By Blog Editor, IOE Digital, on 9 December 2013
It is now taken as a given that an ‘aspirations deficit’ is not the reason so many fewer young people from lower socio-economic backgrounds go to university. This assumption is based on the fact that at age 14 the proportion of young people from less advantaged backgrounds who think they are likely to get into university is much higher than the proportion who will ultimately end up going.
However, just as important is what happens over the next few years. Many young people’s expectations of applying to university change during their adolescent years. There are large falls in the proportion who think they are likely to do so and pre-existing inequalities in expectations grow further.
At the same time, we observe a ‘hardening’ in expectations, as individuals tend to move from saying they are ‘fairly likely’ to apply to university, to ‘very likely’; or – on the flip side – from saying they are ‘not very likely’ to reporting that they are ‘not at all likely’ to apply. Being on the wrong side of this hardening makes it far less likely that potentially qualified young people will apply to university.
Understanding what is associated with these changes may be key to sustaining high expectations, encouraging a wider cross section of university applicants, and hence widening access to university. So, what seems to explain the large changes we see?
Unsurprisingly, some change is explained with the young people’s prior attainment, as individuals with lower performance in national tests at age 11 revise their expectations downwards during their adolescent years. Interestingly, while individuals’ GCSE performance five years later, at the end of compulsory education, is also correlated with changes, there is not evidence that it suddenly causes young people to update their future plans. This is likely because GCSE results do not, for most people, come as a surprise: they are partly anticipated and so changes in expectations are smoothed over time.
This transition at the end of compulsory education is also important for another reason. Almost all the decline in expectations we observe is among those who leave education at this point, or, to a lesser extent, one year later. Among those still in education at age 18, there is no decline in average levels of expectation over the teenage years. This suggests that post-16 education is largely seen as a gateway to higher education, rather than as an aspiration in its own right.
Also of interest is whether other individual characteristics have an effect above and beyond differences in attainment. We find that parental socioeconomic status does indeed seem to have an additional impact, although it is heartening to note that it explains less of the differences in expectations than does attainment at age 11 (however, we should remember that this too is socially graded).
We also find that schools are very important for sustaining high expectations and raising low ones during this period. A tenth of the variation in maintaining expectations is explained by schools, and this rises to as much as a quarter when it comes to raising expectations.
Obviously there are many reasons why schools matter so much, not all of which will be within a school’s control (such as any association between family background and school admissions). Nevertheless, our findings suggest that teachers have a key role to play in keeping bright teenagers from less advantaged households on the track towards university.
Young people’s hopes of applying to university start out high. However, sustaining them, where appropriate, is still too heavily associated with their family background. Without breaking this link it is difficult to see how we can go on also to break the link between socioeconomic status and university attendance.
“Teenagers’ expectations of applying to university: how do they change?” by Jake Anders and John Micklewright is published as a Department of Quantitative Social Science Working Paper available from the Institute of Education, University of London website (pdf, 0.9mb).