I. Introduction
The central theoretical claim of social systems theory is that social systems are self-referential and self-creating and that, like biological systems, the initial characteristics of one generation of members are controlled by properties of preceding generations. According to Humberto Maturana (Zeleny, Reference Maturana and Zeleny1981), writing in relation to biological systems, autopoietic systems constitute networks of production of components that recursively, through their interactions, generate and realise the network that produces them, and constitute, in the space in which they exist, the boundaries of the network as components that participate in the realisation of the network. Maturana labelled these systems ‘autopoietic’, referring to the ‘self-(re)productive operations of organisms that use their own output as input’ (Zeleny, Reference Maturana and Zeleny1981, p. 21; see also Maturana and Varela, Reference Maturana and Varela1980). Luhmann and other legal systems theorists adopted the term, and applied it to social systems, including law (Luhmann, Reference Luhmann1990; Teubner, Reference Teubner1993). Luhmann (Reference Luhmann, Bednarz and Baecker1995, pp. 282–283) went on to propose autopoiesis as a means of accounting for the absence of a unifying set of principles to integrate law, politics, economics and other foundational aspects of society. What is central for each of these subsystems, and at the root of their autopoietic nature, is their self-referentiality. This self-referentiality means that where externally and internally generated communications are responded to, it is in a system-specific way. The description of a system as ‘autopoietic’ may, therefore, be applied to any system which is able to respond to and engage with its environment, but which remains distinct from it, so that a definitional line can be drawn between the system and its context.
Systems exist and create their own boundaries in relation to their environment (autopoiesis); systems organise, reproduce, maintain possibilities, and conditions for other possibilities, through their operation; and the possibilities that are provided by a system's operations are determined by their function. However, the system-specific nature of system behaviour and its interpretation of communications mean that no single system can declare its view as representing a fundamental truth and as binding on all other systems. This ‘essential circularity’ of systems, as it was labelled by Teubner (Reference Teubner1993), means that they face significant difficulties in successfully engaging in intersystem communication, as is considered in the following section. The lack of a common understanding of communications between systems can lead to the development of ‘double contingency’. Double contingency, as described by Luhmann (Reference Luhmann, Holmes and Larmore1982), is a consequence of the confrontation of at least two autonomous systems that make their own selections in relation to one another, but that, because of the complexity of subsystems, are unable to fully and reciprocally understand each other. Consequently, systems ‘concentrate on what they can observe as input and output in the other … They can try to influence what they observe by their own action and can learn further from the feedback’ (Luhmann, Reference Luhmann, Holmes and Larmore1982, p. 110).
The two subsystems – law-making and policing – on which this paper focuses are not standard subsystems, in terms of conventional applications of the theory, and neither subsystem can be characterised as wholly autopoietic. The law-making subsystem, here, is used as shorthand for the relationship between law and politics, and how political policy becomes positive law. As King and Thornhill suggest, drafting legislation may be seen as being ‘part of the political process of policy-making which depends upon the coding government/opposition’ (Reference King and Thornhill2003, p. 44). Once a statute becomes law it sets out what conduct is regarded as lawful conduct and what is unlawful conduct (p. 45). In the hands of the policing subsystem this reflects the subsystem's coding on actions into crime and not crime. As a result of law-making and policing functions, therefore, elements of both the law-making and policing subsystems are more commonly evaluated as constituent parts of the political system (Thornhill, Reference Thornhill2011), rather than as separate subsystems. In the analysis in this paper, however, the two subsystems operate slightly differently in society because they have specialist functions within a functionally differentiated system. The use of such alternatively configured subsystems follows a trend within social systems theory of defining systems in terms of their exhibiting of systems behaviours, as opposed to by reference to a fixed list of recognised social systems or the requirement for wholly autopoietic behaviour. Přibáň and Nelken (Reference Přibáň and Nelken2001) and Lacey (Parker, Scott, Lacey and Braithwaite, Reference Parker, Scott, Lacey and Braithwaite2004), for example, apply systems-based ideas, often viewed at a very general, macro-level, to individual industries and areas of legal regulation. Systems-based analyses have accordingly been applied to a range of functionally specific bodies, on the basis that they are able to respond to and engage with their environment, in accordance with self-developed and self-referential patterns of behaviour, whilst remaining distinct from it (Teubner, Reference Teubner1983, p. 272). It is on this basis that this paper identifies social systems behaviours which are observable in the operation of the law-making and policing subsystems and uses these to inform an analysis of the racial effect arising from the implementation and use of s.44.
This paper focuses on three areas of inter-subsystem communications between the police and law-making subsystems. The first area explored focuses on communications relating to the nature of the powers. Here, it is advanced that a communicative barrier arose between subsystem-specific interpretations of what were intended by the law-making subsystem to be flexible powers subject to strong safeguards against misuse, and the policing subsystem's interpretation of these as highly discretionary. These mismatched interpretations helped to minimise the operational effectiveness of the statutory safeguards, intended to protect against the misuse of the powers. The second area of communications considered relates to expectations for levels of use of the powers. It is argued here that a communicative barrier is identifiable between the police's understanding of law-making subsystem expectations for the level of use of the powers and the law-making subsystem's own expectation of circumspect deployment, with the consequence that the law enforcement subsystem took a blanket approach to use of the powers. The final area of communications focuses on the different subsystem understandings of the requirement for intelligence, through which law-making expectations of police use based on particularised intelligence were understood by the police as including use of the powers based on broadly drafted, race-based profiles, reflective of the national threat assessment. Based on analysis of these three areas of inter-subsystem communications, this paper argues that the relationship between law and politics borne out in the behaviour of the law-making subsystem, generated a particular expectation about the consequences and safeguards of s.44. Such expectations were roundly disappointed because neither law nor politics envisaged the approach taken by the police in using the powers.
Although legal reform has rendered s.44 an essentially moot point in terms of ongoing police operations, the identification and analysis of communicative barriers and instances of double contingency between the law-making and policing subsystems suggest a more generally relevant conclusion: that the lack of external constraints on their behaviour risks being understood by the police as instructing them to act free of any objectively assessed standards of behaviour. At the same time, the analysis regarding s.44 indicates that the police's understanding of their powers and expectations regarding their intended use do not wholly correspond with law-making subsystem expectations of police self-regulation and operation based on professional judgment. Similar concerns have also been raised in relation to Section 60 of the Criminal Justice and Public Order Act 1994 (see Dodd, Reference Dodd2010), suggesting the existence of a systemic problem in police understanding and implementation of legislative powers which systems theory can help to explain. In the case of s.44, this meant that the police powers were deployed without the statutory safeguards, the operational self-restraint or the intelligence-led judgment on which the legislature premised their enactment, accommodating their use in line with race-based profiles.
II. Statutory background
In 2010, following the judgment of the European Court of Human Rights (ECtHR) in Gillan v. United Kingdom,Footnote 1 the UK government suspended and, through Section 59 of the Protection of Freedoms Act 2012,Footnote 2 later repealed, the suspicionless stop and search powers in Section 44 of the Terrorism Act 2000 (s.44).Footnote 3 As well as finding that police use of the powers had infringed the claimants’ rights, under Article 8 of the European Convention on Human Rights (ECHR), the Court in Gillan voiced its concern at the disproportionate impact of s.44 on ethnic minorities and the risk of its discriminatory use.Footnote 4 This judicial assessment corresponded with pre-existing claims of the racially uneven deployment of the powers (see EHRC, 2009; JCHR, 2004, paragraph 47) and their deleterious impact on both community cohesion and effective counter-terror policing (see Travis, Reference Travis2010; Bowling and Phillips, Reference Bowling and Phillips2007; De Schutter and Ringelheim, Reference De Schutter and Ringelheim2008, p. 383). The broad drafting of the s.44 powers and the low evidential standard for their use were treated by the police as being nothing more than pragmatic, necessary legislative choices. However, a range of empirical evidence suggests that the powers had a racial effect. ‘Racial effect’ here is used to denote the disproportionate influence of the suspicionless stop and search powers on a visible, though possibly heterogeneous, minority community, as opposed to a group whose members belong to a single race or share a common ethnic background.
The disproportionate and excessive use of stop and search powers against a particular ‘suspect community’ (Hillyard, Reference Hillyard1993) has been a consistent, statistically established characteristic of this form of policing, albeit that the identity of the community has changed in line with contemporary political and policing priorities (Pantazis and Pemberton, Reference Pantazis and Pemberton2009). The crudeness of the created ‘suspect community’ is suggested by the targeting of Sikh individuals, mistakenly associated with the threat from international terrorism because of their turbans, after 9/11 (Gohill and Sidhu, Reference Gohill and Sidhu2007–08). Statistics collected and published by the Home Office indicate that s.44 was deployed disproportionately and excessively against people belonging to ethnic minority groups, as compared to White individuals.Footnote 5 Between 2001 and 2003 the proportion of White individuals subject to the powers fell from 72 per cent to 63 per cent, with figures published in 2004 showing that Asian and Black people were respectively four and five times more likely to be stopped and searched than White individuals.Footnote 6 In the data year covering 2008/09, of 185,086 individuals stopped and searched by the Metropolitan Police Service under s.44, 58 per cent were self-described as White; around 16 per cent as Asian and around 11 per cent Black.Footnote 7 In 2010, a report by human rights group Liberty (2010) concluded that Black and Asian individuals were between five and seven times more likely to be stopped under s.44 than their White counterparts. That this increase was at least partially attributable to broad-brush, race-based profiling was apparently accepted by Peter Herbert, a member of the Metropolitan Police who stated that ‘intelligence cannot lead to a 1,100% increase; this is just random stop and search’ (Dodd, Reference Dodd2005). When the ‘random’ searches are consistently focused on Asian and Arabic individuals in circumstances where such bias is known, it must be questioned how truly random this result is or whether it is a manifestation of racially biased policing.
In the event that police use of s.44 had resulted in the discovery of significant numbers of terrorists or the uncovering of planned terrorist attacks then there may be grounds for arguing that, despite their racially uneven use, the powers were a justifiable response to the need to safeguard public security. However, this argument is negated by the persistently low arrest and conviction rates arising from use of s.44. In the year 2004/05, only five individuals, all of whom were White, were arrested following a s.44 stop, a hit rate of 1.2 per cent.Footnote 8 Further, out of over 100,000 s.44 stops and searches conducted in 2008/09 there were no terrorism-related convictions (Whitehead, Reference Whitehead2010). Indeed, as reported by Human Rights Watch (2010), in the period between April 2007 and April 2009 there were no successful prosecutions for terrorism-related offences arising from the use of s.44, despite almost 450,000 such stops and searches. As well as failing to secure arrests or convictions, the Independent Reviewer of counter-terrorism powers, Lord Carlile,Footnote 9 expressed his doubts that anything more than ‘morsels of intelligence’, at best, had been obtained from use of the powers. These are arguments against the efficacy of the powers themselves, but also serve to counter any possible suggestion that racial disproportionality in the conduct of suspicionless stops and searches could be justified by a desire to safeguard national security.
Police persistence in targeting s.44 stops and searches disproportionately at minority individuals, irrespective of the growing criticism and lack of operational utility linked to such use, echoes wider patterns of discriminatory policing previously identified and attributed to a variety of possible causes (MPA, 2007; Laville, Reference Laville2002). One such explanation has tied disproportionality to the residual effects of the types of discriminatory behaviour identified by the Macpherson Inquiry (Reference Macpherson1999), and in other studies (see Holdaway and O'Neill, Reference Holdaway and O'neill2007; Cashmore, Reference Cashmore2001, Reference Cashmore2002). These modes of operation, it has been argued, help to sustain an institutional proclivity within the police for race-based behaviours, which in turn played a role in shaping use of s.44 (see European Commission against Racism and Intolerance, 2004, paragraph 83; Foster, Newburn and Souhami, Reference Foster, Newburn and Souhami2005, p. 30; EHRC, 2009). One element of institutional racism is undoubtedly the bias arising from consciously and overtly racist attitudes and behaviours. Aside from this, however, Macpherson recognised the role of unconscious prejudice in fuelling institutional bias. Departing from lines of argument which rely on demonstrating the existence of pernicious motives and consciously discriminatory actions, this paper focuses on unconscious discriminatory behaviours and uses key tenets of social systems theory to suggest a link between the police's use of the suspicionless stop and search powers, including the role of race-based profiles within this, and barriers to communications between the law-making and policing subsystems.
III. Nature of the powers
In Gillan, the ECtHR expressly cited the high levels of subjective officer discretion as contributing to police misuse of the s.44 powers.Footnote 10 During the enactment of the powers, however, the law-making subsystem did not purport to afford the police unchecked discretion in deciding to deploy the powers. Instead, politicians including Michael Howard, Jack Straw and David Wilshire, justified the suspicionless nature of the powers because of the need to afford police flexibility and independence in order to best protect against the threat of terrorist attack.Footnote 11 Michael Howard even sought to reinforce the message that, irrespective of the absence of the reasonable suspicion requirement, s.44 and its predecessor powers were subject to ‘strong safeguards’ against their misuse.Footnote 12 Despite this, the police's interpretation of the statutory requirements, arising from its expectations of parliamentary intention, facilitated the use of the powers without the level of oversight based on which the legislature justified the broad and highly discretionary drafting of the statutory powers. This difference, between intention and effect, points to the existence of a communicative barrier between the legislature and the police, arising from different subsystem expectations regarding the nature of the powers, which, in turn, diminished the protective value of the safeguards against their misuse. The different subsystem understandings of the nature of the powers can be seen more clearly by looking in detail at subsystem discourse relating to the specific statutory safeguards.
The statutory safeguards against the wholly discretionary use of s.44 were designed to take effect in a number of different ways. A key provision promoted by the law-making subsystem was the requirement within s.44(3) of the Terrorism Act for use of the powers to be authorised by a police officer of at least the rank of Assistant Chief Constable. In granting this authorisation, the officer was required to confirm that use of the powers was ‘expedient’ for the purposes of safeguarding against terrorist attacks. The Terrorism Act 2000 also required that the authorisation itself was subject to review by the Secretary of State, whereupon it could be confirmed, cancelled or varied.Footnote 13 A Home Office Circular (2008), designed to provide clarity on the proper execution of the authorisation procedure, indicated that ‘in view of their importance, authorisations are subject to considerable scrutiny before being confirmed by the Secretary of State’. Further, safeguards were present in Sections 44(1) and 44(2) of the Terrorism Act, under which each authorisation expired after a maximum of twenty-eight days, upon which a new one was required for any subsequent use of the stop and search power.Footnote 14 A final safeguard against misuse of the suspicionless powers was the oversight arising from their annual review,Footnote 15 intended to ensure that there was somebody ‘to keep a vigilant eye upon the exercise of these exceptional powers’ (Rowe, Reference Rowe1997, paragraph 11). By virtue of these safeguards, Michael Howard, speaking for the government, maintained that policing pursuant to s.44 would be flexible, whilst subject to significant protections against misuse.Footnote 16
In contrast to the law-making subsystem's intentions, however, the police treated the suspicionless basis for the stop and search powers as representing ‘a significant change in the relationship between the public and the police’ (NPIA, 2008, paragraph 2.3.1). Pantazis and Pemberton (Reference Pantazis and Pemberton2009) have observed that the police thus interpreted parliamentary comments as an instruction to use the powers on the basis of unfettered officer discretion. Exemplifying the highly subjective approach to stop and search, a police report scrutinising the Metropolitan Police Service Standard Operating Procedures relating to s.44 stop and search confirmed that the basis for police use of the powers ‘may not be much more than an intuition on behalf of the officer’ (2004, p. 15). A further indication of the police's understanding of the nature of the powers as highly discretionary is evidence given to the Home Affairs Committee in which officers were described as often appearing to be unclear about why and when it is appropriate to deploy the powers (2005, paragraph 142). Police behaviour, therefore, seems to indicate that removal of the external safeguards against the subjective, discretionary use of stop and search, in the form of the statutory requirement for individual suspicion, was understood by the police as the law-making subsystem's expectation that use of the powers should not be curtailed by any objectively determined standards.
The link made by the ECtHR in Gillan, between the discretionary use of the powers and their misuse, was that police expectations that the powers were highly discretionary limited the protective value of the remaining statutory safeguards.Footnote 17 The Court concluded that the twenty-eight-day authorisation period was rendered effectively meaningless by successive renewals, without any apparent intervention of the Secretary of State.Footnote 18 Further, the impact of the requirement that use of powers be limited to a particular geographical locality was minimised by authorisations routinely being sought and obtained across entire police authorities, enabling the blanket use of powers (see findings of Human Rights Watch, 2010, p. 58). The result identified by the Court was that for some police forces s.44 effectively became a permanent feature of their law-enforcement capabilities.Footnote 19 An example of such use was the continual authorisation of s.44 across the entire Metropolitan Police area from February 2001 until May 2009 (see Home Affairs Committee, 2005, paragraph 44; Cavell, Reference Cavell2009). The police also interpreted the ability of officers to use the powers ‘whether or not the constable ha[d] grounds for suspecting the presence of articles of that kind’,Footnote 20 as allowing an officer to stop and search anyone, provided there was a valid authorisation in place. This mode of use was identified by a Metropolitan Police Authority report, scrutinising stop and search practice, which stated that whilst ‘the law has defined when a police officer may act, … [i]n practice the police can stop and search almost anyone in almost any circumstances’ (2007, paragraph 237). The high levels of subjective discretion with which the powers were used are also suggested by police disregard for the confinement of s.44 solely to counter-terror policing, including their use against photographers and protesters (Evans and Lewis, Reference Evans and Lewis2010; Lewis, Reference Lewis2010).
The statutory safeguards against misuse of the stop and search powers, therefore, had a very different effect in practice than was envisaged by the law-making subsystem, because each subsystem based its approach to the safeguards on its own system-specific understanding of the nature of the s.44 powers and, in particular, the implications of the absence of the requirement for reasonable suspicion. The contribution of inter-subsystem communicative barriers to the misuse of s.44 is also evident through different subsystem understandings of the expectations regarding the frequency with which the powers would be used.
IV. Use of the powers
In extending the suspicionless stop and search powers in 1996, politicians within the law-making subsystem emphasised the ‘circumspection and sensitivity’ with which the police would deploy them.Footnote 21 Setting out the government's plans, Michael Howard offered the police ‘praise and credit’ for the limited and strategic use of the pre-existing powers, and expressed his confidence ‘that they will exercise the additional powers, if they are granted by Parliament, in the same way’.Footnote 22 Michael Howard provided evidence that in the five metropolitan boroughs in which the powers had been used there had been only 1,746 vehicles stopped, 1,695 searched and 2,373 occupants searched, and in the Heathrow perimeter there had been 8,142 vehicles stopped, 6,854 searched and 40 occupants searched.Footnote 23 The government cited such statistics to support its view that the low levels of police use of the pre-existing powers justified its expectations of the infrequent use of the extended powers. In a statement made by Michael Howard on behalf of the government, the suspicionless stop and search provisions were described as ‘exceptional powers … that are needed [only] exceptionally’.Footnote 24 The highly restrained manner in which the police would deploy the powers was also stated as being a key reason behind the Opposition's support, including that voiced by Jack Straw,Footnote 25 of the statutory provisions, despite concern regarding the accelerated enactment process adopted.
Despite the law-making subsystem's expectation of police self-restraint, statistics revealed by Home Office Statistical Bulletins (2009a, 2009b) regarding police use of the powers following 9/11 suggest that this was misplaced. Whilst the increased use of the powers must, of course, be evaluated in the light of the impact of 9/11 on policing priorities, law-making subsystem communications nevertheless adhered to the executive's expectation that the powers should only be authorised, according to a 2004 Home Office Circular, ‘where they are absolutely necessary’, and used with circumspection, caution and proportionality. With s.44, as with the previously existing powers, therefore, there remained a strong consensus within the law-making subsystem, such as expressed by Fiona Mactaggart, that use of the powers ‘must never become the normal run of things’.Footnote 26 In contrast to such expectations, comments such as those of the then Commander of the Metropolitan Police, Rod Jarman, suggest that increasing use of the powers pursuant to numerical targets was an explicit part of police counter-terrorism strategy in some police authorities (Laville, Reference Laville2002). Evidence collected by Choudhury and Fenwick also indicates that in determining how the powers were used there was a subsystem tendency towards ‘going for big numbers’ (2011, p. 32). In line with such use, some police authorities reported the high levels of use of s.44 with apparent pride, using this as an implicit indication of operational success. For example, in 2007, the British Transport Police reported that in the year 2006–07 the force had carried out more than 30,000 stops and searches under s.44, ‘almost as many as every other police force combined’ (BTP, 2007, p. 9). A systems-based approach to this mismatch between expectations and reality of levels of use of suspicionless stop and search suggests two possible explanations: one relating to operational constraints within the statutory powers; and the other relating to the impact of law-making subsystem communications in shaping operationally closed police behaviour.
Looking first at the statutory provisions, there are several features of s.44 which can be distinguished from the powers it replaced. Such changes included those enacted by the Prevention of Terrorism (Temporary Measures) Act 1989, which resulted in the expansion of stop and search from being applicable only within Northern Ireland to applying throughout the United Kingdom,Footnote 27 and from initially only relating to vehicle stops to subsequently also applying to pedestrians.Footnote 28 Alongside the growing range of contexts within which the suspicionless powers could be used, the link between such use and specific terrorist actions was diminished. When suspicionless stop and search was first systematically used by the police it was in response to the bombing of South Quay, in London.Footnote 29 Subsequently, the pre-emptive use of the powers was increasingly advocated within the law-making subsystem, first at a very late stage, such that Lord Lloyd of Berwick (Lloyd, 1996, paragraph 10.21) stated that the terrorists may ‘be intercepted on their way to their target’, but increasingly far in advance of specific terrorist actions, as identified by Richard Shepard MP.Footnote 30 Despite this, Michael Howard, speaking as part of the law-making subsystem, described the extensions and re-enactment of the powers simply as ‘clear practical proposals’, necessary to fill ‘a lacuna in the existing powers’, but as having no substantive impact on their nature.Footnote 31 Parliament's expectation of the circumspect deployment of the powers, therefore, failed to recognise the external constraints within which the police had previously been acting, or that these had been removed or weakened in successive re-enactments of the powers.
The link between the use of the powers and a terrorist attack, or planned attack, was also weakened under the Terrorism Act 2000 because of its expanded definition of terrorism to include individuals ‘concerned in’ terrorism – a wider construction than solely meaning individuals actively involved in, or attempting to commit, acts of terrorism.Footnote 32 This meant that the stop and search powers were operating at a higher level of abstraction than pre-existing powers, accommodating the possibility of an increased level of use (Stone, Reference Stone2006, p. 391). A further change in the external constraints affecting police use of the powers, which may have contributed to the misplaced law-making subsystem expectations of limited deployment, was the need for the earlier powers to be annually renewed.Footnote 33 By contrast, under s.44 the powers were permanently enacted so that, although subject to the oversight of the government reviewer,Footnote 34 they could not be amended or cancelled without specific action by Parliament. The largely unacknowledged effect of the changes to the legislative provisions, in increasing the range of circumstances in which they could be deployed, meant that the law-making subsystem was endorsing the police's ability to constrain and regulate its own behaviour on the basis of its behaviour in circumstances in which there were in fact external factors shaping the subsystem's programme (Teubner, Reference Teubner1993, p. 13). The increased use of the powers once these external constraints were removed suggests that parliamentary reliance on police self-regulation overlooked the police's own expectation of external constraints and its interpretation of the removal of such restrictions as an instruction from the law-making subsystem that it should make far greater use of the powers.
Compounding the impact of the removal of the statutory constraints on police action were parliamentary discourses and other communications regarding the nature of the threat from terrorism and the role of the police in safeguarding against this (Pantazis and Pemberton, Reference Pantazis and Pemberton2011). These helped to shape the context in which the police used the powers. Whilst insisting that the powers should remain confined to exceptional circumstances, the legislature described the level of the national security threat as making the context wholly exceptional. For example, Tom King MP stated that ‘[t]errorism is now a global activity which poses many fresh and serious challenges’.Footnote 35 Descriptions of the exceptional nature of the circumstances faced became even more acute after 9/11, which Tony Blair described as prompting a call to ‘rethink dramatically the scale and nature of the action that the world takes to combat terrorism’.Footnote 36 However, once the exceptional threat became the new state of normality in the law-making subsystem communications, this was interpreted within the policing subsystem's framework of understanding as justifying, even necessitating (Ardau and van Munster, Reference Ardau and Van Munster2009, p. 689; see also van Munster, Reference Van Munster2004, pp. 146–148) that the exceptional counter-terrorism powers were ‘routinely used’ in everyday policing (MPA, 2009). The police echoed law-making communications and described the risk of a terrorist attack as being a ‘daily threat’ (BTP, 2009, p. 23) and the ‘new normality in policing’ (BTP, 2006, p. 5). Responding to what it interpreted as the nature of law-making subsystem descriptions of the terrorist threat, a number of police authorities described themselves as being left with effectively no option but to make the use of counter-terrorism powers a central part of everyday police activities (BTP, 2004, p. 3; see also BTP, 2008, p. 23). Daily use of the powers may have also helped to fulfil the police's ‘need to be seen to be doing something to reassure the public, with little regard for the long-term consequences of what they do’ (Dodd, Reference Dodd2005). Consequently, in the parts of the policing subsystem which were most closely concerned with the terrorist threat, the police interpreted law-making subsystem expectations that they respond to the daily threat of attack as necessitating their use of the powers as an important feature within day-to-day policing.
Despite growing evidence of the highly discretionary use of s.44 and the blanket approach to its deployment, law-making subsystem confidence in the efficacy of police use of s.44 was sustained by a cross-party parliamentary conviction that police operations were based upon detailed and expertly evaluated intelligence.Footnote 37 Such expectations, however, overlooked the fact that the powers were often used by ordinary police officers, as opposed to specialist counter-terrorism operatives (Human Rights Watch, 2010, p. 25). Further, these officers were typically amongst the most junior ranks within the force, and not given adequate training or briefings regarding the counter-terrorism powers (see Choudhury and Fenwick, Reference Fenwick and Choudhury2011,p. 36; Bland, Miller and Quinton, Reference Bland, Miller and Quinton2000, paragraph 9; Dodd, Reference Dodd2002). The reality of the police's particular expertise, and its role in constraining police action, is further brought into question by research such as that undertaken by Ericson and Doyle (Reference Ericson and Doyle2004). In their study of risk modelling in the insurance industry, Ericson and Doyle found that, despite the heavy reliance on the ‘expertise’ of former counter-terrorism officers, these individuals, by their own admission, saw the process of developing intelligence as little more than converting guesses into threats. The policing commitment to deploy the powers on the basis of their ‘professional judgment’, therefore, did not necessarily invoke the narrow and discerning deployment of the powers which the law-making subsystem expected when it enacted the powers. Thus the notion of there being any effective political check upon police use of the powers was largely illusory (Phillipson, Reference Phillipson2007, p. 45). The notion of double contingency gives a good account of the apparent misinterpretation and misapplication of stop and search rules created between the legislature and the police. The legislature ordained measures for the prevention of acts of terrorism which the police interpreted and applied in a racially biased manner. The law-making subsystem created the law with the expectation that its intention would be clear to the enforcer (the police); however, the police were able to make selections from a number of possibilities, and so did not need to ever select the ‘intended’ one. Police action then became contingent on that choice. When problems in application of the law were perceived, the law-maker made new selections in response to the communications of the enforcer. Thus there was two-way communication, even though disparity between intention and outcome remained unresolved.
As well as different system-specific understandings of how the powers would be used, a further inter-subsystem communicative barrier appears to have given rise to different expectations for what use would constitute an expert, intelligence-led use (Moeckli, Reference Moeckli2005; see also Moeckli, Reference Moeckli2007). In particular, both the law-making and policing subsystems appear to have had different understandings of ‘intelligence-led’ policing, as the next section explores.
V. Intelligence-led policing
The law-making subsystem's expectation that s.44 would be based on intelligence reflects the normal police operational mantra that it is ‘intelligence-led policing [which] underpins all aspects of policing’ (ACPO, 2007). When the powers were initially debated this was implicit, by virtue of their largely responsive deployment. When the powers were extended in 1996, parliamentary comments recorded in Hansard, such as those by David Trimble, endorsed use of the powers only ‘when there is intelligence that such an outrage [a terrorist attack] may be committed in a particular area’.Footnote 38 Trimble's expectations were shared by other MPs, including Jack Straw.Footnote 39 Following 9/11, intelligence-based use of the powers continued to be described as vital in ensuring that the powers were usefully and effectively deployed. David Blunkett, for example, stated that ‘[o]btaining good intelligence and being able to target potential terrorists is essential’,Footnote 40 and after the 7/7 London bombings, Charles Clarke reaffirmed that ‘intelligence is our key weapon’ in fighting terrorism.Footnote 41 Communications from within the policing subsystem also expressly advocated an intelligence-led use of s.44, so as to avoid their ‘arbitrary’ deployment (MPS, 2008; NPIA, 2008). However, despite the consistent invocation of the importance of intelligence in using the powers, the meaning it was given, as interpreted by the police subsystem, was very different from the meaning the law-making subsystem assumed intelligence would have (Teubner, Reference Teubner1993,p. 72). The effect of these different subsystem interpretations of ‘intelligence’ was that while the law-making subsystem consistently linked intelligence with specific evidence of terrorist activity, the policing subsystem treated intelligence as potentially existing irrespective of such evidence, thus affording it a far wider construction than that expected by the law-making subsystem. Such pre-emptive use encouraged a purportedly intelligence-led approach, but one that was based on wide net-casting (Walker, Reference Walker2003, p. 387).
Police interpretation of the legislature's communications as not necessitating the existence of particularised intelligence for use of s.44 is demonstrated by National Police Improvement Agency guidance, which stated that ‘[i]f police are in possession of specific intelligence about possible terrorists then searches under s.43 may be more appropriate than under s.44’ (NPIA, 2008, paragraph 2.3.2). This advice does not simply suggest that the policing subsystem understood particularised intelligence as being unnecessary for use of s.44, but that if any such information was present s.44 was not the appropriate policing power to use. Police use of s.44, despite an absence of particularised intelligence, is also indicated by the oral evidence, given to the House of Commons Home Affairs Committee (2005) by Assistant Chief Constable Rob Beckley. Beckley stated that s.44 could be based on either broad or specific intelligence in an area, confirming that the particularised nature of the intelligence was not exclusively adhered to in police use of the powers. Beckley also stated that police used the powers ‘in a pretty random way’ (2005, paragraph 54), further suggesting the absence of a link between their deployment and the existence of specific intelligence (Fenwick and Choudhury, 2011, p. 33).
In line with the policing subsystem understanding that no specific intelligence regarding a terrorist threat was necessary prior to the use of s.44, a British Transport Police Annual Report (BTP, 2007, p. 8) described the primary basis for deployment of the s.44 powers as being the existence of the authorisation, despite this being mismatched with the law-making subsystem's expectation of specific intelligence at both the authorisation and deployment stages. In addition, research by Fenwick and Choudhury cites a police source confirming that use of stop and search was ‘not really intelligence led other than there is intelligence in the system’ (2011, p. 32). Instead of a shared understanding of the intelligence requirements between the law-making and policing subsystems, therefore, it was interpreted in accordance with each subsystem's own understandings of the type of intelligence expected by the other (Luhmann, Reference Luhmann1985, pp. 73–210). The law-making subsystem's expectations for temporally and geographically specific intelligence, prior to police use of s.44, acted merely as an irritant to the police subsystem, and failed to align different subsystem approaches to the requirement. Police understanding of who the ‘usual suspects’ are, through stereotyping, stands as a testament to the self-referential and self-creating autopoietic practices by which the practices of one generation of members is shaped by the practices of preceding generations.
VI. Conclusion
This paper has suggested a link between the ECtHR's criticisms of police misuse of s.44 stop and search powers, in particular its race-based deployment, and the existence of communicative barriers between the law-making and policing subsystems. The analysis provided has suggested that the internal rules of both systems, including each subsystem's communicative redundancies, caused each subsystem to communicate its intentions in one way, while the internal rules and communicative redundancies of the other system caused it to give the communications a completely different significance (Luhmann, Reference Luhmann1986, p. 113). In the case of s.44 these differences in subsystem understanding meant that for the law-makers the absence of formal, mandatory constraints did not mean the absence of constraints in police use of s.44, whereas for the police this absence not only meant that there were no operational constraints, but that they should use the powers more liberally to fulfil law-making subsystem expectations. Such use accommodated police deployment of the powers on the basis of broadly drafted, race-dependent profiles. These profiles rendered the racially disproportionate deployment of the stop and search powers an apparently common-sense approach to counter-terrorism policing, despite its lack of operational utility and its effect in alienating minority communities from police counter-terrorism efforts. The behaviours of these subsystems suggest the existence of communicative barriers between the law-makers and the police. These barriers meant that each subsystem's understanding of, and response to, the terrorist threat was conceptualised in its own self-referential way. Such self-referential behaviours also determined how the two subsystems sought to make sense of what they understood the other to be saying, but the gap between communications as made and as understood demonstrates that, although the law-making and policing subsystems are coupled in some way, they remained very distant and in a state of incommunicability as regards addressing the threat from terrorism.
It took until 2010 for these deleterious effects to be judicially acknowledged, and longer still for s.44 to be repealed. In repealing the powers the UK government recognised the racial effect of their use. Attributing this very much to the context in which the powers were deployed, statutory reform did not occasion any meaningful consideration of law-making and police subsystem communications. As such, it seems almost inevitable that the types of barriers to communications that arose in relation to s.44 will reoccur, and that, irrespective of parliamentary intentions, the use and impact of future police powers will continue to be shaped by autopoietic systems behaviours.
