Rural Education #1: Problems

Prof. Karthik Muralidharan wrote this very long (~900 pages) but excellent book, Accelerating India’s Development. One chapter is about education. This is the first of a two-part blog based on just that one chapter.

 

At independence, India paid more attention to tertiary education over primary education. This was because a technical workforce was considered vital for India’s modernization. In more recent times, while enrolment for primary education has exploded, the outcome (what they learnt) is still very poor.

 

The government’s approach is to focus on inputs (money, infrastructure, teacher count, mid-day meals, books and computers), not on harder to quantify but more important outputs. Take improvements in school infrastructure. While it may make the learning experience more pleasant, it doesn’t translate to better learning if the “teaching and learning processes are ineffective”. Computers are often kept under lock and key and never allowed to be used, for fear of theft. Toilets may be unusable due to lack of water and cleaning staff.

 

Good teachers are critical, but we have no clear ways to identify who they are. Age, experience, salary, qualifications – none of them are good indicators of the quality of the teacher. Besides, there are no rewards (or penalties) for teacher outcomes anyway.

 

The focus on teacher to student ratio, while well intended, doesn’t work. These ratios are determined at a school-wide level, not per-class level. As a result, in most villages, the same teachers end up teaching multiple classes, though each of them requires a different approach. Younger students need more personalized attention, but the one size fits all metric misses that point.

 

A key problem is that how much a student has learnt is low, but he is moved to the next grade anyway, where he can’t cope up. Throwing more teachers at the higher grade thus won’t help, since the root of the problem is something else.

 

Teacher absence is widespread because there’s nobody to check because supervisor roles lie unfilled. Absent teachers make a bad situation worse.

 

The curriculum isn’t designed with a rural student in mind. Her level, her lack of any educational support at home, her probably being the first-generation learner – if the content isn’t designed with such a student in mind, well, it just doesn’t work. Muralidharan calls this graph is “the most important graph for understanding the Indian education system”:

 

Let’s look at what it conveys. The x-axis shows the grade in which the student is. The y-axis shows the actual learning level of the student. Ideally, we need the blue line, which indicates that the two are the same (grade of student = learning level of student). Reality though is the red line – the learning level of the student is below the grade of the student. Those circles in the graph? They indicate how many students in the class are at different levels of learning. Unsurprisingly, “there is enormous dispersion in within-grade learning levels”. This graph clearly tells us that no matter how much we focus on infrastructure, teacher qualification or salary, it won’t help if the students can’t learn to the needed level at the earlier grades. Once they fall behind, it is almost impossible to catch up later.

“One reason well-intentioned policies fail is that they are often designed by elites who do not face the same constraints as the policies’ target population.”

 

Our education system, says Muralidharan, is designed for “sorting” rather than “human development”. Its aim is to “filter” the best for further education rather than teaching everyone the basics. This, he explains, the apparent discrepancy of the same education system producing world class workers in IT and pharma and also producing countless students who can’t even read at 2^nd grade level. It also explains why companies complain that graduates have almost no real-world relevant skills (a filtration over learning system encourages rote learning). The system needs to be changed to focus on foundational skills – you need to “learn to read” early to be able to “read to learn” later.

 

All gloom and doom then? The rest of his chapter (and the next blog) are about possible fixes, based on actual attempts and corresponding data.