GETTING A GRIP: POLITICAL CORRUPTION, DEMOCRATISATION AND REFORM
Mark Philp
Prepared for Princeton University/Central European University joint conference on Corruption
29 Oct-6 November.
I. Corruption
This paper examines issues in the definition of political corruption and links these with a discussion of the problems associated with analysing and controlling corruption in the process of democratisation. The concept of political corruption is predicated on prior judgments and norms of non-corrupt conduct in public office and of the nature of political rule. Transition involves a transformation in the formal political structures of the state and it invokes new norms and expectations for politics, yet only rarely are these norms and expectations widely shared among the political élites and broader populations of these states. Widespread political corruption is one likely consequence, but its control depends very much on the precise character of the corrupt activity.
i. Normative pull: Corruption designates a state of decay, implicitly contrasting it with a 'normal' state, and giving normative priority to the latter over the former. Although corruption is a negative term of appraisal, it is not an unconditional one: political corruption is only as bad as the normative purposes served by politics are good. A policeman who is bribed to turn his back while I escape from an oppressive political order is acting corruptly, but the consequences of his doing so are better than those which follow from his assiduous application of the law. Morally, in this case, we will not condemn the briber. Our attitude to the bribe-taker is more complex: his behaviour produces results which we prefer to the alternative, but we condemn his actions as corrupt because (and in so far) he uses his official position for self-enrichment.[1] The implicit normative weight attached to the conduct of public office and political rule is mirrored in the condemnation which the attribution of corruption carries, but these normative values do not over-ride every other normative claim, and they need grounding.
ii. Grounding the normative pull: Any complex social practice made possible by a set of rules and norms is susceptible to rule breaking. One type of rule-breaking is corruption. Whether a practice is susceptible to corruption depends on whether it has certain characteristics (which I turn to in a moment). But the moral weight of any practice (such as politics) depends on a complex mix of its social functions and consequences, the extent to which generates expectations among its participants and those whom it affects, and on counter-factual elements as to how far it precludes other more desirable practices emerging. To see political corruption as harmful, then, requires both that we identify criteria for, and standards of political conduct, and that we have some sense of the ends or purposes these serve. I have suggested elsewhere that the central criteria for politics is that it involves a claim of a right to rule, where this is grounded in a general, public-regarding justification, and that a minimum account of ends it serves can be given in terms of the management of various forms of social conflict.[2] At the heart of our sense of politics is the idea that it must be justified in terms of its achievement of order, security and stability, and that those who use the political and financial resources which the political system affords them to pursue their own ends are subverting that system and the values it serves. But, again, it must be emphasised that political corruption is only as bad as the ends served by politics are good.
iii. Prescription and description: We can distinguish two dimensions to the analysis of political corruption. a.) There is the, often implicit, prescriptive view of the ends which politics should serve together with an associated conception of what institutional structures, practices, norms and rules are required to realize these ends; and b.) there is the descriptive account of the actual laws, rules, norms and institutional practices of a given political system by which certain practices are identified as corrupt. The descriptive account gives us an initial basis for saying what counts as political corruption within a given system, but it must rest on the prescriptive account to justify the negative connotations of the attribution. Moreover, the prescriptive account must also guide the interpretation of the status quo: every constitutional order requires interpretation, and no order can exhaust the possibilities for interpretation by stipulation in advance, with the consequence that fundamental normative issues lie at the heart of many areas of political practice and have to be resolved in the interpretation and implementation of law and in legislative and political activity. Similarly, since law can itself be the result of corrupt activity, we need some external standard to ground out judgments. On the other hand, the normative account on its own is at too high a level of abstraction and generalisation to be used as a blue-print. Political contexts vary hugely; correspondingly, so does what is politically possible. We have to start with what we have, even if we must also recognise that what we make of it must be informed by the prescriptive dimension.
iv. Corruption and bad politics: Not all forms of bad politics are corrupt. The classical republican tradition treats corruption as the corruption of the state, and identifies practices which lead to the downfall of the state as corrupt. The basic insight of that tradition, that corruption attacks politics and the benefits and values it serves, is valid, but corruption is not the only thing that does so, and we need to avoid inferring from similarity of effects, similarity of cause. In politics there can incompetence, irresponsibility, arrogance, intemperance, and much else besides. These make for bad politics and poor conduct in political office, but they are not identical with political corruption.
II. The Definition of Political Corruption:
i..Three definitions: The three main definitions of political corruption are well known: public office, public-interest, and market accounts[3]. Public office definitions identify corrupt behaviour with officials acting in ways which deviate from their formal public role and with the intent to secure certain private gains. For public interest accounts, corruption exists whenever a responsible official is induced by the promise of certain rewards not legally provided for to act in ways which favour the provider of the rewards and thereby to damage the public and its interests. The two definitions are, however, inevitably interdependent. The public office account assumes that the deviation from the formal duties of the public role take place for private gains, thereby introducing the distinction between public and private interests. Similarly, public interest conceptions depend on there being an understanding of public office in which the substitution of private for public interests is recognised as corrupt. Market definitions offer ways of understanding or analysing corruption but do not provide a way of defining it.[4] An act is corrupt not because it is income maximising, but because it is income maximising in a context where prior conceptions of public office define income-maximising as corrupt. Similarly, if corruption is defined in principal-agent terms, or in terms of trust, we need in each case to say what it is about this sub-set of principal-agent or trust relations which marks them off as corrupt.[5]
ii. The Norms of Public Office/Interest: Although there are clear links between public office and public interest accounts of corruption, one stumbling block to further agreement on a definition of political corruption is the issue of which norms should be authoritative for the conduct of public office and the content of the public interest. Even if the law provides and exhaustive set of standards for conduct within a state there remains the more fundamental question of how far these are the right standards. Empirical analysis simply cannot resolve that issue, nor can it be set aside. Moreover, this undercuts those alternative definitions of corruption which appeal to public opinion or legal norms. Public opinion is rarely agreed on such complex issues, and even when it is consensus is not the same as correctness. Equally, the supreme case of corruption is where those in power set up laws and institutions in ways which serve their individual or group interests. Discussions of the proper conduct of public office and the nature of the public interest must inevitably reach beyond existing practices to more abstract and general values and principles.
iii. Core Features: Despite these difficulties there are cases which we can recognise as incontestably corrupt. Core cases of corruption usually involve four key components:
1. a public official (A), who, acting for personal gain,
2. violates the norms of public office and
3. harms the interests of the public (B)
4. to benefit a third party (C) who rewards A for access to goods or services which C would not otherwise obtain.
The three core elements of A, B and C have been used by Diego Gambetta as a way of identifying a consistent, cross cultural account of corruption which identifies it as a trust relation between A and B which A violates to the benefit of C - where the commodity transferred to C is monopolistic and where A's motive to act in this way is C's bribe.[6] This account allows the content of corrupt acts to vary across cultures and time while the type of action remains constant. However, while he treats the triadic relation of trust as providing the core for a model of corruption, conditioned further by constraints on motive and on the presence of quasi-monopolistic rules regulating the distribution of the commodity in question, I would reverse the order, putting the concept of public office (which is closely connected to the presence of quasi-monopolistic rules) prior to, and not wholly reducible to, the concept of trust. That is, we need the concept of public office to identify which derelictions of trust are politically corrupt; but we also need it to identify which derelictions of trust are corrupt. That is, there are types of trust rel-ationship (such as amorous relationships) whose subversion is not corrupt. For it to be corruption, the trust relationship must a sub-set of such relationships where a set of publicly endorsed rules and norms identify the nature of the trust and its purposes (e.g., the rules and norms of public office, and the underlying principles which identify the point of that office), and which introduce of a separation within the individual between what is required of him in a public capacity and what he may wish to do as an individual agent with certain wants and preferences. It is a further move to resolve adjectival disputes as to whether the case of corruption is, for example, political or economic. The term corruption cannot be predicated of a relationship without reference to the particular set of rules and norms which turn a case of disappointed expectations into a one of corruption.
iv. Core and Periphery Cases: The point of the four criteria is that activities which meet all four are corrupt, although there are also many cases where only three of the four elements are present but where we are still justified in claiming that the action/relationship is corrupt. For example, it is contestable as to whether kleptocracy requires a third party benefitting, yet few doubt that such a regime is corrupt. Similarly, the rewards to C may be something C has a right to, but where the public official levels a tax on access. Alternatively, the public official may act to avoid certain costs, rather than to incur certain benefits. We should also distinguish between A-led and C-led types of corruption: in the former the public official imposes the terms on C (from extortion to informal 'taxation'), in the latter the relationship is reversed (from bribery to systematic subversion of the political domain). (We should also recognise different types of mutual corruption: cases where the exchange is equivalent and the parties equal (as in market transactions); cases where the exchange is asymmetric but the parties equal, as in 'blat'; and cases where the exchange and the parties are unequal, as in patron-client systems).[7] The public interest component may also be weak: the bribed policeman may in fact end up acting in the public interest. On the other hand, we may still be able to use the concept of public interest as a way of helping to clarify the nature of such an individual's responsibilities as a public official. What remains central is the idea of public office as identifying the character and extent of, and the responsibilities associated with, the relationships between A, B and C.
v. Institutional Corruption: The first criterion is further complicated by the distinction which Dennis F. Thompson draws between individual and institutional corruption. He demonstrates that in democratic systems it is often very difficult to hold a line between what one does to obtain political office and how one acts once in office. In many democratic systems, not least the US, the need frequently (and expensively) to ensure re-election creates powerful incentives for individuals to use the resource (in the widest sense) of public office to secure their continuing occupancy of that office. Thus, whereas individual corruption violates rules of office intentionally for personal gain, 'institutional corruption' benefits the office holder politically rather than personally. On Thompson's account, providing favours in return for campaign contributions is corrupt not because it is, in itself, illicit, but because its 'institutional appearance' or 'institutional tendency' is such as to suggest that such services can be obtained in ways which are not in keeping with democratic standards. 'We have to show only that a legislator accepted the gain and provided the service under institutional conditions that tend to cause such services to be provided in exchange for gains'[8] Thompson's account demonstrates the extent to which normative judgments as to the essential nature and purposes of democratic political systems enter into the definition of political corruption.
vi. Comparative judgments: There is an understandable desire to provide a definition of corruption which allows comparative judgments to be made. But if there are differences in the norms of public office between states comparison becomes invidious. The incidence of corrupt rule-breaking or prosecutions for corruption will be a function of how extensively corruption is legislated against and how assiduously it is prosecuted. The difficulties become still greater when we face contrasts between stable Western democracies and societies where strong patrimonial, patron-client, tribal or communal traditions determine access to political power and shape its exercise. Rather than assessing the latter by standards of the former, we should be asking, whether the system works in its own terms, whether it offers the best prospect for ordering conflict in the society, and whether its attempts to order such conflicts are systematically undermined by individual or group activities which that process seeks to constrain. Where there is no recognition of a need for a political order, with public offices, formal rules of conduct and a sense of the public interest, the fact that distributions and allocations take place on non-political criteria does not mean there is corruption: allocations within families rely on other principles, but that does not make them corrupt. It is political corruption only where a political order, which expresses the aspirations of some significant part of the culture and offers a way of reconciling conflicts which alternative modes are acknowledged to exacerbate, is disabled from functioning through its subversion by other orders or systems of exchange. As I have suggested, it is possible in abstract terms to identify when politics, in this sense, is desirable and possible, and when it is being corrupted. But it is always a more local and culturally relative matter as to what in practice necessitates that order, what threatens it, and what sorts of standards are necessary to preserve it.
vii. Conclusion: Some of the difficulties we have identified with the definition of corruption arise from ambiguities and opacity in the way that we identify public officials, their roles, and the rules governing their official conduct; and from the fact that each criterion is intelligible only against the background of a political culture in which there are shared norms and rules governing the conduct both of public officials and of members of the public in their dealings with these officials. A further factor is that the ascription of corruption is partly a normative judgment which acts as an external standard from which critically to appraise local practices and norms. Moreover, a final dimension which needs emphasis is that we can only have political corruption in so far as the institutionalised pattern of norms and rules which delimit public office and regulate its conduct, has an impact on political outcomes. Politics involves the exercise of public office and this implies rule - and political corruption subverts aspects of that rule. Yet this viability, efficacy and impact of that rule is influenced by a very wide range of factors other than the norms of political office, and we need a clear grasp of the point at which such factors corrupt politics.