Over the last decade, education policy in the developing world has been dominated by the issue of access, writes Madeleine Bunting at the Guardian.
A sharp rise in enrollment rates across many African and Asian countries has been the result of the millennium development goals relating to education, which focused on getting children into school and access to school.
But the problem has been the perception that educational achievement has fallen. Despite international donors pledging to increase funding for education and to support improved access, the money often doesn't seem to percolate down to the schools where it is needed.
Enrollment rates are currently running at about 95%, and to pay for universal primary education, a new tax has been introduced. But Rukmini Banerji, director of an NGO, Pratham, discovered that children didn't seem to benefiting. The only assessment the Indian government carried out was a national sample once every three or four years, but it didn't cover all educational districts.
Six years ago, Pratham took on the enormous task of finding out what Indian children were learning. They found that millions of children were falling behind and being consigned to educational failure. 50% of children after five years in school could not read at the level expected after two years of schooling.
Of India's 200 million children, 195 million are now in school but fewer than 100 million are actually learning to read and do basic math. They may be in school for five hours a day, but it is not doing them or the country much good, writes Bunting.
Over the last six years schools in some states may be well-funded but have failed to translate that into effective education.
What is really intriguing is that here is a massive, national-scale assessment exercise that has traditionally been understood in the west to be the job of the state that is being done at a fraction of the cost by engaging an army of volunteers, writes Bunting.
In every district, Pratham works with a partner institution such as a university or women's group. Volunteers are recruited and trained and then spread out to dozens of villages to visit children to conduct the simple tests.
These volunteers also visit schools to check whether teachers are present (a problem in developing countries, where teacher salaries are so low that they often have to have other jobs), and if there is a supply of textbooks and water.
There are several spin-off benefits from this arrangement between Pratham and partner institutes. It broadens engagement in education through its 25,000 volunteers; it helps engage parents in the educational process, and whether the teacher is really doing their job.
The success of the program has inspired a version, Uwezo, in three east African countries, Uganda, Kenya and Tanzania. Uwezo has prompted bitter controversy since it started two years ago, particularly in Uganda because it exposed poor educational standards. Tanzania has just launched this year's survey with a bold bid to maintain the momentum for improved education. Pakistan has now adopted a version of the scheme and plans are in place to take the model to west Africa as well.