Unsurprisingly, in an age where social spending by governments is under tremendous scrutiny and aid flows are under pressure, testing and assessments have found currency in many countries across the developing world. It has also helped civil society put pressure on education systems (whether public or private) to focus on learning outcomes, moving beyond a highly limiting obsession with inputs— classrooms, teachers, textbooks, uniforms, etc. To be clear, the argument is not that one can ignore the need for high quality inputs. Indeed, that would be foolish. However, there is now substantial evidence that on its own, investing in inputs will not yield improved schooling outcomes.
Therefore, one must focus on improving mechanisms for accountability. In order to fix accountability, one has to first set performance standards. In our schools, the performance standard that should matter the most is what children are learning in school. These learning outcomes need not be restricted to reading and numeracy, but at the end of the day, our education system must be able to set grade-appropriate yardsticks for what children will learn in school. And they must be held to account against the delivery of these learning outcomes.
In his column, Anurag Behar uses the example of measuring the temperature of individuals in a population and how creating incentives for improvement will not yield population-wide improvement in health outcomes. A child who is not learning after spending years in a school is not merely suffering from high temperature. She is slowly sliding towards a life of low productivity that will cripple him for life. If policymakers remain unaware of this steady slide that a significant part of our population is trapped in, they will fail to take the appropriate steps required to tackle this crisis.
Assessments speak to the quality of education delivered in our schools. Of course, one can argue about getting these assessments right—in the way questions are designed and how they are administered. Much like one can argue that we need to develop improved mechanisms for measuring health outcomes, but not that critical indicators such as the Infant Mortality Rate need not be measured or used to guide policymakers.
The other argument against assessments is that they are reductionist. The complexity of running an effective school system is not lost on anyone. In order to achieve a setting where teachers deliver lessons to their students in a classroom, one must factor in issues around access to school, child-friendly school structures, teacher recruitment, training, deployment and motivation, funds for equipment, etc. A complex system requires a flexible implementation model—exactly the opposite of our opaque bureaucracy-driven rigid system. By using assessments and shifting the focus to learning outcomes, local administrative units can be allowed to develop their own strategies to improve schooling in their areas. This is a step towards a system where schools would be allowed to decide how to utilise state funds contingent on delivering on learning outcomes—a flexible results-based school financing model with multiple watchdogs.
There is irrefutable evidence that the education sector faces a serious crisis of accountability with organised civil service-employed teachers. This is not an issue that can be overcome by additional teacher training. Research has shown that an isolated intervention at training or re-training teachers do not yield improvements in performance. Also, teachers in public schools in India are already in the upper quintiles of the income distribution of the country. There are also heart-warming stories of teachers who against all odds deliver to the best of their abilities in the interest of their students. This, however, is not the case if we talk of the average teacher in a public school. This crisis needs to be confronted on multiple fronts: by enforcing vertical accountability in the bureaucratic system that runs the education system; as well as providing communities and organised civil society the tools to demand accountability from their schools. Assessments that generate a measure of student-level learning outcomes are a way of doing this.
According to some, the fact that ASER has seen a steady decline in learning outcomes over its own lifetime of promoting assessment-generated evidence to the education sector is an indication that ASER has failed. This is a serious charge that the likes of ASER and Uwezo need to tackle. However, it must be noted that in the absence of a counterfactual, this isn’t a conclusion one can directly draw. The ASER and Uwezo surveys have also taken place in a context where enrolemnt shot up as countries adopted policies to ensure free and compulsory primary education.
But even if we were to accept this charge, assessments and testing need to be continued and indeed, strengthened. On the policy front, the follow-up ought to be in the form of steps to improve its uptake—about what we can do to bring together civil society and local implementers to take concrete steps towards improving learning in our schools.
Education systems are complex. In India, at the national level or even at the state-level, the scale of the challenge can be daunting. It is, however, important not to confuse the ends with the means. Assessments are not the primary lever for systemic change, but they certainly are a key component of the effort required to push the real levers to lift the load of our low-performing education system.
This is a response to a livemint column by Azim Premji Foundation’s Anurag Behar in which he argues that assessments are not a primary systemic lever for improvement in education and that assessments should remain tools that provide feedback to teachers in the classroom. Interestingly, Behar does not make any reference to India’s Annual Status of Education Report (ASER). ASER has been around for a decade, riding on a simple and powerful idea: parents, communities, the wider civil society and policymakers just did not have sufficient information on the levels of learning our public schools deliver.