The black box of who controls and who decides how a family’s monetary and other resources will be spent has long intrigued social scientists. Gender bias in such decisions can negatively affect spending on health and education. Such findings have influenced the design of policies all over the world in the past 20 years to ensure women are targeted as recipients of social assistance.
Education-related decisions have been a focus of attention. Studies have documented, for instance, variation by birth order: Older sisters often have to look after younger siblings. Continuing on our series of blogs posted since International Women’s Day, this post shows that there is ample evidence a child’s gender informs education spending decisions in certain contexts.
Households decide (a) which children to send to school and (b) how much to spend on those enrolled. In the 1990s, for instance, households in most Indian states spent more money on education for boys aged 5 to 14 than for girls, as more boys were enrolled. Pakistan showed a similar bias concerning primary school-aged children for the same reason, while in secondary education, the bias resulted from both higher enrolment rates for boys and higher spending on boys once in school. By contrast, intra-household bias in Sri Lanka favoured girls across age groups, in line with higher completion rates for girls.
Gender bias in household spending often increases at higher education levels. In Ethiopia, household spending on the education of secondary school-aged children in 1994–2004 favoured boys. In Pakistan, the gap in the probability of boys receiving more household resources for education was 13 percentage points for 5- to 9-year-olds but 24 points for 10- to 14-year-olds. In Paraguay, there was a bias towards boys for younger children in rural areas and for children aged 15 to 19 in all areas.
However, more recent surveys show increasing expenditure bias towards girls in some settings. A comparison of surveys in 12 Latin American countries found that households spent more on girls’ secondary and tertiary education than that of boys. In Malaysia, while there was no intra-household expenditure variation nationally, it existed in some regions, favouring 5- to 14-year-old girls, once children were enrolled.
Households in Ghana, however, spent more on male children, conditional on their enrolment at primary school. In India, while gender bias in enrolment fell between 1995 and 2014, bias in spending that was conditional on enrolment rose significantly. This occurred despite an economic liberalization drive around 2005 that opened employment opportunities in the services sector for women. In Thailand, households were more likely to spend on girls’ education, especially at ages 12 to 19, an effect stronger in rural areas. The bias towards girls was more apparent in the amount spent on education than in the decision to enrol children in school.
Even when education is free and boys and girls enrol in equal numbers, household spending on education that is perceived to be of better quality can be biased. In India, a bias towards boys means they are more likely to be privately educated, something we will explore further from a global perspective in the next Gender Report on non-state actors in education. In the Republic of Korea, a study found that parents spent US$23 more per month on private supplementary education and tuition for academic subjects for first-born boys than for first-born girls. These results were driven by parental expectation that boys (especially the eldest) would have higher education attainment and higher-wage occupations.
As these examples show, such biases are not universal; they vary by context. Gendered labour market expectations interact with cultural norms to shape parental attitudes and household allocations. India is a case in point. Households have favoured boys in enrolment and education expenditure in every Indian state except Meghalaya, the only one with a matrilineal system in which women control household resources. Matrilineal structures could also explain the absence of bias towards boys in some of India’s tribal regions. In Thailand, daughters are favoured in education spending decisions because they are expected to be primary caregivers to elderly parents and more likely to send remittances.
The examples show why it is that the GEM Report puts so much emphasis on gender norms as a central piece of the puzzle to achieving gender equality in education. It also explains the World Inequality Database on Education focus on the influence of cross-cutting disadvantages across education levels. Last but not least, it shows that the gender story in education is not always as simple. Country, context and location determine education chances for girls and boys. This is where we should concentrate if we are to design the right policies to bring about greater equality.