Publication - Publication

Commissioner for Fair Access annual report 2019: building on progress towards fair access

Published: 13 Jun 2019

Second annual report from Commissioner for Fair Access, in which he assesses the progress on fair access in Scotland.

50 page PDF

985.1 kB

50 page PDF

985.1 kB

Commissioner for Fair Access annual report 2019: building on progress towards fair access
Chapter 2: Targets and Performance

50 page PDF

985.1 kB

Chapter 2: Targets and Performance


The current targets for fair access, recommended by the Commission on Widening Access and accepted by the Scottish Government, are that, by 2030, 20 per cent of entrants to higher education should come from the 20 per cent most deprived areas in Scotland, as measured by SIMD. Interim targets for full-time first degree entrants have also been set at 16 per cent two years from now in 2021 (and at least 10 per cent in every institution) and 18 per cent in 2026. Progress towards meeting these targets has been discussed earlier in this report.

Two issues continue to be debated:

  • The first, which can be swiftly dismissed, is whether there should be national targets, and institutional targets derived from these national targets, at all. The alternative to national targets is for individual universities to set their own targets using their own indicators of disadvantage. This is essentially the system in England where institutions are required to have access agreements with the Office for Students. Although there is some limited central guidance about the contents of these access agreements, institutions are free to set their own criteria of success. To date, no access agreement has been rejected by the OfS or its predecessor body, the Office for Fair Access, although some institutions have been persuaded to make limited changes in the course of negotiating agreements. The system is not unlike the outcome agreements between institutions and the Scottish Funding Council. The disadvantages of such a system are twofold. There is no 'common currency' for measuring progress and, as the OfS itself has admitted, there is an insufficient sense of urgency. Scotland's progress towards promoting fair access is vindication of a system of national targets;
  • The second issue is how progress towards meeting these targets should be measured. Currently that measure is the percentage of entrants from SIMD20 areas. This has been criticised on a number of grounds - that not all disadvantaged applicants live in deprived areas (and that some students from SIMD20 areas are not themselves disadvantaged); that SIMD20 applicants, regardless of their own individual circumstances, are given priority over other applicants leading potentially to unfair displacement; that universities are forced to compete for a limited supply of applicants from SIMD20 areas; and that there are mismatches between the single indicator used to meet targets, residence in an SIMD20 area, and the multiple indicators that institutions themselves use to identify disadvantaged applicants and target outreach activities.

The Access Data Working Group (ADWG), with representatives from all key stakeholders, was established following recommendation 31 in the CoWA final report that 'a consistent and robust set of measures to identify access students' should be developed. It met four times in 2018 to consider supplementary measures to SIMD. The following discussion reflects the work of that group.

In this section of the report, four different types of measures will be considered:

1. Area-based metrics such as SIMD;

2. Individual-level metrics that are available on a national basis, are robust and reliable and, crucially, available at key decision points;

3. Other measures of individual characteristics that do not pass these tests but which nevertheless are - rightly - used by institutions;

4. Measures that combine area-based metrics and individual-level indicators.

SIMD and community deprivation

SIMD is probably the UK's most sophisticated area-based metric, which has won an award from the Royal Statistical Society. It was originally introduced in 2004, and the latest version dates from 2016. Scotland's 5.3 million population is divided into 6,976 data zones, and each data zone has a population of around 760 people.

It is based on seven indicators which cover various forms of deprivation: income (for example, the number of income deprived people or people on income-related benefits); health (for example, mortality rates, low birth weight and alcohol and drug misuse); employment (unemployment and dependence on benefits); education (school attendance, attainment of school leavers, number of young people not in employment, further education and training, and the proportion entering higher education); crime (rates of violence, vandalism and other offences); housing (over-crowding and lack of central heating); and access (drive and public transport times to key facilities). Using these indicators data zones are ranked into quintiles, from SIMD80-100 (the least deprived) to SIMD20 (the most deprived).

The main alternative area-based metric available for measuring access to higher education is Participation of Local Areas (POLAR), which covers the whole UK. The average population of POLAR medium-level super output areas is similar to that of SIMD data zones. A finer-grain mesh of POLAR is available, although it is not generally used for technical reasons. Like SIMD POLAR is also ranked into quintiles from POLAR Q1, areas with the lowest participation in higher education, to POLAR Q5, areas with the highest participation.

The main difference between SIMD and POLAR is that in the case of SIMD, participation in higher education is only one element of educational deprivation, which in turn is only one of seven aspects of deprivation, while POLAR focuses exclusively on participation in higher education and ignores all other forms of disadvantage. As a result, Scotland, simply because it has the highest higher education participation rate in the UK, has the lowest proportion of POLAR Q1 areas among the UK nations (and Wales the highest). However, Scotland clearly continues to have high levels of social disadvantage. The inescapable conclusion, therefore, is that SIMD is a more appropriate area-based metric to identify deprivation than POLAR, even if the ability to compare performance across the four UK nations is restricted to relative rates of change if different metrics are used.

The main weakness of using SIMD, or any other area-based metric, as has already been indicated, is that it does not describe individual characteristics. As a result there will inevitably be false-positives, less disadvantaged individuals who live in deprived areas, and false-negatives, disadvantaged individuals who live in less deprived areas. For example, only one-third of income deprived people live in the 15% most deprived areas - which means that two-thirds live in less deprived areas. A recent study by Abertay University found that only a third of the students the University admitted on the basis they were disadvantaged actually lived in SIMD20 areas, and a third of their students from SIMD20 areas did not qualify as disadvantaged (according to the University's own criteria, which included attendance at schools which sent few people to university, having care experience, coming from non-graduate families and enrolment on an access course). This is a particular problem in more sparsely populated parts of Scotland where SIMD data zones cover much wider areas with more mixed populations. For example, there are no SIMD20 areas in Shetland, Orkney and the Western Isles. In more densely populated areas, and especially in the west of Scotland, the concentration of social disadvantage makes false-negatives and false-positives less likely.

Individual indicators of disadvantage

The Commission on Widening Access, while arguing that 'the Scottish Index of Multiple Deprivation is currently the most suitable measure of disadvantage', recognised that additional measures could also be used not only to inform decisions about individual applicants but also potentially incorporated in national targets. The Commission identified three types of additional measure - care experience, household income and school environment.

Care experienced

Using care experience as a measure of disadvantage alongside residence in a SIMD20 area is generally accepted. The number and percentage of applicants and entrants with care experience are included in national statistics on access, and all universities use care experience as a marker to identify disadvantage. It is also generally accepted that the needs of applicants/entrants with care experience should be recognised regardless of their socio-economic status, although in practice many may come from socially disadvantaged backgrounds. Any issues about the quality and availability of data about this small group of potential students should be able to be resolved satisfactorily. Universities Scotland has agreed a common definition of care experience, embracing those with experience not only of local authority care but also of kinship care (with institutions free to determine the level of verification they require). Numbers are small but growing. In 2017/18, there were 255 Scottish domiciled full-time first degree entrants who reported care experience, up from 170 the previous year, while the percentage of entrants who reported care experience increased from 0.6% to 0.8%.

Without seeking to diminish the impact of other forms of individual disadvantage, such as those applicants who have been estranged from their parents or are orphans, not all applicants in these - and other - categories are necessarily socially deprived. The proliferation of separate categories of individual disadvantage may tend to detract from a more general definition of deprivation in the context of fair access, such as that which Universities Scotland has developed. In any case some forms of individual disadvantage are 'protected' characteristics, and universities already have a legal duty to make appropriate adjustments. Care experienced applicants are perhaps an exception to this rule - for three reasons. First, there is an overwhelming public and political consensus that they deserve special consideration in the context of fair access; secondly, they are a well defined and comparatively small group; and, finally, the experience of social deprivation and care experience are closely aligned.

Black and Minority Ethnic applicants

There is a case for arguing that black and minority ethnic applicants should also be treated as an exception to the rule that generally fair access should be determined largely in terms of social deprivation. In contrast to England, ethnicity is not such a prominent component in the debate about fair access and widening participation in Scotland. It is my intention to address this issue in greater detail in a discussion paper later this year, along similar lines to the discussion paper on disability.

Free School Meals

Determining a reliable measure of household income as an individual-level metric alongside SIMD is not so straightforward. The obvious indicator is receipt of Free School Meals (FSM), although other candidates include the Educational Maintenance Allowance (EMA) paid to 16 to 19 year-olds who continue in education after the school leaving age. Eligibility for both FSM and EMA is determined by household income thresholds and, for FSM, receipt of certain benefits. There is clear evidence that students registered for FSMs in school are seriously underrepresented in higher education - the ADWG found that around 22 per cent of S5/S6 leavers receive FSM in any year of secondary school but this group only constitutes 11 per cent of those going on to higher education. There is also a considerable overlap between FSM registration and residence in SIMD20 areas, particularly in more densely populated cities.

However, there are four difficulties in using FSM as a reliable individual-level indicator of disadvantage:

1. The individual data collected by schools has only recently met the standards of validation required by official statistics (and some local authorities use additional criteria to assess eligibility);

2. The data only covers FSM registrations, not eligibility, which may be significantly greater;

3. FSM data on individual applicants would need to be shared in a suitably robust form with university admission staff in time to influence decisions;

4. FSMs is not a reliable indicator of individual disadvantage for older students. Other assessments of individual-level household income, such as eligibility for student bursaries and other student support, are currently made too late in the cycle, at the admissions rather than the application stage. To ask SAAS to make these assessments not only for entrants to assess their eligibility for bursaries but for all applicants to determine their financial circumstances would represent a considerable administrative burden.

Despite these issues, the Access Data Working Group has recommended that FSM registration should be used as an individual-level indicator, although this would be determined by registration in any year in secondary education rather than the year when applications were made.

School Environment

A third category of indicators for measuring individual disadvantage which it has been suggested could be used alongside SIMD in national/institutional targets is based on school environment. The most common measure of this type is so-called low progression schools which do not send many of their leavers on to higher education - and, in particular, universities. Much access work is focused around partnerships between universities and low progression schools, whether at the level of individual institutions or through regional groupings. This is an entirely valid - and, indeed, very valuable - approach which recognises the importance of personal links and local knowledge. However, to use attendance at a low progression school as a uniform measure of individual disadvantage across Scotland is not straightforward. The most important difficulty, of course, is that it focuses on schools, not individual pupils. So, like SIMD, it is an area-based not an individual-level metric.

Other Indicators

Other individual-level indicators are available. But none is sufficiently reliable, robust and timely enough to justify inclusion alongside SIMD in national targets. These include the socio-economic status of parents (or, in the case of older students, their own socio-economic status) and the highest level of parental education (and, in particular, whether they were themselves graduates). Both are clearly very powerful influences on access to higher education, and the access gap between advantaged and disadvantaged social groups that fair access policies are designed to reduce. But in both cases the available data is only partial and typically based on self-reporting and therefore cannot be independently validated.

However, the fact that such indicators - along with school environment - are not suitable for inclusion in national targets alongside SIMD does not reduce their value to universities in determining their admissions policies. Rightly universities are free to determine their own indicators for promoting fair access. But three considerations need to be borne in mind:

  • The first is that there is a risk of overlap and redundancy if too many indicators are used to measure individual disadvantage because many may measure the same, or similar, things. It makes sense to focus on a small number of proxies;
  • The second is that, if too many indicators are used to flag up comparatively mild forms of disadvantage, it will dilute the pool of applicants who deserve some form of special consideration, and disguise deep-rooted deprivation. Using too many indicators, even if they are publicly available (as they must be), may also lead to a lack of transparency on the part of potential applicants and their families and also those who advise them;
  • The third is that, although all universities should not be obliged to use the same indicators (because their institutional missions and geographical positions are different, with different balances between supply and demand), there is a case for maintaining broad equivalence - not least for the sake of potential applicants. This will be particularly important if more standardised approaches to contextual admissions/minimum entry standards are adopted across Scotland.

Multiple Equality Measures

The final type of measure is to combine area-based metrics with individual-level indicators. The Universities and Colleges Admissions Service (UCAS) has developed a Multiple Equality Measure (MEM) along these lines, although currently it only covers England. The MEM combines sex, ethnic origin, residence (POLAR in England, although SIMD could be substituted if MEM is extended to Scotland), school type (state or independent) and income (based on registration for FSMs). Like POLAR and SIMD it is divided into quintiles from MEM group 1, the most disadvantaged, to MEM group 5, the least disadvantaged, based on calculations of the probability of going on to higher education via UCAS. Although this 'mixed' metric clearly has potential, it combines people with very different forms of potential disadvantage, for which there are different remedies.

Discussion, conclusions and recommendations

The choice between individual-level indicators and area-based metrics such as SIMD in measuring disadvantage, or some combination of the two, is not simply a technical matter. It also reflects fundamentally different accounts of the causes of inequality in access to higher education and how fair access is best achieved. The first, which can be labelled the 'individual' account, focuses on identifying, and to some degree, compensating for individual deficits in terms of social and cultural disadvantage and, in particular, educational experience. The second, the 'social' account, emphasises deep-rooted, multi-faceted, community-based and (often) inter-generational deprivation.

The 'individual' account steers policy makers, institutional leaders, academic and admissions staff and access practitioners to seeing fair access in terms of opening up pathways into higher education for those individuals who, due to force of circumstances, have been 'left behind' in the race of higher education. In other words, the emphasis is on promoting greater social mobility through improved access to higher education, or the co-option of the 'deserving [educationally] poor'. There are three objections to this interpretation of fair access:

  • It is essentially a continuation of the approach taken to the expansion of higher education opportunities, in Scotland and across the whole of the UK, since the 1960s, although now perhaps with a stronger sense of political urgency. Without underestimating the social and cultural (as well as economic) benefits produced by that expansion, this approach has done little to narrow the access gap;
  • Universities are allowed to remain within their 'comfort zones'. Fair access is largely focused on improving admissions and student support systems to remove hidden or unintended barriers. Universities' core values and practices remain unchallenged. There is less need to ask difficult questions about the extent to which these core values and practices may have been complicit in producing the access gap;
  • There is a risk that improved social mobility, in the absence of seriously addressing the greater need to promote social justice against a background of multiple and entrenched deprivation, may actually weaken still further the social cohesion of deprived communities if fair access produces an exodus of the potentially most talented and motivated young people.

In contrast, the 'social' account steers those policy makers, institutional leaders, academic and admissions staff and access practitioners to seeing fair access in the wider context of social justice (and greater equality of outcomes). My work as Commissioner has inclined me strongly to this second interpretation of fair access. In my view, it is the best way to address the underlying conditions that produce unfair access to higher education; to persuade universities to take a more self-critical approach to their core values and practices (on, among other things, academic progression and graduate attributes) which may reflect and even entrench existing inequalities; and to strengthen communities that are suffering from multiple deprivation (which is why there should be strong links between fair access and community engagement). The following recommendations are made in this spirit:

Recommendations for targets and measures

National targets for the whole HE sector, and minimum targets for individual institutions, should be retained in order to maintain the momentum of progress towards fair access, and to provide transparency and accountability.

SIMD should be retained as the core metric for measuring progress towards fair access, at both whole sector and institutional levels.

It should be supplemented by incorporating a small number of individual-level indicators of disadvantage into fair access policy and monitoring of progress.

The two individual-level indicators that should be incorporated are registration for free school meals, at any time during secondary education, and care experience. Efforts should be made to improve the reliability and timely availability of these indicators. Those with these individual level indicators should not count towards the existing SIMD targets but the Scottish Government should reflect these other groups in future targets.

The case for treating race and ethnicity in a similar way should be actively explored.

The development of individual-level indicators better suited for identifying disadvantage among older applicants should be a priority.

Other indicators such as school environment, socio-economic status of parents and their experience of higher education are not suitable for incorporation in national targets. But this should not downgrade their use by individual universities.

The use of 'mixed' metrics, such as MEM, should be explored, although at this stage they are likely to be more useful as research tools than as policy instruments.

Universities should avoid the proliferation of 'markers' of deprivation, to avoid the over-identification of potentially disadvantaged applicants, and to promote greater transparency (and transferability between institutions).