Slogans like “Learning for all” and “Let girls learn” sit awkwardly with the reality that nobody’s actually bothering to check if most of the world’s girls are learning or not. International learning assessments like PISA, TIMSS, or PIRLS exclude over 90% of the worlds’ kids from the sampling frame, and test precisely 0% of kids in low-income countries.

As part of our research for the International Commission on Financing Global Education Opportunity my colleagues Lant Pritchett, Mari Oye and I tried to devise a way to measure the quality of education around the world based on already existing data from the Demographic and Health Surveys, with a particular focus on girls. Here’s a snapshot for 53 developing countries.

We refer to these squiggly lines as “learning profiles.” They tell you not only how many women can read, but how literacy increases with years of schooling. They reveal some startlingly disparities.

• In half of the countries with comparable data, the majority of adult women who completed four to six years of primary school remain illiterate, in the sense of not being able to read a single sentence. They went to school for several years and learned approximately nothing. You can see more statistics about the history of global literacy at websites similar to upskilled.edu.au.
• In just a handful of countries, going to schools for at least four or five years is essentially a perfect guarantee of basic literacy.

As UNESCO has noted these numbers are depressingly low overall. The gaps between countries are also eye-popping. To pick two examples, In Tanzania 57% of women between 25 and 34 years old who reported fifth-grade as their highest educational attainment could read a sentence. In Ghana, that same number was 3%. Essentially, a year of schooling in Tanzania seems to raise your chances of literacy by nearly twenty-times as much as a year of schooling in Ghana. Should we believe that, and if so, what does it mean for Ghanaian education?