surveys show that FGM occurs among populations that describe themselves as animist as well as in populations of adherents to the three great monotheistic faiths, Islam, Christianity and Judaism (UNICEF, 2013). Because populations that describe themselves as Muslim make up a majority of the population in most countries where FGM occurs, the practice has long been thought of as linked to Islam (Boddy, 1991). In 2007, Al-Azhar University published a religious edict (fatwa) condemning FGM and recalling that the practice is not mentioned in the Koran. This position was echoed by many religious leaders at the national and local levels in a number of countries (UNFPA and UNICEF, 2009). Nevertheless, in certain countries (Eritrea, Guinea, Egypt, Mali, Mauritania, Sierra Leone, and Chad), large proportions of both men and women [58] consider the practice to be a religious obligation (UNICEF, 2013). Several recent studies have shown that the relationship between Islam and the practice of FGM is not systematic, and varies greatly with context. These ethnographic studies show that religious beliefs coexist with other social norms on FGM (Boddy, 1991; Johnson, 2001). A study carried out in Burkina Faso (Hayford and Trinitapoli, 2011), a country with animist (10%), Muslim (60%), and Christian (30%) populations, [59] showed that the impact of religion on this practice (both at individual and collective levels) differs by level of prevalence: in communities where prevalence is high, Muslim religious affiliation is not correlated with the practice of FGM, while the opposite is true in those with low prevalence. The authors explained this in terms of the dominance of group social norms in the first case, independently of religious affiliation, arguing that in the second case, religious beliefs are the dominant influence. Ultimately, the links between religion and FGM are complex and multiform, and ethnographic approaches are needed to arrive at a more precise understanding of them (Boyle, 2005; Johnsdotter, 2007; Johnson, 2007).

The conditions in which FGM is practiced

74FGM has long been described in the anthropological literature in the context of rites of passage, notably for the transition to adulthood (Section I.1). Findings from the DHS and MICS surveys on the conditions in which this practice is carried out [60] reveal that it is now most often disconnected from this ritual dimension. In all countries, virtually all of the women surveyed reported having undergone FGM before the age of 15; in 18 out of 22 countries where data on age at FGM are available, the majority of women were mutilated before age 10 (Figure 7).

Figure 7

Percentage of women aged 15-49 years reporting having undergone FGM before age 15 or age 10, by national prevalence

Percentage of women aged 15-49 years reporting having undergone FGM before age 15 or age 10, by national prevalence

75In Egypt and the Central African Republic, more than half of women (58% and 60%, respectively) underwent FGM between the ages of 10 and 14 years. In only two countries – Sierra Leone and Kenya – were a relatively substantial proportion mutilated at later ages, with 23% and 29%, respectively, undergoing FGM after age 15 (Figure 7). In most countries, age at mutilation also varies by ethnicity. This is the case for example of Kenya, [61] where mean age at FGM among women aged 15-49 ranged from 9 years among the Somali to 16 years among the Kamba and Kalenjin (UNICEF, 2013).

76While FGM continues to be associated with collective initiation rites in certain ethnic groups, in Kenya and Chad for example (Ahmadu, 2001; Droz, 2000; Leonard, 1996), in other regions, such as the Gambia, this is no longer the case: in these regions, FGM is practiced individually rather than in a group, is disconnected from any group celebrations, and may progressively lose its social signification (Hernlund, 2001). Moreover, in half of the countries where mothers’ reports on their daughters’ age at FGM are available, [62] the majority were cut before age 5, suggesting that age at FGM may be decreasing (UNICEF, 2013). These findings must be interpreted with caution, however: this effect could be at least partly due to the fact that certain girls who had not yet undergone FGM at the time of the survey will undergo it at a later age.

77In all of the countries surveyed, mutilation is mainly performed by “traditional” practitioners (women circumcisers or exciseuses, village matrons). There are exceptions, however, as in Egypt [63] and Sudan, where a third of women report having been cut by a health professional: physicians in Egypt, and nurses or midwives in Sudan (UNICEF, 2013). In Egypt, the proportion of girls cut by a health professional has considerably increased over time, from 55% in 1995 to 77% in 2008. This trend towards medicalization [64] of the practice has also occurred in Kenya, where around 40% of procedures were performed by health professionals in the late 2000s, versus a third in the late 1990s (Shell-Duncan et al., 2001; UNICEF, 2013). This recent trend, which in some cases has accompanied a decline in the practice, as in Kenya (Section IV.2), seems to be explained by a counter-productive effect of the first campaigns against mutilation in the 1990s (Section VI.1). These early campaigns focused on the health risks of FGM, notably short-term risks such as haemorrhage and infections, suggesting that they would be decreased if mutilation was performed by health professionals and under more hygienic conditions (Shell-Duncan, 2001).

What is being measured?

81A first approach to measuring trends in the practice is obviously to track how it changes over time. However, as respondents to the DHS and MICS surveys are mainly women aged 15-49, and the procedure in most countries is carried out at early ages (below 15 years), the impact of the campaigns of the last two decades is not immediately visible. Among the 30 countries where surveys have been performed, data covering a period of more than 15 years are available for only five: Côte d’Ivoire, Egypt, Mali, the Central African Republic, and Sudan (Appendix Table A.1). In addition to the limitations inherent to comparing the results of cross-sectional surveys performed at different times on different samples, [70] the main bias is possible under-estimation of the phenomenon, given that the data are drawn entirely from women’s self-reports. In the context of increasing penalization (Section I.2), apparent declines in the practice may simply reflect under-reporting, and not an actual decrease. In Burkina Faso, for example, the prevalence found in the 2010 survey was 4 percentage points higher than in 1998-1999 (72%). The first law making FGM a criminal offense was adopted in 1996, and there were around a hundred convictions between 1997 and 2005. The outlawing of FGM seems to have led to under-reporting of the practice by women respondents in the 1998-1999 survey (Diop et al., 2008).

82For countries where survey data are more recent, it is still possible to carry out a generational analysis by comparing observed prevalence in the youngest and oldest age groups, or by comparing observed prevalence among women respondents (mothers) and their daughters. However, not only is reported prevalence in girls liable to be affected by mothers under-reporting of their daughters’ and their own mutilation (for fear of prosecution); but it is also a poor final measure of prevalence. Depending on the age at which FGM is practiced, some of the daughters of surveyed women (aged 0-14 years) have not yet been cut at the time of the survey, but are still at risk.

83Finally, questions introduced more recently into the DHS and MICS surveys offer information on women’s and men’s attitudes to FGM, uncovering possible ongoing or future changes.

Mixed trends, with contrasts between countries

84As the legal framework on FGM is very recent in most countries (Section I.2), it is difficult to draw any conclusions on the impact of new laws on changes in the practice over time. While legislation seems necessary, it is not sufficient, and programmes to combat FGM also include awareness campaigns (Rahman et al., 2000; Shell-Duncan et al., 2013). These programmes often target local populations at a relatively small scale: in some contexts, at the local level, decreases have been observed following the implementation of programmes based on winning over the community (shifting the norm). The first NGO to implement the theoretical framework developed by Gerry Mackie in its programmes to combat FGM is the association Tostan [71] which has been working in Senegal since 1991, and whose “community empowerment programme” has been deployed since 2007 in a number of other African countries. Actions carried out in Senegalese villages since the late 1990s have yielded positive results, according to numerous field evaluations (UNICEF and Innocenti Research Centre, 2010). However, on a broader scale, trends are uncertain.

85In the 11 countries where multiple surveys have been carried out, the total period covered is more than 10 years. The general trend in all of these countries [72] is a decrease in the practice, but the pace of change differs between countries (Figure 8). In seven cases, decreases were small (less than 5 percentage points). This is notably the case in countries where the practice is nearly universal: in Egypt, over the course of 19 years (1995-2014), prevalence fell from 97% to 92%; in Mali, in 17 years (1995-2012), it fell from 94% to 89%. In contrast, relatively large declines were observed in the Central African Republic and Kenya. In the Central African Republic, the proportion of women aged 15-49 with FGM dropped from 43% in 1994 to 36% in 2000, 26% in 2006, and 24% in 2010. In Kenya, the proportion fell from 38% in 1998 to 27% ten years later (Figure 8).

Figure 8

Changes in the proportion of women with FGM aged 15-49 by national prevalence in 11 African countries

Changes in the proportion of women with FGM aged 15-49 by national prevalence in 11 African countries

86These trends are confirmed by comparing prevalence in different cohorts of women: in all countries, there is a general downward trend over the generations (Figure 9). In the countries where the practice is nearly universal, however, differences remain relatively small, with the exception of Sierra Leone and Egypt, where the prevalence levels observed in the youngest groups (15-19 and 20-24 years) are around 10 percentage points lower than those in older groups. Among countries where prevalence is between 50% and 79%, Burkina Faso and Liberia stand out, showing relatively linear decline with decreasing age, a sign of genuine change in the practice over time. Finally, among countries where FGM is a minority practice, the countries that have shown the most progress over the generations are Kenya, the Central African Republic, and Nigeria (Figure 9).

Figure 9

Percentage of women with FGM in each age group, by national prevalence

Percentage of women with FGM in each age group, by national prevalence

87Another way to capture the social dynamics of the abandonment of this practice is to examine the opinions of women (and men) who express support for its continuation. Questions added to the FGM modules of the DHS and MICS surveys provide a means to assess overall support for FGM amongst all respondents who reported, independently of their own FGM status, being aware of the practice (Figure 10).

Figure 10

Proportion (%) of women aged 15-49 years who reported supporting the continued practice of FGM, by respondent’s FGM status

Proportion (%) of women aged 15-49 years who reported supporting the continued practice of FGM, by respondent’s FGM status

88In all countries, women with FGM are far more likely to favour the continuation of the practice than others: differences by FGM status are often considerable, notably in Mali and the Gambia, where more than 8 in 10 women with FGM favour the continuation of FGM, versus a very low proportion (7% and 3% respectively) of non-FGM women. In two countries, Guinea and Sierra Leone, the opinions of FGM and non-FGM women diverge less markedly (70% versus 49% in Guinea, 69% versus 25% in Sierra Leone), doubtless reflecting greater tolerance for traditional practices among non-FGM women. And finally, in countries where intermediate national prevalence reflects distinct populations with widely varying prevalence, opinions also differ widely, with FGM women much more likely to support the practice than the average (Figure 10). These results reflect the current opinion of adult women, a large majority of whom are no longer in the age group at risk of mutilation. Among women who had undergone FGM, the question was asked well after the actual procedure – that the women themselves had not chosen to undergo.

89Another way to address the question is to look at changes in opinion over time (Figure 11). Overall, the proportion of women who favour continuation of the practice has been decreasing, including in countries where the practice is nearly universal, as in Egypt, where the level of support fell from 82% to 62% in 13 years, and in Sierra Leone, where it fell from 86% to 66% in less than 5 years (Sierra Leone) (Figure 11).

Figure 11

Change in percentage of women aged 15-49 years who reported supporting the continuation of FGM

Change in percentage of women aged 15-49 years who reported supporting the continuation of FGM

90These results partly confirm changes in the prevalence of the practice over the generations (Figure 9). In the countries [73] where results on this question are also available for men on several dates (data not shown), changes of opinion among men over time are fairly similar to those among women. For example, in Guinea and Mali both women’s and men’s opinions have changed little, as support for the practice has remained high. This reflects the persistence of strong social norms in favour of FGM in these societies (UNICEF, 2013).

91Differences between prevalence and percentage of opinion in favour of the continuation of the practice do not completely predict future changes. In contexts where FGM is now condemned, it is more difficult to express support for the practice. To better account for differences between intentions and actual behaviours, the “stages of change” model, originally developed in health psychology to capture changes in behaviour over time, has been applied to FGM (Shell-Duncan and Hernlund, 2006). [74] Starting from the hypothesis that a person’s actual or desired behaviour is influenced by others, the authors identified five categories of readiness for change in the practice of FGM, comparing the opinions of women (with or without FGM) on the continuation or abandonment of the practice and their intentions for their daughters. Women who reported that they supported the practice and that they had had their daughters cut or intended to do so were classified as “willing adherents”; at the opposite extreme, women who supported abandonment of the practice and who said they would not have their daughters cut were considered “willing abandoners” (Appendix table A.3). Applying this model in a qualitative study in three regions in the Gambia and Senegal, Shell-Duncan and Hernlund (2006) showed that this categorization can shed light on trends in FGM, which is not a matter of purely individual decision-making. The most recent UNICEF report presents the distribution of women across these five categories for a number of countries (Figure 12).

92Unsurprisingly, the proportion of women identified as “willing adherents” is highest in countries where prevalence is above 80%, and conversely, in countries with low prevalence, the majority of women are “willing abandoners”. This indicator is consistent with changes in prevalence and opinions over time: in countries where prevalence is high, when support for abandonment increases, prevalence decreases among the youngest women (Figures 9 and 11), with increasing numbers of women classified as “willing abandoners” or “reluctant abandoners” (Figure 12). This is true, for example, of Egypt and Sierra Leone, whereas in Guinea and Mali little change has been observed. [75] Similarly, in Kenya where both FGM and support for its continuation have substantially declined over 10 years (Figures 9 and 11), nearly 6 in 10 women (Figure 12) are now willing abandoners.

Figure 12

Distribution of women aged 15-49 years among the five categories of readiness to change FGM practice, in 9 countries

Distribution of women aged 15-49 years among the five categories of readiness to change FGM practice, in 9 countries

93This categorization could be refined, for example, with data on spouses’ opinions, but it nevertheless provides a relatively clear image of the dynamics of ongoing change. The analysis should be extended to all countries where data on these variables is available from different surveys, with a view to measuring a possible continuum in these stages of change.

Vulvar reconstruction and deinfibulation

124Deinfibulation is a reconstructive surgical procedure performed on the scar tissue caused by the stitching of the labia majora during infibulation. The opening of this scar tissue frees the vagina, the urethral meatus, and the (often intact) clitoral glans, allowing substantial improvements in the patient’s urogenital and sexual health (Nour et al., 2006). [86] Both the surgical act in itself and postoperative management are generally straightforward. These operations can take place at different times in women’s lives. In some cases, the request is made by women who have not yet initiated adult sexual life and who wish to limit or prevent possible complications. In others, deinfibulation is performed during pregnancy or at the time of childbirth. In both cases, multidisciplinary care for women who choose deinfibulation is crucial, as their choice often reflects a desire to distance themselves from or even challenge family practices and community social norms (Abdulcadir et al., 2011).

Clitoral repair

125While the WHO recommends deinfibulation operations for women who have undergone type III mutilations, the current situation with regard to clitoral repair surgery is different. This type of operation remains rare and its clinical evaluation is ongoing (Abdulcadir et al., 2015). It is practiced in a handful of countries, including Senegal, Burkina Faso, and notably Côte d’Ivoire (Ouedraogo et al., 2013; Thabet and Thabet, 2003), as well as in France where it is available in about 20 hospitals and has been covered by the national health insurance system since 2004 (Andro et al., 2010; Antonetti Ndiaye et al., 2015; Foldès et al., 2012; Foldès and Louis-Sylvestre, 2006; Villani, 2009; Villani and Andro, 2010).

126Clitoral surgery following FGM was developed in the late 1990s by Pierre Foldès, a French urologist, as a humanitarian medical intervention for mutilated women with painful complications. The operation consists in freeing the clitoral stump and repositioning it in its anatomical position (Foldès and Louis-Sylvestre, 2006). This operation is carried out in response to a wider range of needs: painful sequelae, but also demands for improved quality of sexual life and/or expectations and demands for physical integrity (“to be a complete woman”). The surgical technique and initial results have been described in various publications, mainly from France (Antonetti Ndiaye et al., 2015; Foldès et al., 2012; Foldès and Louis-Sylvestre, 2006). They show that clitoral surgery significantly improves patients’ quality of life but that it is probably not an appropriate solution for all women.

127The 2,938 patients operated on by Pierre Foldès between 1998 and 2009 represent the largest analysed and published series of “repairs” (Foldès et al., 2012). Patients’ requests are generally part of a multi-factorial process and are sometimes formulated with difficulty. The three main expectations are linked to treatment for pain, improvement of sexual function, and a more complex dimension of “becoming a complete woman”. In the study cohort, the overwhelming majority of patients (821 out of 840) followed up one year after surgery reported that these expectations were satisfied. [87]

128The other studies examine smaller series. Two adopt a wider perspective, analysing the results not only of the surgical intervention, but also of the accompanying multidisciplinary care system (Antonetti Ndiaye et al., 2015; Merckelbagh et al., 2015). One covers 270 women who received care between 2007 and 2012, and the other a separate sample of 169 women treated between 2006 and 2011, in two hospitals in the Paris region. Less than half of the patients ultimately had the surgery. These two studies showed that a large proportion of women requesting surgery have experienced sexual trauma other than genital mutilation (sexual assault and violence). They confirm that “repair” following FGM is not a matter of surgery alone, but that surgery does improve the quality of sexual life.

129In France, the Excision et Handicap (FGM and disability) survey, a general population survey carried out in 2007-2009, also showed that a third of female respondents with FGM reported being interested in surgical reconstruction and that the few who had undergone the surgery (21 out of 685 women) were satisfied with the results (Andro et al., 2009, 2010).