The Role of Civil Society in Political Repression: The UK Prevent Counter-Terrorism Programme

Research on social movements shows a bias towards movements that oppose the status quo. Consequently, state–movement relations are primarily characterised as antagonistic. Where cooperative relationships are considered, the focus is on co-option and institutionalisation of movements. By contrast, this article focuses on social movements that support the status quo and how in their collaborations with governmental actors, they act as para-statal agencies. Drawing on findings from a multi-site ethnography examining the implementation of the UK Prevent counter-terrorism programme, I show how neoconservative think tanks and counter-extremism civil society organisations help to enact and extend Prevent as a distinct form of political repression. As such, this article gives close attention to the otherwise neglected role that non-state actors play in non-violent political repression. My argument builds on and extends emerging work analysing social movement activity beyond the prism of the ‘challengers versus authorities’ paradigm.


Introduction
Social movement scholars have focused on antagonistic relationships between social movements and the state (Snow, 2004;Tarrow, 1998;Tilly, 1978). The dominant traditions of resource mobilisation (RMT) and political process theory (PPT) define social movements as 'collective challenges, based on common purposes and social solidarities, in sustained interaction with elites, opponents and authorities' (Tarrow, 1998: 4). The conventional view is thus that social movement actors exist outside of the state and mobilise extra-institutional resources due to their exclusion from institutionalised channels of power.
This conception of state-movement relations leaves open two unanswered questions. The first is how do we understand the activism of those within institutionalised channels of power, utilising their positions to pursue a social movement cause? The second -in fact, the flip side of this question -is how do we understand the extra-institutional mobilisation of elite groups, who already have reasonable access to institutional channels of power, but also organise extra institutionally as a means of achieving their goals?
In this article, I examine both types of social movement activity. I show how neoconservative think tanks and counter-extremism civil society organisations alongside governmental counter-terrorism actors form an 'enforcement network' (Cunningham et al., 2019) enabling the Prevent counter-terrorism programme as a distinct form of political repression. Moving beyond a 'challenger versus authority' conception of social movement activity, my analysis shows three ways in which political mobilisation is located within and around the state.
First, governmental actors played a proactive role in furthering an ideological agenda by participating in a Prevent enforcement network. Second, the government covertly sponsored civil society campaigns in an attempt to mobilise grassroots support for their policies. Finally, state personnel formed alliances with neoconservative actors. This meant that non-state elite actors played a role in both influencing and enacting government policy.
I argue that existing sociological accounts of repression have a tendency to focus on state-based coercion; this is too narrow to account for the empirical reality of movement suppression. Drawing on and extending social movement literature on the role of nonstate actors as repressive agents (Cunningham et al., 2019;Earl, 2003;Ong, 2018), I suggest that state-supported movement organisations (SSMO) serve as para-statal agencies with quasi-independence from the state.
This has two implications. First, rather than see the state as reactive, I consider the way state actors can proactively engage in political mobilisation. Second, in contrast to a conventional understanding of civil society as autonomous to the state with a countervailing capacity, I show how civil society can be utilised to further advance state power and interests. Moreover, it is their relative autonomy that helps circumvent legitimate means of authority, and evade democratic accountability.

Rethinking the 'Challenger versus Authority' View of Social Movements
Social movement scholars have recently shown greater interest in explaining how social movements influence policies, departing from a previous conception that 'institutionalisation' or co-optation of movements was typical (Tarrow, 1998). Work on brokerage (Vasi, 2011) demonstrates how movements gain leverage over elites. Political mediation models show how specific political conditions mediate movements' resource mobilisation and how both factors influence which causes get taken up by political elites (Amenta, 2008). Others have sought to reconsider the boundaries between state and non-state actors with various attempts at understanding 'institutional activists' (Banaszak, 2005;Pettinicchio, 2012;Santoro and McGuire, 1997), 'sympathetic elites' (Tarrow, 1998) or 'institutional entrepreneurs' (Reichman and Canan, 2003).
While these interventions are welcome, and go some way towards addressing the first question I raise above, social movements are examined insofar as they represent subaltern groups, seeking to gain access to political institutions to advance progressive policies. Social movements studied include those promoting lesbian, gay, bisexual and transgender (LGBT) rights, disability rights, feminism, welfare state politics and affirmative action (Davidson, 2018;Pettinicchio, 2012;Santoro and McGuire, 1997). The focus is almost exclusively on social movement influence in the agenda-setting stage rather than, as I will show, playing a role in both policy formation and implementation. Moreover, with notable exceptions (Pettinicchio, 2012) the assumption is that institutional activists are reactionary, taking up pre-existing causes of outside agitators (Santoro and McGuire, 1997), rather than proactively pursuing their agenda.
Second, this neglects extra-institutional mobilisation by elite groups. Few social movement scholars have studied elite social movements (Boies and Pichardo, 1993). Elite groups do participate in extra-institutional organisations such as elite clubs, policy planning and lobby groups (Carroll and Sapinski, 2010;Domhoff, 1975;Sklar, 1980); though, with few exceptions (Boies and Pichardo, 1993;Sklair, 1995), scholars rarely conceive these as social movement organisations. There is an underlying assumption that elite groups with easy access to institutional channels will not perceive extra-institutional movement organising as beneficial. In fact, institutional channels of power can be disadvantageous. As Pichardo (1995: 24) argues 'since institutional channels of power are governed by fixed . . . rules, they can restrict power as well as enhance it'.
Literature on political violence has long demonstrated the utilisation of extra-institutional mechanisms for militaristic action and para-statal violence. Some of this work has shown how state actors deliberately use para-military, covert or other private actors to circumvent legitimate means of authority (Jamieson and McEvoy, 2005). Janine Wedel's (2009) work demonstrates how 'flexians' -flexible networks of retired military actors, politicians, academics and experts -form shadowy networks of power that can evade institutionalised democratic accountability. Elite groups may, therefore, organise outside political institutions to bypass fixed rules and circumvent legitimate means of authority (Boies and Pichardo, 1993: 64).
Thus, if we lose the assumption that social movements necessarily mobilise for progressive causes (Blee and Creasap, 2010;Cunningham, 2013;Diamond, 1995) and that extra-institutional mobilisation is not only utilised because of a lack of access to power, we may begin to see alternative forms of social movement organisation.

Beyond a Narrow Conception of Political Repression
The 'challenger/authority' dichotomy has shaped repression research. First, following the influence of PPT, it is the reaction of elites and authorities that is assumed to determine the likelihood of movement success or failure. Consequently, the majority of repression research has treated repression as an independent variable -examining how repression has affected the mobilisation of challenger groups (Gupta et al., 1993;Lichbach, 1987;Moore, 1998). There has, however, been a growing area of work that has considered repression as a dependent variable -seeking to understand the allocation (Cunningham, 2003) and nature (Earl, 2003(Earl, , 2011 of repressive activity. With only notable exceptions (Cunningham and Soto-Carrión, 2015;Marx, 1974), sociological analyses of state-led repression have neglected the specific social movement-like behaviours, tactics and strategies of governmental actors, covert agents and officials involved in repressive activities. State-led repression often exhibits direct involvement in social movements. During the Cold War, the CIA infiltrated student groups; the FBI, domestic counterintelligence programme, COINTELPRO planted informants in universities to disrupt student movements (Churchill and Vander Wall, 1988;Cunningham, 2003;Johnson, 1989). Governmental actors are not necessarily passive in these direct interventions; they may use their positions to pursue their own goals and ideological agendas. Gary Marx's (1974) seminal essay on infiltrators called for more understanding of the logic and strategy of infiltration -yet the field has been slow to take up the call. As Cunningham and Soto-Carrión (2015: 158) argue, we should not simply see infiltrators unproblematically as 'agents of the state', it is precisely the 'interstitial position -between policing and activist arena' -that defines the infiltrator role.
Second, the repressive 'authority' in the 'challenger/authority' antagonism has overwhelmingly been assumed to be the state. This view neglects the ways an 'authority' (whether a state, a corporation or even an educational institution) might deploy third party actors to undertake repression (Earl, 2003;Ferree, 2005;Pichardo, 1995). As Earl's (2003) thorough review of the repression literature demonstrates, there is little exploration of the relationship between different types of repression such as state and private repression. Pichardo's (1995) study of the Associated Farmers of California in the 1930s demonstrated how the power elite not only sponsored but directly engaged in mobilising nonelite actors to undermine strikes, unionisation and the New Deal. Diamond (1995: 35, 178-195) shows that during the Cold War state actors began incorporating para-statal institutions -including neoconservative groups -in their efforts to combat communism.
Indeed, as Ferree (2005) demonstrates, even 'softer' forms of repression expressed through institutions of civil society can disarm social movements. Soft repression is a cultural strategy that describes the mobilisation of 'non-violent', often informal, means of 'silencing . . . oppositional ideas' from 'a public forum' (Ferree, 2005: 141). She claims such strategies can often be directed against movement collective identities in order to weaken their mobilisation. Using the example of the feminist movement she identifies three effective strategies that undermine movements -interpersonal ridicule (micro-level), stigma (meso-level) and silencing (macro-level) -each demonstrating different levels of institutionalisation. Socially subordinate groups are regularly subjected to negative forms of external categorisation. As such, liberation movements have mobilised in defiance of such stigma, creating resistant collective identities such as 'black is beautiful ' and 'queer'. For Ferree (2005), the repressive use of stigma is often used precisely against successful mobilisations of collective identity, seeking to break positive associations with the movement. Such approaches often make use of existing forms of stigma in contemporary relations of domination, for example by drawing on sexist or racist categorisations. In the discussion below, I demonstrate how stigmatisation of mobilisations against Prevent rely on racialised definitions of extremism and radicalisation that associate Muslims with terrorism.
Following Earl (2003Earl ( , 2011, I argue that definitions of repression limited to statebased coercion are too narrow to account for the reality of movement suppression. In 1978, Tilly defined repression as 'any action by another group that raises the contender's cost of collective action ' (1978: 55). This not only accounts for repression by non-state actors but also broadens the remit of repressive activity. Channelling, for example, is another form of repression, less accounted for in movement scholarship. Channelling -an indirect means of control -often in the form of regulation, uses 'a reward and consequence structure' that restricts the possibility for certain types of protest (Earl, 2011: 264).
In the following section, I discuss how the UK Prevent programme introduced legal powers as well as institutional policies that restrict the potential for specific civil society organisations. At the same time, Prevent also shapes the terrain of collective action through an opportunity structure of funding and institutional openings. I will demonstrate in this article that Prevent channels an opportunity structure enabling the institutionalisation of 'soft repression'.

Channelling and the UK Prevent Programme
Channelling, although having a powerful impact, does not necessarily have to be designed to quell protest. Its impact can be indirect and hidden. For example, internal organisational disciplinary codes can have a profound impact on the possibilities of collective action, without explicit recognition of this. From the outset Prevent shaped the possibilities for collective action through both regulation and facilitation. It placed restrictions on mosques and civil society organisations through charity regulation and institutional changes in public institutions.
While these policies have had a restrictive impact on forms of political mobilisation, they have provided institutional opportunities for others. Using a more 'dialogical' approach (Meyer, 2005), we can see that while movements do shape policy outcomes, policy changes shape movements. In this case, institutional opportunities have shaped the terrain of political contention of Prevent.
The Prevent counter-terrorism programme aims to 'stop people becoming' or 'supporting terrorists' by 'challenging extremist (and non-violent) ideas that are also part of terrorist ideology' (HM Government, 2011: 6). The policy is built on the assumption that a process of 'radicalisation' causes acts of 'terrorism'. 'Terrorism' is seen as rooted in the circulation of extremist ideologies. As many have shown, the ambiguity in definitions of 'extremism' and 'radicalisation', has meant that in practice Muslims are disproportionately targeted (Kundnani, 2014;Massoumi et al., 2017).
The first iteration of the UK Prevent programme was introduced in 2003. The invasion of Iraq shifted the relationship between the government and Muslim civil society organisations. In 2006, the New Labour Minister, Ruth Kelly, announced the government's intention to actively: 'develop relationships with a wider network of Muslim organisations' especially those 'taking a proactive leadership role in tackling extremism and defending our shared values' (Kelly, 2006). This included a range of groups less critical of the invasion of Iraq than mainstream Muslim opinion -all funded via the Prevent programme. This saw the largest ever injection of funding (£60 million) into Muslim civil society in the UK (O'Toole et al., 2013). Yet, there was also widespread criticism for marginalising organisations critical of government (O'Toole et al., 2013: 56;Rashid, 2014). Prevent funding was also allocated to local authorities on the basis of the proportion of Muslims living within the local population. While 'capacity building' initiatives were focused on empowering what the government perceived to be marginal (and potentially more government-friendly) voices within Muslim communities -namely women and young people -there was also a direct attempt to influence the direction of religious thinking and practice through religious roadshows, training programmes and interventions from the Charity Commission. In this way, the Prevent funding structure shaped the landscape of Muslim civil society. From 2012 onwards, the UK (Coalition) government, shifted away from directly funding Muslim civil society. Instead, they gave alternative, covert, forms of support to Muslim civil society to disseminate pro-government messaging. The implications of this shift are discussed below.
The Charity Commission -the UK charity regulator -was initially tasked with regulating mosques' finances and codes of practice through a charitable status registration drive (HM Government, 2011). This gave greater powers to intervene in the activities of mosques. The Charity Act 2006 required all student unions to register with the Charity Commission, subjecting them also to regulation. The Charity Commission has since investigated 'extremist' 'activities, literature and speakers at charity premises and events in both religious organisations and student unions' (HM Government, 2011: 94). In 2011, the role of the Charity Commission in Prevent intensified. Following the appointment of William Shawcross (formerly a trustee of the neoconservative Henry Jackson Society) as chairman of the Commission in 2012, investigations of Muslim charities under the new extremism and radicalisation code made a steady increase (Ramesh, 2014).
The Counter Terrorism and Security Act 2015 (CTSA), created a statutory duty -the 'Prevent duty' for public institutions to pay 'due regard' to 'prevent people from being drawn into terrorism'. As a result, public institutions have implemented new policies indirectly constraining social movements. Within higher education, universities are required to put in place several policies to enact the duty. Whether officially intended or not, these have placed restrictions on campus activism. Universities now have information-sharing agreements with law enforcement agencies, external speakers are 'risk assessed' and monitored. The programme shares similarities with US state legislative responses to campus protests against the Vietnam War in the 1960s when both criminal and civic statutes restricted access to university campuses and mandated restrictive internal university governance (Gibson, 1989;Price, 1972;Rowland, 1972).
Moreover, these new regulations mean that public institutions have had to make significant changes to policy and practice. Lacking resources and expertise, they often rely on external partners to deliver training and resources. As a result, counter-extremism think tanks and civil society organisations have become key providers. This gave them a platform to disseminate their ideas, as well as access to public institutions.

The Prevent Enforcement Network
The UK Prevent programme was introduced in the context of significant mobilisations against the UK government over the Iraq war. During this time, the anti-war movement expanded to include opposition to counter-terrorism in a broader struggle against the 'War on Terror' (Massoumi, 2015). These existing mobilising structures meant that when the Prevent duty was introduced, it faced significant opposition from a range of campaigning groups (Massoumi, 2015;Massoumi et al., 2017).
In response, the government made alliances with actors already active in this area, including the neoconservative movement -a key opponent of the anti-war movement. The formation of these alliances resulted in an 'enforcement network' comprised of both 'state and private actors seeking to maintain the status quo' who share 'direct connections', 'mutual awareness' and 'shared understandings' (Cunningham et al., 2019: 320). The implementation of Prevent relied on an enforcement network of the following actors: • • State-supported movement organisations, largely counter-extremism groups, which rely on government support (funding, institutional access and other resources). • • Neoconservative groups -these have a longer-standing history with origins in US anti-communist efforts (Diamond, 1995). In Britain, think tanks such as the Henry Jackson Society (HJS) and Policy Exchange have published a series of reports purporting to show evidence of extremism among British Muslims; calling on the government to sever links with particular groups and to expand counter-terrorism measures. As we will see, they have had some success in influencing policy and practice. • • Governmental actors -a specific group of government officials engage in extrainstitutional activism to support Prevent (Verhoeven et al., 2015). The group includes local, regional and higher education Prevent coordinators. In fulfilling the role of Prevent implementation, they engage in extra-institutional activities (social media, participate in public meetings), with their allies in the enforcement network.
The quasi-independence of SSMOs and neoconservative organisations, from official governmental roles, offers the government certain advantages in Prevent implementation. Here Ong's (2018) conceptualisation of 'thugs for hire' -a case of privatised coercion -offers some useful principles for their analysing their para-statal roles. Ong's (2018) case refers to a context in which local states in China use violent non-state actors to bolster their coercive capacity. She identifies three conditions in which the state adopts third party actors as a preferred strategy: (1) when implementing illegal or unpopular policies; (2) when seeking to evade responsibility; or (3) when they are weak in their capacity. Moreover, such third party actors are distinct from the state, but act as extensions of it, yet their lack of permanence means they could be disavowed.
In the case of Prevent implementation governmental actors operated in alliance with non-state actors. Third party actors were mobilised in order to gain legitimacy for the programme, but in such a way that government could distance itself from them. In short, non-state actors are able to circumvent the institutional and political barriers that often constrain governmental actors. The implication of this, however, is that their activities are neither transparent nor accountable, and therefore lack regulation or oversight. The para-statal nature of such actors, therefore, and their shifting but quasi-independence from the state demonstrates the need to rethink state-movement relations and consider the importance of extra-institutional spaces for elite mobilisations.

Methods
This British Academy funded study examines the impact and implementation of counterterrorism initiatives in the UK. Informed by the tradition of Power Structure Research (Domhoff, 2005;Mills, 1981Mills, [1956), I focus on the 'network of organizations and roles . . . responsible for maintaining the general social structure and shaping new policy initiatives' (Domhoff, 2005). Here I used a multi-site ethnography, to 'follow people, connections, associations and relationships across space' (Falzon, 2009: 2) that implemented and responded to Prevent-related activity.
Between September 2015 and April 2019, I attended 31 national and local Preventrelated public meetings. I adopted a purposive sampling strategy to examine the activities of key actors in relation to the Prevent policy, including: 1. counter-terror actors responsible for overseeing the policy; 2. higher education actors (e.g. Universities UK; security officials; university staff) implementing the policy; 3. social movements and campaigners responding to the policy (Islamic societies, anti-war and civil liberties groups, conservative and other groupings); 4. think tanks, policy planning and counter-extremism groups.
I selected the following events: 1. policy seminars and government briefings; 2. Prevent training sessions; 3. anti-Prevent campaign events (e.g. Students not Suspects); 4. events with think tanks and non-governmental organisations (NGOs) supporting Prevent (e.g. Quilliam Foundation, Inspire) (see online Appendix I).
My interest was in alliances, coalitions and conflict networks. Using informal interviews (including some during participant observations) and documentary analysis I sought to examine 'what is said and done within the power network' (Domhoff, 2005). As the 'public script' often tells a different story to the hidden transcript that takes place behind closed doors (Scott, 1990), I approached this from an 'investigative' perspective (Douglas, 1976). I used Freedom of Information (FOI) requests, internal documents and interviews. This gave me access to a number of internal documents. The documentary analysis, discussed below, focuses on internal Home Office documents (see online Appendix II), which provide detail on the relationships with pro-government civil society groups. These internal documents give insight into 'offstage' communications that take place outside of public scrutiny (Scott, 1990). From an ethical perspective, it is both of sociological and in the public interest to examine the internal thinking behind government practices. As Spicker (2011: 124) argues: 'some actions are public in their very nature. The formal actions of governments and public agencies, even in closed rooms, are intrinsically public.' The research reported here was given ethical clearance by the relevant research ethics committee.

The Role of State-Supported Movement Organisations
In this example, I examine how the government sought to mobilise civil society campaigns to deal with opposition to Prevent and help generate (apparent) grassroots support for their policies.
Since 2011 the UK government has shifted the nature of its engagement with Muslim civil society towards a more covert strategy. A strategic communications unit in the Home Office -the Research Information and Communications Unit (RICU) was set up in 2007. In January 2012 Theresa May (then Home Secretary) stated that RICU had a new focus on 'engaging credible civil society organisations in order to encourage these organisations to challenge radical and extreme views in their local communities.' She said: it is more effective to be working through groups that are recognised as having a voice . . . rather than it being seen to be government trying to give a message. Indeed, it's always better to be using those people to whom people look naturally to hear the message, rather than simply doing it as RICU itself. (Intelligence and Security Committee, 2012)

#MakingAStand and Inspire
One example of this work was the government-supported campaign #MakingAStand, led by the organisation Inspire. Inspire described itself as an 'independent non-governmental' 'counter-extremism & women's rights organisation' (Inspire, 2016). Its aim was 'empowering women' (Inspire, 2014). Inspire is said to have started with no funding, being run out of (founder) Sara Khan's kitchen; yet it rapidly rose in profile .
#MakingAStand was launched in September 2014 and was splashed on the front page of the Sun newspaper, featuring an image of a Muslim woman wearing a Union Jack headscarf: 'United Against I.S.' Inspire toured the country aiming to 'empower' Muslim women to 'make a stand' against extremism. However, internal documents described the campaign as one of four 'RICU products' (see online Appendix III: Figure 1). In a section marked 'not for public disclosure', the campaign is described as creating a 'network of British Muslim women across Prevent priority areas' in order 'to transmit HMG [Her Majesty's Government] counter-extremism messages into communities and hard to reach audiences'. The 'resources required' were 'access to women's civil society groups' (online Appendix III: Figure 2).
To undertake this work RICU hired a PR firm, Breakthrough Media (BM), to disseminate counter-extremism messaging appearing to come from Muslim community groups (online Appendix III: Figure 3). BM PowerPoint slides show this included Inspire's #MakingAStand campaign (online Appendix III: Figures 4 and 5) as well as a prominent hashtag campaign #notinmyname that was attributed to the Active Change Foundation (ACF) -a youth work focused community organisation (online Appendix III: Figures 6  and 7). The latter hashtag campaign, also launched in September 2014, was described as a 'rallying call for Muslims everywhere to reject ISIL'.
The target audience for this activity was described in the internal documents as 'Muslim males aged 15-39' (RICU, p. 9). The overall ambition was to find 'credible voices' to promote a 'reconciled British Muslim identity' (Cobain et al., 2016). This relied on the stigmatisation of certain other forms of Muslim identity (via the association with extremism and radicalisation); this can be taken as a clear example of soft repression.
It should be noted here that 'not in my name' was a key slogan of the British anti-war movement. The co-option of this slogan, is evidence of an attempt to appeal to such audiences.
The documents show that RICU required Breakthrough Media to keep the connections between them discreet. In one document, RICU's activity is described as operating in three ways, the third of which is 'Discreet campaigns supported by RICU without acknowledgement of UK Government support. Campaign products are never directly attributed to RICU' (RICU, p. 6).
RICU worked 'at an industrial scale and pace' (Cobain et al., 2016) to create campaigns -promoting pro-government counter-extremism messaging. Unlike infiltration, which is a mechanism for surveillance, discrediting or at times attempts to derail a movement (Cunningham, 2003;Marx, 1974), the primary aim here is to generate (at least the appearance of) grassroots support for government policies. Such 'astroturfing' is a practice often associated with corporations (Stauber and Rampton, 1995). Such campaigns deceive genuine Muslim civil society actors, who are then unable to make informed choices about their collaborations with partners.
Inspire's broader activities include delivery and design of government-approved Prevent training and speaking at policy events and seminars promoting Prevent work. Furthermore, Inspire director, Sara Khan's sister, Sabin Khan is the deputy director of RICU having worked there since April 2009. Kalsoom Bashir (co-director of Inspire between 2014 and 2017) worked as Prevent lead for Bristol City Council for some of this time and was then seconded to the Avon and Somerset Constabulary as regional Prevent trainer . Despite these links to government counter-terrorism activities, actors and sponsorships, Inspire (2016Inspire ( , 2019 continued to describe itself as independent.

Neoconservative Social Movement Organisations
Neoconservative groups also play a para-statal role in implementing the Prevent programme on UK campuses.
Neoconservative think tanks influenced the shift towards non-violent extremism in the second iteration of Prevent discussed above. For instance, a Policy Exchange report in 2009 -Choosing our friends wisely -criticised the Labour Government for, 'stress[ing] law enforcement and strict security concerns over and above everything else' (Maher and Frampton, 2009: 62). It claimed that 'non-violent radicals', were 'indoctrinating young people with an ideology of hostility to western values' (Maher and Frampton, 2009: 5). The revised Prevent strategy, published in June 2011, stated that: 'preventing terrorism will mean challenging extremist (and non-violent) ideas that are also part of a terrorist ideology' (HM Government, 2011). The strategy document cited another neoconservative think tank, the Centre for Social Cohesion (CSC), on five separate occasions (HM Government, 2011;. Neoconservative groups have been active on campus. Student Rights, a project of the Henry Jackson Society, sparked a national debate about the issue of gender segregation at Islamic society events in 2014. Student Rights claimed a quarter of Islamic society events were gender segregated (Sutton and Kassam, 2014). Later that year, Universities UK guidance stated that events in universities could legally segregate on the basis of gender. This was later withdrawn after PM David Cameron made a statement that segregation was not acceptable in universities. In the end, the Equalities and Human Rights Commission guidance emphasised that gender segregation would be unlawful outside religious practice. It is still not clear, however, whether gender segregation is particularly widespread at Islamic society events. The Student Rights report used a highly selective and non-representative sample. Furthermore, the relevance to the Prevent duty is unclearthere is no evidence that segregation on gender is a risk factor for 'being drawn into terrorism'. Yet, the issue of gender segregation features in the CTSA guidance for higher education institutions (HM Government, 2015: 4), showing that Student Rights efforts to create concerns were successful in influencing public debate, policy and practice.
David Cameron's speech, on 17 September 2015 -days before the Prevent duty came into effect -cited 70 extremist events taking place at universities. A Downing Street press release named four universities as most often hosting 'extremists'; six speakers were named as 'expressing views contrary to British values' (Prime Minister's Office, 2015). However, it later transpired that the names had come to the Home Office's Extremism Analysis Unit directly from HJS (Butt v. Secretary of State for the Home Department, 2017; Grove, 2015). This example provides further evidence of neoconservative organisations operating as quasi-independent extensions of the state.
While neoconservative organisations focus on discrediting Muslim civil society actors, SSMOs intervene directly in Muslim civil society activities, as the next two examples illustrate.

Observations from a Muslim Civil Society Event
In the first example, I attended a Muslim civil society event advertised as hosted by a community organisation at a local school. I was surprised to be greeted by a team of people giving out 'Prevent Tragedies' leaflets and watched as they assembled their banners on the stage. Prevent Tragedies is a UK 'police and partners' Prevent initiative. It quickly became evident that the local Counter-Terrorism Intelligence Unit (CTIU) were playing a central role in the event. Police officers made introductory speeches claiming that their work was about safeguarding, not racial profiling.
At least four members of the CTIU were present, but they did not all declare their roles and were not in uniform. Alongside the CTIU, there were two SSMOs present. One was a counter-extremism group focused on women's rights, the other an apparently youth-led counter-extremism group. Both had been in receipt of Home Office funding and had police on their boards. The activists associated with these groups appeared to be both familiar and friendly with the local CTIU and they sat with them rather than the main (community) audience. The SSMOs claimed to be speaking on behalf of the local community, local Muslim women or the local youth community. For example, one speaker said: I have been to work in schools . . . Somali girls that thought they wanted to go to Syria because these girls had been in sexual relationships with young Somali men . . . they felt that the best way of redeeming themselves -of going to Heaven -was to go to Syria and marry a Jihadi fighter . . . it is about us as mothers, as parents having the courage to talk to our children. (Participant A) Participant A was proposing culturalist explanations for why young girls apparently wanted to travel to Syria, implying there was something particular about Somali culturenamely its patriarchal practices -that led to girls travelling to Syria. One man challenged the speaker for undermining Muslim cultural identity: your presentation at the top of your voice . . . [using] unsubstantiated evidence [that] young girls were running from their own community and they want to have boyfriends and children, . . . can you please provide some substantiated evidence, because this is [our] cultural identity, [we are] a Muslim community and this information is very, very damaging, will you provide that please?
The audience applauded loudly and there was an antagonistic atmosphere in the room.
The two SSMOs promoted the Prevent programme more aggressively then the CTIU officers. For example, one of the activists warned against listening to critical arguments about Prevent (which had been made): And I'm meeting next week with a young woman who has come out of prison. She was sent to prison for engaging with terrorist activity . . . She said . . . she was offered help [from Prevent] but she was suspicious of that help because she'd listened to the arguments that have been made here today and she said 'I wish I had accepted that help because now I have got a prison record it's difficult for me to get a job'. (Participant A) At this event, both CTIU officials and SSMOs sought to directly intervene in the physical space of Muslim civil society -both overtly (in trying to 'badge' the event) and covertly (promoting organisations representing a 'grassroots' voice). Both of the SSMOs advocated culturalist explanations for alleged extremism in the Muslim community -criticising the community itself. In contrast, the counter-terrorism practitioners offered a conciliatory approach by offering to help the community and expressing opposition to racial profiling. It appeared that SSMOs were attempting to utilise their semi-autonomous position from the governmental actors to push a harder line on Prevent.
At this event, the strategy did not appear successful. The legitimacy of the SSMOs and the claims made did not appear to resonate with the audience. Nevertheless, some appeared to be intimidated. The two main organisers of the event (representatives from the ostensible host -the Muslim civil society organisation) seemed concerned that CTIU officers were being challenged, and gave them more opportunities to speak than anyone else.
This example demonstrates the ways in which counter-terrorism practitioners and SSMOs work together to promote Prevent and steer the direction of Muslim civil society activities.

Prevent-Related Events: Policy Seminars and Prevent Training
The second example is public training sessions on Prevent. Home Office created sessions are called 'Workshop to Raise Awareness of Prevent' (WRAP). Third party providers run similar events -some, not all, officially approved (HM Government, 2016). In addition, public policy events bring together civil servants, counter-terrorism and higher education practitioners. While these events served different purposes, there were remarkable similarities in the content, activity and even participants of these events.
SSMOs and neoconservative groups are both active in Prevent training but play different roles. Government relies on reports from neoconservative groups to provide (some of) the evidence base and to justify Prevent training (circumventing academic authority). Some SSMOs, consistent with their closer relationship with government, carry out Prevent training themselves. For example, in every live Prevent training I observed, examples were drawn from Quilliam and HJS. The Safe Campus Communities website -the officially sanctioned Prevent resource for higher education -links to material from Quilliam and Inspire (and other pro-Prevent groups). The outcome of these interventions is a highly politicised Prevent training programme -pursued as a social movement strategy.

Institutionalising Soft Repression Strategies
Neoconservative groups use Prevent-related events to disseminate misinformation to discredit Muslim civil society and human rights groups. They claim a link between Muslim organisations and extremism in, for example, an annual report on Extreme speakers and events at universities (Fox, 2018;Sutton, 2015;Wilson, 2017). The evidence underlying these publications is insubstantial and claims often unfounded. Promoting one such report, the Henry Jackson Society states: 'This report warns that public sector organisations and elected officials are partnering with the extremist-linked group Muslim Engagement and Development (MEND)' (Wilson, 2017). MEND is a notfor-profit NGO with the stated aims to: 'empower and encourage British Muslims within local communities to be more actively involved in British media and politics' (MEND, n.d.). This report, alongside another, Profiting from prejudice: How Mend's 'IAM' campaign legitimised extremism were used by HJS activists to undermine a MEND organised Islamophobia Awareness Month event in parliament in November 2018 -by pressuring MPs billed to speak not to attend (Fox, 2018).
Ferree's conception of soft repression is useful in understanding the racialised stigma and silencing strategies used by neoconservative groups and SSMOs alongside governmental actors to associate Muslim civil society actors with 'extremism' or 'radicalisation' (Massoumi et al., 2017). I describe below three examples of these strategies and their institutionalisation in Prevent training.
The first is the use of the stigmatising term the 'anti-Prevent lobby' to describe critics of Prevent. The label was mostly, but not always, reserved for Muslim civil society organisations said to spread misinformation about Prevent. One counter-terrorism practitioner said: 'we hear of Muslim communities being scared of engaging with this agenda [Prevent] -not because of the government but because . . . information that is being fed to them from elements within the Muslim community' (Prevent practitioner). She went on to name a number of Muslim groups, none of which, however, were engaged in any more than democratic activity. Similarly, at another policy event, a Home Office official cited an HJS report on extremism. A Quilliam spokesperson at the event, criticised academics who had signed a letter opposing Prevent as indicating extremism. These examples constitute what Ferree (2005: 144) described as stigma at a meso-level -they act to discredit a collective group for their identities and association with a group as a whole. She argues that the extent to which this moves to the macro-level relates to the degree that silencing becomes embedded in ordinary institutional practices.
For Ferree (2005: 148) while silencing can be a mere side effect of relations of domination, soft repression entails a more deliberate strategy of exclusion. Neoconservative actors mobilise these strategies through activities resulting in a steady stream of press items smearing Muslim activists as extremists -aiming to undermine their participation in university events (Osbourne et al., 2016). Governmental Prevent actors mobilise these strategies within Prevent training, institutionalising them in the practice of the university.
The Prevent duty requires universities to have external speaker regulation policies. Universities use training programmes to maintain consistent application of these policies within their institutions, and the Higher Education Funding Council for England (HEFCE) (the forerunner to the Office for Students (OfS)) commissioned a national training programme in order to ensure consistent practice across universities -the Leadership for Higher Education Prevent (LfHE) training programme.
Data from responses to FOI requests to over 100 universities in England, demonstrate that the majority are using the LfHE programme to provide online training to staff. In a slide in the training, a hypothetical scenario, it is asked whether 'a member of controversial human rights group Cage' should be allowed to speak at the university (see online Appendix III: Figure 8).
In order to get the correct answer, the choice 'Conditional yes -but with appropriate mitigations in place such as another speaker or a neutral chair' must be selected. The following slide then states that 'institutions face criticisms for allowing certain events to proceed where extremist views are represented' (online Appendix III: Figure 9). Cage is a UK-based organisation that campaigns against human rights abuses occurring in the context of the 'War on Terror'. As a prominent critic of government counter-terrorism policy, Cage has been regularly targeted by supporters of Prevent (Mills et al., 2015). The official training programme for higher education identifies Cage as 'extremist' and recommends that they can only speak on campuses with full mitigation -a clear example Ferree's (2005) notion of institutionalised silencing.
In the face-to-face training sessions I observed, some university staff were extremely critical of the racial stigmatisation of Muslims. In one session, three participants heckled the lead trainer, challenged the training material and accused her of focusing too much on Muslims and the 'Islamist threat'.
In addition, a number of research participants have reported difficulties in organising events at universities, particularly with Cage, but also MEND and other politically active Muslims. One safeguarding lead referred to an incident where the Prevent referral team examined an event because a member of Cage was speaking. Another academic involved in organising a conference on 'terrorism' was asked by the police for the full delegate list because a member of Cage was attending. Several conferences on Islamophobia and racism have been cancelled at universities following bureaucratic procedures resulting from new external speaker policies. In each of these cases, organisers I spoke to said that they felt intimidated by the intense pressure from managerial interventions.

Conclusions
I have argued above that the UK Prevent programme is a form of political repression enacted by both governmental and non-governmental actors. Rejecting the challenger/ authority model of confrontational state-movement relations, I have instead shown that the Prevent programme is implemented by an enforcement network which sits within and around the state.
This has two implications. The first relates to the role of state actors in political mobilisation. Rather than consider the state as simply a reactive or even unitary actor, we can observe from this case, how a group of state actors are proactively involved in forms of extra-institutional activity to further a set of ideological objectives. The second, relates to the conception of civil society. Rather than see civil society as some kind of counterbalance to state power, civil society is utilised for the purposes of advancing state capacity in counter-terrorism.
Moreover, the involvement of civil society groups shaped the nature of repression, which took the form of 'soft repression'. Civil society groups do not usually have access to the range of coercive measures at the disposal of the state (this does not always mean that they do not engage in them). Prevent was supposed to be an alternative to 'harder' forms of counter-terrorism policing. The involvement of civil society groups was to bolster the legitimacy of the programme and to undermine ideological and organisational opposition.
Second, the extra-institutional nature of this activity and the (albeit shifting) parastatal nature of these agencies raises questions of accountability. The analysis demonstrates that SSMOs and the neoconservative groups serve as para-statal agencies with quasi-independence from the state. It would appear that their autonomous positions prove helpful in circumventing legitimate means of authority in implementing Prevent as a distinct form of political repression -this question would be a fruitful enterprise for further exploration and give greater understanding and complexity to the relationship between state actors and social movements. Para-statal agencies described in this article are not subject to democratic accountability and therefore, cannot have their actions restrained by formal rules that govern state actors' practice. I have demonstrated that this has resulted in a broadening of the counterterrorism remit with the impact of repressing Muslim political activism. In light of these findings, we need to consider how to regulate security policy formation and implementation to secure democratic principles as well as adding complexity to our understanding of the relationship between state actors and social movements.