The Effects of Primary School Quality on the Educational Participation and Attainment of Kenyan Girls and Boys

In Kenya, adolescent girls fare poorly relative to boys in an educational system characterized by enormous growth, deteriorating quality, and rising costs. Girls are more likely than boys to drop out of school prematurely and are less likely to do well on the primary school leaving exams that come at the end of grade eight. Using data from nearly 600 adolescents aged 12–19 in combination with data collected from 36 primary schools in which those adolescents are enrolled, this paper investigates the effect of school quality on the likelihood of dropping out from primary school in three districts of Kenya. In particular, various elements of schools that are either different for girls and boys or that have a potentially different effect for boys than for girls are explored. The results document both the power of existing gender systems at the level of the family and the potential power of gender systems within the school environment. Although the determination as to whether girls will remain in school rests largely in the hands of the family, school factors also appear to matter. Schools in which boys are favored in class, in which boys are provided with a more supportive environment in terms of advice, in which teachers take the importance of hard subjects such as math less seriously for girls, in which boys are left free to harass girls, and in which girls’ experience of unequal treatment is not recognized by boys, discourage girls’ retention. This material may not be reproduced in any form without written permission from the authors. Few studies of education in developing-country settings have examined the impact of school quality on enrolment and retention. Not only is it critical to determine what school factors affect enrollment and retention generally, it is also important to determine which matter for girls and which for boys. Sex differences in school enrollment often emerge during the teenage years, suggesting the possibility that school quality may have a gender dimension. The same school environment may be experienced differently by boys and girls because of differences in curricular opportunities within the school; differences in treatment by individual teachers; and differences in rules, regulations, and administrative practices. Moreover, even if the school environment is the same for both boys and girls, gender differences may occur in the impact of particular aspects of that environment on school retention. Our interest in studying school quality in Kenya is motivated by these concerns. It is well accepted in the literature that school quality matters for the development of cognitive competencies and ultimate earnings. However, despite the many empirical studies that have attempted to identify which school inputs lead to greater school effectiveness, little consensus exists about which are critical. Much of the controversy in the literature relates to the question of the proper estimation of effects. As statistical approaches become more sophisticated, however, it becomes increasingly clear that the importance of specific types of schooling inputs may vary depending on the context. Indeed, at an early phase of educational development, school effectiveness may depend more heavily on factors that encourage attendance and retention because, without initial enrollment and steady attendance to a basic level, children do not have sufficient exposure to achieve even the minimal cognitive goals set out in most national education plans for basic schooling. At a later phase of educational development, when almost all children complete basic schooling, gains in standardized test scores per grade attended become more important as measures of school effectiveness. Thus, in developing countries where enrollment is not yet universal or where repetition