A unique opportunity to build the evidence base
In 2023, the Mayor of London announced funding to provide free school meals for all primary school children in Greater London. The initiative provided a unique opportunity to learn more about the impact of universal free school meal provision on a range of outcomes.
We run a lot of trials at the EEF, over 240 since we were founded in 2011. Typically, these trials involve deliberately creating two groups: an intervention group that receives a particular programme or approach, and a control group that does not.
The London policy, however, provided a ready-made intervention group — primary school children in London — and a natural comparison group — similar children in schools outside London. By studying differences between these groups, we could learn more about the impact and implementation of universal free school meals.
We are grateful to the Greater London Authority for enabling this evaluation, which helps build the global evidence base on universal free school meals.
Free school meals: targeted vs universal
In England, free school meals have historically been provided on a targeted basis to families who meet specific eligibility criteria, largely linked to income-related benefits. From September next year, this targeted offer will widen to all children whose families receive Universal Credit – an estimated 500,000 more children.
Alongside targeted support, policymakers sometimes consider universal provision, where every pupil receives a free school meal regardless of income. Debates about universal versus targeted models often relate to competing priorities: reducing family costs, tackling food insecurity, improving wellbeing, promoting inclusion, or supporting learning. Attainment is one possible outcome within this broader landscape, but it is far from the only one.
Why policy evaluation matters
The decision to introduce universal free school meals in London came during the height of the cost-of-living crisis. The Greater London Authority highlighted the financial benefits to families, suggesting the policy could “save families hundreds of pounds over the year”. Other hopes for universal provision include improved concentration, better attendance, and enhanced social experiences at school.
Policymakers and the public typically judge success across a range of dimensions, including political popularity, affordability, household impact, and social value. Evaluation, however, has a narrower purpose: to test whether a policy affects specific outcomes of interest and to learn about how policies are implemented. This rigour is essential for building a robust and cumulative picture of “what works”, particularly for improving teaching and learning.
Our evaluation
Our evaluation, carried out by researchers at University College London (UCL), focuses primarily on pupils’ attainment in Key Stage 2 assessments. This outcome aligns with our mission to understand how different approaches influence pupils’ learning, especially for disadvantaged children. However, it represents only one piece of the picture. Universal free school meals could plausibly affect a much wider set of outcomes, including attendance (which we are also looking at through our evaluation), wellbeing, nutrition, social development, engagement, and family finances.
The natural experiment created by the London policy gives us a rare opportunity to examine whether universal provision influences attainment and attendance, while also exploring how the policy is implemented and experienced by families and schools.
Emerging impact findings
We have now published interim findings covering the first 10 months of the policy. These findings should be interpreted with care: they are early results, and the evaluation will continue over two full academic years. Policies of this kind often take time to embed, and effects may only become detectable after a longer period.
At this stage, pupils in London boroughs offering universal free school meals made no additional progress, on average, in their Key Stage 2 assessments compared with similar pupils outside London whose access to free school meals did not change. This finding has a high security rating, but it reflects only the short-term picture and only one possible outcome. It also tells us about this particular universal offer, and not about the targeted provision that all state-funded schools already deliver.
Positive parents and connected classes – implementation findings
Alongside these emerging attainment findings, today’s publication also reports on how the policy was implemented and experienced. These insights help illuminate the broader effects of the policy and provide important context for interpreting future impact findings.
Families – and particularly those whose incomes sit just above the eligibility threshold for the targeted offer – responded positively to the introduction of universal free school meals. Parents commonly reported reduced financial pressure, less stress around preparing food, and improved well-being from knowing their child could access a hot meal each day.
Schools also described using the policy as a springboard to improve the lunchtime experience. Many redesigned dining spaces, streamlined service arrangements, and created new opportunities for pupils to socialise, learn about etiquette, and develop independence. Case-study evidence suggests that shared mealtimes helped encourage inclusion, as pupils tried new foods, interacted more with peers, and experienced a greater sense of belonging.
More broadly, implementation was strong: the vast majority of schools offered a choice of hot meals to all pupils, demonstrating high fidelity to the policy’s aims.
Next steps
A second report will be published in 2026, exploring the policy’s impact in its second year and expanding the range of outcomes considered. It will also look at how universal free school meals are experienced as the policy becomes more embedded.
This EEF-funded evaluation forms part of a wider suite of studies led by different organisations, each examining different aspects of the policy. These interim findings should be understood within that broader evidence-gathering effort. They offer early insights – but they are only the beginning of the story.