II. Individual and Institutional Accounts of Corruption

Corruption has historically been seen as a pathology both of individuals and of institutions. At the individual level, corruption occurs when an individual abuses for her own gain the power and authority she possesses as the holder of a position within either a public or a private institution. Now, in a sense, all office-holders can be seen as using their position for their own gain, if only in that they typically draw salaries as compensation for the tasks they perform within the institution in question. To be more precise, let’s say that individual corruption occurs when an individual office-holder uses her position in order to derive a benefit from holding the position that does not fall within the terms that were agreed to when she took up the position.

A focus on the individual in an account of corruption has a number of evident virtues. First, and most obviously, individual behavior is observable and measurable. In trying to determine the extent of the corruption prevalent in any given setting, being able to focus on individual actions will obviously lower the epistemic hurdles (though, of course, it will not make them disappear altogether). Attempting to identify more subterranean causes of corruption—cases in which individuals are vehicles for corruption, rather than originators thereof—is comparatively much more difficult.

Second, there may be ethical reasons for holding individuals ultimately responsible for corruption. Whatever the institutional pressures that might militate for individuals to act in corrupt ways, it is, some might claim, up to individuals to withstand those pressures and to act with integrity. “Tout comprendre, c’est tout pardoner” (“To understand everything is to forgive everything”), according to the French saying, and the individualistic account may be seen as having the advantage of not providing individual agents with a way of shunting responsibility for corruption onto nebulous underlying institutions and sets of rules. The individualist account allows us to express and to preserve something that is clearly central to any intuitively plausible account of corruption, namely, that it involves behavior on the part of individuals that is in some sense or other undesirable. As an office-holder within either a public or a private institution, an individual has responsibilities toward some specified set of agents built into her very role. When she acts in a way that violates the terms of that relationship for the sake of her own interests, she is clearly at fault for violating the terms of a fiduciary relationship.

Despite the apparent plausibility of individualist accounts, I join a number of other theorists in holding that corruption should be thought of institutionally.^{Footnote 4} Though individual behavior is, to be sure, easier to observe and to evaluate than are deeper institutional forces, by focusing on individual behavior we risk acting like the protagonist of the famous joke, looking for his keys under a streetlight, not because he has dropped them there, but because there is more light there. Specifically, a focus on the individual runs the risk of missing important features of corruption that lie both upstream and downstream from the individual’s corrupt acts. First, to focus on the upstream dimensions, though corruption can originate in the moral turpitude of an individual office-older, it can also have institutional causes. Something about the manner in which institutions are organized can in certain cases lead individuals to act corruptly, even though they are not antecedently disposed in virtue of some character flaw so to act. Second, the corruption of individuals can (though it need not) have institutional consequences. There are circumstances in which individual corruption can be institutionally inert (and in some sets of circumstances even functional) as far as the institution within which it occurs is concerned. My claim here is that political philosophers ought to be particularly concerned with the deleterious impact that individual corruption (whether rooted in individual vice or in institutional design) has on institutions.

It is evident that individual corruption need not yield the corruption of institutions from the not-so-fanciful cases in which benefit accrues to an individual without the distortion of institutional purposes. For example, one can imagine cultures in which figures of authority are lavished with gifts by those over whom they hold sway, but in which neither the members of the public nor indeed the authority figure herself, expects that such gifts are intended to alter, or have the effect of altering, the judgment of the office-holder regarding the manner in which she will confer the benefits (or the liabilities) that she has the responsibility to distribute. One can imagine cases in which all office-holders expect to be bribed, and in which those who depend on them for the carrying out of some official function expect to have to bribe, but in which no one would agree to hold office but for the possibility of enrichment.^{Footnote 5} One can imagine, finally, cases in which office-holders within very unjust states use the power with which they have been entrusted to realize norms of justice that they endorse, and that happen to be superior, from the point of view of justice, to those that they have been entrusted to uphold.

In all of these cases, although office-holders use their offices for personal gain, the principals for whom they are the agents do not suffer any relevant harm from office-holders benefiting from their positions in this manner. In the first case, we are assuming that the gifts in question have no causal role on the manner in which office-holders acquit themselves of their responsibilities. In the second case, principals are actually made better off by accepting a practice of bribes than they would be were rules against such a practice rigorously enforced. And in the third case, although the principals whose interests should have been served by the enforcement by an office-holder are not well served, one could imagine the norms that the office-holder chooses to enforce are more just, in that they serve groups who are unjustly discriminated against by the “official” norms structuring the official role.

Thus, the presence of personal benefit being derived by office-holders is not a sufficient condition for the presence of corruption of the kind that should concern political philosophers (as opposed, say, to ethicists), namely, corruption of a kind that deleteriously affects the functioning of institutions that serve some important public purpose.

Thus, an individualistic focus on corruption can blind us to the institutional consequences of individual corruption, and can lead us to assimilate cases in which individual malfeasance has an identifiable impact upon the functioning of institutions to cases in which it does not. As briefly noted above, it can also blind us to situations in which the incentives to individual corruption (whether it has an institutional impact or not) emanate not from the individual’s weakness of character, but from institutional sources. As we shall see, treating all cases of corruption as if they were the result of individual traits of character rather than of institutional causes blinds us to possibilities for institutional reform that will make corruption less likely.

The next section will be devoted to distinguishing between two different causal patterns that can lead to institutional causes giving rise to individual corruption. But before I turn to that distinction, I want to forestall a concern that my account might raise. As we shall see, the account requires that we broaden the extension of the concept of “individual benefit” well beyond the manner in which it is standardly construed in most accounts of corruption. What might be termed “folk” accounts of corruption standardly view the benefits that individual office-holders derive from the abuse of the prerogatives of their offices as financial. A (morally) corrupt public official accepts a bribe, and in return he grants a lucrative public contract to a suboptimal bid. A physician in a public health system accepts payoffs in order to manipulate the queue for some medical procedure. Thought of corruption elicits images of brown envelopes stuffed with cash being handed off in dark alleys, or of illicit wire transfers. We might at the limit extend our standard conception of corruption to include the office-holder requiring sexual favors in order to deliver a service related to his office.

In what follows I will argue that there are reasons to extend the scope of the concept of “benefit” that we include in our account of corruption. People can benefit from their positions in ways that risk diverting an institution from its purpose in ways that have nothing to do with money or sex. As I will suggest below, they can benefit in ways that endanger the institutions to which they belong by ascribing disproportionate and dysfunctional weight to winning (in legal or business settings) or to the identity-related benefits that stem from their political party thriving. The extension of the concept will give rise in the reader to the concern that in extending the concept of corruption so that it encompasses not just individual moral turpitude but also nonmaterial incentives to act in ways that end up diverting the institutions to which one belongs from its purpose, and by focusing on institutional dysfunction rather than on individual malfeasance, we have not so much extended the concept of corruption as replaced it.

In an important recent article, P. Y. Néron has spelled out a number of challenges that institutional accounts must face. According to Néron,^{Footnote 6} institutional accounts risk hitting up against three interrelated flaws. The first is one of scope: institutional accounts risk labeling all institutional dysfunction as corruption, regardless of the particular etiology of the dysfunction. Second, in so doing, they risk bypassing the most perspicuous explanations of dysfunction, and correspondingly, the most effective routes for the correction of a dysfunction. For example, the failure of the market to internalize the externalities caused by polluting firms is perhaps better thought of as one of market failure than of corruption. Third, and perhaps most importantly for present purposes, Néron points out that any account of corruption risks “essentializing” the institutions which are deemed to have been corrupted, say, by a change of norms. Institutionalist theories of corruption cannot, in other words, claim to be purely descriptive. They must, either explicitly or implicitly, bring a normative theory about the institution’s proper purpose and functioning to bear upon their assessment that it has been designed or constructed in a “corrupt” manner. Such accounts risk being justly perceived as question-begging by those who hold another view about the ends that it is appropriate to view an institution as serving, and about the norms that it is accordingly appropriate to view the functioning of the institution as constrained by. The latter concern is affirmed by Mark Philp, who in the specific context of political corruption, writes that “definitional disputes about political corruption are linked directly to arguments about the nature of the healthy or normal condition of politics.”^{Footnote 7}

Néron’s challenge is, I believe, worth taking up, and I will view my account as subject to its strictures. The functions on the basis of which institutions are said to be corrupt must be argued for rather than simply assumed. As later on in this essay I will be concerned principally with the corruption of democratic institutions, the challenge will be to come up with a general characterization of the purpose of democratic institutions that is sufficiently broad as to be acceptable to a wide range of kinds of democratic theory, without being empty in virtue of being so broadly acceptable. The challenge will also be to be able to characterize corruption as a particular kind of institutional decay, rather than as a category that encompasses all manner of institutional dysfunction.

In order to better understand the specific form of institutional corruption that I want to bring into focus in this essay, I need, however, to make and discuss a set of distinctions that will allow us to distinguish between different kinds of corruption.

III. Three Kinds of Institutional Corruption

There are a number of ways in which institutions themselves can be sources of corruption. We have already briefly referred to two of these, but in the present section of this essay I would like to go into more detail about the way in which institutional corruption can arise in order to distinguish these two kinds of cases from the one that I want to bring into focus.

Let us begin with an informal definition of what institutions are. Institutions, at a first approximation, are sets of rules through which benefits and harms are allocated. These rules are defined on the basis of some sense of how it is appropriate for these benefits and harms to be distributed. Rules have no justification independent of the values and norms on the basis of which an institution is designed in the first place. One way in which corruption can take hold in an institution is when rules are intentionally introduced into the functioning of the institution that foreseeably divert it from those values and norms.

An example of such deliberate institutional corruption is that of the gerrymandering of the electoral map designed to decrease the electoral impact of a group whose electoral behavior is deemed to be deleterious to the electoral prospects of a powerful group that is institutionally situated so as to exercise disproportionate weight on the manner in which electoral boundaries are set up. In establishing that gerrymandering of this kind constitutes institutional corruption, one should be mindful of the concerns voiced by Néron in the article referred to above. One must avoid begging the question by not assuming that some highly specific set of values and norms is criterial for the manner in which electoral districts should be set up. Thus, for example, someone who defined corruption of the electoral system as any intentional departure, say, from the criterion of strict populational proportionality, would be begging the question against views according to which criteria other than proportionality, say regional parity, ought to be integrated into the design of an electoral system. Racial gerrymandering—the redrawing of electoral boundaries in ways that cannot be plausibly explained in any other way than by the desire to water down the electoral impact of specific racial groups—offends against any plausible such set of criteria. Whatever the right mix of norms and values to embody in an electoral system, any reasonable proposal will exclude the deliberate electoral marginalization of specific groups.

Compare deliberate gerrymandering of the kind just described to the effects that might result from electoral redistricting aimed at the achievement of defensible criteria. In Canada, for example, rural ridings are comprised of far fewer voters than urban ones. Although it is a departure from the principle of strict proportionality, the disproportion reflects the importance of balancing the electoral representation of individuals with the representation of regions. The effect of this balancing is to reduce the weight of the vote of ethno-cultural minorities, who tend to concentrate in urban areas.^{Footnote 8} Now, it is far more difficult, on a conception of corruption duly constrained by Néron’s concerns, to argue that the electoral system has been corrupted in the Canadian case, to the extent that the electoral map can be seen as reflecting a mix of reasonable normative concerns, even though it has the regrettable consequence of diluting the vote of ethno-cultural minorities,^{Footnote 9} than it is to do so in the case of an electoral map that cannot plausibly be interpreted in any other way than as reflecting the wish on the part of those who control the process of redistricting to reduce the weight of such votes.

It is clear, moreover, that the example of electoral gerrymandering preserves the focus on individuals as vehicles, though not as originators, of corruption. In a gerrymandered system, the incentive of candidates to win, though in and of itself blameless, ends up serving the purposes of those who have manipulated the electoral system so as to serve the purposes of racists, rather than the public good.

Thus, a first way in which institutions can become sources of corruption has to do with cases in which the rules on the basis of which these institutions operate lead them to derogate from the values and norms on the basis of which they ought to operate. In order to make the condition embodied in the last clause as non-question-begging as possible, the normative grounds on which such assessments are made must be as inclusive as possible, and must not exclude any reasonable construals of the values and norms that can be taken to guide and constrain an institution.

Institutions can also be sources of corruption when they generate perverse or conflicting sets of incentives.^{Footnote 10} Examples abound in this connection. Think of schemes that are in and of themselves blameless, such as those designed to improve the accountability of educational institutions by linking funding and pay increases to educational achievement. On the face of it, such schemes are in line with the values and norms that on any plausible account should be central to the education of children, namely, to ensure that all children receive quality education, and that the interests of all children are counted equally in the attempt to achieve this end. Now, as has been widely reported, schemes such as this, when they are designed without sufficient attention to the foreseeable consequences of at least some modes of implementation, risk undercutting the very laudable goals that they were meant to serve. For example, educational institutions can be led to lowering academic standards in order artificially to increase reported rates of successful completion. High-quality teachers can be disincentivized from taking on the most demanding pedagogical challenges, challenges that they are ex hypothesi best qualified to meet, for fear of having their salaries affected by the predictably lower rate of success that such classrooms might give rise to.^{Footnote 11} Similar sets of perverse incentives that lead institutions to be diverted from the goals that they have been designed to serve can also be detected in health care systems in which quantitative indicators are used to evaluate the effectiveness of physicians, or of health care systems as a whole. The lowering of standards necessary in order to appear to meet quantitative objectives, putatively put in place in order to increase the amount of health that a health-care system produces, ends up giving rise to less health.^{Footnote 12}

Perverse incentives can also be generated when the discrete institutional goals that are written into bureaucratic structures meant to be coordinated within a complex institution come into conflict. Anyone who has worked within a modern university has at some point or another been faced with a “tail wagging the dog” logic in which bureaucratic and administrative logics that are meant to subserve academic ends end up subordinating academic ends to their own, for example by imposing financial reporting requirements so onerous as to disincentivize the pursuit of ambitious funded research projects.

Again, and cleaving to the worries about institutional accounts of corruption put forward by Néron, speaking of corruption in these cases does not involve evaluating institutions on the basis of controversial criteria. Rather, it involves noting the ways in which, for instance, setting up structures and procedures for the measurement of an institution in meeting its own objectives ends up detracting from the attainment of those very objectives. The critique is internal, grounded in the goals and values that constitute the DNA of the institution in question itself. Does speaking of corruption in these cases succumb to the temptation, also warned against by Néron, to express all issues of institutional dysfunction in terms of corruption? Again, I think that this account of the sources of institutional corruption leaves room for other forms of dysfunction including (in the economic realm) market failure, or inefficiency, or what Seumas Miller calls “corrosion”: “Acts of institutional damage that are not performed by a corruptor and also do not corrupt persons are better characterized as acts of institutional corrosion.”^{Footnote 13} He provides as an example the underfunding of an institution such as the court system that can lead over time to the diminution of the quality of adjudication due to, among other things, the lesser resources for the training of magistrates. An account that includes the corruption of institutional purposes due to the perverse incentives that can sometimes be generated by the introduction of institutional mechanisms aimed to abet the achievement of an institution’s goals, but which ends up subverting them can still make room for other forms of institutional dysfunction, and does not run the risk of claiming that all institutional failures are instances of corruption.

I want now to focus on a third source of institutional corruption, which will be the focus of the rest of this essay. To begin to distinguish it from the sources of institutional corruption we have been discussing thus far, let me note the following feature of the kinds of mechanisms that lead to institutions becoming corrupted by perverse incentives. In these cases, the mechanisms in question are largely extrinsic to the goal pursued by the institution. That is, although measurement—along with the “measurement traps” that schemes that have not been well thought out can generate—is a natural concomitant of any goal-oriented institution (in that it is quite natural for agents both within and outside stakeholders of the institution to want to measure the success of the institution in meetings its goals), particular modes of measurement are not strictly called for either by these goals, or indeed by the desire to measure the degree to which they have been met. Measurement techniques that incorporate qualitative as well as quantitative indicators are at first glance equally well-suited to achieving the objectives of measurement. What’s more, they are likely to avoid the production of the kinds of perverse incentives to which I alluded briefly above. Thus, the features of the institutions we have been considering that risk corrupting them are extrinsic to the purpose of the institution itself.

Compare this kind of case to the way in which classical political thinkers envisaged the corruption of political regimes.^{Footnote 14} For Aristotle, for example, each regime had its particular form of corruption, which emanated from a potentiality that was already present even in the ideal form of the regime in question. Thus, as is well known, government by the few can degenerate into oligarchy, rule by the one can give rise to tyranny, and a polity ruled by the many can degenerate into mob rule. These processes of corruption are generated by forces and tendencies that are already present in each form of regime. Since such corruption is intrinsic to the regime in question, rather than resulting from extrinsic mechanisms as in the cases discussed above, it is much more difficult to build bulwarks into the functioning of the corresponding institutions in order to fend it off. That is why classical writers such as Aristotle tended to see the corruption of regime types as inevitable. On his view, it was just a matter of time before the institutions associated with a regime type succumbed to their internal destructive logics.

The classical account of internal forces associated with each regime form leading to its corruption has come to seem antiquated, though there are hints of a resurgence of classical-style thinking in trying to figure out how populism is related to democracy.^{Footnote 15} Part of the reason for this is that political theory has not really taken seriously that there might be noncorrupt versions of the regime types that classical authors still took seriously, such as aristocracy and monarchy. But I would argue, and this will be the burden of the last part of this essay, that there is a tendency inherent in democracy that can give rise to a distinctive kind of corruption that has so far been understudied, or at the very least not studied as a form of corruption. And by this I mean the fact that democracy belongs to the set of what I will refer to as “adversarial systems.”

IV. The Problem of Corruption in Adversarial Systems

We are so accustomed to the central economic, legal, and political institutions of many modern societies that we sometimes fail to notice that they are commonly built around analogous principles of institutional design. They seek to achieve a social good through constrained competition.^{Footnote 16} The market, at least according to the free market theories that underpin the defense of market institutions, makes everyone benefit from the self-serving motivations of those who bring goods to market in order to make a profit. The competition of multiple such individuals and firms moves each market competitor to try to produce better products at a better price, and through the operation of the hidden hand, everyone is made better-off.

Many Western countries, including Canada and the United States, also organize their legal systems on an adversarial basis. Rather than having a “juge d’inquisition” whose main function is to ascertain the truth, legal procedures in adversarial settings pit opposing parties against one another, private parties in civil matters, and accused individuals against the state in criminal cases. Here again, the hypothesis at the basis of the adversarial procedure is that the relevant rights are best protected—and thus the result of a legal process best reflects what justice requires—when each party is zealously defended by counsel, rather than when one legal agent sets out to discover “the truth.”

Democracy is also an adversarial system. As a matter of fact, it is doubly adversarial. Representatives are selected through competitive elections, and the legislative process also pits a governing party against what in Westminster parliamentarianism is referred to as the “loyal opposition.”^{Footnote 17}

The common feature of these three systems is that they all purport to pursue some public good (in the nontechnical sense of the term), and to do so in a manner that is impartial relative to all contending parties and interests, despite the fact that none of the central actors within the system directly aims at that public good. The law, democratic politics, and the market are all geared toward some public good as a “system effect,” one that is best achieved when each protagonist within the system acts on the basis of a competitive “role morality.”

As we have seen, in some cases, the potential for corruption comes from the malfeasance of individual agents, who are impelled by motivations for private gain that are dysfunctional from the point of view of the institution’s ability to achieve its end. In other cases, institutional corruption emanates from norms and rules themselves, either because they have been inscribed in institutions by agents who manage to instrumentalize the institution in question to their (partial) ends, or from the perverse incentives that might be generated by imprudent but morally faultless institutional design.

The particularity of the challenge posed by adversarial systems is that the very partial and self-serving motivations that are required in order to make an adversarial institution operate the way that it should in order to achieve its ends are also the motivations that, if unchecked, risk corrupting the institution and diverting it from these ends. The motivation to achieve profit is not self-limiting. It does not spontaneously adhere to the conditions that must be in place in order for the market to serve its public purpose. On the contrary, the motivation to extract maximum profit in such settings comports quite well with information asymmetries, with the formation of monopolies and cartels, and the like. Similarly, lawyers who have been trained that winning in the competitive settings that trials represent do not spontaneously observe legal rules of discovery—rules about the sharing of evidence with opposing counsel—or the myriad other procedural constraints through which the competition between parties is maintained on a more or less even playing field. Democratic partisans, or so I shall suggest, are also prone to unchecked partisanship that might upset the adversarial logic through which democratic institutions are thought to best pursue the public good.

In Ethics for Adversaries, which is still pretty much the only attempt at dealing with the ethical challenges posed in adversarial systems, Arthur Applbuam argues that despite the fact that they are pitted as adversaries within the institutional settings in which they take part, lawyers (who are central to Applbaum’s account) should not diverge to an excessive degree from ordinary morality. His is an account that minimizes the degree to which these kinds of institutional actors should be guided by specific “role moralities.”^{Footnote 18}

I disagree. The functioning of competitive institutions is premised upon partiality and partisanship. It is because each lawyer is zealously committed to the victory of the party that she represents that she will be motivated to uncover as much evidence as she can that supports her client’s case. If each lawyer in a legal matter does this, then judge or jury will be able to come to an understanding of “what really happened” on the basis of the best available evidence, full stop. The urge to constantly increase market share, shared by all competitors in the marketplace is similarly the motivation through which consumers benefit from the best goods at the best prices. Better to constrain and channel the self-serving motivations of institutional agents who are moved by the desire to win (rather than by the wish to do good, or to achieve justice) than to select and to socialize agents in such contexts into an ordinary morality, one that does not take into account the particularities of the institutional environments in which adversaries operate, and the degree to which the attainment by the institutions in question of their constitutive goals depends upon their comporting themselves as adversaries. Adversarial institutional schemes truly do require institutional design to “economize on virtue,” and adversarial character traits to be channeled in a way that avoids their giving rise to institutional corruption. Thus, these schemes have rules against collusion and the exploitation of information asymmetries in market institutions, and rules about the disclosure of evidence and the introduction of evidence in trial, rather than institutional agents who are spontaneously moved to self-limit the tactical and strategic advantage they can have over adversaries. Needless to say, the capacity of such safeguards to contain and channel self-seeking behavior is one that requires constant vigilance on the part of regulators, as competitors seek to exploit loopholes and blind spots in regulatory frameworks.^{Footnote 19} I shall suggest below that the same is true of the taming of partisan passions.

V. The Adversarial Threat to Democracy

I want to argue that democratic systems are vulnerable to this third kind of corruption. That is, they are prone to the kind of institutional corruption that arises when incentives to placing individual benefit in the exercise of one’s institutional functions are built into the very logic of the institutional design of democracy, rather than being susceptible to institutional corruption due to the injection of a perverse incentive that is external to the incentives normally generated by the system itself (as in the second model of institutional corruption briefly described above).

Representative democracies are, as was briefly stated above, doubly adversarial. First, representatives in legislative bodies are selected through competitive elections that are contested by political parties, part of whose function is to promote the electoral fortunes of party candidates. In a metaphor that is quite telling of the way in which elections are contested, parties set up “war rooms” in the run-up to elections.

Second, representative institutions in modern democracies are themselves set up in an adversarial manner. Representatives do not lose their party identities when they are elected to the legislature, and when they take up official functions therein. The very architecture of most parliaments speaks to the degree to which the functioning of legislatures continues the adversarial logic set in motion by elections. Members of parties typically sit together. The governing party or coalition typically sit across from one another. In Westminster parliamentary systems, for example, there are myriad ways in which the adversarial nature of Parliament is dramatized. Elected officials are not supposed to “cross the aisle,” for example, as newly minted Canadian Prime Minister Justin Trudeau found out when he was criticized for crossing the aisle to the opposition side in order to get members from other parties to return to their seats and consequently prevent the delay of a vote that was scheduled to be held in the House of Commons.

More substantively, the business of government is built around an adversarial logic in which an opposition party is tasked with the responsibility of critiquing the governing party’s legislative agenda (even when they do not have a sufficient number of members in parliament to prevent the government from securing a majority of votes). Opposition parties are expected to oppose the government, at least in the initial phases of the presentation of proposed legislation. For all the talk of “bipartisanship” and “reaching across the aisle” that has become fashionable in the criticism of modern democratic systems, a “loyal opposition” that was too pliant to the will of the governing party risks being seen as not doing its job.^{Footnote 20}

Though political parties have been derided as vehicles through which the competitive logic is institutionalized by political thinkers as different as James Madison and Jean-Jacques Rousseau, they have established themselves as essential features of democratic systems. Many states that are in effect single party states pay lip service to the trappings of the party system by tolerating toothless opposition parties. At the level of theory if not always of practice, democracy in the world today just is multi-party democracy.

The rise of parties in modern democracies can undoubtedly be explained in a variety of normatively uninteresting ways. For example, there are no doubt organizational gains for political candidates in being able to draw upon the resources that parties put at their disposal. Candidates whose ideological differences are not great have a considerable incentive to put those differences to the side in order to affiliate themselves with a political party whose financial and organizational weight is greater than an individual candidate’s could ever be.

But there are also a number of normative arguments that favor political parties, and the particular kind of adversarial system that political parties, as opposed to non-aligned individual candidates and representatives set in motion, as desirable institutional realizations of the democratic ideal.^{Footnote 21}

To get a fix on the important role that political parties can play in modern democracies, consider the complex roles that governments must perform. They must implement policies in a wide range of policy domains, where decisions in one domain often impact decisions in others; this is (i) in part, for budgetary reasons, (ii) because each policy decision involves opportunity costs—that is, some policies could have been implemented in another area but for the decision to implement the policy in question, and also (iii) because the coherence of an overall policy agenda simply makes it the case that decisions in some areas call for policies in another. For example, the decision to reduce greenhouse emissions has logical consequences in the area of urban planning, where denser settlement patterns and a privileging of public transport infrastructure over roads made for private vehicles are natural corollaries of the aforementioned environmental policy.

Another way to make this point is that governments enact platforms, which are integrated sets of policies with respect to the wide range of policy areas concerning which governments are expected to enact policy.^{Footnote 22} Such platforms are integrated at least in the two senses noted above, namely that they reflect the trade-offs which policy choices sometimes require, and attempt to avoid different policy domains being thought of in policy “silos,” that is, without regard for the way in which decisions made in one area have implications for the way in which policies in other areas should be articulated. They are integrated in a third way as well, in that the political parties which, as we will see in a moment, are best situated to produce platforms, articulate their policies, as well as the trade-offs that they often require, on the basis of a number of overarching values that distinguish one political formation from another, rather than in an ad hoc manner.

Platforms are thus epistemically complex objects. Part of the justification for political parties is that they are incubators for policy platforms. Robert Goodin has shown that a democracy in which elections were contested only by individual, non-aligned candidates would be unlikely to generate platforms. Individual candidates would be unlikely to have policy proposals across the full range of policy domains, and even if they did the election to a legislature of a multitude of representatives who all had their own policy platforms would give rise to a cacophony.^{Footnote 23} Political parties, by bringing together people with different interests and priorities, but whose commitment to the same range of overarching values incline them to hash out their differences within a common platform rather than facing the electorate separately, provide the raw material through which platforms are wrought.

Political parties perform an important epistemic function by providing a setting within which platforms through which a party would, if elected, govern, but they also perform an important epistemic function vis à vis the electorate. By organizing the potentially infinite set of policy proposals and combinations thereof into a manageable number of such sets, they make the task of the electorate less epistemically intractable than it might otherwise be. What’s more, electoral competition motivates political campaigns to present their platforms in the best possible light, and to critique the platforms of other parties, a process that leads each party to have to respond to the criticisms that have been leveled at them. Electoral competition among political parties thus serves an epistemic function analogous to that fulfilled by legal adversaries in a trial. Each side is motivated to win, but in putting forward an argument for their case that is tailored with a view to winning, they provide judges and juries with better information with which to arrive at a correct legal judgment than if they were operating on the basis of information and arguments provided outside the context of the adversarial context.^{Footnote 24}

The epistemic function of parties is connected to a moral function. One of the reasons that political parties have been seen with suspicion by political theorists as different as Madison and Rousseau has had to do with their fear of faction, the fear that political parties would reflect and harden existing social fissures, and perhaps even create some new ones. Political parties have been seen by some as symptoms of a hopelessly rifted society. As Jonathan White and Lea Ypi have argued, however, political parties can also be the institutional mechanisms through which faction is transcended. By being forced to bring their case to the general electorate, political parties that may very well have originated in particular sections of society (be they class-based, regional, or whatever) need to present their favored sets of policies as ones that are good for society in general. Though the presentation of what at the outset may have been a set of policies designed with the good of a particular segment of society in mind as generally good may originate in self-serving motivations, the fact of having to present a set of policy proposals as proposals for the general good ultimately has a transformative effect.

If this thumbnail account is plausible, then political parties help democratic systems achieve their purposes, which is to allow for some account of the general good to be operative in the processes of government. It is therefore not just a fact, but also, all things equal, a normatively desirable fact that political parties have come to occupy the central role they have in most actually existing democracies.

Political parties, however, give rise to partisanship.^{Footnote 25} By partisanship I refer to a range of incentives and motivations that lead party members to act on behalf of the political party. At the most obvious and superficial level, partisans want to win because of the benefits that will accrue to them if they do. Candidates want to win because if they do they will be members of government, perhaps even members of cabinet, rather than being stuck on the opposition benches. Party operatives want to win because if they do they will gain the status, prerogatives and powers that come from holding offices within a governing administration.

But at a deeper level, party members identify with their parties. In societies with strong political party traditions, parties are not just disposable instruments that serve the purposes of ambitious individuals. Rather, they are central to the identities of those who work within them, and for many members, parties are an important source of community. Winning elections for the party member is not just a path toward the attainment of instrumental goods of the kinds that were briefly mentioned above. It is at least to some degree a way in which the self and its core commitments are affirmed. Conversely, losing, for the partisan, is felt as a devalorization of self. The deep expressions of emotion that are seen around the world on election nights by both winners and losers reflect the fact that partisans take winning and losing personally.

The set of incentives and dispositions of character that make up partisanship are essential to democratic systems. They provide party members with the motivation to work for the victory of their party. In the same way that the martial traits of character that tend to be selected for among trial lawyers conduce to the good of the adversarial legal system by motivating them to present the best case they can so as to ensure the victory of their client or of their cause, partisanship motivates the partisan to put party ahead of self in working tirelessly to present to the electorate a vision of the common good that is as attractive as possible. When partisans of different parties are motivated in the same way, the electorate can make decisions on the basis of their conception of the result of a debate within which each side has presented the best possible picture of themselves and of the kind of society that they would promote if they managed to form the government.

However, the kind of corruption that is a permanent risk of multi-party democracies results from the very set of motivations, incentives, and traits of character that constitute partisanship and that, as we have seen, are an essential motivational condition for the party system performing the epistemic and moral functions that we associate with it. In the same way the martial traits of character arguably possessed by the modern trial lawyer can cease to serve the cause of truth and justice when these traits are unconstrained, so the traits that make up the moral psychology of the partisan can also, if unchecked, divert the democratic system from the purpose that political parties and the partisans that staff them are in certain sets of circumstances well equipped to promote.

Consider two ways in which partisanship can go awry. The first characterizes the behavior of the partisan during election campaigns. The claim made on behalf of parties and of partisans is that they provide important settings for crafting proposals concerning alternative visions of government, and essential vehicles through which platforms are delivered for consideration and debate to the electorate. The assumption here is that the desire to win will lead parties and partisans to draft platforms that are as attractive and compelling as they possibly can, to present the strongest possible arguments for them, and to subject the alternative proposals of other parties to the kind of scrutiny that will require that these arguments be refined in order to withstand critique.

But the desire to win—to promote the interests of party—can (and in the real world of electoral politics often does) come uncoupled from this epistemic function. If election rules are set up in such a way as to allow political parties to win elections in other ways than by presenting attractive policy platforms to the electorate and by loyally but vigorously critiquing the proposals of opponents, then partisanship can become a corrupting force. It can lead to a form of individual corruption, whereby members of political parties confer benefits upon themselves through their actions as party members and officials (electoral victory and the instrumental and identity-based goods that flow from it). And it can lead to institutional corruption, in that the interests of party can come to subordinate the general good, which engagement by parties in rational contestation and defense of platforms can otherwise reliably track.

Consider a second way in which partisanship can lead to corruption; this way concerns the behavior of partisans, be they elected officials or party operatives, after elections have been held, in the everyday operation of legislative bodies. Putting party first—for example, refusing to consider policy proposals put forward by members of opposition parties simply because they originate with the opposition, or refusing to consider compromises with opposition parties that might give rise to better policies—can in the post-election setting be seen both as an example of individual corruption, and as a source of institutional corruption. At the individual level, putting party first in the ways just briefly described can confer upon the individual the benefit of not having to compromise his identity, of being able to indulge in a form of political purity which, though it may make the partisan feel good about himself, may very well be dysfunctional from the point of view of enacting policy that serves the public good. At the institutional level, moreover, this form of unconstrained partisanship diverts the institutions of democracy from what should quite uncontroversially be seen as their function: namely, to give rise to government that tracks the public good.

This kind of corruption quite plainly satisfies Néron’s test. It is compatible with an inclusive conception of what the purposes of democracy are, which I have characterized here quite broadly as government in the service of the public good. It does not lead to a conception of democratic corruption that is so broad as to exclude other ways in which democratic dysfunction can occur. Rather, it focuses quite narrowly on one source of dysfunction, namely, the operation of the unconstrained motivations, incentives, and traits of character associated with partisanship. Moreover, it corresponds to the second, and arguably the more intractable way in which corruption can emanate not from individual moral turpitude, but from the sets of incentives generated by an “internal” feature of the institutional setting within which office-holders find themselves, namely, that of political parties. The challenge of the kind of corruption we have identified here is that it stems not from a feature of the set of institutions we have been examining that can be easily distinguished from the institutions in question (like the measurement of quantitative indicators can be separated from the education system), but from a feature that on the contrary is quite central to it.

How can the potential for partisanship to corrupt democracy be limited without sacrificing the very great benefits that it affords modern democratic systems?

There are two kinds of strategy that might be adopted. The first identifies excessive zeal of the individual partisan as the source of the problem of democratic corruption, and would envisage ways in which to educate and socialize partisans so that they exhibit the virtues of partisanship without lapsing into its vices. Although I do not want to exclude such educative strategies outright, I am skeptical about them, for generally Madisonian reasons having to do with the imperative to design institutions in ways that “economize on virtue,” but also for the Clausewitzian reason that the kinds of traits of character that are exhibited in adversarial systems are not easily self-limiting. In the absence of external constraints, adversarial processes tend to escalate, as it is not rational for either side in a contest to be the first to exhibit self-restraint.

If this is the case, then the cure for the kind of corruption of democracy that partisanship creates will be institutional rather than individual. We must identify the specific moments in the life of democracies where unconstrained partisanship poses the greatest threat, and identify institutional bulwarks that do not so much extinguish partisan energy as channel it. For example, to revert to the two examples that I briefly described as illustrations of democratic corruption, one can imagine electoral rules being modified in ways that would make it difficult for partisans to uncouple the desire for one’s party to win from the motivation to present an attractive policy platform to the electorate. Electoral commissions might, for example, require as a condition of receiving state funding that parties submit party platforms meeting minimum standards within a certain minimal time frame prior to elections. The second problem, that of “party first” behavior in legislative forums, can be addressed in a variety of ways, relating, for example, to the choice of electoral systems most likely to incentivize cross-party compromise, or with rules governing the composition of legislative committees.

VI. Conclusion

I’ve argued in this essay that corruption is a broader and more pervasive problem than might be thought when focusing on the most spectacular and media-friendly forms of corruption involving financial and sexual inducements and morally compromised individuals. Incentives to derive individual benefits of various kinds often have institutional, rather than purely individual sources. The most difficult to deal with are the cases in which the source of corruption is also central to the institution in question functioning as it ought. Democratic institutions are vulnerable to this form of corruption through the tendency for political parties to breed the motivations and traits of character that make up partisanship. Democratic theorists should, therefore, move the question of how to continue to benefit from the motivation associated with partisanship, and from the epistemic and moral properties of political parties, to the top of the agenda of democratic theory.