Follow us
Publications Commentary Research People Events News Resources and Videos About IFS
Home Publications The crisis in lost learning calls for a massive national policy response

The crisis in lost learning calls for a massive national policy response

Observation

By the time the pandemic is over, most children across the UK will have missed over half a year of normal, in person schooling. That’s likely to be more than 5% of their entire time in school. Absent a substantial policy response, the long-run effects of this learning loss are likely to be slow-moving and substantial. We will all be less productive, poorer, have less money to spend on public services, and we may be less happy and healthy as a result. We will probably also be more unequal, with all the social ills that come with it.

So far, there has been a large focus on how to approach GCSE, A-level and other exams given the large average and differentiated loss in learning time. This makes sense as these exams matter. The potential devastating effect of the crisis on educational inequalities and mental health are starting to be recognized. In what follows, I set out the economic case for a massive national plan to address this crisis. It’s likely to be expensive and hard, but it’s also worth the time and money.

The scale of the problem 

By February half-term, the total loss in face-to-face schooling time will amount to around half a normal school year for children right across the UK. That’s before accounting for lower-than-normal attendance rates in the 2020 autumn term, especially in disadvantaged areas. If most schools don’t go back till after Easter, then children will have lost about two thirds of a normal school year.

The difficulties and inequalities in accessing high-quality remote learning are well documented. Pupils from more disadvantaged backgrounds often lack the required digital equipment and study space to participate in effective remote learning. Younger pupils have found it more difficult to engage in remote learning. Schools, teachers and charities – not to mention parents – have gone to huge efforts to do what they can, but there is no substitute for time with a qualified teacher. 

The unprecedented nature of the current crisis makes it hard to predict the actual effects. Some high-quality, yet depressing, evidence is already beginning to emerge. The best study comes from the Netherlands, where schools closed for 8 weeks last Spring. Despite some of the best digital infrastructure in the world for home learning, the empirical evidence shows that the test scores of Dutch primary school children were significantly lower than previous cohorts. The magnitude is almost exactly equivalent to 8 weeks of normal educational progress, suggesting little educational progress was made during the period of school closures. The negative effects are over 50% larger for disadvantaged children. Reviews of the relevant literature and early evidence on test scores in England also point to big losses from missed schooling and widening inequalities.  

Further testing and data will be required to fully understand the scale, geography and inequalities in lost learning. The initial evidence already suggests this could be enormous. 

The negative effects are also likely to extend beyond educational attainment. We are already seeing clear evidence of reductions in mental health among young people, with 27% of young women showing potential mental health problems. Such evidence has already led to calls for an inquiry into the impact of the pandemic on children’s welfare.

What will be the long-run economic consequences? 

An extensive review of the returns to schooling finds that a year of schooling increases individuals’ earnings by 8% per year, on average, across advanced and high-income countries. By the time children go back to school, most will have lost at least half a year of normal schooling. The lifetime costs of this could be very large indeed. Imagine someone earning £1 million over their working life (not far off the likely average in the UK). For this person, losing half a year of schooling will mean losing £40,000 in income over their lifetime. 

This equates to an astronomical £350bn in lost lifetime earnings across the 8.7 million school children in the UK. Whilst this is just an illustration, rather than a precise prediction, it is important to be clear on the mechanisms that could drive such an effect. All learning is dynamic and builds on skills and knowledge gained at previous stages. Without sufficient catch-up, children will leave school with less knowledge and skills that can be applied in their job or a lower ability to gain further skills. With reduced skills and knowledge, there is also a risk that technological progress and innovation will slow.  

Even much weaker assumptions about the returns to schooling still result in enormous estimated costs. If by some miracle we managed to mitigate 75% of the long-run effects of learning loss, the total loss would still be nearly £90bn. Getting a precise estimate of the likely long-run economic cost would be practically impossible, but such illustrations do help show the massive scale of the problem. The long-run costs are likely to be in the hundreds of billions. 

There are very good reasons to believe that the effects will not be equal across children either. Remote learning has been particularly hard for young children and children from disadvantaged backgrounds. Parents with higher levels of education are likely to have been better able to help their children and, potentially, pay for expensive tutors to help them catch-up. 

There is therefore the clear possibility that the effects of lost learning could be neutralized for those from well-off families and the long-run negative effects could be concentrated amongst those from disadvantaged backgrounds. The net result would be a widening of existing inequalities. These are hard and costly to close once children reach adulthood. 

The long-run risks to the public finances are also severe. If 30-40% of future lifetime earnings ends up as taxes, then lost earnings of £350bn would mean over £100bn less tax revenue over the long-run to spend on public services or paying down the debts we are currently accumulating. 

The inescapable conclusion is that the lost schooling represents a gigantic long-term risk for future prosperity, the public finances, the future path of inequality and well-being.

Prioritizing a return to normal schooling 

Large as these figures are, they do not automatically imply that all schools should go back to normal straight away. It is perfectly possible that the health effects of the pandemic outweigh even these large costs when cases and deaths are high. The high long-run costs of lost schooling do, however, imply a need to prioritise a return to normal schooling when the health situation does allow. It also implies a high return to efforts and resources that enable a faster return, for at least some pupils, be this finding more physical space, use of rotas, more people or more resources. 

We need to think big, really big 

How much would it cost to helps pupils catch-up properly? A good benchmark to start with is the cost of half a year of schooling. By my calculations, half a year of day-to-day schools spending is over £30bn across the UK (£25bn in England, £2.8bn in Scotland, £1.4bn in Wales and £1bn in Northern Ireland). 

I am not necessarily advocating for an automatic increase in school spending of £30bn, but correcting a loss on this scale certainly requires a massive injection of resources. Last summer, the UK government announced plans for a £1 billion catch-up fund in England, equivalent to about £130 per pupil or 2% of current per pupil spending, with plans of a similar scale in Wales, Scotland and Northern Ireland. There is much that is good with these plans and policymakers in England have already announced plans for an additional £300m for catch-up. In total, governments across the UK have allocated about £1.5bn towards catch-up. This is tiny in comparison with the scale of the problem. 

The amount of extra resources for catch-up should be far higher. To prevent inequalities from widening, the distribution also needs to be heavily skewed towards more disadvantaged pupils and/or pupils who have seen the biggest losses in educational progress.  

There is strong evidence for the positive impact of school resources on educational attainment and long-run outcomes. It would be misleading to say that we can extrapolate these positive findings to project the likely effects of a massive injection of resources. However, as a country, we have also already decided that £60bn in spending on schools across the UK each year is a worthwhile investment. During lockdown, we had to spend that money in very different ways to normal. But it does not change the fact that it remains a worthwhile investment. 

Increasing learning time

As the evidence from the Netherlands has shown, lost weeks of learning time is likely to coincide with lost weeks of educational progress. Extra resources must therefore be focused on ways to increase learning time, or the intensity/quality of learning time. 

There is strong evidence behind the National Tutoring Programme (NTP) and I expect it to have positive results. However, it is unlikely to be anything like enough to deal with the seismic loss in learning time. By my calculations, the £250m in the schools component of the NTP can buy about 6 hours of tutoring time for each of 1.4 million disadvantaged students. That’s unlikely to be enough. 

We therefore need to think of big and radical ways to increase learning time. This could be extending the school year, lengthening the school day, mass repetition of whole school years or summer schools. And there is sound evidence that increasing instructional time can yield positive effects. Given we’re trying to compensate for half a year of lost normal schooling, such measures would likely be necessary for a few years.  I am not in a position to advocate for any of these options. Indeed, schools and teachers will probably have a much better idea of what is possible with the right resources. But everything should be on the table and we should be engaged in a national debate about the merits and feasibility of all of them. There are already signs in Wales that such options are being considered. 

Of course, none of these are straightforward and all would need to be properly resourced and staffed. Extending the school year or day could not be done by just asking frazzled teachers and staff to work a little bit more, which would risk a mass exodus of school staff. But if we can use Salisbury Cathedral to deliver vaccinations and 750,000 people are willing to volunteer to help the NHS, then we can use our ingenuity and imagination to find the space and people needed to help children catch-up on their education. A national plan with cross-party support and close partnership with teachers, schools and local government could achieve almost anything. 

Priming the supply side 

Any national plan to increase learning time would need to involve more people, be they teachers, support staff or tutors. There are various ways this could be achieved. Teacher training applications were already up last year and continue to increase this year. Reversing cuts to bursaries and increasing places could yield significant extra teachers. We could seek to engage ex or retired teachers through a wide national campaign. And there are various organisations with huge expertise in this area who could take a significant lead, such as Teach First. 

The NTP is also an excellent model to follow. Part of the programme is about finding and assessing the quality of tutors to ensure value-for-money for the programme. This should be expanded and training provided to would-be tutors to ensure sufficient supply of good-quality tutors. 

Profound costs of inaction 

The long-run costs of the pandemic will fall disproportionately on today’s children. Without significant remedial action, lost learning will translate into reduced productivity, lower incomes, lower tax revenues, higher inequality and potentially expensive social ills. 

The lack of urgency or national debate on how to address this problem is deeply worrying. The necessary responses are likely to be complex, hard and expensive. But the risks of spending “too much” time or resources on this issue are far smaller than the risks of spending too little and letting lower skills and wider inequalities take root for generations to come.

 

Coronavirus podcast

Find out more

Press release
In a new IFS observation, Luke Sibieta, Research Fellow at the IFS, sets out the potential long-run costs of lost schooling. The figures are extremely large and are not intended as precise estimates. Instead, they are intended to illustrate the scale of potential costs and risks we face, and the ...