Despite a growing number of studies, induced abortion in Africa south of the Sahara remains a relatively unknown aspect of the subcontinent's demography.1 Although the postcolonial era has seen a gradual liberalization of abortion legislation in many young nations,2 almost all Sub-Saharan countries continue to ban abortion on request,3 and clandestine abortion is rampant.4 For that reason, most information on abortion comes from hospital records of complications resulting from clandestine procedures. And researchers attempting to study abortion through community-based surveys or ethnographic interviews typically discover that women are reluctant to talk about abortion and are likely to underreport their recourse to it.5

Although officially banned, abortion on request is de facto available in some Sub-Saharan countries, and legal charges are rarely filed against abortion providers or women obtaining abortions.6 It is probably easier to collect credible information on abortion and related issues in these settings than in areas where restrictions are rigidly enforced.

This study focuses on Mozambique, a country with "quasi-legal" abortion services. We explore the profile of women who obtain abortions at an urban hospital and the reasons behind their decisions to terminate pregnancy; we also attempt to determine how representative these women are of the urban female population and what segments of this population have limited access to safe abortion services.

Background and Context

Since colonial times, it has been illegal to provide or obtain an abortion in Mozambique. The nation's criminal code, inherited from the Portuguese colonial code, establishes a punishment of 2-8 years' imprisonment for violations of the law.7 However, abortion has been performed on request at the maternity wards of a number of hospitals throughout the country since the early 1980s. The Ministry of Health has justified this practice as a means of dealing with widespread clandestine abortions and their impact on women's health, given limited contraceptive use.8

Such conflict between the written law and everyday practice is common in societies undergoing revolutionary changes, where politically "nonessential" old laws that remain on the books often are regarded as obsolete.* In addition, militant secularism promoted in the socialist years (the first 15 years after Mozambique gained independence in 1975) undermined the influence of religion, especially that of the once powerful Catholic Church, which otherwise might have rallied a strong opposition to abortion.

In the late 1980s, Mozambique began to move away from the socialist system that had shaped it after independence. The structural adjustment policy of recent years—a reaction to the decay produced by the civil war and economic mismanagement—has all but dismantled the state-run health care system (as well as other elements of the social safety net established by the socialist regime).

Paradoxically, these changes have helped to intensify the discrepancy between the de jure and the de facto status of abortion. Comprehensive free-market reforms have led to an increasing commercialization of the formerly free health care. Legal and ethical proscriptions that stood in the way of this process have become easily negotiable, and both medical personnel and hospital administrators have come to regard abortion services as a source of significant revenue.

As a result, Mozambique has had one of the most liberal de facto systems of abortion on request in Sub-Saharan Africa. Nevertheless, clandestine abortions, performed by curettage or with the use of herbal and other nonpatent abortifacients, remain common.9 These procedures constitute a major cause of maternal morbidity and mortality, and further strain the already scarce resources of the health care system.10

This analysis is based on information from women requesting abortions at the largest hospital in Greater Maputo. Although Maputo is in many respects typical of large and rapidly growing Sub-Saharan cities, it has some peculiar features that are important to bear in mind in contextualizing the analysis. Its population represents a combination of the Portuguese, Christian (primarily Catholic) cultural heritage, particularly strong in the central "cement city," and Bantu cultural elements of rural origin, which dominate in the vast suburban and periurban "reed" belt surrounding the cement city.

Data and Methods

Data for this study come from the records of the maternity ward at Maputo Central Hospital, where by the early 1990s, virtually any woman could obtain an abortion in the first trimester of pregnancy after paying the equivalent of US$16 and completing the bureaucratic formalities. Our sample consists of all women requesting abortions at the hospital in May-July 1993. At admission, these women were asked a series of standard questions by a maternity ward nurse, who recorded their answers on a chart. Both the interview process and the transfer of the collected information to case record forms were designed to assure the confidentiality of the women's responses.

The analysis is based on 394 women whose records included all or most of the following information: age, race, birthplace, residence, marital status, number of children and occupation, as well the name of a person to contact in case of emergency. (Records for four women who requested abortions lacked sufficient information for inclusion.) Considering the delicate nature of the issue, the admission interview had to be short, and some information that might have been useful—such as educational level and abortion history—was omitted.

However, the records included the response given by the woman (or, if she was underage, by her custodian) when the admitting nurse asked why she wanted to terminate her pregnancy. The nurse offered no list of specific reasons; rather, all responses were spontaneous. Since the provision of services was not conditioned on certain reasons and virtually any reason would be acceptable, and because the interview was confidential, most women were probably sincere in their responses.

The limitations of the data notwithstanding, they provide useful insights into the dynamics of and motivations for abortion in Maputo. They also permit comparisons with data from other Sub-Saharan African studies.


Profile of Women Seeking Abortion

Women who requested abortion were, on average, 25.6 years old; their age ranged from 14 to 42 years. Some 13% were teenagers, 62% were in their 20s and 25% were 30 or older (Table 1). The largest single-year age-group was 21-year-olds (10%); this group probably includes several younger women who overstated their age to circumvent parental consent requirements. Assuming that the degree of age misreporting is not excessive, the proportion of women in our sample who were adolescents was considerably lower than the proportion of adolescents reported in studies of clandestine abortions.11

Overall, 40% of the women had had no previous births. By contrast, an estimated 50-80% of women obtaining clandestine abortions in other parts of Africa are childless.12 Age was highly correlated with the number of children; the overwhelming majority of adolescents requesting abortions had had no children (not shown). Interestingly, childless women requesting abortions were younger, on average, than women having their first birth at the hospital;13 this is particularly striking because the hospital specializes in complicated and high-risk pregnancies, and therefore may be expected to have an above-average share of young women giving birth. Thus, in Maputo, as in other Sub-Saharan contexts, women rely on abortion as a means of delaying the start of childbearing.

Some 69% of abortion patients were single, 28% were legally married and 3% were widowed, divorced or separated. However, reflecting the prevalence of consensual unions, 14% said they were single, but named a permanent partner as their contact person in case of emergency. In all, therefore, the proportion who were in a union was 42%. These results generally confirm other researchers' findings that in Sub-Saharan Africa, abortion is less common among married women than among their unmarried counterparts.14 However, the proportion of our sample who were in a union exceeds those reported in most hospital-based studies of clandestine abortion in the region.15

Notably, 23% of abortion patients were white or mixed-race, primarily of mixed European and African extraction. By contrast, an estimated 9% of Maputo's population consists of white and mixed-race individuals.16 This difference implies that demand for or access to abortion services may differ by race.

Although the women were not asked about their religious affiliation, a crude attempt was made to estimate religious background according to a woman's name. By this measure, almost one in 10 appear to have been Muslim or of Muslim background. This considerably exceeds the estimated proportion of Muslims in the city population (3%).17 Muslim and non-Muslim women did not differ significantly with respect to age or marital status (not shown). In a typical Sub-Saharan context, Christianity does not discourage women from obtaining abortions18 or practicing contraception,19 but Islam is associated with reduced abortion rates.20

Most of the women were born in a city. By far the largest group (55%) were natives of Greater Maputo, which includes the city proper and several adjacent municipalities. This figure is comparable to the share of Maputo natives in the city's female population.21 Another 22% were born in other large cities, mainly provincial capitals. With the exception of a small proportion (2%) who were foreign-born, the rest were from small towns and rural areas. Notably, the proportion of city-born women increased with the distance between the birthplace and Maputo, reflecting that residents of urban areas are more likely than their rural counterparts to undertake a long-distance move.

Some 62% of abortion patients lived in Urban District 1, the core of the cement city and its most urbanized and westernized part. Among all female Maputo residents of reproductive age, by contrast, only 18% live in District 1.22 The next-largest proportions of abortion patients lived in Districts 2 and 3 (12% and 6%, respectively), where the cement city gradually gives way to the reed city; only 11% lived in the more distant and rural Districts 4-7 (the outskirts of the city of Maputo). Another 8% were from the even more distant but somewhat more urbanized town of Matola (District 8), the administrative capital of Maputo Province, which has a higher share of more educated and better-off residents than other peripheral areas of Greater Maputo. The remaining 1% did not reside in Maputo permanently.

Reflecting the proportion of young, unmarried and nulliparous women, students constituted the largest occupational group (36% of women seeking abortions); 5% of the women were college students (not shown). These figures are comparable with those reported in other Sub-Saharan studies.23 Another 35% of the women worked outside the home—a much higher level of labor-force participation than has been found in other studies conducted in Maputo,24 and a higher level than those reported for women attempting an illegal abortion in Mozambique and elsewhere who do not benefit from the quasi-legal services in Maputo.25

Interestingly, excluding students, the proportion of workers was similar among married and unmarried women (56% and 53%, respectively). Furthermore, 88% of those who were employed were white-collar workers, while among the city's working women at large, blue-collar occupations predominate.26

Since the decision to abort a pregnancy is usually made in utmost secrecy,27 women probably choose as an emergency contact the person they trust most. Not surprisingly, the husband was the contact person for 80% of married women (not shown), or 33% of the sample overall. This also means, however, that one in five married women did not have enough confidence in their partner to designate him as their contact person. Eleven percent of women, most of them young and single, gave their mother as their contact person, and 10% gave their sister. For 7% of women, most of them young, their father was the emergency contact, and for 9% it was their fianc� or boyfriend.

Among more distant kin, female and male relatives were similarly trusted (9% and 6%, respectively). However, outside the family circle, women's trust was clearly vested in female friends and coworkers (3%). Only two women picked men outside their family; both women were students who named members of their school's faculty. Finally, 10% of women, more than two-thirds of whom were unmarried, did not provide any contact name.

Reasons for Requesting Abortion

In all, 92% of the women gave at least one motive for requesting an abortion. While the reasons they mentioned may not include all motivations women may have for seeking abortion, they probably reflect what these women considered the most important factors. Although worded differently, the answers can be grouped into 12 general categories (Table 2).

Material difficulties were clearly the most common reason for seeking an abortion, cited by 160 women (41%). A significantly higher proportion of unmarried than married women gave this reason (45% vs. 34%—not shown), indicating a more acute perception of material difficulties by the former. However, half of women who were motivated by material hardships also gave other reasons, particularly the wish to continue their studies and the feeling that the number and age of their children made having another child difficult.

Some 120 women (30%) said they wanted an abortion so they could continue their education. This is a frequent reason for requesting abortion among young, unmarried women in the subcontinent.28 However, 22 other women in the sample who were students did not explicitly think of their pregnancy as an impediment to continuing their education. For 74 students seeking an abortion, schooling was the only motive mentioned. Another 27 linked this reason with concern about poor material conditions, and several trade and technical school and college students also made reference to their living children.

Sixty-nine women (18%) requested an abortion because of the young age of their last child, and 36 of these gave this as their only reason. Researchers conventionally interpret this argument as indicating a desire to space births.

Fifty-eight women (15%) thought that they already had enough children. These women had an average of 3.7 children, and 37 of them were in a union (not shown). Understandably, the more children a woman had, the more likely she was to offer this argument. Remarkably, however, more than a third of those who mentioned it had only one or two living children. This does not mean, of course, that the women who stated this reason were determined to end childbearing. But the relatively high share of women who at least contemplated that possibility calls into question the dominant view that married women use abortion almost exclusively to space births or to hide the consequences of extramarital affairs.29 At the same time, it is noteworthy that only 14 women said that their only reason for seeking abortion was that they already had too many children; among other women who gave this reason, 29 also mentioned material hardships, and 11 also said that their last child was too young.

For 25 women (6%), poor health or a chronic illness was behind the decision to seek an abortion. For 14 of them, it was the only stated motive; having too many children was also an important factor for five.

Young age per se was not a major consideration: Only 21 women (5%) cited their young age as a reason for getting an abortion, and just two gave it as their sole reason. Sixteen of the women who wanted to keep studying (or their custodians who answered the question on motives for them) cited their young age as a reason for terminating the pregnancy. These women were aged 15-18—young, but not exceptionally so, as far as the local traditions concerning women's entry into childbearing go. However, it appears that in the urban milieu, as marriage and childbearing are often delayed or forsaken in favor of education and other pursuits, the age limits of "inappropriateness" of starting reproductive life are constantly being pushed upward.

Fourteen women (4%) cited marital problems, but such difficulties were the sole motive only in one case. Typically, women who mentioned marital problems also complained of material difficulties and a large number of children; marital instability and the ensuing uncertainty about the future may have exacerbated some women's perception of material hardships and of the economic burden of children, and therefore enhanced their resolve to terminate the pregnancy.

Thirteen women (3%) cited previous pregnancy or delivery complications (e.g., cesarean section), but only two viewed this reason as sufficient in itself. Other motives were given less frequently. Eight women (2%) sought abortion because they considered their pregnancy mistimed (for reasons other than their young age or that of their last child); five of these did not state any other reasons. This category potentially overlaps with several others, such as the age of the last child, problems in marriage and work-related considerations. Only five women (1%) attributed their desire to obtain an abortion to the failure of their contraceptive method—typically the IUD or the pill. Four of these women were married and in their 30s, with several living children; one was a 21-year-old single student.

For five of the women—all unmarried, childless and young—the partner's refusal to assume responsibility for the child influenced the decision to end the pregnancy. (However, one should not conclude that all other single women enjoyed their partners' support. Rather, in the Sub-Saharan urban context, where young fathers are rarely counted on for financial support after a premarital birth, the abortion decision may be based primarily on the pregnant woman's and her relatives' perceptions of their family's own resources.) Finally, only three women seeking abortion at the Maputo Central Hospital mentioned conflict with work or career plans as a motive for abortion, and all of them gave another reason as well.


In Mozambique, as elsewhere in Africa south of the Sahara, abortion is common. Our observations in Maputo and evidence from other parts of the subcontinent30 indicate that people in general do not regard abortion as murder or as an immoral act, especially if it is performed in early pregnancy. The more serious opposition to abortion stems from deeply entrenched pronatalist attitudes and the enduring perception of the high economic and social value of children. However, these traditional orientations and preferences are rapidly changing, and women who seek abortion at the Maputo Central Hospital, along with those who practice contraception, are among the forerunners of these changes.

The women we studied, who resided predominantly in the city's central areas, were a culturally select group. Obviously, urbanized women would find the hospital's abortion services more accessible, whereas women living on the urban periphery are geographically disadvantaged and less likely to be aware of the existence of such services. Periurban women may not only forgo the hospital's quasi-legal abortion option for more accessible illegal abortion; they also may have fewer pregnancies that they regard as unwanted or that they are willing to terminate, and therefore may have less demand for abortion than women who live in the city core. A correlation between urbanism and recourse to abortion has been observed across Sub-Saharan Africa.31

A disproportionately large share of the women were employed, especially in white-collar occupations. Greater financial resources may help to account for this overrepresentation of working women among these abortion patients. However, in Maputo, where most working women have relatively low incomes, cultural and social factors may provide a better explanation. Thus, as in other Sub-Saharan settings,32 women who reside and work mainly in the cement city, particularly those in white-collar occupations, may be more likely than other women in Greater Maputo to terminate an unwanted pregnancy because abortion is more acceptable in their social milieu, they have greater familiarity with and access to Western medicine and medical institutions, and they have a more acute perception of the conflict between employment and childbearing.

White and mixed-race women were overrepresented among the abortion patients in this sample, perhaps because of their higher educational and income levels. Better-educated women may be more willing to avoid an unwanted or unplanned birth; they may also be better informed about the availability of abortion services and may have the money to pay for the procedure. Educational and socioeconomic differences also help to explain the overrepresentation of Muslims in our sample. Muslims in Maputo—many of whom have traditionally engaged in commerce—are a relatively wealthy and educated minority. Thus, higher schooling (and to some extent higher income) may offset the restrictive premarital sexual code and religious opposition to abortion that commonly characterize Muslim communities.

While a large number of the women were young and childless, this group made up a smaller share of abortion patients than they have in other Sub-Saharan studies of complications after illegal abortion. The hospital's fee for abortion may deter some young women from obtaining the procedure. Moreover, some younger women might not have accurate information about the availability of abortion services at the hospital or might be afraid that these services would not guarantee them complete privacy.

Married women also represented a greater share of abortion patients in this sample than they have in most Sub-Saharan studies. Despite observations that abortion among married women deserves greater attention,33 it is still often assumed that married women in Sub-Saharan Africa seldom seek abortions. Since virtually all women in the region marry, and they typically do so at an early age, this view also implies that abortion has a minor effect on their fertility. Although our study does not permit us to assess the impact of abortion on fertility in Maputo, it demonstrates that married and cohabiting women make up a significant proportion of those who resort to abortion.

Women in official or unofficial marital unions may have more resources to pay for abortion services than unmarried women. Married women, especially those with several children, may also be more likely to see abortion as a legitimate option to ensure a proper interval between their children or to limit the number of their offspring. These findings echo evidence from other Sub-Saharan settings where married women increasingly seek abortion under the pressure of economic hardships and other factors.34

Our results also may reflect differences in hospital utilization patterns between younger, unmarried and older, married women.35 Because of their lack of resources, or out of fear of social repercussions, young, single women are more likely to seek the assistance of unqualified providers or to self-administer dubious and dangerous abortifacients. Therefore, they are more likely to end up in the hospital with postabortion complications. As a result, despite the availability of safe abortion on request at the Maputo Central Hospital, such complications remain a major cause of maternal death at the hospital's gynecology ward, as they do in much of Sub-Saharan Africa.36

Women in our sample justified their request for abortion with a variety of reasons. The data cast doubt on the assertion that abortion in Sub-Saharan Africa is used mainly to delay childbearing and almost never to limit the number of children.37 The intention to stop childbearing is not uncommon among Maputo women, as well as among women throughout the subcontinent.38 Women's fertility desires and preferences are not chiseled in stone, and often are determined by their position within the household, their perception of their economic security, social pressure and norms, and other conditions and circumstances of their lives. Changes and reversals of these desires are likely as women's circumstances change. Yet, for a sizable group of the women, the intention to stop childbearing was strong enough to motivate them to request an abortion.

The future of abortion on request in Mozambique depends to a considerable degree on the evolution of its legal status. The paradox of de jure prohibition and de facto availability is not likely to be ignored forever. Once the more pressing issues of the post-civil war recovery are settled, Mozambique's legislature will have to confront and resolve this puzzle.

As the Zambian experience has shown, legalization of abortion on request does not necessarily eliminate the problem of clandestine abortion, because of insufficient information and outreach and limited technical capacity of legal abortion services.39 The liberalization of the abortion law in South Africa has not eliminated clandestine abortion in that country either.40 But the de jure recognition of Mozambican women's right to obtain an abortion on request would prevent a retreat to complete clandestinity and might help extend access to safe abortion services to the currently underserved segments of Mozambique's female population by lessening the social, informational and emotional costs of abortion. It could also considerably alleviate the financial burden that the treatment of postabortion complications imposes on the national health sector.

However, the complete legalization of abortion is unlikely to lower the pecuniary costs of services. Although increased competition among providers might push costs down, these gains might be offset by the continuing reduction of state subsidies in the health sector. To be sure, abortion is not a typical medical procedure: Most Maputo women see it as an extraordinary need, and even households with moderate resources are able to raise the necessary amount of money for this emergency if there is a firm objection to carrying the pregnancy to term.

If costs do not decline, safe institutional abortion services will probably still remain beyond the reach of poor women—a sizable share of the population in Mozambique. These women would continue to resort to the relatively inexpensive and dangerous services of unlicensed providers and to even less expensive (and more dangerous) self-administered abortions.41 Greater access of the poor to safe abortion services will depend on Mozambique's socioeconomic progress and on the growth of individual incomes. Until such gains start to materialize, unequal access to safe abortion will continue.