In the aftermath of the tragic attack on a school in Peshawar, killing over one hundred children who were merely realizing their right to go to school, the results of the latest Annual Status of Education Report survey in Pakistan launched today turn the spotlight to an on-going simmering catastrophe that is affecting millions of children.

Poor quality of schooling is leaving one in four grade 5 students unable to read a sentence, and one in two unable to read a story in Urdu/Sindhi/Pashtu - simple tasks that they should be able to achieve after being in school for just two years.

It is striking that there has been little change in learning levels in Pakistan over the past three years, according to ASER data. Unless the government meets its commitment to invest more in education, there is a danger that the learning crisis will prevail far beyond the 2030 deadline being established for the next set of development goals.

Too many rural girls don’t go to school

It is important to delve deeper into the ASER data to understand who is in school and, amongst these children, who is learning. In terms of access, our recent analysis finds that while the vast majority of rich girls and boys in rural Pakistan are in school, over 40% of poor, rural girls aged 10-12 have never even been to school.

Amongst those in school, there is a wide variation in the type of school to which children from different backgrounds gain access. ASER data show that private schooling is a growing phenomenon in Pakistan: between 2013 and 2014, the proportion of children enrolled in these schools has increased from 20% to 24%.

Amongst those surveyed, only around 10% of schoolchildren from the poorest households in rural areas are in private schools, compared with 40% of those from the richest households. In addition, there is a clear gender divide amongst the poorest in rural Pakistan: the poorest girls are 31% less likely to attend private schools than are the poorest boys.

More focus on government schools for disadvantaged children

ASER data provide further important insights into who is learning. Amongst 10-12 year olds in rural government schools, who should have reached grade 5, one-third cannot read a sentence.

While those in rural private schools are faring better, it is still a cause for concern that more than one-fifth of children in these schools cannot achieve this task, which they would be expected to have reached by grade 2.

Wealth plays a big role, but it shouldn’t

Overall, wealth is of far greater importance than whether a child is in a government or private school. More rich children in government schools are learning than are poor children in private schools.

And within both government and private schools, poorer children are around three times more likely to be unable to read a paragraph than richer children in the same type of school.

There could be a number of explanations for the wealth gap in learning. One possibility is that children from richer households are more likely to be able to pay for private tuition to compensate for the poor quality of schooling. This is indeed the case in rural Pakistan according to ASER data: 18% of rich boys having access to private tuition compared to just 3% of poor girls.

However, private tuition still does not wipe out wealth differences in learning, suggesting that other factors are important. The poorest performing groups of richer children (those at government schools who do not receive private tuition) outperform the best performing groups of poorer children (those at private schools who do receive private tuition).

Overall, analysis of ASER data from Pakistan alerts policymakers to focus their attention on government schools given that this is not only where the majority of the poorest children are studying, but also where learning levels are lowest.

The fact that rich children in government schools are learning indicates that government schools can play an important role.

What lessons can we learn from ASER in Pakistan?

The ASER data provide important lessons for post-2015 framing of, and measuring progress towards, education goals

The abysmal learning situation in Pakistan is reflected around the world: globally, 250 million children are not learning the basics in reading and mathematics, according to UNESCO’s 2014 Education for All Global Monitoring Report. Around half of these children have spent at least four years in school, suggesting that they are receiving an extremely poor quality of education.

Children who are disadvantaged due to circumstances at birth are most likely to be amongst those who are not learning, whether because they are girls, born into poor households, live in rural areas, or who have a disability.

Given evidence of this kind on low levels of learning, post-2015 education goals need to give greater emphasis to tracking progress on learning outcomes, with an emphasis on children from disadvantaged backgrounds. Yet the ability to hold policymakers to account for these failures, and provide them with guidance on potential solutions, has been stifled by the lack of sufficient data on learning outcomes.

Good data is key

In recent years, ASER Pakistan, along with partner surveys in India and countries in sub-Saharan Africa, has provided fresh perspectives into the scale of education challenges. These data have been significant in informing both national planning and the scale of the global learning crisis and, by extension, the framing of post-2015 goals.

ASER tools can provide a sound basis for the design of measuring learning outcomes. Part of the beauty of ASER is that the tools are simple to use, and that it is straightforward to communicate their findings – prerequisites for measuring progress towards development goals.

It is, of course, also important that the data is reliable and valid. Recent work comparing ASER India data with the country’s national achievement survey data (PDF) shows that they do indeed meet the necessary standards.

Furthermore, ASER data has the additional benefit of differentiating between weaker learners, those who cannot read a sentence or a paragraph, for example. By contrast, national assessments often cater more towards those who are able to perform better on the tests, and so are not as informative for learning lessons on who is being left behind.

When world leaders meet at the United Nations in New York in September to agree new global development goals to be achieved by 2030, it is vital that they take heed of the lessons from ASER. Based on its evidence that millions of disadvantaged children are not learning, they need to make - and keep - promises that all children, regardless of circumstance, are both in school and are at least learning the basics by 2030.

Pauline Rose is an advisor at ASER Pakistan and Professor of International Education at the University of Cambridge in the United Kingdom.