I spent three years trying to fix a hole in a Labour policy and it nearly broke me

Labour and the Liberal Democrats both want to make free school meals universal – and all they needed to do was fix one little thing. How difficult could it be?

Sign Up

Get the New Statesman's Morning Call email.

If there is ever another non-Conservative government, it’s more likely than not that by the end of it, free school meals, at least in England, will have become a universal, rather than means-tested provision.

Since Labour left office, it’s been proposed by politicians ranging from David Miliband in 2010 to Jeremy Corbyn in 2017, and has been a core part of the Liberal Democrat policy programme on and off for most of the past two decades. The Liberal Democrats introduced universal free school meals for children aged five to seven during the 2010-15 coalition government, and pledged to extend it to all primary school children in 2015, 2017 and 2019.

So it seems likely that the policy will eventually be enacted. I like the policy a lot. While it’s hard to disentangle its precise consequences from other improvements in school quality, we can say with a high degree of certainty that it has improved grade outcomes. As well as saving parents money, it allows schools to improve their food offering for less money if all pupils eat school meals, which has important public health and nutritional benefits.  

It is also comparatively cheap in the grand scheme of things – a huge amount of money to you and me, but peanuts as far as government spending is concerned. It comes with an important price tag, however: at the moment, we use free school meals as a way of assessing how effective the British education system is at looking after poor children, essentially from the start of primary school all the way to the completion of the university qualifications. They’re an important way of marking the education system’s homework and testing to make sure that government policies don’t have a negative effect on the poor.

They’re also a useful way for government to assess how large what you might call the UK’s “at risk” population is. Take the recent shutdown in schools. There will be a group of people who aren't known to the police or to social workers, or who haven'y come into contact with a food bank or Citizens Advice, but who will struggle financially if they have to feed themselves and their children three times a day, seven times a week.

We need a metric that can replace free school meals after they become universal. Indeed, although the power to make free school meals free to all pupils is devolved (the SNP government made school meals free for children aged five to eight in 2015), solving this problem is one major barrier to making them free in Scotland and Wales, where Labour and the Liberal Democrats are in coalition.

The policy teams of both Labour and the Liberal Democrats at the time were well aware of this difficulty, and indeed one of the most frequent conversations I’ve had with policy wonks in both parties is what measure you might use to replace free school meals. (Yes, I really am this cool.)

When I wrote about this problem back in 2017, I assumed that within a couple of days someone would have come up with a solution. To be perfectly candid, I assumed that I would come up with a solution and blog about it while the policy was still in the headlines.

Instead, it’s been a lot trickier than I supposed. The question became something of a white whale. Three successive Liberal Democrat leaders came and went – the policy remained in place, and but the question of with what to replace free school meal (FSM) status remained. There are just days lfof the present Labour leadership, and while I anticipate the policy will remain in place indefinitely, I began to fear I would never come up with a workaround.

Here are a couple of possible replacements for FSM status that I considered and ultimately rejected:

Breakfast clubs 

Breakfast clubs provide, you guessed it, free breakfasts to children whose parents cannot do so, whether for reasons of low income or because they have chaotic lives or complex care needs. The problem with using breakfast clubs as a measure for how the education system is serving the poor is that while they are a vitally important state provision, the majority of people who use them have more complex policy needs and it is unlikely that will change in the short-term.

Tracking how the education system looks after children who use breakfast clubs is important, just as tracking how well the government does at reducing absolute poverty is important. But by using it as your sole metric, you’re missing out on the needs of a much greater number of people – just as you would by focussing solely on absolute, not relative poverty.

There are two dangers here: the first is that policies that adversely effected a much larger number of poor people would go unchanged because they had no negative effect on very poor people. The second is that policies which benefited large numbers of poor people might appear to be bad ones purely because of a failure to do much to help very poor people.

You could in theory fix this by expanding the eligibility criteria for breakfast clubs to the existing threshold for free school meals. I was initially attracted to this idea. In practice, however, the longer I thought about it, the more I thought it was impractical. For it to work as a useful metric, you need to have a similar level of uptake for breakfast clubs as for free school meals, not merely the same threshold for eligibility. A lot of people on free school meals – certainly this was the case for me and many of my school friends – had parents working full-time or in shifts. There are two near-insurmountable barriers to having similar levels of uptake for breakfast clubs as free school meals: for some, it means forgoing a rare meal together as a family, and for others it means a much earlier start to the day.

I just don’t think it’s plausible that you can get similar levels of uptake for breakfast clubs as for free school meals, so that wouldn’t work.

Universal Credit 

This solution has the reverse problem – many more people will claim Universal Credit, assuming it is ever fully operational. If the problem with using breakfast clubs is that you risk measuring something too narrow, the problem with using Universal Credit is that you risk measuring something too broad.

Just use information from HMRC 

The government collects lots of information about what households earn, but it does not have the ability to match that information with how children do at school. You could do this, but a) it would be a really expensive IT project, and more importantly b) the challenge I had set myself was to find a policy broadly acceptable to the Liberal Democrats and the Labour party. I don’t think vastly increasing the scope of government surveillance and data-harvesting meets that threshold.

Cognitive aptitude tests 

Many English schools use cognitive aptitude tests (CAT) for two things: to aid their teaching by providing them with a way of assessing the abilities and needs of their pupils when they first arrive, and as a way of assessing teacher performance (you can use a Year 7’s CAT results to roughly predict what they would expect to get in their GCSES – and assess how big of a difference you are making over their time at school).

You could have a mandatory CAT at the start of compulsory education, and track that. The problem here is that you’re measuring something subtly but vitally different: academic attainment versus deprivation. While many people who do poorly in their CAT tests are eligible for free school meals, by no means all of them are and by no means do all people eligible for free school meals perform poorly in their CAT tests.

I’ll be honest: at the point at which I considered this, I was getting desperate. I had eliminated a variety of other things because they were too narrow or too broad (I’m not going to go through all of them but trust me, there were a lot) and was casting around pretty desperately.

Fortunately, on one of my government-sanctioned daily strolls, I had an epiphany. There is an easy and off-the-shelf metric you could use:

School uniforms and school trips 

One of my many fogeyish opinions is that the recent trend of dressing up on World Book Day is a bad thing – it puts further pressure on overstretched parents and is yet another way to highlight the disparity between well-off kids and poor ones.

This is – and if you thought that World Book Day opinion was fogeyish, oh boy, have I got another – why I remain a zealot for school uniform. School uniforms help to blur the distinction between well-off children and the poor. They also provide useful safeguarding functions by allowing passers-by to identify what school a child goes to.

But there have been a handful of rows in which schools have got into stand-offs with parents over uniform codes. In some cases, the parents seem to be a fault; in others, the school has been. In cases like the latter, the Department for Education has struggled to resolve disputes because it lacks statutory powers.

A national uniform code would eliminate those problems, retain the benefits of having a uniform and allow you to use free school uniform – which you could have at the present level for free school meals – as a metric. This would actually be even better than using free school meals, because the number of eligible families participating would be higher.

Most schools provide free or reduced costs for school trips to families in receipt of free school meals anyway, and many schools or local authorities do the same for school uniforms. They just have no statutory obligation to do so and, at present, school leaders are dependent on the mercy of either local authorities or having to raid their own budgets.

If you don’t want to support school uniforms – remember my original challenge was to come up with a solution acceptable to Labour and the Liberal Democrats – then you could simply achieve the same ends by making school trips free at the same point at which we now have free school meals. In both cases, that would allow the opposition parties to retain their commitment to making free school meals a universal provision – without sacrificing the ability to track how well schools are serving poor children.

Stephen Bush is political editor of the New Statesman. His daily briefing, Morning Call, provides a quick and essential guide to domestic and global politics. He also co-hosts the New Statesman podcast.

Free trial CSS