The politics of scandal: Political time horizons, organisational life cycles, and anti-corruption agencies in the ‘new’ EU member states
Agnes Batory Department of Public Policy/Centre for Policy Studies Central European University
[email protected]
Paper prepared for the EGPA Conference, Permanent Study Group on Governance of Public Sector Organisations Toulouse 2010
Abstract A large number of ‘independent’ agencies to combat corruption have sprung up around the world in the past decades. Yet, little comparative research has been undertaken to explain the diversity of their organisational forms or their development trajectories over time. This paper is a preliminary exploration of the following arguments: Firstly, anticorruption and anti-corruption agencies (ACAs) can be studied as regulation of government and as bodies analogous to independent regulatory agencies, respectively. Second, the formal powers and independence ACAs are granted at the outset will crucially depend on whether external or domestic impetuses for setting them up (the external impetus being conditionality and the domestic essentially the need to pacify voters in the wake of a scandal) can counterbalance strong incentives for no action, or only symbolic action. With respect to domestic impetuses, politicians’ time horizons (i.e., the timing of the creation of an agency within the electoral cycle) have major implications for the body’s independence and the scope of its mandate. Finally, the ACAs’ initial formal mandate influences but does not determine how they fare in later life: exigencies, such as support or obstruction from ruling governments, their own ability to use strategic resources, and finally leadership shape the extent to which the agencies are able to carry out their tasks in practice. These arguments are examined through comparison of the establishment and evolution over time of a number of ACAs in the ‘new’ member states of the EU – Latvia, Poland and Slovenia.
1
Introduction Following the example of anti-corruption commissions in Singapore and Hong Kong – both commonly described as phenomenal success stories – ‘independent’ agencies to combat corruption have spread around the world in the past decades. The agencies were initially hailed as major achievements in curbing corruption, and have been propagated by international organisations of various kinds to a wide range of countries. Since then, however, enthusiasm has waned, as particularly in unstable governance contexts such as sub-Saharan Africa the agencies have tended to fail to deliver visible results and thus lose credibility, or in the absence of government support and resources fade into oblivion. In most cases, the design of the bodies seems to have rather obviously failed to strike the difficult balance between agency autonomy and political control/accountability (Christensen and Laegreid, 2007). It is no accident if this short overview sounds familiar to scholars of independent regulatory agencies. Indeed, one contention of this paper is that governmental anti-corruption activity is essentially regulatory activity and consequently many of the themes of the literature dealing with independent (regulatory) agencies (IRAs) are also relevant in the context of anti-corruption agencies. To be sure, there are important differences between IRAs and ACAs, stemming from the fact that the latter are regulators and regulatees at the same time, in the sense that agencies charged with controlling political (and administrative) corruption are often directly confronted with political parties setting them up. There are however also many commonalities in terms of drivers for the creation of agencies, such as policy diffusion and isomorphism (conditionality or problem pressure), the need for specialised expertise, the enhancement of the credibility of policies, or insulation/lock-in effects. Yet surprisingly little work has so far attempted to apply these insights to the anti-corruption field systematically. There have also been relatively few attempts to shed light on the role of electoral incentives, despite the fact that the comparative politics literature (by definition) can contribute major insights into how party politics shapes governmental action, in this case, the setting up and institutional design of a particular public body. The regulation literature is helpful for addressing the shortcomings of the ACA and more broadly anti-corruption literature. Although a lot has been written about why anti-corruption efforts fail -and this is literature is very relevant to this paper – there is no convincing explanation of the cross-national variation, even in countries with similar political systems and governance contexts, in the ACAs’ autonomy and the scope of their mandate. For instance, why do some agencies have investigative powers, while others are set up for policy coordination and/or prevention only? Why are some agencies directly subordinated to bodies made up of, or dominated by, governing parties (cabinets and parliaments), whereas others are carefully removed from the influence of political decision-makers? What is the motivation for political parties to set up agencies that wield the power to discredit them in the first place? And how do the agencies change over time, and what are these shifts (expansion or contraction in powers, resources and independence) in response to? This paper is a preliminary exploration of the following arguments: Firstly, anti-corruption and ACAs can be studied as regulation of government and as bodies analogous to IRAs, respectively. Second, the formal powers and independence ACAs are granted at the outset will crucially depend on whether external or domestic impetuses for setting them up (the external impetus being conditionality and the domestic essentially the need to please voters) can counterbalance strong incentives for no action, or only symbolic action. With respect to domestic impetuses, politicians’ time horizons (i.e., the timing of the creation of an agency within an electoral cycle) has major implications for both the body’s independence and the scope of its mandate. Finally, the ACAs’
2
initial formal mandate influences but does not determine how they fare in later life: exigencies, such as support or obstruction from ruling governments, their own ability to use strategic resources, and finally leadership shape the extent to which the agencies are able to carry out their tasks in practice. These arguments will be examined through comparison of the establishment and evolution over time of a number of ACAs in the ‘new’ member states of the EU, thus combining a national and temporal comparative approach (see generally Levi-Faur 2004). The empirical focus on the new member states is both relatively novel and eminently suitable for a project of this kind. The literature on anti-corruption bodies is dominated by case studies of agencies in post-colonial contexts, whereas little research deals with European anti-corruption bodies.1 In-depth empirical work on anti-corruption agencies in Central and Eastern Europe (CEE) is particularly rare. Yet, the vast majority of European ACAs are located in the EU member states of the 2004 and 2007 enlargements – indeed ACAs are now more common in CEE than in any other region of the world (Charron, 2008). The CEE EU member states produced a wide variety of agencies that offer interesting contrasts and parallels: multi-purpose agencies with law enforcement powers (e.g., Latvia and Lithuania); law enforcement type services (Poland, Romania); and preventive and coordination bodies (Slovenia). At the same time, the countries share a sufficient number of characteristics to control for other sources of variation. These similarities include post-communist legacies, public administration traditions, parliamentary political systems, and most importantly, EU conditionality as an important, if not the chief, driver of reform prior to EU accession (see e.g., OSI, 2002; Batory, 2010). Ideally, all the agencies in the CEEs should be examined to assess the influence of external and/or domestic (electoral) drivers of anti-corruption reform, but data availability limitations only allow for a small number of cases, which will serve as illustrations of the arguments above. 2 Latvia, Poland and Slovenia, the cases selected for this paper, each possess relatively well known anti-corruption bodies that are sufficiently diverse for exemplifying different patterns. The paper relies on empirical material from policy documents and news coverage. The written sources consulted include Freedom House’s Nations in Transit reports that provide a summary of corruption related developments for each year and each of the countries under discussion here; evaluation reports by the Council of Europe's Group of States against Corruption (GRECO); related news coverage wherever an English language source was available, and comment and analysis by local NGOs, most notably Delna and Providus Latvia. The first section below discusses the existing literature on ACAs, and generates a short list of factors that international experience suggests are important determinants of the agencies’ development trajectories. A second section refines this picture by considering insights from the regulation/IRA literature, putting forward a number of tentative propositions linking agency development to particular circumstances present at their establishment and operation. A third section illustrates these patterns in the case of the three agencies selected for this paper. A brief section concludes.
Anti-corruption agencies: State of play The (modern) literature on corruption has taken a meandering path, from a period in the 1960s when the phenomenon was seen as an inevitable and not necessarily harmful by-product of 1
A notable exception is a large data set and comparative analysis by Ancorage-net, coordinated by Luis de Sousa. 2 The intention is to expand the number of cases for a follow-up paper.
3
‘modernisation’, put explicitly in North/South, developed/developing country terms, to the 1990s though to today, when it is (almost) universally described as both omnipresent and in its consequences strongly detrimental (see e.g. Andvig et al, 2000). For the dominant (public choice influenced) strand of literature, corruption is seen as a harm to be managed and mitigated, similarly to other social problems such as poverty or discrimination. 3 Corruption control, in turn, is seen as ‘public policy aiming to reduce the opportunity structures for corruption and to punish deviant and unlawful [behaviour] through the implementation of an integrated set of measures’ (de Sousa 2010: 8). The set of measures, advocated most notably by international financial institutions and development agencies, were informed by principal-agent models, and the idea, put forward perhaps most prominently by Robert Klitgaard (1988: 75), that corruption ‘flourishes when agents have monopoly power over clients, when agents have discretion, and when the accountability of the agents to the principal is weak’. Consequently, at various points in time, a host of anti-corruption interventions were introduced around the globe, from Australia to Ghana and Ukraine, aiming to break up public sector monopolies through privatisation, and increasing bureaucratic controls over public administration.4 This is the context in which anti-corruption agencies – ‘public bodies of a durable nature, with a specific mission to fight corruption and reducing the opportunity structures propitious for its occurrence in society through preventive and/or repressive measures’ (de Sousa 2010: 5) – emerged as perhaps the single most important weapon in the armoury of corruption fighters. The ACA model appeared to offer the way out of an all-important policy paradox: namely, that corruption tends to flourish in countries with weak institutions, i.e., countries where governments are quite literally powerless to affect change, and is at the same time the principal obstacle to creating more effective public institutions. As Rothstein (2007) put it: the problem for corruption control in these contexts is that ‘everything (almost) has to change and this should be conducted as simultaneously as possible’. The ACA model was initially seen to solve this problem, because instead of reforming the existing system, with all the ensuing questions about sequencing, it promised a fresh start: to break up the vicious cycle by ‘wedging’ in a new actor. The expectation was that a new body can be a ‘catalyst and a building block’ for further reforms (Doig, 1995: 163). In the late 1980s and 1990s, ACAs spread like wildfire around the globe. The single most important reason for the speed of policy diffusion is the spectacular success of Hong Kong’s Independent Commission Against Corruption (ICAC), established in 1974, itself following on the footsteps of Singapore’s earlier agency, in visibly reducing levels of corruption within a relatively short period of time (Klitgaard, 1988). The ICAC model was innovative because it centralised hitherto dispersed (and badly coordinated) anti-corruption functions into a single powerful agency, lavishly resourced and equipped not only with strong investigative powers but also a mandate for prevention and public education (Meagher, 2005). The influence of the model was such that establishing an ICAC-like organisation became de rigueur for any government wishing (or needing) to demonstrate commitment to corruption reduction – either because of domestic 3
Needless to say, a powerful critique of this strand has also emerged, contesting the basis and legitimacy particularly of international interventions against corruption, or more precisely interventions guised as anticorruption, termed as ‘global anti-corruptionism’ (see Samson, 2004; and contributions in de Sousa et al, 2009). 4 Although the policy interventions were prescribed mainly for developing and transition countries, they were keeping in with the times: there are clear parallels with OECD countries such as the UK, ‘where the New Public Management [met] the Audit explosion’ (Hood et al, 1998) around the same time.
4
political scandals or to please donors who had made development assistance conditional on such action. The idea that ACAs are all but essential for corruption control was even more firmly established when it entered emerging international norms against corruption. The 2005 UN Convention Against Corruption includes an obligation for signatories to establish specialised anti-corruption bodies with ‘the necessary independence’,5 and there are similar provisions in many regional conventions (OECD, 2007). The international NGO Transparency International’s renowned Sourcebook also refers to independent ACAs as essential ‘pillars’ of ‘national integrity systems’ (Pope, 2000). There were also more direct ways in which the emulation of the Hong Kong experience was promoted, most notably by international financial institutions, which, as concerns over development assistance leaking away in recipient countries grew, pushed developing countries to adopt the model as ‘best practice’ in corruption control. The diffusion of the ACA model was thus assisted by every conceivable international transfer agent (Stone, 2004) and transfer mechanism (e.g., Holzinger and Knill, 2005) known in the policy transfer literature. Among the many ACAs that emerged there is tremendous variation in the organisational forms they take and in public and/or expert perceptions of their performance. The OECD’s (2007) review of models categorizes agencies as: i) multi-purpose bodies with law enforcement powers (essentially, the original ICAC model with a three-pronged approach combining investigative, education and prevention functions); ii) law enforcement type services (specialised prosecution services and police forces); and iii) preventive, policy development and coordination institutions. Multiple examples are provided for each type of agency, and while the first, ICAC-style multipurpose body, is rare, numerous European bodies, such as Belgium’s Central Office for the Repression of Corruption or France’s Central Service for the Prevention of Corruption can be found in the other two categories. There is also a lot of variance in how well ACAs are seen to do their job. Although there is no commonly recognised criteria for determining success or failure (Meagher, 2005; de Sousa, 2010), a number of agencies are well regarded in the literature. Botswana’s Directorate on Corruption and Economic Crime is described as an effective organisation (Heilbrunn, 2006: 142-3), and Latvia’s Corruption Prevention and Combating Bureau, one of the case studies in this paper, features as best practice in an EU compendium of anti-corruption measures (Utrecht School of Governance, 2008). Unfortunately, by now the success stories seem to be a small minority among the many failures. Evidence abounds that the model was, more often than not, transferred to countries where elementary preconditions, such as a degree of administrative capacity and a working judicial system were clearly not in place. Unsurprisingly, in these circumstances ACAs proved unable to produce results in island-like isolation from the rest of the public sector, engulfed by systemic corruption and/or incompetence. Many of the agencies eventually fell victim of the same structural problems – principally, capture – as other public (law enforcement) agencies had done in the past (e.g., Meagher, 2005; Doig et al, 2005). Consequently, in recent years a sense of disappointment has been mounting about the ACAs’ performance (de Sousa et al, 2010). According to a 2004 OECD report, ‘anti-corruption agencies have met with mixed results’ outside their original sites in micro-states of South-East Asia, while a 2005 UNDP report points out that ‘[t]here are actually very few examples of successful independent anti-corruption commissions/agencies’ (both quoted in Doig, 2009). How do ACAs get, as de Sousa (2010) puts it, from empowerment to irrelevance? And conversely, how did the successful cases make this journey, in the reverse? In the original Hong 5
Although the convention leaves open what kind of organisational form the body or bodies should take.
5
Kong case, ICAC literally went from strength to strength: having been established with strong political backing from the governor (Hong Kong being a British crown colony at the time), it was staffed with highly paid professionals, and managed to establish its credibility early on by putting a notoriously corrupt official behind bars in a high profile case (Klitgaard, 1988). Public confidence in the agency grew, which meant more citizens reporting corruption, then over time even more resources and more staff was provided to handle an increasing case load and a wide range of public outreach and prevention activities, and so on. The virtuous cycle has continued until today, leaving ICAC stronger and more powerful than ever: it employs a staggering 1200 staff, and responds (within two days!) to approximately 3500 reports on corruption acts annually. 6 Hong Kong is normally among the 10-15 “cleanest” countries internationally, way ahead of many OECD countries, on Transparency International’s Corruption Perception Index – a profound transformation indeed since the 1970s.7 In contrast, the life-cycle of later time ACAs in developing countries seems to be generally that of gradual decline – albeit not necessarily linear decline. Organisational development is typically ‘sporadic, erratic and vulnerable to disruption by the volatility of government support and by fluctuating donor enthusiasm and fatigue’ (Doig et al, 2005: 5). To illustrate the nature of these disruptions, it is worth quoting Doig et al’s (2005:4) excellent in-depth study of five African ACAs here: The lifecycle of a new ACC is characterized by initially high expectations from governments and donors but the ACC is an infant organization unable to meet the unrealistic expectations imposed upon it. This failure usually means that there is no sustained support for the ACC which limits its capacity to develop as an organization. This failure ‘to thrive’ encourages disillusionment in governments, donors and in ACCs themselves. … ACC Directors are dismissed, authority to prosecute is withheld in sensitive cases and under-resourcing of the ACC becomes the norm, further undermining its capacity and reputation.
The (downward) spiral takes place in the context of electoral cycles, which have strong implications for moving anti-corruption up and down on a given government’s agenda. The lifecycle of governments involves the gradual displacement of anti-corruption as a high priority and indeed political commitment is frequently confined to exposing the crimes of previous regimes. Governments experience periods of instability and, where ACCs investigate corruption at the highest political levels, the response is not often supportive. (Doig et al, 2005: 4)
There is a question as to the generalisability of these observations beyond Africa, but Maor (2004) established similar patterns with respect to a range of anti-corruption mechanisms in countries and time periods as different from each other as the final days of the Soviet Union, mid1990s Australia (Queensland), and the Italian ‘mani pulite’ magistrates of the early 1990s. In addition to predictable the ebb and flow (or rather flow and ebb) of government support, Maor finds that in each of these cases governmental reprisals against anti-corruption bodies also occurred because the agencies or officials (investigators, commissioners, magistrates) invited attacks by becoming ‘over-zealous’. Zealotry is argued to result both from mass media attention encouraging the officials to take further action, and from the officials’ tunnel-vision as a direct consequence of their single issue mandate. However, within these predictable patterns of high hopes and brave ideas slowly giving way to political realities, ACA development trajectories are quite different across cases. Even in a downward swing, there are multiple possibilities for how an agency might fare. One possibility is, as Doig et al (2005: 4) argues, that ‘[w]hen discredited governments are toppled or new leaders 6
See ICAC’s webpages at http://www.icac.org.hk/en/about_icac/bh/index.html http://www.transparency.org/policy_research/surveys_indices/cpi/2009/cpi_2009_table. The usual disclaimers about the methodological limitations of the CPI apply. 7
6
emerge, donors become enthusiastic again and the neglected ACC is reborn or reconstituted’ – i.e., that after periods of neglect and marginalisation, the agencies are given a lifeline and an upward swing follows, until temporary decline starts again and so on. This pattern might be labelled as a rollercoaster. Another possibility is that hostility from governments in office is expressed in the dismissal of the agency’s head, budget cuts, or the withdrawal of cooperation from other (government controlled) public bodies – making the work of the agency very difficult, but not impossible. This pattern might be called stagnation. Yet another possibility is that the agency does not survive. The termination of ACAs is in fact quite rare, but not unprecedented: the three known cases of any kind of anti-corruption body or office being abolished concerns Portugal in 1992, Italy in 1998, and South Africa in 2007 (where the Scorpions were merged into the police) (de Sousa, 2010). What might explain this range of paths, from almost linear development, with an expansion in mandate and resources, as in the ‘gold-standard’ Hong Kong case; a rollercoaster ride shaped by changes of government; stagnation in a hostile political environment; to agency termination? Four factors are suggested by the brief overview of international experience above. One is the agency’s initial official mandate: how strong and independent an ACA is at the outset has important implications for its ability later to fend off attempts to redraw this mandate or isolate itself from adversely changing external circumstances. A second factor is de-facto operational autonomy, shaped by its resources. This is in turn dependent on a third factor, the degree of political support/opposition from changing governments – i.e., the question whether elections change the attitudes of politicians towards the agency and its work. And finally the fourth is leadership – the somewhat ill-defined yet recognisable quality of individuals heading up agencies that enables them to instil a sense of mission and dedication in their staff, and to go after their goals with single-minded determination (‘zealotry’).
The relevance of the regulation literature Before these factors are formulated as tentative hypotheses to be considered in relation to the CEE anti-corruption agencies, a brief excursion to regulatory theory seems to be in order to put them into context. One shortcoming of the otherwise rich literature on ACAs, and more broadly anti-corruption, is that it is relatively insular: the subject tends to be studied by scholars who were drawn to it because of their interest in corruption, rather than in public sector organisations and their management. The assumption seems to be that corruption is just too multi-faceted and, for lack of a better word, special for governmental responses to it to be compared with other policies (this is analogous with a rather extreme version of what Laegreid et al (2008) describe a taskspecific perspective on regulatory agencies). Thus, despite the fact that some early and influential studies explicitly rely on theories of delegation and principal-agent frameworks to explain corruption (and in some cases, corruption control, see e.g., Klitgaard, 1988; Rose-Ackermann, 1999), discussion of anti-corruption in a regulatory context and explicit parallels with independent regulatory agencies are rare.8 This paper clearly cannot (not does it aim to) overcome this limitation at one fell swoop, but perhaps a number of ways in which the regulation literature is relevant can be identified and utilised here. The simple idea behind this is that while corruption is indeed multi-faceted, notoriously amorphous, and contextually dependent, corruption control is a public policy: the provision by governments of a public good, i.e., an ‘acceptable degree of freedom from corruption’, a little like reasonably clean air or health and safety at the workplace. 8
Rothstein (2007) and Cocciolo et al (undated) are notable for doing this, and de Sousa’s large comparative report (2006) is also informed by the regulation literature.
7
Of course, one question that arises here is whether the control of the operation of the public sector can be seen as regulation, normally taken to mean ‘state intervention[s] in the economy or the private sphere designed to steer them and to realise public goods’ (Laegreid et al, 2008: 4). While some ACAs have a mandate to control both public and private sector corruption, it is more common for the agencies to be set up for preventing and rooting out corrupt practices within the state apparatus. However, it is now well-established in the literature that regulation also takes place inside government, as public bureaucracies also face scrutiny by an ‘army of wastewatchers, quality checkers, “sleaze-busters” and other regulators’ (Hood et al, 1998: 61). Common to these relationships is one bureaucracy, in possession of an official mandate, overseeing and seeking to shape the activities of another and organisational separation between regulator and regulatee (and externality of the regulator to direct lines of command) (Hood et al, 1999; Hood et al, 2000; James, 2000) – all of which applies for ACAs. One example of the relevance of regulatory theory is the applicability of now ‘classic’ explanations, of both the rational choice and the sociological institutionalist variety, of why (independent) agencies are established. Many if not all the reasons why IRAs are set up for – synthesised below relying on Thatcher and Stone Sweet (2002), Jordana and Levi-Faur (2004a) and Gilardi (2008) – should also apply in the anti-corruption context. And there is in fact a strong indication that they do, from both various case studies of ACAs and an invaluable survey of heads of ACAs conducted by de Sousa (2006). Enhancing the credibility of commitments. Anti-corruption bodies are commonly created by governments needing to boost their integrity credentials to voters and/or to powerful external actors. African cases in particular point to governments setting up commissions in order to convince the World Bank that they are ‘serious’ about reforms. Similarly, political parties rocked by corruption scandals show to the voters their determination to clean up public life by creating agencies (notable examples are commissions in Australia’s states, e.g. New South Wales or Queensland). Policy transfer through (coercive) isomorphism. The flipside of the coin, from the external actors’ perspective is to push for credible commitments through conditionality in various forms. Aid conditionality used by the international financial institutions in developing countries, and EU membership conditionality in Central and Eastern Europe are prime examples. Reducing uncertainties in the policy area over time. By the creation of an independent ACA governments tie their own hands: it makes it harder for politicians in office to change their minds about combating bribery. Independent ACAs also reduce uncertainly arising from the periodic change of ‘principals’ (i.e. party governments through elections): ‘current politicians … bind future politicians’ (Jordana and Levi-Faur 2004b: 17). ‘To curb corruption without political interference’ was perceived as one of the strongest rationales for setting up agencies by surveyed heads of ACA (de Sousa, 2006: 14). Addressing information asymmetries by utilizing technical expertise. In the same survey curbing corruption in a knowledge-based manner was seen as the most important reason for setting up an ACA. Information asymmetry is a particularly important consideration as corruption involves conscious opponents. The ever more sophisticated modes of private enrichment at public expense mean that ‘control operations [are transformed] into a game of intelligence and counter-intelligence’ (Sparrow, 2008:199). Legitimising policies by choice of a valued institutional model (Jordana and Levi-Faur 2004b: 18). The expectation to reproduce the Hong Kong ‘miracle’ played a part in many agencies’ creation. Lithuania for instance justified setting up its Special Investigative Service by explicit reference to the Hong Kong model (OECD, 2008: 46).
8
Many if not all these motivations for governments to establish agencies apply simultaneously and, apart from coercive policy transfer, all the time. So why are ACAs established when they are established? Again, the theory of regulation has one plausible answer: delegation occurs as ‘response to disaster’, when a ‘spark’ puts an issue on the agenda and opens policy windows (Bernstein 1955, referred to in James, 2000: 335). In the ACA literature too, references to scandals as triggers are common. In each case, including Hong Kong’s ICAC, they exposed the inadequacy of existing control arrangements and demanded immediate counter-action (Maor, 2004; Heilbrunn, 2006; de Sousa, 2010). Anechiarico and Jacobs (1996) argue that US regulation of government is layer upon layer of control, each added in response to successive scandals. What might explain the changes over time in the agencies’ mandate, and the divergent outcomes in their life-cycles (development, stagnation, termination)? Here too, regulatory theory adds depth to the picture by explaining why principals (government parties) are unable to ‘calibrate’ agencies in completely self-serving ways, and why agencies sometimes develop in unfavourable political climates, or even in the face of strong governmental opposition. This is the idea of ‘feed-back effects’. ‘[D]elegation is often a process rather than a one-off event and, moreover, one in which NMIs [agencies] can be active agents’ (Thatcher and Stone Sweet, 2002: 18). Wilks and Bartle’s (2002) study illustrates the unanticipated – and, one may add, unintended – consequences of delegation to independent competition agencies: the fact that, once established, agencies develop their own preferences, influence and, most importantly, legitimacy that is no longer dependent on the delegating politicians or their successors in office. Having become actors in their own right, the agencies then show striking resilience. This might explain why, even in such an exceptionally political field as anti-corruption, there are very few cases of the termination of anti-corruption bodies. Wilks and Bartle’s (2002) study provides another important insight about the nature of the action governments take when a crisis necessitates some sort of policy intervention. Adding to the common explanations for the decision to delegate, listed above, Wilks and Bartle (2002: 148, emphasis added) argue that the creation of agencies and their organisational design is often ‘motivated by a need to reassure and to appear to act’. What takes place is symbolic action, and the newly created agencies are in fact not expected to be actively pursuing the declared policy objective. It is the unanticipated activism (in Maor’s model, zealotry) of the agencies that might make them effective institutions as they find their feet and demonstrate what can be done, turning ‘symbolism into reality’ (Wilks and Bartle, 2002: 171). Symbolic action is a particularly important explanation for the setting up of anti-corruption agencies, because the assumptions about ‘credible commitment’ probably apply to a lesser extent than in other regulatory contexts. This is where the fact that ACAs as agents need to regulate their own principals comes in, i.e., that while governments draw up ACAs’ mandate and approve their budgets, at least in some cases the agencies in turn have the authority to investigate individual politicians and discredit political parties for instance by reporting irregularities in party financing to parliaments, that is, political opponents and the media. Consequently, politicians would need to get over a particularly high hurdle (act against particularly strong incentives) when setting up independent ACAs, were they really to tie their own hands.9 So much so, that it is likely that this hardly ever happens: it is far more common to want to be seen to combat corruption, than actually to combat it. Thus, an agency would only be created if some other influence counterbalances the 9
Krastev and Ganov (2004) refer to this as ‘the missing incentive’: governments either directly benefit from corruption, or they fail to see the benefits of corruption control – at least, within a politically relevant period of time.
9
incentive not to do anything – most likely the need for a (semi-credible) commitment to meet an external actor’s demand (e.g., EU conditionality) or to appease public opinion (domestic constituents). But even then, the odds are that governments will create paper tigers rather than ICAC style independent and powerful organisations. At least, this should be the case when governments expect to be on the receiving end themselves. In terms of the scope of the agencies’ powers, the motivations are fundamentally different when agencies might be used against political opponents. As Doig et al (2005: 4) put it, ‘political commitment is frequently confined to exposing the crimes of previous regimes.’ The incentives would then be to create powerful organisations that are easy to manipulate – not so much paper tigers as attack dogs. (Note that in either case, from the ruling government’s point of view, there is nothing to be gained from granting any more independence to the agencies than absolutely necessary for maintaining some credibility of the action). This sort of strategic calculus can be reasonably expected to be motivated by the time horizons of political actors.10 As Heilbrunn (2006: 143) points out, agencies with strong investigative powers tend ‘to be established in countries with centralised executives free from the institutionalised uncertainties of regular and competitive electoral cycles.’ Indeed, the ICAC was set up when Hong Kong was governed by a non-elected British administrator, and this is the case in many African countries where elections are not free or fair, or simply do not take place. More controversially, it might also be the case in countries with regular competitive elections – if the electoral cycle is predicable. If a new government comes into power in the fairly certain knowledge that they have four years at the helm, they might decide that this is in fact a long enough political time horizon to make the creation of an attack-dog type agency worthwhile. 11 To summarise the discussion so far, the factors identified from the ACA literature might be turned into a number of tentative propositions. The first one, about the initial official mandate of an agency, is that the motivation for setting up ACAs has strong implications for how viable these organisations will be. The desire to please an external actor (international organisation) or flaunt anti-corruption credentials to voters could well result in nothing more than symbolic action – this is in fact the likely outcome, given strong incentives to avoid the risk that the agency will turn against its creators. Unless these incentives are counterbalanced by a pressing need to send a credible signal, then the new agencies created will not be endowed with either too much power or scope of autonomy. Conversely, if a government wants to use the agency against its political opponents (this is most likely at the beginning of the electoral cycle), chances are better that the agency will in fact get substantial powers (including investigative powers), but hardly any distance from its political pay-masters. Second, although governments seek to ‘calibrate’ the agencies with the initial official mandate in predictable ways, a lot depends on how the ACAs, once established, use whatever resources (material and immaterial) they have. This relates to the second, third and fourth factors, 10
If this runs too strongly in the direction of rational choice institutionalism, it is a useful reminder that governments only think of creating an ACA, rather than some other policy intervention, because in their mental frameworks this is the model that has been diffused and legitimated internationally as ‘best practice’. 11 Conversely, it would make sense for outgoing governments to establish very independent agencies to tie their successors’ hands – the problem is only that there is no guarantee that an independent ACA would not go after corrupt politicians past present and future. Not being safe from agency action even after loosing an election removes all incentives for creating one in the first place, unless the current opposition/future government is believed to be more corrupt.
10
operational autonomy, political support, and leadership, respectively, identified in the context of the ACA literature above. How easy or hard an ACA’s life is depends on the disposition of the government in office – e.g., a party, having been given a lot of trouble by the ACA while in opposition is unlikely to look at the agency favourably once in office. The agency’s de facto operational autonomy12 will evolve (even it it’s formal official mandate is not redrawn) in a series of attacks and counter-attacks, where the agency’s chances would be shaped by the extent it has managed to establish itself as an actor in its own right by this time. This in turn has a lot to do with the personal qualities of the agency’s leader. Clearly, these propositions are difficult to test formally, because there are no obvious ways to ‘measure’ the ACAs’ de facto autonomy, nor commonly agreed criteria to assess their performance. Moreover, this paper only covers a limited number of cases. Bearing these limitations in mind, the following short case studies nonetheless provide a useful illustration of broader patterns (the intention is to further examine these propositions in the context of a larger study). The story of three anti-corruption agencies Anti-corruption agencies are very common in the ‘new’ member states: the Czech Republic, Lithuania, Latvia, Poland, Slovak Republic, Slovenia, Romania and Bulgaria each possess an agency, most established in the run-up to the respective country’s accession to the EU (Charron, 2008). (The exception here is Poland, as discussed below). This is no accident: there is consensus in the literature that membership conditionality was a key driver of the spread of the agencies (e.g., OSI, 2002; Dionisie and Checchi, 2007; Charron, 2008). Corruption always featured high on the list of concerns about the CEE countries’ accession and, while the EU did not have an enforceable anti-corruption acquis, it did have a great deal of leverage on the countries aspiring to join. While there was no explicit requirement to set up ACAs (the acquis contains no such obligation), the candidate countries’ governments could often think of no other way to show their ‘determination’ to combat corruption – particularly as the ICAC model was actively promoted to them by various international observers. The agencies that emerged in the CEE countries show a diverse picture, from Lithuania and Latvia that established organisations that resemble ICAC most closely, to law enforcement style organisations in Romania and predominantly preventive bodies such as the Slovene commission. While CEE countries are by no means a homogeneous bloc, there is no obvious explanation for the diversity in organisational forms. Each country has parliamentary political systems where coalition government is the norm, public administrations bearing the trace of post-communist legacies but also some New Public Management inspired reforms, and little tradition of delegation to independent agencies (at least not much preceding EU accession related reforms). Looking at Latvia, Slovenia and Poland, the cases selected for this paper, more closely, they obviously differ in size (the two former being rather small, while Poland is the largest of the new member states), and to some extent, perceived levels of corruption. Although cross-country comparisons are notoriously unreliable in this respect, Slovenia is seen to be on the ‘clean’ end among the CEEs and mid-range within the EU as a whole, and Poland and Latvia more affected by corruption, albeit not as much as countries in South-East Europe.13 Latvia set up its ACA in
12
On the distinction between formal independence (‘enshrined in the constitutions of agencies’) and defacto independence (effective independence in day-to-day activity) see Maggetti (2007) and Yesilkagit and van Thiel (2008). 13 http://www.transparency.org/policy_research/surveys_indices/cpi/2009/regional_highlights
11
2002, Slovenia in 2004 (by transforming an existing body), and Poland in 2006. This is the order in which the agencies are reviewed next. Latvia’s Corruption Prevention and Combating Bureau The idea to create a Hong Kong style agency in Latvia was first suggested by the World Bank, after a WB study showed Latvia to score very highly on an index of state capture (Delna 2009). A 2002 GRECO report pointed out that corruption control suffered from lack of coordination among various institutions responsible for relevant tasks. The EU shared these concerns in its monitoring reports, repeatedly pointing to corruption as a serious obstacle to EU accession. The inadequacy of the existing system was demonstrated all too clearly by a number of high profile conflict of interest cases, including one where the prime minister himself was implicated – the same person who ex officio chaired the Corruption Prevention Council, the very body a newspaper requested to investigate why the prime minister spent his holiday on the yacht of a company the government had given a generous tax break to (OSI 2002: 308). Thus, all was given for the creation of an agency: strong international pressure, principally in the form of EU conditionality; scandals the existing system could not deal with; and an election that brought into power a party stressing a radical break with the past (even the name of the party was “New Era”) and commitment to fighting corruption. The Corruption Prevention and Combating Bureau of Latvia (KNAB) came into being in 2002, finalising plans that had been in the making for some time, with the Hong Kong case serving as explicit model (OECD 2007: 54). Its wide mandate included investigative powers, the authority to verify the compliance of political parties with party financing laws, and preventive tasks – indeed strongly resembling the original ICAC model, apart from one crucial difference: KNAB was subordinated to the Ministry of Justice. Early in the new government’s term, the law regulating KNAB was changed and in April 2003 the Bureau – which had been operational since the end of the previous year – was placed under the direct supervision of the Cabinet of Ministers, and then in 2005 of the Prime Minister personally (OECD 2007: 58). This implies even less distance from party political influence than that was originally the case, although there was (and to some extent, is) lack of clarity as to what ‘supervision’ might entail. The provisions can be interpreted to give the Prime Minister the right to take over a KNAB matter or case (Delna, 2009: 19), or to entail merely the controlling of the legality of KNAB operations, without a right to give orders to the Bureau or its officials (OECD 2007: 58). The appointment and removal of the director of the Bureau is clearly a matter in which the government of the day (may) have decisive influence: the head of the agency is recommended by the cabinet for appointment by parliament. In practice so far most Latvian governments have filled the position by open competition, with the evaluation and selection of the candidates performed by a (non-partisan) selection committee – the recommendation of which was however rejected in three cases (Delna, 2009: 9). One such case occurred in 2004, when a political appointee, Aleksejs Loskutovs was chosen over the candidate recommended by the committee. Mr Loskutovs however seems to have proven ‘too independent’, and after four years at the helm was consequently removed in 2008 (as discussed below). His successor was appointed without convening a selection committee, with the decision taken by the cabinet directly (Delna 2008: 13). Unsurprisingly, the 2008 GRECO report recommends changing a procedures for the appointment and removal of the director, and generally reinforcing the agency’s independence as, given its responsibility for overseeing party financing, KNAB is “in the awkward position of having to supervise its supervisors”(Delna 2009: 20).
12
In spite of these institutional constraints, KNAB’s operations indicate a relatively high degree of de-facto operational autonomy. Already in 2003, the agency demonstrated its independence by causing the removal of a minister on corruption charges from office, and by forcing two parties in the ruling coalition (including “New Era”) to repay large illegal campaign donations (Freedom House [FH] 2003). KNAB maintained this record under the coalition taking office in 2006 too, for instance requiring the leading coalition member People’s Party to repay a large sum of illegal overspending to the central budget (FH 2009). By its activism over time the agency has amassed a large capital of public goodwill: KNAB is one of the most trusted public bodies in Latvia (Delna 2009). This high level of support has in turn served to insulate KNAB from disruptions – the agency has always been relatively well resourced, with no unjustified budget cuts and relatively stable staff numbers, between 128 and 147 (Delna 2009) -- and, to some extent, to protect it when a ruling government wanted to exercise its legal powers for retaliation against the agency. Having initiated three disciplinary proceedings against KNAB director Aleksejs Loskutovs, in 2008 Prime Minister Kalvitis finally dismissed him (Delna 2009: 12). Despite the bad weather thousands turned out to protest against the move in the streets of Riga (the event is sometimes referred to as the “umbrella revolution”), until the PM backtracked, and Loskutovs was reinstated. Eventually Kalvitis succeeded, and the director was fired after news of a case of financial mismanagement within KNAB broke. Political parties remain divided over the status of KNAB – at least some parties in opposition favour giving it more independence, while others, in or close to power, see no reason to remove political influence (Delna 2009). Slovenia’s Commission for the Prevention of Corruption Similarly to the Latvian case, EU and GRECO monitoring reports played an important – probably the most important – part in putting the idea of an anti-corruption body on the agenda in Slovenia, although corruption was seen as a less pressing problem in this country than elsewhere in CEE. It was following a GRECO report that an office for corruption prevention, attached to the Prime Minister’s Office, was established in 2001. The office was tasked with the drafting of a comprehensive anti-corruption strategy, which was eventually adopted, including a decision to create an ACA as an independent constitutional body (a status similar to that of an ombudsman) reporting to parliament. The Commission for the Prevention of Corruption (CPC) was given coordinative, analytical and preventive functions with respect to both the public and private sectors, and while it was not granted power to investigate individual cases, it was given administrative powers with respect to officials’ asset declarations and conflicts of interest (OECD 2007: 110). Members of the Commission were nominated and appointed by a complex procedure involving all branches of power (executive, legislative and judicial) to minimise the influence of any one political actor. In October 2004 the Commission was off to what proved to be a very rocky start, almost at the same time that a parliamentary election replaced a centre-Left government with a centre-Right coalition led by Janez Jansa. Only a few months into the new parliament’s term, an opposition party (the Slovene National Party) proposed closing down the Commission and transferring its competencies to a parliamentary body, i.e. effectively giving oversight over politicians’ economic affairs back to the politicians themselves. According to Nations in Transit reports, there was some indication that the initiative in fact originated from the government, wishing to remove a source of external oversight into its affairs, although the reason publicly cited to support move was the Commission’s alleged ineffectiveness given its lack of enforcement (investigative) powers. In an unusual show of unity between government and some opposition parties, parliament approved the
13
bill and the Commission was shut down – until the Constitutional Court suspended the bill, and the Commission promptly resumed operation (FH 2006, 2007). But political attacks against the body were far from over. In November 2007 a bill was tabled to extend the Commission’s operations until 2009, when the body would be abolished. The bill was not approved but the Commission stayed in a limbo in 2008, possibly at the worst possible time. In that year, just a few weeks before the parliamentary elections Slovenia’s biggest corruption scandal (“the Patria affair”) broke, involving allegations that a Finnish company bribed senior Slovene officials to secure a large military procurement contract. In the Finnish investigation of the scandal prime minister Janza was also implicated and, although no official action was taken in Slovenia itself, the government maintained that the allegations were nothing but a conspiracy to manipulate the elections. Janza and his party nonetheless lost the elections. 2009 then saw further clashes between the Commission and parliament, this time with a centre-Left coalition in power. A parliamentary commission (led by an opposition MP) for instance demanded to have access to all reports filed to the Commission, or in effect, the personal data of any potential whistleblower disclosing acts of corruption involving politicians (FH 2010). Government attacks not only took the form of attempts to redraw (extinguish) the Commission’s original mandate, but also severe budget cuts and a lack of cooperation (almost amounting to a boycott) by other agencies essential for the Commission’s effective operation. The following quote from the Commission’s 2008 annual report nearly illustrates both the severity of these constraints and the morale of resistance the embattled body had developed in response: “The proceedings to adopt a regulation that would again abolish the Commission of course caused problems in the Commission’s work, not so much on account of pressure experienced by the Commission’s staff [footnote: Who became accustomed to such threats from previous years] as due to the non-compliance to the Prevention of Corruption Act by other governmental authorities… It should also be noted that due to very serious issues with the budget which almost led to the staff of the Commission not receiving their monthly salaries, the Commission had serious staff deficiency problems.” (CPC 2008: 37)
In 2008, the Commission was receiving only 50% of the necessary financial resources, and employed only 11 staff instead of the 25 that would have been needed; Parliament had also cut the salaries of the top officials by 25-35% (Dionisie and Checchi [UNDP] 2008: 15). In the circumstances, it is nothing short of a miracle that the Commission carried out any of its tasks at all, but in fact in 2008 it received 661 reports from citizens and resolved 465 cases (up considerably on both counts on previous years), consulted with a wide range of public bodies on so-called integrity (corruption prevention) plans, analysed the asset declarations of senior public officials, processed gift registers and so on (CPC 2008: 27). The attacks on the body seemed only to send a message to the public about its independence. Public opinion was overwhelmingly in favour not only of keeping the Commission, but also of boosting its powers (CPC 2008: 19). This seems to be, in large part, thanks to strong leadership. Drago Kos, CPC’s first and so far only president was a well respected public figure who doggedly – one might say zealously – pursued what he considered was his small organisation’s mission, no matter the obstacles put in his way. As a Nations in Transit report puts it, “Drago Kos… is a thorn in the side of the politicians and public officials … which a priori suggests that he is executing his duties … professionally and lawfully’ (FH 2007). Mr Kos was vice-chairman and then chairman of the Council of Europe’s Group of States against Corruption, and his international professional networks undoubtedly helped to secure strong condemnation of governmental attacks on the Commission by a wide range of bodies from the OECD to Transparency International. By 2010,
14
the Commission finally appeared safe: a new law secures its independence and its (still largely preventive and coordinative) mandate. Poland’s Central Anti-Corruption Bureau The Polish case differs from that of Slovenia and Latvia, or indeed the majority of cases in CEE, in that the creation of the Central Anti-Corruption Bureau (CBA) had nothing to do with conditionality. Plans to establish a special investigative service were announced by Poland’s then new government, led by the Law and Justice Party, shortly after the 2005 elections, with Poland already safely in the EU. As its name suggests, the party was founded largely on a law-and-order and more specifically anti-corruption platform, and indeed corruption was a central issue in the 2005 election campaign (Szczerbiak, 2005). From 2002 on, the country had been rocked by a series of high profile corruption scandals discrediting and fatally weakening the governing Democratic Left Alliance. When Law and Justice came to power in 2005, there was thus much to be gained by the new government’s determination relentlessly to fight corruption – mainly among the former governing party’s ranks. Indeed, it was announced shortly after the election that an ‘elite group of 500 officers’ will be tasked to verify politicians’ asset declarations and investigate suspicions of corruption (FH 2006). The bill on the CBA was approved in May 2006, setting up the Bureau as one of Poland’s five special services. In addition to preventive and coordination tasks, it was given a mandate to fight ‘any activity which may endanger the State's economic interests’14 – a formulation that gave the agency wide discretion in what (and whom) it chooses to investigate. CBA was also granted strong police powers, including the right to monitor accounts and bank transactions, telephone conversations, search company and private premises, detain vehicles and use firearms, and carry out covert operations. In this, CBA closely resembles ICAC’s three-pronged approach and investigative powers – without any of ICAC’s guarantees for operational independence or accountability. CBA sends a report about its activities to parliament once a year and, as is the case with the police, the courts and prosecutor offices supervise the legality of its investigations. The head of CBA reports directly to the prime minister, who also has the right to appoint and recall him/her, with the consent of the President of the Republic, the Committee for Special Services and the Parliamentary Special Services Committee. 15 Law and Justice appointed an active politician from among its own ranks, Mariusz Kami ski, as CBA’s first director – amidst predictable media reactions fearing that the body would be used for a witch-hunt against the opposition. From this perspective it may seem surprising that one of CBA’s first high-profile political ‘victims’ was a member of the government, agriculture minister Andzej Lepper – who was, however, leader of a, for Law and Justice, increasingly awkward coalition partner. (Subsequently allegations surfaced in the press that Lepper was set up by the CBA).16 Lepper was duly sacked by prime minister Jaroslaw Kaczynski, effectively ending the coalition which led to early parliamentary elections Law and Justice may have expected to win. The agency made front page news again when tapes implicating a female candidate of Civic Platform – Law and Justice’s main rivals in the elections – were released in the middle of the election campaign. (Kaminski denied that the timing had anything to do with the upcoming
14
See the CBA’s website at http://www.cba.gov.pl/portal/en/1/2/About.html. Ibid. 16 ‘Daily Dziennik claims Lepper was ‘set up’’ 11 July 2007, Polish Radio External Service. 15
15
elections). The affair may well have backfired for Law and Justice, and it certainly did not show CBA in a favourable light.17 The 2007 elections were eventually won by Civic Platform, and with this the political context in which the agency operated dramatically altered. Having been actively pursued by CBA, and given Law and Justice openly appointing one of their own to head the agency in 2006, the new Civic Platform government may have wanted to replace Kaminski with a ‘friendlier’ successor. However, the removal of the head of the agency would have required consulting, among others, the president of the republic (a Law and Justice politician, the former prime minister’s twin, Lech Kaczynski), and so Kaminski stayed on as head of CBA. The agency was frequently criticised as ineffective and overly partisan, and in June 2009 a Constitutional Court ruling found part of the law setting up CBA, including the wide definition of corruption, unconstitutional, and required the amendment of the law. CBA however seemed to carry on regardless - including a blotched operation in October 2009, which failed to implicate former president Aleksander Kwasniewski of the Democratic Left Alliance. But then in the same month, a CBA investigation did implicate several senior Civic Platform politicians in the so-called Blackjack affair, finding that members of Civil Platform were bribed to influence legislation to avoid a planned increase in taxes levied on the gambling industry. This proved a bridge too far. While accepting the resignation of several of his ministers, Prime Minister Tusk also put in motion the procedure to terminate Kaminski’s appointment as head of CBA, claiming that Kaminski abused his position, and through him Law and Justice used CBA to trip up Civil Platform.18 Kaminski and president Kaczynski both protested against the move, with Kaminski claiming that CBA was only carrying out its mandate and his removal was nothing but revenge for exposing corruption within Civic Platform. 19 Civic Platform supported the investigation of the Blackjack affair by a special parliamentary commission, and rather than finding another active politician appointed a professional, the former director of Poland’s Central Investigation Bureau, as CBA’s new head. The controversy around the agency then abated somewhat, but new leadership alone may not be sufficient for turning around public perceptions of the agency as a former government party’s partisan creation, or of appeasing the many powerful enemies CBA made over the few short years of its existence. Conclusion What do these stories tell us about the establishment and development trajectories of anticorruption agencies? In all three cases, as in most others known from the literature, isomorphism played a major role in the choice of ACAs as a policy response: agency creation was seen as ‘the’ anti-corruption intervention, because the model has been so widely diffused and legitimated internationally. Another common element is high-profile corruption cases providing the spark for the creation of agencies. In this sense, the setting up of agencies is always about the politics of
17
The candidate, immediately dropped by Civic Platform, had been exposed by an undercover CBA agent posing as a businessman, and the whole affair culminated in a dramatic press conference where the tearful ex-candidate begged Kami ski “not to lynch her in public”. ‘Corruption and privatisation dominate final days of Polish election campaign’, 18 October 2007, Polish Radio External Service. 18 ‘More ministers resign over Blackjack affair’, Polish Radio External Service, 7 October 2009. 19 The legality of his dismissal by PM Tusk hinged on a technicality which Kaminski contested in court. The controversy centred on whether it was lawful to remove Kaminski’s security certificate – a move that (would have) required the president’s opinion. Once the certificate is removed dismissal is the prime minister’s prerogative.
16
scandal. Finally, it is clear that in CEE as elsewhere ACAs have a difficult life, being subjected to repeated attacks from politicians seeking to assert their influence. The three CEE cases also tell us that the spark of scandals only ‘ignites’ if other conditions are also present – strong external pressure in the form of conditionality, as was the case in Latvia and Slovenia, and/or policy windows opened by party politics, more specifically, political parties playing the corruption card in response to scandals in an election, as it happened in Latvia and Poland. The different roots, external or domestic, in turn have strong implications for the official mandate and formal independence of the agencies set up. The Latvian and Polish agencies were both given wide-ranging powers (including investigative powers), but little distance from the political decision-makers of the day – although more in Latvia, where the need to send a credible signal to the EU played a strong role in addition to domestic politics, than in Poland where party politics provided the main motivation. In contrast, the Slovenian commission was set up with stronger guarantees of independence but also a narrower scope of mandate in response to conditionality. Yet, of the three, only in the Polish case is the agency’s initial official statute a good predictor of its development trajectory. The Polish government succeeded (at least temporarily) in ‘calibrating’ the agency in a way to serve partisan interests. While the party setting it up was in power, CBA seems to have functioned as an extended arm of the government – thus enjoying government support and pursuing aims perfectly aligned with those of its political pay-masters. Following a change of government, the agency’s partisan leadership had powerful allies (notably the president of the republic) allowing it to continue in the same vein for some time, but the inevitable clash with the new government eventually arrived, and ended with the government’s predictable victory. Given its weak autonomy, the strongest determinant of the ACA’s fate was, and remains, the attitude of ruling parties towards it. It remains to be seen if the professionalisation of the agency will give it sufficient distance from politics to smooth out the roller-coaster ride at least a little. On the other hand, the Slovene commission is not a paper tiger, and the Latvian bureau not an attack dog, as may have been expected on the basis of their formal initial mandate and independence alone. In contrast to CBA, both agencies developed as legitimate, trusted actors in their own right that pursue their goals with a high degree of de facto operational autonomy. Feedback effects are particularly strong in the Slovene case, where the commission survived a long period of stagnation to became an influential player in the prolonged process of drawing and redrawing its formal mandate, until the recently adopted law finally settled its status. This is particularly notable given the Commission’s small size and, in comparison with CBA and KNAB, limited powers. Both KNAB and CBC ‘regulated’ political parties and/or individual politicians in what appears to be an unbiased way, even through governments had a wide array of methods at their disposal to control the agencies, such as the appointment and dismissal of the head of the agency, a particularly strong control method in Latvia, or severe budget cuts and obstruction by other public bodies, as was the case, for some time, in Slovenia. Both agencies earned independent ‘actorness’ by carrying out their functions in the face of, at times, strong political opposition from governments, benefiting from strong leadership and reliance on public support and/or international support garnered through professional networks. Their experience suggests that an ACA’s official mandate influences but does not determine how it fares in later life. To summarise the broader implications of this paper, the three CEE countries’ experience suggests that the decision to delegate to an ACA, and the organisational form the agency would take, is a function of contradictory pressures between, on the one hand, needing to make a credible commitment to meet an external actor’s demand (e.g., EU conditionality) or to appease
17
public opinion (domestic constituents), and on the other, wanting to make sure that the agency subsequently does not turn against its creator. This is another version of the tension between autonomy and control (Christensen and Laegrid, 2007), in rather extreme – that is to say, extremely politicised – circumstances. If one takes on the central assumption of comparative politics – that political parties are goal oriented actors motivated by the pursuit of votes, office and policy influence – then the decks are strongly stacked for political control and against agency autonomy. Thus, even if ACAs are set up, the odds are that governments will create paper tigers rather than an ICAC style independent and powerful organisation that helped Hong Kong significantly reduce corruption. If, against the odds, a strong agency is in fact created, the most likely reason is the intention to highjack the agency from the outset. In other words, governments create powerful anti-corruption agencies not to bind their own hands but rather in the expectation that they can punish electoral opponents. Whether or not this happens seems to be largely determined by the integrity and ‘zealotry’ of the agency’s leadership, which leaves rather a lot to chance. If this really is the case, as it appears to be in Central and Eastern Europe, than anti-corruption agencies should only be created where a lasting consensus can emerge among politicians to delegate – in other words, in countries that do not need an anti-corruption agency in the first place.
18
References Andvig, Jens, Odd-Helgee Fjeldstad, Inge Amundsen, Tone Sissener & Tina Soreide, ‘Research on corruption: A policy oriented survey’; commissioned by Norad, December 2000. Batory, Agnes (2010) 'Post-accession malaise? EU conditionality, domestic politics and anti-corruption policy in Hungary', Global Crime, 11: 2, 164 — 177 Charron, N. (2008) ‘Mapping and Measuring the Impact of Anti-Corruption Agencies: A New Dataset for 18 Countries’ Paper prepared for the New Public Management and the Quality of Government conference. Christensen, Tom and Per Laegrid (2007), ‘Regulatory Agencies—The Challenges of Balancing Agency Autonomy and Political Control, Governance: An International Journal of Policy, Administration, and Institutions, Vol. 20, No. 3. Commission for the Prevention of Corruption (Slovenia) (2008) Report on the Work of the Commission for the Prevention of Corruption in 2008. CPC Ljubljana. de Sousa, L. (2006) European Anti-Corruption Agencies: protecting the Community’s financial interests in a knowledge based, innovative and integrated manner. Final report. Ancorage-net. de Sousa, Luís (2010) ‘Anti-corruption agencies: between empowerment and irrelevance’, Crime Law and Social Change 53:5–22 de Sousa, Luís, Peter Larmour and Barry Hindess (eds) (2009) Governments, NGOs and Anti-Corruption The New Integrity Warriors (Routledge) Delna – Transparency International Latvia (2009) Monitoring the work of the corruption prevention and Combating Bureau. TI Latvia. Dionisie, Dan and Francesco Checchi (2008), ‘Corruption and Anti-Corruption Agencies in Eastern Europe and the CIS: a Practitioners’ Experience’, UNDP Bratislava. Doig, A., Watt, D., and Williams, R. (2005). ‘Measuring “Success” in Five African Anti-Corruption Commissions. The cases of Ghana, Malawi, Tanzania, Uganda and Zambia’. U4 Research Report. Bergen: Chr. Michelsen Institute. Doig, Alan (1995),Good government and sustainable anti-corruption strategies: a role for independent anticorruption agencies?’, Public Administration and Development, 15, 151-165 Doig, Alan (2009) ‘Matching workload, management and resources: setting the context for ‘effective’ anticorruption commissions’ in Luís de Sousa, Peter Larmour and Barry Hindess (eds) (2009) Governments, NGOs and Anti-Corruption: The New Integrity Warriors (Routledge) pp. 65-85. Freedom House (various editions) Nations in Transit, available at http://www.freedomhouse.org/template.cfm?page=17&year=2005 Gilardi Fabrizio (2008) Delegation in the Regulatory State: Independent Regulatory Agencies in Western Europe. Edward Elgar Heilbrunn J.R. (2006), "Anti-corruption commissions", in R. Stapenhurst, N. Johnston, R. Pelizzo (Eds), The role of parliament in curbing corruption, The World Bank, Washington, DC, pp.135 - 148. Holzinger, Katharina and Christoph Knill. 2005. ‘Causes and Conditions of Cross-National Policy Convergence’, Journal of European Public Policy 12(5): 775-796.
19
Hood, C., O. James and C. Scott (2000) ‘Regulation of government: Has it increased, is it increasing, should it be diminished?’ Public Administration 78(2) 283-304 Hood, Christopher, Colin Scott, Oliver James, George Jones, and Tony Travers (1999) Regulation inside government. Oxford: Oxford University Press Hood, Christopher, Oliver James, George Jones, Colin Scott and. Tony Travers (1998) ‘Where New Public Management. Meets the Audit Explosion’. Public Money & Management, Volume 18, Issue 2 April 1998 , pages 61 – 68 James, Oliver (2000) ‘'Regulation inside Government: Public Interest Justifications and Regulatory Failures', Public Administration, 78, 2, 2000, pp. 327-343. Jordana, Jacint and David Levi-Faur (eds.) (2004a) The Politics of Regulation: Institutions and Regulatory Reforms for the Governance Age, Edward Elgar and the CRC Series on Regulation, the University of Manchester. Jordana, Jacint and David Levi-Faur (2004b) ‘The Politics of Regulation in the age of governance’, in Jordana and Levi-Faur, The politics of regulation, pp 1-24. Klitgaard, Robert (1988) Controlling Corruption. Berkeley and Los Angeles: University of. California Press Lægreid, P., Roness, P. G. and Rubecksen, K. (2008) ‘Controlling Regulatory Agencies’. Scandinavian Political Studies, 31: 1–26. Levi-Faur, David (2004) ‘Comparative research designs in the study of regulation: how to increase the number of cases without compromising the strengths of case-oriented analysis’, in Jacint Jordana and David Levi-Faur (eds.), The Politics of Regulation: Institutions and Regulatory Reforms for the Governance Age, Edward Elgar and the CRC Series on Regulation, the University of Manchester (pp. 177200) Maor, Moshe (2004) ‘Feeling the Heat? Anticorruption Mechanisms in Comparative Perspective’, Governance 17: 1, pp. 1-28 Meagher, Patrick (2005) ‘Anti-corruption Agencies: Rhetoric versus Reality’, Journal of Policy Reform 8: 1, pp. 69-103. OECD (2007) Specialised Anti-Corruption Institutions: Review Of Models Paris: OECD Publishing Open Society Institute (2002) Monitoring the EU Accession Process: Corruption and Anti-corruption Policy. Budapest: Open Society Institute. Pope, Jeremy (2000) TI Source Book 2000. Confronting Corruption: The Elements of a National Integrity System. Berlin: Transparency International. Rose-Ackerman, Susan (1999) Corruption and Government: Causes, Consequences and Reform. Cambridge: Cambridge University Press. Rothstein, Bo, Anti-Corruption: A Big-Bang Theory. QoG Working Paper No. 2007:3. Samson, S. (2004) Integrity Warriors: Global Morality and the Anti-Corruption Movement in the Balkans” in D. Haller and C. Shore (eds) Corruption: Anthropological Perspectives. London and Ann Arbor: Pluto Press.
20
Stone, Diane (2004). ‘Transfer agents and global networks in the ‘transnationalization’ of policy’, Journal of European Public Policy 11 (3): 545–566. Szczerbiak, Aleks (2005) ‘Europe and the September/October 2005 Polish parliamentary and presidential elections’, EPERN election briefing 22; http://www.sussex.ac.uk/sei/documents/epern_eb_22_poland.pdf Thatcher, Mark and Stone Sweet, Alec (2002) ‘Theory and practice of delegation to non-majoritarian institutions’ West European politics, 25 (1). pp. 1-22. Utrecht School of Governance (2008) Catalogue of promising practices in the field of integrity,anticorruption and administrative measures against organized crime in the EU. Ministry of the Interior and Kingdom Relations of the Netherlands. Wilks, Stephen and Bartle, Ian (2002) 'The Unanticipated Consequences of Creating Independent Competition Agencies', West European Politics, 25: 1, 148 — 172 Yesilkagit, Kutsal and Sandra van Thiel, ‘Political Influence and Bureaucratic Autonomy’, Public Organiz Rev (2008) 8:137–153.
21
