Will support for low-income families during school holidays improve children’s lives?

Providing supermarket vouchers to families on means-tested benefits to help feed their children during school holidays should increase spending on food, and reduce hunger and usage of food banks. It may also lead to a fall in obesity and an improvement in children’s performance at school.

Thanks at least in part to a continuing campaign by footballer Marcus Rashford, the government has announced that local authorities will receive £170 million to help families meet the costs of food and other bills over the winter – the Covid Winter Grant Scheme.

This is being seen by many campaigners as an extension of the COVID Summer Food Fund, which offered supermarket vouchers worth £15 per week during the school summer holidays to children in England whose families claimed means-tested benefits. If local authorities were to use some of the money to continue this scheme, it should increase spending on food, and reduce hunger and usage of food banks. It may also reduce obesity and improve performance in exams.

So why aren’t families supported to meet food costs during school holidays every year?

What is the context?

Children living in families claiming means-tested benefits are eligible for free school lunches. Around 15% of children attending schools in England are registered for these free meals.

For some children, school is the only place where they get a nutritious hot meal. This means that while schools are closed, in addition to the risk of missing out on learning, some children are also at risk of missing out on nutrition and perhaps even going hungry (also referred to as ‘food insecurity’).

In normal circumstances, children not attending school during term-time – for example, because of illness – would not be provided with food at home. Nor would they be provided with food during school holidays.

Covid-19 has changed this: schools in England were asked to provide meals or vouchers, funded by the Department for Education, to all pupils eligible for free school meals while schools were closed during the first national lockdown. The COVID Summer Food Fund then took over, providing vouchers worth £15 per pupil per week, redeemable at major supermarkets and select other food outlets, for these children throughout the school summer holidays.

Many local authorities continued this offer during the October half-term holiday, and the announcement of the new COVID Winter Grant Scheme means that this is likely to continue during the Christmas holidays and February half-term break as well.

What does evidence from economic research tell us?

Giving food vouchers to food-insecure households with children is likely to increase spending on food. Families are also likely to buy healthier foods.

Economists have not looked at the impact of such programmes on hunger or food bank usage, but the COVID Summer Food Fund and subsequent schemes may reduce hunger and use of food banks during the holidays for at least some families.

What about wider effects? Children who experience food insecurity tend to have lower cognitive test scores and learn at a slower rate than children who do not. There is less evidence that food insecurity leads to higher body mass index (BMI) outcomes or greater obesity rates. Economists have also not considered effects on other health outcomes, including mental health.

Many countries, including England, have recently adopted or improved the nutritional standards that apply to the provision of school meals. Evidence from the introduction or expansion of these nutritious meal programmes suggests that they can lower BMI and reduce childhood obesity rates. They are also associated with higher cognitive test scores.

But the extent to which this evidence can inform the likely impact of providing food vouchers on children’s education and health outcomes may be limited. This is for two reasons:

The content of school lunch boxes (food prepared at home and sent into school) has been shown to be less nutritious than school meals (Evans et al, 2010), and portion sizes tend to be too large. This suggests that the food bought by families is likely to be of lower quality than what is provided by schools, and therefore less likely to have a positive effect on health and education outcomes.
The studies relate to meals delivered over a longer period of time (although there is also evidence that calories consumed can positively affect test scores in the very short term – for example, Figlio and Winicki, 2005).

Evidence from non-economic sources suggests that children’s cognitive test scores and BMI outcomes tend to worsen over school summer holidays, especially among poorer children (for example, Alexander et al, 2016). The same phenomenon may also occur – albeit to a lesser extent – over the shorter winter holidays. It is possible that food vouchers will help to limit these setbacks, even if they do not have any longer-term benefits.

Related question: How might the crisis affect children from poorer backgrounds?

How reliable is the evidence?

What is the extent of food insecurity among families with children?

UNICEF, the United Nations Children’s Fund, finds that in the UK, approximately 19% of children under the age of 15 live with an adult who is moderately or severely food-insecure. A recent survey by the Food Foundation suggests that this has got worse during the Covid-19 pandemic.

There is also evidence that problems of food insecurity are more severe during the school summer holidays: data from the Trussell Trust show that food bank usage was around 7% higher among families with children in July/August 2016 compared with May/June 2016.

Will families spend more on food as a result of voucher schemes?

It may seem obvious that providing families with vouchers redeemable at supermarkets will increase food purchases, but this need not be the case. The voucher could just replace existing spending on food, freeing households to spend more on non-food items.

Overall, however, there is strong and reliable evidence that food vouchers do increase spending on food. A study in New Zealand randomly provided some food-insecure households with children with food vouchers and compared the spending habits of these families with similar families who were not given vouchers. The authors find that providing food vouchers worth NZ$17 increased food expenditure among food-insecure households with children by NZ$15.20 (Smith et al, 2012).

In the UK, analysis of the healthy start vouchers for low-income pregnant women and households with young children also shows that spending on targeted goods increases, but by a somewhat lower amount – a £2.43 per month increase from around a £17 voucher (Griffith et al, 2018).

Will families buy more nutritious food?

The evidence on this question comes from studies that have given families additional income rather than food vouchers, but the impact on food spending from the two is very similar, giving confidence that results on the quality of food purchased would also be applicable to the provision of food vouchers.

Evidence from the United States shows that households receiving Earned Income Tax Credits (which are targeted at low-income families with children) spend more on healthy foods and protein – including fruit and vegetables, meat and poultry, and dairy products – in the month following a lumpy benefit payment (McGranahan and Schanzenbach, 2013).

Does eating more (nutritious) food lead to better education and health outcomes?

The negative effects of food insecurity on children – including on health, growth and cognitive development – have mostly been studied in developing countries. But there is evidence for the United States that children experiencing even marginal food insecurity – at least one episode in the previous 12 months – have lower cognitive test scores and slower learning than children who do not have any experience of food insecurity (Winicki and Jemison, 2003).

The bulk of the economic evidence on whether the provision of more or more nutritious food can improve children’s education or health outcomes is based on studies that look at the effect of providing free or reduced price meals in school, often to children from disadvantaged families. Several US studies show that providing a free breakfast or lunch for children from low-income families increases test scores (for example, Frisvold, 2015).

There is also evidence that extending eligibility for free school meals to more or even all children in school confers additional benefits. Analysis of the effect of a two-year pilot in two local authorities in England finds that compared with offering free meals just to children whose families were eligible for means-tested benefits, offering meals to all children in infant schools leads pupils to make, on average, four to eight weeks more progress than those in comparable areas (Brown et al, 2012).

Following this pilot, all infant school children in England (those aged 4-7) have been offered free school meals since 2014. Research is in progress on the effect of this programme on children’s educational performance and absences from school, but it has already been shown to reduce BMI and obesity rates throughout the first year in school (Rabe and Holford, 2020).

This contradicts earlier evidence on the effect of free school lunches on BMI outcomes, which finds either no effect or an increase in obesity rates. This is likely to be because nutritional standards were low or non-existent at the time that the earlier research was conducted (for example, von Hinke Kessler Scholder, 2013; Schanzenbach, 2009).

Studies focusing on the direct effect of raising school food standards highlight the importance of providing nutritious meals of appropriate portion size. Jamie Oliver’s ‘Feed Me Better’ campaign of 2004 – which replaced low-budget processed meals taken by around 45% of children in Greenwich schools with healthier options – improved scores in English and science by 6-12% (Belot and James, 2011).

Evidence from Sweden (where free nutritious food was introduced in the 1950s and 1960s) suggests that these benefits can be long-lasting, even affecting earnings in adulthood (Alex-Petersen et al, 2017).

Related question: How will lockdown and recession affect the lives of the very young?
Related question: What will be the impact of lockdown on children’s development?

Will these policies reach everyone they need to?

There are a number of barriers and potential points of delay that may prevent support from reaching all children who would benefit from it.

If support is focused exclusively on children who are eligible for free school meals, then it is reliant on families confirming their eligibility with their children’s schools. Not all families do this: research by the Department for Education in 2012 suggests that 14% of children eligible for free school meals are not claiming them.

It is also likely that there are children not eligible for free meals who would benefit from vouchers. Research has shown that eligibility for free school meals is not a precise proxy for family income (Hobbs and Vignoles, 2010). Figures from a recent Food Foundation survey suggest that around 30% of families with children eligible for free school meals – and nearly 10% of families with children who are not eligible for free school meals – are food-insecure.

In terms of potential benefits, evidence from the universal free school meal pilots suggests that gains in attainment were strongest among pupils from less affluent families (Brown et al, 2012). Similarly, analysis of a free breakfast programme in England suggests that benefits were largest for pupils not eligible for free school meals living in disadvantaged areas.

What else do we need to know?

What is the contribution of food insecurity to holiday learning loss?
Do policies that improve nutrition in the short term have longer-term benefits for children’s education and health outcomes?
Are the benefits of this policy likely to outweigh the costs?

Answers to these questions will provide insight into whether these policies should be continued after the Covid-19 pandemic has subsided.