Sutton Trust calls for universities to have better data to target access work and contextual admissions.
Many measures of socio-economic disadvantage that use ‘proxies’ based on local areas are not effective at identifying low-income students and could discriminate against certain groups such as those from BAME backgrounds or those students with young mothers.
This is according to a new report by the Sutton Trust that looks at data from over 7,000 young people in the Millennium Cohort Study, exploring how different measures of disadvantage relate to long-term family income.
The report – by Professor John Jerrim of UCL Social Research Institute – aims to identify the most effective measures of disadvantage – particularly to support universities in their outreach work and in using contextual admissions to widen access. Contextual admissions involve universities looking at additional barriers faced by individual students and giving extra consideration to their application or giving them a slightly lower offer.
A major challenge for universities is access to high quality information on a young person’s background, so as to identify those young people who should benefit from contextual admissions and widening access schemes. Universities often use ‘proxy’ measures, for example looking at the local area someone grew up in based on their home postcode. This report provides a comprehensive overview of how well such measures capture individual family income.
Of the nine measures examined in the research, the most effective indicator of childhood poverty is the number of years a child has been eligible for free school meals. However, universities don’t have access to verified data on free school meal eligibility and rely on students self-reporting their eligibility.
Because of problems accessing data on free school meals eligibility, universities commonly use POLAR (Participation of Local Areas) data in their admissions processes. This measure looks at a young person’s local area and assigns them into one of five groups, depending on the proportion of young people in their area that go on to university. These equal groups, or quintiles, range from the most under-represented areas to the least.
According to today’s report, POLAR is a very poor indicator of family income, in part because it was never designed to measure socio-economic disadvantage. Almost half (48%) of children classified as disadvantaged by POLAR are not from a low-income background. Of all the measures studied in today’s report, it also contains the greatest biases against certain groups.
A focus on POLAR fails to capture many BAME students, those living in London, those with young mothers and those who rent their home. However, other area-based measures performed better than POLAR in capturing long-run family income, including ACORN, a tool which needs payment to access. ACORN uses the Land Registry and various other data sources to classify UK postcodes into 62 different types by level of socio-economic advantage.
The findings are also relevant for employers, charities and other organisations who want to accurately identify low-income young people in order to provide them with further support. The Sutton Trust is making several recommendations in today’s report to improve the targeting of university widening access schemes and the use of contextual admissions.
For its part, the Sutton Trust will review the basket of measures it uses to target its own access work to ensure it is as effective as possible at identifying socio-economic disadvantage. The Trust will also be convening university partners and non-profits working in the access space to discuss the findings and share good practice.
Sir Peter Lampl, founder and chairman of the Sutton Trust and chairman of the Education Endowment Foundation, said:
“In order to widen access fairly and effectively, universities need to know which students would benefit most from outreach programmes and contextual offers. But as today’s research shows, the measures they use are not as effective as they should be at identifying low-income students. They miss out some who deserve support while inaccurately flagging others. It is of particular concern that local area indicators such as POLAR are biased against some groups like young people from BAME backgrounds or those that live in rented accommodation.”
“As a practical next step, the government should make sure that universities have access to data on free school meal eligibility, and target support where it is most needed.”
NOTES TO EDITORS