Hepi TTWI-Web

Tackling Wicked Issues:
Prestige and Employment

Outcomes in the Teaching
Excellence Framework
Paul Blackmore, Richard Blackwell
and Martin Edmondson
Occasional Paper 14
About the Authors
Paul Blackmore is Professor of Higher Education in the
International Centre for University Policy Research, at the
Policy Institute, King’s College London. He has led academic
development and research centres at the Universities of
Warwick, Coventry and King’s. A HEFCE-funded study of 26
institutions worldwide led to a major report and a 2012 book
Strategic Curriculum Change in Universities: Global Trends.
In 2016, he published Prestige in Academic Life: Excellence
and exclusion, which explores difference and motivation
in academic life, and a study funded by the Leadership
Foundation for Higher Education entitled The Role of Prestige in
UK Universities: Vice-Chancellors’ Perspectives.
Richard Blackwell is an Emeritus Professor of Southampton

Solent University, where he was Deputy Vice-Chancellor. At
Southampton Solent University, he led the employability and
enterprise strategy and, later, on student satisfaction. Since
2015, he has been working as an independent consultant
and executive interim manager, primarily on employability,
teaching excellence and the higher skills agenda. Previously,
he worked for the Higher Education Funding Council for
England, the Higher Education Academy and the universities
of Coventry, Nottingham and Oxford.
Martin Edmondson is the Chief Executive of Gradcore, and has

more than 15 years’ experience in graduate recruitment and
employability. He worked for a range of large organisations and
universities before running the Graduates Yorkshire project,
which he then spun out into the social enterprise that is now
www.hepi.ac.uk 1
Gradcore. Martin has led many graduate recruitment and
employability projects, including global graduate schemes for
large employers and delivering employability performance
improvement in universities. He has spoken on graduate
recruitment and employability at conferences around the
world and chairs the UK and European Graduate employment
conference.
2 Tackling Wicked Issues

Contents
Why research trumps teaching and what

can be done about it
by Paul Blackmore 4
hy employment outcomes are important

W
and how they should be measured in future
by Richard Blackwell and Martin Edmondson 39
www.hepi.ac.uk 3
Why research trumps teaching and
what can be done about it
Paul Blackmore
Executive summary
Research has higher status than teaching in academic life

despite many attempts to achieve change. Research has grown
in importance over thirty years and the Research Excellence
Framework (REF) continues to drive priorities in research-
intensive institutions and many teaching-led ones too. Current
Government proposals to recognise and reward teaching
excellence are well intentioned but not likely to be as successful
as hoped. Emphasising differences between teaching and
research and pitting them against each other means that
teaching will lose.
The problem is deep-seated and related to prestige, defined

here as being:
•• relatively scarce;
•• hard to measure;
•• slow to gain or lose; and
•• often decided on by insiders.
Excellence in research attracts prestige, but excellence in

teaching does not. Prestige economies are groups that have
key values in common. They can occur at all organisational
levels. Disciplines can be prestige economies too, transcending

institutions. Individuals know what is valued in their peer group
for a successful career. The success of research excellence
competitions shows that prestige engineering works.
Providing more money for excellent teaching is not the answer.

This will not change many underlying values. Much of the new
funding will not benefit students but will support research, just
as much teaching funding does now.
The Government proposes an even stronger separation

between research and teaching, through an Office for Students
(OfS) and a UK Council for Research and Innovation, with a
Teaching Excellence Framework (TEF) sitting alongside the REF.
Each ‘side’ will proceed with scant reference to the other.
Rewarding excellent teaching requires metrics but prestige

prevents metrics working, for technical and cultural reasons.
Measuring learning gain across institutions cannot be done
meaningfully. Other proxies are too conjectural to be reliable.
In any case, the idea rests on a human capital view about the
value of skills and return on a degree. We often assume that
higher earnings come from increased skills. However, social
and cultural value is signalled by a degree and where it comes
from. Parents and students understand how prestige works.
Claims that better and faster information will drive out prestige
are optimistic. Positional goods have existed for centuries and
will continue to do so. Without reliable and objective measures,
prestige always wins. Even with such measures, signalling
continues. There is a relationship between prestige and
reputation, but not a strong one. Over a long period, prolonged
bad reputational news may dent prestige, but it does take a
www.hepi.ac.uk 5
long time. An institution may be criticised for practices that
have no impact, or a reverse impact, on prestige. For example,
not being inclusive leads to exclusiveness, which can be a
prestige factor.
Prestigious research-intensive institutions will be less affected

by the new current proposals and will do just enough to
trigger increased funding, if it is worth having. Teaching-led
institutions, with private provider competition, may find their
reputation harder to protect.
A more differentiated higher education system tends to be more

expensive because it encourages prestige seeking through
spending on attention-grabbing projects, as can be seen in the
US higher education system and the UK independent school
sector. So quality-related funding will not have the desired
effect, unless extra funding can be ring-fenced for teaching,
which is in practice close to impossible.
An alternative way forward is to link research and teaching

closely together at all levels. The linkage may look very different
because institutions are so varied in their missions. A simplistic
opposition between ‘blue skies’ research and teaching as
transmission must be set aside, with research and teaching
recognised as forms of complex learning. This requires students
to take a more active and engaged part in their learning. Such
an approach is a more constructive response to recent changes
than complaining about consumerisation.
Impact offers another place for connection. The REF emphasises

a requirement for impact, as does the employability agenda
for teaching. A broader view could be taken, with universities

demonstrating their research informs their teaching and
that both have a positive effect beyond the university. The
inevitable conflict with concerns for knowledge ‘for its own
sake’ can be met by emphasising the social and intellectual as
well as economic impact, thereby dealing with a long-standing,
unhelpful and prestige-related tension in universities.
What might this mean in practice? Public funding for research

should require an explanation of the teaching benefit. Funding
for higher education teaching should require evidence of
students learning in research-like ways. At an individual
level, recognition and reward could be centred on a view of
all academic work being about complex learning. This would
enable everyone to make connections between teaching and
research, not just active researchers. Probation and promotion
should be designed in ways that pay attention to how prestige
works. Teaching-only roles, reconfigured more broadly, could
stay in a single academic structure but the greatest reward
would go to those who show they unite teaching and research
in their work.
Teaching must be evaluated differently. The National Student

Survey (NSS) treats students as customers and implies
teaching is telling. Engagement-focused surveys instead ask
what students do and can reveal whether they are learning in
research-like ways.
It cannot all be done at an institutional level. Disciplinary and

professional groups are important parts of the way prestige
operates and should be encouraged to explore ways in which
the climate can be changed.
www.hepi.ac.uk 7
The importance attached to global rankings, success in which
increasingly requires a strong presence in both research and
teaching, presents an opportunity for institutions to ensure
that teaching and research are treated and projected as part of
the same enterprise.
Recent changes to the UK Government mean that

responsibility for higher education policymaking is once more
in the Department for Education (DfE), while responsibility for
research sits in the new Department for Business, Energy and
Industrial Strategy. Even though the two areas have a shared
Government Minister, it is now doubly important to ensure
teaching and research are not driven apart further.

Why research trumps teaching and
what can be done about it
Paul Blackmore
The problem
Research dominates academic life at the expense of teaching.

There have been many attempts to achieve a better balance.
The proposal for a Teaching Excellence Framework (TEF) in
the white paper Success as a Knowledge Economy: Teaching
Excellence, Social Mobility and Student Choice (BIS, 2016) is the
current example. The Higher Education Academy (HEA) and its
predecessor the Institute for Learning and Teaching have aimed
to raise the status of teaching. So have many funding initiatives,
particularly the Teaching Quality Enhancement Fund, which
from 1999 gave universities additional money in return for
teaching improvements. The publication of comparative data
about perceptions of teaching quality, through for example
the National Student Survey (NSS), is a spur to competition in
teaching at a national level. Many universities have changed
promotion criteria to recognise teaching excellence and
introduced new categorisations of staff, with the aim of
providing parallel career routes with parity of esteem.
While some may argue the problem has been solved, a look
across the sector confirms this is still an important issue, and
clearly the Government believes a problem remains. In the
light of earlier initiatives, one must ask whether current policy
proposals are likely to do a better job of raising the status of
teaching or whether they will go the same way. This paper
argues that some fundamental aspects of the problem have
www.hepi.ac.uk 9
not been fully recognised and the proposals may inadvertently
make things worse.
At the heart of the tension between teaching and research

is the influence of prestige seeking in universities. A series of
Research Assessment Exercises (RAEs) has helped research gain
dominance over thirty years, by publicly valuing excellence in
research, in ways that have powerfully shaped universities. So
research-intensive institutions have become even more focused
on their research ambitions. Even teaching-led institutions are
strongly influenced by a cultural bias within higher education
to value research more highly. The Research Excellence
Framework (REF) remains, for at least another round, with the
Government keen to introduce more metrics if it can (Grant,
2016). Thus the REF will continue to be a powerful motivator
in academic life. Global league tables that principally measure
research prestige accentuate the problem.
A TEF may seem an obvious counterbalance to the REF, bringing

metrics to bear to encourage excellent teaching. It remains
unclear whether a TEF can or will achieve its aim of raising the
status of teaching. The task is more difficult and deep-seated
than it at first appears, as the last thirty years of efforts to value
teaching have shown (Gibbs, 2016).
I argue here that the Government’s proposals have not grasped

the size of the problem and have not proposed solutions that
are likely to work. Indeed the white paper proposal to reduce
the ten ‘arms-length’ teaching and research bodies to two
emphasises the separation further, despite the intentions
shown in the brief mentions of the ways research can enhance
teaching (paragraph 44). Excellence in research attracts
prestige, but excellence in teaching does not. The more we
see research and teaching as separate, the more teaching will
lose out in an almost universal desire to achieve prestige. The
impact of global rankings makes it more important than ever
to change attitudes and behaviours.
How prestige and reputation are different
Recent work on academic motivation (Blackmore, 2016) has

focused on two key ideas. Firstly, the generation of prestige and
reputation are two different aims, with differing motivational
consequences. Prestige is relatively scarce:
•• not all can have it;
•• it is hard to measure;
•• it is slow to gain and to lose; and
•• it is often decided by those within a group.
The award of a Nobel Prize to a staff member confers prestige

on an institution: there are few such prizes; its financial value is
not possible to measure; and it is an academically-driven award.
Yet the prestige is immense. Reputation on the other hand can
be held by all, provided all are proficient. It is measurable, may
be gained and lost more quickly and is often confirmed by
external stakeholders (Brewer, Gates and Goldman, 2002). So,
for example, returning students’ assignments in a short time
can add to a university’s reputation; however, every institution
is able to do that if it tries hard enough and it need not take a
long time to achieve it.
www.hepi.ac.uk 11
Secondly, prestige occurs among groups with key values
in common that lead them to accord prestige to particular
phenomena. Anthropologists label this a ‘prestige economy’.
An academic department might be a prestige economy, and
so might a faculty, or an institution or a national system. So too
might a discipline, transcending institutional boundaries.
The ‘script’ that an individual may hold is a powerful shaper of

motivation, listing the achievements that he or she believes
are needed to advance in a career. There is evidence that
these scripts tend to vary by academic discipline and that they
typically focus on research rather than teaching achievements
(Blackmore and Kandiko, 2011). This is consistent with the idea
that prestige trumps reputation. Unless teaching can come
to be seen as prestigious, it will always be of secondary
importance, irrespective of financial incentives.
The idea of prestige economies at different organisational levels

explains why many institutions appear to be unresponsive to
strategic steering and why many innovations fail. There may be
a mismatch between what is valued within prestige economies.
Although universities can send strong signals about what
is valued at an institutional level, academic identity tends to
draw its strength elsewhere, often through disciplinary or
professional affiliation. Inspection and funding changes that
occur at an institutional level may not be noticed much lower
down within the organisation, unless the change is so extreme
as to cause major discomfort. Any change in the respective
valuations of activities has to be owned at a local level.
Anyone who doubts the need to pay attention to patterns of

prestige might consider the powerful effects of the RAE / REF
since 1986 in the UK. This mechanism has provided a way in
which scarce academic prestige can be generated, recognised
and rewarded. It has been outstandingly successful in
confirming the importance of research, often at the expense
of the status of other activities, principally teaching. It has
reinforced the prestigious position of research-intensive
institutions in relation to others and has spurred on ever
greater concentration of research. It has influenced behaviours
in institutions across the whole sector, even those whose main
activity is teaching. Prestige engineering works.
How government can make it worse
In a competition for academic attention, research will generally

beat teaching. Two widely advocated remedies do not work.
Firstly, tinkering with financial incentives leads to game playing
and does not change underlying values. Secondly, treating
research and teaching separately and then trying to boost the
status of teaching sets up a contest in which teaching loses.
A financial incentive assumes that the difference between the

motivation to research and to teach is largely one of resources.
At present, money follows excellence in research, through the
allocation process of the REF and Research Councils’ grant
awards, but excellence in teaching is not so rewarded. This lies
behind the white paper’s proposals to incentivise universities
to provide better teaching by enabling them to charge higher
fees, in return for demonstrating excellence. This may encourage
institutions to pay extra attention to teaching quality to trigger
the increased payment, which would be an advance.
However, such behavioural changes, if they happen, are part of
www.hepi.ac.uk 13
a larger picture. In many institutions it is arguable whether an
increase in teaching income will have much impact on what is
valued and given most attention at departmental and individual
level. Since institutions can do as they wish with their income,
it is plausible that much of the extra funding will not directly
benefit students, as Graham Gibbs has also suggested (2016).
Any new funding could be diverted into support for research.
This is happening already. Current expansion in student
numbers among Russell Group institutions is not driven solely
by a philanthropic wish to educate more people or a progressive
desire to improve social mobility but instead aims to generate
income to spend on other activities, principally research. So
although money is an extrinsic motivator in academic life as
elsewhere, finance is not the principal answer.
The strength of the underlying value system can be judged by

the way that many universities choose to spend their income.
Cross-subsidy of research by teaching happens widely in
the UK and elsewhere, with little transparency in the use of
resources by universities or accountability for it. Under the UK’s
system of dual funding for research, many institutions receive
a block grant for research, its size depending on the quality
of the university’s submission to the REF. Once awarded, the
block grant is set for the period. Other funding comes through
winning research grants and contracts. However, many funders
do not pay the full economic cost of research. The more research
of this kind that universities do, the more the deficit grows.
Thus institutions undertaking research below its real cost often
subsidise their overall research effort from teaching funds.
Additional teaching funding could enable further below-cost
research to be undertaken, generating more prestige.

A study of the Australian system (Norton and Cherastidtham,
2015) suggests A$2 billion of research is funded by teaching,
20 per cent of the total, with one dollar in five coming from
the teaching surplus. The study suggests that activity-based
costing is required, citing the UK Transparent Approach to
Costing (TRAC) system’s ability to separate teaching and
research activities, while noting the difficulties in doing so.
Cross-subsidy of research by teaching is vulnerable to criticism,

especially where one discipline’s teaching subsidises another
discipline’s research. It is remarkable that the practice has
not attracted more attention in the UK. A recent survey of
student opinion (Neves and Hillman, 2016) finds that student
perceptions of value for money have fallen, especially in
England with its £9,000 fees, and that 86 per cent of students
do not want fee increases to be allowed in return for excellent
teaching, while only 18 per cent of students feel they have
enough information about how fee income is spent. The survey
also offers evidence about what many students might think if
they were better informed about how their fees are used. Only
26 per cent of students consider it very important that teachers
should be currently active researchers. Even where being an
active researcher was most often demonstrated, particularly
in Russell Group institutions, students did not think it any
more important. Many more students felt improving subject
knowledge regularly and receiving training in how to teach are
very important. The student view of the world is very different
from that of many academic staff.
In summary, the financial relationship between teaching and

research is complex and increasing the return on teaching will
not necessarily have the effect that the Government might
www.hepi.ac.uk 15
desire. It may instead increase the amount spent on research
and further expose universities to criticism of the ways they use
teaching resources. It may be that prestige factors provide some
insulation, with students recognising that high-profile research
raises the status of an institution and therefore increases the
value of the degrees it awards. However, this attitude cannot be
relied on to continue and is in any case not ethically desirable.
Nor is it an efficient way of funding, from the point of view of
teaching. There are more direct ways of supporting teaching
improvement than by subsidising research which may or may
not make a difference to teaching. Universities wishing to
spend money on research would be wise in future to be able
to show its teaching benefit, or risk severe criticism from more
discerning and better informed students.
Turning to the second problem, both central government and

sometimes even universities set research against teaching.
The Higher Education Funding Council for England (HEFCE)
is involved in the funding of both teaching and research, but
in all other respects they tend to be conceived of as largely
separate enterprises that happen to be carried out in the same
institutions. An absolute division is routinely assumed, quite
understandably in that each is a complex activity in its own
right. The white paper has proposed to increase this divide.
The Office for Students (OfS) and the UK Council for Research
and Innovation will be separate, engaging with institutions
independently of each other. The TEF will stand on one side
and the REF on the other. The white paper’s description of the
role of UK Research and Innovation contains no reference to
teaching (box 3.3).

The Government has recently returned higher education to
the Department for Education (DfE), while research resides in
the Department for Business, Energy and Industrial Strategy.
For now, there is a shared minister, but it nonetheless means
teaching and research are divided in Whitehall too.
Unless the case for research improving students’ learning is

pressed with vigour, we will miss an opportunity to connect
research and teaching, despite this being the only basis on
which the status of teaching can be improved.
What won’t work
Proposals to reward excellent teaching will be frustrated by the

working of prestige, partly because awarding more money for
better teaching requires a robust means of measuring teaching
quality. However, metrics cannot contend with prestige, for
both technical and cultural reasons. Prestige will always be
present, filling much of the inevitable gap between what is
valued and what can be measured.
The Government’s wish to achieve an objective measure of

teaching quality may not succeed because it is not possible to
do in a meaningful and workable way. The ‘Holy Grail’ is a means
of showing how students have progressed from their level at
entry. There are major difficulties in deciding what to measure,
and in undertaking the measurement, especially in ways that
have some meaning across the broad range of disciplinary
and professional curricula. Rhetoric is running ahead of the
capacity to deliver. Schleicher (2016) lists many of the major
difficulties in achieving a measure, which he believes must deal
with individuals’ learning gains, taking into account student
www.hepi.ac.uk 17
and institutional contexts in order to compare like-for-like. His
assertion that it can be done, with the aid of technology and
through co-operation, seems optimistic and at odds with the
substantial methodological challenges he notes.
Because of these difficulties, some argue the environment

provided for learning should be measured because particular
features of a student’s experience are likely to lead to effective
learning (Gibbs, 2016). However, this may not measure the
learning itself. It is simplistic in its assumptions about the
relationship between students’ knowledge and abilities at the
start and end of university and of the impact of the experiences
in between. It is assumed, for example, that it is the institutions’
actions that make the difference. Genuinely robust measures
of this kind, that could justify differential funding, are unlikely
to be found. Paying attention to processes of learning may
reduce the tension between teaching and research, but it is
problematic to rest funding decisions on this highly context-
specific field.
Cultural aspects of the use of metrics are even more difficult.

This is a complex and politically-loaded field. Prestige is central
to the problem. Two kinds of return on higher education
are often cited. A human capital view focuses on the value
of increased skills and the economic return on a degree. A
signalling explanation refers to the social and cultural value
of having a degree from a particular kind of institution. The
latter is not much affected by evidence about teaching quality.
Students and parents know about signalling and understand
how prestige works just as much as universities do.

Some argue that prestigious institutions will in future have a
harder time because better and faster information will drive
prestige-based decision making out. A recent Washington Post
article (Kinsley, 2016) claims that the internet is disrupting the
prestige of higher education. It draws an analogy with the
newspaper industry, which became a global market and had
its business model radically affected by the internet. The article
suggests that as free lectures become more widely available
and interactive, it will be more obvious that parents and
students pay over the odds for an Ivy League education, the
price of which is kept high by its scarcity. So research-intensive
institutions will stand exposed as profiting from having the
ability to give out scarce positional goods and to charge
handsomely for it.
However, we have long known this happens. The complaint was

made by Adam Smith almost 250 years ago about the University
of Oxford (Smith, 2014), and has been used against prestigious
US institutions in more recent times (Boyer Commission, 1998;
Archibald and Feldman, 2011). It applies not only to universities
but to luxury goods that are bought and valued because they
are scarce and exclusive. Parents and students will stop paying
inflated degree prices only if the signalling effect decreases
to the extent that it is not worth paying the money. However,
the way prestige works means this is unlikely to happen. We
know the extent to which a good is desired is often related
to its scarcity. Scarce goods are always in demand if they are
associated with status, and this will remain so for as long as
status is sought in a society. Only when many people have
something does interest move to another place where there
can be differentiation.
www.hepi.ac.uk 19
We often ignore the effects of signalling. A study sponsored
by the Department for Business, Innovation and Skills (BIS,
2013) suggests the Department routinely assumed that the
higher earnings of graduates are explained by the skills they
gain during their degree, ignoring signalling effects. It is not
surprising that this assumption is made. Equally, therefore, the
signalling effect of degrees from particular institutions is likely
to be overlooked. Although the basis on which graduates and
employers make decisions is a complex one, some institutions
clearly have more powerful signalling effects than others.
Many of the existing and easy to administer measures, such as
the job destinations of graduates, may to some extent reflect
the signalling value of particular institutions. They are prestige
rather than reputation measures.
So in the absence of reliable objective measures, prestige will

continue to influence decisions of many kinds. This may seem
to be a spur to devise ever better metrics, to squeeze irrational
prestige-related beliefs out of the system, but this is to attempt
the impossible. Even if feasible measures could be devised, they
would not necessarily guide many students’ choices. It is often
supposed that potential students make institution choices on
the basis of data provided for the purpose. However, many
make an all-round judgement irrespective of what statistics
might say, about what would be their kind of university (Ball et
al, 2002; Croxford and Raffe, 2014), and the signalling benefit
that would result. For such students, a degree from a particular
kind of university is a positional good.
While the internet certainly does overturn business models, and

can do so with remarkable speed by providing comparative data

worldwide, prestige is a powerful countervailing force. Indeed,
slowness to change in the face of current conditions can itself
lend prestige by signalling institutional self-confidence and
the taking of a long-term view.
There is a relationship between prestige and reputation but

it is indirect. For example, statistics can harm prestige. No
institution can be unaffected by a large number of bad ratings
in league tables. However, if a prestigious institution does less
well in one indicator and suffers an immediate reputational hit
in that area, its prestige may not be much affected, at least in
the short-term. Prestige tends to rise and fall relatively slowly,
unlike reputation which is measured against more concrete
items.
The extent to which an adverse indicator will eventually feed

through into loss of prestige probably has to do with the
indicator itself – whether it is one that is valued within the
prestige economy in question. Thus, a metric that shows that a
research-intensive institution is not inclusive in its intake may
have little impact on the prestige of the institution, especially
if its prestige rests partly on its exclusivity. If a research-
intensive institution is shown to be less successful in terms
of its graduates gaining high-status jobs, that may well erode
prestige more quickly, since it damages the interests of parents
and students who buy into the institution’s prestige with the
expectation of that benefit. However, signalling effects ensure
this never happens.
Whether or not the current quest for robust measures of

teaching excellence is successful, the signalling element will
www.hepi.ac.uk 21
remain and institutions of different kinds will therefore face
increasingly divergent environments. Current proposals mean
the more prestigious institutions may be able to increase their
fees without paying much attention to the quality of their
teaching, as has been suggested happens in the US system
(Gibbs, 2016). Research-intensive institutions, with their many
inbuilt advantages, will continue to score well, as they did when
Teaching Quality Assessment (TQA) was introduced in 1993 to
evaluate teaching at a discipline level. When they do not score
well it will not matter much, provided they reach the threshold
for enhanced funding. The entry of more private providers
will provide competition for reputation-led institutions, but
those universities that are prestige-based, buttressed by the
advantages of positioning and pre-eminence through time, will
be largely protected. Indeed, they may prefer to concentrate
on increasing their income through further recruitment of high
fee-paying and prestige-motivated overseas students, than to
jump through yet more UK government hoops for what look at
present to be rather modest returns.
A more differentiated higher education system offers more

room for prestige to work. Prestige is expensive for a system.
An unwelcome consequence of a differentiated system for
government, parents and students is that the cost of higher
education in the top institutions tends to increase rapidly,
as has been the case in US higher education and in the UK’s
independent school sector. A higher education arms race in
the US has been described (Archibald and Feldman, 2011),
adding to other cost inflation pressures. Thus experience in a
system which resembles England, suggests that the increased
income from teaching may be used for another purpose

besides subsidising research. Iconic and attention-grabbing
developments will be funded to generate prestige.
In summary, it is doubtful that the UK Government will drive

improvements in the quality of teaching by making quality-
related funding available, even if a defensible system can be
devised. Beyond what needs to be done to secure funding,
teaching will not receive the benefit of extra income, unless
it is ring-fenced so that it has to be spent on the support of
teaching, as was the practice in Germany before fees were
abolished (Hillman, 2015). This requirement would be difficult
to implement in practice as it would require the specification of
what counts. Moreover, it could be argued that research informs
teaching, so money spent on the former boosts the latter.
Possible ways forward
If prestige engineering works for research, helping to elevate its

importance through successive rounds of research assessment,
then presumably it can help with teaching too, although the
task is much harder. If one makes use of the distinction between
prestige and reputation and works with the idea of a prestige
economy, some possible ways forward can be identified.
Government should plan for research and teaching together
Higher education policymaking has bounced around Whitehall

and been subject to rapid changes in personnel, as Nick
Hillman (2016) has noted. Darian (2016) points to the need for
organisational stability when administering ratings systems. Yet
central government is largely responsible for pitting teaching
against research at a policy level and for ensuring that teaching
loses the unnecessary battle.
www.hepi.ac.uk 23
While there is recognition of the need to enhance the status of
teaching, some of the government’s actions work against this.
The Fulfilling Our Potential green paper (BIS, 2015) mentioned
research only to say what it would not be discussing. It
ignored research-teaching links and the impact on teaching
of increased efforts in research (Bekhradnia, 2016). Research
was left for Paul Nurse to consider (2015), but the Nurse report
ignored teaching.
The subsequent white paper notes that ‘the value of research-

led teaching’ is one of the areas in which an institution can claim
credit, through the qualitative aspect of the TEF (paragraph
44), but this will not affect specific funding. While recognising
the link between research and teaching is welcome, it is
insufficient, when government systems for managing and
evaluating research and teaching are to be entirely separate
and dealt with by unconnected bodies.
The situation will improve if research and teaching are purposely

linked at all levels, from macro-policy through to individual
members of staff. This is an immensely difficult and long-term
task, but will be easier to tackle if the working of prestige is
taken into account rather than ignored. Such a linkage will look
different from one institution to another because the UK higher
education sector is diverse, and increasingly so. The amount of
basic research and undergraduate teaching undertaken will
continue to vary.
Uniting research and teaching
Teaching will gain if it is believed that it shares characteristics

with research. Redefinition may help to emphasise what

teaching and research share rather than what apparently
separates them. To achieve this, it is necessary to broaden
the ways in which research and teaching are often defined.
Blue-skies discovery research is traditionally contrasted with
teaching as the transmission of information, but this is crude
and unhelpful. Ernest Boyer (1990) described four scholarships:
of discovery; integration; application; and teaching. In this way,
the middle ground of bringing research together and applying
it comes into play. What unites all of these is that they are forms
of complex learning. There is of course nothing at all new in
such thinking. What would be novel would be if it informed UK
national policy.
To move forward requires two changes. Again, neither

of these is new, but both need to be worked on to bring
teaching and research together. The first requires a shift from
a transmissive form of teaching that emphasises the lecturer
as the giver of information and the learner as the passive
recipient of it. Instead, students must be encouraged to take
a more active and engaged part in their learning. Staff and
students are all researchers, with a recognition that the skills
of learning are much the same at all levels in a university. There
is a large literature on this and a vast number of case studies.
Undergraduate research is growing in UK universities, with a
British Conference of Undergraduate Research (www.bcur.org)
and Reinvention, an international journal.
The second area for action is also signalled by Boyer, in his

inclusion of the scholarship of application. Impact is a current
concern in both research and teaching. The REF has aimed to
reward evidence of impact as well as excellence in research,
partially redefining excellence to include impact. Teaching
www.hepi.ac.uk 25
has its impact agenda too, which is currently expressed in
the employability agenda of universities, explored in the
accompanying paper. The two could be linked in imaginative
ways, if universities were to show their research expertise
flowed into their teaching and if both together could be
shown to have a positive effect beyond the university. Such
an approach would engage with one of the most entrenched
aspects of prestige in universities: that of knowledge for its
own sake. In teaching, skills gained that can be transferred to
the workplace have for many years, and with many variations
in terminology, been presented as if they were entirely
antithetical to the learning of subject knowledge for its own
sake. What has to be achieved is the imaginative reworking
of this largely false opposition, across all programmes in all
institutions (Phoenix, 2016). At present, active learning and
capability-focused curricula exist despite rather than because
of the ways in which prestige works. In the longer term, the
consistent support for and recognition of curricula of this kind
could reshape perceptions about the nature of teaching and
research, for the good of both.
A major prestige problem that can be predicted in such a

change is that it could be seen as acting in opposition to the
idea of learning for its own sake. If it is emphasised that impact
could refer to economic, social and intellectual impact, and
if these terms are interpreted with imagination, there would
be less ground for such objections. Indeed, it is hard to argue
for either teaching or research that makes no difference to
anybody. Significant resistance to moves of this kind will be
generated if government neglects this aspect in favour of
an exclusive emphasis on employability, narrowly defined.

A member of academic staff whose self-perception is of
being principally a researcher will be unlikely to engage with
employment-focused curricula unless attracting connections
can be made. Prestige engineering has to start from where
people are and what they value, and then connect it to other
desired ends.
The benefits of research-rich teaching have been argued

for many years, famously by Wilhelm von Humboldt who on
founding the University of Berlin in 1810 argued that higher
learning united student and teacher in scholarship. The link
between research and teaching quality is not automatic
(Hattie and Marsh, 1996) but there are major benefits if it is
designed into systems (Jenkins and Healey, 2005). These are
mere pedagogic pipe dreams, as they have often been for
many years, unless they are backed up by concrete action. This
requires government and universities to think afresh about
funding, quality assurance and management of both research
and teaching in higher education institutions. Public funding
for research should require an explanation of the pedagogic
benefit of the research. Public funding for teaching should
require evidence that students are learning in research-like
ways and are benefiting from research funding. In these
ways the interconnection of research and teaching would be
reinforced in institutions. If additional funding is to be given,
then it would be on the basis of a proven interconnection.
Roles, recognition and reward
The nature of the connection between research and teaching

is a political issue. If it is seen as requiring a member of faculty
to pass on the fruits of his or her own research, this model
www.hepi.ac.uk 27
will only work, and then patchily, in some research-intensive
institutions. Such a model risks seeming out of touch, in an era
where many academic roles are rapidly becoming unbundled
and many university teachers are not active cutting-edge
researchers. If, however, the link is one where both research
and teaching are seen as similar processes of learning, then
all institutions and members of faculty can participate. Being
an academic then means being an expert in complex learning,
not just in what is to be learnt. This implies a high level of
scholarship from those who teach, both in terms of their own
relationship with what they teach and their ability to enable
students to learn in research-like ways. Such a requirement
across a rapidly diversifying sector would be a helpful and
timely quality safeguard.
Since teaching is an activity that when done well tends to

generate reputation rather than prestige, institutions must
have strong procedures in place to deal with probation and
promotion, to overcome the bias towards research attainment
that has existed for so long. Recognition of social realities would
be helpful. Entirely separate teaching-only roles will never gain
parity of esteem with research-led ones. That is not to argue
against rewards for those who are excellent in teaching and its
leadership, for they are very helpful signals. However, all forms
of academic work should remain within the main academic
structure. It should be possible for the constituent parts of an
academic role to vary through a career with the least possible
loss of esteem. This requires a unified system and not one of
academic apartheid. In any case, this need not be set out as
teaching versus research. The greatest recognition and reward
could be reserved for those that demonstrate that in their

practice there is a seamless union. If the nexus is conceived
as one largely around learning as a research-like process, all
those who teach can be considered, whether they are active
researchers or not.
Changes in how teaching is evaluated
Just as the inclusion of impact in the REF has led to a greater

focus on what difference research makes, so the way teaching
is thought of will be influenced by the way it is evaluated.
The National Student Survey is central to Government plans
to evaluate teaching. For some years, the NSS, open to every
student completing an undergraduate degree, has attempted
to measure student satisfaction. Unfortunately, it is not at
the moment in a form that is helpful for linking research and
teaching. It has been criticised for encouraging a consumer
culture, by focusing on what universities provide more than
on what students actually do and learn. The implied model
of teaching is that it is largely a matter of telling things to
students. It asks whether staff are enthusiastic, are good at
explaining and have made the subject interesting. If instead
students’ perceptions of the pedagogic quality of what they
are experiencing were evaluated, based on a reputable model
of higher-level learning, then academic’s attention would be
drawn to think more fundamentally about what and how they
teach.
Signs of improvement can be seen in the UK HEA’s piloting

of an alternative survey focusing on students’ engagement
in learning rather than on their satisfaction with what has
been provided by the institution (HEA, 2015). It asks whether
students have: learnt about methods of analysis and research
www.hepi.ac.uk 29
in their subject; learnt about the outcomes of current research;
formulated and explored their own questions; and worked
on a research project. Reform would introduce a full student
engagement survey, along the lines of the USA’s National Survey
of Student Engagement (NSSE) or Australia’s Australasian
Survey of Student Engagement (AUSSE).
The role of disciplines and professions
Claims about the benefits of links between research and

teaching are often made at an institutional level, and
are frequently hubris. A local sense of identity remains a
highly significant socio-cultural aspect of higher education
institutions. Both teaching and research practices vary
significantly across disciplines in an institution. Shifting the
relative importance of research and teaching in universities
will require sustained attention to what it is that academic staff
value, in the communities holding those values.
Reconstructing relationships between research and teaching

has to be done at a disciplinary and professional level, partly
because practice in both teaching and research varies locally,
reflecting the disparate forms of knowledge that are involved.
So, as an example, many members of faculty in the sciences
would argue that their own research is so focused and abstract
that it cannot be usefully taught at undergraduate level, while
humanities and social science members of faculty frequently
thank their students for helping to provide new insights.
An activity such as enquiry-based learning may look rather
different from one discipline to another, a fact hidden by the
use of such a generic term. The quality assurance changes
suggested above would require institutions to show the nature
of their teaching in those terms at a disciplinary level.
An appreciation of the socialisation of academic staff is a helpful

starting point for trying to alter behaviour by changing patterns
of motivation. It has often been said that academic staff have
a greater allegiance to their discipline than to their institution
(Jenkins, 1996). While such allegiances are lessening, including
with the growth of interdisciplinary working and centralised
student admissions, they remain important features of higher
education. The relative success of the HEA’s former discipline-
based subject centres in communicating with academic staff
about teaching, in comparison with more generic approaches,
shows disciplines matter. Messages from within a discipline
receive a more positive reception than those set at system or
institutional levels.
Existing disciplinary and professional associations could be

encouraged to come together to work on some of these identity
and motivational issues, so that these universal themes can be
tackled locally.
Conclusion
Re-engineering prestige by emphasising that research and

teaching are two sides of the same coin offers a better approach
than attempting to buttress teaching so that it can stand against
research. In such a contest, teaching will always lose. Change
needs to be underpinned by a national level policy framework
that regards research and teaching as being closely connected,
for government too often sends two sets of conflicting
instructions to the same institutions and implies that it values
some institutions more than others. The challenge has become
www.hepi.ac.uk 31
greater as a result of the latest machinery of government
changes. Changing attitudes to teaching is very difficult. A
rationalist approach, that assumes top-down structural changes
and financial mechanisms will produce improved practice,
is unlikely to be successful. Alterations in policy and practice
both nationally and locally require attention to be paid to the
motivational consequences of the changes. There needs to be
an analysis of the likely impact on what is valued and how that
will affect roles, relationships and behaviours.
In summary, at a national level we might consider:
•• w
hether a single funding body for teaching and research
might be a better option than the proposals in the Higher
Education and Research Bill – alternatively, ensure in terms
of reference and elsewhere that each funding body proceeds
with full reference to the other;
•• w
hether institutions should be asked to provide a research
and teaching strategy for quality-related funding, showing
how each informs and supports the other;
•• the inclusion of a teaching impact statement as part of all

research applications; and
•• reflecting a research-related conception of teaching in the

NSS.
In universities we might consider:
•• strongly favouring promotion on basis of proven links

between research and teaching;

•• the development of flexible career paths that encourage
a creative blend of research and teaching, imaginatively
defined; and
•• e
ncouraging discipline-led discussions of what research and
teaching might mean and how it can be supported.
The growth of global rankings presents a threat and an

opportunity. A drive to gain and retain places in the world’s
top 100 universities can be seen in many countries, fuelling
Germany’s Excellence Initiative, Russia’s Project 5-100, and
similar ambitions across the world. While this competition
directly involves only a handful of institutions, many more
are influenced, as a particular type of university is constantly
celebrated. We may argue that league tables reward the wrong
things and prefer to generate league tables with different
values, a motivation behind the EU’s sponsorship of the
U-Multirank initiative. However, rankings are also subject to
being valued by their prestige. The iron law of prestige means
that those rankings that reward the best-funded and most
selective research-intensive institutions are the ones that really
count. Again, teaching suffers.
If the grip of a particular kind of ranking cannot be broken,

it can be used to help resolve one of the major tensions that
such rankings make worse. The rules of the game mean that
the winners must be excellent in research, but they must teach
as well, in order to be included. Countries such as Germany,
that fund much of their research in non-teaching institutions,
are disadvantaged. At a time when research and teaching
are in many ways being forced apart, at the prestigious and
influential end of the institutional spectrum it is the age of the
www.hepi.ac.uk 33
university that both researches and teaches. But they must
do both on a large scale in order to have enough resources to
compete. There are two possible responses. This could be very
bad for teaching, which might easily be shouldered out of the
way and starved of resources by prestige-seeking research,
and increasingly assigned to teaching-only staff employed on
short-term contracts. This seems the likeliest outcome without
intervention. However, we might achieve a step change in the
quality of teaching in universities if this moment is to be seized,
to ensure that research and teaching are treated as part of the
same enterprise. Universities cannot control global league
tables; however, they can make an intelligent response to
them, in the ways proposed here, to ensure that teaching gains
the status that it deserves and needs.

References
Archibald, R. and Feldman, D. (2011) Why does college cost so much?
Oxford University Press.
Ball, S., Davies, J., David, M. and Reay, D. (2002) ‘’Classification’ and
‘Judgement’: Social class and the ‘cognitive structures’ of choice of Higher
Education’, British Journal of Sociology of Education, 23 (1) 51-72.
Bekhradnia, B. (2016) Research, in Hillman, N. (ed.), Response to the higher

education green paper, Higher Education Policy Institute 27-36.
BIS (2013) The Impact of University Degrees on the Lifecycle of Earnings: some
further analysis, BIS research paper No 112. Accessed at https://www.gov.
uk/government/uploads/system/uploads/attachment_data/file/229498/
bis-13-899-the-impact-of-university-degrees-on-the-lifecycle-of-earnings-
further-analysis.pdf on 9 May 2016.
BIS (2015) Fulfilling Our Potential: Teaching Excellence, Social Mobility and
Student Choice. Department for Business, Innovation and Skills. Accessed at
https://www.gov.uk/government/uploads/system/uploads/attachment_
data/file/474227/BIS-15-623-fulfilling-our-potential-teaching-excellence-
social-mobility-and-student-choice.pdf on 10 May 2016.
BIS (2016) Success as a Knowledge Economy: Teaching Excellence, Social

Mobility and Student Choice. Department for Business, Innovation and
Skills. Accessed at https://www.gov.uk/government/uploads/system/
uploads/attachment_data/file/523546/bis-16-265-success-as-a-
knowledge-economy-web.pdf on 10 May 2016.
Blackmore, P. & Kandiko, C.B. (2011) ‘Interdisciplinarity within an academic

career’, Research in Post-Compulsory Education, 16 (1) 123-134.
Blackmore, P. (2016) Prestige in Academic Life: Excellence and exclusion.

Routledge.
Boyer Commission (1998) Reinventing Undergraduate Education: A Blueprint

for America’s Research Universities, State University of New York at Stony
Brook, NY for the Carnegie Foundation for the Advancement of Teaching.
www.hepi.ac.uk 35
Boyer, E. L. (1990) Scholarship Reconsidered: Priorities of the Professoriate.
Princeton, N.J: Carnegie Foundation for the Advancement of Teaching.
Brewer, D., Gates, S. and Goldman, C. (2002) In Pursuit of Prestige: Strategy and
Competition in US Higher Education. New Brunswick: Transaction Publishers.
Croxford, L. and Raffe, D. (2014) ‘The iron law of hierarchy? Institutional

differentiation in UK higher education’, Studies in Higher Education, 40 (9)
1625-1640.
Darian, L. (2016) Designing a Teaching Excellence Framework: Lessons from

other sectors, Higher Education Policy Institute.
Gibbs, G. (2016) Teaching, in Hillman, N. (ed.), Response to the higher

Grant, J. (2016) ‘Move to metrics may not bring significant savings to REF
bill’, Times Higher Education, 7 January 2016, accessed at https://www.
timeshighereducation.com/comment/research-excellence-framework-move-
to-metrics-may-not-bring-significant-savings-to-bill on 8 January 2016.
Hattie, J. and Marsh, H. (1996) ‘The Relationship Between Research and

Teaching: A Meta-Analysis’, Review of Educational Research 66, 507 - 542.
HEA (2015) UK Engagement Survey – UKES 2015. Accessed at https://www.

heacademy.ac.uk/sites/default/files/downloads/UKES_2015_items.pdf on
10 May 2016.
Hillman, N. (2015) Keeping up with the Germans?: A comparison of student

funding, internationalisation and research in UK and German universities,
Higher Education Policy Institute.
Hillman, N. (2016) Introduction, Hillman, N. (ed.), Response to the higher

education green paper, Higher Education Policy Institute.
Humboldt, W. von (1970) ‘On the spirit and organisational framework of

intellectual institutions in Berlin’, Minerva 8, 242 - 267.

Jenkins, A. (1996) ‘Discipline-based educational development’, International
Journal for Academic Development 1 (1) 50-62.
Jenkins, A. and Healy, M. (2005) Institutional strategies to link teaching and

research, Higher Education Academy.
Kinsley, M. (2016) ‘How the Internet will disrupt higher education’s

most valuable asset: Prestige’ Washington Post. Accessed at https://
www.washingtonpost.com/opinions/how-the-web-will-disrupt-higher-
educations-most-valuable-asset-prestige/2016/02/05/6bddc1ee-c91e-
11e5-a7b2-5a2f824b02c9_story.html on 1 March 2016.
Neves, J. and Hillman, N. (2016) The 2016 HEPI / HEA Student Academic
Experience Survey. Accessed at http://www.hepi.ac.uk/wp-content/
uploads/2016/06/Student-Academic-Experience-Survey-2016.pdf on 20
July 2016.
Norton, A. and Cherastidtham, I. (2015) The cash nexus: how teaching funds
research in Australian universities, Grattan Institute. Accessed at http://
grattan.edu.au/wp-content/uploads/2015/10/831-Cash-nexus-report.pdf
on 22 July 2016.
Nurse, P. (2015) Ensuring a successful UK research endeavour: A Review of

the UK Research Councils. Accessed at https://www.gov.uk/government/
uploads/system/uploads/attachment_data/file/478125/BIS-15-625-
ensuring-a-successful-UK-research-endeavour.pdf on 9 May 2016.
Phoenix, D. (2016) Making a Success of Employer Sponsored Education,

Higher Education Policy Institute. Accessed at http://www.hepi.ac.uk/
wp-content/uploads/2016/04/Making-a-success-of-Employer-Sponsored-
Education-Report-83.pdf on 10 May 2016.
Schleicher, A. (2016) Value-Added: How do you measure whether universities

are delivering for their students? HEPI 2015 Annual Lecture, Higher Education
Policy Institute.
Smith, A. (2014) The Wealth of Nations. Middletown, DE: Shine Classics.
www.hepi.ac.uk 37
Why employment outcomes are important and
how they should be measured in future
Richard Blackwell and Martin Edmondson
Executive Summary
Employment outcomes are important to stakeholders in higher

education and feature prominently in plans for the Teaching
Excellence Framework (TEF). The Government has proposed
developing a long-term earnings metric drawing on tax data
and is consulting on additional employment and occupational
measures, drawing upon existing sources of data. The main
source of current data is a destination survey of all graduates six
months after graduation, the Destination of Leavers from Higher
Education (DLHE) survey, which is complemented by a follow
up survey three-and-a-half years after graduation, known as
longitudinal DLHE. These surveys are currently under review by
the Higher Education Statistics Agency (HESA).
DLHE is unpopular in higher education. This paper discusses

the key criticisms of the main DLHE survey – timing,
methodology, definition of graduate jobs and cost – as well as
of the longitudinal survey, which lacks influence, and identifies
balancing arguments. DLHE surveys have strengths and there
are opportunities to mitigate the main weaknesses.
The main survey itself is the envy of other countries and

provides a great deal of authoritative data about employment
outcomes down to occupations attained at subject level. The
first destination focus of the main survey and its closeness to
graduation ensure a high response rate and in-depth data, but
www.hepi.ac.uk 39
is unrepresentative of graduates’ early career progress in some
subjects.
Longitudinal DLHE, which occurs only once every two years and
based on a sample survey, is a pragmatic response. It provides
aggregate-level corrective data, but is largely unknown outside
the higher education sector. Longitudinal DLHE could be
enhanced to provide better, annual data to sit alongside short-
term first destination data, giving a clearer indication of early
career paths. Despite the criticism, a suitably reformed DLHE
should have a continuing and important role.
The first major publication that utilises Her Majesty’s Revenue

and Customs (HMRC) earnings data provided new information,
but it took a long time to produce, is at an aggregate level and
covers 10 years before the present fees regime was introduced.
It may be four years or more before such data are available at
subject level and are therefore potentially useful to prospective
students making choices now. This earnings data will get better
and more useful over time but will continue to have limitations
as a metric.
For example, earnings vary considerably by industry and

region. Higher education institutions in modestly-performing
local economies may anchor economic and social activity. Such
institutions face perverse incentives to ship their graduates out
to higher-paying industries and regions, notably London, in
order to improve their employment outcomes to the detriment
of the local economy. This suggests an element of assessment
of institutions’ progress against their own targets – reflecting
their context, mission and priorities for graduate employment
– should be included.
One intention of the TEF is to link the ability to raise fees (by
inflation) to teaching excellance outcomes. In a recent large-
scale survey conducted by HEPI and the HEA, existing students
expressed decreasing confidence in the value for money of
university education and strong rejection of the linking of TEF
outcomes to the ability to raise fees.
There is rising student interest in employment outcomes,

and an indication from a small-scale survey that a link to a
guaranteed graduate-level job might be one condition under
which students would contemplate higher fees. A guarantee
is unrealistic but it may be an important pointer to emerging
student sentiment. This finding needs to be further investigated
before any shift away from first destinations in the DLHE main
survey is contemplated as a result of the HESA review.
A major concern is that employment outcomes largely reflect

non-educational variables, such as entry standards, socio-
economic background and subject studied. Although these
factors are influential, with good data statisticians can and
do control for these variables to enable performance to be
benchmarked. When they do, educational variables that make
a difference, at least to early careers, emerge. Work experience
integrated into the curriculum stands out, especially sandwich
years and placements, alongside obtaining a good degree
(first-class or upper second-class honours). Welcome proposals
to collect more data on placements and work-based learning
are made in the HESA review of DLHE.
The TEF will need to take care that it reflects educational

variables in any judgements about teaching excellence. In this
respect, the proposals in the TEF technical consultation and its
www.hepi.ac.uk 41
general tone are encouraging, although there is a long way to
go in constructing an authoritative TEF.
1. Both earnings and occupational data are required to assess

the labour market progress of graduates in the round.
2. The main DLHE survey should be reformed to mitigate

weaknesses but retained, as should its focus on first
destinations and occupations.
3. Longitudinal DLHE should be enhanced with a view to

providing annual data and in sufficient quantity to enable
subject-based analysis at institutional level.
4. Statistical controls and benchmarking should be applied to

all TEF data to ensure that it reflects educational variables
and educational excellence.
5. There should be an element of assessment against declared

institutional targets to reflect the very diverse economic and
social contexts, missions and priorities of institutions.
6. Some current plans, such as the proposal to collect more

data on work experience and placements, are welcome,
but others run the risk of distracting attention from labour
market outcomes.

Why employment outcomes are important and
how they should be measured in future
Richard Blackwell and Martin Edmondson
Introduction
The Government plans to include employment metrics in

the proposed Teaching Excellence Framework (TEF). This is a
reasonable desire, reflecting the needs of prospective students,
employers and the national economy. The Government’s white
paper, Success as a Knowledge Economy: Teaching Excellence,
Social Mobility and Student Choice, focuses on the development
of a Longitudinal Education Outcomes (LEO) dataset built
around newly available tax data on earnings. The wording,
however, covers broader possibilities through the phrase
‘employment and earnings’, which recurs in the text (BIS, 2016a).
An accompanying technical consultation sought views on
some options for including occupational measures alongside
earnings data, including a ‘highly skilled employment’ metric
based on graduate-level occupations (BIS, 2016b).
This paper argues that it is possible to construct a TEF metric

based on employment outcomes by combining the new
earnings data beginning to emerge with reform of existing
surveys focused on destinations and occupational progress. It
would not be a perfect solution, which is anyway unattainable,
but it would provide the basis for the development of a robust
and educationally-focused assessment.
www.hepi.ac.uk 43
In the face of continuing scepticism, this paper:
•• considers why outcome measures are important;
•• d
iscusses the two main sources of data, which are existing
destination surveys and HMRC earnings data;
•• suggests how these sources may be used together to

provide a system focused on the difference that education
makes to outcomes; and
•• u
rges the current official review of destination surveys to
ensure a continued focus on first destinations alongside
other information.
Context
The Government’s focus on employment metrics is nothing

new, even if the specific proposals are.
Newspaper league tables (published annually by Times Higher

Education, Guardian, The Sunday Times and others) have
included employment metrics for years, all slightly different,
but derived from the same source, the annual Destination
of Leavers from Higher Education (DLHE) survey of recent
graduates.
DLHE has been the subject of considerable criticism and is not

popular in the higher education sector.

Destination of Leavers from Higher Education annual survey
DLHE was largely designed by university career services

to provide feedback on their activities. Over the years, the
main survey has morphed from developmental feedback
for a professional service into a high-stakes institutional
performance measure used in newspaper league tables. All
graduates are surveyed six months after leaving and a high
response rate is achieved. The survey provides substantial
anonymised data about the jobs obtained by leavers, down
to job title, as well as recording subject studied and institution
attended. Pay data is collected but via questions which
achieve low response rates of around 20 per cent. Since 2006
this main survey (frequently referred to simply as ‘DLHE’) has
been complemented by a longitudinal survey of leavers.
DLHE is being reviewed by its owner, the Higher Education

Statistics Agency (HESA, 2016). This sets the high-level
objectives of:
•• f uture proofing the system, particularly changing labour

markets and demands for richer data on graduate outcomes;
•• efficiency, particularly enhanced use of technology to

access and link existing data sets currently held separately;
•• fi
tness for purpose, ensuring the data is usable in new and
emerging contexts; and
•• supporting legislation.
The review floats a number of new areas for graduate self-

evaluation, addresses information gaps and raises questions
www.hepi.ac.uk 45
about survey design. It suggests four new measures of graduate
outcomes that could be alternatives to destinations. They are:
•• student engagement with employability provision;
•• n
et promoter scores, which refers to whether respondents
would recommend their course or institution;
•• questions on subjective well-being; and
•• measures of attributes and skills for life.
It also suggests expanding the range of employment data,

notably to include more information on work and placement
experience. The consultation document issued as part of the
review included a comprehensive list of 128 questions. This
paper focuses on the strategic issues which need to inform the
redesign of the system rather than the details.
The final point of the high-level principles informing HESA’s

review, supporting legislation, refers back to the Government’s
plans. The Small Business Act (2015) has a section on graduate
destinations that enables the ‘Secretary of State to provide
destinations information to the governing body of an institution’
and permits the use of official earnings records. In 2016, the
Institute for Fiscal Studies (IFS) published the first substantial
attempt to make use of such data (Britton et al, 2016). It is
tempting to believe that earnings data made available from
this legislative change can replace other forms of analysis of
graduate destinations, including that previously supplied by
DLHE surveys.

Earnings are a well-understood common currency, key to
graduates’ life chances and can be obtained from an official
source, HMRC. Indeed, the Government proposed in its 2016
higher education white paper to build a new Longitudinal
Education Outcomes (LEO) dataset on these foundations (BIS,
2016a). The supporting technical consultation on the TEF
suggested two further measures:
1. Use of the ‘UK PI employment indicator’, one of a set of

longstanding UK performance indicators published by HESA.
It is derived from DLHE data and measures the proportion
of UK-domiciled graduates that are in employment, work or
further study. This is of limited value since it is really just a
measure of whether graduates are employed or not. It says
nothing about the quality of employment obtained and
so it is generally supplemented by other more in-depth
measures.
2. An occupationally-based ‘highly skilled employment’

metric (BIS, 2016b), which also can be derived from annual
DLHE survey data, although the proposal would involve a
grouping and relabelling of existing occupational analysis.
These proposals have allayed concerns that a crude income
metric might dominate the TEF.
The main concern about an income metric is illustrated by

the IFS study. Although it provides a mass of new description
and original analysis, its main conclusions are in line with
what we already know. Indeed the major issue discussed in
its conclusion, the continuing advantage of having a wealthy
family background in attaining high earnings, lies behind
scepticism over whether the TEF can find an employment
www.hepi.ac.uk 47
metric that reflects educational excellence rather than socio-
economic and other non-educational factors.
Why employment outcome data is important
It is important to develop good outcome data because:
•• outcome measures matter to students;
•• there are academic staff in higher education that reject

employment outcome measures; and
•• t he very wide-ranging HESA review of DLHE risks distracting

attention from outcomes in favour of less salient but
currently in-vogue context data.
There are a number of inter-related reasons why employment

outcomes matter. The first is that in the shift from elite to
mass higher education, the easy assumption that graduates
could step straight into a professional job, prevalent up to
the 1980s, has come under challenge. While the ‘graduate
premium’ remains impressive on average (Walker & Zhu, 2013),
averages can disguise a wide spread of outcomes (Britton et
al, 2016). Most graduates do well in the labour market but
it cannot be assumed that all will walk straight into a good
job. Each year, about 65 per cent of graduates are in graduate-
level managerial and professional jobs after six months, rising
to about 80 per cent after 40 months (Zuccollo, 2015), so a
significant minority of graduates are not in ‘good jobs’ at these
points.
Secondly, the desire of higher education entrants to have

better life chances is clear. Central to this is obtaining a good,
reasonably well-paid job. When asked to identify the most
important measures of teaching quality, students put outcome
measures in first and second place (securing a ‘good’ degree
and ‘graduate employment statistics’ respectively) (Hillman,
2016).
Thirdly, higher education has been actively sold to first-

generation student entrants as a path to the good life
through widening access and participation policies. Many
are from modest means and increasingly take on substantial
debts expecting to acquire first-rate employment. It is open to
institutions to declare that they are focused on transformative
education within campus boundaries and are not seeking
to prepare their students for labour market success. Few
transparently do. So, unless widening access and participation
policies are to be cynical recruitment exercises, there is an
ethical dimension to caring about employment outcomes.
Fourthly, pursuing social mobility is an implicit aim of widening

access and participation and a central policy objective of the
TEF. Social mobility often tends to be treated as synonymous
with entry to university, yet there is evidence that disadvantage
can carry through university life into the labour market,
including for graduates from ‘top’ universities (Bathmaker et al,
2013). Indeed disadvantage can become further entrenched
during study, for example among relatively disadvantaged
students who live at home for economic reasons and are
unable to develop their wider social and cultural capital
(Purcell et al, 2013). There is disturbing data on the gap
between disadvantaged and the most advantaged graduates’
www.hepi.ac.uk 49
acquisition of graduate-level employment of about seven per
cent at six months continuing at the same level through to
40 months when most other gaps close (Zuccollo, 2015). The
implicit equation that ‘access equals social mobility’ is not
tenable at least in its simple form. We need labour market
data, among other things, to assess progress and help devise
strategies that impact positively upon social mobility.
Fifthly, there is the desire to link TEF assessment to the

ability to raise fees at least to ensure that the real income
to institutions is not eroded. The 2016 HEPI / HEA Student
Academic Experience Survey shows rising concerns about value
for money and direct rejection by an overwhelming 86 per cent
of respondents of the proposition that ‘fee rises for excellent
teaching are a good idea’ (Neves and Hillman, 2016). On the
other hand, a recent small-scale survey of applicants and
students by UPP found that nearly three-quarters would be
prepared to pay higher fees if the course to which they were
applying gave a guarantee of a graduate-level job at the end
(UPP, 2016). Employment outcome data, and specifically first
destination data, might have a significant role to play in making
a case for increased fees.
Sixthly, there is scepticism that outcomes do any more than

reflect entry standards, subject studied and reputation-based
institutional hierarchies unrelated to teaching excellence
(Gibbs, 2015). Entry standards matter, performance between
subjects varies significantly and there are a number of
important social variables too, including socio-economic
background, gender and some aspects of ethnicity.
Institutional-level comparisons need to take account of these
factors. However statisticians are quite capable of controlling
for these variables, providing they have the data, and do so
routinely (see Zuccollo, 2015; Britton et al, 2016). Moreover,
research that applies such statistical controls has identified
important educational variables that have an independent
impact on outcomes. These are degree class obtained (Walker
and Zhu, 2013) and work experience embedded in and
surrounding the curriculum – especially four-year degrees
containing a sandwich year placement (Purcell et al, 2013;
Sanahuja Velez and Ribes Giner, 2015), but also including shorter
placements embedded in standard three-year undergraduate
degrees (Thompson, 2016).
One of the proposals to enhance employment data in the HESA

review of DLHE is to close an information gap by investigating
in some depth graduates’ experience of placements and work-
based learning. This is a welcome development, which may
show change is underway, albeit with a time lag before it
begins to impact destination data.
Employment outcomes are only one important lens for looking

at the impact of undergraduate education. It should not be seen
as exclusive or competing with the need for enhanced process
data, including from ‘learning gain’ initiatives (see Millward,
2015; Schleicher, 2016). The alternatives to destinations floated
in the HESA review all have their merits, provided that the focus
on destinations is not simultaneously lost.
It would, of course, be quite possible to have a process-only

system for ‘teaching excellence’ that ignores labour market
outcomes, but it would be a partial account of the impact and
importance of undergraduate education and fail to meet the
needs of key stakeholders.
www.hepi.ac.uk 51
Destinations of Leavers from Higher Education
Although DLHE is frequently referred to as a single survey, there

are in fact two surveys: the main annual survey and a newer
longitudinal survey. It is the main DLHE survey that is used in
league tables. Although it has been the focus of criticism, it has
important strengths.
1. It surveys all leavers (300,000+ each year) and attains a

high response rate (around 80 per cent). There is a large
bank of reasonably consistent data going back many years,
on which other countries look enviously.
2. It provides a great deal of detail on occupational

destinations to higher education institutions and other
users. DLHE gathers occupational data, including whether
graduates are in work or not, in graduate-level managerial
and professional employment (equivalent to the highly-
skilled employment metric proposed in the technical
consultation about TEF), or travelling. Analysis is presented
for a number of different populations annually and although
anonymised, a large amount of data are available on job titles,
ethnicity, gender and so on for institutions (and others) to
access and interrogate. Many higher education institutions
do just that. It is possible to identify all the job titles attained
by students from particular courses and therefore whether
there is a good match between course aims and actual
performance in the near term. This is powerful information
which at present does not appear to be available from other
sources.

3. The detail and granularity of the data enable
benchmarking. For a few hundred pounds benchmark
data can be purchased from HESA at subject level, enabling
performance to be compared with similar courses with
similar intakes elsewhere. This transparent performance
data enables both awards for top performers and a focus on
improvement where it is required.
4. It has provided evidence for innovation in employability

provision in higher education institutions in recent years, an
area in which the UK can claim to be a leading international
player (Pegg et al, 2012; University of Huddersfield, 2014).
However, DLHE has also been the subject of criticism. The main
problems are discussed below.
1. The six-month survey point: The early survey point was

originally intended to reflect the first destinations of leavers
with the thought that this was an outcome higher education
could influence and support. But it has become clear that
the data are not always a good guide to early career paths
and a longitudinal survey was added to supplement them.
Recent evidence from this longitudinal work shows a general
convergence at 40 months between subject outcomes that
diverge at six months (Zuccollo, 2015).
There is a case for retaining first destination data and

accompanying it with longitudinal data to reflect early career
trajectories beyond first destinations, both to inform student
choice and because it provides a more balanced measure
of subject and institutional performance. There is, however,
no agreement on the best measurement point and the
www.hepi.ac.uk 53
further from graduation one assesses graduates’ progress
in the labour market, the more problematic it becomes to
ascribe that to their degree studies alone. Moreover, the less
it tells us about the current state of the course. Graduates’
career progress will be affected by additional skills, qualities
and connections acquired at work after graduation and
independent changes in labour markets, especially in flexible
labour markets, unrelated to their degree. First destination
data may have its imperfections and limitations but so does
long-run data and a sensible balance needs to be struck.
2. Longitudinal DLHE: This second survey was introduced

in 2006 to supplement the annual survey six months after
graduation. The main criticisms are that it lacks salience
and has no impact. Based upon biennial sample data
gathered by external collectors for HESA, rather than higher
education institutions, it surveys graduates three-and-a-half
years after exit. Response rates are typically in the 20 to 30
per cent range (27.6 per cent in 2014 – the fifth such survey)
and numbers are too small to allow drilling down below
institutional level to subject level within institutions in the
way that the annual DLHE survey does. The longitudinal
survey does, however, allow aggregate cross-sector analysis,
for example on the progress of students with particular
characteristics (Zuccollo, 2015). Owing to its limitations, the
longitudinal DLHE survey data are not included in the Key
Information Set (KIS) data published on university web sites
or in newspaper league tables and therefore do not have
great prominence. These factors may lie beneath footnote
17 in the TEF technical consultation which states that ‘We
are not proposing to use the longer term DLHE measure,

taken at 40 months, as it is a relatively small sample reported
at national level and not designed to be representative of
any specific provider’ (BIS, 2016b).
Whether to continue with two surveys or not is one of the

main issues raised in the HESA review. Options are presented
for a single survey or two surveys at various intervals from
six months to 48 months. The questions are framed in
the context of a desire for more longitudinal data, and by
implication, whether there is a better solution to the current
two surveys.
The main risks in moving away from a six-month survey are

to the first destination focus and the response rate. The first
is of interest to students and the second enables detailed
benchmarking at subject and course level. At the same
time it clearly requires at least complementary longitudinal
data. Experience of longitudinal research is that response
rates tend to drop off substantially over time as the sample-
based longitudinal DLHE itself illustrates. In time it may be
that LEO will provide most of the longitudinal data required,
albeit limited to earnings, but in the short-to-medium term
that seems unlikely. Longitudinal DLHE would be much
more influential and useful if it could be reformed to provide
annual data, and in sufficient volume to allow institutions
and users to drill down to subject level.
3. Methodology and allegations of gaming: The origins of

the main DLHE survey explain its in-house methodology, in
which the survey of graduates and coding of responses is
conducted by universities themselves. This has given rise to
questions of consistency and suspicions of gaming. Some
www.hepi.ac.uk 55
institutions, for example, have switched their graduation
ceremonies to coincide with the DLHE survey collection
period in order to achieve high returns in an environment
of excitement and celebration. Whether this affords any
real advantage is debatable. Some institutions contract out
DLHE data collection (for example, in London) and central
data collection is floated in the HESA consultation. Caveats
include: the scale and timeline of the national exercise
(300,000 students annually within three months); the cost
(although current cost is already being covered somewhere
in the system); and the potentially negative impact on
response rates that some fear external data collection might
have. These risks need further assessment, but if they prove
too great then existing controls could be strengthened to
maintain confidence in the system. In particular, the audit
function presently located in HEFCE but presumably to be
transferred into the Office for Students (OfS) in future could
establish a regular cycle of formal activity, perhaps auditing
a sample of returns each year. It could also enhance its
capacity to receive and follow up complaints.
Defining graduate jobs: DLHE uses the Standard

4.
Occupational Classification (SOC). It is a standard way of
classifying jobs that produces a hierarchy of levels or tiers and
enables a boundary between graduate and non-graduate
jobs to be identified. It is independent of DLHE and higher
education, although the initial coding of graduate responses
is done within institutions. The boundary between graduate
and non-graduate jobs is open to argument. However, most
of the criticisms levelled at DLHE are actually SOC coding
issues (Peck, 2015). Occupational boundaries change over

time and some time lags are inevitable in any classification
that is not conducted in real time. For example, at the time
of writing, the police service is debating whether to become
a graduate-only entry profession. The review of SOC by the
Office for National Statistics will no doubt produce some
updates and improvements. In that vein, ensuring frequent
updating of SOC would help mitigate one of the main
criticisms of its use although at the expense of some loss of
comparability across years.
A crucial point overlooked by some critics of DLHE (who

focus on narrow boundary issues) is that although the SOC
may not be perfect, this feature of DLHE is actually one of
its strengths in the present context. It enables occupational
analysis of graduate success that both provides an alternative
to earnings analysis and a relatively simple control for some
of the distorting factors in earnings data.
The HESA consultation floats the idea of detailed skills

analysis by combining HESA desk analysis of course and
subject claims with a new ‘battery’ of questions. Its reduced
list amounts to 12 new questions for graduates (HESA, 2016).
It is left open whether this is intended to be a supplement or
replacement for occupational analysis. As a supplement the
idea has merit, but as a replacement it could be complex and
confusing. A particular caveat is that experience suggests
some questionnaire fatigue among students. Proposals for
additional or substantially expanded student surveys, which
this idea would require, should be tested with students
before implementation.
www.hepi.ac.uk 57
In contrast, the technical consultation on TEF proposed a
simple solution that kept the focus on outcomes. It proposed
a ‘highly skilled employment’ metric based on jobs classified
in SOC groups one to three, roughly equivalent to the
current graduate-level ‘managerial and professional’ job
label used to distinguish graduate and non-graduate jobs.
The intention of the TEF consultation proposal is presumably
to simplify the identification of ‘good jobs’ and destinations
in the interests of clarity for users. It involves a pragmatic
approach and a judgement that puts simplicity for users
above sophistication of analysis. In the absence of evidence
of student appetite for more surveys and/or contextual
data, it seems reasonable, especially as there is scope for
improving the responsiveness of the SOC system to changes
in the labour market.
5. Cost: The annual six-month DLHE survey involves a

considerable commitment of resources by higher education
institutions, as the high response rate is achieved largely
from labour-intensive telephoning of graduates and not
from the online version of the survey. Many institutions
employ temporary staff, often students, to undertake this
telephoning. At the present time, the overall cost of the
exercise is not known but the HESA consultation intends to
address this and gather data on cost. Abandoning labour-
intensive telephoning could be attractive from a cost
perspective, but may have a big impact on the response
rate and hence usefulness of the data. The main potential of
data linking, identified in the HESA consultation as having
potential to improve efficiency and quality, seems to be
around earnings and benefits received by graduates, so at

the moment cost savings from this source could be limited.
Most of the options for change are likely to involve additional
costs at least in the transitional phase.
Changing DLHE
DLHE has strengths that are worth retaining and weaknesses

that need to be addressed. Some of the ideas floated in the
HESA consultation, if adopted at the expense of destination
and occupational data, could leave DLHE providing nuanced
and complex reflective data from graduates that are simply
overpowered by earnings data in assessments and league
tables. DLHE destination data provide an alternative and
complementary basis to LEO’s focus on earnings for evaluating
graduate destinations that should be retained.
A reformed DLHE should keep a focus on destinations and

occupational analysis utilising the SOC. The current two-survey
format should be retained, at least in the short-term, as the risks
of doing otherwise are too high. An early survey to capture first
destinations and ensure a high response rate, at six months
or a similar period, has value for prospective students. New
questions on whether students have received placements
or other forms of work-based learning as suggested in the
HESA consultation, are supported by labour market research
as a key educational variable and could usefully be included.
Other ideas for student self-assessment floated in the HESA
consultation should be treated with caution unless they can
show equivalent labour market salience. Existing longitudinal
DLHE should be improved to provide annual data, preferably at
subject level, to sit alongside short-term first destination data.
If it can be developed sufficiently, over time, to provide course
www.hepi.ac.uk 59
level occupational data, that would be highly desirable and
should be an aim of the longitudinal survey.
There is a pragmatic case for continuing to utilise the SOC

classification of jobs in the absence of a clear alternative,
although ideally with frequent updating to more quickly reflect
changes in the jobs market. The early destination survey could
be centralised or contracted out. If, after receiving consultation
responses from institutions, centralising and contracting out
appear prohibitively expensive or likely to significantly reduce
response rates, the current system could be enhanced by
substantially augmenting the role of audit to protect and give
confidence in data integrity.
Earnings and pay
The main DLHE survey has always included income data but
those data derive from survey questions that are disliked by
graduates and often produce much lower response rates than
other questions. It is an acknowledged area of weakness. The
HMRC data now emerging should be a substantial improvement
on DLHE pay data, although there are important reasons for not
relying on earnings as the sole basis for assessing graduates’
progress in employment.
Jobs should not only be judged on the basis of financial

rewards. For example, cultural and creative industries are an
important sector of the economy and fulfil functions that are
central to both human society and broader creative innovation
but where earnings are typically relatively low. A different but
similar case can be made for some health and caring professions
where roles may have value to users and society not captured

by pay alone. Earnings are not the only, some might even say
main, criterion by which such roles should be evaluated and
are rarely what attract people to work in such sectors. Indeed
it is noticeable that in the recent poll for HEPI quoted earlier,
‘Salaries for new graduates’ features well down the list of
suggested measures of the quality of teaching, unlike ‘graduate
employment statistics’, suggesting that students make quite
nuanced choices, favouring more rounded ‘statistics’ over pay
(Hillman, 2016).
There are some practical objections to earnings as the dominant

metric for employment outcomes too. Firstly, whether earnings
data will provide the granularity and transparency to enable
meaningful comparison below institutional level is still
unknown. Tax data is understandably well protected, and may
not be released to universities with the freedom that exists on
data from DLHE. The white paper promises pilots involving
earnings data at subject level in autumn 2016 (BIS, 2016a) and
it appears that detailed subject level data will not be available
for about four years. This will severely limit LEO’s short-term
usefulness and it is as yet unclear how good the longer term
data will be.
Secondly, there is the issue of relevance for prospective students

and parents. The IFS study mentioned above took years to
emerge, involves a sample of only 10 per cent of data and covers
a period of 10 years before the current fees regime had been
introduced (Britton et al, 2016). In that time, some courses will
have disappeared, completely new ones begun and others
changed beyond recognition. If the LEO earnings metric is very
long-term some nearer-term outcome data reflecting degree
courses on offer to current prospective students will be required.
www.hepi.ac.uk 61
Thirdly, earnings as a sole or dominant metric may create
perverse incentives for some higher education institutions.
There are marked imbalances in pay by occupation, industry
and region. Average salaries vary significantly across the UK
and tend to be lower outside London, in some cases vastly
lower. In April 2015, the Office for National Statistics reported
that by local authority the City of London had the highest
median gross weekly pay of £921 and North East Derbyshire
the lowest at £389 (ONS, 2015). It will be important for any
credible exercise to be able to recognise and control for this
variability when comparing institutional and subject outcomes
from institutions in different parts of the country. One way
of doing this is to compare occupational outcomes within
subjects, independently of the pay that they attract.
An income metric crudely applied would suggest that in order

to maximise their employability outcomes, institutions should
seek to export as many of their graduates as possible into high-
paid industries (stereotypically financial services), located in
London. An example of the potentially negative impact of such
incentives is where higher education institutions play a lead or
‘anchor’ role in a relatively modest regional economy:
•• in developing the skills agenda with Local Enterprise

Partnerships (LEPs) and business partners;
•• a s a major source of innovation through spin-out companies

and graduate business start-ups;
•• in supporting local schools through partnership and

sponsorship (and thus development of the skills pipeline
and school improvement strategies);

•• b
y promoting social enterprise through work with local
communities and third-sector organisations, strengthening
community resilience; and
•• b
y working with local partners on economic and social
development to attract inward investment and forestall the
flight of business elsewhere (Allan, 2015).
Incentivising such institutions to export their graduates

elsewhere would strip the local labour market of the very talent
most likely to lead economic renewal. It also sits uneasily with
other policy objectives designed to rebalance the national
economy, for example devolution.
One approach to mitigating the risk of perverse incentives might

be to encourage an element of assessment around institutions’
own targets. In the main, this would involve bringing into the
foreground and making more transparent existing institutional
goals. Institutions could be encouraged to declare their own
indicators or targets for employment outcomes, based on
their mission and strategic priorities, against which their
performance could then be assessed by stakeholders.
In this context, prospective students, parents, employers,

careers advisers, teachers and other stakeholders could make
their own judgements about the adequacy of the institution’s
employability ambition and its performance against the targets
it has declared. For example, some might openly aim to fill the
City of London with stockbrokers, others to increase the pool
of graduates recruited and retained within local and regional
labour markets, perhaps as an integral part of plans with an
LEP to attract inward investment. Recruiting and retaining
www.hepi.ac.uk 63
high proportions of graduates locally, even if earnings levels
compare unfavourably with London or other regions, might be
an explicit target and, if attained, judged an excellent outcome.
Conclusion
Good HMRC earnings data will be an improvement on DLHE

pay data. For reasons of principle and practice, earnings data
are likely to be a necessary but not sufficient condition for
assessing employment outcomes in the round. A reformed
DLHE system, including an enhanced, annual longitudinal
survey, can offer occupational data to sit alongside pay data,
providing a wider view of labour market performance and the
jobs attained by graduates. Occupational and earnings data
can and should be subject to statistical controls to identify
educational variables and excellence rather than the impact of
socio-economic, subject and other non-educational variables.
Both occupational and earnings data should be available

with sufficient granularity to enable enhancement action
at institutional and subject levels, designed to improve
relative performance and hence overall student experience
and outcomes. Unless or until there is an alternative way of
delivering good occupational data, a reformed DLHE system
should be retained.
At some point the erosion of the unit of resource (per-student

funding) will have to be broached in ways that are persuasive
to existing and prospective students. The UPP survey quoted
earlier suggests that first destination outcomes may be
important for persuading students that in some cases at least,
there is a good case for raising fees following TEF assessment.

Although the UPP findings can only be regarded as indicative,
it is not a propitious time to contemplate a move away from
the first destination focus that the DLHE main survey currently
provides.
This analysis therefore suggests that the TEF should contain

a small basket of measures built on existing foundations
and LEO data. These are:
•• a
n enhanced destination-focused DLHE main survey,
including new information on whether students have
experienced a placement or other form of work-based
learning;
•• an enhanced, annual longitudinal DLHE;
•• u
se of the SOC classification to identify advanced or
‘highly skilled’ graduate outcomes;
•• HMRC earnings data;
•• a
nd an institution-specific element in which performance
is measured against declared targets.
This data can then be incorporated into the UNISTATS website,

which the Government wishes to develop, and/or a revised
KIS already embedded in universities websites. The current
review by HESA would be well advised to focus on the scope
for enhancement, and avoid throwing the baby out with the
bath water.
www.hepi.ac.uk 65
References
Allan, T. (2015) Universities as anchor institutions. http://blog.hefce.
ac.uk/2015/03/16/universities-as-anchor-institutions/
Bathmaker, A-M., Ingram N., and Waller, R. (2013) ‘Higher education, social
class and the mobilisation of capitals: recognising and playing the game’,
British Journal of Sociology of Education, 34 (5-6) 723-743
BIS (2016a) Success as a Knowledge Economy: Teaching Excellence, Social

Mobility and Student Choice https://www.gov.uk/government/publications/
higher-education-success-as-a-knowledge-economy-white-paper
BIS (2016b) Teaching Excellence Framework: Technical Consultation for Year

2 https://www.gov.uk/government/consultations/teaching-excellence-
framework-year-2-technical-consultation
Britton, J., Dearden, L., Shephard, N., and Vignoles, A., (2016). How English
domiciled graduate earnings vary with gender, institution attended, subject
and socio-economic background, Institute of Fiscal Studies Working Paper
W16/06.
Gibbs, G. (2016) Teaching, in Hillman, N. (ed.), Response to the higher

HESA (2016) Consultation on Principles and Future Requirements for the UK’s
Public Interest Data about Graduates, Higher Education Statistics Agency
Hillman, N. (2016) Introduction, Hillman, N. (ed.), Response to the higher

education green paper, Higher Education Policy Institute.
Huddersfield, University of (2014) Enterprise and Entrepreneurship http://

www.hud.ac.uk/enterpriseandentrepreneurship/
Millward, C. (2015) HEFCE seeks the measure of learning gain http://blog.

hefce.ac.uk/2015/09/22/hefce-seeks-the-measure-of-learning-gain-2/

Neves, J. and Hillman, N. (2016) The 2016 HEPI / HEA Student Academic
Experience Survey http://www.hepi.ac.uk/wp-content/uploads/2016/06/
Student-Academic-Experience-Survey-2016.pdf
Peck, E. (2015) Finding new ways to measure graduate success http://wonkhe.

com/blogs/finding-new-ways-to-measure-graduate-success/
Pegg, A., Waldcock, J., Hendy-Isaac, S., Lawton, R., (2012) Pedagogy for
Employability, Higher Education Academy.
Office for National Statistics (ONS) (2015) Annual Survey of Hours and
Earnings 2015: Provisional Results: Section 7. http://www.ons.gov.uk/
employmentandlabourmarket/peopleinwork/earningsandworkinghours/
bulletins/annualsurveyofhoursandearnings/2015provisionalresults#region
al-earnings
Purcell, K., Elias, P., Atfield, G., Behle, H., Ellison, R., and Luchinskaya, D. (2013).
Transitions into employment, further study and other outcomes: Futuretrack
Stage 4 Report. University of Warwick Institute for Employment Research.
Accessed at http://www2.warwick.ac.uk/fac/soc/ier/futuretrack/stage_4_
report_final_30_10_2013.pdf
Sanahuja Velez, G. and Ribes Giner, G. (2015). ‘Effects of Business Internships

on Students, Employers, and Higher Education Institutions: A Systematic
Review’, Journal of Employment Counseling, 52 (3) 121-130.
Schleicher, A. (2016) Value-Added: How do you measure whether universities

are delivering for their students? HEPI 2015 Annual Lecture, Higher Education
Policy Institute.
Thompson, D. (2016) ‘‘I know what to expect next’. How valuable is short
project placement experience to higher education students?’ Journal of
Further and Higher Education (forthcoming)
UPP (2016) Work hard play hard: Lifestyle and employability at university.
University Partnerships Programme.
www.hepi.ac.uk 67
Walker, I. and Zhu, Y. (2013) The Impact of University Degrees on the Lifecycle
of Earnings: some further analysis. BIS research paper No. 112. Accessed at
https://www.gov.uk/government/uploads/system/uploads/attachment_
data/file/229498/bis-13-899-the-impact-of-university-degrees-on-the-
lifecycle-of-earnings-further-analysis
Zuccollo, J. (2015) Most Graduates use their skills but disadvantaged graduates
still lag behind http://blog.hefce.ac.uk/2015/10/07/most-graduates-use-
their-skills-but-disadvantaged-students-still-lag-behind/

Trustees
Professor Sir Ivor Crewe (Chair)
Dame Sandra Burslem
Professor Sir Peter Scott
Professor Sir Nigel Thrift
Advisory Board
Professor Janet Beer
Sir David Bell
Professor Sir David Eastwood
Professor Dame Julia Goodfellow
Professor David Maguire
Professor Dame Helen Wallace
Partners
BPP University
Ellucian
Elsevier
HEFCE
Higher Education Academy
Jisc
Kaplan
Mills & Reeve LLP
Pearson
Times Higher Education
UPP Group Limited
Wiley
President
Bahram Bekhradnia
Successive governments have tried to improve teaching
quality in universities. The proposed Teaching Excellence
Framework (TEF) is the latest attempt.
This new pamphlet explores two of the most difficult issues

that need more attention and offers some solutions.
In the first chapter Paul Blackmore considers why research has

a higher status than teaching, exploring issues of prestige and
reputation to look at what can be done about it.
In the second chapter Richard Blackwell and Martin

Edmondson explain why employment outcomes, specifically
first destination data, are so important and argue that they
should be central to the TEF in future.
HEPI was established in 2002 to influence the higher education debate with evidence.
We are UK-wide, independent and non-partisan.
September 2016 ISBN: 978-1-908240-17-0
Higher Education Policy Institute

99 Banbury Road, Oxford OX2 6JX
Tel: 01865 284450 www.hepi.ac.uk
Printed in the UK by Oxuniprint, Oxford

Typesetting: Steve Billington, www.jarmanassociates.co.uk

Hepi TTWI-Web

Uploaded by

Copyright:

Available Formats

Hepi TTWI-Web

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hepi TTWI-Web

Uploaded by

Copyright:

Available Formats

Tackling Wicked Issues:

Prestige and Employment

Richard Blackwell is an Emeritus Professor of Southampton

Martin Edmondson is the Chief Executive of Gradcore, and has

2 Tackling Wicked Issues

Why research trumps teaching and what

 hy employment outcomes are important

Research has higher status than teaching in academic life

The problem is deep-seated and related to prestige, defined

•• slow to gain or lose; and

•• often decided on by insiders.

Excellence in research attracts prestige, but excellence in

4 Tackling Wicked Issues

Providing more money for excellent teaching is not the answer.

The Government proposes an even stronger separation

Rewarding excellent teaching requires metrics but prestige

Prestigious research-intensive institutions will be less affected

A more differentiated higher education system tends to be more

An alternative way forward is to link research and teaching

Impact offers another place for connection. The REF emphasises

6 Tackling Wicked Issues

What might this mean in practice? Public funding for research

Teaching must be evaluated differently. The National Student

It cannot all be done at an institutional level. Disciplinary and

Recent changes to the UK Government mean that

8 Tackling Wicked Issues

Research dominates academic life at the expense of teaching.

At the heart of the tension between teaching and research

A TEF may seem an obvious counterbalance to the REF, bringing

I argue here that the Government’s proposals have not grasped

How prestige and reputation are different

Recent work on academic motivation (Blackmore, 2016) has

•• not all can have it;

•• it is slow to gain and to lose; and

•• it is often decided by those within a group.

The award of a Nobel Prize to a staff member confers prestige

The ‘script’ that an individual may hold is a powerful shaper of

The idea of prestige economies at different organisational levels

Anyone who doubts the need to pay attention to patterns of

How government can make it worse

In a competition for academic attention, research will generally

A financial incentive assumes that the difference between the

However, such behavioural changes, if they happen, are part of

The strength of the underlying value system can be judged by

14 Tackling Wicked Issues

Cross-subsidy of research by teaching is vulnerable to criticism,

In summary, the financial relationship between teaching and

Turning to the second problem, both central government and

16 Tackling Wicked Issues

Unless the case for research improving students’ learning is

What won’t work

Proposals to reward excellent teaching will be frustrated by the

The Government’s wish to achieve an objective measure of

Because of these difficulties, some argue the environment

Cultural aspects of the use of metrics are even more difficult.

18 Tackling Wicked Issues

However, we have long known this happens. The complaint was

So in the absence of reliable objective measures, prestige will

While the internet certainly does overturn business models, and

Why research trumps teaching and what

hy employment outcomes are important

•• the inclusion of a teaching impact statement as part of all

•• reflecting a research-related conception of teaching in the

•• strongly favouring promotion on basis of proven links