Multi-agency counter-terrorism in Britain and Norway: Intelligence agencies and the administration of welfare

Introduction

What would it mean to be allocated a place at college, or to be given a mentor, through a programme run covertly by an intelligence agency? What would it mean for the norms of democratic societies, if intelligence agencies were able to administer welfare to ‘divert’ people from a path thought too risky, without asking their consent – while harvesting information from a person’s medical, social and educational records in the meantime? Should intelligence agencies, and their counterparts, restrict themselves to investigations and the pursuit of criminal justice? Or should they be able to covertly ‘divert’ persons known to them, through welfare and social policy tools?

To differing degrees, and at differing scales, Norway and Britain ¹ have implemented such programmes. The welfare-administration programmes of these intelligence agencies are not public knowledge; rather, they operate in the non-statutory realm and are collaborations between intelligence agencies and local authorities which manifested in response to lone-actor attacks and fears about ‘returning foreign fighters’ from 2016 onwards. Their names are Operativ Koordineringsgruppe (OKGs; Norway), the Multi Agency Centre (MAC; Britain) and police-led partnerships (PLPs; Britain). The OKGs, PLPs and MAC are not Preventing/Countering Violent Extremism (P/CVE) programmes. P/CVE programmes tend to be owned by local authorities, require the consent of the individual for interventions and limit the role of intelligence agencies within proceedings. The MAC, OKGs and PLPs are, in contrast, all led by (or centrally involve) intelligence and security agencies and operate without the knowledge and consent of the subject.

Despite these differences, ‘multi-agency counter-terrorism’ does replicate some modalities and organizing assumptions from P/CVE. The programmes administer welfare with the intention of ‘diverting’ a person from a path of radicalization, through multi-agency input. This social care is often provided at a level far beyond that available to everyday citizens through welfare programmes. Is it significant, then, that these new programmes allow intelligence agencies to lead multi-agency fora which administer welfare, and to do so covertly?

The significance of these new multi-agency counter-terrorism programmes, I argue, is their disregard for sectoral separation in liberal government. Liberalism is prefaced upon governing through freedom and the avoidance of ‘governing too much’ (Foucault, 1997; Hindess, 2005; Rose, 1999). In performative contradistinction to authoritarian societies (cf. Heath-Kelly, 2022), liberal political systems have traditionally afforded high levels of independence to civil society and fostered the public performance of policing and surveillance as separate from education, health and public administration. This leads to ostensible sectoral separation. Mitchell Dean’s (2002) work provides a typology of the categorization of subjects under liberal governmentality, as requiring variegated involvement from the state to foster self-government; categorization which leads to the constitution of distinct governmental agencies with distinct remits. Those requiring temporary support to render them economically productive are allocated to welfare and employment services; those deemed to require more substantial support are allocated to health, social services and integration services; while those manifesting antagonism to society and state are allocated to the repressive agencies (the police and security services) (Dean, 2002). Such sectoral separation of functions, Dean argues, enables the liberal arts of government to performatively conjure the illusion of a limited policing role within society. In reality, the education, care and health sectors take on regulatory and coercive roles – constituting a ‘liberal police’ which supports and drives the standards expected of the population.

Of course, Dean’s typology is an ideal type rather than a reflection of solid lines which distinguish the functions of governmental agencies. Policing agencies have embedded relationships with healthcare, social work and education – overlaps which have expanded in scope and scale with the rise of multi-agency working. The use of Dean’s typology in this article is not supposed to suggest that health, education and social care are somehow untouched by policing and surveillance rationales and practices; rather that sectoral separation or ‘compartmentalization’ can conceal much of the attendant violence from public view (Pachirat, 2011).

Acknowledging this sectoral overlap between care and policing should not, however, blind us to the exclusion of intelligence and security agencies from educational institutions, healthcare premises, social care and employment agencies. Not all policing work has moved transversally across sectors. The intelligence agencies remained (until 2016) primarily occupied with the business of covert observation, communications interception and running undercover agents. This is what makes the new multi-agency counter-terrorism programmes important: they bring the state’s most powerful enactors of repression into the business of welfare. This undoes the performance of sectoral separation.

This article outlines the existence of these new counter-terrorism programmes, analysing their significance for liberal governmentality. Using Dean’s typology of liberal government, the article argues that multi-agency counter-terrorism marks a distinctive moment in the transformation of liberal governmentality – where normative commitments to freedom and ‘not governing too much’ are maintained in rhetoric, but sectoral separation is abandoned in practice. The interweaving of the state’s most proficient repression agents (intelligence and security agencies) within the means and ends of the social policy sector marks a significant extension of state power. However, these practices of multi-agency counter-terrorism are particularly interesting because they evade conventional definitions of illiberal practice ² (Glasius, 2018), human rights abuse or authoritarian legality ³ (Boukalas, 2017). Individuals are covertly intervened upon not to pursue detention or prosecution, nor to repress their free speech or religious belief through overtly violent means. Instead, intelligence and security agencies appropriate the pastoral means of the ‘liberal police’ to covertly intercede, and intervene, upon a person’s autonomy. This is ‘care’, in its most coercive and disciplinary form, manifested at the behest of the state’s elite repressive agencies.

Sectoral separation: Moving lines within risk, liberalism and the War on Terror

There are many theses applied to understand counter-terrorism within academic research. Postcolonial theory is productively used to identify the historical continuity of policing practices, legal discrimination and discourses which marginalize, brutalize and pathologize racialized people – tracing these practices from imperialism to present-day counter-terrorism (Abu-Bakare, 2020; Dixit, 2022; Kolb, 2021). The science and technology studies and materialist canons are drawn upon by scholars to trace the digital counter-terrorist tools which preemptively securitize people on the basis of data transfer, profiling, patterns of behaviour and algorithmically derived correlations within datasets (Amoore, 2013; Aradau and Blanke, 2015, 2018; Glouftsios and Leese, 2023). And legal scholars have interrogated the surge towards precursor offences, pre-crime and preemptive criminal justice in the counter-terrorism realm (McCulloch and Wilson, 2017; Walker et al., 2022; Zedner, 2007), developing models like ‘authoritarian legality’ (Boukalas, 2017) to capture the implications for rule of law.

Capturing the move of intelligence and security agencies into the administration of welfare is a niche task, however. Programmes like the OKGs, PLPs and the MAC are non-statutory and non-criminalizing, and evade the reach of most theorizing on risk and preemption in the War on Terror. They move ‘risky’ subjects into the realm of social policy rather than subjecting them to exceptional powers such as Terrorism Prevention Investigation Measures (TPIMs, under which individual rights and liberties can be severely curtailed) or ‘no-fly’ listing. And although algorithms and dataveillance play a significant role in much security work, the multi-agency panels convened under the MAC, PLPs and OKGs are resolutely analogue – based upon agency representatives sitting around a table, discussing case notes (Bonelli and Ragazzi, 2014). Furthermore, whereas previous research has shown the significant overrepresentation of racialized Muslims in non-statutory, multi-agency counter-terrorism in the UK (Aked et al., 2021), this article attempts to capture the novelty and strangeness of intelligence agencies administering social welfare. The ‘sectoral shift’ involved in this repurposing of care work speaks most directly to Dean’s description of sectoral separation in his studies of liberal governmentality and its transformation.

Dean (2002) has written insightfully on how sectoral separation reflects the liberal arts of government. Different agencies enact differing styles of intervention upon the public, to ensure proportionality and avoid the danger of ‘governing too much’. Governing through freedom entails that the liberal state specifies the content of individual freedom and turn it towards specific productive goals (Dean, 2002: 38). The liberal state indulges in categorization of its subjects (Dean, 2002; see also Hindess, 2001; Valverde, 1996), according to their capacity for self-government, which determines the agency which will manage them through either pastoral or repressive means.

Governing through freedom, then, necessitates that these different subject groups are managed by different agencies. The prescient conclusion found in Dean’s (2002) examination of ‘liberal authoritarianism’ is that liberal democracies expand their powers of social control through care. Expanding repressive agencies and their powers of surveillance and violence is discouraged, normatively, so the liberal state instead extends social control through its social policy sectors. Despite maintaining a de facto separation of repressive agencies from caring agencies, then, the liberal state enacts authoritarian measures through its social policy agencies (Dean, 2002). Dean states:

liberal rationalities and authoritarian measures are far from incompatible . . . To understand the entwined facilitative and authoritarian sides of liberal government, we need to examine more closely the claim that liberalism seeks to govern through reform and utilize the existing agencies, mechanisms and regulations of civil society. (2002: 39–40)

To that end, he offers us a typology of intervention providers – who practise surveillance, regulation and control across the spectrum of agencies. By nominally separating police and security agencies from care providers, liberal governmentality maintains the illusion of a free civil society space outside government regulation and control. This results in the ostensible sectoral separation of policing from care, health and education, while social policy actually makes significant incursions upon individual freedoms. Indeed, in conversation with Jacques Donzelot, Dean calls these social policy agencies ‘the liberal police’ to highlight their regulatory function (2002: 44).

When the liberal state locates a rational self-governing subject, requiring no significant assistance from the state (group A), it judges that this subject has attained ‘freedom’ and needs no care or repressive intervention. For everyone else, gradated and descending attributions are made of capacity for freedom and self-government – with inversely intensifying levels of management. Those in receipt of employment benefits are afforded slightly less capacity for self-government than group A, because their productivity has stalled. This is often rectifiable through time-limited, disciplinary intervention by employment agencies who will ensure a transition back to work (these are group B, for Dean). ‘Those who are potentially capable of exercising liberal autonomy but who are yet to be trained in the habits and capacities to do so’ (Dean, 2002: 48) are allocated to social care sectors for retraining towards the liberal model of subjectivity (group C); while ‘those who, having reached maturity of age, are for one reason or another not yet or no longer able to exercise their own autonomy or act in their own best interests’ (Dean, 2002) (group D). Dean makes clear that addictions and mental and physical disabilities, as well as infirmity in old age, are categorized within group D by liberal governmentality (Dean, 2002). The police and security services are only involved in cases where criminal subjects, whose agency is delinquent, require repression through force (group E).

Liberal governmentality ostensibly limits the reach of repressive agencies, like police and intelligence and security agencies, to group E – only because it can rely on the intervention of ‘liberal police’ (social policy agencies) to regulate the conduct of all others. Of course, the reliance upon categorizations of subjects to guide this liberal government embeds racism, sexism and ableism into rule. Liberal rule operationalized, for example, the regulation of women’s behaviour through civil society’s policing of their status as second-class citizens (Valverde, 1996). The imperial conquests of foreign lands were justified by John Stuart Mill ([1859] 1975) according to the categorization/racialization of colonized people as ‘backwards’ in seminal works of liberal political theory (15–16), so long as their ruler intended the ‘improvement’ of these people towards the standard of self-government. Categorization within liberal rule is extremely violent, even if – ostensibly – the reach of police and security agencies, and their impact upon individual freedoms, is limited and social policy interventions are preferred in their place.

Does this mean that policing, as an agency, is kept separate from social policy then? Multi-agency working is an important practice within contemporary political administration which is not acknowledged in Dean’s article. Based around assumptions of complex multifactorial causes for delinquency, abuse and addiction (sometimes referred to as ‘wicked problems’ [Sydelko et al., 2021]), multi-agency working brings care, educational and policing agencies together to share information and to develop dynamic solutions. Multi-agency is not a new way of public sector working, particularly in Scandinavian countries where systems like Schools, Social Services and Police (SSP) and Samordning av lokale kriminalitetsforebyggende tiltak [Coordination of local crime prevention measures] (SLT) organize coordination between local policing, schools and social care to identify truancy, drug use and child abuse. The Danish SSP system has been in operation since 1977 when the Crime Prevention Council appointed the central SSP committee. ⁴ The SSP system has since expanded to the vast majority of local municipalities and was supplemented in 2004 by the first Police, Social Services and Psychiatry (PSP) collaboration in Frederiksberg, Copenhagen – itself rolled out nationally in 2009 (Sestoft et al., 2017). In 1989, Norway followed the Danes and established the SLT system, which aims to bring the local police, schools and healthcare agencies together for the prevention of crime (Ellefsen, 2011).

In the UK, multi-agency child protection has been formalized through the Multi-Agency Safeguarding Hub (MASH) system, several decades behind the Scandinavian countries. The first MASH was established in 2011, in response to failures in child protection, serious case reviews and national reports which recommended greater information sharing between agencies (Dunne and Finalay, 2016). The modalities of multi-agency working have integrated police officers with educational, social care and probation networks – creating an overlap between Dean’s separated sectors. Rather than overturning the distinct sectors and their responsibilities, however, multi-agency working has tended to allocate specific police officers, given a ‘preventive’ role, to the pastoral work within education, social care and probation. Multi-agency working has moved some police officers ‘across’ the line which (supposedly) separates criminal justice from care and educational work.

Another gap in Dean’s liberal governmentality framework could be understood to emerge from the developments made to policing during the War on Terror. P/CVE programmes and the popularization of the radicalization discourse have reshaped and extended the policing of racialized communities, such that there exists no sectoral boundary to protect communities from surveillance which can occur from a myriad of community-based and social policy actors (Nguyen, 2019b; Younis, 2022). If Dean’s argument were applied to P/CVE, it would likely produce a very illuminating discussion of the extension of ‘liberal policing’ (the coercion, surveillance and regulation undertaken by social policy agencies) around communities categorized as ‘risky’ (Griffin and Khalid, 2022; Heath-Kelly and Gruber, 2023; Nguyen, 2019a, 2019b). Of course, the tasking of care and education agencies with policing-like tasks is central to Dean’s analysis. Care is not to be regarded as a benign space in his work and often disproportionally impacts racialized communities with its coercive rationale.

Similarly, counter-terrorism policing has itself changed in the War on Terror. The literature on contemporary counter-terrorism policing refers to practices of stop-and-search and intelligence gathering through community policing, for example, which also disproportionally impact racialized communities (Pickering et al., 2008; Quinlan and Derfoufi, 2015). But this article does not explore changes in stop-and-search or local policing brought about by terrorism legislation. Rather, it explores the radical new covert programmes which provide intelligence agencies with welfare-administration powers. Intelligence and security agencies (like the UK’s Counter Terrorism Policing or Norway’s Politiets Sikkerhetstjeneste (PST)) are not synonymous with regular policing. They have a distinct and specialized function, relating to their exclusive focus on hostile state and non-state actors, and their use of secret methods of surveillance.

The agencies traditionally tasked with the highest levels of repression are now covertly distributing welfare, which signals a significant shift in the work of intelligence and security agencies; one which necessitates the application of Dean’s liberal governmentality thesis to capture the overriding of sectoral separation. To repeat, these are not P/CVE programmes; ⁵ OKGs, PLPs and the MAC could be considered the functional opposite of P/CVE, because rather than extending the policing and surveillance work undertaken by care and education professionals, these new programmes bring intelligence agencies into the delivery of welfare. Furthermore, they are covert – rather than requiring the consent of participants.

A note on method

Given the covert nature of new multi-agency counter-terrorism programmes, I was not aware of their existence when I began the research for this project. Their discovery was accidental during fieldwork and case study research for a larger research project, which explores P/CVE programmes in Europe through a comparative lens. The larger project compares eight P/CVE systems: England and Wales; Scotland; Norway; Finland; France; Lithuania; Croatia; and Czechia. In the original project, case study selection is justified according to the different welfare state models of Western and Northern Europe (as well as the policy transfer dynamics affecting Europe’s small states and their implementation and translation of P/CVE discourses and programmes). During the empirical research phase, interviewees in Norway began to remark upon a multi-agency model which operated differently to P/CVE – responding to and threat-assessing people known to security police, rather than being identified by P/CVE reports in the municipality. Originally, as I will describe below, this programme had been established to manage returning ‘Syria travellers’ in the community – but when none returned to a certain municipality, the programme was converted to deal with risks posed by people with psychological disorders as well as potentially radicalized people known to the security police.

As such, the method underwriting research into the Norwegian OKGs includes research interviews and documentary research. Once the details of this programme were established, attention turned to the other country case studies on the larger project. Similar multi-agency counter-terrorism programmes have not (yet) been identified in France, Finland, Lithuania, Croatia or Czechia, but continuing research will be needed to preclude the prospect of their covert presence. The usual methodological difficulties apply when confronting secrecy and trying to draw conclusions about national security work (Bosma et al., 2020). Such conclusions, made on the basis of redacted public-facing reports about national security work, will always be partial reflections of secret systems. However, with knowledge of the Norwegian OKGs, it was immediately possible to detect similar multi-agency programmes in Britain.

The method underpinning research into the British case study was documentary research into inquiry reports which followed the 2017 terrorist attacks in England, from parliamentary committees and appointed reviewers of terrorism legislation, as well as the Independent Office for Police Complaints. Freedom of Information (FOI) releases have proved crucial to documenting the existence of multi-agency counter-terrorism programmes – including releases of policy documentation from the Counter Terrorism Policing Headquarters (CTPHQ), one of which was insecurely redacted (enabling the researcher to make a transcript of the redacted passages pertaining to PLPs).

But is Counter Terrorism Policing to be referred to as an intelligence agency? In Norway, like most European countries, the intelligence agency PST is a specialist police agency. However, in the United Kingdom the domestic intelligence agency (MI5) is not located within the police service. When sufficient intelligence is obtained to enact prosecution or disruption, MI5 does not have legal authority to act upon individuals and require the practical assistance of Counter Terrorism Policing in this regard. ⁶ Counter Terrorism Policing is the agency which cooperates with MI5 to take executive actions. It is not a regular police force ⁷ nor an intelligence agency in-and-of-itself; instead, we might consider it ‘intelligence agency adjacent’. Given the idiosyncrasy of the British security services structure, and the specificity of the Counter Terrorism Policing role, I will refer to Counter Terrorism Policing within the general terms of an ‘intelligence agency’, like the Norwegian PST, throughout this article. However, the reader should understand that this is shorthand which prioritizes readability.

Finally, although the comparison of Norwegian and British counter-terrorism programmes originated organically and in an unplanned way, it is compelling that two countries – with opposing socio-economic traditions – have both afforded intelligence and security agencies welfare administration capabilities. This potentially rebalances the literature which otherwise associates neoliberal economics with preemptive risk-based interventions (Garland, 2002; Zedner, 2007, 2009), by demonstrating the existence of preemptive counter-terrorism in a social-democratic state which maintains significant welfare provision.

Norway: OKGs and multi-agency counter-terrorism

In 2014, the Norwegian Government’s Action Plan against Radicalisation and Violent Extremism tasked the security police (PST) with notifying municipalities about any foreign fighters returning to their local area (Ministry of Justice Norway, 2014). Norway operates through a decentralized system of public administration, with municipalities responsible for the delivery of welfare, integration and all social policy. Since 2014, four municipalities have interpreted the national Action Plan as necessitating OKGs at the local level – Fredrikstad, Sarpsborg, Kristiansand and Larvik ⁸ (Larvik Kommune, 2020; Lid et al., 2016). There is not a national system of OKGs.

Each OKG varies in membership and in its specific functions, but all enact the threat assessment and local multi-agency management of people thought to pose danger to others in the municipality. Lid et al. (2016: 107–108) note in their study that Fredrikstad and Sarpsborg were the first to develop OKGs and these coordination groups became important for coordinating casework, and developing and implementing measures in the management of returned foreign fighters and radicalized persons, as well as being a resource group for the municipality in terms of providing advice on lower-level cases. The work of Kristiansand’s OKG leans more towards contingencies and their management (Lid et al., 2016), whereas others function more as resource centres. The PST contributes significantly to each OKG – but in different capacities. In Larvik’s Action Plan, a PST officer is noted as being part of the formal cooperation structure (Larvik Kommune, 2020: 9). In other municipalities, interviewees noted ‘tight cooperation with the PST’ in cases involving radicalization concerns (Interview 1) or were very aware that indirect cooperation was occurring with PST, through the local police officer assigned to the OKG (Interview 2). In the latter case, it was emphasized that the OKG deals only with radicalization-related concerns and most cases are referred in through local services – but that information, and occasional referrals, would arrive indirectly from the PST through the local police representative. This had caused some consternation in the group:

‘We have discussed this sometimes ethically and legally – and asked the police to be open if this is the case. We like to know who we are talking to’. (Interview 2)

The OKGs were originally set up as an extra coordination structure to deal with returning ‘foreign fighters’. But the expected flow of returnees from Syria did not materialize, whether due to them being killed in action or deterred by the domestic criminalization of their actions abroad. In their absence, OKGs were not abandoned – rather, they shifted their purpose to broader categories of residents, such as radicalization concerns, or crisis management in general. The Kristiansand OKG’s mandate was extended in 2017 to include: ‘events of high severity; that require urgent notification of the municipality; and cannot be managed through the ordinary interaction fora between the police, municipalities and hospitals’ (Kristiansen and Ringereide, 2020). As the municipality had experienced several extremely violent events involving violence from people with psychiatric disorders, it was felt that a mental health practitioner should be incorporated within the multi-agency OKG. The group’s mandate expanded to ‘all types of dangerous persons’ but excluded situations of domestic violence or suicide (Kristiansen and Ringereide, 2020).

In their study of the Kristiansand OKG, Kristiansen and Ringereide (2020) write that protagonists frame the group as a ‘mailbox’ into which risk assessments and intelligence from the PST can be deposited. Where such a risk assessment indicates a ‘red’ level of threat, the crisis manager of the OKG will coordinate a meeting of the group to share data and interpretations, ‘triage’ the risk of violence through the application of the Historical and Clinical Risk Management-20 (HCR20) risk assessment process and decide upon a course of action (which can extend to detention within a psychiatric facility or more intensive management by the welfare agencies of the municipality). To be clear, the intelligence agency provides a risk assessment of a person to the OKG, without the person’s knowledge, which can lead to detention in hospital. The individual will not be made aware of the OKG mechanism behind their involuntary detention.

In line with Dean’s (2002) arguments that liberalism extends repressive social control measures through care and social policy agencies – because its norms preclude the overt expansion of police and intelligence surveillance – this transformation of the OKG’s subject group was understood to be socially progressive in nature. The academics invited to review the Kristiansand OKG pilot project, Aslaug Kristiansen and Hans Otto Ringereide, represent the aims and intentions of the ‘catchment group’ expansion as meaning that

any returning foreign fighters would be ‘handled’ within the framework that has been established for ordinary business . . . It also implies that a returning home fighter will be handled based on the same democratic, social and human rights that apply in a democratic society. (Kristiansen and Ringereide, 2020: 5)

Here, a national security project was originally designed to respond to the supposedly unique threat posed by returning foreign fighters. But when this threat did not materialize, the group was not disbanded and there was no recognition that normality, relative safety and unexceptionality prevail. Security discourses and actors frequently fail to admit that threats have dissipated – relying as they do on imaginaries of ever-present, even escalating, danger (Heath-Kelly, 2018). The protagonists of the particular OKG (the crime prevention coordinator, police station chief, mental health clinic director, the emergency coordinator for health and social care, the municipality coordinator and municipal superintendent) took the opportunity to ‘normalize’ the multi-agency forum as a continuing mechanism of crisis management and care – for multiple groups, not just foreign fighters. The protagonists show clear signs of desiring democratic transparency – having invited academic researchers from the University of Agder to evaluate their activities between 2017 and 2020. This is not a contradiction of the expansion of their remit; rather, it underlines Dean’s (2002) point that liberal governmentality does not shy away from authoritarian practices but rather embeds them within the language of individual rights, transparency and limited government.

Furthermore, the various OKGs are very sensitive to the data protection and privacy laws which structure different agencies in Norway. ⁹ Norwegian police are enabled by law to share some personal information with other agencies, when conducting their work – far more than health professionals are normally allowed to share (Dean, 2002). Indeed, a member of an OKG emphasized that when organizing meetings of the participants, he conducts a complicated dance around confidentiality laws. He is not informed of the subject’s name when calling a meeting but can ask the police if they will be able to share personal information with the OKG members. The ‘dance’ around personal information manifests in the following way:

I send a meeting invitation to the hospital, to the services in the municipality and to the police. The police will send ID to the right persons in these services that need-to-know identity. Because before the meeting, they have to go into [their systems] and find out more about, ‘Do we provide any services? What kind of services? Do we need to find more information about it?’ and so on, before we have the meeting. And this will, could, happen within one hour. (Interview 1)

We can see here that data protection laws have become a modality of security governance (Bellanova, 2017), rather than functioning to centralize privacy concerns. For example, the OKGs will access and discuss sensitive information about a person’s relevant diagnoses, traumatic experiences and living situation, painting a very vivid picture of their subject’s life upon which to make a reading of potential risk to the public, but will anonymize or pseudonymize details such as the name in meeting invitations. To unlock this information, the police are required to communicate the name and date of birth with the representatives of the agencies. Such multi-agency information gathering and sharing is (believed to be) legitimized by the ‘Emergency Clause’ of the Criminal Code, which states that normally illegal acts will become legal if performed to protect life, health or property – and there is no other reasonable means (Kristiansen and Ringereide, 2020).

The practice of the OKGs is being recognized (at the time of writing) as a valid direction for national policy. The innovation of local municipalities in their coordination efforts with intelligence agencies (directly or indirectly), and the ‘dance’ they must do around confidentiality laws, is being used to articulate a need for legislative change in Norway. In August 2022, the PST released a report titled ‘Extremism and mental disorders – as well as cooperation between the health services, police and PST’ which describes existing cooperation between PST and local health services. The document makes a strong call for Norwegian lawmakers to amend the laws on information sharing to enable PST to access medical information, in order to ensure the protection of society from radicalized persons and serious crime. The PST grounds its request for rights to access medical information on: the need for healthcare interventions to prevent the risk of violence (PST referrals to health, presumably to seek involuntary hospital detention); the need for co-assessment of risk by health and PST professionals in specific cases (where health assists the PST with assessment of threat posed by radicalized individuals); and for ‘ensuring appropriate notification routines in the event of a worrying development’ (where health and intelligence cooperate on covertly monitoring ‘nominals’ in the community through their local access to healthcare) (PST, 2022).

Significant sectoral blurring has developed in Norwegian public administration, across the sites of the four OKGs. The OKGs take referrals from the intelligence services of Norway and manage those people in the community health services or regional psychiatric hospital. Services are provided, openly, but the public protection dynamics underpinning that provision – and the involvement of PST – remain unspoken. Rather than considering this to be in tension with liberal governmentality, where the social policy sectors are ostensibly firewalled from the activities of intelligence and security agencies (Dean, 2002), OKG participants described their work in a focus group conducted by Kristiansen and Ringereide as

less about uncovering, combating, controlling or monitoring behaviour, but about providing help. To provide help - this notion not only helps to form a basis for the conversations in the group, but it also gives a direction to the activities that OKG will carry out. (Kristiansen and Ringereide, 2020: 27)

Multi-agency counter-terrorism demonstrates the fundamental blurring of repression and care agencies in contemporary liberal states. Dean’s (2022) description of liberal governmentality points to the functional coercion involved in care work, to instantiate self-governing on the part of the individual concerned, but also to the ostensible barrier maintained with criminal justice and security agencies. The performance of a sectoral distinction is necessary, to avoid any tension with liberal governmentality’s norms of individual liberties and governing through freedom. Yet the novelty of multi-agency counter-terrorism is that intelligence agencies have entered the space of care work.

This appears to be a development of the liberal governmentality dynamic, rather than a break from it. The intelligence agencies are not pursuing a criminal justice resolution to cases overseen by the OKGs, but rather ‘reallocating’ them from intelligence-led surveillance to local monitoring and healthcare (under supervision). The ostensible separation of security agencies from the social policy sector is simultaneously invoked (by transferring the individual from the intelligence agency roster to the OKG) but also revoked (by the very involvement of intelligence agencies in the securing of care pathways for individuals). The OKGs have brought the most senior agencies of repression into collaborative relationships with care. Tellingly, Kristiansen and Ringereide (2020: 27) reflect that ‘[a]fter 17575 words and 39 pages of transcription [of focus groups run with OKG members from health, social care and police], the word “disagree” does not appear once’. Because who could disagree with care?

British multi-agency counter-terrorism: PLPs, the MAC and vulnerability support hubs

Multi-agency counter-terrorism in England and Wales is not decentralized to local authority control, as it is in Norway, but is organized by the national Counter Terrorism Policing across three ¹⁰ semi-covert programmes: PLPs, the MAC; and Vulnerability Support Hubs (VS Hubs) which, in April 2024, were put on a permanent footing as the Counter Terrorism Clinical Commissioning Service. For increased readability, I will refer to them as VS Hubs from hereon. These programmes are (partially) described in documents produced by inquiries into the 2017 terrorist attacks in Britain, ¹¹ FOI requests, ¹² CONTEST (counter-terrorism strategy; HM Government, 2018, 2023), unpublished reviews undertaken within MI5 and Counter Terrorism Policing into the failures of 2017, and discussion of those reviews by the Intelligence and Security Committee of Parliament (2018) as well as the Independent Reviewer of Terrorism Legislation (2020).

The MAC is the most publicly acknowledged multi-agency counter-terrorism programme. In a Home Office press release accompanying the 2018 CONTEST strategy, the MAC was described by the Home Secretary:

The MAC brings together existing expertise within the Counter-Terrorism community, including Prevent and Pursue and works in partnership with experts in offender management, social care and safeguarding. Bringing these fields together helps make MAC a new capability to reduce the national security risk posed by these individuals. (Javid, 2018)

The press release presents the MAC as bridging the previously separate domains of CONTEST: Prevent (early intervention, based in the local community) and Pursue (Counter Terrorism Policing and the criminal justice pathway). This bridging is framed as generating an increased capacity to manage threats to the public through both social care and intelligence-led policing. As the statement continues, the Home Secretary explicitly endorses the integration of MI5 (the UK’s domestic intelligence agency) within the business of local social care and multi-agency management of individuals. This is an important moment for Dean’s rendering of liberal governmentality around the ostensible separation of repression agencies from the care sector, in the name of governing through freedom. Here, the UK government acknowledges that the separation of agency remits has been replaced with cross-cutting working.

In combination, PLPs, the MAC and VS Hubs constitute a multi-agency system which manages a wide variety of people, covertly. The system and its components are presented in Parliamentary reports as responses to the 2017 terrorist attacks in Westminster, Manchester, London Bridge, Finsbury Park and Parsons Green (Independent Reviewer of Terrorism Legislation, 2020; Intelligence and Security Committee of Parliament, 2018) – although internal evaluation reports for the VS Hubs, released in response to FOI requests, demonstrate that parts of the multi-agency system pre-dated the 2017 attacks.

Multi-agency counter-terrorism in Britain is oriented around the sharing of intelligence agencies’ data with ‘local authorities and other government departments’ as well as ‘enrichment of that data from the databases of multi-agency partners, including local authorities’, followed by risk assessment and the delivery of a risk management plan by a multi-agency panel (led by Counter Terrorism Policing) (Independent Reviewer of Terrorism Legislation, 2020: 26–27). Specifically, the Independent Reviewer is referring here to the MAC programme – which covertly manages ‘closed subjects of interest’, no longer under investigation by MI5, through their local services and care providers. These people have not been charged with any crime, nor are they aware their welfare packages are arranged through counter-terrorism policing structures. Through the MAC, they are ‘diverted’ through the covert administration of welfare and social care – such as apprenticeships, places at a local college or educational establishment, mental health support, etc.

In the 2023 CONTEST strategy, the MAC was designated under the Pursue strand of work (criminal justice) rather than Prevent (HM Government, 2023: 66). This is fascinating because the MAC maintains precisely the same relationship with social care providers (facilitating care packages for those covertly identified as possible threats to public safety) as PLPs. PLPs are not directly acknowledged in CONTEST, apart from careful nods towards work which will ‘better connect the counter-terrorism system with expertise in healthcare, education, social services and the criminal justice system to respond to the complexity of the terrorist threat’ (HM Government, 2023: 5).

However, in FOI releases by Counter Terrorism Policing (one of which was poorly redacted and has been unredacted for the present analysis), PLPs are associated with the Prevent stream of CONTEST. The MAC and PLPs undertake the same processes for risk-assessing individuals in the community and covertly apportioning social care, yet one is positioned within Pursue and the other with in Prevent.

Of course, ‘linkage’ does not mean that PLPs are actually part of the British Prevent programme. Indeed, PLPs contravene several foundational principles of P/CVE referral work – including obtaining the consent of the referred person, ¹³ and partnerships being led by the local authority. PLPs are intelligence and security agency-led multi-agency processes that cater to people who are ‘unsuitable’ for the standard Prevent pathway because they have, for example, refused consent for a Channel process. Refusing consent actively positions a person on the PLPs pathway, emphasizing the distinction of PLPs from P/CVE. Alternatively, to ‘qualify’ for PLP, a person can have been identified as linked to someone under Pursue investigation, or ‘the individual is a “hardened” extremist/radicaliser/extremist groomer whose activities need disrupting’, or ‘the case involves institutions, venues, or “leafleting/stickering” activities (etc.) that have Prevent relevance’ (CTPHQ, 2020: 29).

Where people referred to Prevent are identified as ‘unsuitable’ for Channel (for the reasons above), ‘the PLP process will provide management through disruptive activities, or multi-agency support, or a combination of these’ (CTPHQ, 2020: 29). Here we can locate the fundamental ambiguity of PLP, as it pertains to sectoral separation in liberal governmentality. ‘Disruption’, for Counter Terrorism Policing, is a repressive technique of pervasive legal intervention upon a subject. It focuses on non-terrorist offending and is designed to exhaust a subject (and deter them from a potential terrorist pathway) while also reducing their support in networks and communities (CTPHQ, 2020). Multi-agency support, however, is the covert, but deluxe, application of social care resources. By utilizing both pathways, sometimes in combination, the ambiguity of PLPs as a liberal governmentality modality is profoundly apparent – the once separated agencies of repression and social care have united, within one programme.

The business of PLP case management involves the frequent risk assessment of the referred person, using information provided by any partner agency – or from police intelligence gathering – to assess the risk of a mental health crisis, foreign travel or terrorist offending by the subject (CTPHQ, 2020: 33). This leads to the grading of subjects as ‘red’, ‘amber’ or ‘green’ (CTPHQ, 2020). The documentation asserts that ‘any noted change in the individual’s complex needs’ or ‘significant changes in the individual’s life more generally’ should also trigger a reassessment of risk at a PLP panel meeting (CTPHQ, 2020: 34).

The transfer of health information to MAC and PLP is facilitated by the VS Hubs: a Counter Terrorism Policing program which places a consultant psychiatrist, a consultant psychologist and mental health nurses in Counter Terrorism Police Headquarters in Manchester, Birmingham and London to obtain confidential medical information from National Health Service (NHS) trusts and provide clinical formulation services to Counter Terrorism Case Officers (Aked, 2022; Aked et al., 2021; Younis, 2022). The clinicians in the VS Hubs formulate the relationship between mental health conditions and engagement with extremism in a subject’s life, which leads to a RAG rating (Red, Amber Green) of concern. Attribution of ‘red’-level risk can lead to direct involuntary hospitalization, whereas amber leads to ‘tripwires’ being set up with a GP or mental health professional (an agreement to contact Counter Terrorism Policing should an individual stop taking their medication or disengage from treatment). The VS Hubs provide advice to the MAC, as well as police-led panels, on strategies to increase protective factors and reduce risk in a person’s life. Again, no consent is obtained from the data subject for the transfer or additional processing of this information.

Unlike Norway, Britain’s multi-agency counter-terrorism is not limited to localised intelligence agency participation in the assessment of healthcare needs and apportioning of care. Rather, the MAC, PLPs and the VS Hubs fully transfer welfare and social care administration capabilities to intelligence and security agency programmes. It is Counter Terrorism Policing who own the PLP and MAC processes, acting upon advice from VS Hubs and making decisions on a person’s management in the community and the welfare resources to be given them (to protect the public from harm).

Intriguingly, however, these programmes are ambiguously positioned between repression and care work. The MAC is positioned within the Pursue stream of the 2023 CONTEST strategy, despite operationalizing the same processes and relationships as PLPs. PLPs, however, are internally referred to by Counter Terrorism Policing as their contribution to the Prevent strategy (despite PLPs being a non-statutory programme not openly admitted by the UK government). This sectoral ambiguity reaches its zenith in PLPs’ combination (sometimes in the same case) of legal disruption tactics alongside multi-agency social care packages. PLPs practise both criminal justice-oriented repression and Counter Terrorism Policing-led covert care.

The significance of multi-agency counter-terrorism in Britain relates to the dismantling of distinctions between sectors and agencies in liberal governmentality. The categorization of subjects would once lead to their management by either a care or repression agency – an ostensible distinction made to performatively constitute, and reconstitute, the liberal adherence to limited government and individual liberties (Dean, 2002). Of course, care agencies would also ‘police’ and coerce their subjects, but through means other than the criminal justice system.

In the contemporary multi-agency counter-terrorism practices in Britain, the ostensible separation of sectors has been disavowed. The intelligence and security agencies now utilize the welfare system and social policy to manage some of their nominals, while continuing to investigate and prosecute others. The combination of these methods, and the overlap between the intelligence agencies and care sector in general, speaks to a dynamic modelling of the subject under contemporary liberal governmentality. Moving beyond the static attribution of categories to subjects (such as Dean’s identification of ‘group a’, ‘group b’ and so on), multi-agency counter-terrorism has rebuilt categorization around dynamic understandings of identity and risk – such that a subject can move fluently between the mandates of intelligence agencies, the criminal justice system and social care. In response, British multi-agency counter-terrorism has removed sectoral separation between agencies, enabling intelligence actors to pick and choose their methods from the full spectrum of governmental techniques. Liberal governmentality’s processes of subject categorization and resulting sectoral separation have been transformed.

Conclusion

The Norwegian and British examples in this article demonstrate significant transformation in the methods deployed by liberal governmentality, to further the cause of self-government and government through freedom. Where sectoral separation was the natural response to static categorizations of subjects in the liberal governmentality model (as requiring care, or temporary welfare support, or repression) (Dean, 2002), multi-agency counter-terrorism has undone the restriction of intelligence agencies to particular modalities of intervention. They can implement both repression through criminal justice and engage with multi-agency social care pathways, as relevant. This reflects the development, over decades, of the ‘vulnerable’ subject position – one who is simultaneously wounded or beset by deficit and needs, while maintaining a potential to become dangerous in the future (Aradau, 2004; Brown, 2016; Heath-Kelly, 2013; Heath-Kelly and Gruber, 2023). The dynamic potential attributed to the vulnerable subject (who is attributed the potential to move from ‘need’ to ‘dangerousness’, abruptly) has led to the creation of dynamic governmental counter-terrorism strategies (Javid, 2018), which dispense with the traditional liberal performance of ostensible separations between policing and care sectors.

The ambiguity of these multi-agency programmes is deeply troubling, for intelligence agencies are provided with access to private medical and personal details about non-criminal individuals and use them to organize health and social care interventions in their lives. Individual autonomy is clearly threatened here, by such covert pastoral intervention. However, we must note that multi-agency counter-terrorism programmes have emerged at a time when some of the most controversial repressive strategies of the counter-terrorism apparatus are in (slow) decline. The extremely invasive regime of ‘Terrorism Prevention and Investigation Measures’ (TPIMs, which replaced ‘Control Orders’), are High Court-enforced measures which can remove an individual from their area of residence and relocate them to an area where they have no ties, under curfew. TPIMs are now infrequently used – in comparison to rates of use in previous years. ¹⁴ In conversations with a senior counter-terrorism officer, it was suggested that the co-occurrence of multi-agency counter- terrorism’s rise with the declining use of TPIMs is not a coincidence, and that intelligence and security agencies seek to embody the norms of liberalism, where the least invasive repression method is used to manage each subject and ensure public safety.

Here we return to the significance of Dean’s (2002) argument relating to governing through freedom, categorization and sectoral separation, specifically that liberal governmentality operates around an ostensible commitment to limiting government, in service of freedom. Even the intelligence and security agencies, it seems, would prefer to use social care methods to protect the public where it is possible to avoid the imposition of criminal justice pathways and inevitable prison sentences of great severity. And yet the covert application of social care by intelligence agencies simultaneously extends its repressive and coercive qualities, sidestepping the individual autonomy that liberal governmentality would purport to respect. Multi-agency counter-terrorism is one civil liberties nightmare replacing another.

To conclude this consideration of multi-agency counter-terrorism, it is important to note its simultaneous commitment to the liberal governmentality project (of categorization and limitation of government interventions) alongside its reversal of liberal governmentality’s methods (the sectoral separation by which limited government of individual subjects was to be obtained). This represents the transformation of liberal governmentality around the dynamic subject categorization – where an individual is attributed the potential to move, abruptly, between categories of need and dangerousness, rather than sticking to a static rendering of their character by governmental apparatuses. Multi-agency counter-terrorism generates (and then responds to) new subject categorizations of a dynamic ‘vulnerable’ individual (Heath-Kelly and Gruber, 2023), requiring ideological and psychological rehabilitation alongside welfare and continuing threat assessment by police, simultaneously.

Two conclusions can be drawn from this. Multi-agency counter-terrorism is a transformed modality of liberal governmentality, still structured by the ideological commitment to limited government through gradations of invasive interventions upon subjects. It prefers to practise social control through modalities of care, rather than criminal justice, wherever possible. Concomitantly, multi-agency counter-terrorism simultaneously marks a substantial departure from liberal governmentality with regards to sectoral separation. The most elite purveyors of repression (intelligence and security agencies) are now permitted to distribute welfare and social care, where the attribution of dynamic, vulnerable subjectivity overrides traditional liberal commitments to ostensibly limited policing.

This transformation of liberal governmentality speaks to its capacity to enact increasingly sophisticated regimes of control, while concealing its violence through an ostensible commitment to welfare and social care (Dean, 2002). Furthermore, to find intelligence agencies acting as care providers in both Britain and Norway speaks to the potential pervasiveness of this liberal transformation around dynamic, vulnerable subject positions. British social and economic neoliberalism positions the country at a distance to Norway’s comparably social-democratic political culture – and yet the commitments to liberal governmentality across both nations have led to the abandonment of sectoral separation in multi-agency counter-terrorism.