The Different Sources of Dirty Hands: Episodes, Rules, and Careers

A. Examples of Dirty Episodes

DE actions thus must maintain a kind of sui generis character and cannot (or rightly should not) be generalised into broader rules. Consider deviations from democratic procedure, such as a security official leaking important documents in violation of democratically authored rules: such practices are unlikely to ever be justified as a rule, and indeed it will be difficult if not impossible to establish rules in the first place that capture only the ATC-justified authorisation of such acts. Given the fact that such acts are necessarily departures from the norms, DE analysis and its associated demands of justification will be the appropriate framework for those rare instances where they are ATC-justified.¹³

The most famous example of an episodic DH dilemma is the ubiquitous ‘ticking-bomb scenario,’ where torturing a captured suspect emerges as the lesser evil in the face of a catastrophic and otherwise unavoidable terrorist attack. However, DH theory has focused too much on this case, and its flaws as a thought experiment make it a poor example to base DE on. It is often treated as the paradigmatic case of DH, with other cases such as grubby politicking and noble-cause corruption merely being less dramatic examples of the same principles at play in this case.¹⁴ Given that torture, and the broader policy decision to maintain an institutional capacity for it, falls outside of what democratic politics can self-justify, this assumed lack of qualitative difference is implausible.

As Shue notes, just as ‘hard cases make bad law’, perhaps ‘artificial cases make bad ethics’, offering little to no relevance to real-world cases.¹⁵ Empirical complications and real-world considerations are replaced with infeasible assumptions: the act of torture will deliver the intended results, the necessary institutional capacity for torture is somehow in the hands of otherwise good agents, the notorious slippery-slope effect can somehow be prevented. The omniscience required for situations such as the ticking-bomb scenario to be justified in practice thus will never exist, especially for those individuals who are not seeking them out, and invoking the logic of the ticking-bomb is apt to open the door for the wholly unjustifiable instances of torture that motivate an absolutist prohibition in the first place. The same can be said for ideas such as a ‘coup for democracy’ or a ‘benevolent despotism.’¹⁶ Once we assess moral dilemmas from the viewpoint of the agents who actually face them, as opposed to from some all-knowing external position, the abstractions needed to make these thought experiments work destroys any real-world applicability.

In theory, the ticking-bomb scenario serves as an example of DE. However, its flaws as a thought experiment render it an undesirable candidate for illustrating part of our moral universe. Indeed, I confess a secondary goal of this project is to help displace the ticking-bomb scenario from its central position in DH thinking. Coady provides more promising, and context-sensitive, candidates for paradigmatic DE cases.¹⁷ These three cases–compromise, extrication, and moral isolation–serve as more useful starting points than the ticking-bomb scenario, rooted as they are in failures of diurnal features of democratic politics and cooperative endeavour. Firstly, compromises, while often a professional necessity and sometimes a positive moral good, can reach a dirtying level. For example, compromising with evil actors and evil causes. While the disposition to compromise forms part of the political DC, specific dirtying decisions often merit episodic treatment.¹⁸

Next, extrication, when an agent is implicated in an ongoing immoral project, such as an illegal and unjust war, where both continued participation and a ‘clean break’ inflict serious moral costs. Using the logic of ‘lesser evils’ here helps to ensure that such participation is not normalised, and that the moral dirt reflected in each option is given its due weight. There is a prima facie duty to extricate from such practices and endeavours, yet that duty can be overruled in terms of action-guiding, though not evaluative, content. The ongoing participation in this endeavour remains wrong but remains a lesser evil.

This leads us to an underappreciated aspect of Walzer’s ticking-bomb scenario: the new political leader inherited the entire situation, including the institutional capacity to torture. Walzer himself neglects this aspect of how he constructed the problem, instead opting for a time-slice narrative where a single lone-wolf politician is the focus of the moral conflict. It is thus far more likely that a ‘moral’ politician must extricate from arrangements and agreements that involve (complicity with) torture than they are to ever encounter a static ticking-bomb scenario as conventionally described.

Finally, moral isolation produces many DH conflicts: how to do good when others, unscrupulous or even evil, are behaving badly? These cases usually fall under DR or DC. However, they can still count as DE under the right conditions, namely when the moral justification for such an act needs a ‘lesser evil’ justification that cannot be derived from the self-justified moral rules of politics or when the conventions that define DR and DC are themselves in a state of flux. These considerations all matter for Ned. There is an asymmetry between being the first to cross a line and continuing to cross it once the line has already been eroded,¹⁹ with the former meriting the heavier demands of justification and deliberation associated with DE.

B. The Limits of Dirty Episodes

What is the analytical or normative error in placing DR and DC cases, such as Edna’s and Paul’s, within the DE framework? Why is this not the be-all-and-end-all of DH analysis? I offer three answers: firstly, we cannot fully assess an agent’s moral character through DE alone. Secondly, reducing DH to DE means that relevant cases will be either regrettably excluded from DH analysis or distorted beyond recognition. Thirdly, attempting to explore a DR or DC case through DE will lead to misleading evaluative and action-guiding prescriptions. These errors will apply, too, to forcing the DC or DR framework onto an ill-fitting example, but I focus on DE as that category is the most alike traditional DH views. The three scenarios that Coady outlined, then, do generate many interesting episodic moral conflicts: important sources of dirt can still emerge outside of the context of a distinct choice.

One of Walzer’s claims is that DH dilemmas arise for ‘someone with clean hands getting them dirty’: there is an assumption that the agent facing a DH dilemma must have ex ante clean hands, and will likely have to strive to restore that innocence after the dirtying act.²⁰ In contrast, I contend that DE decisions can arise for agents who are not on that baseline (including, as we shall discuss, because they have adopted a DC or participated in a DR) and, for reasons such as political expediency and security, may not be able to cleanse their hands in the manner that Walzer calls for. DE alone cannot account for these other sources of dirt. Thus, the first limitation of DE: assessing an agent’s moral character and political profile cannot be reduced solely to the evaluation of single acts and decisions, nor can it be calculated with reference to a single baseline of moral ‘cleanliness.’ Many agents who face DH decisions will already have, as the price of effective political action, eschewed innocence as a way of life.²¹ This should not then mean that they cannot face difficult moral dilemmas. They eschewed innocence, not all moral lodestars in their totality. Some matters necessarily sit outside of the scope of DE.

This time-slice analysis cannot on its own facilitate a full moral assessment of a politician. Reynolds provides the example of a politician who believes that tackling climate change requires radical solutions yet chooses to join a less radical centrist party with greater hopes of achieving power to achieve change from within.²² Whether that politician’s approach was justifiable, whether on the grounds of moral integrity or political efficacy, cannot be assessed solely by looking at the individual decisions. The necessary fuller evaluation must include their skills and attributes, ideological commitments and constraints, and their relationships to the roles that they occupy and other political agents.

We could say that DH theory has no role in this fuller evaluation: however, this removes the possibility to investigate the many cases of ATC-justified dirt that do not meet this time-slice dynamic. Walzer emphasises the ‘critical moments’ and ‘hard choices’ that inflict dirt.²³ These are not, however, the only sources of justifiable dirt in political life. This leads to the second criticism: reducing DH to DE either excludes relevant cases that share the feature of ATC-justified dirt or distorts them beyond recognition. We cannot, as we have seen, extend the logic of the ticking bomb to other scenarios merely by lessening the stakes. But Ned’s case is also disanalogous to Edna’s and Paul’s in an additional respect: the ATC-justifiability of what he does is abnormal and aberrational. It sits outside the self-justified rules of political conduct, both in idealised and actual terms. The corruption thus requires a stringent lesser-evil justification that Edna and Paul need not produce.

This leads to the third and most important consequence of trying to reduce all DH cases to DE, that it leads to incorrect normative conclusions. The true costs of dirtying one’s hands, the overall ATC-justifiability or necessity of such moves, and other aspects of the case can be misdiagnosed, resulting in inaccurate calculations: for example, if Ned were to neglect or grossly underestimate the threat of such offers becoming institutionally normalised (or erroneously assume they already were), this would increase the risk of him dirtying his hands without ATC-justification.

Politicians like Ned, Edna, and Paul may question why they should behave decently at all in a world that seems so vile.²⁴ Breaking down DH allows agents to explore to what degree this vileness can be challenged or circumvented, and to what degree the agent’s own decency might play a role in obstructing or facilitating such change. Breaking down dirt would also help determine to what extent the agent could have, or should have, anticipated the possibility of such a scenario, and thus how liable they can be held by for their level of preparedness.

Without distinguishing between sources of DH, political actors will have a poorer understanding of what their moral and political obligations in such contexts are. Breaking down dirt therefore structures the process of moral imagination that should surround such dilemmas. Moral imagination is useful for rebutting the charge that DH theory too often treats as fixed and immutable the assumptions and premises that render dirt-free responses impossible.²⁵ Agents can critically assess where the dirt falls, whether the costs are worth bearing, and what morally accessible reform options exist, if any. A monolithic understanding of DH thus greatly hinders this exercise of moral imagination.

To illustrate this role for moral imagination and how the tripartite typology of DH facilitates it, consider Ned. Ned thinks that he must commit a determinate wrong (accepting the corrupt deal regarding school construction contracts) and that this is the ‘lesser evil’ compared to electoral defeat. Ned must meet a certain burden of proof before he can confidently justify this move. He might be wrong in thinking that his situation is a DH predicament: the dirtying deal might not be necessary for electoral viability, or the consequences of losing the election might not be dire enough to render such a deal as the lesser evil.²⁶

This is the first role for moral imagination: to determine if an agent is in a DH conflict in the first place. For this first role, the tripartite typology already comes in handy: a monolithic account of DH excludes some cases. If Ned’s situation turns out to be more like Edna’s, for instance, and the justifying logic behind his dirty conduct moves from lesser evil reasoning to abiding by the rules of the competition, then Ned might neglect the suitability of DH reasoning altogether. A more nuanced understanding of what the dirt from DH looks like helps a broader range of actors understand the moral messiness that their actions entail. A monolithic view of DH might encourage actors such as Ned to deny the moral remainders of their actions or, on the other extreme, to seek false refuge for their immoral deeds within the narrative of DH. Strengthening the tools for moral imagination can only be helpful here.

The second role of moral imagination is then in identifying what kind of DH conflict it is, and this decision has tangible impacts on how an agent such as Ned should approach their situation. All DH cases involve the recognition of some uncancelled moral remainder, but a moral remainder that stems from an exogenous source demands different treatment than those that stem from an inescapable reality of politics or the contingent rules of politics. The stakes of this distinction will hopefully become clearer as I sketch out the other two categories, but to summarise, the degree to which such dirtying conduct can and should be normalised will vary dramatically based on what category said conduct falls under.

Behaviour which might be justified in a discrete DE scenario, for example, might invite disaster if it is treated as merely ‘the way things are done’: Ned’s dealmaking is so much more corrosive to healthy democratic functioning than Paul’s unsavoury choices or Edna’s own deals (which are at least publicly understood and operate under some legal constraints). Ned needs to therefore treat his behaviour as more exceptional. Edna, for her part, needs to remember that her situation is not Paul’s. It is possible, and desirable, to change the rules of the game in such a way that future agents no longer face her conflict. Strengthening some kind of taboo or restraint regarding such practices, then, would be morally necessary in a way that it is not for Paul. A monolithic depiction of the sources of DH will obscure all these ambiguities.

A. How Politics Becomes a Dirty Career

Four aspects of the political vocation contribute to making it a DC: politics serving as its own way of life; with largely self-justified internal norms and rules; certain behaviours inherent to governing and politicking creating their own moral dirt; certain vices becoming politically virtuous as a result. The analysis in this section owes much to, and is all compatible with, insights from political realists. However, one need not be a political realist to acknowledge these features. We need not accept that politics is governed by its own entirely separate normativity to accept that politics, as a profession, invites distinctive kinds of normative reason and forms of normative reasoning.

Firstly, politics represents its own way of life, standing at least some distance away from ordinary morality. The problem for aspiring political actors is not occasional clashes between the different moral worlds that they subscribe but choosing which moral world to subscribe to in the first place. They must decide whether to embrace the necessary functions and responsibilities of political life, which requires adopting ‘entire systems of value’³⁰ and ‘different ways of life’.³¹ The aspiring politician must suppress ‘private qualms’; replace traditional norms of integrity with political variants; and choose ‘experience’ over ‘innocence’ to comprehend ‘the deeper demands of politics’.³² This neo-Machiavellian insight revives Machiavelli’s call for aspiring politicians to ‘learn how not to be good,’ rather than ‘merely ‘learn how not to act well’’.³³ Traditional morality’s standards and requirements remain constant but cannot be fully adhered to.

For example, in Paul’s case, consider a value such as a democratic belief in respecting others. As a citizen, this ideal seems to demand avoiding disrespectful behaviour, including being an honest and fair participant in debates. As a politician, he must actively promote respect as a political good. This could entail treating individuals disrespectfully to win a legislative battle, being less than honest about one’s true motivations, or delaying showing full respect to a group to better secure the broader societal expression of respect. Imagine a politician who privately supports same-sex marriage in a socially conservative society but decides to advocate civil unions to help secure more immediate wins. Alternatively, a politician in a wealthy society believes that the demands of global justice require massively increasing foreign aid spending–but for reasons of political efficacy, they instead defend their stance in terms of national self-interest. This can create dirtying moral remainders–not as severe as those usually associated with DH, but noteworthy nonetheless.

Secondly, politics is a world of ‘largely’ self-justified rules.³⁴ ‘The appropriate standards of evaluation’ emerge from within politics, not through abstract external moral frameworks.³⁵ Other moral concerns and considerations, with freestanding importance, are pursuable within politics, but the metrics for political success and the instrumental goods used to achieve such goals are determined ‘internal to the practice’ of politics.³⁶ Politicians such as Paul must ‘accept the authority of those standards’, even if some, such as ‘a degree of ruthlessness in the pursuit of power’,³⁷ can be dirtying.

This does not mean ‘anything goes.’ Vices cannot be wielded flippantly or recklessly, and certain morally important goods such as stability, rights and civility are also self-justified by politics. There are, besides, limits to what a role morality can justify. Ordinary moral concepts continue to exist, and political roles and institutions must be justified by reference to said broader morality.

Thirdly, several core aspects of governing and politicking involve actions that cause ATC-justified dirt. Political power, for example, encompasses ‘the art of permissible [state-sanctioned] violence,’³⁸ drastically increased vulnerability to moral luck, and responsibility for governing over pluralistic societies with clashing interests. Even if the context fully legitimises such acts, the politician is changing the ‘margins’ at which they experience moral conflict,³⁹ requiring a shift in moral thinking.

Possessing fine political ideals does not provide an escape. Realising the lofty goals of ‘common life, trust, and justice’ requires political tools such as ‘power over others,’ but these values are deeply vulnerable to the very same political tools needed to advance them.⁴⁰ Being responsible for this fragile balancing act is part of the democratic condition. Similarly, those politicians who are best placed to advance and embody political convictions are also often uniquely positioned to fight the ‘bad apples’ of politics.⁴¹ Democratic integrity may demand not living up to one’s stated values but instead doing what is necessary to oppose the greatest threats.

Politicians such as Paul do not stumble into positions of power by accident, but instead proactively and enthusiastically seek them out, willingly subjecting themselves to the ‘stress and frequent indignities’ of ‘the greasy pole’.⁴² To draw an analogy with war fighting and police work, ‘the morally legitimate use of [certain] harmful methods is constitutive of these occupational roles’.⁴³ Treating such methods as DE will be the right approach for someone not ensconced in such a role, but certain professions must deploy them with greater ease. This does not eliminate the moral risks that come with such practices, but it does recontextualise them.

The combined outcome of these features is that the political world values different dispositions and skills. Some of these virtues may be recognised by those outside the profession: honour and valour are usually far more important for soldiers than civilians, but the latter can understand and appreciate why such virtues matter for the former. The same goes for politicians, with duties of closed impartiality and public service not demanded of ordinary citizens. For the private citizen, sacrifices to promote the public good are often supererogatory: not so for the (decent) public servant.⁴⁴

However, prioritising these virtues can inflict dirt. There is an ‘inherent socio-psychological tension’ between roles: those in dirtying roles are very rarely expected to completely abandon their previous moral status as ordinary citizen.⁴⁵ The politician who focuses on the needs of their polity may feel that they must neglect some of the moral demands, both closer to home and to those further afield, that they still feel as ordinary citizen.

But political virtue might break from ordinary virtue in even more fundamental ways. While political virtue is not wholly alien to ordinary virtue, nor to change or contextual variance, even in the best of environment a politician must develop dispositions that require stepping away from usual moral education. This is the crucial point for Paul. The demands of public virtue will increase as the stakes of political activity increase. They may even sit in opposition to what their political goals would seem to suggest: ‘the good conscience of liberal proceduralism’ cannot serve as a perfect guide for individual behaviour.⁴⁶ The political leader must adopt ‘the qualities of a lion and a fox,’ force and fraud, and practise ‘the leadership qualities of resilience, guile, flexibility, decisiveness, courage, and the ability to master the capriciousness of luck or fortuna’.⁴⁷ Paul must ‘inhabit’ the system and display an adaptable ‘resilience’.⁴⁸ Without accepting this, Paul will not be an effective political actor.

B. The Results of Adopting a Dirty Career

What are the consequences of adopting a DC? I focus on three. A monolithic focus only on DE, as the orthodox DH formulation presupposes, would overlook these consequences.

Firstly, a DC shifts an agent’s moral baseline, changing what counts as a DE. Acts that merit conscious deliberation for ordinary citizens become mundane, if not outright virtuous, for politicians. This is not an unwelcome normalisation of shady behaviour, but a necessary and ATC-acceptable adjustment to the profession’s demands. For example, getting someone of integrity and competence fired is gravely anomalous for most people. For democratic politicians waging election campaigns, it is the end goal of many campaigns. This is not to say that Paul is committing something dirtying in trying to get able politicians from opposing parties fired. Provided Paul adheres to certain norms of behaviour, Paul’s opponent has no morally interesting complaint against his efforts to get her fired by the electorate. Rather, Paul is adopting strategies and conduct that would be dirtying in most contexts yet are perfectly acceptable within the role morality of politics. What would be a DE conflict for most is wholly legitimated under the auspices of his role.

Secondly, adopting a DC changes how politicians navigate DR and DE. The DC affect how clean politicians are before the conflict, how prepared they should have been for such dilemmas, how clean they should try to be afterwards, and how best to navigate second-order dilemmas such as those around disclosure and hypocrisy.⁴⁹ Adopting a DC also helps determines their values and goals, conceptions of integrity, specific responsibilities or role obligations they might have, and the capabilities, dispositions, and therefore options they possess.

Take the demands of disclosure as an example: the orthodox DH view, as espoused by Walzer, requires politicians to own up to the dirt that they have incurred as part of their efforts to cleanse this same dirt. Such advice would often be career suicide. Still, it would be too quick to then say that there is no compelling reason for politicians to acknowledge the dirt that they have done. By viewing politics as a DC, we can best contextualise the demand for disclosure as one demand among many. Disclosure might, in some scenarios, be the right thing to do politically–suppose someone like Ned or Edna seeks to shine light upon these deals as part of their political project. Still, whether it is the ATC-best thing to do will be determined, at least in part, from considerations internal to the profession.

Thirdly, a DC can, through the actions and dispositions mandated, inflict its own dirt. Paul, for instance, must adopt certain vices as virtues, including hypocrisy and partiality. Understanding these moral costs, and their accumulative and personal nature, requires looking beyond a decision-by-decision perspective. DC provides this framework. Several of the costs and risks associated with being a ‘career politician,’ for instance, might be best understood using this framework.⁵⁰

C. The Limits of Dirty Careers

Why can we not reduce the DH discussion to the choice of whether to embrace a DC or not? To begin, the arguments against overextending the scope of DE theorising work here as well. Due to the normative limits of what a role morality can decently justify and the impossibility of a DC capturing all crises and predicaments that a politician might face, reducing the scope of DH to DC leaves us unable to comprehensively diagnose many instances of ATC-justified dirt in politics. It would thus be mistaken to place categories such as Edna’s entirely within the DC category. To do so would be to commit a mistake commonly found in the DH tradition, whereby the conditions that render dirtying acts seemingly necessary are treated as fixed and immutable without sufficient justification and without exploring alternative solutions.⁵¹

Two other reasons are worth brief description: firstly, there is not a singular DC. A ‘politician’ is a nebulous and tricky thing to define, and the same therefore goes for the corresponding role morality. A British local councillor, an Indian campaign organiser, a Brazilian congressional Deputy, and the US President are all democratic politicians: the demands of their roles are vastly different, however, and this is even before we discuss those in politics-adjacent professions such as public administration and issues advocacy. Even within the same context, different politicians will favour different skillsets and dispositions and thus relate to their political roles in fundamentally divergent ways.

Secondly, restricting our discussion solely to DC makes it harder to morally distinguish between politicians and their political projects and actions. Morally important choices are often taken within the bounds of the profession. It is part of the DC of politics to be okay with handling certain types of value conflict. However, that does not then mean that all choices in those conflicts are equal. Different options will inflict varying levels of (ATC-justifiable) dirt. An ability and willingness to compromise, for instance, is a vital part of a political role morality, yet some individual compromises, such as compromises with particularly evil actors and causes, demand a higher burden of moral proof more associated with DE.⁵² As Applbaum⁵³ argues, the moral licences of such roles, whatever their broader justification, do not legitimise acts that would otherwise be strictly morally prohibited — or, within the context of DH, do not wipe out the dirt and moral taboos that come from such transgressions. Certain acts that are more likely in the political realm still require the same DE justification that they would for anyone. Furthermore, if certain acts become necessary as part of the rules of the game but remain unjustifiable overall, then DR rather than DC will be the more apposite framework.

A. Necessary Conditions of Dirty Rules

To explore the necessary conditions of a DR dilemma, it is worth first defining the ‘dirty rulesets’ that are necessary for such problems to arise:

1. There must be a serious flaw in the ‘rules of the game’ that political agents must abide by. The ‘terms of the competition’ include wrongful behaviour.

2. This flaw is neither necessary nor desirable and should be fixed.

3. Complying with these rulesets inflicts moral dirt upon the agent.

Dirty rulesets can, as well as covering the terms of the competition and explicit constitutional structures, include political ‘settled wisdom,’ ‘third rails’ of politics that cannot be touched, and instances where the placement of the Overton window prevents effective advocacy for just positions. On an institutional design level, DR dilemmas can be solved ‘simply’ by changing these conditions, and such change is feasible and desirable as a political goal. Failing that, a collective action solution where actors refrain from exploiting such structural shortcomings, at least to such a degree that they no longer become necessary to compete, can also dissolve the DR dilemmas. Remove the dirty rulesets, the DR dilemmas go with them.

DR dilemmas, where agents must either abide by their principles or play by these corrupt rules, arise for individual agents, who cannot simply wish away these unwelcome political constraints on their own. Only by inhabiting the system as it currently exists can politicians begin to make their own ethical commitments, including structural reforms, ‘authoritative’ for others.⁵⁵ This is not to deny the role for political creativity and moral imagination, but instead that politicians must start from politics how it is, rather than as it should be. Otherwise, reform efforts will almost always fail and, given the ethical unacceptability of not reforming, this cannot be an acceptable ethical conclusion.⁵⁶

The necessary conditions of DR dilemmas are therefore as follows:

1. Political success requires (ATC-justified) engagement with dirty rulesets.

2. Engaging with these practices is morally dirtying, and the agent should not use these practices as if they were morally benign political practices. There is a moral good associated with acting as if cleaner rules are not in place.

3. To achieve worthwhile political goals, including changing these rulesets, requires political success, and thus engagement with the rules as they currently stand.

B. Where the Dirt Comes From

Unlike DE acts, each individual decision that feeds into engaging with DR practices need not be discretely justified on its own as a distinct lesser evil. In an ideal world, they would require such distinct justification. Dirty rulesets, however, ask politicians to treat these decisions as if they were more akin to being part of a DC, as part of the terms of the competition. The fact that broader society has created these dirty rulesets does not absolve Edna of all resulting moral responsibility for her own contributions.⁵⁷ She has a moral duty to work to change these rules, and shares in the wrongdoing that comes from how these rules operate. Why is this the case? Why can she not escape individual rebuke by subsuming DR within the DC of politics, or placing the moral blame entirely at the institutional level?

There are the specific harms or wrongdoing caused by the specific dirty rulesets, of course. As mentioned, the harms caused by DR arise not solely because the institutional opening exists, but because political actors are using these openings to set unsavoury terms of competition. In Edna’s case, her own discrete contributions are partly constitutive of the overall harms of the system, as well as corrupting to her own sense of integrity. This does not, however, explain what is distinctively dirtying about engaging in DR practices qua DR practices.

To do this, we need to identify the intrinsic good in rising above dirty rulesets and not stooping down to the grubbiest contingent realities of politics. There are several plausible candidates for why politicians have duties to challenge DR institutions by example as well as by reform: a moral responsibility to engage in collective political action to ‘prevent or eliminate injustices’; a natural duty of justice to defend just institutions where they do exist and to create them where absent; a focus on individual democratic rights to self-authorship and self-ownership that are threatened by DR institutions; the need to protect the instrumental benefits of democracy in terms of fairly mediating disputes and ensuring accountability and answerability; and the need not to cultivate dispositions and talents that frustrate good democratic functioning.⁵⁸

As politics is an ongoing institution and a continuous project, one that political agents are striving to help co-author, with ‘many hands’ responsible for many of the flaws,⁵⁹ it is individual behaviour as well as collective action that helps constitute the wrongs of DR. Edna is helping maintain this system, in her own small way, via her participation. Those harmed by the system, such as by losing their right to equal democratic participation, have reason to object to her behaviour too. Because of this, Edna has moral reasons to seek change and to try and mitigate the consequences of her own engagement with DR, ideally through some material commitment to reform and higher standards of behaviour. But she also gets more dirt on her hands than she would if she were merely playing by DC rulesets.

C. Distinguishing DR from Other DH Instances

A key benefit of acknowledging the dirt in DR as real, and as distinct from DC, is that it helps emphasise that these dirty rulesets are not inevitable and that the desire to replace them should be kept alive. By distinguishing DR from DE, we acknowledge the inescapable role that political agents have in perpetuating the ongoing existence of DR. We also avoid tying political agents’ hands too severely by not requiring, as DE does, stringent ‘lesser evil’ justifications for every dirtying action. Edna can thus flexibly pursue the internally justified goals and currencies of politics, while still adopting a proactively critical attitude towards the tools she is using, making sure dirty rulesets are engaged with in a proportionate and reform-minded manner.

By this, I mean that Edna does not treat this campaign finance system as yet another benign part of the political toolkit, as DC analysis would suggest, but nor does she require her every engagement with it be justified as a distinct lesser evil, as DE analysis demands. Picture a minor political issue that Edna cares passionately about–choosing between different municipal park designs, say. It would be acceptable for Edna to use the dispositions and talents that come with her DC as part of her response to this issue. What would be unacceptable is for Edna to use the system of legalised bribery in its full weight to win on this issue.

Furthermore, part of her political calculations should be to consider how she could change this dirtying ruleset. With moral imagination, she can look for ways to change the rulesets: for example, through collective action solutions, promises to use power to obtain legal reforms, and risk-taking with her own political career to shift the Overton window.

Dirty rulesets are prevalent in modern democratic states. Of course, democracies often legitimately require ‘dirtying’ behaviour of the sort that we discussed under the umbrella of DC, and there will always be anomalous decisions that require sui generis episodic treatment. However, DR cases are different. If we want to improve the democracies that we live within, we need to recognise when a dirtying scenario is neither an acceptable part of the democratic way of life nor an aberrant episodic dysfunction. Though a full discussion of a particular context is not possible here for reasons of space, consider the United States of America. In a moderately unjust democracy such as the United States, institutions such as the filibuster, the electoral college, and the redistricting system create dirty rulesets.⁶⁰

These rulesets need not be formal institutions, but instead the political lay of the land. Consider the politics of the nineteenth century, when the still-nascent Republican Party adopted a free-soil platform as opposed to outright abolitionism. Such a move was dirtying, but arguably defensible given the political realities of the antebellum era. Or consider the politics of today, where the Overton window on issues such as immigration and policing reform is further to the right than what many would consider just. Shifting it leftwards will, of course, require making the normative argument in explicit terms, but it will also require getting sympathetic politicians elected, who may have to adopt the more conservative rhetoric in the interim. Framing one’s political projects in a politically palatable way is part of the political DC, but sometimes this framing can be significantly more dirtying.

Take an international example: the logic of nuclear deterrence and mutually assured destruction (MAD) can be seen as a lesser evil compared to what the world might be with lopsided nuclear proliferation or an explosion in the number of conventional wars, but one does not need an absurd amount of optimism to think that more robust peace politics could be attainable. In the meantime, the logic of MAD serves as a lesser-evil dirty ruleset and one that requires DR behaviour. The distinction between DR and DC can help structure debate about disagreements within professions. For example, the laws of war more broadly can be seen as a DR, if one believes that less morally dirtying alternatives are viable (in line with just war revisionists), or as DC, if one believes that they are not.

This is thus how I distinguish DR from what Corvino describes as ‘Structural Dirty Hands’ (SDH), which unlike DR focus on paradoxes of action with lesser-evil justifications.⁶¹ Corvino uses the example of an individual business continuing to participate in the evils of sweatshop labour when the alternative (closing sweatshops and withdrawing from that market altogether) is greater destitution for the erstwhile workers and leaving the market entirely to unscrupulous actors. Unlike traditional DE, extricating oneself from an instance of SDH requires collective action.

Unlike DR, however, there is no presumption that there is a ‘permissible’ right to stay in. Extrication is the prima facie moral demand in SDH cases, only alleviated by how continued participation serves as the lesser evil.⁶² There is no moral right to benefit financially from sweatshop labour, and it is usually not a precursor to pursuing other admirable political ends. DR cases do not have this ethical obligation to quit and do not require lesser evil justifications. Fleeing the political (or geopolitical) sphere when such serious issues are at stake is, if not impossible, hardly a serious suggestion for every well-intentioned aspiring political actor.

Both DR and DC represent the self-justified rules of the game and the terms of the competition. However, while DC covers the desirable or core features of the political role morality, DR covers the features that can and should be changed. Whether an institution is remediable or not, and therefore counts as DC or DR, can be determined by the costs of changing the institution. Certain factors that concern Paul, for instance, could in theory be changed, but doing so would require changes to how his democratic fellow citizens behave. Enforcing such changes is highly likely to have high moral costs. Abolishing the political need for duplicity might be theoretically possible, but not without enforcing changes to human nature that leave it morally prohibited. Therefore, Paul remains in a DC.

Some theorists, such as Coady, are optimistic about the ability to craft rules and institutions in such a way as to all but eliminate DH scenarios. Others, such as Hollis, would regard such an endeavour as foolish; many political realists would similarly view an attempt to deviate from the self-justified demands of political morality as a recipe for disaster. The truth, I suggest, is likely somewhere in between. Some DH cases can be prevented. Others cannot. By distinguishing between DR, DC, and DE, we can begin facilitating that necessary further analysis.

Furthermore, eliminating DR practices may not be the most pressing priority in any given context. A political machine that is rife with corruption will, ideally, be eliminated: if said machine is the only avenue through which marginalised citizens can achieve political goods, then zealously pursuing its elimination might be premature at best. It is still helpful, however, to designate such practices as DR, even if their continued presence is ATC-justified the best option. In these cases, a DR dilemma can arise at the institutional level about whether to allow such norms to persist, as well as the usual DR dilemma at the individual level about how to engage in such rulesets.

Determining whether a particular ruleset falls within DC or DR can also be a subject of heated political debate: take the choice between a proportional representation voting system and a first-past-the-post system, for instance. Both systems will incentivise different virtues and dispositions. Still, democracies must pick a system. The act of making this choice all but assures that some democratic agents will have to learn how to prosper under systems that they view as containing remediable structural flaws. Given the importance of integrity to political actors, there is thus a degree of agent-relativity in terms of whether something constitutes a DC or a DR. By acknowledging the system’s contingent nature, however, political actors can better structure, in accordance with their own values and normative commitments, how they want to grapple with these decisions.

Abolishing DR in their entirety will, even putting aside the potential role of agent-relativity, be impossible. Take the issue of state capture, the use of state power to advance private interests at the expense of broader national concerns. One can argue, as Bagg does, that no state can fully eliminate the potential for such capture to arise. The risk of capture, then, seems to be part of the DC of politics. However, though we cannot ‘eliminate’ the prospect of capture, we can still make judgements about ‘relative degrees’ of capture and treat it as the target of a ‘constantly evolving struggle’:⁶³ the issue therefore merits more of a DR treatment. Analysis of issues such as state capture is enhanced by being able to distinguish between DR and DC, allowing us to identify the risks that are endemic to politics (DC) and those that can, if not now then at least in a plausible future, be challenged (DR).