The whole point of the awards is to highlight successful approaches we didn’t even know existed

Scientific evidence shows that well-designed prizes can be powerful drivers of behaviour change. For example, the Ansari X Prize – which offered $10 million for whoever could get a manned spacecraft into space, twice, in two weeks – prompted investment of over ten times the prize fund; a boost to R&D that traditional commissioning would never have achieved.

The new Pupil Premium Awards borrow from this rationale and similar successes. The size of the prizes will turn heads (indeed that is the point), but the total prize fund of £4 million is the equivalent of just 0.16 per cent of the cost of the pupil premium policy itself.

If distributed evenly, schools would receive just £160 each. Instead, national awards of £250,000 for secondary and £100,000 for primary school are commensurate with the importance of improving the attainment of disadvantaged pupils.

They will highlight the art of the possible and encourage successful schools to collaborate with those that need help tackling this vital challenge.

Jack makes some thoughtful points, but appears to have missed some important features of the new awards. So, let’s first run through how they actually work.

In stage 1, qualifier awards are won by the biggest improvers. Eligible schools are automatically entered and ranked by a three-year average improvement in value-added attainment of premium pupils.

In stage 2, national award entries are judged by an esteemed panel including pupil premium champion Sir John Dunford and Education Endowment Foundation (EEF) chief executive Kevan Collins.

One secondary will be awarded £250,000 and one primary £100,000, with up to eight regional winners taking £100,000 and £50,000 respectively.

There are three key features about this process that warrant elaboration. One, automatic entry means more schools will engage with the awards.

Two, Jack correctly highlights the dangers of inferring too much from one-off test scores and we are delighted to see him quote Daniel Kahneman, whose work we hold in the highest regard at the Behavioural Insights Team.

We agree that raw attainment will vary from year to year due to factors beyond a school’s control. To mitigate this the stage 1 qualifier awards are to be made on the basis of a three-year improvement in value-added attainment. One random year of great results will not be enough to ensure a qualifier award.

Finally, the national awards will be judged by looking not just at quantitative attainment improvements but also at a qualitative explanation of how that improvement was achieved.

Of course, the top-ranked school in this metric may not be truly ‘better’ than, say, the 34th. But from here, the role of the judges is key. They will ensure national awards are won by schools that not only show statistical improvement but also a commitment and passion for helping other schools make the same improvements.

Indeed, the whole point of these awards is not just to reward schools that are already making progress but also to inspire schools to engage with the EEF Toolkit or other research and perhaps tell us of approaches we didn’t even know existed.

The qualifier schools will be inspired to push harder and perhaps win the national prize or one of the regional awards. Thousands of other schools that narrowly missed out on qualifier awards will be told how close they were and encouraged to learn from the EEF Toolkit how to improve further still.

This is what the awards are really about – bringing to life and inspiring the other schools to turn around the lives of their most disadvantaged.

But Jack is right to challenge the current funding and the awards in one fundamental sense. At present the pupil premium is essentially distributed by input criteria (the number of disadvantaged students) and schools are right to be sceptical of simple ‘payment by results’ systems that can entrench the position of the high performers and hold back those who are struggling.

However, there is a compelling case to be made for rewarding improvers, and far more than the 500 schools that will win at this year’s Pupil Premium Awards.

As Jack argues, there may well be others that are equally good but will, unfortunately, narrowly miss the statistical threshold for qualification. But the solution to the statistical problem that Kahneman wrote about would be to increase substantially the number of qualifiers for the national awards, not to abandon rewards for improvers altogether.