Kids’ educational test scores are a major cause for concern across the world. Learning plummeted nearly everywhere during the Covid pandemic – but even before that, standardised test result measures in mathematics, science, and reading were heading in the wrong direction.

Education truly unites parents across the world, although the level of challenge differs, with American and rich world results stagnating at relatively high levels, whereas children in the world’s poorer half struggle with even reading a simple sentence or doing basic math.

But after years of experience, it has become clear which policies don’t work at all – even if they have loud backers.

Spending:

Increasing per-pupil spending sounds like a no-brainer, but it can deliver little or no learning at all if the money isn’t used wisely.

India showed this when it increased spending per primary school pupil by 71 percent over seven years, but reading and math test scores still declined sharply.

The go-to policy for many teacher unions and politicians is making classes smaller.

This sounds like it should make a big difference: teachers can devote more time to individual needs. However, analysis finds that reducing class sizes is one of the least cost-effective ways to improve student learning.

A 2018 review of 148 reports from 41 countries found smaller classes had “at best a small” effect on reading proficiency and no effect on math.

Paying teachers more is another favoured approach. But even dramatically increasing salaries can have little effect on learning. Indonesia embraced all these popular policies at once: it doubled education spending to achieve one of the lowest class sizes in the world while doubling teachers’ pay – yet a landmark randomised controlled study showed “no improvement in student learning”.

The awkward truth is that the commonly promoted approaches – increasing salaries for teachers, lowering class sizes, and building more schools – are costly and do little or nothing for learning.

But there is a promising policy, which comes from an unexpected place. Malawi – one of the poorest countries in the world – has suffered from overcrowded classrooms, lack of learning materials, and a shortage of trained teachers.

It’s not somewhere we might expect to find innovative solutions. Yet it is now embracing an educational policy that shows hope in turning things around – and is even being scaled and adapted elsewhere.

When my think-tank the Copenhagen Consensus Centre and the Malawian National Planning Commission joined forces to identify the most powerful, cost-effective policies to boost Malawi’s well-being and growth, one educational policy emerged above all others.

“Technology-assisted learning” sounds deceptively basic. But it solves an often intractable problem. Almost universally, schools put all nine-year-olds in one grade, ten-year-olds in another, and so on. But many of the children in each class are either far behind or far ahead.

Kids in Malawi are now for one hour a day using personalised, adaptive software on a tablet. It first identifies where each child is at, and then teaches it reading, writing, and numeracy at his or her exact level.

Teachers describe how amazed they were to start using the software and discover that their entire classroom of kids would become fully engaged. Children have described the relief of not having to worry about being embarrassed about getting a wrong answer in front of their peers or being forced to compete for time with the teacher.

The policy is incredibly cheap – costing as little as $15 per pupil per year in Malawi, partly because using the tablet just one hour a day means it can be shared among many students. Extensive studies show that just one year of one hour each day can deliver an astonishing three years of normal learning.

Skilled:

Greater learning translates eventually into more skilled adults who will be more productive in the workforce and command a higher wage.

Using standard economic estimates, this will mean that one year of kids spending one hour a day on a tablet will command a lifetime income increase of roughly $16,000.

Given that most of this income will arrive decades into the future, the present-day value of that benefit is some $1,575. That is a phenomenal 106-fold return on a $15 investment.

Malawi is in the process of scaling up the policy to all its 6,000 primary schools, and encouragingly costs are proving to be even lower, making the policy even better.

Currently, almost 300,000 children work at a tablet one hour a day, with the goal to reach all 3.8 million children in grades 1-4 by the end of this decade.

Already, Sierra Leone and Tanzania are starting to work on implementing the same approach. The determination of Malawi shows how the entire poorer half of the world could improve learning for nearly half a billion primary school children.

Our analysis shows that across the poorer half of the world, an educational investment of $10 billion in delivering this approach would deliver more than $600 billion in benefits annually, boosting future productivity.

This approach could also be useful in rich countries, with early evidence from trials in the United Kingdom looking promising.

Parents around the world are desperate for policies and approaches that could turn around poor test scores and ensure that kids are better equipped for the challenges of tomorrow. Tablets to teach at each student’s own level offer a powerful way forward.