India has enrolled more children in secondary schools than ever before, but is failing to teach them what they should be learning, while the most vulnerable are falling further behind the rest, as IndiaSpend reported on September 20, 2017.

The Annual Survey of Education Reports (ASER), brought out by the education advocacy Pratham since 2005, have corroborated this fact. Rukmini Banerji, Pratham India’s chief executive officer, says policymakers set unrealistic educational goals for children, and argues for more meaningful outcome indicators. Focusing on the school experience is more important than measurable outcomes, she says, adding, “If I look back, I don’t remember the maths that I learnt but I remember many other things which presumably have gone into shaping who I am.”

In this interview, Banerji, a Rhodes scholar who trained initially as an economist but shifted to education for her PhD at Chicago university, discusses the demographic dividend, pedagogy, teacher training, children’s own aspirations, and the role of the central government in improving the quality of education. Here are some edited excerpts:

Until 2016, ASER surveyed children up to the age of 16. ASER 2017 focuses on adolescents. Year after year, your surveys have shown children in higher grades have a poor grasp of skills taught in lower grades, including adolescents, as we reported on January 16, 2018. These are the children expected to deliver India’s “demographic dividend”. Have we lost the battle already?

The Right to Education Act (RTE) makes education mandatory for children up to age 14, that is, Grade VIII. But our usual survey covered children till age 16 because there are still a lot of students in that age group who are finishing Grade VIII. Those surveys showed how children move into higher grades without having learnt the requisite reading and writing skills.

However, this year we focused on adolescents above that age and out of RTE’s ambit. The motive was not to underline the same failures in learning but to go beyond--that’s why we called it “Beyond Basics”--and ask what children are doing in real-life settings, outside the classroom.

Our usual ASER is very basic and fairly straightforward--you could say ‘academic’ in some sense: the maths is numerical, just like in textbooks. This year’s ASER was an attempt not just to explore what adolescents can do but also to get a sense of what they are thinking, what they want to do, what they feel capable of doing and so on, so that their ability to do certain things is placed in a larger context of how they view what lies ahead. Several results were very interesting.

One is that many children continue after Grade VIII--beyond which RTE does not apply--and there is no big drop in enrolment. It’s clear that the push for more years of schooling continues well beyond the compulsory stage. That is probably coming from both the supply side (there are more schools) and from the demand side (people’s aspirations are higher). Faith in education or, I would call it, in years of schooling, is quite deep.

More kids staying on beyond Grade VIII is an opportunity because if young people are still enrolled in educational institutions, it’s easy to reach them. If, for example, we think that some skills still need to be taught--English, for example, as most students’ level of English is below what is expected as per textbooks--we can reach the kids who are still enrolled somewhere.

It is important to examine what the “demographic dividend” is. In what way are we absorbing a potentially more ready set of young people not just in the workforce but also in other aspects of life? Ten years ago, you had about 12 million children getting up to Standard VIII; today that number is 22 million. So whether they are able to read and do maths or not, they have had the experience of eight years of schooling.

If I look back, I don’t remember the maths that I learnt but I remember many other things which presumably have gone into shaping who I am. Reading and maths are just the tip of the iceberg. It’s important to convert this eight-year experience into a useful experience--socially, emotionally and academically. Definitely, a student who has had eight years of schooling is different from a student who has had four. We would like that exposure to translate into many things.

So I look at the demographic dividend in this way: how prepared are young children? ASER tells us one story but there may be much more to it. Data don’t capture some of it, which is explored in our short film. They are quite articulate about their hopes, their dreams, the challenges they face. I assume that 10 years ago, those that didn’t make it that far [up to Grade VIII] may have not been that articulate. We have to figure out what these kids can do, what they are able to think about and what they would do given certain opportunities. Are we as a society, as an economy, able to create those opportunities?

Another thing ASER 2017 shows is that large numbers of kids are working while still in school and they are working usually in their family enterprise or farm. Now, that’s not what they want to do; they want to do something else. The fact is: if a relevant opportunity were to show itself, they would probably find time while they are going to school to actually do that. To utilise these young people’s potential, are we able to provide that kind of opportunity for them to learn something new?

Have you analysed how household characteristics affect learning outcomes?

Yes, our director, Wilima Wadhwa, has used ASER’s data to explore how students perform in government and private schools depending on household characteristics. She has found that there is no difference between students’ performance in government and private schools if you control for household characteristics.

However, we find it’s important to leave the main ASER report as straightforward as possible, otherwise people impute all kinds of motives. For instance, an article in the Economic & Political Weekly said the covert purpose of ASER is to run down government schools and bring in privatisation.

One study showed that children from poorer families working as shopkeepers in informal markets, who might perform well below their grade-level expected performance, are able to solve basic arithmetic problems, because their job requires that on a daily basis. Other research has also shown children perform better in natural settings than in classrooms.

In such settings, the question is how can we get them to do the normal maths. But even better is how does one bring the school curriculum into that context.

We have been working with the Abdul Latif Jameel Poverty Action Lab (J-PAL), a global impact evaluation centre, in Delhi. What we are asking is that if you have been exposed to certain kinds of maths games in pre-school age or even in the first couple of years in school, does that enable you to do maths better later?

These games do not use numbers. For instance, you have two cards. One has more dots than another and you ask kids to identify which one has more. These were developed at Harvard University and tested out only in laboratory settings on tablets initially. J-PAL took those ideas from that setting and put them into south Shahdara and Pratham bālvādis (children’s centres) in Trilokpuri [both in Delhi] to test outcomes in developing countries.

The interesting thing is that the performance of children in Trilokpuri versus children in Cambridge, Massachusetts, US, is quite similar at the pre-school level. But when you follow them through after a couple of years in school, you find that our children do a lot worse. Duflo’s interpretation is that school maths in India doesn’t build on the intuitive, instinctive maths skills at the pre-school age. That again outlines that the problem is not with our kids but with what we expect them to do once they enter school.

The government is planning to reduce the national board syllabus in two to three years, according to this February 26, 2018, statement. Do you think that will help?

I need to understand first what “reducing syllabus” means. For me, it’s not a quantity thing. What ASER’s or MIT economists Abhijit Banerjee’s or Esther Duflo’s work shows is that whatever we are expecting of children in school tends to target the top of the class.

If I look at ASER data, the reading test is a Grade II test. If you ask how many students can do that in Grade III and look at the data from 2014 across states, you see a huge variation. So, Himachal is the best, where 50% of students are at this level or higher (we don’t know what higher is because we don’t measure it) and another 25% are able to read paragraphs quite frequently. Hence, teaching at grade level makes sense in Himachal. But at the bottom of the pile is Uttar Pradesh (UP), where only about 7% kids can actually read Grade II texts. Now, in UP’s case teaching at grade level makes no sense. As many as 93% of students are nowhere close to learning what you are teaching them.

Given this, we need to think what the role of the first couple of years in school is. If it is to build a solid foundation, you have to define what that foundation consists of and do whatever it takes to get them to that level. You may need to use different methods, different kinds of groupings.

Pratham has been doing pedagogical experiments where we have successfully trained practically everybody by the time they reach Grade II: they are able to not only read at that level but also write stories. It’s not a question of quantity or reduction or expansion. If my goal is that every child should read, then I will work towards that goal. If my goal is that a child should be able to say what she thinks, then we create opportunities for that to happen. So I don’t know what that announcement meant and we need to go deeper into it. Sometimes headlines give you a short form of what was meant. I don’t want to comment on it without knowing exactly what is being planned.

Also, I have seen an announcement asking for our thoughts on this. So this is an opportunity to say that we feel quite strongly that there should be stage-wise, clear goals. There is no big rush. Everybody now has a long life. So if by the end of Grade V you are able to read most things, ask questions on them, formulate your own thoughts and do some basic computations, I think that foundation is quite strong to build on.

It’s also important--from [what we understand after] all these years of doing ASER and what we do in Pratham--that parents understand what their children’s goals are. Only then will parents support and participate in a big way. We wouldn’t have got universal schooling if everybody didn’t understand what going to school meant. And so, I think we are at a stage where everybody needs to understand what learning means. Learning cannot be explained to parents in bullet points about outcomes. You might do that for schools, but for parents you have to explain it in broad terms.

Very often we find the ASER tool is extremely useful [in explaining that]. Parents when told for the first time say: “Acchā, mere bacche ko yeh ānā chāhiye? Mujhe patā hi nahi thā. Mujhe lagā thā ki us kitāb mein jo-jo cheezein hai vo sāri usko āni chāhiye (So this is what my child needs to know? I didn’t know. I assumed he must know all that there is in the textbook).” Then you tell them: “Kitāb mein yahi hai (This is exactly what is in the textbook).” This process of explaining to and taking along parents is important.

Schools, governments, parents are all thinking in terms of grade. What is a grade? It is a way to organise our schools into age groups. But if children in the same grade are distributed across age groups, then you have to do something to help them to get to the same learning level quickly. The method that we use in Pratham--even in a state like UP, where the baseline is very low--is to teach [children who are behind what their grade expects of them] separately about a couple of hours a day for 40-50 days. It’s a very doable task. Teaching algebra to Grade VIII may be difficult but teaching everybody to get up to a basic level is not tough. Just because you have been left behind does not mean you can’t catch up. But you cannot make kids catch up in the way it’s routinely done. You have to do something slightly different.

In ‘teaching at the right level’, we take, let’s say, [children in the] third, fourth, fifth [grades]. And instead of organising them by grades, we organise them by learning levels [that is, whether they are proficient at reading alphabets, words or sentences in ASER tests]. Those who are at the word level are put together and so on. The results--which J-PAL has evaluated--show that substantial and significant change happens in a very short period of time.

But it means that you have to: one, accept that there’s a problem, and two, be willing to move away from your usual age-grade organisation of schools to, at least for a part of the school day, grouping children by learning level. Which the children take to very naturally because that’s how they play.

To me the best ideas are always commonsense ideas. When your older brother teaches you, he doesn’t care about your grade-level curriculum. He starts from wherever you are stuck.

India has spent billions on teachers and their training, yet learning outcomes are worsening, we reported on May 19, 2015. Is India investing in the right kind of teacher training? Is the rising contractualisation of teachers the right way to go?

We’ve worked with teachers in many states. Let me say [in their defence] that our teachers are like us. Are we a highly skilled country? We are not. Are we terrible? No.

What I just described as ‘teaching at the right level’ is to start with a student at whatever level s/he is at, and use a combination of activities to get her/him to the next level. I find that teachers across the board--whether para teachers [those hired on contract] or permanent teachers--are able to do that if they are freed up to do that. Freed up from the pressures of the grade-level curriculum. If teachers are [only] expected to finish a textbook by the end of the year, then that is what they will do. But wherever we are working with the government, such as in Karnataka and Himachal, and the goal is to bring third, fourth, fifth standard kids up to the basic level and the school allocates two hours a day for that, we find teachers do it quite well.

As for training, no matter how much we are trained, all of us learn a lot on the job. But we don’t learn alone. Unfortunately, teachers don’t have recourse to such learning. You need to build a cadre of people who are going to help the teachers. And we have such people in many states, called ‘cluster coordinators’ [each cluster has 10-20 schools in a block] or block resource people or, in general, whoever is above the level of the teacher. Wherever Pratham works with state governments or district administrations, we say that we first want to work with those people. Because those are the people who can really help the teachers.

I think the top thing the country has to do is turn these cluster coordinators--whose job is to check inputs such as buildings, textbooks, teacher recruitment, scholarships--into academic people. That is not a hard thing to do because the education system has one huge asset: children. As soon as you make some progress with the children, you get instant gratification from them. But you can only get that if the goals are doable, not if you have to teach them something so tough that kids fail to comprehend.

Another important thing is to have goals that are achievable and acknowledging that, for whatever reason, a lot of kids are well below grade level. And if you don’t have fundamental skills, you can’t move forward. So, first, spend the time in a school year to help kids come up to this level. Then you can train the teacher.

You speak about building cardres of support staff for teachers. But what can be done about teacher shortages (as we reported here and here)?

Those cadres already exist in the system. And, yes, we have teacher shortages. Obviously you need teachers, but it’s not like if you don’t have enough teachers, nothing can happen. But it’s also because we have gone to one extreme of universalisation that demands there be a school in every small pocket. I remember a large village in Rajasthan that has five government and private schools each. Each government school has 70-80 kids and two-three teachers, and everybody complains about shortage.

I think sensible consolidation of schools to maximise teacher availability can be done where possible. So, yes, teachers need to be hired but there are lots of other ways.

Contract teachers have a bad name and I’m not commenting on their working conditions; they should get paid adequately. But in the studies that we have done on these teachers, their performance is nothing worse than regular teachers. Sometimes--maybe because they are younger--they perform better than regular ones. So, we need to reorganise the resources available today to achieve our goals.

In Bihar once, when we were working with a district to do ‘teaching at the right level’, what they did was that within clusters [10-20 schools in a block], if there was a shortage, they would move a teacher from one school to another.

Yes, for going up to the highest grade level and teaching children everything that’s in the curriculum, we may need more specialised training. But at the level of which ASER is talking, it can be done [even without waiting for more teacher training].

We have a hierarchy of government schools: Navodaya and Kendriya Vidyalayas. Do you think that adds to the problem? Do we need a common-school system?

Common-school system is an idea that has been talked about a lot. If you look at some rural areas, what exists is effectively a common-school system. Especially in parts of eastern India, where there aren’t enough private schools.

Those are like common schools in two ways. One is that all the kids go to the same school. And then, in the village, all children go to the same set of tutors. So, it’s common both formally and informally.

You cannot impose a common-school system when everything else is differentiated. Where kids go to school is a function of how society is organised. You need to have a common health system, a common public transport system. Common-school system is an appealing idea but implementing it is difficult. The hindrances are coming from society.

Also, if the government school is good, chances are everybody wants to send their children to that school. There will always be aspirational reasons why people want to differentiate themselves. As long as choices are available, people will take them. But wherever government schools have begun to perform well, the drift into private schools has slowed down.

Besides, the school education market for much of India is a grey market, because this distinction between government and private is faulty. There is a big unorganised, undocumented tuition space where all these actors are present. If you analyse that, you will see Indian families have a foot in both spaces--formal and informal. And you have this at different affordability levels. Typically, this private-government division is posed as a false binary.

(Vivek is an analyst with IndiaSpend.)

We welcome feedback. Please write to We reserve the right to edit responses for language and grammar.