Can we rank social mobility?

Whether we like them – or not – league tables are here to stay. Students (and their parents) use them to make choices about which university to go to. Employers use them to decide where to focus their recruitment efforts. Universities use them to decide what initiatives to concentrate on. The rankings are produced using a basket of metrics based on education (employability of graduates, student satisfaction) and research that are then weighted by the team producing the table.

Incentivising better performance

There is one important aspect of the university mission that is not captured in any league table; social mobility. Universities have long been engines of social mobility, and the government is (quite rightly) asking universities to do more in terms of improving equitable access, and to ensure that students have equal opportunities for success once they get to universities. However, social mobility is never a factor in published league tables, and indeed success in attracting a broad range of students is likely to make it more difficult to score well in many of the standard metrics used to derive rankings.

Without having a measure of social mobility, league tables will reinforce social division. Take just one metric that is central to many league tables; employability. It is recognised that students from less advantaged backgrounds are less likely to get highly-paid graduate-level jobs [pdf] than their more privileged contemporaries. There are many reasons for this, and much work is being done by universities and employers to mitigate this, but that is the situation. Therefore, universities that have a high proportion of students from disadvantaged backgrounds will, all other things being equal, do less well in rankings. That in turn will reduce the attractiveness of the university to employers. Students from privileged backgrounds find it easier to access high ranked institutions, and so there is a positive feedback loop reinforcing inequality.

What I have applied to employability can just as equally apply to degree outcome, student satisfaction and other measures used in the league tables.

Diversity improves performance

At present, the exclusion of metrics measuring social mobility represent a potential social injustice. However, there are other reasons for measuring how good universities are in this respect. The first is that diversity is central to the educational mission of universities. Mixing with people of different backgrounds, beliefs and cultures drives a wider education that benefits all members of the university. Many young people recognise this – on a purely anecdotal level I went with my son to the open day of a high-ranking university. On the train home he said that he could not see himself there as all the other candidates were too like him. He, like many other young people, sees university as an opportunity to broaden their horizons by interacting with people from many different backgrounds.

That is not to deny that some students and their parents see university as a place to reinforce their exclusivity. These families may see social mobility as a threat to their privileged position.

Employers are also realising that the more diverse their workforce the better decision-making can be and the more successful their business. They are keen to seek out diversity not just as a tick-box exercise but because they see it as vital for their companies. The efforts that many employers are making are recognised in the Social Mobility Employer Index, published by the Social Mobility Commission.

Improving league tables

If the university league tables are therefore going to remain relevant to the needs of society (and the students and employers who use them to make choices) then they must find a way of including social mobility into the equation.

This can be done. One way is by benchmarking. In August 2017, the Economist published a league table showing salaries 5 years after graduation. They were able to benchmark the expected earnings, so producing a league table of ‘value added’. It showed a very different picture from many rankings, with just two Russell Group universities (Nottingham and Oxford) joining Portsmouth, Aston, Newman, Bournemouth, Robert Gordon, City, Harper Adams and Brunel in the top 10. Similar benchmarking is used as part of the Teaching Excellence and Student Outcomes Framework (TEF) assessment of educational quality, in which each university’s data is compared to benchmarked data. The Guardian league table, while not measuring social mobility, does include a value-added component derived by comparing the entry qualifications with degree outcome.

A second approach is to measure the diversity of universities. Data on various aspects of diversity are submitted annually and form part of the access agreement that universities who charge more than £6,125 fees make with the Office for Access. They are therefore readily available. The Institute for Policy Research at the University of Bath recently analysed the ethnic diversity of universities, comparing the racial segregation within and between universities and that of their surrounding area. This resulted in a ranking that, while concentrating on just one measure of diversity, highlights the astounding differences seen between universities.

Finally, it is possible to measure how successful universities are in ensuring that all students have an equal chance of success. The size of the attainment gap between different groups of students is a measure of just that.

Including social mobility in league tables is going to be necessary if the rankings that they produce are going to remain relevant to the needs of society, students and employers. Those involved in producing these tables will need to produce a basket of metrics that can serve to indicate how universities are performing in this respect.

post list Latest articles

4 responses to “Can we rank social mobility?

  1. Really thought provoking blog. The Academy appears rife with homophilic institutions. Real measures of social mobility and levels of diversity would be very useful and quite telling, provided they weren’t used in a fatalistic manner to excuse universities from seeking meaningful social change

  2. My understanding is the social mobility commission was working on a university social mobility ranking project. I don’t know where this stands now after Alan Milburn’s departure.

  3. The phrase ‘engines of social mobility’ (EOSM) has been linked to practically every aspect of UK education. Examples of engines in policy-wonk-discourse include: schools (in general), Free Schools, Grammar Schools, 6th Form Colleges, University Technical Colleges, adult education, Academies, Universities. It has even been applied to Public Schools (!) by Dominic Carman.

    The best known example is Michael Gove’s 2011 speech introducing the public premium:

    ‘Schools should be engines of social mobility. They should provide the knowledge, and the tools, to enable talented young people to overcome accidents of birth and an inheritance of disadvantage in order to enjoy greater opportunities’.

    Though this was not even the first time Gove used the phrase.

    As a political sound-bite, EOSM seems as ubiquitous as ‘centres of excellence’ (COE). Indeed, all of the candidates for engines have at one time or other, by one politician or another, or by one VC/Head/Principal or another also been labelled as centres of excellence. The correspondence is so exact that many now equate engines of social mobility with centres of excellence: EOSM = COE. So to rank excellence, simply rank by engine-ness.

    The author of the post suggests diversity (D) as a measure of engine-ness. Hence:

    D = EOSM = COE.

    It sounds beautifully simple – but go back to paragraph one (above) and consider how much disagreement there actually is over the merits of Free Schools, Grammar Schools, UTCs, Academies, and Public Schools. All are defended as EOSM, all claim to be COE, but does it follow they are all D ?

Leave a Reply