ºÚÁϲ»´òìÈ

Free school meals as a measure of disadvantage

Whether a child is recorded as receiving free school meals (FSM) is an influential indicator of disadvantage in England. Our previous blog highlighted the strengths of the measure, and lack of appropriate alternative. However, in our , we find many issues that limit the usefulness of FSM as a measure of disadvantage, which call for caution when using and interpreting this measure.

An inconsistent measure over time and across places

One of the strengths of the FSM measure is that it has been collected over time and across all of England in the National Pupil Database. Yet we find that FSM is not identifying equivalently disadvantaged children across time and place.

Changes in policies (such as FSM eligibility criteria) and events (for instance the Covid-19 pandemic) have led to changes in which children and the number of children who are registered as FSM.

For example, the introduction of transitional protection during the rollout of universal credit meant that from 2018, any child already receiving FSM kept their eligibility until the end of their education phase, regardless of family income. New applicants, however, had to meet the new income threshold. As a result, some children continued to be recorded as FSM despite no longer meeting the income criteria, while other children with lower household incomes – who were above the threshold and not previously registered – were ineligible and therefore not counted.

Differences in registration processes across local authorities also mean that even children who meet the same FSM eligibility criteria have a different likelihood of being registered in different areas. For example, in some areas parents must complete an application for FSM, while other areas use local administrative data to identify and automatically register entitled children.

‘When using FSM to compare rates of disadvantage over time and across areas we are not comparing like for like.’

When using FSM to compare rates of disadvantage over time and across areas we are not comparing like for like. What may appear to be a trend in increasing/decreasing levels of disadvantage, or higher/lower rates of disadvantage across areas, may be an artefact of changes in policies, and/or different registration practices.

FSM excludes many disadvantaged children

A of children in poverty are not registered for FSM. One reason for this is the FSM eligibility threshold is low and inflexible, (for example, the current income threshold of £7,400 per year does not account for family size – and larger families are more likely to be in poverty).

The FSM measure also represents some disadvantaged children better than others because certain groups face greater barriers to registration. As a result, these children are underrepresented in the data despite experiencing equal or greater levels of disadvantage. For example, applying for FSM can be more difficult for parents who speak English as an additional language.

Research-informed best practice using FSM

Until we have a better measure of disadvantage, FSM will continue to be used. However, there are ways we can increase the validity of the measure:

• Treating periods of non-enrolment as FSM equivalent – that children who have periods of not being enrolled in state school are less likely to be registered for FSM, even though they are just as likely to be in long-term income poverty as their peers never ‘missing’ from education.
• Using multiple measures to identify disadvantage in the early years – for pre-school children there are multiple indicators of disadvantage: FSM for children attending state-maintained nurseries; Early Years Pupil Premium which is paid for disadvantaged children attending any setting; and whether a child is accessing hours funded on the basis of low income (from age two). Although pre-school children face a higher risk of poverty, they are in the FSM measure. Combining the measures, though still imperfect, captures more disadvantaged children that are otherwise excluded.
• Analysing and interpreting FSM in interaction with other characteristics – of those children who meet the eligibility criteria for FSM, the likelihood of being registered is patterned by characteristics including ethnicity and having English as an additional language. Therefore, where data allows, including these characteristics alongside FSM, for example when evaluating the effectiveness of education policies, can help illuminate the nuances of disadvantage.
• Adding area-based measures of disadvantage – area-based measures, such as the indices of multiple deprivation, provide an additional data source for measuring disadvantage that has been found to have in terms of educational outcomes.

Conclusion

In the long-term we need better data to accurately measure disadvantage in a way that is useful and accessible across policy, practice and academic research. Upcoming to FSM eligibility will only to find a better measurement solution.

In the meantime, the suggestions above may support researchers and policymakers to make use of the advantages of the FSM measure, while accounting for some of the issues which can result in misleading estimates of disadvantage.