Free school meal (FSM) eligibility is one of the most common ways to identify pupils from socio-economically disadvantaged backgrounds. However, many questions remain about which measures best capture socio-economic disadvantage, especially once pupils leave secondary school.
In line with common policy and research practice, the EEF uses FSM6 – pupils who have been eligible for free school meals at any point in the last six years – to define our target group.
But FSM6 only applies up to the end of Year 11. Beyond that, there’s no clear agreement on how to define socio-economic disadvantage. As we commission more evaluations in the 16 – 19 sector, this poses a challenge.
To help us make an informed decision about the most appropriate measure of disadvantage, we commissioned FTT Education to do a piece of work exploring the existing and derived indicators of disadvantage in two key datasets: the National Pupil Database (NPD) and the Individualised Learner Record (ILR).
The goal was to understand how the size and make-up of the 16 – 19 disadvantaged group varies according to which measure is used.
Which measures did we explore?
- FSM6: Eligible for FSM at any point in the six years before the end of Key Stage 4 (KS4). Used by the DfE to allocate Pupil Premium funding and available in the NPD for use in research.
- FSM6 CLA: FSM6 plus students ever looked after by the local authority or other state care. Used by the DfE when publishing statistics about the attainment and progress of disadvantaged pupils at KS4 and available in the NPD for use in research.
- FSM6 Y12: FSM6 plus free meal eligibility in October of Year 12. This is a derived indicator, calculated by combining FSM6 with either FSM eligibility from the NPD’s Autumn School Census (for school-based providers) or free meal eligibility from the ILR’s October data return (for all other providers).
- Ever FSM Y11: Eligible for FSM at any point up to the end of Year 11. While there is an indicator available in the NPD, it is based exclusively on the Spring School Census in Year 11. For completeness, the researchers opted for deriving the indicator based on all historical School Census records for each student.
- Ever FSM Y12: Ever FSM Y11 plus free meal eligibility in Year 12. This is a derived indicator, calculated by combining the researchers’ derived Ever FSM Y11 indicator with either FSM eligibility from the NPD’s Autumn School Census or free meal eligibility from the ILR’s October data return.
- IMD27: Living in one of the 27% most deprived areas in Year 12, based on home postcode, according to the Indices of Multiple Deprivation (IMD). This indicator is used by the DfE to calculate the economic deprivation component of disadvantage funding for 16 – 19 providers, and is a derived indicator based on the deprivation level of students’ home postcode.
What did we find?
1. Different measures give groups of different sizes
Depending on the measure used, between 23% and 34% of the 16 – 19 population are identified as disadvantaged. FSM6 identifies the smallest number of disadvantaged students, with each subsequent indicator adding more students to the disadvantaged cohort. The biggest increase comes from switching from FSM6 to Ever FSM Y11, with Ever FSM Y12 only increasing the disadvantaged cohort by a relatively small amount.
2. There are limitations to free meal data in Year 12
Only about half (51%) of pupils who were eligible for free meals in Year 11 are still recorded as eligible in Year 12. The picture is uneven too: 86% remain eligible in school-based providers compared with just 37% in other providers.
This suggests the data has limitations. For example:
- The NPD records students who applied for FSM and met the criteria.
- The ILR only records students who are both eligible and actually take up the meals. On top of that, some students may not claim free meals in 16 – 19 providers because the process is more complicated.
3. Area-level and student-level measures don’t always match
Using IMD27, 31% of 16 – 19 students are classed as disadvantaged. But this doesn’t always align with FSM-based measures:
- 18% of students not captured by FSM6 are included under IMD27.
- 11% of FSM6 students are missed by IMD27.
- In fact, only just over half (53%) of FSM6 students would also be identified as disadvantaged by IMD27.
So what’s the best option?
If our goal was to have the best chance at powering an evaluation for the disadvantaged sub-group, we would opt for either the IMD27 Y12 or Ever FSM Y12 measures. But both risk pulling in students who don’t quite match the profile we’re most interested in. This is also true of FSM6 CLA. For example:
- IMD27 includes some students who aren’t FSM-eligible and misses some who are.
- Ever FSM measures include students who may have been disadvantaged in early childhood but whose circumstances later improved.
- FSM6 CLA includes students who may not be socio-economically disadvantaged.
These leave two contenders: FSM6 and FSM6 Y12.
- FSM6 is consistent with our wider work in schools and provides a clear benchmark.
- FSM6 Y12 goes a step further by capturing a slightly larger group and offering a more up-to-date (though imperfect) picture of disadvantage at age 16.
Where does this leave us?
Deciding on a definition of disadvantage at 16 – 19 is not straightforward. Each of these two options – FSM6 or FSM6 Y12 – comes with trade-offs between accuracy, coverage, and consistency. Our next step is to weigh these carefully against the purpose of our evaluations, so we can be confident that we’re reaching the students who most need support.
Which means that now is a great time for evaluators and the wider sector to weigh in with their perspectives! Please reach out to me at maria.rodrigues@eefoundation.org.uk if you have any thoughts to share.