Maternal literacy and health care in three countries: a preliminary report.

Abstract

Is literacy a missing link between female schooling and reduced child mortality in developing countries? An affirmative answer might be based on Preston's (1989) model of mortality reduction, which gives a central place to literacy in the spread of improved health knowledge to mothers of young children. Other analysts of the demographic evidence (Le Vine 1987; Cleland 1990 ) have been sceptical about literacy and other skills in the school curriculum, finding it implausible that the level attained in a few years of attendance at low-quality schools during her childhood could be retained by the average woman in her childbearing years and have an impact on her child health care and reproduction sufficient to affect birth and death rates. But answering the question requires measuring literacy directly, which has not been done in demographic research to date. This report presents the first results of literacy assessment in three community-level studies, indicating that literacy skills acquired in schools of rural Mexico, rural Nepal and urban Zambia are retained to some extent into the childbearing years and may affect the reproductive and health behaviour of women with young children. Figure 1 shows a theoretical model of the pathways that might link maternal schooling with child survival through a causal impact of school attendance on literacy, health skills and health-care practices. National censuses and demographic and health surveys have shown robust associations between Boxes 1 and 5 without clarifying the intervening variables that might connect them. There is also a good deal of evidence indicating associations between maternal schooling and health care (Box 4), suggesting that maternal use of health and contraceptive services, as well as domestic sanitary and dietary practices, may be involved, but this still leaves unknown the processes by which attending school during childhood and adolescence affects maternal practices during the childbearing years. Figure 1 proposes a model by which schools do what they are officially intended to do, transmit the literacy skills (Box 2), enabling women to acquire health information and interact effectively with medical personnel (Box 3) on the basis of which they engage in health care practices (Box 4) which reduce the risks to their children's survival. This has seemed an unlikely causal sequence to analysts aware that in many developing countries schooling is of poor quality, there is little apparent support for women to retain literacy skills in their adult lives, and the associations of formal education with mortality and fertility in many populations do not exhibit the thresholds and asymptotes one would expect if they were attributable to the curriculum at a particular level of schooling (such as learning to read in primary school). In other words, the link between Boxes 1 and 2 has seemed implausible, leading analysts to consider more seriously processes other than literacy acquisition to account for the links of schooling to health care and child survival. But the model has not been tested empirically, and that was the aim of the research reported here. The overall strategy of our research has been to conduct approximate replications of a community-level study, including an interview survey and home observations, in diverse settings throughout the world that have a substantial population of childbearing women who had attended school. Mexico, Nepal and Zambia were chosen for the first studies because of