I. INTRODUCTION
The central goal of this paper is to show how police act in ways that are not obviously illegitimate, but fall under the moral standards applicable to punishment. This may strike some as surprising. The explicit job of the police does not include punishment. No official document will mandate punitive police action. Punishment is meant to occur downstream in the criminal justice system. Indeed, the wrongness of some of the paradigm cases of police malpractice is often described with the implication that such action is illegitimate because it is punitive.Footnote 1 For instance, one author notes that in arresting people who have recently been violent, it is not uncommon for police deliberately to cause some physical suffering, by, for example, overtightening handcuffs. He continues: “. . .however deserved such measures may sometimes appear to be, they are improperly imposed and in that respect are unjustified. . . It is not for the police to inflict punishment.”Footnote 2 Faced with this kind of case, we are led to the intuitive conclusion that punitive police action is, of necessity, an injustice. Nevertheless, I will argue, standard police tactics do in fact fall under any normal definition of punishment (or in any case, under the normatively relevant elements of any normal definition of punishment).Footnote 3
In short, some police and state action is not punishment, but is relevantly like it, with respect to the applicable normative standards. It may also be the case that some state action is punishment even though it is not labeled as such, and that some police action is punishment. That we sometimes express our belief that some noncourt act is excessive and unjustified with the claim that it is “punitive” should not distract us from the point that some noncourt acts must, in order to be justified, meet the same standards as those applicable to acts of punishment. What does this mean for our attitudes toward the criminal justice system? It presents a dilemma: either police routinely act in an illegitimately extrajudicial fashion, or the stringent standards to which we hold criminal court procedures are, if they are grounded, not grounded directly in the fact that the courts administer punishment. I will argue that neither option is attractive.
In Section II I focus on some specific police strategies that seem to fall under the definition of punishment. It will be clear that the strategies I discuss are not the only possible instances of police action that fall under the definition of punishment. The claim I make naturally expands to other strategies. It will also be clear that not every instance of the strategies I describe falls under the definition of punishment. Nonetheless, this is not the place to determine precisely which police actions do, and which do not, count as punitive. The approach here is ecumenical. I hold that however one understands punishment, from a broad range of plausible possibilities, some everyday police action can constitute it.
Having set up the prima facie case for my view, I consider a series of objections to it: a series of ways of holding that police merely enable criminal justice and do not enact it. Thus, some will object that police do not (properly) intend to punish and that their purposes lie elsewhere. I discuss that view in Section III.A, arguing that the intention behind an institution is not fully determinative of its normative status. Some claim that legitimate police actions do not carry the forms of hard treatment or deprivation that are characteristic of punishment. In reply, in Section III.B, I discuss how the deprivations of normally understood punishment compare with the deprivations of police actions. Section III.C addresses the idea that punishment is distinct because it involves an intention to impose a deprivation for its own sake. Noting the retributivist idea behind this objection, I argue that a built-up theory of punishment with such a backing will impute a relevantly similar kind of activity to police and the rest of the criminal justice system. Section III.D explores the relevance of censure, responding to the objection that normal police strategies lack the reprobation that is essential to punishment. Finally, in Section IV I argue that this presents a dilemma, threatening attractive views about either policing or punishment.
II. TWO STRATEGIES
A number of police strategies cause a deprivation or disadvantage to an individual in response to a perceived infraction. This much, at least, they share with punishment. In addition, many such strategies lack the goal of bringing the individual into the remit of a higher institution that will judge the case and sanction an authoritative response to the infraction. For the purposes of this paper I will focus on two examples.
The first is the recruitment and management of informants by threat of prosecution. In recruiting informants, it is a well-established practice for police to offer to refrain from bringing a prosecution for a known or suspected crime in exchange for cooperation with a further investigation. In one recent study of informants in the United Kingdom, “84 per cent of police informers had either been in custody or facing charges when they were recruited and in 85 per cent of the examples it was the handler who launched the recruitment.”Footnote 4 In the words of another study:
In practice, “informal immunity” is granted if the informer keeps his or her part of whatever deal is struck, with participating informers only rarely prosecuted, and then not usually successfully.Footnote 5
By its nature this practice is difficult to measure and make public, and one might reasonably speculate that institutions would instinctively resist transparency about it. The offer of immunity may be formal, and on a par with a plea bargain. But it is also the case that police will make use of people's legal vulnerability in order to recruit them for investigative purposes. Thus, the study continues:
Where doubts arise as to the role of an informer, the Crown Prosecution Service should normally be involved in the decision-making process. The extent to which this happens in practice is unknown, although some officers do not even disclose the existence of informers to their supervisors, let alone the courts.Footnote 6
This practice is especially well documented in the United States, where guidelines for police making arrests of drug groups urge that arrests be coordinated to make it easier for potential informants to cooperate while remaining undercover. Police and prosecutors may achieve this by lowering charges against individuals who are not expected to act as informants, so as to make the informant indistinguishable from the others from the perspective of the criminal organization.Footnote 7 It should be emphasized the extent to which the process is typically police-driven: it occurs “immediately after, or sometimes in lieu of” arrest; it is run on agreements that are “often informal and rarely memorialized in writing”; and before charges are filed, “police can truthfully promise not to disclose [a possible informant's] most recent crime to the prosecutor.”Footnote 8
The second type of police action is public-order maintenance by threat of prosecution. This involves the appropriate use of discretion with regard to minor transgressions. John Kleinig argues:
The complexities of social life . . . call for situational judgements, and mere warnings or a legalism that feeds every formal infraction (misdemeanour) into the hopper of the criminal justice system are likely to be inadequate or inappropriately heavy-handed. This is especially the case with loitering with intent, disorderly conduct, vagrancy, public drunkenness, and the like. There is also a range of ill-defined minor offences, especially involving young people, to which police may (but perhaps ought not to) appeal for purposes of making arrests. The police officer may confiscate contraband, seek to ban a person from a particular venue for a time, or impose some other social restriction or requirement when arrest and processing might not be called for. At certain times and places, that role for the beat officer was recognised and valued.Footnote 9
The appropriate use of discretion is such that—ideally—a formal charge or warning is not brought against every infraction. Rather, by loitering or being drunk and disorderly in public, you make yourself liable to a formal procedure in criminal justice, but whether or not you are charged should depend on a police officer's judgment of the value of doing so, given the context. Since both you and the officer have mutual knowledge of the possibility of a charge, and of what it depends on, the balance of power between you changes. The officer may decide merely to request that you move elsewhere. Knowing that you may otherwise be charged, you comply with the request, even though you would prefer not to. And the criminal law governing police is sensitive to those facts. As with the case of informant recruitment, possible but undischarged uses of the law are alluded to by those holding power, with the goal of achieving a better outcome.
One of the complexities of theorizing about the proper role of police is the matter of police misconduct. There is extra room for abuse in areas that permit discretion and judgment, and these are the areas that form the focus of this paper. The world of informants, in particular, gives rise to horrifying stories.Footnote 10 One response to the above kinds of cases will be to argue that police power to act with discretion should be severely limited or rescinded entirely. I discuss that view in Section IV. For the moment, we should distinguish between objections to the powers of an arm of the state that are based in the need for safeguards against possible or likely abuse, and objections that are consistent with proper use of the stated powers. The arguments I discuss fall into the latter category; this does not intrude on any argument concerning the “dangers to virtue” in the application of punishment.Footnote 11
III. HOW POLICE PUNISH
The two strategies I described—recruitment and management of informants by threat of prosecution, and public-order maintenance by threat of prosecution—involve causing a given individual a disadvantage or deprivation. They do so intentionally. They carry censure. They are enacted by state representatives. They are carried out in response to perceived wrongdoing, specifically, wrongdoing that involves a violation of established and announced laws. They have all of the core features of punishment. What distinguishes them from punishment? I believe that there is not a useful distinction. How might one distinguish police action? I will consider and reject four ways to do so. These appeal in turn to the intention to punish, the nature of punitive deprivation, the intention to deprive, and censure.
A. Intention to Punish
The intention or goal behind an act or practice is generally significant in understanding its nature and justification. Some will emphasize that, even when they use a great deal of coercion, police do not intend punishment; it is not how they perceive their acts. Police would be surprised to learn that this is what they do.
How can we capture this idea? It is tempting to claim, “X is punishment only if X is intended to be punishment by those enacting it.” Taken as definitional, this is unhelpfully circular, since the object of the relevant intention (a “punishing” intention) remains unspecified; insofar as the object of the intention is specified, it refers back to the term being defined. If it is a definitionally necessary condition for an act to constitute punishment that it is intended to be punishment, then in order to punish, one must have an intention that the thing one intends is a thing one intends to be punishment. One is reminded of the problems faced by a crude intention-based theory of meaning: just as the meanings of words cannot be fully determined by the intentions of those speaking them, neither can the proper designations of acts be fully determined by those performing them.Footnote 12 Furthermore, if not fully punitive, police actions can be punishment-like; they can possess all of the features of punishment except a tendency normally to label them as punishment (if the tendency to label as punishment is indeed a feature of punishment).
In further support of this claim, consider an argument about how definitions function in the debate on justifications of punishment. Some object to utilitarians that punishment of innocents is not really a kind of punishment; punishment is essentially, by definition, something that is meted out only to those it is believed have transgressed.Footnote 13 But, continues the objection, utilitarians are committed in some conceivable circumstances to punishing known innocents. Therefore, utilitarianism has an inadequate theory of punishment: it fails by definition. H. L. A. Hart responds—and I agree—that the utilitarian can successfully reply that it does not matter whether some practice is called punishment; what matters is whether the practice is justified.Footnote 14 If necessary, the utilitarian view could punish the guilty (where socially useful), and it could “shpunish” the innocent (where socially useful).Footnote 15 The institutions of punishment and shpunishment are sufficiently similar in their purported justifications, forms, and outcomes, that we usefully discuss them under the same heading. Alternatively, the utilitarian can reply, in the same spirit, that, rather than rejecting utilitarianism, we should revise our definition of punishment so that it might, conceptually, be aimed at the innocent. Either way, the complex of punishment and shpunishment (if the latter exists) deserves our attention. And Hart goes on to give reasons, not stipulated definitions, for why punishment should only be meted out to those have done wrong.
Implicit in Hart's argument is conceptual room for the existence of a practice that is (1) relevantly similar to punishment, but that (2) we are hesitant to label as “punishment,” while nonetheless is (3) on reflection ultimately justified. Furthermore, it is a practice that (4) we are hesitant to label as “punishment” because, prior to reflection, we sense it is unjustified (and not merely because we sense it is not punishment). If we go this far with Hart, then the first part of my argument, the establishment of the possibility of legitimate quasi-punishment, succeeds. There might be activity that may not normally be called punishment, but that falls under the same normative standards, and is justified or unjustified against those standards. In his case, Hart argues against the punishment of the innocent—but accepts it is a possible category. In my case, I suggest that punitive police action is also a possible category, and also suggest that such action may be justified.Footnote 16
Don Scheid provides another approach to the definition of punishment.Footnote 17 According to his view, punishment is a “reducible concept,” whereby, according to context, normally essential elements of its meaning can sometimes fail to be present. For instance, one may say that punishment is necessarily treatment of a person who has transgressed. But when one says, “an innocent has been punished,” the concept itself instantly reduces. The “transgression” element of the definition is, given the context of one's utterance, no longer present; one is not simply saying something incoherent, equivalent to the claim, “this is a four-sided triangle.” Scheid's view is amenable to the argument I offer here. If all of the elements of the unreduced version of the concept are normatively relevant, then they should be present in police action too; if they are not, then we should not be distracted by the label “punishment,” especially in light of its distracting reducibility.
B. Hard Treatment
Some may object that police-imposed deprivations are not the kinds of deprivations that are characteristic of punishments. Punishment involves prisons, community service, and so forth. Police alone cannot coerce people into those things.
In response, note, first, that punishment is not distinct from other kinds of treatment by the hardness of the treatment involved: punishment can involve a small or trivial kind of hard treatment, including (if one is willing to include them in the definition) suspended sentences, whereas actions by the state that are nonpunitive, including investigations that lead to trials, can be seriously disadvantageous.Footnote 18 Police action in general can involve hard treatment, including hard treatment that is qualitatively similar to that meted out by courts: pretrial detention involves a deprivation of liberty that is burdensome in ways that full-blown detention is burdensome. As Seumas Miller says, “harmful and normally immoral methods are on occasion necessary in order to realize the fundamental end of policing, namely the protection of (justifiably enforceable) moral rights.”Footnote 19 Whether or not Miller is correct about the ultimate grounds of legitimate policing, it seems clear that it involves imposing burdens.
Thus, in the case of recruitment of informants by threat of prosecution, the act of informing is burdensome. It is time consuming, mentally demanding, and carries a risk of other kinds of harm. Like prison, it involves an objective element, measured in terms of a period of incarceration, and a subjective element, a varying effect that the incarceration will have on a prisoner. Similarly, public-order maintenance also involves imposing deprivations. These may often be small deprivations. But it should be clear that this does not in itself disqualify an action from being a form of punishment; if there is a difference it must lie elsewhere.
Now, a different and more powerful version of the objection that I am considering here could be mounted. It may be claimed that the offer to a suspect of an opportunity to turn informant is not a kind of hard treatment. If anything, it is a lessening in the level of hard treatment that a person would otherwise face. The offer provides a new option that would avoid a trial and possible custodial sentence. Police in these cases, continues the objection, do the opposite of punish. They ease the degree of deprivation received by those who have made themselves liable to receive more. On this view, even if some police coercion can impose greater deprivations than some punishments, the particular kinds of police action that I have proposed involve, on reflection, the reduction of deprivation, not the imposition of it.Footnote 20
However, consider the following case. A person is being sentenced, and is in fact liable to go to a certain high-security prison. In exchange for a certain commitment to cooperation, the judge recommends a different prison with a less strenuous regime. The convicted person accepts the offer. We still want to say that the new, less arduous sentence is a case of punishment. The police actions that I am considering have the same structure. One cannot, therefore, hold that some police action is not punishment merely on the grounds that it involves a reduction in the deprivation to which a person would otherwise be liable. The set of options imposed on the would-be informant, taken together—face trial, or act on behalf of the state in the ways that we request—is itself a form of hard treatment. Amalgams of unattractive options can be conceptualized as hard treatment. For instance, the choice between two different prison cells itself is an application of a severe disadvantage on an individual. And the option one selects on accepting a plea bargain remains a punishment.Footnote 21 Finally, it remains sensible to say that the individual or institution that makes the offer of a lesser deprivation nonetheless does impose that deprivation.
To be sure, not every case of burdensome questioning, arrest, or search need count as a case of punishment. My claim in this section is that it is not sufficient to distinguish police action from punishment to assert that police action does not impose hardships (since that is false), and that it also is not sufficient to distinguish police action from punishment to assert that the deprivations that police action imposes are of a kind that disqualifies them from counting as punishment (since that is arbitrary). Further, my claim is consistent with it being the case that hard treatment by the police will generally be less severe and of shorter duration than deprivations that are imposed at sentencing: it remains possible that some punishments are less severe than some police actions, and it remains the case that both are impositions of deprivation by arms of the state. Now, it might be responded that police action is distinct because it carries a different kind of goal, or expresses a different attitude or claim. I turn to those responses in Sections III.C and III.D in turn.
C. Intention to Deprive
Several authors propose that in the case of punishment, hard treatment, or its accompanying deprivation, is directly intended.Footnote 22 In contrast, there are many cases of foreseeably depriving, but justified, state action that lack a full-blooded intention to cause harm or deprivation. These include the charging of fees, civil commitment following a defense of insanity, quarantining, and large construction projects that are justified on balance but involve unwelcome forced purchase orders or deleterious effects on local businesses. Those practices involve either deprivation as a foreseeable side effect, or as a means to some other end, but do not have deprivation as their goal. Indeed, continues this line of argument, it is plausible that what makes punishment a distinct moral category, and marks it as falling under a distinct set of norms, is that it is especially difficult to justify an intended harm as opposed to a merely foreseen but unintended harm. Furthermore, there are coercive police practices that do not intuitively seem usefully categorized as punishment, because punishment has a goal-based or functional element, namely, an intention that a deprivation takes place. For example, a normal arrest is not in itself punishment, since the coercion it involves is backed by an express goal of enabling a procedure that will pass judgment on the act that instigated the arrest.
This point might be mobilized against the claim that I am urging in this paper. Thus, it will be argued that the intention behind any legitimate police action is not the direct imposition of harm or hard treatment. The proper purpose of coercive police activity is to make it less likely that people will be harmed, and to apprehend those who have acted criminally. And where it is possible for police to deploy strategies without their subjects being burdened or deprived, the police are right to choose that option. In the examples that I have offered as cases of punishment-like police activity, it might be objected, deprivation is merely a foreseen but unintended consequence of legitimate police action.
1. The Relevant Intention Is Elusive
The intention that we pick out in identifying punishment cannot be a literal one. The contents of trial judges’ minds are not decisive of the nature of the practice or its normative status. The intention should be understood, rather, with respect to the function of the practice. In grappling with this issue, Robert Nozick leaves the following question unanswered:
The theory [of punishment] we have outlined. . .places intention within the institution of punishment, not simply behind it. Whose intention, then, is relevant: that of the prison guards (which shift?), the warden, the judge, the jury, the legislature, the citizenry? Or (like Durkheimean social facts), are certain intentions built into the institution, apart from the ascertainable intentions of particular participants?Footnote 23
If we take as relevant the actual intentions of the state agents participating in the practice of criminal justice in this context, we produce some odd results. Imagine that a trial judge or magistrate has had a long week of near-identical road traffic cases; negligently, but accurately, and out of habit, on autopilot, she follows through the process of sentencing. She is barely cognizant of the distinguishing features of the defendant in front of her, and responds to them on a level that is not properly conscious, even if it is legally correct. The probation officer is in a similar state of mind, and the prison officers notice little about the person convicted, processing him as part of a batch of others. In the process only the slightest glimmer, if any, concerning the specifics of the convicted crosses the consciousnesses of those state agents who are tasked with inflicting the punishment with regard to an intention to cause a deprivation. To be sure, the person on the receiving end of this treatment has a complaint. He is mistreated, or is treated correctly only by accident. But the point is this: however we assess the complaints in this case, it is clear that it is a case of punishment; whatever intention is necessary for punishment is present, and it is not the case that the punishment is diminished because the intention is dim and inchoate. The institutional arrangements are such that a deprivation has been deliberately imposed on the individual, in response to a transgression, and the deliberateness of this deprivation, insofar as it exists in a way that is constitutive of punishment, resides in the structure of the process rather than in the minds of any of the agents who implement it.
Here is another case that establishes the same idea.Footnote 24 Imagine two defendants who receive similar sentences from different judges. Both judges are fully competent in the jurisprudence of sentencing. The first judge has, in addition, a great deal of experience and understanding of the prison system, whereas the second does not. In sentencing, the contents of the two judges’ minds will inevitably differ. The first will be conscious of details of the process of incarceration, and will (or ought to) intend that these be imposed on the convicted person. Ignorant of those details, the second judge can only intend the broad outlines: the number of years, the probation conditions, and so forth. Supposing both defendants receive the same sentence, are we to say that the first receives a greater punishment than the second? We have three options: (1) accept that the punishment is greater; (2) deny that all punishment is intentionally applied deprivation; or (3) claim that the intention behind punishment exists elsewhere than in the mind of the trial judge. Option (1) is unacceptable, and option (2) contradicts the basis of the objection that I am considering here. We are left with the third option.Footnote 25 In addressing the objection that police treatment is not punishment because it does not involve intentionally applied deprivations, then, we should understand the “intention” as one that exists in the structure, or in the function, or in the institution itself.
Some may take this point further, denying that punishment necessarily involves an intended deprivation.Footnote 26 One might include within the category of punishment the foreseen but not intended effects of actions, or their likely effects, or the effects that the actor ought to take into account even if they have not in fact been taken into account. If this is the case, then a fortiori one cannot distinguish between court practices and police practices on the grounds that only the former intend deprivation. Nonetheless, let us grant that punishment involves an intentionally applied deprivation, but that the relevant intention behind state punishment rests in the institution of the state itself, rather than in any of its agents. I will argue that, with that premise, there is still reason to attribute that intention to some police action, if the relevant intention does exist in archetypal cases of punishment (Section III.C.2).
2. Distinguishing Police and Court Intention
The informant is not accidentally recruited. It is intended that the informant go through the process of informing. And it is known what this typically involves. Similarly, public-order police do not (or should not) accidentally threaten force or arrest, and the consequences of such exercises of power are clear. How can we make sense of the idea that the deprivations imposed by police actions are not intended? The process of informing involves taking time, risk, effort at deception, gathering and interpreting information, disruption of a person's social network, and the provision of information to the handler. We might split the intention in this case apart. Perhaps the policy behind informant recruitment carries only the intention that information is provided to the handler. On this interpretation, the hardships that go with informing are foreseen but unintended consequences of the recruitment. If the informant could provide the information without going through those hardships, then this would be preferred. The deprivation itself is not intended. A similar strategy might be used for other police activity: insofar as it imposes any disadvantages, such disadvantages are not intended but are collateral effects of justified activities. The courts, on this view, intend the deprivations that they impose, and this is what distinguishes them as punishing authorities.
This strategy walks a tightrope between placing so much state activity in the category of “foreseen but unintended consequences” that court-imposed punishment may be included, and being sufficiently ecumenical about “intentional” state activity that police-imposed deprivations can be counted as such. Note first that the test for intentional harm that is implicit in the argument is difficult to pin down. On some ways of making the test precise, police-imposed deprivations of the kind I am discussing will be counted as intentional. We might, for example, say that a harm is unintended if and only if, counterfactually, were it possible for the same outcome to be achieved without the harm, then the actor would choose that course. Despite its surface plausibility, this test diminishes the category of intended harms beyond our normal usage. Compare the case of a tactical bomber, whose bombs that are aimed at military infrastructure foreseeably kill civilians, and a terror bomber, whose bombs are targeted at civilians with the goal of lowering enemy morale. It has been noted that it may be the case that both the tactical bomber and the terror bomber would choose to drop their bombs and achieve their ends without killing civilians, were this option available.Footnote 27 The difficulty is that this way of separating merely foreseen from intended consequences does not capture all that we want from that distinction. It is also not obvious that the institution of punishment passes this test: according to this view, were it possible to achieve deterrence, reconciliation, and rehabilitation effects without imposing a deprivation, then, supposing this path would be followed, the setbacks involved in custodial sentences are deemed unintentional in the relevant sense, and thereby (according to this view) the sentence is not a form of punishment.
More generally, there is reason to doubt that the strategy of arguing that police harms are unintended will succeed. In debates about the doctrine of double effect, the “terror bomber” is taken to be a paradigmatic case of intended harm, since the deaths caused are a direct means to a military end. The tactical bomber, on the other hand, does not use the civilians as a means; they are (in some sense that remains controversial) unconnected to the bomber's purposes.Footnote 28 The police strategies of informant recruitment, and many cases of public-order policing, are a much closer analogue of the former case than the latter. The targeted individual is directly involved in the strategy; the inevitable setbacks that are imposed by the police are part of an activity that is causally useful for the police in achieving their ends.
Let us try a somewhat different tack. It might be argued that the difference between police action and punishment is that the justification behind police action does not rightly include the fact that the person has done wrong.Footnote 29 On this view, police sometimes may intend deprivation, but they never rightly intend deprivation only on the grounds that it makes a person worse off. Only in the case of courts is the institution properly set up to aim at that. And, continues the objection, such an intention is distinctive of punishment. Thus, the police may aim at deprivation-for-the-sake-of-crime-prevention. This includes prevention by means of disincentivization of others and the prevention of criminal acts that would otherwise be carried out by those on whom police impose the deprivation. Police may also intend deprivation-for-the-sake-of-enabling-justice, including the imposition of bail and pretrial detention. But if they aim at deprivation-for-its-own-sake, they step beyond their proper role. That is the role of the courts alone.
This argument has the disadvantage that it defines nonretributive punishment out of existence, and to do so is to remove the grounds for discussing existing controversies. The objection supposes a retributivist interpretation of the criminal justice system. It imputes an intention to deprive behind the normal practice of punishment, and asserts that this intention is properly absent in police action. It is doubtful that punishment by definition has an intention to deprive-for-its-own-sake behind it, because an institution that conceives of itself as punishing on nonretributive grounds still imposes punishment, even if on mistaken grounds. If the goal of the administering institution is purely to deter, then its activities may remain punishment (even if they are not justified).
Some might respond by offering a separate name for impositions of hard treatment in response to a wrongdoing that have no intention of imposing a disadvantage for its own sake. Perhaps these might be called cases of “telishment” rather than punishment.Footnote 30 This would be an unhelpful narrowing of a live debate, in which it is now clear that one need not in principle commit oneself to a simplistic consequentialism in order to hold that the existence of a deterrence effect might warrant the imposition of hard treatment.Footnote 31 If it is legitimate to define out of existence “punishment” that is imposed on grounds other than intrinsic retributivism, then it is equally open for the nonretributivist to assert that the retributivist is not truly seeking to justify punishment, but rather some other concept, say, deservingment. Both moves are dialectically unhelpful. One does not need to think that retributivism is false (or indeed true) to think that it is a mistake to define away punishment that is not motivated on purely retributive grounds.
D. Censure
It is common to include in a theory of punishment an expressive or communicative element. For instance, for Joel Feinberg, punishment “has a symbolic significance largely missing from other kinds of penalties.”Footnote 32 In this section I discuss accounts of punishment that contain a communicative or expressive element. On such views, punishment is sanction and censure. It is hard treatment that carries reprobation for the breach of a rule.Footnote 33 Standardly, the element of reprobation distinguishes punishment from taxation and fees, which are an expected part of doing business. My purpose here is to argue that it is possible for police action to express public censure.
First let us develop the expressive view a little more. Punishment conveys blame or reprobation. In doing so it announces to the recipient of punishment a general social assessment of his or her actions and announces to the community, including the (purported) victim of the actions of the person who is punished, the judgment that a wrong has taken place. It might be said that the purpose of criminal sanction is to create censure.Footnote 34 If punishment is not justified by its communicative function, or its expressive function, it is still plausible that it nonetheless is essentially communicative or expressive. The justification of punishment will in that case include a justification of that element of it. The expressive part justifies, or else it is in need of justification. Either way, it seems to be an element of the concept.
The recruitment of an informant case expresses censure. The action says, “you have acted in this wrong way and you have thereby made yourself liable for punishment; if you like, you can accept a different burden, imposed on you by the state, instead of the formal punishment.”Footnote 35
To be sure, censure by police is not always public, in the sense that it is not always known by or accessible to members of the public. Some might use this fact in order to object that censure by police is thereby not punishment, or in any case has a different purpose or function, and should therefore be understood with respect to a distinct (and less restrictive) set of values. Thus, continues this objection, in police investigations in general, the censure is, even in the most open case, not public—and sometimes it is manifestly private, for good operational reasons. Informants provide an extreme case: by their nature their status must be cloaked.
Nonetheless, the potential privacy of censure does not trouble the position I am urging here. It is one thing to claim that punishment should be carried out in public, and it is another to claim that purported punishment that is carried out in private is not punishment, or in any case falls under different and qualitatively weaker standards. The latter claim seems difficult to defend. Concealed state punishment can exist. Its justification is a further question. It is thus important to distinguish between censure and stigmatization: the former implies the communication of reprobation and may or may not be public; the latter implies effective public reprobation.Footnote 36
The expressivist theory of punishment cannot focus merely on the words and writings of the trial judge; in order for that view to justify custodial sentences and other kinds of deprivation, it is necessary to argue that the hard treatment typical of punishment itself says something reprobative. My claim is that, insofar as we can make sense of this idea (and one key line of criticism of the expressivist theory voices doubts about this),Footnote 37 we can see the same in police actions.
IV. A DILEMMA AND A PROJECT
A. Dilemma
I have argued that some police strategies that are not obviously illegitimate are cases of punishment. I note how those strategies have all the features of normal accounts of punishment: they are censorious, they impose deprivation, they are carried out by those in authority in response to a transgression, and so forth. I have considered a series of ways in which one might differentiate police activities from the activities of the courts, and found them wanting, for our purposes. I conclude that police, as we conceive of them, punish not just in cases of corruption or brutality or vindictiveness, but as part of the normal workings of their profession.
This presents a dilemma. Either police routinely act in an illegitimately extrajudicial fashion, or the stringency of the standards to which we hold criminal court procedures is, if it is grounded, not grounded directly in the fact that the courts administer punishment. The dilemma arises because certain norms surround the practice of punishment. Not least among these is the idea that punishment of the innocent is an egregious wrong, and that we should therefore institutionalize steps to avoid it, such as the standards and burden of proof that exist at trial and the requirements of the publicity of criminal justice. Police action is not carried out in a way that provides direct access to those procedures. So if and when police action is punitive, we must either condemn it on the grounds that it fails to provide access to the proper procedures that necessarily attach to legitimate punishment, or we must limit the purview of those procedures as applicable only to a subset of cases of state punishment. This dilemma is of great significance for the way we theorize about the norms surrounding the criminal justice system: on one hand, it hardly needs stating that it is intolerable for police to act coercively beyond their proper role; on the other hand, our commitment to certain ideas about the proper application of punishment is deep.
B. First Horn
The first horn of the dilemma accepts that police routinely engage in activities that are properly described as punishment, and adds that those activities should therefore cease, perhaps to be reimplemented later only with a robust, court-like set of procedures. Some will be attracted to this conclusion, perhaps motivated in part by an awareness of patterns of police misconduct. Nonetheless, given what I have argued so far, it is an unattractive position. This is because so many police actions would be rendered illegitimate. The shared features of the kinds of cases that I have focused on are these: there is sufficient evidence for an arrest or prosecution, and there is a decision not to carry this out, but some other, lighter form of hard treatment is applied instead, usually with the consent of its recipient. Many police decisions have this structure, and the order-maintenance and informant-recruitment cases are extremely common.
In the case of order maintenance, rather than an informal expression of censure or a demand that a person step away from a volatile scene, the options for police would be to set in motion the procedures of the formal criminal justice system (which would in many cases be heavy-handed and counterproductive), or to do nothing. Consider, furthermore, the following goal of community policing:
The key. . .is that officers must identify and take action against disorder and minor crime, including the immediate conditions that encourage them. But taking action need not be limited to enforcement—actions should be more preventive, substantive, collaborative, and creative whenever possible. The public should still see their police addressing disorder and incivilities, but police methods should go well beyond. . .easy, simplistic, and possibly counter-productive zero-tolerance enforcement campaigns.Footnote 38
This ideal expresses one popular account of the function of police. It seems, however, on the present view, that when “preventive, substantive, collaborative, and creative” solutions involve the imposition of some burden in response to a perceived transgression, the police go beyond their proper role.Footnote 39
Similarly, informant recruitment backed by the possibility of prosecution would no longer to be a viable strategy. It would be necessary for police to have the kind of certainty of guilt that would be established by a trial before putting pressure on a person to become an informant. Proper procedure, indeed, would suggest that the process should normally be public; this sits in tension with the nature of undercover informant work. The practice is, nonetheless, currently ubiquitous.Footnote 40 And on reflection, the judicious use of its legal powers is one ideal to which we hold the police force.Footnote 41 I do not mean to suggest that the practice should be unregulated or is not in need of reform or new guidelines. Rather, my claim is that an effort to apply the normative limits to this practice that we apply to standard cases of punishment would wipe the practice out entirely, in a way that significantly reduces police ability to investigate networks of serious crime.Footnote 42
The first horn of the dilemma, then, would greatly limit police activity: in acts that involve the strategic reduction of deprivations to which people have made themselves liable, the police would be held to standards applicable to normal cases of punishment; such acts are so common and central to police activity that we would need to engage in a wholesale reconceptualization of the police function, placing novel and powerful limits on it.Footnote 43
C. Second Horn
So much for the first horn of the dilemma. Let us consider the second. This horn abandons the premise that punishment is a matter only for courts, thereby opening up space for an expansion of the theory and justification of punishment into the realm of police actions. This view is attractive not only because it avoids the conclusion that we must radically limit police activities, but also because it is in one way natural to put courts and prisons in the same normative field as policing: all involve state coercion in response to crime. Nonetheless, it is not easy to hold this view, if one also wishes to hold a conventional view of the grounding of our procedures surrounding punishment. The conventional view says:
The procedural standards to which we hold court activities (such as the standard and burden of proof, and the requirements of publicity) are stringent because courts are in the business of administering punishment.
In order to see the depth of commitment to the conventional view, consider the following idea, which is given due by a diverse range of theories of punishment: it is a grave wrong to punish those who are not liable to be punished.Footnote 44 Call this the “seriousness of misapplication.” This expresses deep-seated beliefs about legitimate punishment. At the level of principle, the possibility that it will warrant or merely countenance the punishment of the innocent is a serious charge against a theory of punishment. The idea of the seriousness of misapplication is often seen to warrant directly the general high standards of fairness to which we hold the trial process.Footnote 45 The serious wrong of misapplied punishment can be understood as directly grounding the procedural protections provided by the requirement that people be assumed innocent until they have been proven guilty beyond reasonable doubt. Footnote 46 In particular, it is prominently expressed in the criminal justice system in its procedural deployment of the presumption of innocence at trial. For instance, the principle of presumption of innocence has been described as “a fundamental normative precept of the Anglo-American conception of justice.”Footnote 47
Of course, not all subscribe to this view.Footnote 48 Thoroughgoing utilitarians will not hold it. Some may hold that it is a wrong to punish the innocent but also that it is a wrong to fail to punish the guilty, and that these wrongs balance one another. Some will argue that it may be a wrong but not a grave wrong to misapply punishment where the punishment is small, and furthermore that the punishment (if we may call it that) that is administered directly by police typically is small. Be that as it may, some police activity of the kind I am describing can impose a serious setback on a person. Furthermore, it will make a difference for many theorists whether such activity falls under the heading of punishment, and it would be a radical reform to put police and court coercion on the same footing.
The conventional view becomes difficult to sustain if one also believes that there is a wide range of cases of punishment that need not be held to such standards. That is just where we are taken by the second horn of the dilemma that I am describing. According to this position, we must deny that the seriousness of misapplication of punishment grounds our strict court procedures. Instead, legitimate state punishment can exist without those procedures: it exists in everyday police actions. Normal expressions of a commitment to high standards of proof for the imposition of punishment do not admit of easy exceptions. On the view I am describing here, such exceptions are numerous. Someone might complain, “How dare the state punish me without full proof or procedure?” That complaint is now incomplete.
D. An Escape: Legitimate Punishment Requires Access to Procedure, not Enactment of Procedure
Both horns of the dilemma involve the idea that punishment can only legitimately take place in the wake of a stringent procedure. It may be responded that the standards that are applicable to legitimate punishment are not standards that must be met in every case, but describe procedures to which the person charged must have sufficient access. For instance, where a defendant pleads guilty it is legitimate to impose punishment without calling a full array of witnesses. Therefore, continues the objection, just as there are circumstances in which court process may legitimately impose punishment without a full-blown trial incorporating the presentation and interrogation of all of the evidence, so also are there circumstances in which the police may impose punishment-like deprivations without such procedure. Indeed, in the courts, such situations are the overwhelming norm: in the United States around 97 percent of federal cases (after those that are dismissed) end in plea bargains,Footnote 49 and defendants in Crown Court cases in the United Kingdom plead guilty at a rate of around 70 percent.Footnote 50
Now, continues this objection, our abhorrence at the wrongful conviction of the innocent does not give direct reason for court procedures, such as a presentation of proof of guilt beyond reasonable doubt, to be manifested prior to every case of punishment; it requires only that those to be punished are able to activate those procedures should they choose to do so. Especially in the case of minor infringements, often all sides will reasonably prefer that the matter be settled before reaching full trial: both defendant and prosecutor will take the view that their resources are better expended on other matters.Footnote 51 And the police strategies that I have described do in fact provide the target with access to full court procedure should they prefer it. The proposed informant can refuse to inform and instead face the charges that the officer proposes will be brought; the protester can refuse to follow the officer's instructions and face a public disorder charge. Thus, concludes the objection, it may be the case that certain standard coercive police action falls under the rubric of punishment, but so long as those subject to it have the opportunity to challenge the police and to pursue the issue in the courts, such police activity is not rendered illegitimate. If so, the second horn of my dilemma—that we struggle to explain why court-imposed sanctions are restricted in a way that police action is not—falls away, since punitive police action can permissibly take place where it leaves the option of a pathway to the courts.
In response, note first that that we do not tend to think that mere formal access to procedure is sufficient to achieve its purposes. Plea bargains provide a useful case study. For instance, many find that formal access falls short in cases in which the threat of greater charges provides people with an overwhelming incentive not to invoke the procedure. In exchange for a commitment to plead guilty to a lesser charge, a prosecutor might agree not to bring to court a more serious possible charge. Alternatively, in exchange for a commitment to plead guilty to a subset of possible charges, a prosecutor agrees not to bring the full set of possible charges to court.Footnote 52 In seeking a just system of plea bargains, Richard Lippke argues as follows:
If those guilty of crimes can be brought to confess them openly and honestly, then not only will we save precious resources investigating crimes, we will also obtain a greater level of assurance that when we inflict punishment, we do so justifiably. Given the considerable moral hazards involved in punishing individuals, this higher level of assurance is not to be gainsaid. Granted, criminal trials require state officials to demonstrate beyond reasonable doubt that defendants are guilty of the crimes with which they have been charged. If that standard is truly and fully met, then the officials who impose punishment act defensibly in inflicting punishment on those convicted of crimes. Yet the standard of “beyond reasonable doubt” might be met more fully or even exceeded if defendants themselves candidly admit and fully disclose their crimes. In doing so they may reveal things that definitively affirm their guilt. We might therefore be able to say with some confidence that our punishing them is justified to a higher degree—“beyond all doubt,” as it were.Footnote 53
It is notorious that plea bargains may be structured so that an innocent defendant faces a strong incentive to plead guilty to a smaller charge, borne of the possibility of facing a very serious charge and augmented by poor standards of indigent legal services and wide prosecutorial discretion with regard to the charges brought.Footnote 54 Such incentive structures render the guilty plea as poor evidence, contrary to the function that Lippke outlines.
Just as innocents faced with plea bargains can have perverse incentives to accept them, so also can the balance of power in cases of police activity be such that mere formal access to court procedures is insufficient to make compliance with a police offer good evidence of wrongdoing. In order to urge the suspect to inform, the option of facing prosecution may be made yet less attractive, with greater investigative assiduousness leading to greater or better-substantiated charges.Footnote 55 One writer on this subject agrees that the decision to inform may be less than autonomous for these reasons, but argues, further, that this coercive element does not render the process unjustified: “the potential informant's vulnerability to the State's unequal bargaining power results from the informant's voluntary and knowing decision to engage in criminal conduct.”Footnote 56 As that author notes, this is not the whole story, since it assumes away the possibility of misapplication.Footnote 57 Where the state erroneously uses such coercive capacity in order to impose a deprivation on somebody who has not in fact made herself liable to such treatment, our sense that a person has been wronged is akin to our sense of wrong in cases of misapplied punishment.
Nonetheless, there will be cases in which the decision to cooperate would provide sufficient evidence of wrongdoing that we would sense that the police imposition of hardship is justified. Where the evidence against an individual is strong, and the perverse incentive effects of charging and investigative structures are minimized, we will be able to describe a set of cases in which police legitimately deploy their powers in order to obtain cooperation instead of arrest. These will include cases in which the cooperation is experienced as a hardship that is imposed as a response to a wrongdoing.
We can see, then, the beginning of a way out of the dilemma that I have described. If this route is pursued, it will have implications for the way that we understand and theorize about the police role. The escape does not deny my characterization of some normal police activity as punishment. Rather, it seeks to provide a framework for legitimizing it. This invites a novel way of understanding the theory of police, one that is closer to theories of punishment than is usually considered. Further, this view speaks in favor of the need for certain kinds of judgment on the part of police, who may in some cases only legitimately apply force or coercion where there is an effective (and not merely formal) opportunity for a person to avoid it.
V. CONCLUDING REMARKS
I have argued for a claim that seems at first blush counterintuitive: certain standard police practices are normatively on a par with punishment. The thesis is less surprising when we notice other areas outside of penal institutions in which both our practices and our theorizing reach for principles standardly applicable to punishment. For example, Douglas Husak argues that relegislating preventive detention as criminal punishment would have “the virtue of candor.”Footnote 58 If it is possible for punishment to exist outside of the criminal process in the form of preventive detention (even if uncandidly), then one naturally wonders what other actions by prosecuting authorities and police could amount to cloaked forms of punishment. Against the spirit of Husak's argument, it may be the case that actions that fall under the normative standards of punishment legitimately exist outside of direct court authority. Police work as we (upon reflection) conceive of it exhibits a large amount of discretion; this is exemplified by the strategies of prosecution-backed informant recruitment and public-order policing on which I have focused here.
If this is correct, then we should feel a jolt, both theoretically and practically. At the level of theory, my argument poses a threat to the traditional divide between the courts as the enactors of criminal justice, and police as mere enablers. Our assessment and justification of the criminal justice system ought to be a great deal more alive to the effects of policing, including acts that fulfill the normal criteria of punishment. And thus, in the realm of practice, we face a dilemma that threatens either a popular conception of legitimate policing, or a popular conception of legitimate punishment. Finally, I have sketched a view that would escape this dilemma, while still endorsing the characterization of some normal police activities as punishment.
