REF consultation – follow the conversation here!

Des Fitzgerald, Cardiff University - 18.09.18

I am surprised to see that the word ‘critical’ has proven controversial in the proposed descriptor. This controversy seems to me to rest on claims that could use further discussion. The first is that sociology can be still be characterised through a binary in which work is either concerned with schism, on the one hand, or solidarity, on the other; the issue is not that these forms do not matter – as Prof. Bhambra rightly says below: yes, let’s have division, solidarity, and indeed many other forms of relation too – the error, in my view, is rather in thinking that sociological work is characterised by an affiliation to one *or* the other, and that inclusivity therefore requires a commitment to both (or neither). But if that split was once important to sociology, it is no longer clear to me that it plays any significant role in the mainstream of the discipline today – where, in fact, it seems much more usual to understand some scene or process as simultaneously, variously and ambiguously solidaritous, schismatic, conflicted, inclusive, and so on. Such are the entangled dynamics of social as I think (?) most sociologists would understand them today – not the grand, summary divisions of conflict and ‘social order.’ The second issue that warrants discussion is that this presumed division is then marked by a commitment to critique or the critical, and that the same term, ‘critical,’ marks a conflictual or (in Prof. Hammersley’s terms) ‘negative ‘perspective. There is a great deal that might be said about this debate (and again, I agree with Prof. Bhambra that this entire discussion needs to be situated in a more expansive attention to the history and politics of knowledge-making) but I note briefly that this specific concern seems to reflect an identity of the ‘critical’ with ‘critical theory’ deriving from the Frankfurt school – but the term as it is used today, of course, has many genealogies, not least its roots in literary interpretation, where it identifies a kind of rigorous and care-ful attention, a commitment to fine-grained judgment, and, yes, too, a willingness to take and argue for particular ethico-political positions. I don’t find that controversial: when we say to our students (as I invariably do at least) that we want them to think critically about some text or situation, we are obviously not, I think, asking them to adopt a ‘negative attitude’ to it – rather we are asking for just this kind of close and detailed attention, precisely this same form of rigorous interpretation, the same willingness to adopt and inhabit a stance and so on. In addition, I am surprised by the claim that Science and Technology Studies (STS), in particular, would be discouraged by a more critical descriptor. Of course I think we should take Prof. Dingwall’s report of his experience in 2008 very seriously. But just to note that I am an STS scholar, and I am writing this having just left the world’s largest STS conference, having already attended Europe’s largest STS conference this year – and in both places I found a deeply and unproblematically critical discipline, concerned with environmental and health justice, with the encoding of inequalities of race and gender into new technologies, with democratic accountability in innovation and science policy, and so on. I am hardly offering a scientific analysis, but for what it’s worth I did not leave either event sharing Prof. Dingwall’s view that work in STS would have difficulty aligning itself with the ‘critical’ discipline outlined in this descriptor.      


Paul Bagguley, University of Leeds - 17.09.18

Upon reading some of the contributions to this debate I must admit to not really recognising the draft Sociology UOA descriptor from some of the contributions. Yes there are serious issues around representation on the panel, and I think the work around racism, ethnicity, diasporas and migration could be handled in a more subtle and nuanced way. Furthermore, a comparison of the Social Work statement and the Sociology statement makes it clear that sociological work on sciences and technology clearly belongs in the Sociology panel. It specifically mentions ‘social studies of science and technology’ and ‘the digital’. Nor am I convinced about the concerns about a lack of concern with social solidarity, as the text refers to various phenomena such as culture, media and inter-group relations which are often concerned with the processes through which social solidarity is generated and reproduced. I fear that the descriptor has been read by some rather too literally and narrowly, and not in the broad-minded and ecumenical way in which it should. Concerns about the expertise of the panel in relation to specific specialisms should be addressed in the next phase of appointments to the panel, and is particularly challenging given the fluid character of specialisms that emerge, decline and mutate quite rapidly. At least they are all sociologists, unlike the Social Work panel, and that expertise really counts.

Also running through the comments is a kind of ‘shroud-waving’ about the future of the Sociology panel based around the usual themes of departments choosing to move to the Social Work panel, due partly to those departments’ interdisciplinary nature but also the myth of the reluctance of the Sociology panel to award ‘top marks’. First and foremost the Social Work panel’s name is to me quite striking. If both disciplines were equally represented surely it would have been ‘Social Policy and Social Work’. For me this speaks volumes about the politics and culture of that panel. For example six out of the top ten SW and SP departments in 2014 had significant Social Work components. Secondly there are a huge variety of reasons why particular departments were submitted to the different panels, and it is not always a departmental decision. Thirdly, we seem preoccupied with descriptors, publications and panel members’ expertise. I think we also need to consider the aspects of environment and impact and what these might mean for Sociology as distinct from Social Work. Impact for example could be more broadly interpreted for Sociology than for Social Work and the research environment would also be very different. So to the myth of Sociology’s reluctance to award top marks in the REF and previously the RAE – something I hear repeatedly asserted. Below is a simple table of the percentage of overall top marks from 1992 to 2014 that is publicly available. Of course the data is not really comparable and one could dig around in it in various more complex ways, but I do think it says something about the culture of the two panels over time and the new moderating role of Panel C. In 1992 what was than Social Policy was more generous, but since then became less generous than Sociology, and from 2008 onward we see the effect of the role of panel C ensuring consistently. In short, decisions about which panel to go to should be made on academic grounds, and not spurious assumptions about generosity. Is your mix of outputs, impact and environment more Sociology than Social Work?

REF/RAE comparison of Sociology and Social Work (previously Social Policy)

Overall scores

1992 RAE %

1996 RAE %

2001 %

2008 %

2014 %












Social Work






(previously Social Policy)         


John Holmwood, University of Nottingham - 14.09.18

In the context of pressure to provide 3*/4* PhD students and research fellows will potentially come under pressure to allow co-authorship, especially if they do not themselves meet the eligibility criteria (because of not being in a post). Perhaps there should be a requirement that for all all co-authored papers, the author on whose behalf they have been submitted meets appropriate criteria for being recognised as an author. The International Committee of Medical Journal Editors, for example, sets out the following:

  • Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND
  • Drafting the work or revising it critically for important intellectual content; AND
  • Final approval of the version to be published; AND
  • Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Their full justification for the position is here.


Muslims in Britain Research Network - 06.09.18

Our specific area of concern relates to the portability of outputs and redundancy. We at MBRN have had conversations with staff at institutions who have been through processes of redundancy in the last 12 months, and they raise questions about the current portability proposals. I have copied our response to the on-line consultation form immediately below. I would be interested to know if others have similar concerns, and if any potential solutions are being considered.


Consultation question 10

MBRN is broadly supportive of the proposal to make staff who have been made redundant ineligible for submission to REF 2021, but is concerned that it does not give sufficient protection to those placed at risk of compulsory redundancy. Redundancy is, needless to say, a drawn out process requiring formal notification, consultation and notice: it typically lasts 6 months or more. During this time staff have every incentive to move to a new employer. Employer and employee also often come to a voluntary settlement which may speed up the process. Under the current portability proposals, in both these circumstances staff would be eligible for REF submission, even though those leaving their employment would not be leaving voluntarily in any meaningful sense.

MBRN has in fact been made aware of members who have gone through redundancy proceedings and who have been offered voluntary severance amounting to their total wages for the period of consultation and notice. Due to ex gratia redundancy payments being untaxed, this gives staff a financial incentive to leave 'voluntarily' while costing the HEI little or nothing extra (due to the reduced pensions contributions). These staff have been informed that the HEI intends to submit their outputs if they are judged of sufficient quality. 

The negative incentives rightly identified by the research councils, then, do not appear to have been entirely removed by the proposal in 206.b. Indeed, we have been presented with evidence that this proposal is already influencing HEIs’ judgements about redundancy decisions. Given this, we would recommend considering measures to rule as ineligible not only those who have been made redundant but also those who have been placed formally at risk of compulsory redundancy.


Gurminder Bhambra, University of Sussex - 04.09.18

In March 2018, around 200 sociologists signed a letter to the Chair of Main Panel C, Professor Jane Millar, expressing our concern with the lack of representation of Black and Minority Ethnic colleagues both on Main Panel C and on sub-panels 20 (Social Work and Social Policy) and 21 (Sociology). In short, there are no BME colleagues on the Criteria Phase panels of either of these Units of Assessment. Out of twenty-three colleagues on the assessment panel for sub-panel 20 (which includes those on the criteria phase panel) only one is BME, while for sub-panel 21, made up of sixteen colleagues, only two are BME. By contrast, the gender ratio is 52:48 (F|M) for sub-panel 20 and 44:56 (F|M) for sub-panel 21.

In the ‘Consultation on the draft panel criteria and working methods’, it is stated that the four Main Panels, in consultation with the Equality and Diversity Advisory Panel, are responsible for equality and diversity issues. The requirements for equality and diversity, however, seem to be more passive than proactive. The issues are covered in length only in paragraph 37 which sets out the following:

  • “Where further nominations are invited, the REF team will again ask nominating bodies to provide information on how equality and diversity considerations were taken account of during the nominations process.
  • The Equality and Diversity Advisory Panel (EDAP) have reviewed the templates provided by nominating bodies in the 2017 round of nominations, and will produce a report summarising good practice identified in the nominations process.
  • Nominating bodies will be invited to consider this report when making nominations in 2020.
  • In recommending further panel members and assessors, sub-panel chairs will give consideration to enhancing the extent to which the overall body of members reflects the diversity of the research community.
  • The REF team are currently undertaking an analysis of the representativeness of panel membership, which will be able to further inform chairs’ considerations in this regard”

It is not clear how this guidance moves us forward from HEFCE’s earlier call for special attention to be given to equality and diversity issues especially in the light of our letter mentioned above which noted serious failures in addressing such issues in the initial stages for sub-panels 20 and 21.

Delineating sociology, as Graham Crow sets out, has always been a contentious issue. For my part, this is a consequence of sociology being both a system of knowledge oriented to history and constituted by that history. I agree with Crow that attention to the politics of knowledge also requires recognizing that there are differences over how knowledge is understood to be produced and to function. So, yes, there should be reference to solidarities as well as to division. However, there is also an issue of inclusion of other constructions of the politics of knowledge – constructions of its racialized character and not just its association with inequalities of gender and sexuality.

The displacement of the racialized structures of colonialism from sociology’s standard accounts of modernity, for example, contributes to the racialized structure of sociological thought in much the same way that colleagues have argued for the gendered character of knowledge production. The proposed descriptor for sociology – as well as its panel composition – seems to recognize that issues of gender and sexuality are not simply substantive topics of research, but also provide us with different ways of understanding the social world. This has been so over the various REF cycles. In the light of this, it is odd that there has been, and is, so little engagement with race, racialized structures, or associated issues of epistemology. Race is mentioned only once in the descriptor for sociology as a topic of research and the ‘history of social thought’ has been removed.

Much critical work has been done by those within sociology, likely more than in any other discipline, to address issues of race and racialization. Much of this has come about as a consequence of sociology opening up to diverse demographics, specifically to scholars from social locations not typical of those previously entering higher education. In the case of feminism and gender, it was clear to colleagues that the address of inequalities associated with gender required not only having expertise on gender, but also required women to play a full part within the academy and its broader structures. It was thus disappointing for Professor Millar in her response to our letter to state that “on the Sociology sub-panel 4 out of the 15 appointed so far have ethnicity, racism, multiculturalism, religion or nationalism as their current primary research, providing expertise on the assessment of these fields. While representation is not the same as expertise, it is important to note this coverage.” This is compounded by the fact that there has been no response to the raising of these concerns by either sub-panel.

Representation is indeed not the same as expertise. Sadly, the current sub-panels seem flawed on both representation and their range of expertise. Perhaps HEFCE places limits on the number of panel members that now make it difficult to resolve these issues. However, that would place bureaucratic constraints over the very commitment to equalities – their monitoring and address – to which the consultation is avowedly committed.


John Holmwood, University of Nottingham - 03.09.18

Across 5  cycles of the Research Assessment Exercise (RAE) and its successor the Research Excellence Framework (REF) the number of submissions to the Sociology Unit of Assessment (UoA) ave more than halved (from 62 in 1992 to 29 in 2014), while those for Social Work and Social Policy (SW&SP) have increased (the two areas were combined in 2008 and so precise comparisons cannot be made). Between 2008 and 2014 staff submitted to Sociology UofA fell from 927 FTEs to 704. Unless this trend is reversed there are likely to be as many sociologists submitted to the SW&SP UoA as there are to Sociology.

In this context, the BSA has an interest in subject descriptors and representativeness for both Panels and also in the working relations between them (which are more acute than those of other Panels on Main Panel C). The proposed descriptors for SW&SP are very specific. While it is accepted that the UoA is an interdisciplinary one and that there will be considerable overlap with Sociology, all descriptors refer to a focus on social work or policy. Many outputs of sociologists included in SW&SP submissions will not have this focus,

At the same time, the Sociology Panel as it is currently comprised does not have (self-declared) expertise in major areas of the discipline, many of which are likely to form part of a sociological component of SW&SP submissions – for example, the sociology of health and illness, science and technology studies, education, family and household.

The subject descriptors for this REF exercise contain a specific clause indicating the possibility of cross-referral (there was no mention at all in the REF2014 descriptors for SW&SP).  The justification in the SW&SP case is that this “arises from the large number of academic units that combine the constituent subject areas and that may make a combined submission in UOA 21 or UOA 20.” However, cross-referral only functions where there are no substantial gaps in the coverage of the panels. In addition, if there is recognition of ‘combined submission’, why is there no reference to relevant subject descriptors for sociology separate from policy and social work?

Of course, individual panel members are able to peer review in areas where they do not have direct expertise and no-one is questioning their integrity to do so. However, as Michèle Lamont suggests in her study of interdisciplinary peer evaluation, How Professors Think, such judgements are made possible by the representatives of different fields or epistemological positions within the panel seeking to make a collective judgement – relevant excerpt here. See, Graham Crow’s piece on how the descriptor seems to favour a particular, contested view of the discipline, and Gurminder K. Bhambra’s piece on how an implicit recognition of the politics of knowledge is skewed toward gender and sexuality with no equivalent recognition of race and ethnicity.

These issues cannot be resolved through policies of advanced calibration across panels. Indeed, in REF2014, it was clear that there were serious discrepancies in scoring outputs between the Sociology and SW&SP Panels, with the latter giving proportionately more 4* scores than the Sociology Panel and more than available metrics would suggest were appropriate. The idea that processes agreed in advance can resolve these difficulties is the same as suggesting that an exam board would meet in advance to agree its marking policies, but would not reflect on how they had been applied in practice.

Main Panel C needs to take up the issue of a proper review of outcomes and state how possible discrepancies will be mitigated. But, as sociologists, we also need to address the serious issue of the representativeness of the two Panels and the possibility that large sections of our discipline will fall between the gaps of each. It does not augur well that the joint HaPS/BSA meeting on 1st October to discuss REF2021 does not have on its agenda a discussion of the SW&SP descriptors, despite the fact that a very substantial proportion of the BSA membership are likely to be submitted to that UofA.  


Robert Dingwall, Nottingham Trent University - 03.09.18

I have noted the comments by Graham Crow and Martyn Hammersley on the REF descriptor for Sociology. It seems to me that if this form of words stands, compared with 2014, it will drive more units of assessment towards Social Policy as a more inclusive reviewer of their work.

All human social institutions can be considered in terms of their efficiency, their effectiveness, their equity, and their humanity. Do they use resources appropriately? Do they do what they claim to do? Are they fair to everyone involved with them? Do they treat everyone involved with them in ways that are considerate? Sociology has things to say about all of these questions. At any given time in any given project, any one may be more important than the others. As a discipline, we may be particularly sensitive to the equity issues but they do not define us in the way the REF descriptor implies. We also recognize that attempts to remedy perceived injustices often create disadvantages elsewhere and that planned interventions, however, well-intentioned, frequently have perverse outcomes. To the extent that we are a 'critical' discipline, it is only in that much wider sense.

If this descriptor had been in use in 2008, when I was on the panel, it would have discouraged submission of much of the science and society work that was identified as a distinctive strength in UK sociology. STS people are not indifferent to issues of equality and social justice, but they also recognize the challenges of innovation and its importance for economic prosperity, environmental quality and national interest. As Martyn implies, it is also hard to see how much of the important work being done by ethnomethodologists and others in partnership with computer scientists would qualify. Do we really want to suggest that if sociological research moves nearer market or towards advanced technologies that it ceases to be sociology?

As far as I can see the contentious second sentence of the 2021 descriptor adds nothing and could well be removed, leaving the identity of the discipline defined in more or less the same terms as 2014 (or 2008 as I recall).


Martyn Hammersley, The Open University - 02.09.18

Given the amorphous character of the discipline, statements about the nature of sociology are contentious: they will almost always betray sins of omission and/or commission, at least from some point of view. This is perhaps even more true today than it was in the past. The REF 2021 sub-panel have been forced to enter this minefield for a quite specific and highly bureaucratic purpose. And, in terms of this purpose, it seems to me that they have provided an account that is, overall, accurate and inclusive. However, I agree with Graham Crow that there is one part of their statement that is problematic. The second sentence of the ‘descriptor’ states that sociology ‘is a critical discipline which focusses on and is concerned with issues of social inequality, division and justice’. While this captures the orientation and focus of much sociological work today, it appears to exclude quite a lot that nevertheless belongs to the discipline. Graham rightly objects to the implication of an exclusive emphasis on the study of division, inequalities, and conflict, apparently leaving out work that focuses on, and is concerned about, social order and solidarity. But, for me, there is an additional problem: the description of the discipline as ‘critical’. Not all sociology is critical in the sense presumably intended here (indicating a negative attitude towards the phenomena being investigated). In this context, the point is not whether sociology should be ‘critical’ in this manner, it is simply a fact that not all of it fits this description. Indeed, there are approaches that specifically turn their face against such an orientation: ethnomethodology is perhaps the clearest example, but there is much other work that does not claim to know how the world ought, or ought not, to be on sociological grounds. The offending sentence appears to imply that such work is not sociological, and therefore would not be eligible. I suggest it be deleted.


Graham Crow, University of Edinburgh - 31.08.18

The delineation of the scope of Sociology has been a matter of on-going debate since the foundation of the discipline. The latest effort of the REF 2021 sub-panel 21 criteria group to describe Sociology is more extensive than the 2014 descriptor in terms of word length, but the result is a more narrowly circumscribed conception of what the discipline is. Of the three sentences added to the new version that is out for consultation, it is the second that narrows the field. The statement that “It is a critical discipline which focusses on and is concerned with issues of social inequality, division and justice” may appear unexceptionable until it is asked what such wording omits. Pondering this took me back to my 2006 inaugural lecture which set out to ask ‘If Sociology is the answer, what is the question?’ Taking inspiration from influential publication titles, I argued that Sociology’s defining questions were ‘Who are we?’, ‘Can we live together?’ and ‘Who gets what?’ The sub-panel’s specification of Sociology as a discipline concerned with ‘social inequality, division and justice’ confirms the value of the third question which John Westergaard used as the title for his 1995 book, but it sidelines the second (to my mind equally-important) question which Alain Touraine used for his 2000 book. Societies and other social entities are made up of both divisions and unities, of separation and togetherness, of conflict and cohesion, of schisms and solidarities (as David Lockwood’s 1992 book Solidarity and Schism flags as a key conceptual pairing). It is ironic that solidarity should be lost from the sub-panel’s view in the year that the BSA Annual Conference theme was ‘Identity, Community and Social Solidarity’. It is to be hoped that the wording of the sub-panel’s proposed descriptor can be revised to include both halves of the sociological tradition. Conflict sociology has its place, but it is not synonymous with Sociology as a whole.