The inevitable fallibility of policing

ABSTRACT

The title of this paper is taken from the final sentence of the book How People Judge Policing (Waddington et al. [2017]. How People Judge Policing. Oxford: Oxford University Press.) which, though it had four authors, was really the brainchild of the late Tank Waddington. The paper picks up the book’s final observation and seeks to develop it, examining the problematic core of policing, and using this as a basis for thinking more generally about issues of trust, legitimacy and reform. There now exists an increasing body of research which shows how the delivery of policing can influence perceived procedural justice, the popular legitimacy of the police, and a variety of public behaviours such as compliance with the law and co-operation with the police (Tyler, [2017] Procedural justice and policing: A rush to judgement? Annual Review of Law and Social Science, 13, 29–53.). Such work is having increased impact on debates around police conduct and legitimacy and is increasingly seen as central to police reform efforts in Anglo-American policing, and in some other jurisdictions. Though accepting the broad thrust of such research, as well as its importance, this paper suggests that there are dangers in over-reading the potential of procedural justice, not least in forgetting some crucial lessons from the history of police research. The argument here focuses on the inherent complexity of policing and the inevitability of error within it. The simple but often overlooked lesson is that controversy and dissent are the norm rather than the exception in policing, and that much dispute and disagreement rather than reflecting a failure of approach or procedure, derive from the nature of policing itself.

KEYWORDS:

• Police
• legitimacy
• reform
• procedural justice

Introduction

My concerns here are with policing and with questions of reform. The study of the police and policing has become one of criminology’s larger sub-divisions, and within that field the issue of reform has tended to be a focus of almost ever-present concern. Although reform initiatives have proliferated, they have often been followed by a sense of limited impact at best (Sparrow et al. 1990). Indeed, where America is concerned, we find ourselves at a time where, in some circles, the lack of faith in police reform has led to increased demands for something close to abolition (Vitale 2017). In such circumstances it behoves us to think carefully about what we want from policing and how best that might be achieved. My underpinning argument here is that there is a tendency in reform circles to overlook some of the more significant lessons from the history of police research and, often, to over-simplify both policing and, consequently, the reform task itself. My more particular focus is procedural justice – one of the more recent and most prominent developments in this field and one to which many current reformers pin their hopes. In this context, procedural justice, at heart, focuses on police-citizen interactions and the possibilities for their improvement. Important as this is, this paper cautions against an over-reliance on, or over-reading of, the potential of procedural justice. Police-citizen interactions not only occur within a wider social, legal and organisational context but are shaped and moulded in important ways by that context. Reform efforts that focus on the interactional level, and ignore or minimise the importance of policing’s structural context, will necessarily be of limited utility. Second, even in the narrow context of police-citizen interactions, the very nature of policing places profound limits on the extent to which reforms focused on improving quality of ‘service’ can hope to improve citizen experience.

I must stress at the outset that the broad approach taken here is based on arguments that have a long pedigree and which in many ways are well-established and well-rehearsed. Nevertheless, the tendency in some contemporary criminological debates to overlook, or at least to under-appreciate or underplay them, means they bear repeating. A fundamental aim of this article, therefore, is to act as a reminder, a prompt, in this case to embrace one of the basic lessons of extant sociological research: that policing is unavoidably complex, and error is an inescapable part of it. In what follows I will be relying almost entirely on what these days is often referred to as the ‘classic’ policing literature¹; the foundational material that gave rise to the extraordinarily large industry that is now policing studies. This is a conscious choice taken for two reasons. First and foremost, it is in this literature that, to my eye, these arguments have been most convincingly made. Second, the choice is a deliberate attempt to avoid the chronocentrism – the pronounced custom in criminology and other disciplines to cite what is recent and neglect what is older – that Paul Rock (2005) has so memorably documented. In what follows, I therefore draw heavily and unashamedly on the works of Michael Banton, Egon Bittner, Maureen Cain, Carl Klockars, Peter Manning, Jerome Skolnick, James Q. Wilson and many others, much of whose work was published between the 1960s and 1980s. This approach brings with it a fairly obvious limitation and a danger. The limitation is that by relying on ‘classic’ policing studies I am restricting myself to an Anglo-American literature and tradition, with the danger that, if I am not careful, the article will fall prey to a fairly standard form of ethnocentrism (Manning 2005).² The consequence of the approach I have taken is that the observations I make below most obviously apply to Anglo-American policing contexts, and possibly to some cognate systems. Where possible, however, I make reference to, or draw contrasts with, other countries, jurisdictions or regions.

What follows proceeds in four parts. The initial section looks briefly at the primary arguments of the literature on procedural justice and police legitimacy, outlining the core claims such work makes. Second, I turn my attention to the nature of policing, focusing, again briefly, on what the classic literature has to say about the police task and its unavoidable complexity. Third, and central to the argument here, I examine the nature of police-public interactions, arguing that it is the constraints and intricacies of such encounters that are what makes policing fundamentally fallible. While it is undoubtedly the case that police practices can be improved, and reforms that seek to increase professionalism (depending on how that is interpreted) are to be welcomed, the reality is that dissent and controversy in policing cannot be reformed into non-existence and that the recognition of this should temper discussions of what we want from policing by ensuring that we do not lose sight of what is possible.

Police legitimacy and procedural justice

The literature on policing, particularly in western liberal democracies, has long been concerned with the issue of police-public conflict, and with problems of public dissatisfaction, lack of trust and confidence as well as challenges to police legitimacy (Rumbaut and Bittner 1979). Different challenges face police services in developing democracies and transitional societies where widespread political and cultural change is often a necessary precursor to, or part of, the process of police reform (Hinton and Newburn 2009). Explanations for trends in police legitimacy in liberal democracies have focused on matters such as general attitudes, demographic differences, as well as broader matters of politics and political economy (Reiner 2010). In the last few decades, the most influential body of research has arguably been that under the banner of procedural justice. This suggests, and an increasingly large body of research demonstrates, that citizens who believe those in powerful positions to be acting fairly and to be treating people with dignity are more likely to see such authority as legitimate. Legitimacy is closely linked with compliance and, though the evidence here is less strong, it is suggested by some researchers that increased perceived legitimacy is associated with greater likelihood of compliance with instructions from such sources of authority (Tyler 2017). Growing out of the work of the American social psychologist, Tom Tyler (Tyler 1990) on why people obey the law, procedural justice research has in recent years had a sizeable impact in the world of policing. As Bottoms and Tankebe (2017, p. 75) note, ‘a powerful conclusion from the existing empirical research is that, when they make overall normative judgements about policing, citizens often prioritise procedural fairness (how they are treated) over the perceived favourability or fairness of the outcome of their encounter with the police’. This is especially important, it is argued, for while it is impossible in process terms for the police always to provide citizens with the outcomes they desire, it is ‘almost always possible to behave in ways that people experience as fair’ (Tyler 2004, p. 89). In at least one respect observations along these lines are far from new. Peter Manning, for example, in his observational studies of policing in the 1970s, noted that displays of fairness were crucial, ‘not so much for the ultimate moral value persons attach to it but because it is often the most effective means to ensure compliance’ (Manning 1977, p. 238).³ One will immediately see here, however, that while Manning observes that procedural fairness by police officers is often the best bet in terms of getting citizens to do what is asked, he does not go as far as to imply that behaving fairly is something that is straightforward to achieve. Indeed, I want to argue that behaving fairly, or more accurately being perceived to behave fairly is, unlike the suggestion from Tyler above, a far from easy outcome for police officers to achieve, and that much of the reason for this lies precisely in the nature of police work and police-public interactions.

Procedural justice research is increasingly influential in various parts of the globe (and in Anglo-American and Australian policing most particularly).⁴ This large and still rapidly expanding body of research has demonstrated, in often impressive detail, the ways in which procedural justice affects perceptions of legitimacy and compliance with authority (Jackson 2012, Hough et al. 2013). By contrast, the experience of disrespect by the police appears to reduce compliance (Mastrofski et al. 1996). Moreover, the broad findings from procedural justice research also appear to have quite wide potential. That is, they hold across a number of jurisdictions (though Tankebe et al.’s (2016) research suggests that there may be considerable cultural variation in the relationship between legitimacy and social order), and across social groups within jurisdictions – being relatively undisturbed by class, gender or ethnicity within those jurisdictions where the results have been tested. Research has demonstrated broader effects, including increased ‘trust and confidence’ in police (Jackson 2012), improvements in police-minority relations (Murphy 2013) and in enhanced co-operation in crowd control (Stott et al. 2012). A systematic review of research relating⁵ to procedural justice was able to conclude that, all in all, ‘the procedural justice/legitimacy approach to policing is essential for encouraging co-operation, compliance and satisfaction with police and is beneficial across national contexts’ (Mazerolle 2013, p. 248, emphasis added), though it is important to note that the review was almost entirely dominated by U.S. research studies, with only occasional references to the UK, and to one-off studies in Australia, Israel and Ghana.

Notwithstanding such claims, it is important to acknowledge that expectations of the police, the standards to which they are held, and the general rules and norms regulating their conduct are generally socially specific. It is for this reason that cross-national studies are vital (see Roché and Oberwittler 2018b), that the gaze of police scholars moves beyond its usual preoccupation with Anglo-American contexts, and that the dominant assumptions that underpin observations of police-public interaction are made explicit. Police studies, like criminology more generally, have hitherto been a largely northern hemispheric enterprise (Brogden 1987). To a significant degree this has been true of the procedural justice literature also (though, see Murphy 2009, Tankebe 2009). Many of the concerns of the procedural justice literature are based on terms such as ‘fairness’ and ‘respect’ that vary culturally, often significantly (Morris and Leung 2000), and it is crucial not to presume that ‘western’ standards are somehow universal (Sato 2018). Indeed, Dahal’s (2019) recent research which illustrates high levels of tolerance of police violence by victims of crime in Nepal, Belur’s (2010) documentation of widespread political and public acceptance of police extra-judicial killings in India, and Xu et al.’s (2020) examination of corruption, use of extra-legal powers, delegitimation of the victim, and selective law enforcement in policing in parts of China, are all powerful illustrations of this. In short, these examples act as reminder to us of the need to contextualise claims about both public expectations and police conduct. Policing is intrinsically complex and comparative policing doubly so.

The positive results from research in the US and the UK has meant that police departments in these jurisdictions have been drawn to procedural justice-influenced approaches as a means by which they can ‘build their legitimacy; increase voluntary deference to their authority; motivate more compliance with laws or rules; and heighten cooperation in achieving community-level goals, such as the maintenance of social order’ (Tyler 2017, p. 30). It is a not an unreasonable assumption to imagine, rather akin to developments in community policing in recent decades (Ellison 2007), that successes in Anglo-American policing will likely result in attempts at wider promulgation of procedural justice, certainly within developed liberal democracies and quite likely beyond (Reisig et al. 2012). Testing procedural justice theory in a wider set of policing contexts will undoubtedly help illuminate how situationally variable its key components are and to what extent, notwithstanding current prescriptions, it is possible to identify generalisable lessons for police reform. One of the strengths of procedural justice research is that it acknowledges the importance of perception in police-public interactions, in particular looking at the ways in which citizens understand these exchanges, something which has often been absent from research in this field (though, see Brown and Benedict 2002). As I shall come on to argue, however, it is also important ‘to keep in mind that police-public encounters are interactions between people who are actively constructing meanings about themselves, each other, their circumstances and the context. What applies to members of the public, applies equally to the police’ (Waddington 2017, p. 29). In practical terms, the question that then arises is what is procedural justice – and, crucially of course, does it mean the same thing in different places or is it, rather like community policing, apparently almost endlessly malleable? What are the factors that police reform efforts need to prioritise? In fact, certainly as far as Anglo-American policing is concerned, it now appears that a fairly prescriptive model has emerged, offering four principles that are seen as underpinning procedurally just policing, and guiding the exercise of police ‘authority in a fair and just way, through the ‘quality of treatment’ and the ‘quality of the decision-making process’’’ (Mazerolle 2013, p. 246). Tyler (2004) has identified the four factors as: participation, neutrality, being treated with dignity and respect, and trust in the motives of decision-makers.

Though acknowledging that such matters are culturally specific, there are good reasons for thinking that, in broad terms, communication and participation, neutrality, treating citizens with dignity and respect, and seeking to behave with honest motivation, are goods in their own right. Procedural justice research, therefore, is to be welcomed for its potential contribution to increasing a variety of positive practices within policing. That they might be related to increased trust and compliance is, for some at least, a bonus (but see Sarat (1993) for an important dissenting view). Increasingly, however, there is something of a tendency within the police reform world to see procedural justice as a potential solution to an ever-widening set of policing challenges; to see it, if not as a ‘silver bullet’ (Roché and Roux 2017), then at least as something that can be relied upon to help bring about fundamental change in a number of different contexts (Presidential Task Force, 2015). But the history of police reform hardly makes for optimistic reading, the twin challenges of substantial officer discretion and allegedly resistant police (sub)culture(s) being regularly cited in accounts of limited policy success (Mastrofski 2004, Loftus 2009). Too often, reformers over-simplify policing, airbrushing some of its inherent complexity, attaching goals to reform efforts that are overly complex, unrealistic, inappropriate to the circumstances, or simply not based on solid evidence. In this regard, one might point to the current trend in some jurisdictions toward the promotion of police training programmes based on procedural justice theory, despite the general lack of data to show that such activity improves perceptions of police legitimacy, reduces crime or alters the behaviour of officers in the field (Nagin and Telep 2017). To hold any genuine chance of success, reform proposals must be based on readings of policing that understand the intricacies of the world they are seeking to transform, not least in remaining cognizant of the ‘craft’ of policing (Willis and Mastrofski 2017, and see Wilson 1978). In what follows, I want to focus on two areas of complexity within policing: first, the general issue of the police role and police capacity and, lastly but centrally, the inherent complexities of police-citizen interaction.

Complexity: police role and capacity

In many liberal democracies, it is not only difficult, but essentially impossible to alight upon an agreed prescription of how policing should be undertaken, and what objectives it seeks to achieve. I will turn to objectives shortly, but first the issues of capacity and style. Police bodies themselves vary markedly, historically and culturally, in what they are established to do (officially), and the means they are given to achieve such ends (legal capacities for example). Comparative studies of policing – though these remain relatively rare – and historical analyses, all illustrate the varieties of policing forms that exist (notwithstanding the commonalities that are also identified). As de Maillard and Roché (2018, p. 389) observe, ‘Classic definitions of police, inspired by the study of American policing practices, centre around the notion of the use of legitimate force. But when faced with such examples as ‘predatory policing’ in Russia or corrupt police in Latin America, such notions take on different meanings’. Policing was originally understood very broadly as something relating to the whole field of the internal government of a territory. Thus, for Patrick Colquhoun, whose name is closely associated with the development of formal state policing capabilities in Britain, in the early nineteenth century the term police was to be taken ‘to understand all those regulations in a country which apply to the comfort, convenience, and the safety of inhabitants’ (quoted in Neocleous 2000, p. 720). It is only more recently that policing has become associated with the narrower set of matters to do with public order and social control. Even within this narrower framework, however, policing itself has proven difficult to define.⁶ Attempts at definition necessarily illustrate the extraordinary complexity of policing activities, and even where the focus is solely upon public or state policing bodies, there exists huge variation in how such work is both understood and undertaken. As William Westley (1970, p. 16), in one of the earliest studies of policing, and in this case talking only of US police departments, and that in Gary, Indiana most particularly, put it, the ‘exact function of the police … is difficult to delineate … [It] not only differs from department to department depending on the city in which they are located, but also changes with the perspective in which it is described’ (see also, Greene 2010). In short, it is possible to take very different, even quite widely contrasting, views of this institution and its function, depending on one’s viewpoint.

That said, within liberal democracies it is possible to discern general patterns in what policing is conceived to be and in the ways in which it is undertaken. David Bayley’s wide-ranging international research, much of which was conducted from the 1970s to the 1990s, offers a picture of considerable similarity in the nature of the policing task across a range of liberal democracies. Though there are variations of course, he shows patrol work to account for at least half of what the police do, with much of this activity itself dominated by public calls for service. As against their public image, ‘very little of the work patrol officers do has to do with crime’ and ‘[m]ost of the genuine crime the police are called upon to handle is minor’ (Bayley 1994, p. 17) not least because such a significant proportion of police activity is citizen-initiated (Reiss 1971, Black 1973, Greene and Klockars 1991). After patrolling, criminal investigation and traffic duties are the next biggest components, though in some countries such as England, Germany and the Netherlands there have been moves to establish other agencies to deal with elements of the traffic function (Wilson and Chappell 1975, Mäkinen and Zaidel 2003). Though there are commonalities, there is also considerable variation in which public policing is organised and delivered. That policing is malleable so far as role and capacity are concerned, means that there is nothing that is obviously ‘fixed’ about it, thus reinforcing the observation that police and public definitions or understandings of role and capacity may differ markedly, and this may lead to confusion, misunderstanding and, potentially, conflict.

If, at the level of general organisation, there is little that is fixed, and there is potential for misunderstanding, this is as nothing to the complexity of policing in practice. Here the lack of clarity becomes more pronounced, stemming in the main from the gap that exists between many of the assumptions that are made about the role of the police and its reality. Peter Manning, among many others, examined this dissonance, noting that ‘one is struck with a contradiction between the public symbolism of the role and the private or, more specifically, the actual occupational activities of police patrol’ (1977, p. 13). This has consequences internally and externally for the police: for officers on the one hand, and for their various audiences – public, political elites for example – on the other.

The reality is that there tends to be both ambiguity and vagueness in public understandings of policing. Manning again: ‘The equivocation about the police role that characterizes most communities permits and encourages the police, especially the administrators, to respond to this public ambiguity and to present and maintain multiple definitions of the police role and function. Typically, they maintain a ‘private’, or covert, and a ‘public’, or open and accessible, version of their policies and practices. Audiences are segmented in this way, and performances are differentially directed and symbolized’ (1977, p. 317). Most obviously the external contradiction in policing is seen in the gap between the police’s image of crime-fighting and law enforcement and the reality that, as Michael Banton put it ‘ … the policeman [sic]⁷ on patrol is primarily a “peace officer” rather than a “law officer” (1964, p. 127, see also, Cain 1973, Holdaway 1977). This observation, he said, was based less on how officers spend their time, though this can also be used to reinforce the point, and more upon a consideration of how officers respond to the matters that they are called upon to deal with. The most striking thing about police patrol, he suggested, was ‘the high proportion of cases in which policemen do not enforce the law. Usually, they have good reasons for not invoking it, and they act within the limits of their discretionary power as this is recognized by custom if not by statute’ (Banton 1964, p. 127).

For a variety of reasons that are well-rehearsed in the literature then, ‘under-enforcement’ of the law is a key characteristic of police practice (Bittner 1970, Wilson 1978).⁸ At base, it is practically impossible for all laws to be enforced all the time: ‘the penal codes the police are presumed to enforce contain thousands of titles. While many of these titles are obscure, unknown, or irrelevant to existing conditions, and the administration of criminal justice is concentrated around a relatively small fraction of all proscribed acts, the police select only some, even from that sample, for enforcement’. (Bittner 1990, pp. 241–42).⁹ Moreover, there is then the need for officers to avoid making matters worse; the ‘fact that a police officer is himself a member of the community whose norms he has to enforce, accounts in large measure for the difference between the everyday actions of the peace officer on patrol and the more exciting deeds of his colleagues on the screen’ (Banton 1964, p. 150). The consequence, as James Q. Wilson (1978, p. 31) put it, is that officers tend to approach ‘incidents that threaten order not in terms of enforcing the law but in terms of ‘handling the situation’. Where intervention in what appears to be criminal activity takes place, it is not that it is a crime per se but ‘because the particular crime was a member of a class of problems the treatment of which will not abide’ (Bittner 1990, p. 249, see also, Bittner 1967). In this way, and like many institutions, at the heart of everyday policing is the matter of judgement or discretion. Where policing differs from many other organisations is that it ‘has the special property … that within it discretion increases as one moves down the hierarchy’ (Wilson 1978, p. 7). The study of organisations, Wilson argued is replete with examples of the ‘bureaucracy problem’: we want front-line workers to do the right thing, to deal with problems effectively, but we also want them to do other things, such as save money and be fair. So, we want them to exercise judgement, ‘but we rarely tell him what “judgement” is or how to use it and we stand ready to criticize him after the fact if he uses it in ways we dislike’ (1978, p. 3). The police officer on patrol, as Wilson observed, ‘is neither a bureaucrat nor a professional, but a member of a craft. As with most crafts, his has no body of generalized, written knowledge nor a set of detailed prescriptions as to how to behave – it has, in short, neither theory nor rules’ (Wilson 1978, p. 283).¹⁰ Officers, as has long been described, tend to develop ‘recipes’, hard-earned knowledge of how best to handle situations (for an important elaboration see Smith 1986). However, as Banton (1964, p. 178) observed, ‘because they acquire it unconsciously and are not given to examination of their own reactions, they usually cannot explain quite how they deal with awkward cases or why they employ one approach rather than another’ (see also, Bittner 1970).

If police practice tends to be characterised by considerable discretion, and by ‘under-enforcement’, then it is also an area of decision-making in which enforcement decisions are characterised by ‘low visibility’ (Bittner 1990). They are low in visibility because it is generally difficult to keep them under review (Banton 1964, p. 129). Although it remains the case that the police frequently invoke both bureaucratic and quasi-military imagery of command and control to frame the ways in which the organisation operates, and the ways in which performance is managed, ‘the actual modes of control possible, given the discretion of police officers, are very weak’ (Manning 1977, p. 193). Or, as Egon Bittner (1990, p. 253) put it, ‘though police departments are highly bureaucratized and patrolmen are enmeshed in a scheme of strict internal regulation, they are, paradoxically, quite alone and independent in their dealings with citizens’. The core for the officer on the ground, therefore, becomes a question of identifying what it is that requires action. The answer, according to Bittner’s famous formulation (1990, p. 249) is matters that involve ‘something-that-ought-not-to-be-happening-and-about-which-someone-had-better-do-something-now! (Bittner 1990, p. 249). The identification of this something is, however, necessarily variable and unpredictable. Prescriptiveness in policing is all but impossible.

At any time, and in most liberal democracies, the police mandate is both extraordinarily broad and very varied. Indeed, it effectively has no limit. As Egon Bittner observed, the fact that ‘the definition of the police mandate escaped Ockham’s Rasor’ should come as no surprise, and anyone familiar with the practicalities of policing will know this cannot be avoided. That the duties of the police officer are ‘of a mind-boggling variety’ was an ‘obvious’ conclusion, Bittner suggested, and yet this is something which still often escapes the attention of politicians, the public, and sometimes even police officers and their representatives. As a consequence, we continually have to remind ourselves of Bittner’s (1990, p. 250) observation that ‘no human problem exists, or is imaginable, about which it could be said with finality that this certainly could not become the proper business of the police’. The police mandate is believed, or assumed, to stretch to every kind of emergency: ‘of human relations, of health, of nature – the police are among the first to be called in, and generally people expect them to do something about it’ (Westley 1970, p. 19). This absence of strict boundaries has a variety of consequences both for officers and for the public – not least in further stripping away any possibility of certainty in policing – and is an important source of police-public tension and of the unavoidable fact that policing regularly involves public dissension (Waddington 2015).

The apparently boundary-less nature of Anglo-American policing is not some simple reflection of a very broad mandate; it is, as a wide range of early authors on policing noted, the outcome of the existence of a unique competence that police officers bring to the things they are asked or required to do. Thus, as Bittner puts it, even when social workers call the cops, they are mobilising a form of intervention that is police work even when it appears to be social work. Why? Because of their authoritative position. His ‘theory’¹¹ of the police – one, to my eye, that has never been bettered, is that they ‘are empowered and required to impose or, as the case may be, coerce a provisional solution upon emergent problems without having to brook or defer to opposition of any kind, and that further, their competence to intervene extends to every kind of emergency, without any exceptions whatever. This and this alone is what the existence of the police uniquely provides, and it is on this basis that they may be required to do the work of thief-catchers and of nurses, depending on the occasion’ (1990, p. 233, emphasis added).

Complexity: police-public encounters

A further area of complexity, and the one that I want to focus the remainder of my attention on here, is to be found in the nature of police-public interaction. Much policing is necessarily improvisational, creative and open-ended (Shearing and Ericson 1991). Indeed, it is inherently unpredictable, and this means that controversy and conflict are never far away. Police-public interactions can be thought of, à la Goffman (1961), as ‘encounters’. They are, in Manning’s terms, ‘a coming together of strangers that is relatively ephemeral and unrehearsed’. It is a focused gathering of people, physically present and engaged in what is generally a single activity where those involved continue their participation and interest only as long as the encounter happens. As he further notes, and as I will illustrate below, encounters involving the police have a ‘special character’ not least because, following Bittner, ‘the police enter the encounter as authority figures with a special property of role involvement, or embracement, meaning that they see themselves fully in terms of the imagery of that role and confirm in their expressive behaviour their acceptance of it’ (Manning 1977, p. 230). The focus of my attention here is on what I have referred to as the ‘inevitable fallibility of policing’. That is to say, I take it to be the case that the practice of policing has a number of properties which make error, conflict, controversy and dissension unavoidable. Such things, of course, have the potential to be reduced or mitigated, but they cannot be entirely avoided. The matters around which there will inevitably be conflict, controversy and dissension include those values held to be central to procedural justice, including neutrality, dignity and respect, and trust in motivations. In what follows I briefly outline eight interconnected characteristics of police-public encounters that contribute to policing’s inevitable fallibility.

Limited information

The first concerns the limited information that is often available to the participants in the encounter, and to police officers in particular. Much police work comes as a result of calls from the public, hence why so much of the early literature noted that public demands were generally a much greater restraint on police autonomy than either the law or police policy (Waddington 1993). Many calls to the police are, at best, ambiguous, requiring considerable interpretation. Those taking calls have necessarily to make a series of judgements, including whether what is being said is generally credible (from a source that appears ‘trustworthy’), is valid (and not a delusion) and is something that might reasonably be thought to require the police (which is capacious but not limitless in practice). Finally, there is the question of whether sufficient information has been provided to pass it on to officers who may respond (Waddington 1993 Manning 2003,). A ‘message’, Manning (2003, p. 217) suggests, is a relative term. It must be assessed and stripped of what he refers to as ‘equivocality’ and ‘noise’. The former refers to those uncertainties about source already referred to but may also include issues of trust relating to the channel of communication – such as who has passed it on. Noise refers to other, wider unwanted or unnecessary data. Stripped down, the officer then proceeds with what is left; something that is necessarily partial and incomplete. It is this uncertainty, this absence of information, that leads to the ritualisation of much policing, with officers relying on ‘what might be called ‘recipes’ for the sequences of conduct that he will engage in’ (Manning 1977, p. 236).

Anticipating the future

Uncertainty in policing has a further dual aspect where officer-citizen encounters are concerned. As William Ker Muir (1977, p. 173) originally put it, an officer’s intellectual challenge is twofold, involving both judgement and understanding. That is to say officers have ‘not only to predict the future but … also … to explain the past’. In anticipating the future, and with limited information, the police are regularly forced to make ‘rapid judgements about the probable behaviour of a person that often, and necessarily, rest as much on appearance as on past behaviour’ (Wilson 1978, p. 38). As a consequence, the work of these ‘streetcorner politicians:

… were beset by uncertainty. He could not avoid acting on the basis of predictions, and often he did not have adequate information to make an accurate judgement at the outset. In short, he was likely to make many mistaken first judgements about a matter. Thus, a policeman had to learn how to confirm his first predictions. Confirmation involved two distinct operations. He had to doublecheck whether his observations were accurate. That was the question of reliability. Had he correctly scanned the harbingers? He also had to be sure that the harbingers denoted the concept. Was the association between cue and consequence sufficiently uniform to support the inference from part to whole? In short, were the presuppositions about the connection between indicator and what was indicated correct? That was the question of validity (Muir 1977, pp. 164–65).’¹²

Suspicion and cynicism

A third complicating feature of police-public encounters stems from the suspicion generally displayed by officers. As Jerome Skolnick (1966) famously observed, the pervasiveness of danger and unpredictability in policing makes officers alert to matters that appear suspicious. In confronting citizens, officers are ‘especially alert to two kinds of cues: those that signal danger and those that signal impropriety’ (Wilson 1978, p. 39). A police officer who is not suspicious is unlikely to make much of a success of their career. But suspicion is a further source of conflict within police-public encounters. As James Q. Wilson put it, ‘ … the tendency of the patrolman to be and act suspicious arises not simply from the danger inherent in his function but from his doubts as to the ‘legitimacy’ of the victim’ (Wilson 1978, p. 27). From long experience, and as a product of the uncertainties already described, officers will necessarily apply some judgement to claims made by any member of the public with whom they come into contact. Not only are the officers suspicious, but as the literature on police culture from Niederhoffer (1967) onward has described, they tend toward the cynical. Again, this can be of little surprise. What is unusual, perhaps even extraordinary for the citizen, quickly becomes mundane and routine to the police officer. Their cynicism, born of long experience is, at least in one sense, ‘an indispensable guide to the inevitable process of ‘reading between the lines’ … to imaginatively reconstruct what is occurring at the scene and how best to respond to it’ (Niederhoffer 1967, Waddington 1993, p. 119). James Q. Wilson again: The police are like various professionals, without themselves being a profession, in that they handle on a routine basis what to others are emergencies. When the police arrive to look for a prowler, examine a loss, or stop a fight, the victim and suspect are agitated, fearful, even impassioned. But the police have seen it all before and they have come to distrust victim accounts (to say nothing of suspect explanations) of what happened. Instead of offering sympathy and immediately taking the victim’s side, the police may seem cool, suspicious, or disinterested because they have learned that ‘victims’ often turn out not to have been victimized at all. A genuine victim, of course, is dismayed by the routine manner in which his crisis is being attended to and irritated because the police do not instantly and fully accept his version of what happened (Wilson 1978, p. 25, see also, Rock 1998 on victims of homicide).

Celerity of decision-making

In addition to working with necessarily limited information, being asked both to figure out the antecedents of any particular incident as well as its likely future consequences, the suspicious and potentially cynical patrol officer is also regularly required to take decisions quickly. Not only is it assumed that officers should make the right choices – notwithstanding the difficulties listed thus far, but often it is expected, indeed necessary, that the decisions are taken without delay. The operation of police discretion often involves the weighing up of a number of alternatives. Is the matter at hand something that the officer feels warrants their attention? If so, how is it to be understood? ‘Does a particular commotion constitute a “disturbance of the peace”? Does a person who questions a police arrest decision “interfere with police operations”? If a car radio is playing Beethoven loudly, does it constitute “excessive noise”? Does a domestic disturbance call for intervention of a social work or law enforcement kind? These decisions must often be made rapidly, there may not be a clear set of precedents, and if no arrest is involved, there may not be a meaningful review procedure’ (Kleinig 1997, p. 5).¹³ As a consequence, authority has a problematic quality in police-public encounters, for these are ‘normally brief interactions between strangers with the possibility of coercion being exercised, where there is seldom a previous experience in such encounters for citizen participants, and where the course of exchange is limited in time’ (Manning 1977, p. 231).

Competing or overriding organisational priorities

The fifth factor affecting police decision-making is the existence, and potentially complicating consequences of, organisational priorities. Although acknowledged that these are unlikely to be a primary determinant much of the time, police procedures (such as they are), organisational imperatives and policy priorities will necessarily come into play some of the time, and they will affect officers’ choices. Anything that influences what officers choose to do may also impinge on citizens’ experiences – for good or ill. Taking departmental priorities to be a complicating factor in officers’ decision-making, and assuming that at least some of the time that citizens on the receiving end of officers’ decisions will be unaware of departmental priorities, they must contain at least some potential for confusion and conflict. They are, in this sense, yet another factor leading to uncertainty in police-public encounters, heightening the potential for dissension and conflict.

Servants and masters

Yet a further complicating factor emanates from the nature of the relationship between the police officer and the citizen. This takes a dual form. As long observed, the uncomfortable reality is that officers simultaneously are objects both of fear and awe; we expect them to be at the same time servants and masters (Bayley and Mendelsohn 1969, p. 45, see also Banton 1964). At heart, as Peter Manning observed, interaction with citizens tends to involve ‘attempts at interpersonal control’ (1977, p. 231). Even where no formal powers are invoked, the quality of the relationship, as observed earlier, relies upon the existence of a special form of authority. Punch illustrates this using the example of a common challenge faced by police officers: dealing with a call to a domestic dispute without resorting to arrest or some other formal intervention. In undertaking such work, Punch (1979, p. 147) notes, the officer ‘still brings to the situation a uniform, weaponry and a battery of resources charges which can be called upon if he fails to negotiate a satisfactory outcome’ (Punch 1979, p. 147). A police officer’s powers and authority cannot be hidden; they cast a shadow over any encounter, reminding the citizen that a fundamental power-imbalance lies at the heart of all police-public encounters: ‘If a subject stops a policeman to ask him the way, he is free to withdraw when he wishes; but should his behaviour be suspicious the policeman might detain him, and he would have to offer an acceptable explanation of his conduct to win release. This element of constraint distinguishes police-public interaction from nearly all other forms of social relation’ (Banton 1964, p. 167). And yet, of course, though holding considerable power over the citizen, in many jurisdictions the officer is also a representative of an organisation whose mandate places them in a service role, with an expectation that they may be called upon to offer help, up to and including their life, when there is an emergency. The ‘role conflict’ that researchers have regularly found at the heart of policing, is perhaps most frequently seen and expressed in the ways in which officers have traditionally embraced their law enforcement role, while simultaneously disparaging and seeking to distance themselves from the ‘demanding, onerous, and sometimes self-demeaning tasks’ (Manning 1977, p. 12) that make up so much of their job. This role conflict, involving both internal and external contradictions, is a further source of misunderstanding, potential confusion and conflict in police-public encounters (Ward 1970).

Citizens as reluctant performers

The authoritative, powerful position of police officer makes citizens reluctant performers in the interaction. The citizen ‘has little tendency to embrace the citizen role, as he is likely under stress, is an irregular performer, and is perhaps slightly reluctant to participate (whether as informant, suspect, offender or bystander) (Manning 1977, p. 230). Worse still, their reluctance may lead them to behave in ways that are well-known to lead to conflict. In Goffman’s terms, such citizens, showing embarrassment, lack of clarity or, worse still, insufficient deference, ‘fail to be properly demeaned’ (1956, p. 489, see also, Sykes and Clark 1975, Black 1980). The policing literature is replete with examples in which failures by members of the public to show officers sufficient ‘respect’ leads to conflict, and escalation of tensions and perhaps to the use of police power. While it is tempting to dismiss such actions as simply an unfortunate and negative characteristic of policing practice, and one to be eliminated wherever possible, it ignores the fact that it is a product of the power-imbalance that lies at the heart of all police-public encounters, and ‘comes to the fore in the way that police officers routinely exercise their powers on the basis of suspicion’ (Waddington et al. 2017, p. 26). This power-imbalance relates to a further complexity, and that is the fact that officers undertake much of this work, with all these attendant constraints and difficulties, in what Egon Bittner (1970, p. 40) called ‘the native habitat of the problem’. This can be a source of difficulties for a number of reasons. Most obviously, perhaps, the ‘working environment of the police is not only charged with emotion and suspicion, it is often, in the eyes of the police, hostile and uncooperative’ (Wilson 1978, p. 27), and this is something, as is well documented, that varies significantly by context (Antrobus et al. 2015).

The fluidity of public attitudes

The final characteristic I want to raise in this regard, and one that reflects much of what has been said above, relates to the well-established finding that public attitudes to the police are fluid (Manning 1977). What is perhaps less well established, and which is illustrated by recent research by Waddington et al. (2015, 2017), is the idea that considerable individual divergence in perceptions and attitudes toward the police is the norm and, moreover, is an unavoidable characteristic of policing. Different individuals can and do view and interpret precisely the same police-public encounters in wholly different and sometimes quite contradictory ways. In doing so they rely on their wider cognitive resources not only to interpret what they witness but also to make inferences about what cannot be seen, whilst simultaneously imagining ‘past occurrences, current possibilities and future potentialities far removed from what they witnessed’ (Waddington et al. 2015, p. 232). That is to say people disagree about policing, often strongly, not because of disinterestedness or failures of understanding – though both of these may exist some of the time – but because radical differences of view have always, and will always, exist and are intrinsic to the nature of police-public interaction. That this is so is partly a consequence of the eight characteristics of police-citizen interaction outlined here. Understandably, widely differing views of what is thought appropriate or optimal is a source of great difficulty for officers, and is one of the reasons, as Donald Black (1980, p. 188) observed, that ‘people – especially those of low status – generally do not receive what they want from the police’. As James Q. Wilson (1978, p. 75) observed in his classic study of police work, the ‘absence of agreed-upon standards for how the police should behave makes it hard for the patrolman, in his opinion, to do his job properly; the presence of many procedural rules makes it easy to penalize him for doing it, in somebody’s opinion, improperly’ (Wilson 1978, p. 75).

Reform attempts and the inevitable fallibility of policing

Research in the broad field of procedural justice has become enormously influential both inside and outside the academy. Rightly so, for it has much to offer in helping understand aspects of police-public interaction, and how citizen evaluations of policing, and citizen willingness to respond positively to authoritative command, are related to the ways in which policing is conducted. Nevertheless, and in part to protect what it has to offer, we must guard against dangers of over-interpretation; of using procedural justice-related ideas too uncritically or assuming that what appears appropriate in one context will necessarily be in others, asking such ideas to bear a greater burden than they can manage and, more particularly, failing to treat policing with the nuance, complexity and, crucially, breadth and contextualization that it requires. In the context of thinking about policing, politics and legitimacy, there are two concluding points I wish to make, one related to the specifics of police-public interaction and the other linked to the wider realm within which such interaction takes place.

First, as I have hoped to demonstrate above, policing has a number of features – some related to its inherently complex nature, others that are a consequence of the complexity of police-public encounters, that limit the extent to which it will, and in all likelihood can, demonstrate those characteristics most closely associated with procedural justice. Participation, neutrality, being treated with dignity and respect, and trustworthy motives are all broadly desirable features, and are all sensible and legitimate objectives, but it is important to remember the limit to which it is possible to achieve them. James Q. Wilson beautifully captured an element of this half a century ago, when he said:

‘The patrolman’s substantive and distributive conception of justice influences both his decision whether to intervene in a potentially disorderly or law-violative situation and his decision how to intervene. In both cases he must necessarily act in ways that deviate from the strict procedural conception of justice that applies, in principle, in the courtroom. This is because officers are routinely called upon to “prejudge” persons by making quick decisions about what their behaviour has been in the past or is likely to be in the future. The line between prejudging them purely on the basis of police experience and prejudging them on the basis of personal opinion (showing “prejudice”) is often very thin (Wilson 1978, p. 38).’

That is to say, on the one hand, there is much that the police officer is required to do that is unlikely, on occasion, to conform to the principles of procedural justice. Moreover, and crucially, even when attempting to act in ways that correspond precisely to such a model, it is very possible officers’ actions will not be perceived or received this way. Dissension is inescapable in policing. Citizens see and experience things differently; what is ‘fair’ to one person, will not necessarily be so to another, and is very unlikely to be to everyone. This is true everywhere: within individual liberal democracies, between different liberal democracies, and arguably especially so when comparing assessments of fairness within and between democratic, transitional and non-democratic societies (see, in particular, Sato 2018).

As a consequence, producing ‘a prescriptive list of ingredients of “competent policing” is futile because what is said to constitute competent policing may vary according to the circumstances in which such comment is given’ (Fielding 1984, p. 573). Indeed, ‘the ‘quality’ of an encounter between two people is an intangible and subjective thing and, to a great extent, can only be judged by reference to particular circumstances and individual personalities: ‘what one person sees as abrasive behaviour another may find perfectly normal and acceptable in a given situation’ (Southgate 1987, p. 183). Public and political demands for greater certainty in policing – assuming that officers can be trained or somehow persuaded to act in ways that will necessarily reduce conflict and criticism or enhance legitimacy and compliance – are based on a false picture of what is involved in police work. As Thacher (2019, p. 280) observes, there is something about police practice which is ‘irreducibly improvisational [involving as it does] a set of discretionary choices that practitioners make by exercising judgment in situ, rather than objective scripts that can be specified ex ante’. It is not the intention of this paper to argue that reform attempts which focus on enhancing procedural legitimacy are unimportant – far from it – merely to indicate the necessary limits of such activity. Realistic and practical reform efforts need to acknowledge that policing is necessarily creative, that even when delivered with the purest of intentions and motives it will not necessarily be received and perceived in that way and, finally, that making mistakes is inescapable in policing. Approaches to reform which recognise this fallibility, and these limitations, are likely to lessen the disappointment that so often follows overambitious or over-sold initiatives. Indeed, such realism may even enhance the possibilities of success.

The second point concerns the importance of putting all this in a wider context – something that has not been the primary focus of this article but whose importance matches that of the narrower concerns relating to the nature of police-citizen encounters. Too narrow a focus on police-public interaction or on public attitudes toward police activity risks neglecting broader factors that shape policing generally and police legitimacy more particularly. At the meso-level this includes matters of organisation and culture, matters which frequently fall into the ‘too difficult’ category where police reform efforts are concerned. On those occasions police culture is discussed, all too often it is framed within a very reduced and particular understanding. Sklansky (2007, p. 19) refers to this as the ‘Police Subculture Schema’, one in which ‘police officers think alike; that they are paranoid, insular, and intolerant; that they intransigently oppose change; that they must be rigidly controlled from the outside, or at least from the top’. This picture, based on a mis – or very thin reading of the classic literature, side-lines the social structural influences on police occupational culture (see inter alia Bittner 1967, Cain 1973, Holdaway 1983). In fact, as Reiner (2017, p. 239) puts it, although the foundational studies of policing clearly illustrated how police attitudes, perceptions, and the nature of their (sub)cultures, were vital in understanding policing, ‘they shape practice only in interaction with a plethora of other factors, and not in a unidirectional way’. As Reiner has consistently argued (inter alia 2007, 2021) this means widening the gaze even further beyond police-citizen interaction, and organisation and culture, to macro-level matters of political economy – the ways in which political and economic institutions, and the state and the market most particularly, frame and shape the nature of policing organisations and activities. Understanding this means recognising that the source of much we would like to reform derives from its context rather than from within the organisation or its culture. Policing, as Reiner observes, has a ‘bifurcated mission’. It is ‘concerned with the protection of order, but order has two faces: general—the universally beneficial achievement and safe-guarding of social co-ordination and co-operation; and particular — the reproduction of the inequalities of power and privilege that characterizes all complex societies that have existed so far’ (2017, p. 240). Meso – and micro-level police reform cannot be the solution to macro-level structural problems.

Police legitimacy must be understood at a social or collective level as well as at an organisational or, indeed, individual or interactional one (Smith 2007). In this regard thinking comparatively or historically about police legitimacy inevitably highlights how institutional arrangements, political systems, and the nature of civil society and of citizenship are all matters of great significance, framing and influencing narrower matters of public attitudes and conduct. But this is more than a problem of explanation, for a failure to locate such analyses within their wider social and political contexts risks turning what purports to be about increasing police legitimacy into something more akin to police legitimation. As Sarat (1993, p. 666) memorably put it, this is the danger of turning ‘issues of difference into questions of procedure, to want to turn away from arguments about substantive justice, and to fetishize the spirit of the game’. In this context, and for all its particular promise, the privileging of procedural justice can be seen as a further illustration of what Sewell (2005) has referred to as the ‘thinning of the social’ and what in a different, but cognate context, Rose (1996), rather more melodramatically identified as ‘the death of the social’. Such claims refer to the substantive shifts in ways of imagining and talking about public problems and their solution that accompanied the dramatic changes in political economy, affecting liberal democracies in the latter decades of the twentieth century. This shift, as Rodgers (2011, p. 5) observed in the Age of Fracture, meant that in contemporary times we heard ‘less about society, history, and power and more about individuals, contingency, and choice … Ideas of power thinned out and receded’. Where once concerns about the social structural context of policing, and questions of power were generally always present, even when the dominant frame of reference was symbolic interactionist (see Manning 1977) or even focused on the micro-politics of police accountability (see Jefferson and Grimshaw 1984), much contemporary scholarship has become more narrowly focused. The individualising effects of economic thinking, the rise of rational choice, and the increasing dominance of consumer-oriented approaches, have had profound consequences for what we might think of as police studies, driving it ever further toward evaluation, policy-relevance, questions of effectiveness and ‘what works’.

Egon Bittner (1970) noted long ago the fact that the nature of policing – involving the power to use coercive, even deadly, force – is so generally offensive to democratic social values that it is cloaked in ‘circumlocutions’. Such notions obfuscate rather than clarify and do so precisely in order to hide uncomfortable truths about the nature of policing. Reform movements in policing frequently end up playing such a role. Thus, as Carl Klockars (1988, p. 257) observed of community policing, while ‘one cannot take issue with its extremely powerful and unquestionably good aspirations’, what is problematic is ‘the creation of immodest and romantic aspirations that cannot, in fact, be realized in anything but ersatz terms’ (1988, p. 257). The same danger surrounds ‘procedural justice’, and increasingly so as it takes on the characteristics of a reform movement within policing (Presidential Task Force, 2015, Worden and McLean 2017). That is to say, not only may the claims made for it become too grandiose and too generalised, but via its preoccupation with the detail of professional practice those utilising it may lose sight of or, worse, divert attention from the essential character of policing.

These tensions can be seen in the growing dissonance between the enhanced status of procedural justice in official reform circles on the one hand, and the growing visibility, on the other, of a radical ‘defund the police’ movement that increasingly views something akin to at least ‘partial abolition’ as the only realistic way forward (Bautista Duran and Simon 2019, p. 100). Leaving the shadow of abolitionism there in the background, let us conclude by returning to the narrower domain of police-citizen ‘encounters’. Even here I have suggested there are good reasons to be cautious about procedural justice. There can be no doubt that citizens’ subjective experiences of policing can be affected by the ways in which the police conduct themselves. And, as acknowledged earlier, there are good reasons for considering the general values associated with procedurally just conduct to be valuable, worthy of promotion and likely to lead to considerable improvements in professional practice and citizen experience. The problem is that there can be no uniformity in what is likely to be considered procedurally just within any single jurisdiction, let alone across jurisdictions. At heart, we have to accept that even the most thorough-going and intensively implemented reform efforts cannot rid policing of its inevitable fallibility.

Acknowledgements

I am grateful to Jack R. Greene, Robert Reiner, Paul Rock and the two editors of the journal for their enormously helpful comments and thoughts on an initial version of this paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 For one set of views as to what might constitute the ‘classic’ literature in the field, see for example the series of articles published in the journal Policing and Society in 2015–16 under the general heading ‘Revisiting the classics’: Brown (2016); Holdaway (2016); Newburn (2016); O’Malley (2015); O’Neill (2016); Ponsaers (2016); Reiner (2015); Stott (2016).

2 As will be clear from the names of the authors of these ‘classic’ studies, and reflecting the times they were written in, this is potentially a rather male view of the world also.

3 See also (Rubinstein, 1973, p. 233ff).

4 According to Google Scholar Tyler’s (1990) Why People Obey the Law has received over 9,300 citations and his (2003) article with Jason Sunshine (The role of procedural justice and legitimacy in shaping public support for policing) over 2,300 times.

5 As the authors note, actually little of the research they reviewed had been designed originally to test procedural justice in policing; rather the research was related to the processes considered central to procedural justice techniques.

6 See Jones and Newburn (1998, pp. 18–19) for one attempt at definition.

7 As I have noted, much of the literature to which I will refer comes from the 1960s and 1970s, an era when the use of the masculine pronoun was considered unproblematic in relation to policing (and much else) and, indeed, in many ways reflected the reality of policing. I have put ‘sic’ here to indicate that modern sensibilities would question such usage. Having noted it here I will refrain from doing so in future quotations in order to avoid unhelpful repetition.

8 Research on wider regulatory enforcement has produced very similar observations. See, for example, Keith Hawkins’ (2002) classic study, Law as a Last Resort: Prosecution decision-making in a regulatory agency, Oxford: Oxford University Press.

9 As James Q Wilson put it when making a cognate point, ‘ … in at least the larger or more socially heterogenous cities, the patrolman encounters far more problems of order maintenance than opportunities for law enforcement, except with respect to traffic laws’ (1978, p. 18).

10 Indeed, as Banton noted, ‘Experienced police officers develop very considerable skill in handling different sorts of people but because they acquire it unconsciously and are not given to examination of their own reactions, they usually cannot explain quite how they deal with awkward cases or why they employ one approach rather than another’ (1964, p. 178).

11 Some might not necessarily consider his explication of policing a ‘theory’. See, for example the discussion in Brodeur (2007).

12 See more general description of these problems in Muir (1977, pp. 168–69).

13 There is research that shows that having to make quick decisions is reported as a source of considerable psychological stress for police officers. See, for example, Violanti and Aron (1994).