Making a meal of school kids’ diet

‘Jamie Oliver’s healthy school dinners continue to boost learning, study shows’, declared a report in the Guardian. But are our children really being held back by bad food? Can improving school meals really save the day?

The Guardian story is based on work by researchers at Essex University and Oxford University to be published in the Journal of Health Economics. The researchers, Michèle Belot and Jonathan James, compared exam performance in the London borough of Greenwich – the first borough to take up Oliver’s ‘Feed Me Better’ campaign and menus – with a number of nearby boroughs to see what effect the school meals changes may have had on exam performance.

Comparing marks in Science, English and Maths at Key Stage 2 (usually children aged 11), the report finds that educational outcomes ‘did improve significantly in English and Science’ while there was an improvement in Maths that wasn’t statistically significant. The report also finds that ‘authorised absences – which are most likely linked to illness and health – fell by 14 per cent’.

While that sounds like a clear win for the new school meals, things are a little more complicated. The first problem is that this is not a randomised clinical trial with a clear control group and an intervention group. The study was examining a series of year-groups in different boroughs with changing compositions. For example, the number of children from low-income households or who do not speak English as a first language varied both over time and between boroughs. Perhaps more importantly, the number of children entitled to free school meals varied, too.

Take just one school year-group as an example: In Greenwich, in 2006-07, 37.96 per cent of children were from poorer households, 34.98 per cent were eligible for free school meals, and 69.96 per cent used English as a first language. In the non-Greenwich boroughts, 44.69 per cent were from poorer households, 40.24 per cent were eligible for free school meals, and just 44.69 per cent used English as a first language. The intervention group in Greenwich is, therefore, significantly different from the control group outside Greenwich.

The study uses econometric techniques to try to control for these differences, but these differences already suggest that making strong claims about the role of school meals will be tricky. We then need to take into account that, based on the raw scores and the percentiles for the study schools’ performance in a national context, both Greenwich and non-Greenwich schools improved over this period. The claim is simply that the Greenwich schools improved more.

There’s also the problem that the researchers weren’t able to distinguish between children who did and did not eat school meals – they only had data for the performance of the school as a whole. Since around 45 per cent of the children took school meals, this means that the majority of children were not eating the new school meals.

Given these different factors, some caution needs to be taken about reading too much practical significance into fairly small improvements in exam results. Moreover, it would be wrong to assume that any improvement must have been caused by the nutrients in the school meals. The ‘Feed Me Better’ campaign devoted a lot of effort to trying to convince parents and pupils to eat these school meals, building food-related activities into the school curriculum. The most marked improvement in scores was in Science, which at Key Stage 2 includes a significant amount of material on nutrition. On the other hand, in Maths, where there would have been no such effect, the improvement in scores was not statistically significant.

The researchers also note that the improvements in performance have not been uniform. ‘Our results show that the changes have been more pronounced among some groups of pupils than others. Specifically, the improvement in test scores is more pronounced among girls than boys and among children from middle and high socio-economic status. The second effect is not necessarily expected, because children from lower socio-economic status receive school meals for free and are therefore more likely to have been affected by the menus.’ While the authors suggest that this could simply be a negative, short-term reaction to the changes among poorer children, it does raise the possibility of alternative explanations for these results.

The researchers do themselves no favours by making some fairly sweeping statements about the state of children’s diets. In the introduction, they write: ‘Children’s diet has deteriorated tremendously over the last decades and has become a major source of preoccupation in developed countries, in particular in view of the rising rates of obesity among young children, observed across all developed countries.’ It is true that here has certainly been a rising preoccupation with food and obesity, but does that mean that diets have necessarily worsened? In terms of the affordability and variety of foods available, the opposite is true. Moreover, childhood obesity rates have actually plateaued in recent years, even before the roll-out of the new school meals.

Improving school meals is no bad thing, though at a time of competing demands on school budgets and parents’ incomes, it may not be as much of a priority in the next few years. The real danger with such a study is that it will be used to bolster a reductionist view of education: we don’t need to devote resources to teaching children better, so the thinking goes, we can improve education standards simply by tinkering with nutrients. That would be a very negative educational outcome.

Rob Lyons is deputy editor of spiked and blogs at Panic on a Plate.