Independent report

The national research environment for the study of extremism in the UK (accessible)

Published 14 December 2023

Prepared for the Commission for Countering Extremism (CCE)

This report has been independently commissioned. The views and opinions expressed are those of the authors and do not necessarily reflect the official policy or position of the UK Government or the Commission for Countering Extremism.

Copyright © 2023 Daniel Allington

Executive summary

As a result of systemic problems both in studying extremism and in communicating the findings of such study, there are likely to be substantial gaps in the knowledge base around extremism in the UK.

Above all, there appears to be a lack of research on specific extremist movements in the UK today — especially when it comes to Islamist extremism. This suggests that it would be unwise to assume that publicly available research on extremism provides a sound basis for UK government policy.

Findings of the two studies reported here suggest that:

• The study of extremism is highly politicised, and its politicisation presents clear potential for silencing and exclusion of certain perspectives
• Projects supported by the major public research funders appear to be skewed towards studies of extremism in general, as well as towards studies of the far right, especially with regard to the UK of the present day and the recent past
• There may also be comparable skews in research carried out without such support (for example, within think tanks)
• There are many obstacles to collecting relevant data by conventional means, including when trying to access data and research participants via state agencies, such as HM Prison and Probation Service (HMPPS), and when seeking approval from risk-averse research ethics committees
• The use of ‘naturally occurring’ online data in place of more conventionally collected data raises problems of representativeness, and also does not always avoid difficulties with regard to ethical approval processes
• Lack of data-sharing makes it difficult for stakeholders to seek second opinions, and leads to duplication of effort
• Dissemination of research findings on extremist groups and their associates and supporters is hampered by several factors, including online intimidation, credible physical threats, and strategic lawsuits against public participation (SLAPPs, also known as ‘lawfare’ or ‘intimidation lawsuits’)

The skew towards generalism appears problematic in that policy applications for research on general phenomena (for example, radicalisation) may be unclear in the absence of research on how those phenomena are manifest ‘on the ground’ (for example, in the recruitment practices of specific organisations). The skew towards studies of rightwing extremism, and away from studies of Islamism (especially in a contemporary UK context), could also perhaps be seen as problematic, given that Islamist extremism proportionally represents a far greater terror threat in the UK (see, for example, the Independent Review of Prevent), although that is an inherently political matter.

This report further suggests the existence of multiple social and ideological pressures which could potentially distort the research field:

• The skew towards general studies may be partly explained by an academic tendency to devalue ‘descriptive’ studies of specific phenomena: ambitious researchers are incentivised to target high prestige journals which tend to favour general and theoretical research, and it may be that such research is also favoured by peer reviewers of grant applications

• It may also be partly explained by the risks run in researching specific organisations which may respond with legal, physical, and other threats (see above)
• The skew towards studies of the far right may be partly explained by a tendency to focus on short-term government priorities, with one ‘hot topic’ at a time attracting funding (right-wing extremism having been the most recent ‘hot topic’)
• There is evidence of possibly justified concern that studying Islamist extremism might lead to a researcher’s being labelled as racist
• There is also evidence of possibly justified concern that being perceived to be critical of actions supporting ‘progressive’ or left-wing causes might lead to negative professional consequences (‘cancellation’)
• Given another study’s finding that a significant minority of academics may discriminate against funding applications on ideological grounds (whether from a leftor right-wing perspective), there appears to be a plausible risk of silencing and exclusion through peer review, especially as the typical approach to peer review of grant proposals is one that appears particularly vulnerable to bias
• Beyond this, it seems likely that fear of ideological discrimination may have a general chilling effect, potentially leading researchers to avoid engaging in projects that they suspect could lead to controversy that might be harmful to their careers

The problem of possible ideological bias in peer-review could be mitigated through adoption of open peer review (increasing reviewer accountability) or through a transparent process of reasoned adjudication between contradictory reviews on the part of an identifiable individual (increasing funder accountability). Awarding small grants through open competition with quotas for particular kinds of extremism might help to ensure that study of and expertise in key areas continues to be supported despite cyclical shifts. Universities and HMPPS could assist by ensuring that their approval processes do not obstruct public interest research on extremism.

A governmental commitment to notifying researchers when their work is cited in briefing documents or makes any other contribution would also help to ensure the continued supply of useful research, especially where this is not funded by the government. Value could be added to government-procured research through engagement of entities from across the research ecosystem in the design, analysis, interpretation, and dissemination stages of research projects commissioned from generic research providers, as well as through requirements for data sharing.

Lastly, and perhaps most urgently, the government and the Solicitors Regulatory Authority (SRA) could publicly acknowledge the threat which SLAPPs present to the sharing of information about extremist groups, their associates, and their supporters. Potential remedies might include legislation, as well as guidance issued and action taken by the SRA.

1. Introduction

1.1 Background

Although there are many fields of enquiry in which ‘pure’ research is necessarily dominant, the study of extremism is an applied social science. Extremism researchers aim to influence and inform policymakers, practitioners, and the public. To take just one area in which research on extremism may serve to inform policy and practice, the National College of Policing advocates for the use of ‘the best available evidence . . . to inform and challenge policing policies, practices and decisions’, and notes that research can help stakeholders not only to ‘develop a better understanding of an issue — by describing the nature, extent and possible causes of a problem or [by] looking at how a change was implemented’ but also to ‘assess the effect of a[n] . . . intervention — by testing the impact of a new initiative in a specific context or exploring the possible consequences of a change’ (College of Policing, n.d.). Extremism is also a frequent topic of interest for the mass media, with researchers based in universities and think tanks often contributing interviews and op-eds, or simply providing quotes as subject-matter experts. The research community thus provides a vital component in a society’s attempts to respond to the threat of extremism, however it is to be conceived. As a result, the question of how that community carries out the task of researching extremism — of the ways in which individual researchers choose directions for study, and of the ways in which they are supported or frustrated in their efforts to pursue knowledge in those directions and then communicate their findings to stakeholders and the public — is of more than academic interest. In adapting to structures of incentives and disincentives (opportunities for funding or exposure, for example) and navigating obstacles (difficulties with access to research subjects, for example) researchers jointly construct the evidence base on which policy, practice, and public opinion rely when it comes to this key area. Thus, the institutional conditions for research — the research environment, in this study’s terminology — can be seen to set the parameters within which knowledge of extremism can emerge. At a further remove, what is known about extremism is one of the principal determinants of what can be done about extremism. Understanding the national research environment for the study of extremism should thus be understood as a necessary first step in the development of an extremism research policy designed to produce an evidence base adequate to support a democratic, wholesociety approach to extremism.

1.2 Definitions

This report adopts the UK government’s official definition of ‘extremism’ as ‘vocal or active opposition to fundamental British values, including democracy, the rule of law, individual liberty, and mutual respect [for] and tolerance of different faiths and beliefs’ (Secretary of State for the Home Department 2011, 107). In common with UK Research and Innovation (UKRI), this report defines ‘research’ as ‘as any form of disciplined inquiry that aims to contribute to a body of knowledge or theory’ (UKRI 2023).

1.3 Structure and purpose

The current report presents two studies. The first of these is a quantitative survey of research grants competitively awarded by UK funders in recent years. It aims to provide a high-level summary of the extremism-relevant topics that this form of funding has supported. The second of the two studies is a qualitative analysis of semi-structured interviews with UK-based extremism researchers at all levels of seniority, both inside and outside the university system. Although the interviewees are not claimed to constitute a representative sample (indeed, it would be hard to define the population which such a sample would ideally be representative of ), their perspectives enable exploration of the complex tangle of incentives, disincentives, and obstacles negotiated by extremism researchers. These interviews are supplemented by an unstructured interview with a Home Office research programme lead, who provided insight into the interface between extremism researchers and government research stakeholders.

The report ends by trying to draw together the findings of the two studies, and by making recommendations derived from those findings.

2. Study I: Quantitative survey of competitive public research grants

The first study to be presented in this report is an attempt to map out statistical patterns in grants supporting research into extremism and awarded by UK public funders through open competition. By design, this study relies entirely upon open-source data drawn from the websites of the funders themselves.

It is acknowledged that a great deal of research is not publicly funded. Indeed, much academic research is not ‘funded’ in any formal sense, being the work of scholars whose posts are primarily supported through teaching-related income: a point to which Study II shall return (see in particular Section 3.3.3). Moreover, much public funding for research is not awarded through open competition but through procurement tendering, while a further proportion of research is funded by non-public means, including donation. This study can therefore only survey a proportion — perhaps a very small proportion — of the UK’s total extremism-related research output. But, however small that proportion is, it has special importance. Firstly, such funding is one of the levers by which the research activity of individuals and organisations may be (perhaps inadvertently) nudged in particular directions. Secondly, it can facilitate researcher-directed activities that would otherwise be very difficult to sustain.

To expand upon the first point, public research funding is one of the primary means by which research is incentivised, both at the institutional level, and — thanks to intrainstitutional reward structures — also at the level of the individual researcher. For example, a typical set of academic promotion criteria for a research-intensive university requires ‘[c]ontribution to successful funding applications’ at grade 7 (Lecturer A), ‘highly rated [independent] grant applications’ at grade 8 (Lecturer B), and ‘[r]esearch income in excess of the Russell Group median for the discipline’ at grade 9 (Senior Lecturer) (University of Glasgow, n.d.c, 2), with successively more onerous requirements at higher levels (University of Glasgow, n.d.b, 2; n.d.a, 3). Whether funders and government agencies intend it or not, competitive awards of funding thus send out a powerful signal as to the kinds of projects which researchers should attempt to replicate if they wish to be successful in their careers.

Moving on to the second point, there are many research activities which depend to a greater or lesser extent on funding. For example, some of the most important forms of data collection (such as fieldwork interviews, experiments, and representative sample surveys) may be unfeasible without some level of financial support. Where this support is an expense budgeted into a tender submission, the researcher or research team acts as a supplier and must meet client requirements very closely. A grant, however, typically allows researchers greater freedom to follow emerging insights — and will often allow them to do so over a far longer time period. Moreover, the findings of grant-supported research are likely to be published openly, often in peer-reviewed venues, contributing to the general store of knowledge on a topic rather than simply satisfying a single research client’s immediate need for information. Grant-supported research thus has many advantages over other forms of research, and, in studying the grants awarded to extremism research, one may learn something important about the conditions which make particular extremism research projects possible or impossible.

Altogether, three funders were studied. These were UK Research and Innovation (UKRI), the Centre for Research and Evidence on Security Threats (CREST), and the Commission for Countering Extremism (CCE) — the latter of which also funded this study. UKRI receives funding from general taxation via the Department for Business, Energy, and Industrial Strategy; at £7.9 billion, its 2021/2022 budget accounted for about half of the UK’s total public spend on research and development, although just £1.5 billion of this took the form of research grants (UKRI, n.d.a, 3, 8). CREST receives funding via UKRI from British intelligence and security agencies as well as from the Home Office (CREST, n.d.a); it received £4,859,620 for the period running from September 2020 to June 2024, of which £900,000 was earmarked for research commissioning (CREST, n.d.b). The CCE is funded via the Home Office, and does not primarily exist to support public research, although it chose to do so in 2019, when, under the leadership of Dame Sara Khan, it opened a one-off, rapid-turnaround public competition for multiple small grants to study particular aspects of extremism (CCE 2019). Charitable funders such as the Airey Neave Trust (a provider of relatively small grants exclusively for the study of terrorism and extremism) and the Leverhulme Trust (a provider of often very large grants for all forms of non-medical research) were not studied. The British Academy (a public provider of fellowships and small grants across the humanities and social sciences) was also not studied, because too little information was publicly available for awards made prior to 2022.

Although UKRI has been referred to above as a single funder, it is in fact an umbrella body incorporating nine research councils. In the research presented below, these are analysed as separate entities alongside UKRI, which makes an appearance only thanks to the rare instances where it, rather than one of its constituent bodies, was the specified funder of a project. Of the nine research councils, five were relevant to Study I. These were the Arts and Humanities Research Council (AHRC), the Engineering and Physical Sciences Research Council (EPSRC), the Economic and Social Research Council (ESRC), the Natural Environment Research Council (NERC), and Innovate UK. The other four research councils — the Biotechnology and Biological Sciences Research Council, the Medical Research Council, the Science and Technology Facilities Council, and Research England — were not found to have funded any research that could be included in the study.

Think-tank research is largely excluded from this study, because (while there have been exceptions) think tanks do not typically apply for research funds from these sources, for reasons that will be touched on in Study II. However, as Study II will argue, this kind of funding can act to sustain expertise within the university system, which can then be drawn upon by think tanks and stakeholders. The precise nature of the research that public funders choose to support may thus indirectly act to shape the pool of experts available for consultation or for involvement in further research not necessarily funded in the same way.

2.1 Research questions

Study I was designed to answer the following research questions:

• What roles do different UK public research funders play in supporting research on extremism?
• Which forms of extremism, in which time periods, and in which geographical locations, have UK public research funders facilitated the study of?

The study was originally intented to address the additional question of which research methodologies most frequently received public funding. However, it was found that publicly-available information on funded projects was often insufficiently detailed to enable the methodology of a research project to be identified.

2.2 Methodology

2.2.1 Search strategy

An initial search for UKRI grant-funded projects was done via the UKRI web portal because search via the Gateway to Research API (UKRI 2021) — a UKRI software interface provided in order to facilitate automated data collection — offered fewer documented options for refinement. Search strings are listed below, and were applied to title and abstract, then filtered by date range (requiring a start date between 2018 and 2022, inclusive), with results being downloaded in table format. After this, data were processed and prepared using scripts (i.e. short computer programs) written in the statistical programming language, R.

Both via the API and via the portal, the UKRI system appeared to recognise ‘extremist’ (although not ‘extremism’) as a synonym of ‘extreme’, and this apparently undocumented behaviour resulted in thousands of hits unrelated to extremism, such as ‘Extreme laser-driven hydrodynamics’ (Dior 2021-2025) or ‘Multipoint sensors for extreme environments’ (Fells 2020-2025), both of which were funded by the EPSRC. These spurious hits were removed by requesting full information for all projects via the Gateway to Research API, and then filtering out entries which did not actually feature the string ‘extremist’. Hits for the ‘extremist’ search string were only integrated following this cleaning process, which was again accomplished through coding in R. The use of a separate data-collection procedure for that particular search term was time-consuming, and ultimately added few projects to the sample, but to have avoided it would have raised concerns about completeness, as there would have been no way of knowing how many relevant projects had been missed. Total hits are given in Table 1.

This strategy located a total of 42 documents featuring the term ‘extremist’, although the initial search pulled in 3169. All other search terms received a total of 376 hits; however, after duplicates had been removed, the true total was 298. The search term ‘extremist’ increased this total by just two, with the result that the total number of UKRI-funded projects to request for download came to exactly 300.

CREST appears not to offer API or portal access to a database; instead, it maintains a website with funded projects organised by manual tagging. However, the number of funded projects in the time frame (two starting in 2018, 11 in 2019, 15 in 2020, 12 in 2021, and two in 2022) rendered manual data collection feasible. Projects were filtered by tag, selecting in turn ‘Countering violent extremism’, ‘Extremist actors’, and ‘Radicalisation’ (there was no category for ‘Terrorism’ or similar), recording project URLs where the start date fell within the range of interest, and using a web scraper written in R to download and extract project information from the files located by the URLs. Relevant details were stored in a table, with the project summary and the information contained in the ‘Show more’ drop-down being combined into a single text to be treated as an abstract. The number of hits per tag is given in Table 2. Once duplicates were removed, a total of 15 CREST-funded projects remained.

The CCE funded several research projects following an open call issued in February 2019. These were included in the current study via their published outputs, 17 of which were found on the UK government website (any uncompleted projects or projects whose outputs did not pass peer review will not have been published). All were considered relevant by virtue of having been specifically funded as studies of extremism. The only publicly available information on the projects consisted of the published papers themselves, plus the sparse information contained in their pages on the government website. Titles, abstracts, and other information were copied and pasted from these two sources into a table. Where an abstract was not included in the paper itself, the introductory section from the paper was used as a substitute. In the single case where there was neither an abstract nor an introductory section, the brief description of the paper provided on the relevant page of the UK government website was used instead.

Table 1: Hits for search strings, UKRI portal

Search string	Hits
extremist	3169 [42]
terrorism terrorist	126
communism communist “revolutionary socialism” “revolutionary socialist”	70
stalinism stalinist trotskyism trotskyist maoism maoist extremism	46
radicalism radicalisation	43
“far right” “extreme right wing”	43
islamism islamist “islamic state”	25
anarchism anarchist	10
jihadi jihadism jihadist	6
“white nationalism” “white nationalist”	3
“neo-nazi” “neo-nazism”	2
“far left” “extreme left wing”	2

Table 2: Hits for tags, CREST website

Tag	Hits
Countering violent extremism	7
Extremist actors	8
Radicalisation	6

2.2.2 Inclusion criteria

Data on a total of 332 projects were collected. However, the majority were entirely irrelevant for the purposes of this study. For example, a toxicological investigation of certain chemicals whose abstract notes that ‘these compounds [have been] used as chemical warfare / terrorist agents’ (Bury 2020-2024) is not a study of extremism or terrorism: it is a piece of natural sciences research whose importance to society is advertised partly through allusion to the potential relevance of its findings in the event of a possible future terrorist attack. Similarly, a study of cinematic representations of the Holocaust which characterises the ‘current political climate’ in terms of ‘resurgent far-right elections across Europe’ (Rixon 2020-2023) is not a study of the far right: it is, rather, a study of Holocaust memory which notes for context (again, partly as a way of advertising its wider social importance) that representatives of the broad political grouping whose past representatives include the perpetrators of the Holocaust have enjoyed recent electoral success in some European nations. There was also the question of what to do about studies concerning state terrorism when conducted at home, or states whose official ideologies would be recognised as extremist: a study of Nazi death camps or Soviet gulags is unlikely to have great relevance to UK counter-terrorism policy, for example, and it would clearly be unhelpful to include e.g. a historical study of the Iran-Iraq war simply because both sides were led by rulers whose views would be regarded as extremist in the UK. In addition, the grant which supports CREST was excluded, as CREST is here analysed as a funder in its own right (moreover, the funds comprising that grant have a different source).

A study whose details were collected as part of the search could be included in the quantitative analysis only if:

1. It prominently and explicitly identifies itself as a study of terrorism, terrorists, extremism, extremists, radicalisation, counter-terrorism, counter-extremism, or de-radicalisation (regardless of how these are defined), OR
2. It centrally concerns acts of terrorism (as defined in UK law) or individuals, groups, or movements who carry out or endorse such acts, except where the acts of terror are conducted by state agencies within the borders of the states in question, OR
3. It centrally concerns individuals, groups, or movements characterised by extremism (as defined in the UK counter-terrorism strategy) or conventionally identified as extremist (e.g. Islamist, far right, far left, etc), except where the ideology of those individuals, groups, or movements is the official ideology of the states in which they operate

For these purposes, the so-called Islamic State (i.e. the entity also referred to as IS, ISIS, or Daesh) was not treated as a state, as it was not internationally recognised as a state. That is, studies of Islamic State were regarded as studies of a terrorist organisation operating within the borders of Syria and Iraq. In order to avoid debates as to what should count as ‘extremist’ in historical contexts, projects were automatically excluded if their period of study entirely preceded 1918, i.e. a century prior to the earliest start date for the research projects themselves.

Altogether, 118 projects were found to satisfy the above criteria and thus included in the study. This set of projects is treated as the sample for the analysis presented in the remainder of that part of this report which is devoted to the current study. It is acknowledged that other inclusion criteria would have been possible and that human or technical error may have led to the unintentional exclusion of some projects, and that the data collection strategy may have failed to identify some projects that could potentially have met the inclusion criteria: for example, additional search terms might have pulled in relevant studies that were never examined here. However, it is argued that the sample can be assumed to provide a reasonably complete snapshot of UK extremism research projects which began in the five years from 2018 to 2022 and were supported by public funds allocated through open competition.

It is further emphasised that not all projects included are primarily focused on extremism. Indeed, the project with the greatest funded value was not (see below). Moreover, even those studies which were primarily intended to study movements, organisations, or ideological currents here characterised as extremist did not necessarily characterise their objects of study as ‘extremist’ — a term which many scholars of certain such movements would be likely to regard (at best) as unnecessarily pejorative. An example of a funded project on a movement regarded as extremist for the purposes of the current study, but not apparently treated as extremist by the researcher or researchers involved, is ‘Women as Transnational Agents in the Development of Iberian Anarchist Thought and Practices Relating to Female Bodily Autonomy in the Interwar Period’ (Turbutt 2021-2025). If anarchism is defined as a form of extremism, or as relevant to the study of extremism (as it is for the purposes of this study), then the outputs of research projects looking at anarchism (such as Turbutt 2022) can be considered to contribute to the knowledge base on extremism, for example by serving to inform potential literature reviews, regardless of whether or not they treat their objects of study as ‘extremist’. Indeed, such works might even contribute by helping to establish that some particular movement, organisation, or ideological current should not be regarded as extremist.

For a complete list of included projects, with start dates, funders, banded levels of funding, and titles, see Appendix I. Perusal of the titles may give the reader a perspective which goes beyond the largely statistical analyses which follow; combined with a search engine, they may perhaps serve as a useful index to recent extremism-related research projects in the UK satisfying the criteria above.

2.2.3 Data preparation, extraction, and analysis

An R script was used to extract data from the various tables and to create a uniform HTML page for each project. A Google Forms questionnaire was created on which to record relevant information for projects that met the inclusion criteria (see Appendix II). Extracted data were then downloaded in the form of a new table and combined with automatically harvested information for visualisation and analysis using R. Confidence intervals and tests of statistical significance were considered to be unnecessary (and would indeed have been meaningless) as this is in effect a ‘whole population’ study: that is, the numbers calculated do not relate to a random sample from which statistical inferences may be drawn, and with regard to which a degree of statistical uncertainty can be mathematically estimated; rather, they describe the characteristics of the complete set of funded studies discoverable via the search strategy at the time of data collection and subsequently found to satisfy the inclusion criteria. Human error and unanticipated technical issues may have resulted in the exclusion of funded projects which might have changed the analysis in some way, for which reason, statements of findings are appropriately hedged. However, the approach throughout has been to follow clearly described procedures as rigorously and transparently as possible, in order to minimise ambiguity.

Discussion of statistical patterns in the following sections is in many cases supplemented by brief focus on the project in each category with the greatest funded value. This means that CREST-funded projects, whose funded value is not public, and CCE-funded projects, which were all supported by small grants, were not used as examples. In one case, it was found that details available via the UKRI website contradicted details collected via the API. This contradiction is noted in the text below. Due to time constraints, information available via the API was otherwise treated as accurate in the main analysis, as it would not have been possible to cross-check every piece of information against the UKRI website. For projects supported by the two non-UKRI funders, there was only a single information source, so the possibility for contradiction did not arise.

2.3 Findings

2.3.1 Projects by funder

For total numbers of projects funded by each funding body, broken down by size of grant, see Figure 1. The most important funder by far, with more than three times as many projects supported as its nearest competitor, and with six times as many to have received grants of £100,000 or more, appeared to be the ESRC: one of the seven research councils incorporated within UKRI. However, it should be borne in mind that studentships constituted the great majority of projects funded by all three of the research councils found to have funded extremism research during the period in question.^{[footnote 1]} The largest number of non-studentship projects were funded in the CCE’s single round of funding, although the grants in question were capped at £10,000. The size of CREST grants is not announced via the CREST website, and is therefore recorded as ‘Unknown’.

Figure 1: Funded extremism projects by funding body, 2018-2022

Two large grants were attributed directly to UKRI, rather than to any of its constituent funding bodies.

It should be noted that the numbers are quite small — as was the total spend (so far as it can be determined). When studentships are excluded, the UKRI and its constituent research councils funded 26 projects that could be included in this study — an average of 6.5 projects per year — as compared to the total of 17 that were funded by CCE and the total of 15 that were funded by CREST. The research council-funded extremism research projects included some that were relatively high-cost, with the median value being £113,569, but the total value over the four years (again, excluding studentships) was just £6,491,101. Although this may seem at first glance to be a very substantial sum, it represents just £1 in every £1000 that UKRI and its constituent research councils awarded in the form of research grants during the same period. Moreover, 19% of it went to support a single project which was not primarily focused on extremism (more details below).

2.3.2 Projects by form of extremism

For number and funded value of projects by form of extremism studied, see Figure 2. Abbreviations used for forms of extremism are explained in Table 3. Studies specifying more than one form of extremism could be counted in multiple categories.

Figure 2: Funded extremism projects by form of extremism studied, 2018-2022

Absences are as important as presences. Thus, the reader’s attention is drawn to the fact that several forms of extremism which were treated as possible categories in data collection did not appear in the data. It may of course be that some eligible projects focusing on such forms of extremism received funding, but failed to be captured by the search terms used. However, this would only apply to UKRI-funded projects, as potentially eligible CREStand CCE-funded projects were not identified through keyword search.

In particular, the data collection strategy pulled in no projects focusing on ethnic forms of extremism other than that of the far-right, no projects focusing on forms of misogynist, sexual, or gender-related extremism other than those looking at Incels, no projects focused on single-issue extremism, and no projects focusing on extremism motivated by what is officially referred to as ‘mixed, unstable, or unclear’ ideology. The project which focused on a political form of extremism other than that of the far left was a studentship entitled ‘The role of tribalism in sub-Saharan terrorist groups’, which involved the use of fieldwork interviews and surveys in Nigeria and Kenya in order to gather new data on Boko Haram and the Mau Mau (Micheni 2018-2022). While Boko Haram are an Islamist group, the ideology of the Mau Mau was not Islamist, and thus, this project was double-counted. As an anti-colonial nationalist movement, the Mau Mau could have been treated as left-wing extremists or as ethnic extremists other than of the far right, but either of these analyses would have been problematic, so ‘Other political extremism’ was regarded as the most straightforward categorisation.

The largest group of projects comprised those that did not specifiy a particular form of extremism, of which there were 48 (i.e. 41% of all included projects). This category was particularly well-represented among the highest known-value projects: 59% of projects in the £100,000+ category, as compared to 38% of projects across the other four categories, concerned no specific form of extremism. This category also encompassed the single highest-value project included in the current study: ‘Governing Democratic Discourse: Social Media, Online Harms, and the Future of Free Speech’, funded value: £1,222,449, which (as mentioned above) was directly funded by UKRI. This project is a philosophical study of ‘a wide variety of harmful speech that is pervasive on social media networks, including hate speech, terrorist incitement, disinformation and misinformation, cyber-harassment, advocacy of self-harm, micro-targeted electioneering, foreign propaganda, and divisively uncivil political speech’ which will involve the production of ‘a combination of traditional academic outputs and innovative public resources, including an informational project website, an educational podcast, a major policy report, and a pilot civic education curriculum for secondary school students’ (Howard 2022-2026, n.p.).

The second largest group consisted of projects studying right-wing extremism. Among projects of known financial value, the highest-value project in this category was ‘Everyday transnationalism of the far right: an interdisciplinary study of Polish immigrants’ participation in far-right groups in Britain’, funded value: £682,950 (Garapich 2023- 2026). This was a three-year ESRC-supported project led by Roehampton University and involving ethnographic observation and semi-structured interviews with members of far right groups in six locations within the UK. The total number of funded projects concerning Islamist extremism was considerably lower, although the gap was mostly accounted for by the much greater number of studentships focused on the far right. The highest-value project of known value in this category was a two-year ESRC-funded project led by the University of Glasgow, ‘Mobilisation of foreign fighters in the former Soviet Union’, funded value: £472,751: a fieldwork-based study involving interviews with ‘active, former, and aspiring foreign fighters in Ukraine, Russia’s republic of Dagestan and amongst the Chechen Diaspora in Western Europe’ (Aliyev 2022-2024, n.p.). The fourth-placed category, i.e. left-wing extremism, fell far behind, with only nine studies, five of them studentships. None of the studies of known value in this category reached six figures in the sum awarded, the highest-value award being the ESRC grant to support ‘The Politics of Women’s Agency: Gender and Peacebuilding in post-conflict Nepal’, funded value: £98,687. This two-year project built on the ethnographic fieldwork carried out by its principal investigator in the course of her PhD research on the Nepali Maoist movement, and involved the production of peer-reviewed journal articles and the organisation of ‘a one-day workshop in Kathmandu . . . [and] six workshops in three rural districts in Nepal, engaging women ex-fighters and women activists’ (Ketola 2018-2020, n.p.).

There were only two funded projects looking at other forms of religious extremism, and just three looking at Incels. Each of the latter two totals included a single highvalue grant. The high-value award relating to non-Islamist religious extremism was an ESRC grant to support the seven-month project, ‘Explaining non-state armed groups perpetration of mass atrocity crimes’, funded value: £321,901 according to the API and £399,826 via the UKRI website. This project was double-counted because its list of relevant non-state armed groups comprised one Christian extremist organisation, the Lord’s Resistance Army, and four Islamist organisations: Al Qaeda, Islamic State, Boko Haram, and Al-Shabaab (Hinkkainen 2021, n.p.). Led by the University of York, the project aimed to collect systematic evidence on such groups in six countries in Sub-Saharan Africa and the Middle East, focusing both on ‘group characteristics, such as age, ideology, and external support’ and on ‘interactions, for example, between the non-state armed group[s] themselves, other actors such as the government, and external actors such as UN peacekeepers’ (Hinkkainen 2021, n.p.). The high-value award relating to Incel extremism was an AHRC grant to support a project led by University College London and entitled ‘Understanding the Cel: Vulnerability, Violence and In(ter)vention’, funded value: £201,027. This two-year project ‘seeks to save the lives of potential victims as well as perpetrators’ of Incel violence by using ‘[c]reative research methods’ to develop ‘new knowledge about the culture of Incels, the identities and experiences of this complex community, and the factors contributing to the risk of extreme violence and hate crimes’, and to use this knowledge to develop training resources in partnership with the Metropolitan Police Force and Police Scotland (Regehr 2022-2024, n.p.).

Table 3: Abbreviations for forms of extremism used in visualisations in this report

Abbreviation	Meaning
XRW	Right wing extremism
XLW	Left wing extremism
Other pol.	Other political extremism
Islamism	Islamism
Other rel.	Other religious extremism
Incel	Incel / misogynist
No spec.	No specific form of extremism

2.3.3 Projects by focus on counter-extremism or counter-terrorism

There were 44 included projects concerning counter-extremism or counter-terrorism (see Figure 3). These account for 62% of projects not focused on any particular form of extremism. Among projects of known funded value, the one to receive the largest grant was ‘The Transformation of Transatlantic Counter-Terrorism 2001-2025’, funded value: £631,587. This four-year project, one of the few which appeared to be funded directly by UKRI, is a study of organisational change in Usand EU-based counter-terrorism, using documentary sources, interviews, and ethnographic observation (Bury 2020-2024). Also included in this category were several critiques of the Prevent programme and a number of projects aiming to develop technological systems for detecting extremist speech on social networking services.

Figure 3: Funded extremism projects by focus on counter-extremism or counter- terrorism, 2018-2022

Although it would be unreasonable to go into the same level of detail on all categories discussed in this study, the above discussion should suffice to provide an indication of the diversity of publicly-funded research on extremism, as they range from projects looking at specific extremist groups to projects looking at no form of extremism in particular, from projects focused on data collection to projects focused on sharing theoretical knowledge or findings arising from prior studies, and from projects focused on specific real-world locations to projects primarily dealing with the Internet. There is, moreover, a great deal both of creativity and of scholarly rigour evident in the design of many of the projects. However, the question of whether these projects are optimally targeted to contribute to actionable knowledge about UK terror threats remains open.

2.3.4 Projects by location and by period

For funded projects by location of object of study, see Figure 4. This reveals that studies of the UK context were the most common by some margin, although it also shows that the difference in frequency between those studies and studies of the rest of the world is almost entirely accounted for by projects of low or unknown value (which is for the most part to say, projects funded by CREST and the CCE). Interestingly, there were only two projects in the highest funding band which gave no specific geographical location. For funded projects by time period of object of study, see Figure 5. By far the largest group of studies concerned extremism in the 42 years from 1980 until the time of data collection. Studies of extremism at earlier points in history were relatively rare. This suggests that the study of extremism is primarily concerned with the present and the recent past, although it is noted that studies of the Nazi and Soviet regimes were excluded from the data collection by design. It is further noted that all of the high-value grants were received by projects identifiably linked to the most recent time period, and that this category also accounted for the bulk of the projects of low or unknown value (see above).

Figure 4: Funded extremism projects by location of phenomena studied, 2018-2022

Figure 5: Funded extremism projects by time period of phenomena studied, 2018-2022

Figure 6: Funded extremism projects by funder and form of extremism studied, 2018-2022

	AHRC	EPSRC	ESRC	NERC	Innovate UK	UKRI	CCE	CREST
XRW	9	1	18	1	0	0	5	5
XLW	5	0	2	0	0	0	2	0
Other pol.	0	0	1	0	0	0	0	0
Islamism	2	0	16	0	0	0	4	1
Other rel.	0	0	1	0	0	0	1	0
Incel	1	0	1	0	0	0	0	1
No spec.	1	3	26	0	1	2	7	8

2.3.5 Intersections between categories

For funded projects by funder and form of extremism studied, see Figure 6. All funders except for the AHRC and NERC (the latter of which funded only a single study) funded more projects with no specific form of extremism than with the most commonly specified form of extremism, but this tendency was strongest at CREST, where 53% of funded studies did not concern a particular form of extremism, compared to 40% of ESRC-funded and 37% of CCE-funded projects.

The most commonly specified form of extremism was in all cases right-wing extremism. However, the ESRC and CCE funded almost as many projects looking at Islamism as looking at right-wing extremism. Thus, the large imbalance towards the extreme right is mostly accounted for by the set of projects funded by the AHRC, which funded just two studies looking at Islamism, and by CREST, which funded just one: ‘Mapping terrorist exploitation of and migration between online communication and contenthosting platforms’, a one-year project of unknown funded value at King’s College London which focused exclusively on the online realm (Maher 2020). Studies of leftwing extremism were mostly funded by the AHRC — a point which may be explained by the time period on which they were focused (see below).

For funded projects by time period and form of extremism studied, see Figure 7. This shows that left-wing extremism was the only form of extremism primarily featured in projects focused around the years prior to 1980, where it was more often represented in funded projects than right-wing extremism, and where projects looking at no specific form of extremism were not represented at all. This may account for the unusually frequent appearance of left-wing extremism among AHRC-funded projects, as the academic discipline of History falls within the AHRC’s remit. Studies of Islamism during the period were effectively absent, as the only project to fall at the intersection of the two categories was the aforementioned studentship focused on Boko Haram and the Mau Mau: this was correctly double-counted for both period and form, but the Islamist group in question came into existence in the post-1980 period, while the pre-1980 group was not Islamist.

Figure 7: Funded extremism projects by time period and form of extremism studied, 2018-2022

	Pre-1980	1980-present	Unclear / not applicable
XRW	6	34	1
XLW	7	5	0
Other pol.	1	1	0
Islamism	1	23	0
Other rel.	0	2	0
Incel	0	3	0
No spec.	0	41	7

For funded projects by location and form of extremism studied, see Figure 8. Interestingly, the aforementioned dominance of projects investigating the extreme right is much more pronounced with regard to UK-focused studies: among the latter, more than twice as many funded projects looked at the far right as at Islamism, which was in turn the focus of twice as many funded projects as the next-largest category, the far left. By contrast, in studies looking at identifiable locations outside the UK, there were nearly as many studies focused on Islamism as there were focused on the far right. (The sole CREST-funded project on Islamism had no particular geographical location, being the aforementioned study of how Islamist extremists have migrated between online platforms, Maher 2020.) Perhaps unsurprisingly, projects associated with no particular time period or location were also overwhelmingly associated with no particular form of extremism. Interestingly, these projects appeared to have attracted few grants confirmed to be of large or medium size — although projects with no clear location included a disproportionately high number of grants of unknown value (which is to say, grants awarded by CREST).

Figure 8: Funded extremism projects by location and form of extremism studied, 2018-2022

	UK	Rest of world	Unclear / not applicable
XRW	20	17	9
XLW	4	6	0
Other pol.	0	1	0
Islamism	8	14	3
Other rel.	1	1	0
Incel	1	0	2
No spec.	27	9	20

Fully representing the intersections between forms of extremism, locations, and time periods in a single visualisation would be excessive. However, if we look only at projects dealing with the UK from 1980 until the present day, as in Figure 9, we find that there were 55 publicly-funded projects, of which 49% looked at no specific form of extremism, while 35% looked at far-right extremism, and 15% looked at Islamist extremism, with 7% looking exclusively at other forms of extremism than the top two (note again that percentages may fail to sum to 100, as some projects focused on multiple forms of extremism, e.g. both Islamism and on the far right). Excluding studentships from the total does little to change that pattern, leaving 32 publicly-funded projects, of which 53% looked at no specific form of extremism, while 28% looked at far-right extremism, and 12% looked at Islamist extremism, with 9% looking exclusively at other forms of extremism.

Figure 9: Funded extremism projects focusing on the UK from 1980 onwards by form of extremism studied, 2018-2022

Thus, when the focus is on the UK of the present day and the very recent past, there appears to have been little publicly-funded research focusing on specific forms of extremism other than that of the far right. Indeed, if we discount studentships, all such research identified via this study’s data collection strategy was funded by the CCE — with the exception of a single high-value project focused on Incels (i.e. Regehr 2022-2024, discussed above).

2.4 Discussion

Perhaps the first finding to note is the small number of publicly-funded studies of extremism, and the very low percentage of public research funding which appears to be directed towards such study. Although it has been acknowledged that some funded projects may have been inadvertently left out of the sample for technical reasons or as a result of human error, and also that one provider of mostly small grants (i.e. the British Academy) could not be included in the study because of data availability issues, the true figures are unlikely to be much higher than those found to be through the methodology followed here. This is not necessarily a bad thing: whether extremism research has received too much, too little, or the right amount of public financial support is essentially a political question on which it would not be appropriate for this report to comment.

The next finding to note is that, at least as represented in the sample analysed here, publicly-funded studies of extremism appear to have a tendency not to focus on any specific form of extremism. For example, there are studies of ‘extremist’ online communication or of ‘extremist’ recruitment that do not specify which form or forms of extremism they will investigate. Moreover, where studies in the sample do specify a particular form of extremism as being of interest, it tends to be right-wing extremism that they focus on — at least when they are looking at the UK of the present day or the recent past. Again, whether this is or is not as it should be is essentially a political question, and, as such, outside the scope of this study.

It should be acknowledged that projects which do not concern any particular form of extremism may still be useful: for example, a number of such projects which were included in this study focus on British counter-extremism policy, or on processes such as radicalisation in the abstract. Nonetheless, we can still ask whether there are enough detailed studies of specific extremist movements and organisations, or of policy responses undertaken in relation to these, being carried out. That question cannot be addressed here, but this study has at least served to establish that, if such research is underway, it appears likely to be underway with little support from the UK’s public research grant infrastructure — especially where the topic is Islamist extremism in the contemporary UK.

This brings us to the question of whether the apparent hierarchy of forms of extremism, with right-wing extremism at the top and Islamism in a distant second place (unless we consider the remoter past, or other parts of the world) is as it should be. Whether it is or is not is, again, a political question. However, it is perhaps worth recalling the observation, by the Independent Reviewer of Prevent, that ‘[i]n the years since the [Islamist] 2017 Westminster Bridge attack, the vast majority of realised and foiled [terror] plots have been Islamist in nature’, and that, at the time when the independent review was conducted, ‘80% of the Counter Terrorism Police network’s live investigations [we]re Islamist, while 10% [we]re Extreme Right-Wing’ in nature (Shawcross 2023, 14). There may, of course, be very good reasons for publicly-funded research on extremism in the present-day UK which concerns specific forms of extremism to focus primarily on the UK’s second-greatest present-day terror threat, and only secondarily on the UK’s primary present-day terror threat (study of which appeared to be funded, not only through a far smaller number of grants, but also through grants that were exclusively in the lowest financial bracket). However, if the intention was not to produce such an outcome, funders and stakeholders might wish to consider means by which to achieve a different balance of funded research, and, in this light, it is perhaps worth noting that CREST — a body set up specifically in order to study security threats to the UK — appears in recent years to have funded exactly as many studies of the Incel movement, which is not regarded as a terrorist threat in the UK (Shawcross 2023, 53), as of the UK’s principal terror threat.

No criticism of the funding bodies themselves should be read into this. Whether the relative proportions of funded research accounted for by various forms of extremism are regarded as appropriate or not, it is impossible to know whether patterns associated with particular funders are better explained by biases in the selection process or by the research interests of the communities which typically apply to those funders for support. That is: if a particular funder very rarely supports a particular kind of research, that might be because its selection procedures result in a high rate of rejection for applications for grants in the service of such research, or it might alternatively be because it receives low numbers of such applications in the first place, whether because researchers carrying out such research have other sources of funding which they tap into preferentially, or because they do not consider themselves to need funding for such research, or because they perceive (rightly or wrongly) that their applications would be unlikely to succeed. Thus, the lack of studies of certain forms of extremism in certain contexts may have many potential explanations, whether relating to the peer-review process, to the nature of the applications submitted for review, or to something else entirely.

In this context, it may be worth recalling the observation that ‘[m]any [counter extremism] practitioners who wish to focus on the principal terror threat to this country . . . find themselves viewed with suspicion even by colleagues’ (quoted Shawcross 2023, 101): if similar pressures operate within the extremism research community, it may be that researchers simply choose not to submit proposals for funded projects looking at Islamism in order to avoid negative judgements from their peers (potentially including their peer-reviewers). The current study can provide no insight into that possibility (although Study II, below, attempts to do so). However, it is noted that a recent survey of university staff found that 9% of academics would give what they perceived to be a left-wing grant application a lower rating, and that 22% would do the same for what they perceived to be a right-wing grant application (Adekoya, Kaufmann, and Simpson 2020, 64). If research on right-wing extremism is understood to be ‘left-wing’, while research on Islamist extremism is understood to be ‘right-wing’, and research on ‘extremism’ in general is perceived to be politically neutral, this might explain the dominance of projects focused firstly on no specific form of extremism and secondly on right-wing extremism.

In this connection, it is observed that, where grant applications are assessed by calculating a numerical average of ratings across a panel of peer reviewers, with only the very highest-scoring proposals going forward — as at CREST and the UKRI funders — even a single negative review will generally be sufficient remove a project from contention. If roughly 9% of potential peer reviewers would discriminate against a proposed study of the far right for being ‘left-wing’, then, given a standard ESRC panel of three peer reviewers (UKRI, n.d.b), the probability of such a proposal’s encountering at least one such reviewer is around 25%, equating to odds of approximately one to three, and if roughly 22% of potential peer reviewers would do the same against a proposed study of Islamism for being ‘right-wing’, then the probability of such a proposal’s encountering at least one such reviewer is around 53%, equating to odds of approximately one to one.

Academic journals also employ peer review, of course, but they usually have higher success rates, and in any case do not typically make decisions on the basis of a numerical average, instead requiring editors to adjudicate between contradictory reviews. This would not necessarily eliminate bias, but it would at least ensure accountability for any bias which remained.

Here it is also perhaps worth noting that grant applications are typically only subject to single-blind review. This is a form of peer review in which the identity of the authors or applicants is known to the reviewers at the time of review, but the identity of the reviewers remains permanently confidential. Studies of academic journal and conference peer-review processes have found that, under a single-blind system as compared to what is usually considered the gold standard in academic publishing, i.e. the double-blind system where author identities are confidential at the time of review, ‘reviewers may be [negatively] biased towards authors that are not sufficiently embedded in their research community’, thus restricting the advance of knowledge by limiting the influx of new ideas (Seeber and Bacchelli 2017; see also Tomkins, Zhang, and Heavlin 2017). Doubleblind peer review may not be practical for grant applications, given that applicant curricula vitae and publication histories are an important component of the applications themselves. However, a further alternative exists in the form of open peer review, in which the identity of the reviewer is revealed to the community at large on the ethical grounds that ‘it seems wrong for somebody making an important judgment on the work of others to do so in secret’ (R. Smith 1999, n.p.). Early studies in medical and psychiatric journals found that open peer reviews were at least equal in quality to double-blind peer reviews (Rooyen et al. 1999; Walsh et al. 2000), but the main advantage of open peer review is simply that identification gives reviewers ‘skin in the game’ (Haffar, Bazerbachi, and Murad 2019, 673), ensuring that they may be held accountable for any bias which they might display: a consideration which may have special importance in a field with such potential for ideological conflict as extremism studies.

Whether research on specific forms of extremism is indeed perceived in ideological terms is a question to which Study II shall return. In the meantime, it is observed that anti-fascism has been argued to be such a key element of progressive political identity as to constitute the very ‘core of left-wing consciousness’ (Diner 1996, 124), while a report by a British counter-extremism think tank both identified Islamophobia as a characteristic of the political right (HNH 2019b, 3; 2019a, 22, 24; Lowles 2019, 8) and appeared to blame its prevalence on public consciousness of Islamist terror attacks (HNH 2019b, 3; Carter 2019, 14, 16). A more recent report led by a researcher affiliated with a UK civil society organisation also asserted that, in the UK, ‘[t]he stereotyping of Muslims as terrorists and extremists [by the far right and the centre right] has fed into dangerous government decisions and policy that have targeted and stripped Muslims of basic rights’ (Syeda and Molkenbur 2023, 12–13). This study cannot attempt to assess the validity of such claims. Instead, it simply offers them as evidence of the political battlefield into which researchers may be perceived (and may understand themselves to be likely to be perceived) to sally forth in proposing to study specific forms of extremism, rather than ‘extremism’ in general: a topic about which more shall be said in Study II. The suggestion made here is that, even if the Scylla of alleged anti-progressive ideological discrimination presents less of an anticipated risk than the Charybdis of alleged anti-conservative ideological discrimination, the optimal strategy would still be to avoid mentioning anything that might conceivably fall afoul of either — for example, by putting together a proposal which addresses ‘extremism’ in the abstract, i.e. without giving any suggestion that some particular form of extremism might be a focus.

Given that the CCE has so far only issued a single call for funding proposals, it is remarkable that it was found to have funded the largest number of non-studentship projects, and indeed all included non-studentship studies of forms of extremism other than those of Incels and the far right as manifest in the UK since 1980. It is, however, important to recognise that the sums awarded by the CCE were small: indeed, there were UKRI-funded projects whose funded value exceeded that of all CCE-funded projects combined. Large grant-supported projects can potentially achieve a great deal more than smaller ones, with members of staff devoted to multiple workstreams over extended periods of time. For example, the largest grant included in this study — identified as relevant because of its focus on terrorist incitement alongside a number of social problems less directly associated with extremism — will run for four years, and involve production of a wide range of outputs. By contrast, lower-value grants are in many cases sufficient only to support a short programme of activities designed to share existing knowledge, or to facilitate some specific activity, such as travel to a location where fieldwork or archival research will be carried out. But for scholars not belonging to large, permanent research teams, the stitching together of a succession of such grants may be crucial to their production of knowledge and development of expertise. This is to say that, while large grants have a special importance, both large and small grants are important to the research ecosystem, with each having a distinct role to play.

In context of its potential to become an important supporter of small projects on extremism, the author of this study notes the CCE’s achievement of near balance between Islamism and right-wing extremism in the studies which it funded through its open call. This balance is easily be explained by the terms of that call, which specifically requested projects addressing a range of specific forms of extremism. For example, applicants were invited to submit proposals for studies of the far-right group, National Action, and the Islamist group, Al-Muhajiroun (CCE 2019). It is noted that calls for funding proposals from CREST tend to be far more general, with titles such as ‘Evaluating the cumulative impact of HMG’s counter-terrorism communications’, ‘Contagion of extremism’, ‘Environment and interventions’, and ‘Conspiracy theories and extremism’ (CREST 2021, 14–16). It is observed that, if very general proposals have an advantage at peer review over those which mention specific forms of extremism, they may tend to win out unless they are excluded by the terms of the call — regardless of whether their advantage consists in reduced potential for ideological disagreement or simply in an academic preference for the abstract and the theoretical.

That said, it should be acknowledged that the ESRC also achieved near-parity between Islamism and right-wing extremism in terms of number of projects funded, without having explicitly called for studies of particular groups — although apparently also without awarding a single non-studentship grant to a study of Islamism in the contemporary UK or in the UK of the recent past. While it is possible that such grants were awarded but did not come to light as a result of the methodology employed here, their absence from the sample studied here suggests that they are likely to be rare at best. Possible reasons for this apparent rarity have been discussed above, and shall be returned to in Study II.

3. Study II: Qualitative analysis of interviews

Although the first study in this report has provided a description of certain general patterns in publicly-funded research into extremism, it can only offer speculative but hopefully plausible interpretations for those patterns. In talking to extremism researchers and stakeholders, however, an interviewer may elicit explanations of why things happen in the way that they do. The second study presented here is more freeform than the first, consisting of a thematic analysis of a number of semi-structured interviews. Interviewees have been allowed to speak in their own words, as far as possible, with connections drawn between their accounts where points of comparison or contrast emerged.

3.1 Research questions

Study II was designed to answer the following research questions:

• What sort of considerations do UK-based extremism researchers take into account in deciding what to focus on in their work?
• How do UK-based extremism researchers understand the external factors which facilitate or obstruct them in their work?
• To what sort of ideas do UK-based extremism researchers appeal in addressing the question of whether certain areas are underor over-researched?

3.2 Methodology

3.2.1 Selection of research subjects

Potential research subjects were initially identified through the researcher’s personal network, with some additional subjects being suggested by the Commission for Countering Extremism. The list of interviewees was then diversified through searches of UK university and think tank websites for individuals with a specialism in extremism research. Interviewees were contacted via the researcher’s personal email account. Non-response, and problems in scheduling interview times and dates, limited the range of interviewees in ways that would not have been possible to predict. Altogether, 16 researchers were interviewed (ten male and six female), along with a further interviewee based at the Home Office and representing the perspective of a major institutional research stakeholder. No claim can be made that these interviewees constitute a representative sample of the UK extremism research community. However, the sample included both university and think tank employees, and a mix of early-career, mid-career, and senior researchers, as well as a single independent researcher. Some interviewees focused primarily on Islamist or Jihadi extremism, some on right wing extremism, some on the Incel movement, and one on a form of religious extremism other than Islamist extremism. There was considerable crossover between these groups, and one interviewee researched extremism in general, without focusing on any particular form, while another directed research projects on multiple forms of extremism.

3.2.2 Conduct of interviews

Interviews were conducted online, in most cases through the researcher’s personal Google Meet account. Interviews were digitally recorded and professionally transcribed. The schedule of interview prompts is presented in Appendix III.

3.2.3 Analysis of interview data

Emergent themes were identified in interview transcripts. An initial list of 13 themes was narrowed down through merging of smaller themes until six remained for summary and quotation.

3.3 Findings

3.3.1 Contested borders

One of the most important themes to emerge from the interviews concerned the boundaries of extremism studies itself. For example, some interviewees made an explicit case for the study of non-violent extremism to be treated as a core part of the field, arguing that it has been neglected, and some called for extremist crime to be acknowledged as a subject of enquiry, regardless of whether or not it was violent. One spoke as follows:

[R]esearch on non-violent extremism has been always relegated to the footnotes and academic literature, although it is the identity that definitely sparks specific behaviour. So I do believe that more focus on that will be beneficial to understand terrorism itself . . . [but] for many years, it was considered as not relevant. I remember in 2014, when, you know, I was researching on Hizb ut-Tahrir, some people were telling me ‘So why don’t you research Al Qaeda”, or — it wasn’t the time of ISIS — then 2015, right into focus on ISIS. And my stance was, ‘ISIS has almost the same ideology as Hizb ut-Tahrir and it’s one group. But there are so many groups that share the same ideological foundation, they’ve . . . been around since the fifties, but have inspired different terror groups all over the world. And still nobody talks about them. So it’s about time to focus on the ideologies . . . inspiring . . . other organisations.

In other respects, however, arguments or accounts of conflicts over boundaries were more pointed. Several interviewees expressed the view that simply studying Jihadism can result in a researcher’s being labelled as racist or Islamophobic by others in the field, with one of the most senior suggesting that this had caused university-based researchers to avoid researching the topic. Conversely, one interviewee suggested that, in some cases, it may indeed have been Islamophobia that motivated a focus on Islamist extremism:

It’s clearly a priority to understand [Islamist extremism], but I do wonder if there is still somewhat of . . . an element that’s racist, but maybe that’s too strong, but you know, because maybe the kind of far-right white supremacist misogynistic stuff is obviously more likely from kind of white communities and we’re a British country. I do just wonder if it’s less prioritised as a kind of research? Sometimes, I’m not sure. I suspect maybe.

Two interviewees complained of an obligation among extremism researchers to create a false equivalence between Jihadism and right-wing extremism, with one scholar of right-wing extremism arguing that the appearance of equivalence is buttressed through ‘number acrobatics’, for example by counting incidents rather than fatalities. A researcher primarily focused on Jihadism argued as follows:

There’s like a conceptual, artificial manufacturing of equality . . . A few years ago, you’d have to do two slides on ISIS or something in your presentation and then two slides on the English Defence League. And it’s like, well, English Defence League, as unpalatable as you might find them, is not a terrorist group, and certainly not a genocidal group. So already, you’re kind of morally and intellectually muddying the water a little bit there.

A more senior researcher argued that this had resulted in relative neglect of Islamist extremism as an object of study:

It’d be hard to sit here and say that there isn’t enough research or focus on Islamist terrorism and extremism. There is a lot of research. But I think [that], proportionate to the threat that it poses in this country from both a terrorism and extremism perspective, possibly there isn’t enough.

On the other hand, several interviewees argued that Islamism had been over-researched, with one explaining as follows:

I would say, . . . we really went down a radicalisation rabbit hole when it came to radical Islam, to a degree that was — considering, considering the impact on broader society, and to — considering, I would also say, the security risk, it’s an over-researched area.

Although some of the people to whom the researcher spoke emphasised the greater physical danger posed by Jihadis, this interviewee viewed that as a misleading argument, as no form of extremism poses a particularly great physical threat to the UK population, as compared to other threats:

There’s no doubt [that the focus on Islamism] was a reaction to the threat. [But] also, it’s also a reaction to the priority that’s put on it.  [T]he actual threat in terms of human life, for example, is   often not proportional to where that threat lies in the priorities of government. And it depends on a whole series of political reasons, initiatives, initiatives, or considerations. . . [T]here’s a snowball effect that happened, that is not, certainly in terms of terrorism, not proportional to the threat to human life, for example, that is represented in the UK.   [T]here are few countries where you can say terrorism really represents a threat to the daily life and functioning of a country.   Iraq would have been one of them, Afghanistan would be another. But in most countries, and in most Western countries, it doesn’t challenge the functioning of society. But we decided that it represented a major threat to our national security. And that had a huge effect on the next twenty years.

This viewpoint (about which more will be said below) emphasises that the question of which extremist threats should be researched is always political, and cannot be answered by a simple appeal to numbers. In addition to extremism researchers’ apparent frequent reliance on funding from state agencies, this implies that the field itself has an inherent relationship to politics which may subject researchers operating within it to ideological conflicts of a more overt kind than would be likely to be encountered in many other fields of study. The consequences of the field’s politicisation were emphasised by two senior think tank researchers. One gave the following characterisation of the field as a whole:

The research landscape is also slightly drawn on political lines. And I think, you know, the kind of, where the focus is slightly reflects your institutional politics. That’s probably as much the case with academics as it is for think tanks as it is for civil society organisations. And so, . . . it’s the threat from Islamist groups that have insidious sort of activities within communities in the UK, or whether it’s looking at big national security type threats, whether it’s . . . that Russia is sort of supporting the far right extremist movements across Europe, or whether it’s . . . looking at the sort of creeping influence of far-right populism . . . [and] all of these things kind of sit within the boundaries of extremism as a sector . . . [so] that the challenge is that . . . what you’re focused on reflects your own perspectives on what is important and significant   So I think [that] all these things are the sort of subtext to . . . how the funding works [and] how the conversations happen. It’s . . . a very political set of topics that we’re dealing with. And that’s quite . . . hardwired into the approaches that are taken.

The other suggested that the politicisation of the field explained its greater current focus on the far right than on other forms of extremism:

I think probably there’s also a political interest, you know, a lot of young people are left-leaning liberals. And so   studying the far right is our generation’s version of fighting the Nazis  fighting the good fight. I think you see that a lot in advocacy groups, activist groups, like [REDACTED], where there is a blurring, I think, sometimes of legitimate research, and — well, it’s kind of exposés, isn’t it, which I have mixed feelings about, of the far right — bleeding into political activism, promoting a [left-wing] vision of the world, and not wanting a right-wing vision of the world. So there’s probably an element of of that, as well.

The anxiety that politicisation might be distorting the field was widespread among interviewees. Moreover, a degree of scepticism about the validity of exposé-type research on the far right carried out by activist researchers was also expressed by the other think tank researcher quoted above, who, despite representing an organisation which other interviewees recognised as left-wing, and which maintains connections with ideologically-motivated independent researcher-activists, emphasised the importance of maintaining a separation from such work for the sake of credibility. One scholar of the far right who explicitly identified with the political centre-left went so far as to argue that a damaging orthodoxy was emerging within the research community, with left-wing or progressive politics becoming a prerequisite for study of the far right:

[The assumption is that] you have to be centre-left to work on the far right, or far-left to work on the far right. You can’t be somebody who’s a conservative who’s disgusted by the far right, you’re not welcome. And I think that message is very clearly pushed that . . . you have to sign up to certain values . . . you have to hold these views, and holding these views is a prerequisite to doing the work. . . .

Antifa, anti-fascist action, was formed by the Communists in 1932. To some extent, it still contains . . . still maintains that sort of international network of brotherand sisterhood of the far left. Even though ‘anti-fascist’ for some people can simply mean what it says on the tin, ‘I’m against fascism’, . . . I think most people rightly would say, it also means something specific: an ideological activism against fascism and the far right.  It’s ‘I hate the fascists, so I’m going to research them as a better way of crushing them’. And   you know, people who research child pornography, I’m sure it’s not a big stretch to say they’re also against it, but they don’t necessarily need to say ‘I’m a campaigner against this, that, and the other’, because it only just undercuts the dispassion which even those people would be expected to bring to their work.

So I think that, now, one of the things   [that] I would say to you, I have really been hearing from bigger stakeholders, governments, . . . the UN . . . [is] ‘Can we even trust the research that comes out of universities any more on far right extremism? It looks so biased, that it doesn’t seem like we can trust it any more.’  And that to me, would be a catastrophe for the study of extremism. If the people who are writing on extremism aren’t trusted to be unbiased and dispassionate — I’m not so naive as to say ‘objective’ or ‘true’, I think we’re past all that — but people who are striving to be dispassionate, to represent counter-arguments faithfully, that sort of stuff. If that is evacuated from the subject, it would be a catastrophe.

Such complaints echo the argument made in Section 2.4, that patterns of attention and neglect in funded research might be to some extent be driven either by ideological discrimination, or even by anticipation of ideological discrimination. Political differences were also apparent in discussions of far-left and ecological activism, over which interviewees were divided, with some feeling these shaded into forms of extremism unjustly neglected by scholars, and others fearing an official drive to associate such activism with extremist violence. One interviewee — an insecurely-employed researcher focused on far-right extremism — perceived that, within the extremism research community, there were strong professional incentives not to look at forms of extremism associated with ‘progressive’ causes, and suggested that colleagues might refuse to work with a researcher who had done so.

There was also division among interviewees over whether misogyny should be treated as a form of extremism. Some interviewees suggested that the far right had been defined too broadly, especially in an alleged willingness on the part of some researchers to identify mainstream conservative figures with the far right. However, others suggested that it might in fact have been defined too narrowly, with one interviewee expressing concern that subtle expressions of extreme right-wing ideology, as embodied, for example, in selective readings of British history, were passing under the radar and not being recognised as extremist. Given that the interviewees were effectively one another’s peers, these divisions may be taken as evidence for the proposition that peer review in the field is politically charged. This may have implications both for publication and for grant awards, since both depend on peer review as a quality control measure: while there was nothing to suggest that any of this study’s interviewees would rate papers or proposals unfairly, their comments provide evidence that certain kinds of work are perceived as ‘left-wing’ or ‘right-wing’, and, as discussed in Section 2.4, there is evidence that a substantial minority of academics are willing to discriminate on precisely such ideological grounds. This possibility is returned to in the following section.

3.3.2 Funding

Although the researchers interviewed for this report included senior academics based at leading British universities, none had received UKRI funding. One interviewee went so far as to state that ‘it’s very unusual to hear that somebody is conducting research on [extremism], and [that] it’s funded by UKRI’. A number of interviewees raised issues over peer review. For example, one senior researcher took the view that the inherently interdisciplinary nature of extremism research made assessment by an adequatelyqualified peer review panel unlikely when applying to a UKRI research council. Some academic interviewees had only ever received small grants, but nonetheless appeared to have played an important role as expert witnesses and unpaid providers of advice to various state agencies. A number seemed to have given up on UKRI, with one stating as follows:

I’ve been trying, trying for years. I applied . . . to the ESRC.   And it was a well-articulated research project   it was well structured, well built, everything was logically linked. And also it had support letters from practitioners across the country. The data were there. Access, the gatekeepers, were there. I think I’m well qualified to   run the project. The research team was brilliant, and   the [feedback] said it was great, but it just didn’t pass the [numerical] threshold.

It is hard to say whether this experience differs from that of academic researchers in other subject areas. The experience of receiving positive feedback and yet failing to achieve a high enough average score once the ratings awarded by all peer reviewers have been aggregated may indeed be universal, as it is a consequence of this particular way of assessing proposals, which — when combined with the high threshold implied by a low ratio of available funding to applications — effectively gives the power of veto to each individual reviewer (see Section 2.4).However, it has been argued in that this power of veto may be particularly problematic for peer review of grants related to the study of specific forms of extremism, as these may have an unusual (and possibly asymmetrical) propensity to attract discrimination on an ideological level: an interpretation that receives support from interviewee comments quoted and summarised in Section 3.3.1. In this context, it is noted that one interviewee went so far as to suggest that research on forms of extremism associated with causes perceived as ‘progressive’ would not be supported by conventional funders, as grant applications would be rejected by peer reviewers.

From where, then, does funding for extremism-related research typically come? A senior think tank researcher stated that private donors often ‘aren’t happy to say they’re funding extremism work’, seeing extremism as a ‘contested, difficult area’ with which they would not want their names to be associated. Several university-based interviewees had been funded by CREST, but another senior think tank researcher found its application process to be excessively ‘cumbersome’ adding, ‘we really can’t afford to dedicate that much time for [such] a low success rate’. A number of smaller funders were mentioned, but it appeared that governmental and (in some cases) police and military agencies were the most important funding source for extremism research, in some cases even where such research was carried out under the auspices of Critical Terrorism Studies: a field centrally concerned with critiquing state responses to terrorism. The author of this report notes that these findings suggest a marked difference from other research fields with which he is familiar, and may perhaps serve to set extremism studies apart from UK research culture in general: funding assigned directly by governmental or intergovernmental agencies, as opposed to bodies such as UKRI, appears relatively unusual throughout most Humanities and Social Sciences disciplines. In this context, it is worth drawing attention to the views of the Home Office interviewee, who repeatedly emphasised that the management and funding of research on security and extremism was under-resourced:

The research base is often not well-funded. It’s often not a priority. I mean, certainly, if you’re looking at Homeland Security, more operational activity is necessarily a higher priority. It’s more urgent, and the consequences are more serious. So things like . . . longer-term research, that might support and be fundamental to policy development [are] naturally not going to be in [the] top three items in anybody’s list.   [T]here isn’t the resource to put the effort in.

The latter interviewee stated that the effects of this under-resourcing were manifest not only in the relatively small sums available for commissioning external research on the topic, but also in the low number of civil servants available to manage the procurement process or to keep track of relevant research carried out independently: a key problem given the applied nature of most extremism research, whose social value often depends upon its reaching an audience in the policy community. That is, the study of extremism in the UK is highly dependent both for funding and for impact on an entity which is structurally unable to prioritise it in terms of investment or even attention. The interviewee continued as follows:

So for us, if we’re going to commission money, commission research and spend money, our choices are effectively to go through an existing contract, or an existing framework, or to compete the requirement openly. And that last option takes time and absorbs resources to do because you have to create the requirement, you have to do expressions of interest, you have to you have to determine in which commercial framework it goes on, you then have to put it out to tender. There are timeframes waiting for bids to come in, and you then have to assess and score all the bids, you then have to notify all the partners, write contracts – you lose a lot of time if you’re gonna compete each individual requirement as it comes along, which is why we have both frameworks and call-off contracts. . . . In our space, we have contracts with people like [REDACTED] and [REDACTED] and [REDACTED], [REDACTED], people like that. So those frameworks have been as call-off contracts and those cases have been competed. They are then given for a value-added amount of time, they’ll be like two million pounds for three years. As long as the work you want to do fits in that space, you can go directly to them. And they price it according to a rate card, and it’s a much faster process. But it’s — it’s narrowing your choices, to ‘Well, we’ve got these . . . frameworks, or you have to go and do the whole competition thing yourself’. You know, [INAUDIBLE] will go ‘There’s the easy option or there’s the hard option’. People go: ‘That’s an easier option’. And they’re not always the best fit. But they are convenient. And it’s — reasonable quality. So that, that’s — that’s that model.

So the other one is, is frameworks. The one that is, is public is ACE for the Home Office. So the Accelerated Capability Environment. So that, that is, it’s effectively a subcontract framework.   And we take our problem, our requirements to them, they put it out to a pre-approved community, it will come back with shaped bits and costed bits. And then we will select from them to provide what we want. Mainly that’s for the data and digital space, because that’s what they were created for. But they have an academic research community as well. So that provides a mechanism to go out to the academic community to the university community without us having to do an invitation to tender ourselves. So we’ve got, moved the contractual challenge. One step removed, so it’s quicker for us. Again, that framework was competed as a large scale framework. And the winning bidder is then operating on a rate card to deliver a service. But what we really really can’t do is go ‘Oh, we’ve got this piece of research, it’s about 125,000 pounds for a year’. We can’t just go to the university and go, ‘You guys have got the expertise. You’re the known experts in this, we’re just going to give you the contract’. Because it’s anti-competitive, it doesn’t meet government transparency requirements. And equally, if we put that out as a tender to all universities   often, they’re quite, they’re relatively small pieces, so the universities will often go ‘The timing doesn’t work’   or they go ‘The amount of money isn’t worth it to do all the work’.

In this context, the importance of CREST becomes particularly clear, as, from the government’s point of view, it effectively plays the role of a subcontractor for the research contracting process, and as such not only reduces the need to rely on large generic research providers whose expertise may not be ideally suited to every task

(a point on which another interviewee, a former government agency researcher, also expressed doubts), but does so without imposing an excessive administrative burden. Moreover, CREST provides an infrastructure for the direct communication of research findings to stakeholders in government. (A further advantage, of which the interviewee may not have been aware, is that the problem of not being able to recoup full costs for small projects is less likely to arise for universities under this approach. This is because funds distributed by CREST are treated as grants, and universities account for grant-funded research differently, not regarding themselves or their employees as contractors for the grant-awarding body, but instead regarding the latter as providing financial assistance to them in their mission to produce publishable research.)

The same interviewee suggested that a further solution to some of these problems might be greater data-sharing, with commissioned datasets being made available to multiple teams of researchers. This, the interviewee emphasised, would help to aoid duplication of effort and expenditure (roughly equivalent datasets having been collected multiple times and yet never directly compared), but also of providing second and third opinions with regard to the analysis and interpretation of data. This might serve to mitigate the problem which appeared to be hinted at in the faint praise which this interviewee gave to the work done by large non-specialist research organisations (‘reasonable quality’). Such an approach might also solve a problem highlighted by almost all of the researchers interviewed for this report, i.e. lack of access to data (especially given the limited availability of direct funding with which to support the collection of fresh data).

3.3.3 Timeframe

Multiple interviewees drew attention to problems relating to the time frame in which research can be completed. One of the highest profile interviewees described a research process involving several years of data gathering and analysis and culminating in the publication of ‘an ideas book, where you lay out your argument in an accessible and compelling way’. His or her experience was that ‘[n]obody reads articles’, but ‘a surprising amount of people either read books, or flip through them, or want to be aware of the latest books’, and that this can often open doors for the author of a book that has successfully attracted interest at the right level. This particular interviewee had achieved an impressive level of influence in this way, but it was clearly an approach requiring a high level of commitment over a long period of time. By contrast, many interviewees presented themselves as being much more pressed for time. The Home Office interviewee summarised the problem as follows:

I’ve seen research that’s more sort of fundamental underpinning and long term. And it is, in my limited experience . . . government has a problem waiting for the answer   Going ‘Well, I’m gonna have to wait three years for that’ is not terribly appealing in terms of funding, because funding by government is always more difficult when it goes across more than one financial year. That’s just the nature of government financing rules, in terms of what you can fund, and also   being really blunt, what will often happen is the person who commissioned the research will have left the role by the time that research is completed. And there is always the risk that the . . . new incumbent does not see the same value or importance to that work. Or, quite frankly, the world has moved on, and it’s not as important as it was when it was commissioned. It’s the nature of work that you can’t do in a shorter period of time. Everybody would like the answer in about three months, and, particularly in the academic space, three months, is still gathering data, let alone doing any analysis or providing any insight. But the policy drive is ‘We want an answer next week, to our question’. And that doesn’t line up very well, with the academic environment.   So there’s that question of how do you fund stuff that’s short and dirty, because that’s what the policy customer wants? What do you actually — extremism is a classic example. There is some really important, foundational, fundamental research that needs to be done to support an issue that is not going away anytime soon. But for the people who are commissioning it to fully understand that, they’re gonna have to wait for those answers, because to do the research well and to have integrity and a solid, suitable evidence base, they’ve got to be patient. And it’s . . . not in their nature, and it’s not in policy’s nature. And ministers will go, ‘This is a problem now, tell me [the] answers now’. Ministers like a nice evidence base behind those answers. But if you go, ‘We’ve been getting an evidence base, but it’s nine months to 12 months before we’re gonna be able to tell you the answer’, that doesn’t — that doesn’t land well. In terms of responding to issues, there’s a definite tension between those two things. Good research takes time, particularly when it’s foundational. The customer who we are . . . designed to support   is not looking at that same timeframe. It doesn’t work for policy requirements.

Think tanks, on the other hand, are explicitly set up to provide the policy community with research insights in a form and on a time frame suited to the latter, and interviewees who had spent time both in that world and in universities recognised the difference between the two in this regard. However, one interviewee who had given up a think tank job to study for a PhD argued that the pressures of the fast-turnaround think tank world resulted in lower quality research:

I guess the actual style of work in a think tank is very different. So [my think tank work] was research, but it was also putting events together   It was a real mix of you do your own research, but you also organise a roundtable next week and you also have to — get funding, so I was really actively working on . . . big bids [for funding].  It was just very, very busy. So in terms of the actual work, . . . the actual research output  definitely wasn’t as academic, you do much more kind of, y’know, briefing reports, 2000 words here, 5000 words there, they’re that much quicker turnaround. But I felt that it was much more quickly fed in to the people that we wanted it to be fed into. So every week   we’d be hosting some kind of roundtable which had 20 different stakeholders from that industry that we’d written a report about. So maybe quality of research maybe wasn’t as — maybe wasn’t as good. But ultimately, it had all these recommendations that were maybe quite useful that were getting fed in — whether or not we made much change, I’m still not convinced really   So I did become a little bit disillusioned with that style of working, I actually felt like, I don’t think the research is that rigorous. And sometimes I thought we were all just saying the same thing.

The ‘sameness’ of much research output on extremism was referred to by other interviewees, and might perhaps be explained by the time pressures of the policy world, to which extremism research is (as this report has already noted) very closely tied — even when it is not conducted by think tanks. One think tank researcher complained that extremism research ‘happens in cycles, like there’ll be a year or two, a year or a few month period where everybody publishes articles on the same subject’, while a senior figure in a different think tank complained that because research is ‘driven by the funding landscape, which [doesn’t] really ha[ve] the capacity to do multiple things at the same time’, there had been ‘a huge run on Salafi jihadist research’ between 2015 and 2018, followed by ‘a huge flip to the far right . . . as if . . . the  baton was passed from ISIS to   groups like National Action’, with the result that ‘there really is relatively little focus on what — what the [Islamist] threat looks like now after the territorial decline of ISIS’. As that particular interviewee observed, this made it difficult for think tanks to build up expertise in a wide range of areas, with the result that it is difficult for them to be prepared for changes in policy priority resulting from unpredictable future events. These various objections suggest that, at any one point in time, most forms of extremism will be under-researched: an issue that might perhaps not matter if extremist threats stood still (assuming that each form will get its time in the spotlight, eventually). However, multiple interviewees emphasised that such threats continue to evolve.

It is thus arguable that society and the state rely on university-based and independent researchers to pursue their individual research interests without public support, or with only minimal public support, as insurance against the blindsiding that inevitably happens when ignored or forgotten threats come to the fore: academic interviewees in particular appeared to operate by searching for gaps in the intellectual market, which they would then seek to fill over a period of many years, drawing on whatever sources of financial support they were able to find (see Section 3.3.2). Such long-term investments of effort pay off in the event that a chosen area of expertise becomes a matter of policy interest, but the very conditions of employment which enable university-based researchers to operate in such a way may also prevent them from capitalising effectively on such opportunities. That is, academics are able to pursue non-income generating activities during their allocated research time precisely because this accounts only for a minority of their working hours, the majority of which are taken up with activities associated with teaching and administrative work. Thus, university-based researchers whose expertise — developed over the long term — has become relevant to government, military, or security service needs often struggle to satisfy the corresponding demands for evidence and analysis, because the same arrangements that make it possible for them to develop that expertise also make it harder for them to comply with stakeholder temporal expectations. One interviewee spoke as follows:

[O]ne of the difficulties — and this is why this is why  full-time staff [in] universities don’t tend to enter in this area — is that there tend to be very, very short deadlines    So, NATO will need something within the next three months or within the next six months. They don’t, they can’t wait three years for it. Because that’s not how their funding cycle works, right? So a lot of it depends on how much time I actually have in that window. And I’ve had to sign up for a lot of work. Because   I have teaching that term, or because I have a series of other projects   that I need to finish. So I think   that’s definitely something that is a serious consideration that I have to take. Can I do this within the next three months? The second thing is, can the university free up my time within the next three months? And is it willing to do so? And that’s a really difficult one as well, because we have increasing . . . teaching [and]   administrative responsibilities, [which] means that  it’s very difficult for us to just say, ‘Well, I’m going to spend a month doing this’, or ‘I’m going to spend the next two weeks in Mali’.

On the other hand, another interviewee complained that, even in the academic world, relative short-termism is imposed by the demands of the Research Excellence Framework or REF, and lamented that ‘[t]he temptation is always   to knock out a small self- contained, you know, research article  on a small kind of self-contained study’ rather than a big, richly descriptive book such as that which the first interviewee quoted with regard to this theme saw as the ideal means through which to communicate ideas. A period of several years’ research, potentially spanning more than one REF period, can be provided by a UKRI grant, but, as Section 3.3.2 of this report has indicated, these were perceived to be relatively inaccessible to extremism researchers.

It should be noted that the situation was very different for the independent researcher who was interviewed for this study. Provided with a roving brief by anonymous backers, he or she was able to work in a highly agile fashion, moving flexibly between multiple related projects over an extended period of time, with the confidence that funding would continue as long as results continued to be demonstrated.

3.3.4 Access to data and research subjects

As noted above, the Home Office interviewee lamented the way in which datasets were collected at great expense, employed in single projects, and then lost to view. Access to data and research subjects from whom to collect it was raised as a problem by multiple interviewees, with availability, or potential availability, of data being a primary consideration in deciding whether or not to research a particular topic. One of the most senior academic interviewees even opined that, when it comes to the radicalisation process, ‘almost all the questions  [in which] the public and the politicians are [most] interested   cannot be answered’ because of difficulties in collecting or accessing relevant data. One think tank researcher cited clearance to access government data as a motivation for accepting government work that otherwise lacked interest. Three researchers — one employed by a think tank, and two employed by universities — complained of difficulties in collecting data with regard to one of the most important groups of research subjects in the UK from the point of view of the study of extremism, i.e. the offender population, due to problems in working with HM Prison and Probation Service (HMPPS). In some cases, these were exacerbated by the time pressures imposed by government funding, which have already been discussed in this report:

I’ve experienced a few issues with data access, regarding offenders. Especially where you need approval from HMPPS to . . . interview practitioners working in prison intervention services, as well as to access specific records on offenders. So, yeah, I think this put me off a little bit, from kind of extend[ing] my research also, as much as I could in the offending population. As you know, it’s very difficult. I’ve been rejected a few times. And then I realised that it’s got nothing to do with the research . . . [but is] because . . . I’ve asked for data such as, you know, health records or family background, that apparently [are] not stored in one place. So, personally, the people, the staff would have to go and retrieve . . . this data from all the places where they are, and they don’t have resources, and they don’t have time to do it. And when I [am working on] a government-funded piece of research, they want — their timeframes are shocking, it’s three months, or six months. So basically, . . . this is very hard for me to kind of approach practitioners and say . . . ‘I need this’, because if I had maybe one year, they would say yes, but then, as the time — the timeframe is very [short], it creates issues with data access, and all of that.

In other cases, however, interviewees suggested that access would be impossible regardless of the timeframe:

In Britain, to my knowledge, there’s no one interviewing . . . the multitude of imprisoned Jihadist extremists, which strikes me as . . . crucial. Like, because because we don’t really know who the Jihadists are on the outside, right? Like we don’t, we don’t know the supporters of Jihad are in the general population. . . . Whereas . . . [in a prison], you’ve got people that we know have been decided in a court of law to be Jihadist extremists. Obviously, it doesn’t use that terminology. And the limited reporting information we get out of the out of the prison system, in Britain doesn’t paint exactly an encouraging picture of what’s going on in there. So I think it’s crucial, actually, that people are allowed to access, or at least some people, or at least one person is allowed to access . . . that population.

Another researcher interviewed for this report had attempted to interview incarcerated extremists, but been refused access:

I approached the Home Office, the Prison Service. And they . . . cited a whole bunch of reasons………………………. They said ‘If you wrote, it would risk re-traumatising victims, if you then write an article on — on a particular offender and he refers to his crime’, whatever……………………… Yeah, so I thought that was odd……… I guess, if you talk about their crimes maybe and you write about their life story, maybe victims, family members of people they . . . hurt, they would……. think that you were excusing their behaviour or whatever. I didn’t really understand it, but they’re just very cautious about letting outsiders into prison to actually speak to these people. But in many cases, they don’t have direct victims, because it’s people who’ve attempted to go to Syria, or have come back and don’t have any immediate victims in the UK who would then be traumatised by any academic report that’s written on them. I found that very odd. So I just think the Prison Service is incredibly risk-averse and protective and . . . doesn’t want outsiders sniffing around, I guess.   [So] I haven’t been able to get any kind of access. But . . . [Hugo] Micheron in France has been interviewing loads of these guys.

Other interviewees suggested that the difficulty of accessing extremists explained the prevalence of studies of online extremist discourse, which some interviewees considered to be excessive. That said, university-based researchers dealing with online data may also face roadblocks — albeit that these are more likely to be placed by their own employers’ ethics boards than by gatekeepers such as HMPPS. One computer scientist complained as follows:

My university told me that   if I look at comments made by individuals, that’s personally-identifiable information . . . [And] I’ve seen   the idea that we need to get permission from all the individuals that appear in a Twitter scrape that we do. And that’s just not feasible. I’ve had a colleague in another university, she was a female researcher, she wanted to write a book about incels, and   one of the admins on one of the incel websites agreed to post an anonymous survey onto the forum   And the university told her that she needs to speak to each of the individuals to fill in the survey to get their consent to use the data. Now they said this to a female researcher, who’s researching a group who hates women.

Although not all university-based interviewees had experienced problems with ethical approval, perhaps indicating the diversity of practices between universities, refusal of or delay in ethical approval was repeatedly cited as an obstacle to access to necessary data. In some cases, refusal appeared to be motivated by paternalistic concern for research subject wellbeing. One interviewee had to base an entire PhD thesis around publicly available data because the extremists who were its subject were deemed to be too vulnerable to speak to:

Obviously, you had to go through the whole kind of ethics system with [the university] to make sure everything was legit.   There were things I wasn’t able to do, for instance, like actually directly approaching someone online and asking to interview them. And that was something that was [judged to be] kind of too dangerous . . . for potential participants.   I suppose I was able to get enough [data]. But perhaps in a different world, you know, I’d have been able to directly access them, it might have been more insightful. . . .

When I was approaching them and anything went wrong, could I trigger something bad happening? . . . Perhaps if I hadn’t [flagged that],   [the ethics office] might have just thought it’s a very straightforward research [project]. But because I flagged it, it then became a much bigger process of concern. And  the risks were seemingly too big.

A more senior researcher at a different university was similarly prevented from issuing a questionnaire to self-identified members of one particular movement on the grounds that ‘ask[ing] participants whether they agree with the actions of [a particular mass killer] . . . [could prompt them to] go out and physically harm [a member of the same group targeted by that killer]’. As one of the researcher’s colleagues observed, there is no reason to think that such an outcome is more likely than the reverse, i.e. that answering the questionnaire might prompt vulnerable or potentially dangerous participants to seek help. However, this argument apparently had no impact. The same interviewee is likely to have spoken for many extremism researchers in analysing the process as follows:

One of the things I study is something called ‘motivated reasoning’. So it’s this idea that people . . . apply vigorous rational standards in different ways, depending on whether they like the thing or agree with the thing . . . that’s in front of them. And we’ve really hit a brick wall with some of our research for, I believe, precisely those reasons. So the issue that we have with ethics is that you have someone who is a human being who is not an expert in what you’re studying, who gets to make a choice about the risk level of that research. And they, essentially, are not asked to justify that decision. So if they’re given some research they don’t agree with and that they don’t like, they can say, ‘I feel like this is very high risk’, and therefore it becomes very high risk. And therefore a two-week ethical process, in the case of our work, turns into a four-month ethical process where it gets . . . gets criticisms levied at it . . . that [would] apply to all psychological research.

Drawing an analogy with the ‘Just So Stories’ of Rudyard Kipling, in which fanciful scenarios are invented in order to explain why the world is as it is, this particular interviewee argued that such criticisms originate ‘where someone has gone, “I don’t like this topic, I think it’s scary, I don’t understand it, so I’m going to make up some sort of story as to why it’s very high risk” ’. The implication was that, rather than impartially assessing the risks presented by a research project in order to inform a decision, individuals involved in the ethical review process sometimes produced such assessments of risk post hoc, in order to justify a decision that had already been made on different grounds.

On the other hand, a number of interviewees (including the interviewee who provided the quote above) suggested that university ethics boards were so preoccupied with bureaucratic processes designed to minimise the potential for legal exposure and reputational damage that they under-emphasised researcher wellbeing, or failed to address more genuinely ethical issues such as how to present potentially controversial findings so as to minimise the possibility of malicious use. The paradox of ethical approval processes that fail to engage with fundamental ethical questions was set out as follows by one interviewee:

Often, ethics is taken as a technical exercise of, you know, ‘Here’s my consent form and here’s what I’m gonna say’. And that’s fine. But I think, when you’re dealing with questions, such as the ones I’m dealing with, there are more fundamental questions, such as ‘Which kind of actors do you engage with and which kind of actors do you not engage with?’ Right? . . . I did a project for [REDACTED] on [REDACTED] on . . . the potential for negotiations with [REDACTED].   [But] engaging in a negotiation with a group such as [REDACTED] presents other issues, right?   I’m not overestimating the impact of my research, but I think you still have to ask yourself the question of . . . ‘What are the objectives of this group?’ . . . ‘What are the dangers or potential ethical problems with, uhm, facilitating the possibility of negotiations with this group?’ And that becomes a far more complex political and ethical question than ‘Where are my consent forms?’

Overall, there thus appeared to be a sense that some universities and state institutions, including within government, could engage better with questions of the public interest, and of how to promote this with regard to the collection of data.

3.3.5 Devaluation of the particular

The tension between the general and the specific in extremism research was a theme emerging from several interviews, sometimes in combination with a sense that there was too little research being done that really engaged with the specifics of particular extremist groups or movements. One senior think tank researcher complained that ‘mechanisms of understanding [that] were [initially] applied to   how a group like ISIS operated’ were ‘almost   transferred wholesale to the far right’ in ‘an attempt to   create a sort of established frame that moves from one threat to the other’. This complaint was presented in relation to the field’s already-noted tendency to cycle from one topic to the next: a tendency which some interviewees attributed in turn to the nature of the funding upon which it relies. Because funding from state agencies is awarded on a short-term basis with one topic at a time having priority, multiple spheres of extremism-relevant expertise may be difficult for think tanks to sustain. Thus, there is an incentive to treat work done on one form of extremism as a ready-made framework for understanding another, once the priority changes. This could potentially imply relative neglect of anything particular to some specific threat.

The pressures faced by academic researchers are different, but may have a similar outcome. One such pressure simply related to the forms of publication which carry greatest prestige within the university system: as one of the most senior academics interviewed lamented, the top peer-reviewed journals look for work that is ‘applicable to a wide range of domains’, and will thus publish theoretical studies and even ‘very, very technical improvements to a statistical model . . . [that] can be used in   many different areas’, but will dismiss studies of the specific and the particular as purely ‘descriptive’. As that interviewee emphasised, ambitious researchers tend to engage in research that is more likely to lead to publication in high-prestige venues — which is to say, generalist and theoretical research.

Given such a situation, it is perhaps telling that the only interviewee with no complaints about the structure of the extremism research field and its surrounding infrastructure was an academic whose research was at the highest level of generality, and was therefore valued not only by state agencies, but also by the academic system. Some other academic interviewees appeared to regard communication with the world outside the university as an alternative to academic recognition that would not be forthcoming, with one stating that he or she had given up on all prospects for promotion. However, there was considerable uncertainty with regard to the question of whether and when such communication had been effective. For example, one interviewee only discovered that a particular state agency had been trying to apply his or her work only contacted for advice at a relatively late stage.

The independent researcher spoken to for this study was the only interviewee whose work was focused entirely on specifics, being concerned not with larger narratives or theoretical accounts, but simply with the identification and analysis of active extremist networks in the contemporary UK, and with the exposure of key actors within those networks. This was explained as follows:

There is a lot of quite important long-term and theoretical work to be done. But . . . I think that in some cases, [my kind of work is] what’s going to be really quite useful, because some of these people have never really faced strong criticism before. . . . [I]t’s kind of like . . . ‘Look at this guy over here. . . . Let’s actually put [the theoretical work] into practice. . . . We’ve all come up with definitions of extremism, we’re concerned about extremism. Well, here is an extremist. And here’s what you might be able to do about him.’

That particular researcher’s methodology consisted in collecting evidence, compiling it into reports, and then presenting those reports to relevant authorities and to trusted journalists. Such research does not aspire to provide a general or overall understanding of any particular extremist phenomenon, but has the potential for direct and immediate application of a kind that much academic and think tank research does not. Indeed, the independent researcher’s answers to questions were illustrated with multiple examples of cases where his or her work had achieved direct, practical impact.

3.3.6 Obstacles to publication

Interviewees identified some extremist groups as having a tendency to act to silence any unwelcome press. A particular concern was the alleged use of spurious legal threats in order to keep certain kinds of research out of the public domain. Such threats, sometimes referred to as ‘lawfare’ or ‘intimidation lawsuits’, are perhaps most accurately described as strategic lawsuits against public participation (SLAPPs). Interviewees emphasised that SLAPPs do not need to result in a court victory in order to achieve their goal of silencing criticism or eliminating unwelcome exposure. Indeed, SLAPPs are not intended to reach court at all: defined by the Solicitors Regulatory Authority (SRA) as ‘the bringing or threatening of proceedings in order to harass or intimidate another who could be criticising or holding [the complainant to] account for their actions and thereby discouraging scrutiny of matters in the public interest’ (SRA 2022, n.p.), SLAPPs typically involve cases that have little or no legal merit, but would simply prove too expensive or exhausting for the complainee to defend. One interviewee lamented the effectiveness of SLAPPs in censoring negative stories about extremist groups in the press, warning of the impact that Islamist pressure groups in particular had on newspapers, especially given falling revenues in the digital age: while his or her experience with the BBC had been that it possessed ‘the resources and the money and the determination and the competence’ to defend itself against lawfare, he or she emphasised cases where newspapers had ‘trashed’ their own journalists’ work rather than face the expense of fighting a SLAPP. As he or she explained, the result of such a capitulation is that ‘[y]ou end up in a situation . . . which is worse than where you started, because [now] you’re seen as this Islamophobic liar’.

A senior interviewee who had worked both in government and in multiple think tanks argued that legal challenges have a chilling effect on what can be said by researchers in many different contexts:

I once was involved, when I was really junior, in a piece of work which was conceived as being a kind of mapping of Islamist organisations . . . And we did some first drafts. Again . . . I was really junior at this point. And they got sent out to organisations, and . . . I had a file . . . yay high . . . of threatening legal letters. In the end, nothing came of that project. . . .

[I]f the government were to [carry out a piece of work like that], you have got parliamentary privilege . . . and you can say [things] in Parliament . . . and be immune from legal action. But . . . you can’t then put the Home Secretary on the Today programme to repeat the claims, because they’re not immune . . . from that once they’re speaking outside of Parliament.  The thing is, it’s not libellous if it’s true. But that doesn’t stop people making a claim. And fighting a claim is incredibly expensive and time-consuming and demanding on resources. And the Civil Service is not geared up for it.

Silencing attempts are not limited to the legal realm, however. One interviewee spoke of the risks that a female researcher faces with the publication of personal information, stating that this sometimes had ‘almost made [her] think maybe [she] shouldn’t be doing this’. Another spoke of being targeted online by supporters of an extremist group after publishing a report about that group, and explained that this had led to a decision to publish on ’broader phenomen[a]   instead of the groups themselves’from that point onward, in order to avoid attracting their attention.

However, the worst harassment by far was suffered by an interviewee who had not only received death threats from members of extremist groups, but also threats in relation to his or her elderly parents, who were known to the members of some of the groups that he or she was writing about. That particular interviewee spoke as follows:

And you do question yourself . . . We have   security cameras put up on the outside . . . But it’s one of those things   if you constantly fear that your work is going to lead to more threats or intimidation etc. You completely step back and say ‘I’m not researching this community at all anymore. I’m going to do something completely different.’ Or you just kind of dig your heels and say ‘I’m gonna carry on’. I am carrying on, I know, [but] I am sitting on certain . . . research papers that I’ve written because I am too worried about the fallout from them.

In one case, this interviewee had received a threat which directly referenced the names of individuals who had been assassinated by members of the movement to which the individual issuing the threat belonged. The risks faced by the independent researcher were also highly credible, because of his or her work investigating organisations controlled by a hostile state that has issued specific threats with regard to its critics in the UK. His or her main defence against targeting is anonymity — which is not available to researchers who publish directly — and a level of training to which few academics or think tankers will have had access:

There’s some risk, but I’m not so worried. . . . I did a lot of . . . counterterrorism work [when I was young]. So . . . I know how to stay safe.  I know how to recognise covert surveillance.

It thus seems that multiple factors could act to prevent the publication of research that mentions concrete individuals and groups: a problem which may interact with the above issues faced by anyone who wishes to research specific rather than general phenomena. Findings can be presented to stakeholders in confidence, but (as noted above) academic researchers are incentivised to engage in research that will lead to publication in highprestige venues, while think tank researchers need to publish regularly in order to remain in the public eye. One researcher even argued that media coverage was necessary in order to motivate the authorities to take action on problems of which they were aware, or to ‘jolly things along’ in the face of delays. Moreover, the Home Office interviewee observed that speaking to researchers directly is time-consuming from a stakeholder point of view and relies on the unremunerated labour of the researcher, also emphasising that information shared with one government entity does not necessarily reach others. While no interviewed researchers complained about lack of compensation for addressing stakeholders — indeed, they appeared to regard an audience with stakeholders as a form of compensation in its own right — the preferred option was in most cases to do so on the basis of an organised body of research that had been or would soon be published (for example, in a form such as the ‘ideas book’ referred to by one interviewee; see Section 3.3.3).Difficulties with publication therefore have major implications for the study of extremism.

3.4 Discussion

Think tank researchers and academics alike emphasised different factors affecting choice of research direction. All researchers spoken to hoped to achieve impact or influence either on policy or on public discussion. However, those working in think tanks appeared more proactive when it came to achieving such impact and influence, and had both the ability and the obligation to ensure that knowledge transfer happened rapidly and in a highly targeted manner, while those employed by universities generally played a longer game, each aiming to develop a unique area of expertise over many years. Both of these approaches appeared to have disadvantages: on the one hand, there was unease about the quality, originality, and diversity of much rapid-turnaround research; on the other hand, the academic approach takes years to bear fruit, by which time, the need for the ‘fruit’ in question might no longer be so pressing. Indeed, even the typical threeand six-month turnarounds required for government-funded research often appeared to be too long for politicians, while even the multi-year cycle of the UK’s national research quality audits, now known as the REF (the most recent three iterations of which were carried out in 2008, 2014, and 2021), could be considered too short-termist for true scholarly excellence.

The depth of expertise developed over such an extended period provides a pool on which state agencies can draw, although their ability to do so may be compromised by the mismatch between the timescales involved in government contracting and the processes and working practices employed in the university sector. Moreover, the breadth of the pool of academic and think tank expertise may be threatened by lack of support for areas at times when their study does not align with current policy priorities. Simple acknowledgement that impact has been achieved in some way — for example, through incorporation in a briefing — may in some cases be sufficient to reassure researchers, their employers, and their funders, that effort and investment has not been wasted in those wilderness years. Given the potential importance of such feedback — which not only signals to the researcher that his or her work has proceeded along useful lines, but also enables him or her to claim credit for having served the public interest, for example to an employer or funder, or even in the REF (which is partly designed in order to reward research which achieves ‘effect on, change, or benefit to the economy, society, culture, public policy or services, health, the environment or quality of life, beyond academia,’ REF 2021, n.p.) — it is striking that no formal expectation or mechanism exists in order to facilitate its transmission, with researchers in some cases only learning that their published work has had impact only when they are invited to share findings directly, or to provide expert advice.

There were deep concerns with regard to funding. Grants awarded through open competition skirt many of the problems associated with government funding, but particular difficulties were perceived in attempting to secure such funding through the UK’s primary research funders, i.e. the constituent bodies of UKRI. It should be noted that there was no clarity about what these difficulties were, although problems with peer review were raised, along with suggestions of ideological bias on the part of potential reviewers. Moreover, complaints about the onerous nature of CREST applications should be put in context of the far more onerous UKRI application process, which requires completion of a 25-page online application form with multiple multi-page attachments, and whose success rate has now fallen to 21% (Baker 2021): CREST does not publish its success rates, but application is via a comparatively simple offline form which (not counting researcher curriculum vitaes and stakeholder letters of support) might run to a mere dozen or so pages when complete. Difficulties in securing funds from private donors may perhaps combine with these issues to produce an unusual degree of dependence on government, police, and military funders, which is in turn felt to be problematic, both in terms of the short-term turnaround required for the resulting work and in terms of the necessary focus on a narrow menu of rapidly changing policy priorities, rather than on a diverse range of ongoing issues.

The greatest expressed anxieties arose over the alleged politicisation of the field. Such politicisation would be very hard to avoid, given both reliance on the sources of funding mentioned above and the inherently political nature of extremism. This point was raised by those interviewees who argued that, while Islamist extremism represented a greater numerical threat to life in the UK than other forms of extremism, this alone would not justify the attention paid to it. It should be emphasised that, while this point of view may appear counter-intuitive, it is coherent: a country and its inhabitants may face many different threats at any given time, and the question of how to prioritise and respond to those threats is never self-evident, always political. A government could, for example, theoretically choose to ignore all extremist threats in favour of the far greater numerical threat presented by traffic accidents: there were 1560 reported road deaths in 2021 (HMG 2022, n.p.), which is exactly 30 times the 52 deaths by terrorism in 2005, considered to have been the country’s deadliest year of the 21st century so far with regard to terror attacks (Casciani 2018). That this is not done is a political choice — and if it is a political choice to focus on extremism at all, then to prioritise one extremist threat over another is even more so. With regard to number of fatalities, Islamist terrorism leads the field, but it is a political choice to give this metric greater weight than others, such as (for example) numbers of Prevent referrals (that these numbers may to some extent be politically determined in their own right, see Shawcross 2023, 50–51, may serve to underline this point), or indeed to give any numeric measure greater weight than more abstract notions, such as perceived threat to community cohesion (however that is to be assessed). The politicisation of extremism studies, noted by a number of interviewees, was therefore probably inevitable, especially given the inherently political nature of the category of ‘extremism’ itself: all extremist groups have political goals, and ‘extremism’ can only be defined in relation to a politically-determined set of norms, whether identified with the values of a particular nation state, as in the UK (see Section 1.2), or with some more universal notion, such as human rights (see Webber and Struthers 2019). However, interviewees appeared to orientate towards a more specific form of politicisation, their evidence appearing to support the speculation, presented in Study I (Section 2.4), that patterns of research attention in the field might reflect an ideological conflict between the political left, right, and centre.

Given such a level of politicisation, and given what appear to be fundamental ideological disagreements between interviewees, it is perhaps unsurprising that, both in the academic world and in the think tank world, extremism researchers did not necessarily trust other extremism researchers to provide even-handed judgements. This lack of trust has obvious practical implications, not only for peer review of publications and grant applications (a point raised in Section 2.4), but also for the very notion of expertise, which by its nature depends on a degree of consensus: what one sociologist has called ‘[t]he recognition, socially marked and guaranteed . . . which the competitor-peer group bestows on each of its members’ (Bourdieu 1975, 25). This is to say that, if researchers in a given field recognise or refuse to recognise each other’s contributions for ideological reasons, it may be unusually difficult for outsiders (for example in the policy community, or the media) to identify the ‘experts’ whose opinions will best represent the state of knowledge in that field. Because the question of who is to be regarded as an expert in the field is political, not only in relation to the politics of the field itself (as in all fields of scholarship), but also in relation to the politics of the state (which funds the work done in this particular field in an unusually direct way), and of the conflict between that state and the groups who are the field’s object of study (and are defined as such by their rejection of a politically-determined set of values), there is probably no non-political solution to this problem. Thus, if stakeholders were (for example) to agree with interviewees who suggested that the research literature on Islamist extremism was becoming out-of-date as a result of recent focus on the far right, they might be better advised to stimulate such research through targeted funding than to assume that the research ‘market’ will correct itself naturally. The alternative effectively amounts to a strategy of waiting for the next terrorist incident and then commissioning research reactively.

Interviewee suggestions that researchers could face professional consequences for taking the wrong political line should be taken seriously in the light of recent events at the Centre for Analysis of the Radical Right (CARR), whose director resigned after a campaign sparked by his approval for publication of a blog post expressing concern at what the post’s author described as the ‘infiltrat[ion]’ of the counter-extremism field by ‘activists . . . who advocate for committing criminal offences in furtherance of their opposition to the radical right’ (McCann 2022, n.p.). The blog post was deleted and its author’s fellowship at CARR was apparently terminated. Whether or not the reader agrees with the position expressed in the offending blog post, the apparent success of the campaign indicates that interviewee concerns about what is popularly termed ‘cancellation’ may not be entirely misplaced. This is, of course, an extreme example, but attention is drawn to the argument that ‘[w]here someone’s views are manifest in their work, and the political composition of a sector is tilted in one direction, there is . . . a danger of structural discrimination’ arising from the cumulative effects of ‘a slightly lower mark on a grant application here, a bit of resistance to a promotion application there’ (Adekoya, Kaufmann, and Simpson 2020, 70, 68). That argument was made with regard to British universities, and to the social sciences and humanities as a whole, but it has particular relevance for those who study a topic as inherently political as extremism — whether in universities, in think tanks, or in the position of a researcher who operates independently, yet still needs a degree of institutional support in order to achieve optimal impact. It will, after all, be rather more of a challenge for researchers to avoid manifesting their political views in their work if that work concerns contemporary extremist movements of any specific variety than if it concerns, say, the archaeology of the early Neolithic period.

Difficulties collecting relevant data were often raised. These were largely seen to be caused by institutional factors, such as lack of understanding of the particular needs of extremism researchers by university ethics committees, and difficulties accessing information on or potential research participants via the prison and probation system. These difficulties might be partly alleviated through better data sharing, especially where collected at public expense: various research organisations are contracted to collect and analyse data which could helpfully be subjected to secondary analysis. However, there currently seems to be no way of making such data available, nor even of advertising its existence — thus, institutional change may be required. The sometimes apparently insurmountable challenges which interviewees encountered in attempting to access extremist offenders or data thereon through the prison system may likewise remain unless HMPPS can be encouraged to see the facilitation of legitimate research on extremist offenders as part of its remit. It should be acknowledged that research does happen in prisons, and sometimes addresses extremism: a secondary concern, for example, in a recently-published monograph on Islam in British and other European prisons (Wilkinson et al. 2022). However, the author of this study’s limited personal experience of carrying out prison-based research matches that of interviewees in suggesting that access to inmates is impeded by what appears to be an imperative for prison authorities to avoid any potential controversy, even on seemingly innocuous topics. It seems plausible that a research project which specifically announced itself as a study of extremism amongst inmates might swiftly encounter roadblocks from an institution with understandable reasons for reluctance in anything that might generate publicity. This must be seen in context of the Independent Reviewer of Terrorism Legislation’s report on terrorism in prisons, which not only raised alarm about the problem of inmate extremism, but complained that reliable information on the scale of that problem was impossible to acquire, and warned of the lack of a mechanism by which even ‘a senior official within HMPPS, the police or MI5 [would be able] to see where Islamist group activity occurs, or is on the rise, within England and Wales’ 104 public prisons’ (Hall 2022, 14).

These difficulties may go some way to explaining the prevalence of studies of online discourse. There are definite advantages to working with public online data, due to their low cost and ease of acquisition relative to interview data or representative sample surveys. Moreover, in common with other ‘naturally occurring’ data, their collection ‘does not flood the research setting with the researcher’s own categories (embedded in questions, probes, stimuli, vignettes, and so on)’ and ‘does not put [research subjects] in the position of disinterested experts on their own . . . practices, thoughts, and so on . . . (as many interview and questionnaire studies do)’ (Potter 2002, 540). However, unless they are collected from private forums (raising major ethical issues), the content of online extremist messages will be limited to what extremists consider appropriate and interesting to discuss publicly, and without data from other sources with which to triangulate, there is no way to gauge the representativeness of what they choose to discuss with regard to ideas held by or circulating among real-world extremists. One recent study argues that, because participation on major social media platforms is substantially determined by demographic variables, ‘[s]ocial media data cannot be used to generalise to any population other than themselves’ (Blank and Lutz 2017, 741), and a systematic literature review has concluded that, with regard to public opinion research, ‘social media data should supplement, but not replace, traditional methods and data sources’ (Reveilhac, Steinmetz, and Morselli 2022, 10119). If online discourse cannot be assumed to be representative of opinion within large and comparatively well-understood social groups in the absence of corroborative evidence gathered by more conventional means, still less can it be assumed representative of opinion within extremist movements or organisations.

The qualitative evidence examined here supplements the numerical evidence provided in Study I by likewise suggesting that the knowledge base may be skewed towards very general studies of extremism. Because interviewees provided such detailed accounts of their research practice and their understanding of the context within which that practice has developed, the current study was able to go beyond Study I in providing direct (albeit anecdotal) evidence of factors which may have contributed to the production of this skew. These factors include the fast-changing nature of the funding landscape on which think tanks rely, which practically requires the rapid repurposing of expertise, as well as the reward structure of academia, in which specific studies of anything may be perceived as having lower scholarly value than widely applicable theories operating at a high level of generality. Other factors may include awareness of the difficulty of publication in the face of possible legal threats (even if spurious), and the possibility of becoming the object of harassment and physical threats — even death threats.

In view of the importance which several interviewees attached to the problem of legal threats, it is worth noting other areas in which SLAPPs are routinely used to prevent information being shared which particular groups or individuals would prefer to stay out of the public sphere. At the time of writing, legislation is currently being debated on the use of SLAPPs to suppress journalistic and academic reporting of economic crime: a tactic which has proved effective because of the extreme wealth of some of the perpetrators or alleged perpetrators. In a recent debate, David Davis MP used parliamentary privilege in order to expose one of the most egregious examples, one aspect of which had (unusually for a SLAPP) reached as far as a court decision, at the same time illustrating the general problem by highlighting how a single person represented by what he called ‘the go-to law firm for every bad actor seeking to undermine or misuse British justice’ had issued spurious but potentially very expensive threats to The Guardian, Private Eye, Chatham House, King’s College London, and the BBC, as well as to a former MP (Davis 2023, cols. 527–528).

Interviewee testimony suggests that the kind of litigative onslaught described by Davis may mirror the experience of extremism researchers who attempt to publish on specific individuals and organisations, especially in the Islamist sphere. Although a court victory for the defendant may serve to give the statements that provoked a SLAPP better credibility than they previously held (as is arguably the case in e.g. Shakeel Begg vs British Broadcasting Corporation 2016), fighting this kind of challenge requires resourcing and institutional support which is often lacking, and this may be among the factors keeping much published extremism research at a high level of generality. While more generalist research on a social phenomenon can certainly be high value, its use will often be uncertain without up-to-date studies of specific aspects of that phenomenon, which the findings of this study (in common with Study I) suggest may have been systematically under-researched —- especially when it comes to the Islamist movement in the contemporary UK. In this context, it is perhaps worth noting that papers published as a result of the single round of CCE public funding (which Study I suggests was for several years the only major public organisation that funded research on recent Islamist group activity in the UK via open competition) triggered at least two legal challenges, although neither was ultimately successful in keeping the papers in question out of the public sphere.

While it may never be possible to publish all research findings on extremism, obstacles to publication should be recognised as a serious limit to the potential impact of such research, and even to the conditions that enable it to be carried out in the first place: publication is one of the chief means by which academic and think tank researchers demonstrate their track record in a particular area, without which, they may be unable to secure funding or even employment. Furthermore, the backers of the independent researcher interviewed for this study also expected to see stories appearing in the press (albeit written by named journalists rather than by the anonymous researcher) as evidence of that researcher’s effectiveness. These are important considerations. Even where it can be carried out, research whose findings cannot be published will not be subjected to peer review (except for the internal peer review practices of some organisations) and can neither be built upon nor critiqued by the wider research community. Unpublished findings cannot, for example, be incorporated into a literature review, instead potentially needing to be shared with different stakeholder groups on separate occasions. This is not only inefficient but unreliable: where research is non-discoverable by conventional means, opportunities for knowledge exchange may simply fail to arise, as a result of lack of stakeholder awareness that the research has been done.

Thus, whether we consider scholarly publications, think tank reports, newspaper articles, or government grey literature, there seem to be real difficulties in getting findings on specific extremist groups into the public sphere: difficulties which appear likely to have a negative impact both on the application and on the production of knowledge. From a stakeholder point of view, it is clearly preferable for private communication to supplement rather than replace public dissemination of research findings, providing early access to knowledge and an opportunity to seek clarification of or expansion on particular points of interest rather than being the sole means by which to learn about research projects otherwise doomed to obscurity. As for researchers, they may simply not bother to carry out investigations whose findings they anticipate may never see the light of day.

4. Conclusion

The two studies presented in this report jointly suggest that there may be substantial structural issues regarding the study of extremism in the UK. In particular, the author of this report wishes to draw attention to the quantitative finding of an apparent lack of grants awarded via open competition for the study of forms of extremism manifest in the UK since 1980 other than those associated with Incels and the far right. It may be that extremism research stakeholders do not consider this to be a problem. However, if they do, then the qualitative part of this report has been able to suggest several factors which might be acting jointly to introduce distortions into the knowledge base: distortions perhaps reflected in or exacerbated by funding patterns as previously mentioned.

Politicisation of extremism research emerged as a key theme in those interviews. Interviewees did not appear to be confident that other researchers in the field would always be even-handed and concerns were repeatedly raised (including by a representative of an organisation widely seen as being ‘on the left’) regarding the credibility of individuals carrying out ‘activist’ research on the far right. Several interviewees expressed the opinion that studying Islamism could lead to accusations of racism, while at least one — by no means the only interviewee to suggest that Islamism was over-researched — appeared to take the view that such accusations might not be entirely wrong. Some interviewees took the view that ‘left-wing’ or ‘progressive’ forms of extremism (for example relating to environmentalist demands) were under-researched for political reasons, and even that negative professional consequences might follow from trying to study them, while one expressed alarm at the prospect of official pressure to associate such causes with violent extremism. Recent events taking place at the UK’s leading think tank focused on the far right, i.e. CARR, indicate that the stakes of such ideological conflicts are very real. This report does not take a view on who is right or wrong in relation to these events, but notes that scholarly detachment may be unusually difficult to find in a field centrally concerned with political issues and openly conflicted along political lines.

Given the widespread employment of an approach to peer review that seems especially vulnerable to bias (i.e. single-blind review with decisions based on a numerical average of reviewer scores), the political climate in the field may help to explain why the balance of funded projects appeared to lean so heavily firstly towards general studies of extremism (perhaps seen as politically neutral, and therefore ‘safe’) and secondly towards studies of the far right, which, given the different arguments made in both studies presented here, might be expected to receive less hostile reviews than, say, studies of Islamist extremism. This pattern appeared to be funder-dependent: the dominance of studies of the far-right was less pronounced among projects funded by the CCE (which supported a relatively large number of small projects) and by the ESRC (which supported by far the most projects overall, in some cases with very large grants). However, only the CCE appears to have funded studies of recent (i.e. post-1980) Islamist extremism in the UK. Political tendencies might also explain why the dominance of studies of the far right was reduced among publicly funded projects on extremism in countries outside the UK and in the fairly distant past, as these are at a greater remove from political conflicts in which members of the UK extremism research community are likely to feel invested. However, many other problems have been identified, including various difficulties around access to or collection of data, the low academic prestige accorded to non-theoretical research on highly specific phenomena, and the possibility of legal and physical threats to researchers. All of these may have acted to discourage the study of particular extremist movements or groups in the contemporary UK.

Taking all these factors into consideration, it seems plausible that, where researchers are given the opportunity to apply for funds, they might feel incentivised to keep their proposed projects at the highest possible level of generality that can still satisfy the terms of the call. Furthermore, where some degree of specificity is needed, researchers might plausibly feel that it was wisest to take what would appear to be least controversial approach from the point of view of likely peer reviewers —- that is, to focus on the extreme right. This study has highlighted the very high generality of topics specified in CREST calls compared to those in the only CCE call to date: a level of generality which could be argued to render the former of these two approaches particularly attractive. Again, however, it must be emphasised that there is no objective standpoint from which to judge that the balance of funded projects is or is not as it should be: there is no scientific law which states that research topics must be prioritised according to the threat level that they present, and the question of whether there are ‘too many’ or ‘too few’ studies of the extreme right, the Islamist movement, or any other form of extremism is a political matter, as is the question of whether there are ‘too many’ or ‘too few’ studies of extremism in general (as opposed to, say, road safety).

Interviewees characterised the extremism research field as dependent upon a rapidlyshifting funding landscape within which long-term research projects are hard to develop. Several researchers suggested that the current configuration of this landscape has recently favoured study of the far right, leading to widespread failure to keep track of recent developments in the Islamist sphere. However, it seems likely that, given such a situation, every form of extremism will sooner or later find itself under-researched. This view was corroborated by the Home Office interviewee, who described extremism research as poorly funded and subject to short-term political imperatives. The intellectual entrepreneurship of academic researchers, who operate by identifying an area that has been insufficiently worked over and then ploughing that furrow for a period spanning many years, may mitigate these problems. However, the institutions which make such entrepreneurship possible, i.e. universities, are not well-equipped to support their staff through the type of fast-turnaround projects which may suddenly become possible once policy interest shifts to a previously neglected area, and may see these as financially unviable when supported through the government’s usual contracting system, although they regularly accept smaller sums when these are structured in the form of research grants rather than fees to be paid on a work-for-contract basis. Think tanks are better placed to take on such contracts, but even their representatives suggested that they would prefer to develop and maintain expertise over a longer term than current government research contracting practices can support: the cost of employing experts over many years may be impossible to sustain when the demand is for fast-turnaround research on a rapidly-changing set of policy priorities. Moreover, the practical difficulties of research procurement ensure that much government-commissioned research will continue to be carried out by generic research providers contracted to give insight on multiple topics but lacking in specialised expertise for any one specific project.

Large public research grants provide give researchers and research organisations with a potential way out of the cycle of short-term government, police, and military funding, but these are extremely time-consuming to apply for and have a low success rate. Moreover, the projects which which they support appear to have a tendency towards a very high level of generality, which would seem unlikely to aid the development of a wide pool of expertise focused on a diversity of specific extremist threats. In addition, where such projects buck this apparent trend, there seems to be a current bias towards study of one specific particular extremist threat, i.e. the far right. One interviewee also complained that private donors are often reluctant to be associated with extremism research, which suggests that philanthropy may be unlikely to fill any gaps that exist.

Long-term maintenance of a diverse pool of extremism-related research expertise on which stakeholders may draw appears to depend on university staff and independent researchers voluntarily pursuing interests in a wide range of directions despite often uncertain incentives. Independent researchers will follow their own agendas, especially where financial backing is available, although both they and their backers may need to hide their identity to avoid legal and physical reprisals. As for researchers working within a framework in which promotion and even continued employment may depend both upon publication and upon funding acquisition, extremism research —- especially on specific forms whose study their peers might consider to be ideologically suspect —– may come to seem highly unattractive.

In this context, CREST clearly plays a vital role in providing government funding for extremism research, helpfully structuring this funding in grant form (which removes the need for universities to recover full costs). Under the terms of their arrangement with CREST, government agencies can set priorities for CREST open calls, and also participate in the peer-review process by assessing bids for policy impact, but they do not have to issue contracts to individual teams of researchers, and can delegate responsibility for quality control to CREST’s community of peer reviewers. However, this puts members of that community in the position of gatekeepers, which is important to remember, given the impression which this report has gathered of a highly politicised field which may be distorted by several various forms of bias and self-censorship. In this context, it should be recalled that CREST’s funded project pattern of funded projects was particularly skewed towards research on the extreme right. Whether this arose because applications for grants to support other projects were rejected or because they were not made in the first place, it can be argued that more specifically targeted topics such as those employed in the CCE’s sole call could be used to alter the overall pattern, if that were to be felt desirable.

Issues around accessing data and potential research participants are very important. Experienced researchers will not attempt to carry out projects which they know will be made unfeasible by lack of such access, with the possible result that questions of interest to research stakeholders are not addressed. ‘Naturally occurring’ public online data are important, but may not be representative of anything other than themselves, and would ideally supplement data collected by conventional means, rather than being used as a substitute. Obstacles to data collection highlighted by interviewees include risk-averse ethical reviewers, uncooperative gatekeeper institutions, and the lack of a culture of data sharing, both with regard to research projects commissioned using public funds and with regard to key state institutions such as HMPPS.

Difficulties in publishing research on specific extremist groups — which might help to explain why generalised extremism studies dominate publicly funded projects — should also be acknowledged as serious. Researchers may of course present unpublished or unpublishable research findings when called on for advice. However, this raises several problems. Firstly, this kind of research has neither been peer-reviewed nor subjected to public criticism, which means that its value may be hard to gauge. Secondly, it is not discoverable, which means that individuals, organisations and sub-organisational units which might benefit from it may be unaware of its existence. Thirdly, there is little reputational benefit to individuals and institutions carrying out such research, diminishing incentives greatly. When these issues are compounded by legal challenges and threats to researcher safety, researchers and their employers (both in the university sector and in the faster-paced think tank world) may simply avoid this kind of work as risky and unrewarded. As for independent researchers, venturing into such treacherous waters without institutional protection —- as some evidently do —- must clearly require exceptional courage. It is thus to be regretted that Parliament is at present only debating legislation on SLAPPs which relate to economic crime (HM Government 2023). However, it is noted that the SRA has indicated that it is potentially able to take disciplinary action against solicitors who are judged to have issued SLAPPs even where they have not broken the law (SRA 2022).

In summary, this study has found several apparent issues which could be argued to impede or distort extremism studies in the UK. The field seems deeply ideologically conflicted, with the suggestion that expertise is muffled or even excluded due to an alleged requirement in some quarters to conform to emergent political orthodoxy; indeed, the possibility that researchers might be ‘cancelled’ for stepping out of line appears real, and it is emphasised that anticipation of discrimination may have a chilling effect extending beyond any specific implicit or explicit censure. There also seem to be major problems with funding, data collection, and even the dissemination of findings; in some cases, extremists have used legal or physical threats, as well as online harassment, to silence researchers. Considering these issues, it appears likely that key aspects of extremism may be under-represented in the knowledge base. All considered, it would seem unwise to assume either that a sufficient understanding of the current threat environment may be obtained from reading the published literature or that a range of expertise matching the full range of extremist threats will always exist in the UK.

5. Recommendations

Some of the problems identified in this report could be addressed through small changes to stakeholder practices. For example, government agencies could probably do more to leverage the expertise of external researchers if it were recognised, firstly that they often see access to data and stakeholders as a form of compensation, and secondly that their employers may not always expect to recoup full costs of activities which help them to fulfil their institutional mission. For example, universities generally charge full economic costs on research services which their staff provide for clients on a work-forcontract basis, but may charge considerably less where the funder is understood instead to be providing a grant which will support university staff in carrying out public-interest research that will lead to publication in a scholarly venue.

Substantial value could also be added to conventionally-procured research through promotion of greater cooperation between entities across the research ecosystem. For example, engagement of university-based researchers as consultants in assisting in the design of research projects or in analysing the resulting datasets, or engagement of think tanks in the identification of policy implications and the communication of findings, could bring an infusion of subject-specific expertise to work carried out by generic research providers without compromising on the flexibility which the latter can bring.

A simple commitment to informing individual researchers or research teams that their work has been useful (for example, when it is cited in briefing documents) would also help greatly, both because it gives researchers feedback by indicating that they have been working along beneficial lines, and because it gives them a documented way of claiming credit. This would in turn help those researchers to make a case for continued support from their employers or funders — which is also in stakeholder interests.

However, to solve some of the underlying problems discussed in this report —- or at least, to mitigate their potentially distorting impact on the knowledge base —- it is suggested that an extremism research policy may need to be developed and communicated, identifying research goals which are meaningful to researchers and stakeholders, acknowledging the likely difficulties to be faced in achieving these goals, and setting out concrete measures which might plausibly help these difficulties to be overcome.

Such a policy could, for example, require that, where government agencies issue data collection and analysis contracts, they include clauses requiring both raw and processed data to be deposited in a curated repository, with access by invitation or application where full openness is considered impossible or inadvisable. To give a further example, to be eligible for public funding, British universities could reasonably be expected to demonstrate that their ethical review processes were not neglecting public interest in favour of managing their reputations or avoiding exposure to legal threats. A defined pathway could also be created for accessing data and potential research participants via state agencies, including HMPPS, as it would be of obvious direct benefit to publicinterest research.

As an explicitly political act, the setting of quotas for funded projects might also have the advantage of foregrounding the (probably inescapable) politicisation of the field itself, and thereby stimulating debate about which forms of extremism research are actually required. A move away from single-blind peer review on the part of the major research funders might also serve to increase confidence in the fairness of the peer review process, and, while it seems unlikely that UKRI will would shift to open peer review for the sake of a field of research field accounting for only a small fraction of a percentage point of its distributed funds, it might perhaps perhaps be considered by a specialist funder such as CREST. Alternatively, simply giving an identified individual responsibility for adjudicating between contradictory reviews might reproduce some of the advantages of open peer review, as compared to the ‘computer says no’ approach to decision-making wherein a numerical average is taken across multiple anonymous reviews, potentially enabling a single biased reviewer to torpedo a proposal. Additional measures could be employed as desired: for example, empirical studies of specific extremist groups could receive extra weighting in the approvals process, if stakeholders indeed consider such studies to be valuable, thus ensuring that such studies continued to be supported and incentivised in the face of what this report suggests may be contrary incentives towards generalism.

Joined-up thinking will be essential, however. There is little point in funding empirical studies of specific extremist groups if data on those groups is impossible to collect, or if the groups themselves put researchers and their families at such risk of physical harm as to make it impossible for them to continue in their work, or use spurious legal threats to block the publication of whatever research outputs are eventually produced. With regard to physical threats, one might hope to see employers providing training for those facing a credible threat, [erhaps along with financial support for installing security systems where appropriate, but the misuse of legal process appears to be a more widespread problem. It is therefore welcome that the SRA has issued solicitors with a statement emphasising that they are ‘expect[ed] . . . to take reasonable steps to satisfy [themselves] that a claim is properly arguable before putting it forward’ (SRA 2022, n.p.), but a voluntary approach to abandoning SLAPPs would seem unlikely to succeed, given the suggestion (both in interviewee testimony and in parliamentary debate) that certain legal firms have come to specialise in such actions. The SRA may wish to consider its potential role both as a stakeholder and as an instrument of extremism research policy, proactively seeking out or publicly asking to be notified of cases where SLAPPs have been used to suppress the dissemination of information about extremists. However, it seems likely that a legislative solution will be required, just as it has been found to be required in relation to SLAPPs issued in relation to allegations of economic crime. For this reason, a consultation on the use of legal and other silencing tactics by extremist groups and their associates and supporters would seem particularly urgent.

Technical acknowledgements

This report was typeset with the use of knitr v. 1.39 (Xie 2014), kableExtra v.

1.3.4 (Zhu 2021), bookdown v. 0.26 (Xie 2016), Pandoc v. 2.5 (McFarlane 2018), and pdfTeX v. 3.14159265-2.6-1.40.20 (Thành 2019). Code and text were edited in RStudio v. 2023.03.0+386 and earlier (Posit team 2023). Quantitative analysis was carried out in R v. 4.2.2 (R Core Team 2022), with the additional use of a number of libraries, notably GtR v. 0.3.0 (M. Smith 2022) for accessing UKRI data via the API, rvest v. 1.0.2 (Wickham 2021) for scraping data from web pages, dplyr v. 1.0.10 (Wickham et al. 2022) for data manipulation, and ggplot2 v. 3.3.6 (Wickham 2016) for data visualisation.

References

Adekoya, Remi, Eric Kaufmann, and Thomas Simpson. 2020. “Academic Freedom in the UK.” London: Policy Exchage. https://policyexchange.org.uk/publication/academic-freedom-in-the-uk-2/.

Aliyev, Huseyn. 2022-2024. “Mobilisation of Foreign Fighters in the Former Soviet Union.” Swindon: UKRI. https://gtr.ukri.org/projects?ref=ES%2FV003453%2F1.

Baker, Simon. 2021. “UKRI Success Rates Fall as Grant Applications Ramp up: Overall Funding Decisions More Than Double to Almost 40,000 Thanks to Covid-Related Grants.” Times Higher Education. https://www.timeshighereducation.com/news/ukri-success-rates-fall-grant-applications-ramp.

Blank, Grant, and Christoph Lutz. 2017. “Representativeness of Social Media in Great Britain: Investigating Facebook, LinkedIn, Twitter, Pinterest, Google+, and Instagram.” American Behavioural Scientist 61 (7): 741–56. https://doi.org/10.1177/0002764217717559.

Bourdieu, Pierre. 1975. “The Specificity of the Scientific Field and the Social Conditions of the Progress of Reason.” Social Science Information 14 (6): 19–47. https://doi.org/10.1177/053901847501400602.

Bury, Patrick. 2020-2024. “The Transformation of Transatlantic Counter-Terrorism, 2001-2025.” Swindon: UKRI. https://gtr.ukri.org/projects?ref=MR%2FS034412%2F1.

Carter, Rosie. 2019. “State of the Nation: 2018.” In State of Hate 2019: People Vs the Elite?, 14–18. London: HOPE not hate. https://hopenothate.org.uk/wp-content/uploads/2019/02/state-of-hate-2019-final-1.pdf.

Casciani, Dominic. 2018. “Was 2017 the Worst Year for UK Terrorism?” BBC. https://www.bbc.co.uk/news/uk-42420358.

CCE. 2019. “Commission Calls for Academic Papers.” London: Commission for Countering Extremism. https://extremismcommission.blog.gov.uk/2019/02/04/commission-calls-for-academic-papers/.

College of Policing. n.d. “Evidence-Based Policing.” Coventry: College of Policing. https://www.college.police.uk/research/evidence-based-policing-EBP.

CREST. 2021. “Research for Understanding, Mitigating, and Countering Security Threats: Fifth Broad Topic Announcement, Call Specification Version 1.1.” Lancaster: Centre for Research; Evidence on Security Threats. https://crestresearch.ac.uk/download/3597/crest_call_specification_2021_v1_1.pdf.

CREST. n.d.a. “About.” Lancaster: Centre for Research; Evidence on Security Threats. https://crestresearch.ac.uk/about/. ———. n.d.b. “Centre for Research and Evidence on Security Threats (CREST).” Swindon: UK Research and Innovation. https://gtr.ukri.org/projects?ref=ES%2FV002775%2F1.

Davis, David. 2023. “Lawfare.” Hansard 735. https://hansard.parliament.uk/Commons/2023-06-29/debates/AB9B8F81-7AE0-446A-81E8-6F50DFE7BCC2/Lawfare.

Diner, Dan. 1996. “On the Ideology of Antifascism.” New German Critique 67 (Winter): 123–32. https://www.jstor.org/stable/827781.

Dior, Lawrence. 2021-2025. “Extreme Laser-Driven Hydrodynamics.” Swindon: UKRI. https://gtr.ukri.org/projects?ref=studentship-2607617.

Fells, Julian. 2020-2025. “Multipoint Sensors for Extreme Environments.” Swindon: UKRI. https://gtr.ukri.org/projects?ref=EP%2FT00326X%2F1.

Garapich, Michał. 2023-2026. “Everyday Transnationalism of the Far Right: An Interdisciplinary Study of Polish Immigrants’ Participation in Far-Right Groups in Britain.” Swindon: UKRI. https://gtr.ukri.org/projects?ref=ES%2FW010151%2F1. Haffar, Samir, Fateh Bazerbachi, and M. Hassan Murad. 2019. “Peer Review Bias: A Critical Review.” Mayo Clinic Proceedings 94: 670–76. https://doi.org/10.1016/j.

mayocp.2018.09.004.

Hall, Jonathan.      2022.      Terrorism in Prisons.        London:    HM Government. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1071318/terrorism-in-prisons.pdf.

Hinkkainen, Kaisa. 2021. “Explaining Non-State Armed Groups Perpetration of Mass Atrocity Crimes.” Swindon: UKRI. https://gtr.ukri.org/projects?ref=ES%2FV000292%2F1.

HM Government. 2023. “Factsheet: Strategic Lawsuits Against Public Participation (SLAPPs).” London: HM Government. https://www.gov.uk/government/publications/economic-crime-and-corporate-transparency-bill-2022-factsheets/factsheet-strategic-lawsuits-against-public-participation-slapps.

HMG. 2022. “Reported Road Casualties Great Britain, Provisional Results: 2021.” HM Government. https://www.gov.uk/government/statistics/reported-road-casualties-great-britain-provisional-results-2021/reported-road-casualties-great-britain-  provisional-results-2021.

HNH. 2019a. “Attitudes Towards Muslims in Britain.” In State of Hate 2019: People Vs the Elite?, 19–28. London: HOPE not hate. https://hopenothate.org.uk/wp-content/uploads/2019/02/state-of-hate-2019-final-1.pdf. ———. 2019b. “Executive Summary.” In State of Hate 2019: People Vs the Elite?, 3. London: HOPE not hate. https://hopenothate.org.uk/wp-content/uploads/2019/02/state-of-hate-2019-final-1.pdf.

Howard, Jeffrey William. 2022-2026. “Governing Democratic Discourse: Social Media, Online Harms, and the Future of Free Speech.” Swindon: UKRI. https: //gtr.ukri.org/projects?ref=MR%2FV025600%2F1.

Ketola, Hanna. 2018-2020. “The Politics of Women’s Agency: Gender and Peacebuilding in Post-Conflict Nepal.” Swindon: UKRI. https://gtr.ukri.org/projects?ref=ES%2FS011188%2F1.

Lowles, Nick. 2019. “Overview.” In State of Hate 2019: People Vs the Elite?, 3–8. London: HOPE not hate. https://hopenothate.org.uk/wp-content/uploads/2019/02/state-of-hate-2019-final-1.pdf.

Maher, Shiraz. 2020. “Mapping Terrorist Exploitation of and Migration Between Online Communication and Content-Hosting Platforms.” Lancaster: Centre for Research; Evidence on Security Threats. https://crestresearch.ac.uk/projects/mapping-terrorist-exploitation/.

McCann, Craig. 2022. “Expelled from a Progressive Think Tank — for the Crime of Denouncing Antifa Violence.” Quillette. https://quillette.com/2022/02/22/expelled-from-a-progressive-think-tank-for-the-crime-of-denouncing-antifa-violence/.

McFarlane, John. 2018. Pandoc (Version 2.5). https://pandoc.org.

Micheni, Makeni. 2018-2022. “The Role of Tribalism in Sub-Saharan Terrorist Groups.” Swindon: UKRI. https://gtr.ukri.org/projects?ref=studentship-2098309.

Posit team. 2023. RStudio: Integrated Development Environment for r. Boston, MA: Posit Software, PBC. http://www.posit.co/.

Potter, Jonathan. 2002. “Two Kinds of Natural.” Discourse Studies 4 (4): 539–42. https://doi.org/10.1177/14614456020040040901.

R Core Team. 2022. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. https://www.R-project.org.

REF. 2021. “Guide to the REF Results.” Swindon: Research England. https://www.ref.ac.uk/guidance-on-results/guidance-on-ref-2021-results/.

Regehr, Kaitlyn. 2022-2024. “Understanding the Cel: Vulnerability, Violence and In(ter)vention.” Swindon: UKRI. https://gtr.ukri.org/projects?ref=AH%2FW007398%2F1.

Reveilhac, Maud, Stephanie Steinmetz, and Davide Morselli. 2022. “A Systematic Literature Review of How and Whether Social Media Data Can Complement Traditional Survey Data to Study Public Opinion.” Multimedia Tools and Applications 81: 10107–42. https://doi.org/10.1007/s11042-022-12101-0.

Rixon, Chloe. 2020-2023. “Memory, Cinema and Polish Deathscapes: Re-Membering the Holocaust.” Swindon: UKRI. https://gtr.ukri.org/projects?ref=studentship-2420728.

Rooyen, Susan van, Fiona Godlee, Stephen Evans, Nick Black, and Richard Smith. 1999. “Effect of Open Peer Review on Quality of Reviews and on Reviewers’ Recommendations: A Randomised Trial.” British Medical Journal 318 (7175): 23–27. https://doi.org/10.1136/bmj.318.7175.23.

Secretary of State for the Home Department. 2011. Prevent Strategy. HM Government. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/97976/prevent-strategy-review.pdf.

Seeber, Marco, and Alberto Bacchelli. 2017. “Does Single Blind Peer Review Hinder Newcomers?” Scientometrics 113: 567–85. https://doi.org/10.1007/s11192-017- 2264-7.

Shakeel Begg vs British Broadcasting Corporation. 2016. Royal Courts of Justice, London, no. EWHC 2688 (QB). https://www.judiciary.uk/wp-content/uploads/2016/10/shakeel-begg-v-bbc-judgment-final-20161028.pdf.

Shawcross, William. 2023. “Independent Review of Prevent.” London: House of Commons. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1134986/Independent_Review_of_Prevent.pdf.

Smith, Matthew. 2022. GtR: Access Gateway to Research Data via the API. https:

//github.com/MatthewSmith430/GtR.

Smith, Richard. 1999. “Opening up BMJ Peer Review” 318 (4). https://doi.org/10.1136/bmj.318.7175.4.

SRA. 2022.        “Warning Notice:      Strategic Lawsuits Against Public Participation  (SLAPPs).”    London:                        Solicitors  Regulatory  Authority.           https:

//www.sra.org.uk/solicitors/guidance/slapps-warning-notice/.

Syeda, Seema, and Titus Molkenbur. 2023. Mainstream and Far Right Islamophobia: A UK-Germany Comparison. Bonn: Friedrich Ebert Stiftung. https://library.fes.de/pdf-files/bueros/london/19913.pdf.

Thành, Hàn Thê. 2019. pdfTeX (Version 3.14159265-2.6-1.40.20). https://ctan.org/pkg/ pdftex?lang=en.

Tomkins, Andrew, Min Zhang, and William D. Heavlin. 2017. “Reviewer Bias in SingleVersus Double-Blind Peer Review.” PNAS 114 (48): 12708–13. https: //doi.org/10.1073/pnas.1707323114.

Turbutt, Sophie. 2021-2025. “Women as Transnational Agents in the Development of Iberian Anarchist Thought and Practices Relating to Female Bodily Autonomy in the Interwar Period.” Swindon: UKRI. https://gtr.ukri.org/projects?ref=studentship-2603231. ———. 2022. “Sexual Revolution and the Spanish Anarchist Press: Bodies, Birth Control, and Free Love in the 1930s Advice Columns of La Revista Blanca.” Contemporary European History published online first: 1–19. https://doi.org/10.1017/S0960777322000315.

UKRI. 2021. Gateway to Research API 2: Version 1.7.5. Swindon: UKRI. https://gtr.ukri.org/resources/GtR-2-API-v1.7.5.pdf. ———. 2023. “Key Terms Glossary.” Swindon: UK Research and Innovation. https://www.ukri.org/councils/esrc/guidance-for-applicants/research-ethics-guidance/useful-resources/key-terms-glossary/.

UKRI. n.d.a.  “How We’re Funded.”  Swindon: UK Research and Innovation.                           https://www.ukri.org/wp-content/uploads/2022/08/UKRI-050822-FundingExplainer-HowWereFunded.pdf.

UKRI. n.d.b. “Peer Review Process.” Swindon: UK Research and Innovation. https://www.ukri.org/councils/esrc/guidance-for-reviewers/peer-reviews/peer-review-process/.

University of Glasgow. n.d.a. “Academic Promotion Criteria: Professorial Zone Descriptors, Research & Teaching.” Glasgow: University of Glasgow. https: //www.gla.ac.uk/media/Media_497802_smxx.pdf. ———. n.d.b. “Academic Promotion Criteria: Readership Descriptors, Research & Teaching.” Glasgow: University of Glasgow. https://www.gla.ac.uk/media/Media_503278_smxx.pdf. ———. n.d.c. “Academic Promotion Criteria: Research & Teaching Grades 7-9.” Glasgow: University of Glasgow. https://www.gla.ac.uk/media/Media_829567_smxx.pdf.

Walsh, Elizabeth, Maeve Rooney, Louis Appleby, and Greg Wilkinson. 2000. “Open Peer Review: A Randomised Controlled Trial.” British Journal of Psychiatry 176: 47–51. https://doi.org/10.1192/bjp.176.1.47.

Webber, Diane, and Alison Struthers. 2019. “Critiquing Approaches to Countering Extremism: The Fundamental British Values Problem in English Formal Schooling.”  London: Commission for Countering Extremism.  https://www.gov.uk/government/publications/critiquing-approaches-to-countering-extremism-the-fundamental-british-values-problem-in-english-formal-schooling.

Wickham, Hadley. 2016. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. https://ggplot2.tidyverse.org. ———. 2021. rvest: Easily Harvest (Scrape) Web Pages. https://CRAN.R-project.org/package=rvest.

Wickham, Hadley, Romain François, Lionel Henry, and Kirill Müller. 2022. dplyr: A Grammar of Data Manipulation (Version 1.0.10). https://CRAN.R-project.org/package=dplyr.

Wilkinson, Matthew, Lamia Irfan, Muzammil Quraishi, and Mallory Schneuwly Purdie. 2022. Islam in Prison: Finding Faith, Freedom, and Fraternity. Bristol: Policy Press.

Xie, Yihui. 2014. “knitr: A Comprehensive Tool for Reproducible Research in R.” In Implementing Reproducible Computational Research, edited by Victoria Stodden, Friedrich Leisch, and Roger D. Peng. Chapman; Hall/CRC. http://www.crcpress.com/product/isbn/9781466561595. ———. 2016. bookdown: Authoring Books and Technical Documents with R Markdown. Boca Raton, Florida: Chapman; Hall/CRC. https://bookdown.org/yihui/bookdown.

Zhu, Hao. 2021. kableExtra: Construct Complex Table with kable and Pipe Syntax (Version 1.3.4). https://CRAN.R-project.org/package=kableExtra.

Appendix I: Included grant awards (Study I)

Start year	Funder	Value of grant	Title
2018	AHRC	Studentship	Use of Folktale Adaptations in Exploring the Role of Gendered Social Narratives in the Far-Right Radicalization of Online Communities.
2018	EPSRC	Studentship	A Radicalisation Framework for the Monitoring and Understanding of Extremist Content Online
2018	ESRC	Studentship	Counter-terrorism and militarism in education
2018	ESRC	Studentship	Fear, identity and security in post-9/11 American foreign policy
2018	ESRC	Studentship	Lost in Translation?: How do state-funded secondary schools and teachers in England perceive and operationalise the ’fundamental British values’ polic
2018	ESRC	Studentship	MSc Investigative and Forensic Psychology
2018	ESRC	Studentship	Stolen, enslaved, sold, spoiled, or saved? Sahelian sex trafficking as militarised experience.
2018	ESRC	Studentship	Terrorism Expertise and Judicial and Legislative Practices in Bulgaria
2018	ESRC	Studentship	The role of tribalism in sub-Saharan terrorist groups
2018	ESRC	Studentship	Who Controls the Border? An Analysis of The Nordic Far-Right
2018	ESRC	£10,001-£99,999	The Contemporary Euro-American Far Right and the New White Nationalism
2018	ESRC	£10,001-£99,999	The Politics of Women’s Agency: Gender and Peacebuilding in post-conflict Nepal
2018	Innovate UK	£10,001-£99,999	Raven: to Locate and Identify Online Extremist Multimedia
2019	AHRC	Studentship	Narrative logics of hate speech: algorithms of fascist storytelling in the digital age.
2019	AHRC	Studentship	Performing Stragismo and Counter-spectacularisation: Italian Terrorism and Its Legacies
2019	AHRC	Studentship	The Fascist Temptation: Students and Young Intellectuals in Interwar Romania
2019	AHRC	£10,001-£99,999	European Fascist Movements, 1918-1941
2019	EPSRC	Studentship	Radical Right Extremism, Online Propaganda and Hybrid Human-Automated Content Removal
2019	EPSRC	£100,000+	GCRF Gender and Violent Extremism Network
2019	ESRC	Studentship	Between welfare and warfare: an International Political Sociology of anti-radicalisation as a technology of government
2019	ESRC	Studentship	Identity Fusion in Virtual Groups: The Impact of Shared Dysphoric Experiences in Global Echo Chambers on Online Radicalisation
2019	ESRC	Studentship	Ideological Change in the Italian Communist Party and Austrian Socialist Party, 1964-1991 (Provisional)
2019	ESRC	Studentship	Investigating the Effect of Territorial Control on the Micro-Dynamics of Domestic Terrorism
2019	ESRC	Studentship	Preventative approaches in radicalisation: a comparative study of early interventions for extreme far right and ’Islamist’ ideologies.
2019	ESRC	Studentship	Russia and Foreign Fighters: from import to export
2019	ESRC	Studentship	Tackling Extremism Through Philosophy In Schools
2019	ESRC	Studentship	The impact of militarisation on perceptions of state legitimacy in Niger
2019	ESRC	Studentship	The role of historical narratives in extremist propaganda.
2019	ESRC	Studentship	Tracing the emergence of far right discourse in the UK mainstream
2019	ESRC	Studentship	Understanding Enagement with Religious Actors Under the Channel Programme from 2005-2015
2019	ESRC	Studentship	Visual Extremism Online: A Discursive News Values Analysis of Social Media Images from Radical Right Groups
2019	ESRC	£100,000+	’The Crime-Terror Nexus: Investigating the overlap between criminal and extremist practices, narratives and networks in Tripoli, Lebanon
2019	CCE	£1-£10,000	Belief, Attitude, and Behavior Change: Leveraging Current Perspectives for Counter-Radicalization
2019	CCE	£1-£10,000	Critiquing Approaches to Countering Extremism: The Fundamental British Values Problem in English Formal Schooling
2019	CCE	£1-£10,000	Critiquing approaches to countering extremism and terrorism via Prevent
2019	CCE	£1-£10,000	Drivers of Extremism: Global Political Antagonisms reproduced in Cypriot and Italian Insurgencies
2019	CCE	£1-£10,000	Embedding human rights in countering extremism: reflections from the field and proposals for change
2019	CCE	£1-£10,000	Exploring Radicalisation and Extremism Online - an Experimental Study
2019	CCE	£1-£10,000	Extremism Online - Analysis of extremist material on social media Professor Imran Awan, Hollie Sutch and Dr Pelham Carter
2019	CCE	£1-£10,000	Mainstream Islamism in Britain: Educating for the “Islamic Revival”
2019	CCE	£1-£10,000	Mainstreaming Islamism: Islamist Institutions and Civil Society Organisations
2019	CCE	£1-£10,000	Modernising and Mainstreaming: The Contemporary British Far Right
2019	CCE	£1-£10,000	National Action: links between the far right, extremism and terrorism
2019	CCE	£1-£10,000	Overview of the Far-Right
2019	CCE	£1-£10,000	Talking Our Way Out of Conflict: Critical reflections on ’mediated dialogue’ as a tool for secondary level CVE
2019	CCE	£1-£10,000	The Moral Ecology of Extremism A Systemic Perspective
2019	CCE	£1-£10,000	The changing nature of activism among Sikhs in the UK today
2019	CCE	£1-£10,000	Violent extremist tactics and the ideology of the sectarian far left
2019	CCE	£1-£10,000	What is to be Done about al-Muhajiroun? Containing the Emigrants in a Democratic Society
2020	AHRC	Studentship	Divine Sovereignty in Jihadi-Salafist Thought: An Intellectual Genealogy, 1990-2015
2020	AHRC	Studentship	Global Connections, Local Struggle: Transnationalism and Spanish Anarcho-Syndicalism, c.1910-1939.
2020	AHRC	£100,000+	#ContestingIslamophobia: Representation and Appropriation in Mediated Activism.
2020	EPSRC	Studentship	Expanding and Assembling Approaches to Improve Decisions on Identification and Classification of Online Terrorist Content
2020	ESRC	Studentship	A comparative study of European local Prevention of Violent Extremism (PVE) policies
2020	ESRC	Studentship	A melting pot of hate: Researching the alt-right on Reddit and Voat during US 2020 Presidential Election
2020	ESRC	Studentship	Eco-fascism and political ecologies of the far-right in Germany
2020	ESRC	Studentship	Empirically explore how practitioners working on the prevent strand of the UK strategy - passing the violent threshold to constitute counter terrorism
2020	ESRC	Studentship	Exploring the link between ’intimate’ terrorism and mass public violence
2020	ESRC	Studentship	Gender in the Online Far Right: Mapping Transnational Linkages and Fractures
2020	ESRC	Studentship	Online Radicalisation and Alt-Right Poitical Violence
2020	ESRC	Studentship	Populism and the British Far-right:
2020	ESRC	Studentship	The implications of changing migration discourses and the rise of the far-right in Europe on peripheral, rural communities in Wales and Sweden.
2020	ESRC	Studentship	Who Defines Terror?: A Quantitative Analysis of Interstate Deliberation on Terrorism
2020	ESRC	£10,001-£99,999	Using Psychological Insights about Stigmatized Groups to Inform Policymaking and Institutional Design
2020	ESRC	£100,000+	Cancellation of citizenship and national security: a comparison between France and the UK
2020	ESRC	£100,000+	Covid-19: What are the Drivers of the Islamophobic Infodemic Communications on social media
2020	ESRC	£100,000+	Prebunking: Psychological ’vaccines’ against online misinformation, extremist recruitment and vaccine hesitancy.
2020	UKRI	£100,000+	The Transformation of Transatlantic Counter-Terrorism 2001-2025
2020	CREST	Unknown	Constraining Violence
2020	CREST	Unknown	Development of Risk Assessment Schemes for Channel
2020	CREST	Unknown	Disengagement and Desistance: A Systematic Review
2020	CREST	Unknown	Factors That Deter Threat Actors And Reconnaissance
2020	CREST	Unknown	Knowledge Management Across the Four Counter-Terrorism ’Ps’
2020	CREST	Unknown	Mapping terrorist exploitation of and migration between online communication and content-hosting platforms
2020	CREST	Unknown	Memetic Irony and the Promotion of Violence within Chan Cultures
2020	CREST	Unknown	Mining the Chans: Exposing the Visual and Linguistic Dynamics of Radicalisation in Far-right Image-boards (MineChans)
2021	AHRC	Studentship	Challenging ’far-right ecologism’ through contemporary environmental fiction
2021	AHRC	Studentship	How does the victimhood identity of right-wing populist parties in France, Germany, and Austria contribute to the rise in anti-Semitic hate crime?
2021	AHRC	Studentship	Strained Solidarities: Black Power and the British Left, 1960-80
2021	AHRC	Studentship	The Freedom Defence Committee: Anarchism and Libertarian Thinking in British Thought from 1940 - 1960.
2021	AHRC	Studentship	The cultures political participation in UKIP. An Ethnographic exploration of English Nationalism in a coastal town in South East Essex
2021	AHRC	Studentship	Women as Transnational Agents in the Development of Iberian Anarchist Thought and Practices Relating to Female Bodily Autonomy in the Interwar Period
2021	AHRC	£10,001-£99,999	Rethinking International Communism
2021	AHRC	£10,001-£99,999	Rethinking the Philosophy of Terrorism
2021	ESRC	Studentship	An Exploration of the Linguistic and Visual Practices of Far-Right Militia Groups in Online Messaging Boards
2021	ESRC	Studentship	Analysing media portrayals of far-right inspired terrorism
2021	ESRC	Studentship	Between Politics and Law: The Question of the European female Islamic State returnees
2021	ESRC	Studentship	Computational approaches to detecting in-groups and out-groups in extremist communication
2021	ESRC	Studentship	Countering online terrorist propaganda whilst respecting the right to free speech
2021	ESRC	Studentship	Education, extremism and utopian thinking
2021	ESRC	Studentship	Exploring the Complex and Contradictory Role(s) of Women within the UK Far Right
2021	ESRC	Studentship	How misinformation in the aftermath of terror attacks has been used to influence public reactions by the UK far-right.
2021	ESRC	Studentship	Identity Based Trauma and Engagement with Extremism: How does trauma shape identity and engagement with extremism?
2021	ESRC	Studentship	The Application of Digital Forensics and Machine Learning to Automatically Detect and Flag Modified Online Terrorist Propaganda
2021	ESRC	Studentship	The Role of Civil Society and Actors in Narrating Online Against Terrorism/Violent Extremism (TVE): Narratives and Practice
2021	ESRC	Studentship	Understanding Countering Terrorism Policies in Global South: A Case Study of Rohingya Refugees in Bangladesh
2021	ESRC	Studentship	YouTube, 4Chan and Incels: Investigating the mechanism and process of indoctrination and radicalisation of misogynistic terrorists to inform formation
2021	ESRC	£100,000+	Atmospheres of (counter)terrorism in European cities
2021	ESRC	£100,000+	Conceptualising Threat: Implementing the Prevent duty within Greater Manchester’s Further Education Sector
2021	ESRC	£100,000+	Explaining non-state armed groups perpetration of mass atrocity crimes
2021	ESRC	£100,000+	The effect of terrorism on public attitudes and individual well-being in Great Britain
2021	NERC	£1-£10,000	Comparative Data Analysis in Political Aesthetic Strategies: The Persuasive Propaganda of the Italian Populist Right on Facebook and Twitter
2021	CREST	Unknown	Con.Cel: Tracking the Online Contagion of InCel and Male Supremacist Ideology
2021	CREST	Unknown	Detecting Hybrid Social Identities: A Computational Analysis of Influence and Resilience in Online RWE
Communities
2021	CREST	Unknown	Environment and Interventions: Assessing the environmental risk of terrorism
2021	CREST	Unknown	Mapping and Modelling Influence Interactions between Conspiracy Communities and Extremist Entities (MICE2)
2021	CREST	Unknown	Misogyny, hostile beliefs and the transmission of extremism: a comparison of the far-right in the UK and Australia
2021	CREST	Unknown	Prosecuting Extremists in the United Kingdom: An Exploration of Charging, Prosecution and Sentencing Outcomes
2021	CREST	Unknown	Trauma, Adversity and Violent Extremism
2022	AHRC	Studentship	Extremist and counter extremist narratives in the digital sphere and their impact on domestic information ecosystems
2022	AHRC	£100,000+	Understanding the Cel: Vulnerability, Violence and In(ter)vention
2022	ESRC	Studentship	Radicalization and Polarisation in New Media: The Case of YouTube
2022	ESRC	£10,001-£99,999	Rethinking Salafi/Wahhabi transnational and the limits of state power in North Africa
2022	ESRC	£100,000+	Constructing the illiberal citizen? Radicalisation prevention, counter-terrorism, and the media in the UK
2022	ESRC	£100,000+	Everyday transnationalism of the far right: an interdisciplinary study of Polish immigrants’ participation in far-right groups in Britain
2022	ESRC	£100,000+	MOVING FROM NETWORKED TO PATCHWORKED SOCIETY: MOTIVATIONAL UNDERPINNINGS AND SOCIETAL CONSEQUENCES
2022	ESRC	£100,000+	Mobilisation of foreign fighters in the former Soviet Union
2022	UKRI	£100,000+	Governing Democratic Discourse: Social Media, Online Harms, and the Future of Free Speech

Appendix II: Data extraction form categories (Study I)

• Time period (tick all that apply): 1980-present; Pre-1980; Unclear / not applicable
• Geographical location (tick all that apply): UK; Rest of world; Unclear / not applicable
• Focus on counter-extremism or counter-terrorism? Yes; No
• Forms of extremism focused on (tick all that apply): Islamism; Right-wing extremism; Left-wing extremism; Incel / misogynist; Single-issue; Mixed, unstable, or unclear ideology; Other religious extremism; Other ethnic extremism; Other political extremism; Other sexual / gender-related extremism; Form of extremism not covered by the above; No specific form of extremism

Appendix III: Interview prompts (Study II)

• Which forms of extremism do you study, and in which settings?
• Which specific aspects of extremism do you study?
• Which methodologies do you use to study extremism?
• Talk me through a typical research project for you: what is involved in carrying it out?
• Which funders have you worked with (if any)?
• Who would you see as the stakeholders for your research?
• How do you choose what to research?
• Could you tell me about any factors that might encourage you to pursue a particular project, or that might make it easier to carry through?
• Could you tell me about any factors that might make it difficult for you to carry through a research project, or that might lead you not to try in the first place?
• Could you tell me a bit about your experience of communicating your findings: to other researchers, to stakeholders, or to the public?
• Are you confident that your findings are reaching (or will reach) the people who need to be informed by them?
• What would you see as the priority areas for research on extremism in the UK?
• Are some areas overor under-researched?
• What would you see as the most useful methodologies for studying extremism?
• Are some methods overor under-used?
• Is there anything that you would like to add?

About the author

Daniel Allington is Reader in Social Analytics at King’s College London. Organisations which have supported his past research include the Arts and Humanities Research Council, the Economic and Social Research Council, and the Commission for Countering Extremism. He is Deputy Editor of the Journal of Contemporary Antisemitism, and a Research Fellow of the London Centre for the Study of Contemporary Antisemitism.

1. These were identified by the financial value returned for them via the API, which was in all cases zero. Studentships are not, of course, zero cost, but they are accounted for separately. ↩