Ethical Theory and Moral Practice

, Volume 16, Issue 4, pp 777–790 | Cite as

Dirty Hands and the Complicity of the Democratic Public



The alleged problem of the dirty hands of politicians has been much discussed since Michael Walzer’s original piece (Walzer 1974). The discussion has concerned the precise nature of the problem or sought to dissolve the apparent paradox. However there has been little discussion of the putative complicity, and thus also dirtying of hands, of a democratic public that authorizes politicians to act in its name. This article outlines the sense in which politicians do get dirty hands and the degree to which a democratic public may also get dirty hands. It separates the questions of secrecy, authorisation, and wrongfulness in order to spell out the extent of public complicity. Finally it addresses the ways in which those who do and those who do not acknowledge the problem of dirty hands erroneously discount or deny the problem of complicity by an appeal to the nature of democracy, a putatively essential need for political openness or to the scope of ideal theory.


Dirty hands Secrecy Authorisation Bad faith Openness Non-ideal theory 

1 The Problem of Dirty Hands

The problem of dirty hands has been extensively discussed in the wake of Michael Walzer’s original, seminal article on the topic (Walzer 1974). In the subsequent debate there has been disagreement as to whether there is any such problem (Nielsen 2000; Gaus 2005) and as to whether or not the problem is exclusively or predominantly a problem of political action (Stocker 2000). I shall accept both that there is a phenomenon of moral note which merits the description ‘dirty hands’ and that political actors can get their hands dirty.

My concern is less with the problem as it allegedly affects those, politicians, who must dirty their hands. Instead it is with the implications of the problem for those, the public, who might appear to escape its grip. Put another way, I am interested in the situation of those who give others the licence to do what results in dirty hands, but who might not themselves get sullied. This is the problem of the complicity of non-political actors, those who democratically authorize political actors. In what sense are the hands of the former also dirty? I shall argue that those who do acknowledge the problem of dirty hands oversimplify or misconstrue the problem of complicity, and that those writing within Anglo-American political philosophy are unable to acknowledge the problem as it affects either politicians or their public.

Let me first lay out the terms of the original problem of dirty hands. It is that of being required on occasion to do what is necessary but what is also wrong at the same time: ‘For sometimes it is right to try to succeed, and then it must also be right to get one’s hands dirty. But one’s hands get dirty from doing what it is wrong to do’ (Walzer 1974 p. 68). The puzzle thereby identified lies in the fact that it is prima facie inconsistent to describe an action as both right and wrong.

Those ill-disposed to the existence of paradoxes will refuse to acknowledge that the problem is a genuine one. They say one of two things: either it is the case that what the person does is right all things considered; or it is the case that what the person does is wrong absolutely and unconditionally. Each response to, and purported dissolution of, the problem of dirty hands seem to fit with the demands of, respectively, consequentialist and deontological moral theory (Lukes 2006). For the first holds that it is appropriate to weigh the wrongness of the action in the balance of morally relevant consequences and find that it is outweighed by the beneficial outcomes secured. Whilst the second theoretical approach holds that the action has breached a constraint upon permissible actions which should never be breached.

Those who favour the first response and who hold that a purportedly dirty hands action is in fact permitted (or indeed even morally obligatory) all things considered will not deny that people do sincerely feel contrite for the ‘wrong’ that they have done. They will however assert that a person should only feel regret at finding themselves in a situation where they have to perform the action in question; they should not feel guilt at having in fact done something wrong (Cullity 2007). It could be held that such feelings of guilt, even if ultimately unwarranted as a sign of wrongdoing, do nevertheless serve good purposes. The propositional content of such a feeling (‘I have done wrong’) is mistaken, but the feeling itself has some instrumental value. These good purposes might, for example, be those of ensuring that individuals take great care when considering whether to do what is apparently very wrong. Thus, we would not wish individuals to take lightly the possibility of killing a human being even if there are occasions on which the killing of another is justified all things considered.

The puzzle of dirty hands arises in the following manner. A contemplated action breaches a moral constraint thought normally to be absolute. Yet the action in question needs to be done in order to bring about a highly desirable outcome, or to avoid a disastrous outcome. Walzer’s much discussed example is that of torturing a terrorist to disclose the whereabouts of the bomb which will otherwise kill many innocent people. We are to suppose both that torture is wrong but that the discovery of the bomb and the resultant prevention of a calamitous loss of lives are necessary. It is not that the torture is somehow justified overall by a balance of considerations in which its wrongness is outweighed by the good of saving so many lives. Rather it is that the torturing of the bomber is wrong and remains so. Yet the torture of the terrorist was also necessary if the havoc he intended to wreak was to be averted.

In sum, the terrorist must be tortured. Yet even after the torture has achieved its end, there is a ‘remainder’, a wrong that is not expunged or silenced by the weight of countervailing good.1 Others speak of the ‘moral residue’ bequeathed by our ‘solution’ of moral dilemmas (McConnell 1996). Bernard Williams writes of the ‘uncancelled moral disagreeableness’ of some actions that may be necessary and overall justified (Williams 1978).

Walzer himself thinks that it is appropriate for the person with dirty hands to feel that the dirt still clings to hands. He did something terrible but at the same time he had no real choice but to do what he did. Such dirt is evidence of real as opposed to misplaced even if useful feelings of guilt. Hence the politician who dirties his hands can only cleanse them by a process of ‘self-punishment and expiation’. Walzer’s favoured literary representation of those who do undergo such a process are the executed ‘just assassins’ of Camus’ play of that title. They are willing to die, and do die, because of what they have done: ‘On the scaffold they wash their hands clean and… they die happy’ (Walzer 1974 p.80).

For my purposes I shall assume only the following. Those who have ‘dirty hands’ are those who do what may be necessary and justified overall but who can only do what they morally feel great reluctance to do. They feel such reluctance because what they do is prima facie very wrong. Having done what they had to do they feel that something is left over, something is not ‘cancelled’ or silenced by the weight of reasons that tell in favour of the overall outcome. Their own contrition and sense of wrong-doing is not expunged by any sense, however warranted, that they had no choice and that circumstances demanded that they do what they did. They feel the stain of that deed not to be expunged. They may be mistaken to feel as they do. They may be right to feel it inasmuch as they have acted impermissibly in violating an absolute constraint. It may, as Williams insists, be important that they were disinclined to do what was morally disagreeable but necessary, insofar as they are the kinds of person who will not do what is disagreeable when it is not necessary (Williams 1978). The point is that some individuals will do what they feel (rightly or wrongly) to dirty their hands and others will not. Moreover the former may be acting in the name, and interests, of the latter.

2 The Nature of Political Action

Having outlined what the problem of ‘dirty hands’ amounts to I turn to the question of whether and why ‘dirty hands’ are to be found in politics especially. There are three important, relevant features of the political. The first two serve to explain why politicians’ hands, more than those of non-politicians, may get dirty, whilst the third generates, in conjunction with the first two, the particular problem to which I wish to draw attention. These three features are those which Stuart Hampshire identifies (in more or less similar terms) as the elements of ‘Machiavelli’s problem’ and as distinguishing public from private life (Hampshire 1978).

First, the stakes are much higher in politics than elsewhere. It is the sphere in which very great good can be done and very great harm can be avoided, in which the fate of a whole community may be gravely damaged, saved or lost. Political policies affect the fate of all those who live within the boundaries of a state. Political decision-making ranges over questions that determine the economic, social, physical and psychological well-being — that touch indeed on the very future existence and physical integrity — of whole populations.

Not all politicians will make momentous decisions. Some will spend a lifetime in comparatively minor positions executing insignificant policy; others will never be anything more than members of a legislature. Even those who occupy the highest offices of state might never have to go to war or declare a national emergency. Yet the political responsibility of making major decisions does not only fall upon unelected, supreme, all-powerful leaders. Even the democratically chosen politicians of a country confront the prospect of making critical decisions of enormous import. They may need to act to avert (or to initiate) a military action, to prevent a coup d’état or an invasion, to forestall social collapse in the wake of a catastrophe, or to take measures that ensure the continuation of stable democratic governance.

The good that can be done or the evil that is avoidable in the canonical dirty hands dilemmas conform to this template of political action — momentous, unavoidable and wide-ranging in its effects (Coady 2012). Politicians — perhaps uniquely amongst the professions — make extraordinary historic choices that affect everyone and perhaps for all time. In consequence the burdens of office are considerable and hugely demanding.

Second, politicians can readily dirty their hands. Politics is that sphere in which it is most likely or commonplace that very wrongful acts may be done, or matters can be arranged for others to do bad things. In the first place the state is the site of force and violence either actual or in prospective use. The state, as Max Weber avers, ‘claims the monopoly of the legitimate use of physical force within a given territory’ (Weber 1948 p.78). Politics is in the last and decisive analysis the exercise of officially sanctioned violence (Walzer 1974). Politicians command the resources of the state in pursuit of their ends. They command not only those means of securing the continuing stability of their own regime, but those means — armed forces, intelligence, counter-intelligence and anti-terrorist personnel, diplomatic staff, foreign agents, immigration officers — that are necessary to defend the integrity of its borders and its national security. In sum, politicians have the means at their disposal to do considerable wrongs both to their own citizens and to those of other states.

Furthermore, their profession provides them with regular if not necessarily predictable opportunities to do such great wrong. The British Prime Minister, Harold Macmillan, famously responded to a journalist who asked him what was the greatest challenge facing a statesman and most likely to divert a government from its current course of action, ‘Events, dear boy, events.’ It is an endlessly quoted (and misquoted) epigrammatic summary of the radical contingency of those circumstances in which politicians must act. The provenance of the famous Macmillan quote is uncertain. However Abraham Lincoln certainly said something very similar: ‘I claim not to have controlled events, but confess plainly that events have controlled me’ (Lincoln 1864). Politics is the sphere of certain unpredictability (Sorell 2000). Politicians know, even if they know nothing else, that they will almost certainly face sui generis situations of great moment and consequence, ones for which no previous occasions will have prepared them.

Moreover, political choices tend to be especially acute and dramatic. Politicians face situations in which the choices are clear and often sharply opposed; they face emergencies which call for immediate, robust, and unequivocal responses. Thus when ‘events’ do demand a policy or a course of action the politician is most often confronted with simple alternatives of momentous consequence.

In sum politicians have the means, the opportunities and the pressing need on occasion to do what is ‘morally disagreeable’. Such wrongs can be done in two very different ways. Politicians use their political power to pursue various ends, and in pursuit of those ends wrongs may be done. Wars kill, laws coerce, economic and social policies impoverish some and enrich others. Yet politicians can only use power if they retain it. As Williams notes, ‘trying to stay in office’ is a central and critical political activity (Williams 1978 p. 57). There are, as he says, many morally (and politically) wrong ways of doing so (such as buying votes). There are others which are also wrongful but may be justified if the power thereby retained is used to good ends. The reality of quotidian politics is that politicians must frequently deceive, break promises, lie, cheat, bully, and compromise.

The third important and relevant feature of the political is that politicians act in our name and on our behalf (Walzer 1974; Hollis 1982). They are the representatives of the collective. They may be empowered to do so by democratic procedures, or they may do so as our self-appointed rulers. Of course politicians can pursue power for its own sake or for that of their own glory and personal interests. Those who do act on our behalf may fail in various ways and to varying degrees by dint of incompetence, venality, or the unavailability of suitable means. However politics as a profession, a calling, a vocation, a craft or skill is distinguished by its collective purposes and ends. Politicians are our guardians. They are empowered and normally expected so to act.

3 Collective Political Action: Whose Hands are Dirty?

From the fact that politicians act on our behalf and that, in democratic societies, we explicitly empower them to so act, it seems to follow, as Martin Hollis argues, that ‘when [the politician’s] hands get dirty, so do ours’ (Hollis 1982 p. 396; emphasis added). This is the simple view: the dirtiness of hands that is a feature of political activity extends in a straightforward, unaltered and unreduced form, from empowered political to empowering non-political actors. However matters are complicated and to see why this is the case, and fully to appreciate the problem I am concerned with we need to clarify various issues.

4 Authorization

The democratic public authorizes its politicians to act on their behalf. The authorization is both of a set of persons and of a class of possible actions. Politicians are identified as the appropriate source of political acts and yet constrained in what they are permitted to do in the name of those who gave them power so to act. This is obviously a very simple summary and many difficulties attach to the matters of exactly who can do exactly what (and only what) is authorized. It may, for instance, be thought to make a difference whether politicians act as representatives or as delegates of those who authorise them.

Yet crucially however exactly authorizations operates its overall effect is such that the public ‘owns’ the actions of its politician. Non-politicians are the original source of political actions. As the etymology of ‘authorization’ indicates the public is the true author of the actions performed by others.

There are interesting questions as to how far and in what manner a collective such as the public might be held morally accountable for the actions of some individuals, namely politicians, acting in their name. These include whether or not the democratic public can be accounted a collective, whether or not any collective can be held morally responsible, and whether or not the processes of political representation ensure that moral blameworthiness passes from the actions of the politicians to the collective that elected them (Pasternak 2011). Nevertheless, these questions are at a tangent to those of the present piece. Debates about collective moral responsibility arise when there is a question of ascribing accountability for some egregious wrongdoing such as the pursuit of an unjust war. The dirty hands problem is one of necessary but ‘morally disagreeable’ action. Second, as will become clearer in the next section, this article argues that there is a moral difference between politicians and their public, not in respect of their degree of responsibility for wrongdoing but in respect of the extent to which their hands are sullied by doing what is necessary yet such as to leave a moral ‘residue’.

5 The Division of Labour

The simple view introduced above fails to acknowledge the moral division of labour which follows from a basic (‘political’) division of labour. According to the latter some individuals fulfil the roles of political agency and others do not. By political agency I mean merely such forms of activity as are associated with the coercive exercise of state power or in short with ruling. I do not intend political agency to be particular to democratic governance even if this is the presumed form of polity under consideration here. Only anarchists, defenders of mass direct democracy and those who believe with Marx that under communism the ‘administration of things’ supplants the ‘government of men’ will be disposed to dispute the necessity of such a division of labour. Even a utopian socialist may concede that only some can be political administrators.

The claim of necessity for this political division of labour is not a substantive claim about the inevitability of oligarchy that appeals to psychological or sociological facts demonstrating differences between those disposed to rule and those disposed to be ruled (Michels 1962). Rather the necessity of a political division of labour follows from the overwhelmingly obvious efficiency, even in the most democratic of societies, of only some occupying those offices and positions that are associated with the discharge of specific political duties.

The basic political division of labour, conjoined with the ‘dirty hands’ problem, yields a putative moral division of labour. Only politicians can do those things that sully hands; only politicians can order the torture of suspects or the bombing of targets with known civilian casualties. Of course the torturer faces his difficult decisions: Should I inflict more pain and risk death in the hope of finally securing a confession? However the torturer does what he does under the command of his political master from whom the initial order to torture comes in the first instance and with whom the buck ultimately stops.

There is a moral division of labour in the relevant sense because the exercise of political agency, even if fully authorized, is special in at least the following respects. First, the politician is closer to the act both in terms of being the person from whom the decision so to act comes and in terms of knowing clearly and fully what is being ordered. Second, the politician runs the risk of getting it wrong. The first two features of the political were that it is predictably highly unpredictable and that the opportunities to do or avert great harm are unique. A politician must take decisions in the knowledge that he may make one that does enormous but unnecessary harm. He may dirty his hands needlessly; he may do a wrong that is not only not ‘uncancelled’ but has no gains to set against it.

If only the politician does (and should properly) feel the pain of doing uncompensated and ‘residual’ wrong (even when getting it right) then his role is an unappealing one. It is all the more so precisely because of the scale of the wrong that can be done, and perhaps only done by someone who exercises the collective power of the politician. Indeed what we expect of politicians is even greater, and the moral division of labour even more pronounced. We demand of our politicians both that they carry the weight of decision-making that leaves them with dirty hands (and on occasion hands that cannot ever be cleaned) and that they be of a certain moral character. We wish them to be persons who clearly and distinctly recognise the moral ‘disagreeableness’ of what they do. They must be able to dirty their hands on our behalf but acknowledge exactly what it is to have dirty hands. To repeat Bernard Williams’ important point: ‘only those who are reluctant or disinclined to do the morally disagreeable when it is really necessary have much chance of not doing it when it is not necessary’ (Williams 1978 p. 64).

6 Secrecy

It might be thought that when politicians dirty their hands on our behalf they do so secretly. To that extent then we cannot dirty our hands inasmuch as we do not even know what is done in our name. However it is important to clarify what must be kept secret and what, if done secretly, might be thought to dirty the hands only of those in the know.

Politicians keep some matters secret, and on occasion they are justified in doing so. There are then different reasons for political secrecy (Yeo 2000). Let me offer an indicative but not exhaustive list of political decisions that, prima facie, ought to be kept from the public. There are financial matters, such as the impending devaluation of a currency, whose full disclosure would be economically destabilising. There are matters pertaining to the conduct of an on-going war or military action whose disclosure would assist the enemy. Some of these might, further, involve or have as a consequence the dirtying of the politicians’ hands. A frequently cited example is the decision taken in the Second World War by the Prime Minister of the United Kingdom, Winston Churchill, not to act on the information received that the German Luftwaffe planned to bomb Coventry. To have effected a systematic evacuation of that city’s civilian population would have revealed to the enemy that the United Kingdom was in possession of the means to decipher their messages. Keeping that secret, and thus continuing to having effective access to the Germans’ encrypted conversations, could be judged greatly to expedite the successful pursuit of victory. Indeed it might be viewed as essential to the Allies’ eventual military triumph. As a result the keeping of that secret saved many more future lives than were lost in a single night’s bombing of one city. Arguably — and in the terms that Walzer understands the problem — Churchill acted wisely and as he had to but he still had blood on his hands after the event (Mendus 2009 pp. 52–3).

In addition to financial or military matters there are judicial proceedings or criminal investigations that touch on state security. Here again politicians might judge, and with justification, that secrecy is the best course. There are diplomatic negotiations with a foreign power whose conduct is delicate and whose successful outcome could be jeopardised by premature disclosure of their content. A particularly interesting example is that of negotiations with an enemy. Think of talks by the British Government with the Provisional IRA during the Northern Ireland ‘troubles’; or of possible talks between Western allied states and the Taliban in Afghanistan. Their progress might be imperilled by publicity and in these cases the peril would derive from reputational risks and threats to the legitimacy of the negotiating parties. Thus neither party to such negotiations might wish it to be known that it was talking with an enemy for fear that its respective supporters would withdraw their support from the continuation of negotiation..

Finally, within the category of warranted political secrets is information whose full revelation might be expected to have consequences fatal to or very damaging of the resolution of particular crises or emergencies. Consider how, for instance, making known the nature of a contagious disease could cause civilian panic of a kind that rendered it impossible to contain. Or consider how letting the public know how badly a war was going might occasion a general demoralisation that undermined the collective will to win what could otherwise be won.

Clearly falling outside the category of warranted political secrets are those maintained for narrow party sectarian or personal advantage; and those that touch on the illegal or immoral behaviour of politicians, their taking of bribes to expedite policies for instance.

7 Wrongfulness

Politicians can do wrong in the exercise of power and in its retention or pursuit. The wrongs that politicians can do range in severity. At the least severe end of spectrum and almost as a requirement of quotidian politics politicians break promises, lie, and deceive. To stay in power even the most estimable of governments may have to blackmail those whose support they need and to blacken by false rumour the reputations of their opponents.

At the other end of the spectrum politicians’ wrongdoings can extend to violations and suspensions of basic rights. These in turn might be isolated transgressions or be more systematic and sustained throughout the conduct of a particular course of political action. Rights violations can comprise causing civilian casualties, torture, detention without trial, non-consented quarantining of individuals, and the unauthorised gagging of media personnel. Politicians may also sanction, permit, endorse, encourage, or assist in the violations of rights by states over whose actions they have significant influence.

8 Complicity and Bad Faith

Bringing together the points made we can conclude that not all political wrongdoing needs to be secret, just as not everything that is kept secret is wrongful. Equally politicians may be authorized or not authorized to do wrong or to act in secret. Finally what is done openly or in secret, is or is not authorized, need not be seriously wrongful. There are thus many possible combinations. The interesting one is an authorized use of political power to do something that is ‘morally disagreeable’ and which leaves the politician with dirty hands. The simple view is that defended by Hollis: our hands are just as dirty as those of our politicians. In fact our hands are dirty but not as dirty. Inasmuch as we authorize politicians to act on our behalf and to get their hands dirty we thereby escape full but not all complicity in what is done. The moral division of political labour is a consequence of the fact that the politician does what we would not or could not do ourselves. That moral division of labour ensures that the dirt of politics is not simply and straightforwardly communicated from the hands of politicians to those of their authorizing public.

Since some dirt does transfer it would be bad faith for non-politicians to assert that their hands are clean. At its simplest someone acts in bad faith if they do something whose true import they know and yet disown or discount. In one of Jean-Paul Sartre’s more famous examples the woman whose prospective lover puts his hands on hers leaves her hand where it is but does not invest her passivity with the meaning of active acceptance of his gesture. She does so by effectively disowning her own hand. ‘The young woman leaves her hand there, but she does not notice that she is leaving it’ (Sartre 1958 p. 97).

In similar terms the public may be guilty of something akin to Sartrean bad faith, both knowing and affecting not to know that something bad has been done in its name. We allow our politicians to dirty their hands on our behalf, although we deny that our own hands are as a result sullied to some degree. The wrongful action is thereby falsely represented as not being owned by anybody but the political agents who represent us. A public disowns actions of those to whom it has given power and ignores its own complicity in actions it has authorised and is thus the ultimate author of.

Machiavelli requires of the Prince that he acts immorally to retain power but represent himself to those he exercises power over as being innocent of wrongdoing. Martin Hollis glosses this as demanding that the politician ‘get his hands dirty and wear clean gloves’ (Hollis 1982 p. 389). Others sympathetic to this idea suggest that it may indeed be a failing of liberal political philosophy that it cannot see that politics must be dirty (Bellamy 2010). This may be so. The point being made here is that the public seeks to view itself as not complicit in the wrongdoing of its authorized political representatives. It is thus not so much a case of the politician getting his hands dirty but being required to wear clean gloves, as one in which he gets his hands dirty as do non-politicians but the latter get to wear the clean gloves. The problem is not a denial that politics is dirty but a denial that those who are not politicians are sullied.

It of course helps that many wrongful political actions are secret. In this way a public can simply deny that it knows what is being wrongfully done and in this manner escape complicity. However the public can authorize secret actions, as can anyone. The authorization of secret actions is not the explicit and open authorization of particular acts; it is the authorization of such kinds of action, within specified limits, as may be necessary at some point.

Not knowing exactly what is done in our name is perfectly consistent with its being owned by us in the sense of authorized in advance. Note thus the ambiguity in Steven Lukes’ comment about the taking of unpalatable decisions by our political representatives:

I suspect that in any feasible world we must continue to pay people …… to take on the burden of making decisions that we would prefer not to know about. Perhaps in these cases we democratically endorse being non-democratic (Lukes 2006 p.15).

The ambiguity lies in the use of the word ‘prefer’. It could mean that we cannot know about what is going to be done if it is to be effective. Secrecy is a condition of the success of the actions decided upon. Or ‘prefer’ may mean that although we know that bad things are done in our name we would rather not know. In the latter case we are surely guilty of the collective bad faith adumbrated. We need not democratically determine that we shall be non-democratic. In delegating political decisions to our representatives we already have decided that those decisions are taken by politicians and not by ourselves. Yet they are still our decisions to the extent that both the identity of the politicians and the scope of their decision-making are authorized by us. We might discount or deny that these decisions are those of a democratic public. However if we have authorized our politicians to act in our name then their actions are also and in the last analysis our own. We do not thereby escape getting our hands dirty, even if they are not as dirty. The inauthentic donning of clean gloves should not disguise our complicity.

9 Dirty Political Hands, Clean Public Hands

Apart from the simple view defended by Hollis, my claims are directed at two kinds of error both of which in different ways deny complicity. The first is made by those who acknowledge the problem of dirty hands but who restrict its application to the class of politicians; the second is made by those who do not acknowledge the problem of dirty hands but who would misrepresent the problem of political dirtiness. I will take each in turn.

The principal discussions of ‘dirty hands’ characterize it as a problem that besets exclusively the profession of politics. Thus Williams’ piece is addressed to the moral character of politicians as a distinctive class of persons that ‘we want and need’ (Williams 1978 p. 55; my emphasis). ‘We’ of course are not politicians. Similarly Walzer is criticised for regarding dirty hands as the burden of solitary, isolated and mostly single political actors (Sutherland 2000).

Yet, even those who concede that, in principle, the dirt of politics might transfer believe that it does not and cannot in a democracy. Thus Shugarman argues that a connection between dirty hands and democracy is ‘only sustainable given a highly truncated, narrow, elitist version of democracy’ (Shugarman 2000). An early piece which recognises the significance of the use of ‘we’ in the discussions nevertheless simply affirms that ‘we’, the public, would in a properly participatory democracy hold our politicians to moral account (Howard 1977). Or again the successful dirtying of hands is viewed as necessarily ‘democratically illegitimate’ (Coady 1993 p. 537)

Sutherland for his part restricts the role of democracy to the retrospective review of the dirty actions of politicians acting beyond the reach of immediate democratic control.

The public does not so much suffer dirty hands as ‘the headache of the morning after’ in knowing what has been done in its name (Sutherland 2000 p.213). Both Hollis and Lukes regard the problem similarly as one in which democratically authorised political agents must do what they do ‘beyond our ken and control’ (Hollis 1982 p. 398).

In other words a true democracy is either one in which everything political agents do is rendered open to the authorising public — and ‘we’ would not get our hands dirty — or a somehow imperfect one in which political agents must act secretly and outwith our democratic mandate — and thus ‘we’ cannot get our hands dirty. However this is to ignore the facts of authorization, the distinctions between secrecy and wrongfulness, and the proper nature of the moral division of political labour.

10 Liberal Political Philosophy and Clean Hands

The other error is one made by those who do not acknowledge the problem of dirty hands. Here the illusion of our clean hands is maintained by views that support a contrast between ‘dirty’ and clean politics, somehow thereby attributing the dirtiness of any political actions to the failings of politicians (or of politics). This error is made by those within Anglo-American post-Rawlsian political philosophy who largely if not totally ignore the problem of ‘dirty hands’. Relevant pieces on the topic are not anthologized in a standard collection (Goodin and Pettit 2006) and the topic is not even mentioned in the most authoritative introduction to the subject (Kymlicka 2002).

Even the emergent criticism of the naïve, unworldly moralism of much Anglo-American political philosophy in the name of a socially and historically attuned realism about the political world (Williams 2007; Geuss 2008; Galston 2010) does not, as such, appeal to the problem of political dirtiness.

There are two obvious ways in which political philosophy in its dominant contemporary Rawlsian form may obfuscate our complicity in the dirtiness of politics — by representing political secrecy as itself wrong, and by simply seeing a dirty politics as belonging to non-ideal theory.

11 Political Openness

It is often maintained that openness is a requirement of political life. This is evidently not so and as has already been argued politicians may justifiably keep secrets, albeit only certain kinds of secrets. Moreover those who cite the authority of John Rawls in favour of openness should acknowledge exactly how he understands his ‘publicity condition’ as a constraint on the concept of the right (Rawls 1999a). ‘Publicity’ in the Rawlsian sense is both broader and narrower than the openness of political life it might be invoked to support. It is broader in that Rawls construes publicity as more than merely accessibility and as something closer to comprehensibility and acceptability (Wall 1996). For a reason to be ‘transparent’ in this sense it is not enough merely that citizens should know why a state acts as it does, but they should also understand and endorse (or at least not reject) the reason in question. It is narrower inasmuch as Rawls thinks of the publicity condition as a constraint on the adoption of a conception of justice. He does not aver nor is it plausible to think that each and every political action that the representatives of a democratic people take in its name should be made public. As shown already, some political actions can only achieve their ends if they are not disclosed to the public.

Politicians are accountable for their actions to the democratic public. Accountability, strictly speaking, presumes that an account can be rendered. To this extent transparency is a generally warranted norm of democratic political life, even if it may be possible for politicians to provide the account of what has been done only after the fact. Crucially we cannot reasonably demand of politicians that they are always completely open and honest about everything they are doing in the name of the public. Moreover, to repeat the point made earlier, acts done secretly need not be unauthorised ones, nor indeed wrongful ones. We continue to own and thus to be complicit in those political deeds whose secret performance we have authorised

12 Non-ideal Theory

The identified problem of public complicity in political dirtiness is not one of non-ideal theory. The distinction between ideal and non-ideal theory derives from a brief outline in Rawls’s A Theory of Justice (Rawls 1999a §39) and is now extensively discussed (Stemplowska 2008; Swift 2008; Phillips 1985). Non-ideal theory, at least so far as Rawls understood it, is designed to deal with those ‘unhappy’ situations in which individuals are not generally compliant with principles of justice, or in which natural circumstances make adoption of such principles problematic.

However even in a just and well-ordered society whose citizens are generally compliant politicians may still have to sully their hands; they may have to do what is ‘morally disagreeable’ in order to secure great goods or to avoid terrible harms. Of course this necessity may press only inasmuch as other states or political actors are not motivated by a sense of justice. The leaders of just states may need to dirty their hands in defending themselves against the aggressive actions of unjust states.

Nevertheless Rawls also understands that a well-ordered society is one in which there is general, if not universal, compliance. Thus it is possible that even within such a society the behaviour of only a few could jeopardise society’s continued existence and necessitate the taking of extreme measures. Enormous damage can be done, for instance, by very small terrorist cells.

Emergencies — due to natural catastrophes and not the actions of any person — might also require politicians to dirty their hands. Such disasters need not be anything other than extremely rare events. They are not an effect of an enduring failure of the natural order. Yet they fall within that domain of certain unpredictability that defines the scope of political action. This action in turn could involve hard choices and the dirtying of hands.

In short even within the scope of ideal theory as Rawls understands it there is a problem of ‘dirty hands’. It arises from the fact that other states are unjust, that the unreasonable few within a just state can do enormous harm, and that even well-ordered societies may fall victim to the depredations of events that lie outside the scope of justice. Well-ordered and just societies may need their politicians to dirty their hands. This necessary concession is distinct from the admission that actual, historically existent well-ordered societies may sometimes do what is less than morally ideal. Thus Rawls’s reference to ‘dirty hands’ in The Law of Peoples is not an acknowledgment of Walzer’s problem. It is an entirely proper recognition of the fact that real well-ordered societies may have behaved in less than morally ideal fashion and yet still be justified in their course of action (going to war with non-well-ordered societies) (Rawls 1999b p.94 n.7).

13 Concerns and a Conclusion

This piece has been concerned to identify clearly the degree of complicity a democratic public bears for the dirty hands of its politicians. That complicity is ignored or misrepresented by both those who do acknowledge the problem of dirty hands and those who fail properly to do so. What is crucial to my discussion is the existence of a division of moral labour that follows from the division of political labour. This moral division means that politicians get their hands much dirtier than the public whom they represent, but that the democratic public is still complicit to a degree in their representatives’ actions.

Two dangers threaten. The first is to the integrity of the politicians. Apart from the personal costs of having to do bad things, there is moral jeopardy. The person who is prepared to do bad (perhaps a great deal of bad) albeit in order to achieve good may be willing (increasingly) to do wrong for no good or simply to fail to see what is good. What threatens is both a greater preparedness to conceal and to deceive, where secrecy may not be strictly necessary; and, further, the increasing opaqueness to the politician of the crucial distinction between what is and what is not for the good.

The second danger is to the conditions of democratic self-government. It threatens not the integrity of politicians but the manner in which the public authorises its politicians to act in its name. A public that delegates to its politicians the discretion to act on its behalf may increasingly fail to recognise the proper limits of such delegation. Politicians may be more disposed to see it as unnecessary to disclose what it is that they are doing in the public’s name; the public for its part may be increasingly unwilling to demand an account of what is being done in its name. It may be increasingly incapable of discerning that the discretionary powers exercised by politicians are being abused or extended beyond what is strictly required.

To summarise: the original problem of dirty hands is that politicians must do what is morally ‘disagreeable’ but morally necessary. They are left with an ‘uncancelled’ wrong. The problem I have addressed is the extent to which we, the democratic public, also get our hands dirty, and how we might deny as much in bad faith. Contemporary political philosophy is wrong to deny the problem; it cannot appeal to the essential need for political openness, to the character of democratic governance, or to the scope of ideal theory to support that denial.2


  1. 1.

    This of course assumes that torture can be justified by the overall good it secures or harms it avoids. This is controversial. Henry Shue allowed in an early, and influential, article on the subject that it might be. (Shue 1978). He has come to believe that this is not the case (Shue 2009).

  2. 2.

    I am very grateful to two anonymous referees for the journal whose comments have helped me greatly to improve this article.


  1. Bellamy R (2010) Dirty hands and clean gloves: liberal ideals and real politics. Eur J Polit Theor 9(4):412–430CrossRefGoogle Scholar
  2. Coady CAJ (1993) Politics and the problem of dirty hands. In: Singer P (ed) A companion to ethics, 2nd edn. Basil Blackwell, Oxford, pp 532–540Google Scholar
  3. Coady CAJ (2012) The problem of dirty hands. Stanford Encyclopaedia of Philosophy Last accessed 04.07.2012
  4. Cullity G (2007) The moral, the personal, and the political. In: Primoratz I (ed) Politics and morality. Palgrave Macmillan, Basingstoke, pp 54–75Google Scholar
  5. Galston W (2010) Realism in political theory. Eur J Polit Theor 9(4):385–411CrossRefGoogle Scholar
  6. Gaus GF (2005) Dirty hands. In: Frey RG, Wellman CH (eds) A companion to applied ethics. Wiley-Blackwell, Oxford, Chapter 13Google Scholar
  7. Geuss R (2008) Philosophy and real politics. Princeton University Press, PrincetonGoogle Scholar
  8. Goodin R, Pettit P (2006) Contemporary political philosophy, 2nd edn. Blackwell, OxfordGoogle Scholar
  9. Hampshire S (1978) Public and private morality. In: Hampshire S (ed) Public and private morality. Cambridge University Press, Cambridge, pp 23–54CrossRefGoogle Scholar
  10. Hollis M (1982) Dirty hands. Br J Polit Sci 12:385–398CrossRefGoogle Scholar
  11. Howard WK (1977) Must public hands be dirty? J Value Inq XI:29–40CrossRefGoogle Scholar
  12. Kymlicka W (2002) Contemporary political philosophy, an introduction, 2nd edn. Oxford University Press, OxfordGoogle Scholar
  13. Lincoln A (1864) Letter to A.G. Hodges. In: The Oxford dictionary of quotations. Oxford University Press, OxfordGoogle Scholar
  14. Lukes S (2006) Liberal-democratic torture. Br J Polit Sci 36(1):1–16CrossRefGoogle Scholar
  15. McConnell TC (1996) Moral residues and dilemmas. In: Mason HE (ed) Moral dilemmas and moral theory. Oxford University Press, New York, pp 36–45Google Scholar
  16. Mendus S (2009) Politics and morality. Polity Press, OxfordGoogle Scholar
  17. Michels R (1962) Political parties, a sociological study of the oligarchical tendencies of modern democracy. Translated by Paul, E. and C., Introduction by Lipset, S.M. The Free Press, New YorkGoogle Scholar
  18. Nielsen K (2000) There is no dilemma of dirty hands. In: Ryman P, Shugarman DP (eds) Cruelty and deception: The controversy over dirty hands in politics. Broadview Press, Peterborough, pp 139–155Google Scholar
  19. Pasternak A (2011) The collective responsibility of democratic publics. Can J Philos 41(1):99–124CrossRefGoogle Scholar
  20. Phillips M (1985) Reflections on the transition from ideal to non-ideal theory. Nous 19(4):551–570CrossRefGoogle Scholar
  21. Rawls J (1999a) A theory of justice. Revised edition. Harvard University Press, CambridgeGoogle Scholar
  22. Rawls J (1999b) The law of peoples, with “the idea of public reason revisited”. Harvard University Press, CambridgeGoogle Scholar
  23. Sartre J-P (1958) Being and nothingness: An essay on phenomenological ontology. Translated from L’Être et le néant: Essai d’ontologie phénoménologique [1943] by Hazel Barnes. Routledge, London, p 97Google Scholar
  24. Shue H (1978) Torture. Philos Publ Aff 7(2):124–143Google Scholar
  25. Shue H (2009) Making exceptions. J Appl Philos 26(3):307–322CrossRefGoogle Scholar
  26. Shugarman DP (2000) Democratic dirty hands? In: Rynard P, Shugarman DP (eds) Cruelty and deception: The controversy over dirty hands in politics. Broadview Press, Peterborough, pp 229–249Google Scholar
  27. Sorell T (2000) Politics, power, and partisanship. In: Rynard P, Shugarman DP (eds) Cruelty and deception: The controversy over dirty hands in politics. Broadview Press, Peterborough, pp 67–86Google Scholar
  28. Stemplowska Z (2008) What’s ideal about ideal theory? Soc Theor Pract 34(3):319–340CrossRefGoogle Scholar
  29. Stocker M (2000) Dirty hands and ordinary life. In: Ryman P, Shugarman DP (eds) Cruelty and deception: The controversy over dirty hands in politics. Broadview Press, Peterborough, pp 27–42Google Scholar
  30. Sutherland SL (2000) Retrospection and democracy: Bringing political conduct under the constitution. In: Rynard P, Shugarman DP (eds) Cruelty and deception: The controversy over dirty hands in politics. Broadview Press, Peterborough, pp 207–227Google Scholar
  31. Swift A (2008) The value of philosophy in nonideal circumstances. Soc Theor Pract 34(3):363–388CrossRefGoogle Scholar
  32. Wall SP (1996) Public justification and the transparency argument. Philos Q 46(185):501–507CrossRefGoogle Scholar
  33. Walzer, M (1974) Political action: the problem of dirty hands. In: Cohen M, Nagel T and Scanlon T (eds) War and moral responsibility. Princeton University Press, Princeton, p 62–82Google Scholar
  34. Weber M (1948) Politics as a vocation. In: Girth HH, Mills CW (eds) From Max Weber: Essays in sociology. Routledge & Kegan Paul, London, pp 77–128Google Scholar
  35. Williams B (1978) Politics and moral character. In: Hampshire S (ed) Public and private morality. Cambridge University Press, Cambridge, pp 55–73CrossRefGoogle Scholar
  36. Williams B (2007) In the beginning was the deed: Realism and moralism in political argument edited by Geoffrey Hawthorn. Princeton University Press, PrincetonGoogle Scholar
  37. Yeo M (2000) Dirty hands in politics: On the one hand, and on the other. In: Ryman P, Shugarman DP (eds) Cruelty and deception: The controversy over dirty hands in politics. Broadview Press, Peterborough, pp 157–176Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2012

Authors and Affiliations

  1. 1.Department of Politics, Philosophy, and Religion, County SouthLancaster UniversityLancasterUK

Personalised recommendations