Flying over the UK on a POLAR expedition. The distant cracks in university access are widening.

Focus help on individuals and social mobility will take care of itself.
It seems that everyone is circling around the problem and avoiding getting onto the ground and making changes.

This week the Higher Education Policy Institute (HEPI) published a ranking of all UK universities according to the distribution of its student intake from different areas of the UK with varying historical rates of participation higher education [1]. It was reported in Times Higher Education as “Elite university intakes ‘as imbalanced as poor nations on income” [2]. On the face of it, this looks like a fair way to inform everyone that there is a wide variation between our universities with regard to social inclusion. However, a closer look reveals an inherent ‘circular logic’ in the methodology. This is tied up with radical changes in how such statistics have been gathered in recent years. The data used to make the rankings comes from those of the Higher Education Statistics Agency (HESA) [3]. It looks at one parameter only and makes a comparison to generate a university ranking against a score from 0 to 1.0 with the higher score indicating a greater social inclusion ‘imbalance’. It comes as no surprise that Oxford and Cambridge and other ‘Russell Group’ universities get higher scores. This predilection toward ‘ranking’ is dangerous when it has the capacity to impact negatively upon individuals in need of help. I will argue below that asking universities to address these imbalances based upon post codes, that are linked to larger neighbourhoods associated with low participation in higher education, is a false demand that is a potentially damaging challenge. The roots of low social mobility go much deeper into our communities than any university, elite or otherwise, can be expected to address. The answer to improving social mobility is not to be found in any university ‘gaming’ the rankings. It is to be found in the individual. The more recent methods 'fly high' over a social map of the UK and look at areas rather than individuals. Remember they can see government up there but government may not see them as individuals. Focus help on individuals and social mobility will take care of itself.

The POLAR methodology brings a cold chill to some ‘neighbourhoods’

For the first time this year, HESA has reported only the participation of students from different geographical areas classified according to the rates of Participation in Higher Education known as POLAR groupings. The boundaries that define each geographical area are approximately similar to the established council ward areas.  The POLAR 3 classification method was used by HEPI to generate their ranking [1]. POLAR3 is based on people who were aged 18 between 2005 and 2009 and entered a HE course in a UK HE provider or English or Scottish further education college, aged 18 or 19, between academic years 2005/06 and 2010/11 [4]. A student’s postcode is assigned to a ward or ‘neighbourhood’ that is classified as belonging to a quintile from one to five. If under 20% of those identified in POLAR 3 participated in HE then the ward is allocated a Quintile 1 score. Conversely, at the other end of the scale, if over 80% participate in HE then the ward is allocated a Quintile 5 score. This is a very simple idea that enables councils and HE planners in the university sector to identify areas that need encouragement to participate. The HEPI report takes this a step further by ranking the universities using a well-worn social sciences measure known as the GINI ratio (a simple method developed by Corrado Gini from 1912, defined clearly by the Office for National Statistics and used to assess the distribution of income and wealth [5]).  Simply put, in an equal society it might be expected that an individual university would take students equally in proportion from each of the POLAR 3 quintile classified areas 1-5.  This would merit a perfect score of 0. In the real world, all universities diverge from this ideal and score a higher GINI ratio. A score of 1.0 would indicate that the University was taking all of its students from Quintile 5 wards. The flaw in this approach is the underlying assumption that universities with high GINI ratios take most of their students from wealthy areas. However this is the converse for 11 universities that are not identified in the HEPI report. This has consequences for the potential’ gaming’ strategy used by university senior managers.

POLAR 4 and the effects of changing geographical boundaries.

This approach presents us an even more alarming prospect when making such comparisons into the future. This is because of the introduction of POLAR 4. The more recent POLAR 4 classification is basically similar to POLAR 3 but with one significant change in that: “The methodology for POLAR4 differs from previous releases in the choice of geographical location and because of changes in the availability of data" [6].

Figure 

To illustrate this in one familiar case, I refer to an area of Coventry where I grew up and still have family connections. It has changed little since I left there as a student in 1973. The ward that I lived in was socially mixed and, whilst attending primary school between 1960 and 1965, I became aware of fellow pupils coming from poor housing and poverty. We had very little at home but I was cared for and set off each day well clothed and having had breakfast.  We did not qualify for benefits or a free school meal since my father worked hard in a factory for relatively low pay. 

By contrast, many pupils lacked very basic needs whilst others were much better off than I was. This council ward in Coventry qualified as POLAR 3 Quintile 2 and is illustrated in Figure 1.  

However, POLAR 4 apparently leads to a classification change upwards to Quintile 3. This might suggest a sudden improvement in the fate of people there. Instead, an apparently minor boundary change has occurred as noted in Figure 2. 

Figure 2

The post code area circled in blue has been allocated to another area in POLAR 4. This area has dropped from Quintile 2 in POLAR 3 to Quintile 1 since the burden of the area circled in blue is shifted their way. The area in question was very well known to me as housing the children most impoverished when I was at school. The latest multi deprivation data in 2015 indicate that the area is still much less advantaged than the rest of the council ward [7].


Little has changed apart from the boundary. This shift in boundaries is likely to cause such effects in other cases and illustrates the danger in relying on POLAR classifications on their own in comparisons year on year.

Is Scotland so different?

It came as major surprise that the HEPI report ranking had included universities from Scotland. It is established that POLAR methodology gives a genuinely false picture of neighbourhoods where there are other factors including greater social mixing. HESA itself notes that:  “The relatively high (in UK terms) participation rate in Scotland coupled with the very high proportion of HE that occurs in FE colleges means that the figures for Scottish HE providers could, when viewed in isolation, misrepresent their contribution to widening participation.” As a result, in addressing social inclusion and widening access in Higher Education in Scotland, there has already been a different approach (see TEFS blog February 2018 [8]). This divides the population down to much smaller numbers than POLAR. Students are divided into groupings based upon where they come from and the Scottish Index of Multiple Deprivation (SIMD) [9] associated with that. Because SIMD is applied to relatively small population groups of several hundred people it is relatively ‘fine grained’ unlike the much larger sized POLAR groupings of 5000 or more used in the rest of the UK. Disadvantaged individuals are thus less likely to get lost in areas where the majority are better off. One could argue that the Scottish approach is more valid and ought to be applied to the whole of the UK; including Northern Ireland.

Circular logic: University gaming and the POLAR game changer.

The circular nature of the logic in using POLAR methodology in universities cannot have escaped most observers.  If universities are to improve their ranking in the GINI ratio table, they must recruit equally from each of the POLAR quintile areas. If they manage this it will spiral back into the rankings that compress accordingly. Working alone, they can only achieve this by changing their entry qualification requirements; effectively  by post code. Those institutions with more from Quintile 1 areas might consider taking their ‘foot off the pedal’ in these areas. Those with the opposite problem would use the converse logic.  Regardless of any apparent improvement in the situation, as a result of any changes in HE recruitment, deprived areas will continue to exist unless government intervenes at other levels. Thus universities using the POLAR methodology spiral in a circle around the vastly greater problems that exists below in the communities.

The real danger lies in university managements starting to look at improving their position in the GINI Ratio ranking and not at an improvement in social mobility. This would mostly entail seeking out the better off students from lower POLAR Quintile areas to make up the numbers. The most disadvantaged students in these areas could be simply passed over as they go for the ‘low hanging fruit’.  Rectifying this problem should fall onto the councils and educators working prior to university selection. This involves better resources targeted onto areas that are already well known to them. Instead, the universities might then concentrate their efforts on supporting the individual students that they recruit as a result. The individual is the unit of concern in the university, not the POLAR area that they come from.

Supporting the individual.

In the past, the background of the individual student was considered in relation to which socio economic classification (SEC) group the student came from [10]. This was achieved by asking questions during the application process, such as in UCAS applications, which profession or job or otherwise their parents have.  A review by HEFCE in 2013 [11] concluded that “concerns that the data underlying the NS-SEC based indicator was of poor quality.” This concern was compounded by a change in the question asked in academic year 2008/2009 that appeared to inflate the figures for SEC 4-7 students. Thus the SEC system has been discontinued from 2017. Up to now, universities were trying to achieve targets for intake from students from the lower SEC groups 4 to 7. These include: Small employers and own account workers, lower supervisory and technical occupations, Semi-routine occupations and routine occupations. The rest that never worked or are long-term unemployed comprise group 8. This is in contrast to students that make up the vast majority of the University student population from SEC groups 1 to 3 from: Higher managerial, administrative and professional occupations, large employers and higher managerial and administrative occupations, higher professional occupations and lower managerial, administrative and professional occupations. The move to POLAR rankings soars over the social landscape at a much higher altitude and then miss a significant number of individuals.

However not all is lost. The new Office for Students has set out: “The regulatory framework for higher education in England [12]. Universities there wishing to increase fees will need to submit an Equality Impact Assessment and it essentially leaves it up to the universities to decide how they do it. But diverting increased fees from the better off students across to students that are less advantaged is an inherent assumption.  It means that the individual student needing help will have to be identified. However, the anarchic approach by government means that there will be a divergence in the scale and types of support offered at different institutions. This creates a ‘market’ and more of the inevitable ranking tables will emerge to feed it.

The Office for Fair Access and HEFCE [13] together provide some insight into what the priorities of the Universities should be: “….disadvantage tends to follow students throughout higher education, and therefore institutions should analyse their own data to assess how best to support access, student success, and progression for this group”. They acknowledge that there is a: “…lack of nationwide data on student success and progression based on ethnicity, gender and socioeconomic status combined”.

Conclusion.

Support for the individual must lie at the core of help for students going through school and beyond. Flying over the social landscape with a POLAR view is most likely to miss many needy individuals.  The society that we all are part of is ultimately composed of individuals. Any teacher or lecturer in a university or college regularly meeting and helping students knows that individual students have problems that vary widely. If more disadvantaged students enter universities, it will be as a result of hard work by teachers in schools with disadvantaged pupils alongside the hard pressed social workers. Only then should the task be handed to universities and colleges that will see their challenges increase. This will  further exacerbate the problems in an equality environment that is polluted by fees, maintenance loans that are insufficient and the stress of precarious part-time jobs for some students.

In my capacity as a university teacher, whilst also serving on the governing body of a Russell Group university, I observed how senior management had lost touch with the situation on the ground in their own institutions; both with students and ordinary staff. Their ‘myopia’ was associated with their obsession for tables and rankings coupled with their ‘altitude’ that makes them remote from the situation on ‘on the ground’. The tension between improving access versus improving the quality of students entering, as measured by UCAS Tariffs, was evident. These two rankings are logically opposed to each other.  It seems that everyone is circling around the problem and avoiding getting onto the ground and making changes.

The problems of equal access in universities reflect wider problems and are symptomatic of these. There is a widening gender gap in favour of women students entering universities. Those that are disabled need more help. Those that cannot cope have mental health problems and many are less advantaged students that struggle financially. The latter are often the product of sustained poverty at home. Reports in the media this week of starving children in our schools [14] is nothing new and this is not going away fast without meaningful action on the ground as opposed to observation from the air.

Mike Larkin, retired from Queen's University Belfast after 37 years  teaching Microbiology, Biochemistry and Genetics. 

References:

[1] Benchmarking widening participation: how should we measure and  report progress? Professor Iain Martin, Vice Chancellor of Anglia Ruskin University. http://www.hepi.ac.uk/wp-content/uploads/2018/04/HEPI-Policy-Note-6-Benchmarking-widening-participation-FINAL.pdf

[2] Elite university intakes ‘as imbalanced as poor nations on income’. Times Higher Education 5 April 2018. https://www.timeshighereducation.com/news/elite-university-intakes-imbalanced-poor-nations-income

[3] HEFCE POLAR MAPS.  http://www.hefce.ac.uk/analysis/yp/POLAR/

[4] HESA definitions. https://www.hesa.ac.uk/data-and-analysis/performance-indicators/definitions#low-participation-neighbourhoods-polar3-applicable-t1-t2-t3

[5] Office for National Statistics GIMI coefficient. https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/families/methodologies/theginicoefficient

[6] HEFCE POLAR 4.  http://www.hefce.ac.uk/pubs/year/2017/201729/

[7] English indices of deprivation 2015: Ministry of Housing, Communities & Local Government. https://www.gov.uk/government/statistics/english-indices-of-deprivation-2015. The Index of Multiple Deprivation (IMD) is a measure of relative deprivation for small areas. It is a combined measure of deprivation based on a total of 37 separate indicators that have been grouped into seven domains; income; employment; health deprivation and disability; education, skills and training deprivation; barriers to housing and services; crime; living environment.

[8] TEFS Blog February 2018. http://studentequality.tefs.info/2018/02/the-betrayal-of-generation-of-poorer.html  and the Scottish Funding Council http://www.sfc.ac.uk/

[9] Scottish Funding Council Report on Widening Access 2015-16 http://www.sfc.ac.uk/publications-statistics/statistical-publications/statistical-publications-2017/SFCST082017.aspx

[10] Office for National Statistics Socio economic classifications. https://www.ons.gov.uk/methodology/classificationsandstandards/otherclassifications/thenationalstatisticssocioeconomicclassificationnssecrebasedonsoc2010

[11] HEFCE Review 2013 How should we measure higher education? A fundamental review of the Performance Indicators. http://www.hefce.ac.uk/pubs/rereports/year/2013/ukpireview/#d.en.85232

[12] The regulatory framework for higher education in England. https://www.officeforstudents.org.uk/advice-and-guidance/regulation/the-regulatory-framework-for-higher-education-in-england/

[13] HEFCE Policy Guide 2017: Tackling inequality. http://www.hefce.ac.uk/sas/inequality/differential/ and Office for fair Access https://www.offa.org.uk/universities-and-colleges/guidance/topic-briefings/topic-briefing-white-british-students/#individuals

[14] BBC report 2018  http://www.bbc.co.uk/news/education-43611527

and survey. http://www.cpag.org.uk/sites/default/files/Child%20poverty%20and%20education%20-%20A%20survey%20of%20the%20experiences%20of%20NEU%20members.pdf

Guardian April 2, 2018. https://www.theguardian.com/education/2018/apr/02/teachers-warn-of-growing-poverty-crisis-in-british-schools

Guardian 2016.  https://www.theguardian.com/education/2016/jan/15/hungry-pupils-breakfast-teachers-survey

and BBC Scotland 2015,   http://www.bbc.co.uk/news/uk-scotland-35196452