Implications of Formal Schooling for Girls' Transitions to Adulthood in Developing Countries
Cynthia B. Lloyd and Barbara Mensch
The main interest of demographers in the relationship between education and childbearing has been the negative association between years of schooling and achieved fertility among adult women, an interest that has spawned an extensive empirical literature on fertility differentials. The experience of being in school, however, occurs prior to adulthood, during childhood and adolescence, and involves much more than years of exposure to school. It is therefore surprising that demographers have paid little attention to the process and content of formal schooling and its more immediate implications for the maturation of girls. Specifically, there has been no investigation of the determinants of schooling duration and its attendant implications for the timing of entry into adulthood-a life cycle phase that, for females in most societies is synonymous with marriage or motherhood, whichever comes first. The timing of girls' entry into adulthood takes on particular demographic significance among most populations of the South, where rapid population growth is explained largely by the size of the population under 20 and the space between generations. However, no demographic study of which we are aware has attempted to open the schoolhouse door to see what happens inside and how this might affect the timing of girls' transitions to adulthood.
This chapter draws on insights from two literatures on education in developing countries in order to place the traditional demographic literature on education and fertility into a larger conceptual framework that focuses on successful transitions to adulthood. In reviewing these literatures, we take a particular interest in what they can tell us about how school provides different experiences for girls and for boys, as well as the role of formal schooling in gender role socialization.
We define a successful transition to adulthood not only as a delayed transition, but also one in which a young person is allowed to grow to develop her/his full potential physically, intellectually, and emotionally before taking on adult responsibilities such as bearing and rearing children or providing for their material support (or the support of other family members). Furthermore for girls, a successful transition to adulthood in a world in which adult men continue to hold greater power requires, in our view, not only the development of their human capital, but also the acquisition of a sense of self-esteem and personal mastery that will be necessary if they are to realize that potential in their public and private life.
The first education literature we review relates to school quality (or school effectiveness) and academic achievement in low-income countries. It is primarily an economic literature prepared from a Western perspective. Here, schooling is viewed as a production process. The goal is to identify critical inputs to formal schooling that contribute positively to the development of students' cognitive competencies as reflected in various standardized test scores. Few insights can be found in these studies about the nature or context of adolescents' daily lives in school, nor do they address how schooling might reinforce or change adolescents' learning and socialization, with consequences for their continuation in school or the timing of their entry into reproductive roles. However, much can be learned from this literature about elements of school quality that potentially have implications for school retention and transitions to adulthood.
The second literature relates to the role of formal schooling in the socialization process in traditional societies undergoing development. This literature is less easily characterized than the first and draws from anthropological studies as well as comparative education studies emerging from schools of education. It is particularly rich in African materials from which we draw many of our examples. From these studies one can develop a culturally oriented understanding of formal Western school as an institution situated within the larger society and of its role in the socialization of boys and girls. While this body of research offers a more textured description of schools than that provided by the schooling effectiveness literature, systematic comparative assessment of the schooling experience is rare, and rigorous empirical analysis linking life within the school to outcomes other than test scores has not, to our knowledge, been undertaken. However, as with the economic literature on education, much can be learned from this literature about potential characteristics of school that might be important in encouraging girls to remain in school into their late teens.
This chapter begins with some background on what is known from the demographic literature about the relationship between years of schooling and the subsequent demographic behavior of adolescents. This is followed by a discussion of the more immediate linkages between teenage pregnancy and continuation in school. From this background on the demographic importance of years of formal schooling, we shift our focus to the potential demographic implications of school
quality and school experiences by reviewing in turn the literature on school effectiveness and that on schooling and socialization. Key points are illustrated with recent data on the primary schooling experience of adolescents collected in 3 of Kenya's 50 districts—Kilifi, Nakuru, and Nyeri—expressly to bring some empirical content to the many hypotheses that emerge from these literatures (Ajayi et al., 1997).1 We conclude the discussion by proposing a broadened approach to the study of schooling that includes various dimensions of successful transitions to adulthood as the outcomes of schooling, and that extends the range of inputs traditionally used to capture school quality to include inputs having implications for other aspects of maturation beyond the development of cognitive competencies.
Schooling And The Subsequent Demographic Behavior Of Adolescents
The literature on schooling and the subsequent demographic behavior of adolescents focuses largely on the relationship between years of formal schooling and critical demographic events in girls' transition to adulthood. The duration of schooling or the achievement of critical levels is seen as a factor that can influence the timing of these events, a view that is based primarily on extensive empirical evidence of strong statistical associations. Evidence consistently shows that women with no or less than primary schooling tend to have earlier ages at marriage or first birth and higher subsequent fertility than those who have completed primary schooling (United Nations, 1995; Jejeebhoy, 1995; Ainsworth et al., 1996). The ages at marriage of secondary school graduates are usually higher still, with completed fertility being sharply lower. At the same time, the age at marriage and fertility of women with equal levels of schooling vary greatly among different societies, as does the relationship between years of schooling and age at marriage and fertility. For example, the extensive survey data assembled by Jejeebhoy (1995) show age at marriage among women with no schooling ranging from lows of 14.4 and 14.5, respectively, in Uttar Pradesh and
Maharashtra, India (1992-1993), to a high of 24 years in the Philippines (1978) and Costa Rica (1976). Small percentage differences (less than 10 percent) in age at marriage between women with no schooling and those with completed primary schooling can be observed in countries with both high and low ages at marriage.2 Larger percentage differences in age at marriage (more than 25 percent) between women with no education and those with more than primary schooling are more typical and tend to be associated with lower ages at marriage among the unschooled (below 18); there are, however, numerous exceptions to this finding as well. While these findings clearly point to the importance of culture in explaining cross-country variations in the relationship between years of schooling and timing of marriage, they might also suggest the potential importance of variations in the content and quality of schooling.
In most developing countries, marriage or childbearing and continued schooling are typically incompatible; girls who become pregnant or marry are asked to leave school as a matter of policy. Such policies create the potential for a direct link between age at leaving school and age at entry into marriage and childbearing. In most societies, however, there is a gap between age at leaving school and age at marriage, so that many girls experience at least a few of their premarriage years of adolescence out of school. Therefore, the relationship between years of schooling and the timing of demographic events cannot be understood as a purely mechanistic one.
While it is clear that schools play an active role in the education and socialization of the next generation, little progress has been made to date in identifying what aspects of education are transforming in ways that ultimately matter for the timing of marriage and childbearing. Several recent studies hypothesize that education delays marriage and childbirth by enhancing the autonomy of women (Jejeebhoy, 1995), giving them more influence in marriage decisions and, through employment prior to marriage, greater control over resources (Jejeebhoy, 1995; see also Diamond et al., this volume). Using the same logic, these gains in autonomy should also give girls a greater say with their parents in prolonging their schooling, thus increasing their exposure to those aspects of schooling that affect these demographic processes.
Other researchers are more skeptical of the autonomy-enhancing effects of schooling for girls, given the strong gender role messages conveyed by teachers and textbooks and the fundamentally conservative nature of schooling. "The content of the curriculum and the way it is taught hold up an image of docility and modesty rather than self-assertion and reward girls who conform to it" (Jeffery and Basu, 1996:20). Moreover, education may reduce women's value in the
marriage market as a result of the universal tendency for women to marry "up." For example, in Bangladesh there is a "strong cultural preference for very young brides reflected in the common Bengali adage kuri te buri ('old at twenty').... A girl who is married late requires a great amount of dowry to compensate for her reduced value to that groom's family" (Amin, 1996:197; see also Basu, this volume).
Possible outcomes of schooling other than enhanced autonomy may be more important determinants of demographic behavior (see also Diamond et al., this volume). For example, education leads to greater earning capacity, thereby increasing the opportunity cost of women's time. Other salient effects of schooling relate to the acquisition of specific bodies of knowledge or particular skills. The literature on maternal education and child mortality has struggled with identifying and documenting the mechanisms through which schooling of women improves the health outcomes of their offspring (see, for example, Cleland and van Ginneken, 1988; Kaufman and Cleland, 1994); some of the same pathways that are hypothesized in this literature may operate in the linkage with marriage and fertility. Even if schooling does not always produce literacy, time in a classroom may improve literacy skills, giving women the ability to understand decontextualized language, which in turn alters behavior (LeVine et al., 1994). Such language is used in health and family planning messages broadcast over the radio; it is also the means of discourse in clinics and pharmacies. Still another possibility is that education broadens one's outlook, "making citizens of those whose horizons had been largely confined to the family" (Caldwell, 1980:235). In exposing women to Western middle-class values and norms, education may challenge preconceived notions about the importance of early marriage and childbearing.
Teenage Pregnancy And Continuation In School
In general, as noted earlier, continued attendance in school is possible only for young women who can avoid pregnancy or childbirth while in school—even while still in primary school, where most girls are when they reach adolescence because of delayed entry and/or grade repetition (Mensch and Lloyd, 1998). At the same time, the risks of pregnancy for unmarried adolescent girls may be affected both positively and negatively by schooling. For example, the availability of family-life education may provide girls with information about sexuality and contraception that would help them avoid pregnancy. On the other hand, schools may provide an environment that fosters sexual harassment of girls by both teachers and fellow students, thus increasing their pregnancy risk. In focus groups involving young people in Kenya, conducted as a prelude to our broader examination of schooling and the experience of adolescents, girls revealed how they were sexually harassed in school, while boys discussed how they harassed girls (Mensch and Lloyd, 1998). Recent reports of a mass rape of schoolgirls in
Kenya, by no means the first to be publicized (Mwati and Munyiri, 1996), reveal symptoms of a larger social problem.
In Africa, as elsewhere in the developing world, there is concern about the growing percentage of teenage births among the unmarried. Furthermore, the rapid expansion of education in the region has led to an increasing focus in the literature on the link between premarital childbearing and girls' dropping out of school, particularly out of secondary schools, where adolescents are assumed primarily to be enrolled (Ferguson, 1988; Meekers et al., 1995). However, there is no evidence that pregnancy is the principal reason for African girls' early withdrawal from school. The only study of which we are aware that has attempted to quantify dropouts directly attributable to pregnancy focused primarily on secondary schools in Kenya and did not collect data on dropouts for other reasons; thus comparisons with other causes of dropout were not possible (Ferguson, 1988). Yet this study did show reported dropouts due to pregnancy affecting only about 1 percent of girls in school each year (Ferguson, 1988), so that pregnancy is unlikely to be a leading proximate cause of dropping out despite the concern raised by the author of this report.
Furthermore, the literature on schoolgirl pregnancy in developing countries implicitly assumes that girls who are forced to withdraw from school because of pregnancy would have continued in school if they had not become pregnant. Yet there are many other reasons a girl might withdraw from school during her adolescence. Moreover, for those girls who do become pregnant, a nonsupportive school environment may increase the chances that they will give birth instead of seeking an abortion and continuing in school. Indeed, rather than pregnancy causing girls to drop out, the lack of social and economic opportunities for girls and women and the domestic demands placed on them, coupled with the gender inequities of the education system, may result in unsatisfactory school experiences, poor academic performance, and acquiescence in or endorsement of early motherhood.
In marked contrast to the absence of studies from developing-country settings, there is a considerable literature on the consequences of teenage childbearing for school completion (as well as other socioeconomic outcomes) in the United States. Early U.S. studies concluded that childbearing disrupts the educational careers of adolescents, and, as in current reports from Kenya, identified pregnancy as the primary reason for dropping out among females (Furstenberg, 1976; Trussell, 1976; Mott and Marsiglio, 1985). While these studies attempted to control for confounding factors, they did not consider the possibility that teenage reproductive behavior may be endogenous to school completion, leading to inflated estimates of the effect of early fertility (Ribar, 1994; Hoffman et al., 1993; Geronimus and Korenman, 1992; Ahn, 1994). The question is whether the association between adolescent childbearing and dropping out of school is due at least in part to the birth itself interrupting education, or whether it is due to some underlying set of attributes—measurable or unobserved—that induce both early
birth and early dropping out. Such issues led us to explore the other literatures on education noted earlier for what they can tell us about the association between the quality of the schooling experience—as regards both cognitive competencies and gender equity—and schooling outcomes.
School Quality And Academic Achievement
In the economic literature on the effects of school quality in developing countries, researchers have focused primarily on the direct relationship between school quality (variously defined) and academic achievement as measured by student performance on standardized tests (Fuller and Clarke, 1994; Harbison and Hanushek, 1992; Lockheed et al., 1991; Fuller, 1987; Heyneman and Loxley, 1983; Tan et al., 1997) and only secondarily on unraveling the links among school inputs, enrollment and retention, and ultimate achievement (Card and Krueger, 1996). These production-function studies conclude that school quality matters for immediately measurable school outcomes (Hanushek, 1995), but there is little consensus about how to define school quality or about what dimensions of quality actually make a difference. Indeed, quality is often measured in terms not of specific inputs such as the number of desks, the credentials of the teachers, or the availability of laboratory equipment, but of correlates such as financial resources per student or class size.
Furthermore, this literature gives little attention to teacher attitudes or classroom dynamics as elements of school quality or to those aspects of the school and classroom environment that may result in different experiences for boys and girls. One notable exception of which we are aware from our comprehensive review of the African literature is Appleton (1995), who studied gender differences in performance on exams in Kenya and found both parents' and teachers' attitudes about the natural ability of boys and girls to be significantly correlated with differentials in performance. At the same time, it may be noted that the concern with regard to including attitudinal variables in a production function of school inputs and outputs is that they may be jointly determined, in that gender differences in actual performance may be a factor in shaping teachers' attitudes.
The benefit an adolescent derives from a good school is directly related to the amount of time he or she spends there. In studying the links between school quality and adolescent educational achievement, it is necessary to understand not only the various elements of good schooling, but also which elements in particular encourage enrollment, attendance, and continuation. Studies of the determinants of school enrollment and attainment, however, have focused primarily on the measurement and assessment of family factors because of their clear importance (e.g., Hill and King, 1993; Lloyd and Blanc, 1996) and have paid little attention, beyond the anecdotal, to school characteristics. The relationship between school characteristics and the length of time adolescents spend in school is rarely explored directly (for two exceptions see Glewwe and Jacoby, 1994, and
Hanushek and Lavy, 1994; see also Glewwe, this volume); the evidence cited, even from the U.S. literature, is primarily indirect and suggestive, rather than conclusive (Card and Krueger, 1996). Indeed, it was with that research gap in mind that we designed a study in Kenya to systematically collect data on the school environment, as well as to collect data from a sample of adolescents and parents in the same communities (see Ajayi et al., 1997).3
Adapting the framework developed by Lockheed et al. (1991) to assess school effectiveness, we concentrate on evidence related to three broad elements of the educational process that have some independent support in the literature as being salient for academic achievement: (1) time to learn, such as hours spent in class and time spent using available facilities; (2) material inputs, such as books, desks, libraries, laboratories, and playing fields; and (3) effective teaching, including pedagogical practices and teacher competency. Our interest is in identifying which elements might have different effects on boys and girls and as a consequence might have implications for school retention.
Time to Learn
There is broad consensus in the education literature that the amount of time effectively dedicated to learning in school is directly related to positive educational outcomes (Lockheed et al., 1991). Many factors can detract from learning time, and some can affect boys and girls differently. School-based factors include disruptions due to teacher absence, ceremonial events, and time out of class for domestic chores or punishment. Anthropological evidence from West Africa suggests that girls are sometimes asked to do more domestic chores than boys, with the consequence that their learning time is reduced relative to boys (Biraimah, 1980; Anderson-Levitt et al., 1998). In our Kenyan survey, we collected data on various aspects of learning time. From our primary school data, based on a sample of mainly mixed schools where boys and girls can be compared within the same school settings, we found that a slightly higher percentage of girls than boys performed chores during the day prior to the survey (72 versus 64 percent in Kilifi, 79 versus 71 percent in Nakuru, and 77 versus 64 percent in Nyeri), and on average, girls performed slightly more chores than boys (.95 versus .80 in Kilifi, .96 versus .83 in Nakuru, and 1.02 versus .80 in Nyeri). However, these differences are not large. Rates of punishment for girls varied by district, being highest in Kilifi, where overall exam scores are lowest, and lowest
in Nyeri, where school performance is much higher. On the other hand, rates of punishment for boys did not vary systematically by district. As for sex differences in punishment, girls were slightly more likely to be punished in Kilifi and Nakuru and boys more likely in Nyeri. That girls in any of the three districts were more likely to be punished came as a surprise, given that girls are generally thought to be better behaved than boys. However, in Kenya punishment in school is often the result of poor performance, and girls in Kenya do worse than boys academically. On balance, there does not appear to be strong evidence that boys and girls in Kenya differ in quantitatively important ways with regard to the time available for learning within mixed school environments (Ajayi et al., 1997).
One additional factor that is sometimes seen to increase learning time, particularly for girls, is sex-segregated classes or single-sex schools. Studies in Nigeria and Thailand have shown higher math achievement for girls in single-sex relative to mixed schools, but lower achievement for boys when schools with similar resources are compared. One factor identified is the greater amount of time spent on instruction in all-girl relative to mixed schools. A problem that plagues these studies, however, is the high-income elasticity of demand for single-sex schooling for girls, making it difficult to disentangle the role of the school from the role of family factors in determining these outcomes. In our Kenyan study, we found that the educational background of parents of girls in single-sex secondary schools (such schools are rare at the primary level) was significantly higher than that of parents of boys in single-sex secondary schools, which was higher again than that of parents of boys and girls in mixed secondary schools. Furthermore, we found no evidence of differences in the length of the school day in all-boy versus all-girl schools.
Other dimensions of school quality can affect time to learn indirectly to the extent that they influence daily attendance rates. For example, inadequate toilet facilities could deter girls from attending on days when they are menstruating. We found no evidence that this was the case in Kenya, however, where less than 3 percent of adolescent girls currently enrolled in school who had reached menarche reported staying home when menstruating. This pattern exists despite descriptions of appallingly dirty toilets with little privacy even in high-performing schools (Mensch and Lloyd, 1998). Another dimension of time to learn is the degree of participation in teacher-student classroom interaction. The literature indicates that girls participate in such interaction less frequently than boys and thus have fewer opportunities to be engaged in active learning (e.g., Grisay, 1984). Our own Kenyan data show fewer teacher-student interactions for girls than for boys in a standard-sized class of normal duration (20.2 versus 16.7 in Kilifi, 16.7 versus 14.7 in Nakuru, and 16.7 versus 15.6 in Nyeri).4
Various material inputs affect school quality, including (1) the availability of instructional materials (e.g., textbooks, desks, library, science laboratory), (2) the condition of and access to basic facilities (classrooms, toilets, and playing fields), (3) the availability of certain amenities (water, electricity, transport), and (4) school-specific elements of the curriculum that go beyond the core (such as sports, clubs, and family-life education). Over the past decade, researchers have found that the availability of textbooks and other instructional materials has a consistently positive effect on student achievement in developing countries (Heyneman and Loxley, 1983). For example, girls' schools have been found to have inferior science equipment relative to boys' schools, which may undermine the content of the science curriculum and contribute to girls' poorer performance on examinations (Herz et al., 1991; Kinyanjui, 1993).5
A recent study evaluated the impact on average test scores of increasing the supply of textbooks in poor schools in Kenya (Kremer et al., 1997). The study found that the program stimulated such an increase in enrollment in the experimental schools that the positive effects of additional textbooks on test scores were negated by the negative consequences of the increased enrollment. This study provides direct evidence of the potential responsiveness of enrollment to changes in school inputs and underscores the importance of disentangling the effects of school quality on cognitive competencies from its effects on enrollment and retention.
There has been less attention in the literature to other material inputs to schooling, which, while less directly linked to academic learning, may have important implications for gender differences in the quality of the school experience and may affect girls' continuation in school. We have already mentioned the potential consequences of inadequate toilet facilities for opportunities to learn. Lack of privacy in toilets can also provide opportunities for sexual harassment, a further discouragement to girls (see Anderson-Levitt et al., 1998, for recent evidence from Guinea). In our Kenyan sample, witnesses observed harassment of girls by boys around the toilets in 62 percent of the schools visited in Nyeri, 20 percent in Nakuru, but none in Kilifi, suggesting that the problem arises only in certain settings (Ajayi et al., 1997).
Family-life education, even if provided to both sexes, can have differential benefits for girls given their greater vulnerability and the risks they face in terms
of pregnancy and school continuation. In our Kenyan survey, we found that topics such as sexuality and family planning were rarely taught as part of family-life education, and even where they were taught, appeared to have no effect on the students' knowledge or understanding of pregnancy risks (Ajayi et al., 1997). Only about a third of students in primary schools were able to give correct answers to questions on pregnancy risk.
Other material inputs with a potential gender dimension include sports facilities for girls. Even the casual traveler in developing-country settings is struck by the common sight of groups of boys kicking a ball around during their free time after school or on weekends. Girls are rarely seen in similar circumstances. The question arises as to whether these differences are reinforced within the school environment through the sports curriculum and facilities provided for boys and girls. Evidence from developed countries suggests that sports are important for adolescents in developing a sense of teamwork, goal setting, competition, and the pursuit of excellence in performance, as well in improving health (Brady, 1998). A particular benefit of sports for girls that is hypothesized in the literature is the development of self-esteem. While we found no evidence from our Kenyan data that sports facilities within the school environment were less available for girls than for boys, we expect this finding is atypical of most developing-country settings, where athletic facilities for girls are generally inferior, and their use is often discouraged. Indeed, in Pakistan ''from grade 1 onwards public school girls are taught to cover their heads and refrain from participating in sports" (Shaheed and Mumtaz, 1993:70-71). Our Kenyan findings can probably be explained by the widespread availability in schools of netball, a sport played only by girls that is common in Britain and its former colonies and requires less space than other sports.
The most consistent finding with respect to teacher credentials and effective teaching is the importance of teachers' knowledge of the subject matter and their verbal proficiency (Cleghorn et al., 1989; Fuller and Clarke, 1994). There is no consistent evidence that girls perform better with female than with male teachers (Abraha et al., 1991; Fuller et al., 1994; Appleton, 1995), except possibly in single-sex schools, where female teachers deal exclusively with female students (Lee and Lockheed, 1990). In the case of Kenya, where better schools (at least insofar as performance on the primary school leaving exam is concerned) appear to have a higher ratio of female-to-male teachers than is found in the weaker schools, there appear to be no differences between male and female teachers in either credentials or attitudes. This finding suggests that any positive effects female teachers may have on girls may be explained largely in terms of serving as a role model of a professional woman, rather than providing any special encouragement.
The production function approach to the study of school effectiveness in developing countries has led to many frustrations. School inputs that appear statistically important in one context are unimportant in others. Furthermore, gender differences that appear potentially important in one context do not bear out systematically. In a recent review of the literature, Fuller and Clarke (1994) make a strong appeal for the recognition of culture and context in the linking of inputs to outputs. They also draw attention to another group of education researchers whom they classify as "classroom culturalists":
These observers of schools focus on the normative socialization that occurs within classrooms: the value children come to place on individualistic versus cooperative work, legitimated forms of adult authority and power and acquired attitudes toward achievement and modern forms of status. (p. 120)
In the discussion that follows, we seek guidance from this second body of literature relating to the role of formal schooling in the socialization process in traditional societies undergoing development.
Schooling, Gender, And Socialization
In a modernizing society, formal Western-style schooling provides a new context within which socialization takes place. While there is strong evidence that the formal school curriculum is becoming increasingly homogeneous across societies, reflecting the emergence of a global culture (Meyer et al., 1992), school administrators and teachers can be creative in finding ways to adapt, reinterpret, and transmit traditional gender systems and modes of learning in this new educational context. Bledsoe (1992) uses Sierra Leone to illustrate how Western education has been transformed to reflect the traditional culture. In traditional Mende society, knowledge is power; the elders within the community, who have special knowledge, control access to that knowledge and seek gain in exchange for sharing it. Knowledge does not have value in itself unless it is imparted in the appropriate way. Through an exchange, the recipient is properly "blessed." and the giver is recompensed. "Since blessings legitimate rights to certain domains of knowledge, how children learn is as important as what they actually learn" (Bledsoe, 1992:192). In a village study conducted in rural Madagascar, Bloch (1993) found that the uncritical acceptance of knowledge acquired in school could be explained in terms of its association with the wisdom of elders in the community. The knowledge of elders is seen to be absolute and morally true because it emanates from ancestors and beyond, but at the same time it is seen as irrelevant for practical day-to-day activities. The ''chalk and talk" (Fuller and Snyder, 1991) approach to teaching in most classrooms, where the teacher is vocal, dominant, and often punitive, takes on new meaning when one under-
stands the link between traditional authority structures and the transmission of knowledge in different cultural contexts.
This approach to the transmission of knowledge in traditional societies gives school principals and teachers special power and authority and potentially makes girls, particularly during their adolescence, especially vulnerable. School principals, teachers, and students bring their knowledge and experience with gender systems in the traditional culture into the school and into the classroom. These systems are adapted and reinforced through the daily interactions between teacher and students and among students, as well as through the formal curriculum; the content of required texts; and formal administrative rules and regulations, which operate through the distribution of rewards, punishments, and duties. In states that were formerly colonies, some of the values and educational goals of the early missionaries and colonial administrators may also be reflected in administrative and teaching practices (Yates, 1982).
Certain school policies and administrative practices convey strong messages to boys and girls about their respective roles and social status. Policies requiring the expulsion of girls who are found to be pregnant (but not of the boys who are equally responsible) and preventing their readmission to the same school suggest to both boys and girls that society values the education of boys more than that of girls and gives girls a disproportionate responsibility for pregnancy prevention in a context of unequal power in sexual relations. In Malawi, for example, there are school policies that either differ for girls and boys or are differentially enforced. To illustrate, children are traditionally supposed to kneel when speaking to their parents, a courtesy that is also extended to teachers. However, while girls are compelled to do so, "male students were seldom observed kneeling when they spoke to teachers and even those who did were more likely to crouch than actually kneel" (Hyde, 1994:19). As another example, to "protect" girls at boarding schools (where about half of girls attending secondary school in Malawi are enrolled) there are rules confining them to crowded dorms in the evening and early morning, whereas boys are allowed to use classrooms and other rooms to study (Hyde, 1994). Other practices that embody messages to boys and girls may include uniform policies, the assignment of different types of school duties to boys and girls, the provision of different extracurricular activities, or the differential assignment of honors and awards. In our Kenyan study, we found evidence that boys receive twice as many of the academic prizes as girls on average, but this is in a context in which boys outperform girls in most subjects (Mensch and Lloyd, 1998). Nevertheless, while awards may have been given fairly to the top students, who are typically boys, the practice of giving awards in mixed schools in such a context conveys strong and discouraging messages to girls, who may already suffer from low self-esteem.
Messages conveyed through school policies and administrative and teacher practices are heavily reinforced by centrally designed teaching materials, which rigidly reinforce cultural notions of appropriate gender roles. "In developing countries, textbooks transmit heavily stereotyped images of men and women, with women adopting low profiles and having traits of passivity, dependence on men, low intelligence and [a lack of] leadership" (Stromquist, 1994:2409). An analysis of the gender content of Kenyan textbooks found images of women appearing much less frequently than those of men; when women are depicted, they typically are in a position subordinate to men and are portrayed in fewer types of roles, and their physical appearance is assigned more importance than their achievement (Obura, 1991). A survey from the Middle East found that females were most frequently depicted as mothers and little girls in Arabic textbooks (El-Sanabary, 1993). Latin American textbooks also portray women as housewives and mothers; "when working women are shown, they hold jobs traditionally associated with female nurturing (teaching, nursing, and domestic service)" (Bustillo, 1993:193). Sometimes the stereotyped images of females go beyond the traditional and are quite negative, particularly by comparison with the images of men. In Pakistan, while "attributes such as bravery, rationality, respectability and humaneness are all associated with males .... women, especially in Urdu textbooks, are quite prominently portrayed as cunning, careless, non-cooperative, and repentant" (Shaheed and Mumtaz, 1993:69).
While textbooks reinforce traditional images of women, the curriculum available to girls has been said to limit their options (see, for example, Hyde, 1993, and El-Sanabary, 1993). Unfortunately, however, the literature on the content of education has failed to investigate systematically whether girls and boys actually take different courses or are discouraged by teachers from pursuing specific academic subjects. For example, Benavot and Kamens (1989) conducted an exhaustive review of the curriculum policies of primary schools in developing countries, but did not address the issue of gender differences in curricular content. It has been asserted—although not documented—that in technical schools in Africa, the curriculum for girls is restricted to secretarial and home economics courses, while boys have the option of pursuing carpentry, welding, mechanical drawing, electronics, and the like (Njeuma, 1993). In some countries, a "home science" curriculum is designed exclusively for girls for the express purpose of reinforcing gender stereotypes by preparing girls for their socially prescribed roles (Herrera, 1992).
Family-life education, discussed earlier in the context of school quality, has the potential to convey empowering messages to adolescent girls and boys. Moreover, the content of the material presented is of great importance for the reproductive health and well-being of adolescents. A lack of information about sex and contraception may not only increase the risk of sexually transmitted diseases
for both boys and girls, but also, if it leads to pregnancy, jeopardize girls' continued enrollment in school (Mensch and Lloyd, 1998). While there has been no exhaustive survey of family-life education programs in schools, the United Nations Population Fund reports it has supported such programs in 79 countries. Most of these programs are didactic in nature and "focus on helping young people plan productive lives" (McCauley et al., 1995:23). Although systematic evaluation of the content and efficacy of family-life education in developing-country settings has not been undertaken,6 there is suspicion that sensitive issues, including gender-power dynamics and sexual relations, are ignored. Thus one observer has asserted that "when sex education evolved into family life education, much of the key sexuality content was removed" (Senderowitz, 1995:30). Perhaps expectations about the efficacy of family-life education programs are unrealistically high. A review of such programs in several African countries is quite pessimistic about their ability to affect attitudes and behavior, arguing that the solution to early and unprotected adolescent sex is not to be found "in the education of the young" because "male promiscuity and, by extension, female availability has remained unchallenged in most of our societies" (Hyde, 1997:22).
To the extent that there are empirical studies of teacher attitudes toward girls, they tend to be purely descriptive in nature, lacking analytical links to outcomes such as performance and retention. The picture that emerges from this body of research is not particularly complimentary. School-based studies often report negative expectations and attitudes toward girls on the part of both male and female teachers. In Togo, for example, teachers disparaged their female students, characterizing them as "disruptive" or as "lack[ing] interest in school," whereas they described male students as "responsible," "hardworking," and ''scholarly" (Biraimah, 1980). Positive attributes for girls related mainly to their appearance. In Malawi, teachers depicted girls as less "serious" and capable (Davidson and Kanyuka, 1992), less interested in their schoolwork, and sometimes lazy (Hyde, 1997). In Guinea, teachers described boys as able to learn lessons well, more likely to participate in class and provide "good" responses to teachers' questions, and "ambitious," whereas girls were typically described as well behaved but timid and not as hardworking as boys (Anderson-Levitt et al., 1998).
The reasons teachers give for boys' better academic performance often relate to negative attributes of girls rather than positive attributes of boys (Davidson and Kanyuka, 1992). Of particular interest in this regard are the responses of Kenyan primary school teachers in Appleton's (1995) study to the question: "Girls tend to do less well in the primary leaving exam. Why do you think this is?" The largest number of responses were related to the effects of adolescence on girls, who become disturbed by their bodily changes, lose interest in school, become more interested in boys and in their own appearance, and suffer from mood swings. Other responses were related to sexuality, immorality, and pregnancy. Male and female teachers gave similar responses, although female teachers were slightly more likely to mention adolescence, whereas male teachers were more likely to mention girls' interest in boys. It may be noted that Appleton was careful to use general terms in phrasing questions to teachers and parents that applied to men and women or boys and girls in order to identify attitudes formed outside the immediate context of the study.
In our Kenyan study, the most striking findings about teacher attitudes emerged from questions about teachers' preference for the teaching of boys or girls (Ajayi et al., 1997). Of those primary school teachers who said they preferred teaching one sex over the other, the overwhelming preference was for boys. Overall, an average of 22 percent of primary school teachers said they preferred teaching boys, while only 5 percent said they preferred teaching girls. This discrepancy was most marked in Nyeri (one of the best-performing districts in Kenya as measured by school exam scores), where 33 percent preferred boys and 6 percent girls, and least marked in Kilifi. Overall, an average of 39 percent of Nyeri teachers expressed a preference for one sex over the other, while only 10 percent of Kilifi teachers did so; on the other hand, 63 percent of Nyeri teachers felt students learn better in mixed classes, while only 38 percent of Kilifi teachers held this opinion. Perhaps surprisingly, female teachers in every district expressed a stronger preference for boys than did their male counterparts.
We also investigated teachers' gender role attitudes by asking about the importance of various subjects for girls and boys and the relative ease with which boys and girls learn each subject (Ajayi et al., 1997). In general, primary school teachers believe math is somewhat more important for boys to learn than for girls and that English is about equally important for both. Teachers also appear to believe very strongly that English is easier for girls than for boys: an average of 52 percent of teachers overall hold this opinion, while just 4 percent consider English easier for boys. Opinions about gender differences in math are even stronger: 67 percent of teachers, averaged overall, said math is easier for boys, while no teacher said math is easier for girls. These relationships hold, with some variation, in each of the three districts.
School and Classroom Dynamics
How might the attitudes described above affect the interactions within the classroom between teachers and students, as well as between boys and girls? Bellew and King (1993:311), note that "empirical evidence is lacking from developing countries to support or refute the hypothesis that teachers' interactions with female students discourage girls' attendance or achievement." Yet there is some evidence in the literature that girls do indeed participate less in the classroom because they volunteer less (Biraimah, 1980; Anderson-Levitt et al., 1998). Although there is no systematic evidence that teachers are actively biased against the classroom participation of girls, in that they appear to select fairly from among those who volunteer, their passive response to the sexual differential that emerges results in uneven treatment nonetheless (Biraimah, 1989). This fair but passive behavior of teachers may be particularly powerful in reinforcing gender attitudes and expectations among students in a context where the culture prescribes different forms of knowledge and different styles of learning for boys and girls (Fuller et al., 1994). It may also explain why the introduction of teaching practices that are viewed positively in the West, such as more open-ended questioning and discussion and the use of programmed teaching and instructional materials, may sometimes appear to accentuate rather than alleviate gender differences. For example, Fuller et al. (1994) found that teachers' greater use of open-ended questioning reduced girls' advantage in learning English. Similarly, the introduction of a programmed teaching approach in Liberia resulted in improved performance on average for both boys and girls, but greater gender differentiation in favor of boys than was observed with more conventional approaches (Boothroyd and Chapman, 1987).
Inspired by the work of Sadker and Sadker (1995), who documented the subtleties of gender bias in U.S. classrooms through many hours of observation of the quality and nature of student-teacher interactions, we designed a classroom observation instrument for our Kenyan study that involved counting and assessing student-teacher interactions during a class period:
Our goal was to assess whether teachers pay more attention to boys and provide them with more encouragement or whether they treat girls and boys equitably. In constructing variables to measure "good interaction" we tried to include all events recorded by our observers that had a positive or supportive tone—or at least those that did not have a negative one. Thus, we included instances of students reading aloud; students making presentations in front of the class, teachers instructing or explaining; teachers acknowledging, extending, amplifying or praising correct answers; teachers completing explaining or seeking responses to student questions; and teachers positively acknowledging, expanding upon or encouraging student comments. (Mensch and Lloyd, 1998:176)
Our results indicated small but systematic differences between boys and girls in the number of positively toned interactions with their teachers (in a class stan-
dardized for duration and composition, see footnote 4). Teachers had, on average, 15.4 positively toned interactions with boys and only 13.8 such interactions with girls, with consistent differences found in all three districts.
Gender relations in the larger society are also reflected in the way girls and boys interact with each other in school and what behaviors are tolerated by the school administration. We have much anecdotal evidence from the field assistants in our Kenyan study that the schools worked hard to show their best side to our observation teams despite our assurances that we were not part of a school inspection team. Nonetheless, we were able to observe systematic differences in the extent of bullying and harassment of girls by boys and of boys by girls. In the majority of schools in each district, boys were observed harassing girls. On the other hand, in only 10 percent or less of the schools in each district were girls observed to be harassing boys (Ajayi et al., 1997). In such an environment, girls may come to develop a sense of inferiority and powerlessness vis-à-vis the opposite sex.
Physical Maturation and School Attendance
Tensions between the traditional culture and formal schooling are particularly visible when children stay in school past the point of physical maturation or the onset of puberty. Parents in traditional societies exert control over their daughters' sexuality; initiation ceremonies for both boys and girls that are tied to puberty define the beginning of social adulthood. The knowledge gained during these ceremonies readies young men and women for marriage and a sexual life (Hyde and Kadzamira, 1994). In such settings, the persistence of girls in school causes social confusion; a schoolgirl is viewed as a child from a social point of view, whereas a girl who has been initiated (or circumcised) is considered a social adult. One reason many school systems in developing countries are reluctant to introduce family-life education may be that they see it as conveying privileged adult information to pupils who must remain children if they are to stay in school.
Bledsoe (1990) emphasizes the symbolic importance of the school uniform, which, like the clothes worn by girls during initiation ceremonies, conveys their status as initiates or trainees who should be recognized as belonging to a protected class. In many settings, however, wearing a uniform is not sufficient to protect girls from the sexual advances of fellow students and teachers. A schoolgirl's sexuality can be exploited by powerful teachers who are able to manipulate their position of privilege to seek recompense for their support or "blessings." At the same time, in a period of rapidly rising school fees, a schoolgirl's sexuality can become an asset, if carefully managed, to help finance her school fees. Bledsoe (1990) describes the complex symbolism of school fees in the marriage negotiations of the Mende of Sierra Leone and the simultaneous risks and oppor-
tunities conveyed to girls through their participation in school as sexually mature adolescents.
School As A Factor In Successful Transitions To Adulthood
The socialization literature and the economic and demographic literatures previously reviewed appear to be saying quite different things about the benefits of schooling for girls. In the demographic literature, formal schooling is seen as uniformly positive in that it leads women to delay marriage and childbearing, and ultimately to bear fewer children and invest more in each. In the economic literature, good schools enhance cognitive competencies, with lifetime benefits in terms of higher earnings. On the other hand, in the socialization literature, schools are seen as conservative institutions that reinforce the gender inequalities present in the surrounding society. It would seem that schools are simultaneously reinforcing existing gender bias and inducing more modern forms of behavior that have the potential to help women acquire marketable skills, useful information, and a more modern and global outlook.
To understand the processes that underlie these apparently contradictory outcomes, the production-function approach needs to be broadened to allow for the assessment of a wider range of inputs and outputs. On the input side, not enough attention has been given to teacher attitudes and classroom dynamics. On the output side, not enough attention has been given to school attendance and retention on the one hand and reproductive outcomes on the other. As a first step in this direction, we have estimated the effects of various dimensions of primary school quality, including gender differences in student perceptions and experiences, on dropout rates in Kenya. We have found that, while household factors remain overwhelmingly important in explaining girls' higher dropout rates, gender differences in treatment within the school are also statistically important explanatory factors (Lloyd and Mensch, 1998).
How can formal schooling make a positive contribution to successful transitions to adulthood as defined at the beginning of this chapter? First, it can provide sufficient support and encouragement to parents and students so that students can attend formal schooling for the number of years necessary to acquire critical knowledge and skills.7 Second, it can convey a range of academic and practical knowledge and teach both marketable and other life skills. Third, it can provide a protective environment that removes students from the risks of harassment and sexual exploitation while attending school. Fourth, it can treat boys and girls equally and teach values of fairness and equality. The first of these at-
tributes of school effectiveness becomes a prerequisite for the other three in that each additional year of exposure to school allows those other attributes to operate over a longer period of time. Thus school factors affecting retention become critical to school effectiveness. For this reason, it is surprising that the school effectiveness literature has had so little to say on factors affecting retention.
What might those factors be? Clearly the traditional three factors of time to learn, material inputs, and effective teaching play a role. Students will be encouraged to continue in school when they perform well. But within any school, there will always be students at the top of the class and others at the bottom.
What are the school factors that might encourage continued attendance among middling and lower-ranked students and continued support among their parents? Here teachers' attitudes are likely to be critical. In Kenya, where girls underperform boys in all subjects except Kiswahili and English in the primary leaving exam, girls begin to drop out of school at an earlier age than boys because of their poorer performance. Yet we know that this poorer performance cannot be explained by underlying differences between boys and girls in basic aptitude because in many other settings, girls' exam performance is equal to or better than that of boys at this level of schooling. Therefore, forces must be at work in the schools and/or in the larger society that discourage girls, leading simultaneously to poorer performance and increased dropout rates. Again the most likely culprit within the schools is the gender role attitudes of teachers, identified by Appleton (1995) as a key factor in girls' poorer performance on exams in Kenya.
Because of the strength of the association between education and both fertility and child mortality, demographers have stressed the critical role of years of schooling and hypothesized at length about the pathways of influence.8 In contrast to demographers, economists have begun to realize the importance of opening the schoolhouse door. They increasingly recognize that inadequate attention to school quality can bias estimates of the effects of education on a range of outcomes, including cognitive competencies, school attainment, earnings, fertility, and mortality. To date, however, they have focused their energy on identifying which school inputs are most effective in raising academic achievement. This particular outcome by its very nature produces a list of inputs that is too narrow. Throughout this chapter and our previous work in this area (Mensch and Lloyd, 1998), we have argued that school is a critical institution not just because of its traditional role in expanding knowledge and cognitive skills, but also because of its more intangible role in socializing adolescents to be productive adults. If we are interested in successful transitions to adulthood, we must widen our definition of school quality and pay particular attention to the ways in which school as a social institution delays transitions to adulthood and empowers or undermines the next generation.
Abraha, S., A. Beyene, T. Dubale, B.S. Holloway, and E. King 1991 What factors shape girls, school performance? Evidence from Ethiopia. International Journal Educational Development [ 1 (2): 107-118.
Ahn, N. 1994 Teenage childbearing and high school completion: Accounting for individual heterogeneity. Studies in Family Planning 26(1): 17-21.
Ajayi, A., W. Clark, A. Erulkar, K. Hyde. C.B. Lloyd, B.S. Mensch, C. Ndeti, and B. Ravitch 1997 Schooling and the Experience of Adolescents in Kenya. Nairobi, Kenya: Population Council.
Ainsworth, M., K. Beegle, and A. Nyamete 1996 The impact of women's schooling on fertility and contraceptive use: A study of fourteen Sub-Saharan African communities. World Bank Economic Review 10(l):85-122.
Amin, S. 1996 Female education and fertility in Bangladesh: The influence of marriage and the family. In R. Jeffrey and A.M. Basu, eds., Girls' Schooling, Women's Autonomy and Fertility Change in South Asia. New Delhi, India: Sage Publications.
Anderson-Levitt, K.M., M. Bloch, and A.M. Soumare 1998 Inside classrooms in Guinea: Girls' experiences. In M. Bloch, J. Beoku-Betts, and R. Tabachnick, eds., Women and Education in Sub-Saharan Africa. Boulder, Colo.: Lynne Rienner Publishers.
Appleton, S. 1995 Exam Determinants in Kenyan Primary School: Determinants and Gender Differences. Washington, D.C.: McNamara Fellowships Program, Economic Development Institute of the World Bank.
Bellew, R.T., and E.M. King 1993 Educating women: Lessons from experience. In E.M. King and A.M. Hill, eds., Women's Education in Developing Countries: Barriers, Benefits and Policies. Baltimore, Md.: Johns Hopkins University Press.
Benavot, A., and D. Kamens 1989 The Curricular Content of Primary Education in Developing Countries. Working Paper Series 237. Washington, D.C.: The World Bank.
Biraimah, K.C. 1980 The impact of western schools on girls expectations: A Togolese case. Comparative Education Review 24:S197-S209.
1989 The process and outcomes of gender bias in elementary schools: A Nigerian case. Journal of Negro Education 58(1):50-67.
Bledsoe, C.H. 1990 School fees and the marriage process for Mende girls in Sierra Leone. Pp. 283-309 in P.R. Sanday and R.G. Goodenough, eds., New Directions in the Anthropology of Gender. Philadelphia, Pa.: University of Pennsylvania.
1992 The cultural transformation of western education in Sierra Leone. Africa 62(2):182-202.
Bloch, M. 1993 The uses of schooling and literacy in a Zafimaniry village. Pp. 87-109 in B. Street, ed., Cross Cultural Approaches to Literacy. Cambridge, UK: Cambridge University Press.
Boothroyd, R.A., and D.W. Chapman 1987 Gender differences and achievement in Liberian primary school children. International Journal of Education Development 7(2):99-105.
Brady, M. 1998 Laying the foundation for girls' healthy futures: Can sports play a role? Studies in Family Planning 29(1):79-82.
Bustillo, I. 1993 Latin America and the Caribbean. In E.M. King and A.M. Hill, eds., Women's Education in Developing Countries: Barriers, Benefits and Policies. Baltimore, Md.: The Johns Hopkins University Press.
Caldwell, J. 1980 Mass education as a determinant of the timing of fertility decline. Population and Development Review 6(2):225-251.
Card, D., and A.D. Krueger 1996 School resources and student outcomes: An overview of the literature and new evidence from North and South Carolina. Journal of Economic Perspectives 10(4):31-50.
Cleghorn, A., M. Merritt, and J.O. Abagi 1989 Language policy and science instruction in Kenyan primary schools. Comparative Education Review 33(l):21-39.
Cleland, J.G., and J.K. van Ginneken 1988 Maternal education and child survival in developing countries: The search for pathways of influence. Social Science and Medicine 27(12): 1357-1368.
Davidson, J., and M. Kanyuka 1992 Girls' participation in basic education in southern Malawi. Comparative Education Review 36(4):446-466.
El-Sanabary, N. 1993 Middle East and North Africa. In E.M. King and A.M. Hill, eds., Women's Education in Developing Countries: Barriers, Benefits and Policies. Baltimore, Md.: The Johns Hopkins University Press.
Ferguson, A. 1988 Schoolgirl Pregnancy in Kenya: Report of a Study of Discontinuation Rates and Associated Factors. Nairobi, Kenya: Ministry of Health, Division of Family Health.
Fuller, B. 1987 What factors raise achievement in the third world? Review of Educational Research 57:255-292.
Fuller, D., and P. Clarke 1994 Raising school effects while ignoring culture? Local conditions and the influence of classroom tools, rules, and pedagogy. Review of Educational Research 64(l): 119-157.
Fuller, B., H. Hua, and C.W. Snyder 1994 When girls learn more than boys: The influence of time in school and pedagogy in Botswana. Comparative Education Review 38(3):347-376.
Fuller, B., and C.W. Snyder 1991 Vocal teachers, silent pupils? Life in Botswana classrooms. Comparative Education Review 35(2):274-294.
Furstenberg, F. 1976 The social consequences of teenage parenthood. Family Planning Perspectives 8(4): 148-167.
Geronimus, A.T., and S. Korenman 1992 The socioeconomic consequences of teen childbearing reconsidered. Quarterly Journal of Economics 107(4):1187-1215.
Glewwe, P., and H. Jacoby 1994 Student achievement and schooling choice in low income countries. Journal of Human Resources 29(3):843-864.
Grisay, A. 1984 Analyse des inegalites de rendement liees au sexe de l'eleve dans l'enseignement primaire Ivoirien . International Review of Education 30:25-39.
Hanushek, E.A. 1995 Interpreting recent research on schooling in developing countries. World Bank Research Observer 10(2):227-246.
Hanushek, E.A., and V. Lavy 1994 School Quality, Achievement Bias, and Dropout Behavior in Egypt. Living Standards Measurement Study Working Paper 107. Washington, D.C.: The World Bank.
Harbison, R.W., and E.A. Hanushek 1992 Educational Performance of the Poor: Lessons from Rural Northeast Brazil. New York: Oxford University Press.
Herrera, L. 1992 Scenes of schooling: Inside a girls' school in Cairo. Cairo Paper in Social Science 15(1). American University of Cairo Press.
Herz, B., K. Subbarao, M. Habib, and L. Raney 1991 Letting Girls Learn: Promising Approaches in Primary and Secondary Education. World Bank Discussion Papers 133. Washington, D.C.: The World Bank.
Heyneman, S.P., and W.A. Loxley 1983 The effect of primary-school quality on academic achievement across twenty-nine high-and low-income countries. American Journal of Sociology 88(6):1162-1194.
Hill, A.M., and E.M. King 1993 Women's education in developing countries: An overview. In E.M. King and A.M. Hill, eds., Women's Education in Developing Countries: Barriers, Benefits and Policies. Baltimore, Md.: The Johns Hopkins University Press.
Hoffman, S.D., E.M. Foster, and F.F. Furstenberg, Jr. 1993 Reevaluating the costs of teenage childbearing. Demography 30(1):1-15.
Hyde, K. 1993 Sub-Saharan Africa. In E.M. King and A.M. Hill, eds., Women's Education in Developing Countries: Barriers, Benefits and Policies. Baltimore, Md.: The Johns Hopkins University Press.
1994 Barriers to equality of educational opportunity in mixed sex secondary schools in Malawi. In S. Erskine and M. Wilson, eds., Gender Issues in International Education. Malawi: University of Malawi, Centre for Social Research.
1997 Sexuality education programmes in African schools: A review. Paper presented at National Academy of Sciences' Workshop on Adolescent Sexuality and Reproductive Health in Developing Countries. Washington, D.C., March 24-25, 1997.
Hyde, K., and E. Kadzamira 1994 Girls' Attainment in Basic Literacy and Education Project: Knowledge, Attitudes and Practices Pilot Survey. Final report. Malawi: Centre for Social Research, University of Malawi.
Jeffrey, R., and A.M. Basu 1996 Schooling as contraception? In R. Jeffrey and A.M. Basu, eds., Girls' Schooling, Women's Autonomy and Fertility Change in South Asia. New Delhi, India: Sage Publications.
Jejeebhoy, S. 1995 Women's Education, Autonomy, and Reproductive Behaviour: Experience from Developing Countries. Oxford: Clarendon Press.
Kaufman, G., and J. Cleland 1994 Maternal education and child survival: anthropological responses to demographic evidence. Health Transition Review 4(2): 196-199.
Kinyanjui, K. 1993 Enhancing women's participation in the science-based curriculum: The case of Kenya. In J.K. Conway and S.C. Bourque, eds., The Politics of Women's Education. Ann Arbor, Mich.: The University of Michigan Press.
Kirby, D. 1995 A Review of Educational Programs Designed to Reduce Sexual Risk-taking Behaviors Among School aged Youth in the United States. Santa Cruz, Calif.: ETR Associates.
Kremer, M., S. Moulin, D. Myatt, and R. Namunyu 1997 Textbooks, class size, and test scores: Evidence from a prospective evaluation in Kenya. Institute for Policy Reform Working Paper Series. Washington, D.C.: Institute for Policy Reform.
Lee, V.E., and M.E. Lockheed 1990 The effects of single-sex schooling on achievement and attitudes in Nigeria. Comparative Education Review 34(2):209-231.
LeVine, R.A., E. Dexter, P. Velasco, S. LeVine, A.R. Joshi, K.W. Stuebing, and F.M. Tapia Uribe 1994 Maternal literacy and health care in three countries: A preliminary report. Health Transition Review 4(2):186-192.
Lloyd, C.B., and A. Blanc 1996 Children's schooling in Sub-Saharan Africa: The role of fathers, mothers and others. Population and Development Review 22(2):265-298.
Lloyd, C.B., B.S. Mensch, and W. Clark 1998 The effects of school quality on the educational participation and attainment of Kenyan boys and girls. Presented at the Annual Conference of the Population Association of America. Chicago, Ill. April 2-4.
Lockheed, M.E., A.M. Verspoor, D. Bloch, P. Englebert, B. Fuller, E. King, J. Middleton, V. Paqueo, A. Rodd, R. Romain, and M. Welmond 1991 Improving Primary Education in Developing Countries. Oxford: Oxford University Press.
McCauley, A.P., C. Salter, K. Kiragu, and J. Senderowitz 1995 Meeting the Needs of Young Adults. Population Reports 41. Baltimore, Md.: Johns Hopkins School of Public Health, Population Information Program.
Meekers, D., A. Gage, and L. Zhan 1995 Preparing adolescents for adulthood: Family life education and pregnancy-related school expulsion in Kenya. Population Research and Policy Review 14:91-110.
Mensch, B.S., and C.B. Lloyd 1998 Gender differences in the schooling experiences of adolescents in low-income countries: The case of Kenya. Studies in Family Planning 29(2): 167-184.
Meyer, J.W., D.H. Kamens, and A. Benavot 1992 School Knowledge for the Masses: World Models and National Primary Curricular Categories in the Twentieth Century. Washington, D.C.: The Falmer Press.
Mott, F.L., and W. Marsiglio 1985 Early childbearing and completion of high school. Family Planning Perspectives 17(5):234-238.
Mwati, M.W., and S. Munyiri 1996 Gang rapes schoolgirls. The Nation 40.
Njeuma, D.L. 1993 An overview of women's education in Africa. In J.K. Conway and S.C. Bourque, eds., The Politics of Women's Education. Ann Arbor, Mich.: University of Michigan Press.
Obura, A.P. 1991 Changing Images: Portrayal of Girls and Women in Kenyan Textbooks. Nairobi, Kenya: ACTS Press.
Ribar, D.C. 1994 The Socioeconomic Consequences of Young Women's Childbearing: Reconciling Disparate Evidence. State Park, Pa: Pennsylvania State University, Department of Economics and Population Research Institute.
Sadker, M., and D. Sadker 1995 Failing at Fairness: How Our Schools Cheat Girls. New York: Simon & Schuster.
Senderowitz, J. 1995 Adolescent Health: Reassessing the Passage to Adulthood. World Bank Discussion Papers 272. Washington, D.C.: The World Bank.
Shaheed, F., and K. Mumtaz 1993 Women's education in Pakistan. In J.K. Conway and S.C. Bourque, eds., The Politics of Womens' Education. Ann Arbor, Mich.: University of Michigan Press.
Stromquist, N. 1994 Gender and education. Pp. 2407-2412 in T. Husen and T.N. Postlethwaite, eds., The International Encyclopedia of Education. Second Edition. Tarrytown, N.Y.: Elsevier Science/Pergamon.
Tan, J., J. Lane, and P. Coustere 1997 Putting inputs to work in elementary schools: What can be done in the Philippines? Economic Development and Cultural Change 45(4):857-881.
Trussell, J.T. 1976 The social consequences of teenage parenthood. Family Planning Perspectives 8(4): 184-192. United Nations
1995 Women's Education and Fertility Behavior: Recent Evidence from the Demographic and Health Surveys. New York: United Nations.
Yates, B.A. 1982 Church, state, and education in Belgian Africa: Implications for contemporary third world women. Pp. 127-151 in G.P. Kelly and C.M. Elliott, eds., Women's Education in the Third World: Comparative Perspectives. Albany, N.Y.: State University of New York Press.