South Asia is home to a growing youth population and widely considered to benefit from the “demographic dividend” in the coming decades. The United Nations Population Fund’s State of World Population 2014 report The Power of 1.8 Billion: Adolescents, Youth and the Transformation of the Future therefore calls for increased investment in youths and adolescents.1
Schooling versus learning
Most governments in South Asia have already invested heavily in education to achieve the Millennium Development Goal (MDG) target of universal primary education for all children by 2015 and have succeeded in closing school enrollment rate gaps vis-a-vis other developing regions.
However, two influential reports published in 2014 have singled out South Asia as a region undergoing an education crisis. According to the United Nations Educational, Scientific and Cultural Organization’s Education for All Global Monitoring Report 2014, schools in South Asia are failing in terms of learning outcomes. In terms of global ranking, students from India and Pakistan perform only better than those in sub-Saharan Africa.
A World Bank report published in late 2014 on student learning in South Asia has gone one step further by looking into emerging evidence from new micro datasets.2 The report, by far the most comprehensive to analyze the performance of South Asian educational systems in terms of student learning, confirms the low level of student learning across the region: “Up to one-third of those completing primary school lack basic numeracy and literacy skills.”3
Failure to learn in school
Recent research conducted in Bangladesh, however, indicates that the learning crisis is much more severe than previously thought—learning achieved against years spent in school is very low across all grades.
The World Bank study Assessing Basic Learning Skills by Vincent Greaney and Shahidur Khandker was the first to document the case of schooling without learning in South Asia. Conducted in 1992 in Bangladesh, the study assessed basic literacy and numeracy among 5,200 individuals and found that the majority of those who had completed primary schooling failed to attain minimum standards in four areas: reading, writing, written mathematics, and oral mathematics. The level of competency in basic numeracy skills was particularly low, even when focusing on graduates of primary school enrolled in secondary school.
In collaboration with Nazmul Chaudhury of the World Bank, I replicated Vincent Greaney’s seminal work 16 years later in 2008 as part of another World Bank research project on the state of student learning in rural Bangladesh.4 We focused on children 10–17 years old and tested them using a subset of Greaney’s written and oral numeracy tests. They were examined irrespective of whether they were in school, dropped out, or never chose to enroll. The findings are striking. Children on average increase their written math scores (i.e., percentage of correctly answered questions) by 6.4 percentage points per year of schooling (grades 1–9). Once we control for child attributes, parental characteristics, and Raven’s scores (a measure of cognitive ability), the figure is even lower—4 percentage points per year.
Five years of primary education only raises the percentage of correct answers by 31.6 percentage points or 6.3 percentage points per year when averaged over the 5 years; the figure is lower, 3.6 percentage points, in case of oral math.
Given that the tests are designed to assess rudimentary numeracy skills taught at the primary level, these figures highlight a very low level of achievement in rural Bangladesh. The learning profile is unusually flat and reflects a deep crisis in Bangladesh’s education sector.5
In collaboration with BRAC, the largest Bangladeshi nongovernment organization (NGO), I revisited the state of learning among rural adolescents in 2012. Among sample adolescents who have completed 5 years of schooling (i.e., primary school graduates), 30% do not have basic numeracy skills, while 33% cannot read two simple sentences in Bangla and 66% cannot do so in English.6 A large proportion of adolescents go on to post-primary schooling to attain these basic numeracy and literacy skills, even though they are supposed to have achieved this by the end of primary school. This once again points to a shallow learning profile. Similar evidence is also coming out from assessment exercises carried out by ASER in India and Pakistan.7
Preparing children for labor market success
With 30% of the total 158.5 million people in the age group 10–24 in Bangladesh,8 education reforms are urgent if the country is to capitalize on the window of opportunity presented by its bulging youth population. Equally, South Asia faces a very difficult challenge for future economic growth and productivity if the region fails to equip youths with adequate marketable skills.
The emerging evidence on flat learning profiles challenges the conventional wisdom that spending time in school is valuable preparation for labor market success. Much of the progress in improving access to education in South Asia may not have created tangible gains in cognitive skills. Therefore, future strategies aiming to prepare millions of youths for a life of work must start by fixing the school system. Otherwise, South Asia risks overemphasizing single-focused policy prescriptions (such as enhanced access or improving physical infrastructure at school) and may reproduce the past pattern of schooling without learning.
1 See UNFPA.org 
2 H. Dundar, T. Beteille, M. Riboud, and A. Deolalikarl. 2014. Student Learning in South Asia: Challenges, Opportunities, and Policy Priorities . Directions in Development: Human Development. Washington, DC: World Bank Group.
3 See worldbank.org 
4 M. Niaz Asadullah and N. Chaudhury. 2013. Primary Schooling, Student Learning, and School Quality in Rural Bangladesh . CGD Working Paper 349. Washington, DC: Center for Global Development.
5 The schooling–learning profile, or simply learning profile, refers to the distribution of learning outcomes across years spent in school or grades, e.g. children who just complete primary school (grade 5) do marginally better in terms of math competencies than those with less than primary or no schooling at all. See CGADEV.org 
6 See BRAC.net 
7 See Aser Centre .
8 See The Daily Star