The Changing Faces of Europeanisation: How Did the European Union Influence Corruption in Slovakia Before and After Accession?

The paper looks at the European Union's impact on corruption in Slovakia, both before and after the country's accession to the European Union. It shows that even though pre-accession channels of influence diminish or disappear, membership brings new means of influence. Our methodology looks beyond institutional changes and explicit anti-corruption mechanisms. Before 1999, the European Union's influence on corruption was close to zero, and between 1999 and 2004, its influence was strong and positive. Post-accession developments are more ambiguous, but despite an apparent worsening of corruption after 2004, we observe a transformation rather than a loss of European Union influence.

THE ENLARGEMENT OF THE EUROPEAN UNION (EU) TO THE EAST has been accompanied by scepticism about the ability and willingness of the post-communist countries to fully absorb and implement EU institutional requirements. Initially, this fear manifested itself in an accession process of unprecedented length and intrusiveness, with the Union using the general Copenhagen accession criteria to push institutional prescriptions, even in areas such as the civil service or corruption control, where there was no (or little) acquis to rely on (Grabbe 2003). The actual accession of the ten post-communist countries in 2004 and 2007 brought worries about potential backsliding and weakening EU influence once membership was assured.
Post-accession research has produced mixed findings. While the new members continued to transpose EU legislation assiduously, their lacklustre performance on implementation of EU legislation led to the conceptualisation of a new model of compliance called the 'world of dead letters' (Falkner & Treib 2008). Additionally, there seemed to be an actual backsliding in areas without an explicit European legal framework, such as civil service reform (Meyer-Sahling 2009).
In this paper, we examine the changing faces of Europeanisation in the area of corruption before and after accession. Corruption is a key concern with regard to new member states and, due to the limited formal acquis, it is an excellent candidate for reversal (or at least some backsliding) following accession. We contend that while it is true that pre-accession channels of influence do disappear or diminish, membership brings new channels that should also be taken into account. To obtain an accurate picture of the EU's influence, one needs to look beyond regulation or even its implementation.
The paper consists of five sections. After the introduction, we start with the view from the literature dealing with the EU's influence on domestic institutions in general, and specifically in the area of corruption prior to and following accession. After discussing the methodology, the paper presents its findings on the five potential forms of EU influence on Slovak corruption. The article concludes with a synthesis of the findings and a discussion of their potential wider application.
EU influence on corruption and domestic institutions before and after accession: the view from the literature In this section, we first provide a brief conceptualisation of corruption for the purposes of the article, then look at what the literature has to say about the process of Europeanisation during EU accession and in its aftermath, focusing on the area of corruption in the new member states as a group. We pay particular attention to how the EU conceptualised the fight against corruption in its prescriptions for the candidate countries, and whether the issue of backsliding has emerged since accession, both in general and specifically with regard to corruption.
There is no universally accepted definition of corruption, as it is a phenomenon that takes various forms, depending on time and place. Miller (2006) implies that it is like pornography-difficult to define, but easy to recognise. In practical policy making, two approaches prevail. One looks at corruption in terms of legality. The problem here is that such a definition is endogenous, since policy makers (and law enforcement as well as the judiciary) in any given country determine what is legal and what is not.
The other approach is to emphasise the distinctions between the public and private roles of an individual and to define corruption as an abuse of public office for private gain. Another formulation of this view is that corruption occurs when an official acts according to private rather than public interest. This approach can also be reconciled with the economic perspective of the principal -agent theory, where acts of corruption are generally unobservable and performed by agents (officials) that contradict the interests of the principal (the electorate) to provide gain for the agents (Becker & Stigler 1974). As Rose-Ackermann (1999) points out, however, these definitions presuppose a clear distinction between the public and private roles of a given person. This condition is not fulfilled in many countries, but is a limited problem in the new member states where the distinction between public and private roles is relatively well established.
The term 'Europeanisation' has many definitions and levels of application and can be rather nebulous, not to mention highly contested (see Mény et al. (1996) for a discussion; also Radaelli (2000)). In this paper, we are concerned with what Olsen (2002) identified as one of the five faces of Europeanisation-the process of change affecting the domestic situation in those countries that are EU members. The focus is on how the European Union 'hits' the member states (Trondal 2002), both prior to and during membership. In this context some authors (for example Wallace 2000) prefer the term EU-isation.
Europeanisation influences member states through both coercive and non-coercive instruments. Knill and Lehmkuhl (1999) recognise three mechanisms of Europeanisation. The first is prescriptive, where specific institutional requirements are prescribed. The second alters domestic opportunity structures. The third modifies the beliefs and expectations of domestic actors. It should be noted that even though these authors discuss policies, their intellectual framework is in practice largely concerned with regulatory action and 'soft' instruments such as policy documents without legislation; this is an approach that will prove to be somewhat limiting in our case. Sedelmeier (2004, 2005) use a similar, albeit distinct conceptual framework from Knill and Lehmkuhl (1999) to explain Europeanisation in the area of rule transfer to the Central and Eastern European candidate countries prior to their accession. They conclude that an 'external incentives model of governance' is the best explanation for rule transfer to these countries as opposed to two alternative explanations-social learning and lesson-drawing models. The external incentives model, unlike the other two, is coercive in the sense that the reward of EU membership is conditional on the rule transfer. They also conclude that conditionality generally worked strongly with regard to the so-called acquis conditionality (Schimmelfennig & Sedelmeier 2004, p. 672).
In practical terms, this conditionality works through a formalised enlargement pathway. As a part of this pathway, developed in the 1990s, the EU not only conceptualised accession stages based on membership conditionality, 1 but also hardened the conditions for joining. The membership path for post-communist states has thus been relatively onerous compared to the past (Grabbe 2003).
Specifically with regard to corruption, the Copenhagen mandate enabled the European Commission (EC) to push the candidate countries to adopt various anti-corruption policies. Despite the somewhat vague conceptualisation of the Copenhagen accession criteria, the Commission developed detailed prescriptions, which it enforced during the accession process (OSI 2002). The anti-corruption measures are defined within the chapter on Justice and Home Affairs and are the so-called 'soft' anti-corruption acquis. They concerned, for example, harmonisation of definitions of corruption within the Penal Code and making bribery of foreign officials a penal offence. The 'soft' anti-corruption acquis also includes a number of international agreements, for example the Council of Europe Criminal Law Convention on Corruption (Council of Europe 1999a) and its Civil Law Convention on Corruption (Council of Europe 1999b), as well as the European Convention on Laundering, Search, Seizure and Confiscation of the Proceeds from Crime and on the Financing of Terrorism (Council of Europe 2005) and the OECD Convention on Combating Bribery of Foreign Public Officials in International Business Transactions (OECD 2011).
Other chapters of the acquis contain a number of legal requirements relevant for corruption control in areas such as public procurement (candidate countries had to adopt procurement directives), requirements concerning judicial reform (the European Commission pushed for judicial independence and overall reform of the judiciary), financial control and audit (the Commission asked for changes in the system of financial control to provide some assurance for the use of European funds) and civil service reform. 1 Democratic (political) conditionality was first formally raised in 1962 by the Birkelbach Report of the European Parliament, which became official policy in relation to Spain's interest in joining the then EEC (Pridham 2002). Corruption in EU candidate countries has been one of the European Union's biggest concerns since its initial 1997 assessment in the 'Agenda 2000' report on CEE countries' application for membership (OSI 2002). The Commission released its Regular Reports concerning each candidate's progress towards accession, including an assessment of corruption. Regular Reports constituted one of the basic tools of the Commission when challenging corruption practices in candidate countries and when asking for improvements. When developing Regular Reports, the Commission used a checklist on monitoring corruption that focused not only on the existence and implementation of anti-corruption policy, but also on institutional arrangements for implementation and division of tasks among institutions; codes of conduct and training programmes for public servants; cases of corruption in government and public administration and how the authorities react to these cases; and ratification and implementation of the relevant international conventions (OSI 2002). However, as we will show in this research, this was used as an ex-post evaluation tool of progress without clear ex-ante signals on the need to pass specific measures (with very few exceptions).
Pre-accession Europeanisation can result in a divergence between formal and informal rule transfer and between short term achievement and long term sustainability (Milanese 2001). Following accession, the former pre-accession relationship that relied on coercive instruments and enhanced the superiority of the EU is being supplemented by new forms, and the one-way reception of EU laws and policies is being replaced by softer forms of integration (Bulmer & Burch 2000;Dyson & Goetz 2003).
Empirical studies of post-accession behaviour have not demonstrated major problems or any regression in transposing legislation (e.g. Dimitrova & Toshkov 2009;Sedelmeier 2009;Trauner 2009). On the contrary, the new member states have remained among the most assiduous compliers with EU requirements in regard to legislation-even those countries considered to be most problematic members overall by the European Commission, notably Bulgaria or Romania (Gateva 2010;Trauner 2009).
However, two problems have arisen. First of all, analyses of civil service reform in postcommunist new member states showed that there is measurable institutional backsliding in this area. This example demonstrates what can happen to pre-accession requirements imposed without clear reference to the legally binding acquis (Meyer-Sahling 2009). This is highly relevant for our research since the anti-corruption acquis is also largely of this 'soft' nature. Secondly, even when there is a requirement to transpose legislation, this can be ineffective if there is no or insufficient enforcement. Falkner and Treib (2008) talk about a specific model of compliance in new member states with regard to working time and equal treatment, which they call the 'world of dead letters' and, in their view, includes, in addition to all four new member states they studied, Ireland and Italy.
To partially compensate for these weaknesses, the EU has negotiated more extensive asymmetric powers vis-à-vis the new entrants for the second wave (Bulgarian and Romanian) accession in 2007. These powers, embodied primarily in the so-called Cooperation and Verification Mechanism, included both the enhanced powers of the Commission with regard to European funding provided to the member states, and the more extreme and temporary power to suspend recognition of judicial decisions by the new member states after accession if their judicial system is deemed to be unsatisfactory (Gateva 2010). More recently and informally, several large member states, including France and Germany, have been making entry into the Schengen zone conditional on improvements in the area of corruption. Gray (2009) notes that the EU has also had an impact on financing costs for national governments, with impending and actual EU membership associated with substantial drops in perceived credit risk. Her quantitative analysis shows that a large part of this effect is related to the 'seal of approval' phenomenon, whereby the Union endorses a pre-existing policy mix. She speculates that removal of the 'seal of approval' through explicit negative messages could serve as a disciplining device for members. There is some anecdotal evidence to support this point (e.g. Hungary in the second half of 2010), but it is insufficiently conclusive, as the global crisis made EU institutions extremely reluctant to further disadvantage the most fragile member states. Table 1 shows developments in corruption perception for the new EU member states prior to and after accession. The data used are the Corruption Perception Index (CPI), a metaindex compiled from a variety of other sources that measure perceptions of corruption in a given country. While the CPI has its limits (having to do with the fact that it is based on perception rather than the reality of corruption) and should be carefully interpreted as shown by Anderson and Heywood (2009), it is broadly based, robust and the only available source of corruption data on an annual basis for all new member states over the relevant period. Warner (2002) documents why a comparative attempt to measure corruption levels in individual EU countries through objective measures-ranging from reports of EU-level institutions to national prosecutions or discoveries of irregularities in the use of EU funding-is largely futile. Lambsdorff (1998) and Montinola (1997, 2001) make a convincing argument for why CPI is successful in overcoming most of the problems related to perception-based indices.
CPI for a given year relies on surveys conducted over the previous two to three years. Therefore, we chose the 2006 CPI as the best approximation of 2004 accession data. We chose 1999 because it is the first year when data are available for all countries and also because 1997 -1998 was the period when accession negotiations started in earnest with the first group of post-communist states. The 2011 CPI is the latest one available, reflecting the situation between 2009 and 2011. We can indeed see broadly based improvement in the pre-accession period, with eight out of ten accession countries seeing improvement in their score. We can also observe some backsliding in the post-accession period, with only two out of ten countries registering improvements and seven out of ten experiencing worsening of the score. Only Poland saw a major improvement after accession.
The literature and the data show that the general worries about post-accession backsliding are not unfounded, but the real picture is much more complex. The new members continue to perform well on the transposition of legislation, but badly on implementation and also in areas without an explicit European legal framework. Specifically with regard to corruption, the perception of backsliding is borne out by data, but that does not show a causal relationship. It is worth noting that the existing research devotes little attention to the role of other potentially important, but indirect, channels of EU influence, namely the Structural Funds and the direct EU powers in some areas.

Methodology
This article is a case study as defined by Gerring-'an intensive study of a single unit with an aim to generalize across a larger set of units' (Gerring 2004, p. 341). Single country design was chosen to allow a holistic view of the EU influence in one area, comprising all major policy instruments and comparing the situation prior to and after accession.
As noted by Yin (2003), case studies are appropriate instruments for social science research when the phenomenon being studied is highly complex and 'when "how" or "why" questions are being posed, when the investigator has little control over events, and when the focus is on a contemporary phenomenon with some real-life context' (Yin 2003, p. 1). In such a context, Gerring (2004) and Van Evera (1997) emphasise the complementarity of single-unit and cross-unit research designs rather than the superiority of one or the other.
We started by identifying the potential channels of influence through which the Union and its institutions influenced the situation in Slovakia with regard to corruption. Our approach was inductive rather than deductive. We reviewed in-depth research published on corruption in Slovakia and, to a limited extent, in other Central Eastern European countries (Beblavý 2009;Beblavý & Mesežnikov 2009;Grødeland 2007;Kollár et al. 2006;Mesežnikov & Kollár 2003;OSI 2002;TIS 2007). These were complemented by literature on Slovak political developments, EU accession and the use of conditionality during the late 1990s and 2000s (Haughton 2003(Haughton , 2007Pridham 2008).
The academic sources were complemented by use of the European Commission's Regular Reports on Slovakia's Progress towards Accession from 1999 to 2003, as the principal EU instrument for assessing a candidate country's progress in a given year (European Commission 1999. However, there are important gaps in existing research, which are filled through a limited number of personal interviews and also the use of newspaper accounts where appropriate. We were able to identify five channels of influence: political/democratic conditionality and composition of the government; policy conditionality and anti-corruption policy; anticorruption technical assistance; Structural Funds and other EU funding; and direct EU intervention related to competition in the single market (state aid, procurement).
Other potential channels-for example the role of the European Commission European Anti-Fraud Office (Office européen de lutte antifraude-OLAF) or the effects of EU-funded research and advocacy-can either be subsumed into one of the five areas above or have not been identified as relevant in any of the periods we examine.
To link our inductive categorisation with the literature and clarify our conceptualisation, we place, in Table 2, the five channels into the framework by Knill and Lehmkuhl (1999) on how Europeanisation influences the domestic situation. The table has three rows, one for each mechanism of Europeanisation as identified by Knill and Lehmkuhl. We see that it is relatively easy to identify three of the channels as prescriptive. However, financial aid and technical assistance are not so easily classified in our view, since they both alter domestic opportunity structures and contribute to changing beliefs and expectations of domestic actors. We also felt it was important to distinguish between mechanisms with an explicit focus on corruption/anti-corruption and those with other apparent objectives.
For each of the researched potential channels described above, we present an analysis of developments during the four stages of the accession calendar: pre-negotiation (before 1999); pre-accession (1999)(2000)(2001)(2002)(2003)(2004); immediate post-accession (2004)(2005)(2006); and later postaccession (2006)(2007)(2008)(2009)(2010). This division is a slight adaptation of the stage-structured conditionality model (Gateva 2010), taking into account the Slovak electoral cycle. In the pre-accession phase, the Mečiar government (1994)(1995)(1996)(1997)(1998)  The paper rates all five channels of influence, assessing their positive (þ to þþþ), negative (2 to 2 2 2) or neutral/irrelevant (0) contribution to corruption in Slovakia. It then aggregates the ratings to form a comprehensive picture of EU influence and demonstrates how this has changed after accession. To alleviate the problem of weighting all areas equally, we present a differentiated number of positive or negative signs depending on our assessment of the importance of a given area for corruption developments in Slovakia. It should be noted that the added value of the paper is not primarily in the quantitative ratings, as they are only a useful abbreviation of the narrative analysis presented in the following text. Ratings are based on the utilisation of the same sources that were used for identification of the channels themselves (see above).

Evaluation of potential channels of EU influence on corruption in Slovakia
Political conditionality: composition of the government During the 1994-1998 period, Slovakia was ruled by a government coalition dominated by the Movement for Democratic Slovakia (Hnutie za demokratické Slovensko) of Prime Minister Vladimír Mečiar, together with the xenophobic Slovak National Party (Slovenská národná strana) of Ján Slota and the ultra-leftist Association of Slovak Workers (Zväz robotníkov Slovenska) of Ján L'upták. As a result of government policy, Slovakia became excluded from further integration into the key Western 'clubs'-the European Union, NATO and OECD. This negative development was not caused by corruption per se, but rather by an overall concern about the state of democracy and the rule of law (European Commission 1999, p. 11). Slovakia thus became a regional integration laggard, on a par with, for example, Croatia. The EU thus imposed a double political conditionality on Slovakia. At first, it tried to make accession conditional on a change of behaviour by the Mečiar government. In this respect it was unsuccessful, for, in what (Pridham 2002, p. 209) calls a 'clear-cut instance of failed response to democratic conditionality', the government refused to modify its behaviour. As a result, the European Union effectively imposed political conditionality, whereby further integration was made conditional on the exclusion of Vladimír Mečiar and his party from government (Schimmelfennig & Sedelmeier 2004;Schimmelfennig et al. 2003). This had an important influence on both electoral results and coalition composition in Slovakia. Fisher and Haughton (2008) explain the link between government policies, EU conditionality and the electoral results in Croatia and Slovakia as follows: 'Where appeals to the nation were combined with illiberalism and were received unfavourably by strategically important international clubs, public support for such parties declined' (Fisher & Haughton 2008).
While such conditionality could not be and was not legally binding (unlike the acquis), it was made quite explicit in political communication (Pridham 2002). As a result, Mečiar and his party remained in opposition during the subsequent 1998 -2006 period (as did the Slovak National Party, while the Association of Slovak Workers disappeared from the political scene). The effectiveness of this conditionality can be seen both in its effect prior to the accession and its disappearance afterwards. In 1998, a complicated coalition of both left and right was assembled to exclude Vladimír Mečiar from the government, while in 2002, there was a readiness to continue this coalition with a new left-wing party SMER (Direction) rather than contemplate a government including Mečiar (in the end, this was not necessary thanks to an unexpectedly good showing of the centre-right parties).
After the accession-in 2006-both major parties (the Slovak Democratic and Christian Union -Democratic Party (Slovenská demokratická a kresťanská únia -Demokratická strana-SDKÚ -DS) for the centre-right and Direction -Social Democracy (SMER -Sociálna demokracia-SMER -SD) for the centre-left), publicly endorsed coalition with Vladimír Mečiar, and consequently SMER-SD successfully formed a coalition government with the parties of both Mečiar and Slota. In other words, no mainstream party recognised or accepted Mečiar and Slota as coalition partners prior to accession, but both SDKÚ -DS and SMER-SD found them acceptable afterwards. This shift in politics is a strong indication of the effectiveness of the political conditionality vis-àvis government composition.
Since the Movement for Democratic Slovakia and the Slovak National Party and their leaders were prone to political clientelism and corruption in comparison with major political parties, their presence in the government or exclusion from it had a significant influence on the level of corruption in Slovakia. Provision of thorough evidence for this claim would greatly exceed the space available here. However, the following anecdotal evidence can be given. Firstly, at the end of the Mečiar government, the gap between outsiders' perceptions of corruption in Slovakia and in the rest of Central Europe according to Transparency International was the highest in history-3.7 for Slovakia versus 4.6 points of the Corruption Perceptions Index for the other three Central European countries in 1999 (Transparency International 1999). Secondly, interviews with 20 members of the political and economic elite in 2009 showed that there is a dominant view among them that corruption in this period was higher than subsequently (Beblavý & Mesežnikov 2009). Thirdly, during the lifetime of the Fico government (2006)(2007)(2008)(2009)(2010), there were two corruption scandals leading to a minister's resignation in the case of Mečiar's Movement for Democratic Slovakia and a further five concerning Slota's Slovak National Party, but only one case for the dominant SMER-SD. This is despite the fact that the two smaller parties together held only five out of 16 ministerial seats (TASR et al. 2009). Finally, amongst the leaders of six major political parties in Slovakia, Mečiar and Slota were considered to be the richest politicians by a wide margin (SITA 2010).
Therefore, regardless of domestic dynamics, EU accession had an important positive effect on the corruption developments in Slovakia by temporarily excluding these two individuals and their parties from the government. The other positive effect pointed out by Pridham (2002) could be noticed in the new scope that the opposition acquired to reassert itself and defend democratisation. As shown later in the article, this reassertion also entailed an obligation to fight corruption. From this point of view, the political conditionality regarding the composition of the government can be seen as an indirect positive yet unsustainable channel of influence.

Policy conditionality-anti-corruption policy
During the Mečiar government (1994)(1995)(1996)(1997)(1998), EU conditionality with regard to specific anticorruption policies was ineffective due to the reasons described above. This altered after the change of government in 1998 and the subsequent opening of accession negotiations in 1999. The anti-corruption measures were defined in the chapter on Justice and Home Affairs and are described above in the methodology section of this article. All the requirements were fulfilled by Slovakia prior to accession. Therefore, policy conditionality clearly worked in getting Slovakia to adopt the required legislation and policies. However, it is difficult to argue that these measures had a major impact on the state of Slovak corruption. An evaluation of 12 major anti-corruption reforms in Slovakia during the 1998 -2006 period identified only one of these reforms as linked to accession conditionality (changes in financial control and audit at central and local government level) and this reform was rated as having 'low' impact (Beblavý 2009). Some of these other measures were also mentioned by the Regular Reports (for example in the second regular monitoring report Slovakia was praised for adopting anti-corruption measures (Pridham 2002), but always in an ex-post fashion rather than as ex-ante requirements (European Commission 1999. The slew of anti-corruption programmes and reforms introduced by the two Dzurinda governments was tightly linked to its domestic policy agenda (Beblavý 2009). In June 2000, as a result of early corruption scandals within the new government that resulted in public outcry, the first Dzurinda government approved and established a National Programme for the Fight against Corruption (Národný program boja proti korupcii), developed by the Office of the Deputy Prime Minister for Economic Policy, Ivan Mikloš, who also became responsible for the anti-corruption agenda. The National Programme for the Fight against Corruption endorsed a reduction in discretion and increase in transparency, but it also sought to improve the functioning of the judiciary and prosecution, combat conflict of interest at all levels of government, and change norms of behaviour through education and public information campaigns. After the 2002 election, a second Dzurinda-led government came to power. The anti-corruption agenda, however, was shifted to another minister-Daniel Lipšic who served (between 2002 and 2006) as the Deputy Prime Minister for Legislation and the Minister of Justice. In May 2003, the government adopted Lipšic's proposal for a more technical document laying out specific legal measures to implement the government anti-corruption objectives (Mesežnikov & Kollár 2003, p. 644).
The anti-corruption strategy of both Dzurinda governments combined specific sectoral reforms (e.g. bank privatisation) with horizontal policies (e.g. freedom of information legislation). Both types of policy change took place within an overall environment of rapid reforms in Slovakia on all fronts after 1998. This changed political environment was the result of many factors, but EU accession and fiscal pressures were among the most relevant. Rapid policy change contributed to an overall feeling that when change is possible in so many areas, it should also be possible to do something about corruption. More importantly, deep structural reforms of the key elements of the public sector made it possible to remove or diminish, at low political cost, some of the structural underpinnings of corruption.
Therefore, one can argue that the pressure exerted by the Commission with regard to explicit anti-corruption policies was a minor element of the anti-corruption reforms undertaken in Slovakia in the pre-accession period, though the general emphasis placed on the issue by the Regular Reports created an imperative 'to do something' stronger.
After accession, the implementation of requirements following from the 'soft acquis' slowed down or even stopped in some cases, but did not go into reverse.  show that for the approval and ratification of the Council of Europe's anticorruption conventions, EU political conditionality was crucial. However, once Slovakia became a member, the urgency of implementation receded. Specifically, in April 1999 the Slovak Republic became one of the founding members of the Group of States against Corruption (GRECO), the Council of Europe's anti-corruption monitoring body. GRECO's objective is to monitor its members' compliance with Council of Europe anti-corruption standards.
As of January 2011, Slovakia had completed three GRECO evaluation rounds. In the first round, GRECO concluded in 2003 that 15 out of 19 recommendations had been implemented. During this period, the European Commission closely monitored the implementation of GRECO recommendations. To fulfil the first round of GRECO recommendations did not require any significant changes in legislation, and many of the recommendations were defined in more general terms-such as to create an institution to fight corruption. In the second evaluation round in 2006, Slovakia fully implemented only eight out of the 17 GRECO recommendations while in the third round (held in 2010), it fully implemented only one out of 16. This fact can be attributed not just to the loss of EU conditionality, but also to the progressively more 'difficult' nature of the recommendations and an increase in domestic compliance costs . Pridham (2008) states that EU pressure in the post-accession period was evident through the initiation of new policies and the establishment of new agencies, but that its relevance was rather symbolic and limited. The EU has remained largely silent, including over the abolition of the civil service office established to protect the profession from political interference by introducing open competition for posts (Pridham 2008). Therefore, with regard to the effect of EU conditionality on anti-corruption policy in Slovakia, we can conclude that firstly the EU was successful in imposing its policy preferences in anti-corruption policy prior to the accession, but less so after the accession; and secondly that the EU requirements and subsequent rule transfer constituted only one element among many that impacted on corruption in Slovakia during the period.

Anti-corruption technical assistance
Technical assistance for anti-corruption reforms is another means through which the Union can influence anti-corruption policies and their implementation and, ultimately, the state of corruption in a country. The impact increases with the number of reforms thus supported and their individual importance.
Until 1999, EU-funded projects in Slovakia were neither directly oriented towards structural reforms in this area, nor did they result in any anti-corruption regulations. 2 We therefore cannot speak about EU influence in the form of policy transfer and learning as a factor decreasing corruption in this period.
In the pre-accession period (1999)(2000)(2001)(2002)(2003)(2004), several structural reforms were conducted using policy transfer and technical assistance provided by the EU. In 2001, the Commission provided a grant under the PHARE scheme to support the fight against corruption in Slovakia. 3 The project funded technical assistance to change civil service legislation, as well as an experienced Spanish prosecutor who worked as the long-term expert on the creation of a Špeciálny súd (Special Court) and Ú rad špeciálnej prokuratúry (Special Prosecutor Office), which were set up to deal with corruption investigations and trials (Kollár et al. 2 Interview with a former employee of the EC Delegation to Slovakia, Peter Muška, Bratislava, 23 May 2009. 3 PHARE (Poland and Hungary: Assistance for Restructuring their Economies) was a European Union grant scheme targeted at institution building and investment support in Central and Eastern Europe during the preaccession period. 2006, p. 672). Further examples include the development of training courses for investigative journalists and provision of assistance in the assessment and further development of data collection and processing among law enforcement agencies (PHARE 2001).
After joining the EU, the government continued to implement anti-corruption projects programmed or started earlier, but in the 2006 -2010 period it did not initiate major technical assistance projects in anti-corruption policy. 4 The only exception is a project from the EU Transition Facility, which was implemented between 2006 and 2008; however, even this project was an explicit part of a funding modality related to Slovakia's accession rather than to its regular membership. This project focused on the training of law enforcement institutions such as the police, prosecution, and courts. In this period, there were no projects financed by the EU that focused on policy transfer to conduct structural reforms and decrease corruption. Policy transfer and training is thus not significantly present in the period 2006-2010.
As mentioned above, the assessment of the role of external actors in Slovak anticorruption efforts concluded that foreign technical assistance played an important role in six out of 12 anti-corruption reforms studied, but that the European Union played a role in only one of them, which was assessed as having 'low' impact (Beblavý 2009).
EU technical assistance therefore played a moderately positive role in the 1999-2004 period, but was largely irrelevant before and thereafter.

EU financial assistance
When looking at the role of EU financial assistance in Slovak corruption, we consider the Structural Funds and pre-accession funds to be of a similar type. We proceed from a description of changes in the form and volume of aid over time to an analysis of their likely influence on the issue of corruption. The existing literature does not have much to say about the impact of EU aid on corruption, considering corruption to be constant and then assessing how high corruption influences the effectiveness of EU programmes in generating economic growth (Beugelsdijk & Eijffinger 2003;Royo 2007). We argue that a combination of volume of aid and some of its structural features make it likely that there was actually an increase in corruption.
Before 1999, the volume of EU aid flowing into Slovakia was minuscule and generally limited to the PHARE programme. The decision-making system concerning allocation of EU contracts to Slovak entities was directly managed by the Commission and the local Commission staff. In the run-up to EU accession, the flows began to be scaled up and to deliberately mimic the Structural Funds and other programmes for member states in order to prepare Slovakia for membership. Two programmes-the Special Accession Programme for Agriculture and Rural Development (SAPARD) and the Instrument for Structural Policies for Pre-Accession (ISPA)-were particularly prominent, designed to prepare for, respectively, the Common Agricultural Policy and the Cohesion Fund, which also meant that their administration was shifted to domestic agencies accredited by the EU. The gradual increase in domestic control coincided with a scandal in which accusations of impropriety 4 Interview with Marek Kalavský, a former employee of the Anticorruption Unit functioning within the Office of the Slovak Government, Bratislava, 26 May 2009. were made, by an ex-wife, against the senior civil servant responsible for aid coordination. Even though the accusations could not be substantiated, an EU audit uncovered structural weaknesses (European Commission 2001, p. 15). The scandal led to the temporary suspension of EU aid and the dismissal of Deputy Prime Minister for EU Affairs, Pavol Hamžík, from the government in 2001 (Zsilleová 2001). On the other hand, it should be noted that, from a macroeconomic and fiscal point of view, the flows were still quite low.
Accession brought two dramatic changes. The first concerned the decision-making process. As a member of the Union, Slovakia began to receive Structural Funds, thus bearing primary responsibility for programming, project selection and project management. The European Commission and other EU institutions retain powers of oversight, but these powers are significantly limited compared to the pre-accession period. Accession therefore increased the ability of various Slovak public institutions (particularly managing authorities) to use and abuse the funds.
Even more importantly, the volume of aid was scaled up dramatically-in two stages. In the 2004-2006 period (projects from this period were financed until 2008), the volume was increased to an annual flow of about e400-500 million. In the 2007 -2013 period, it jumped to e1.5-2 billion annually (TIS 2007).
The challenges were compounded by the Slovak political and policy strategy vis-à-vis the Structural Funds. Due to a combination of fiscal pressures and negative experience, Central and Eastern European countries had, prior to EU membership, gradually reduced the amount of subsidies overall and, more importantly, dramatically reduced the share of subsidies that is allocated in a discretionary manner rather than based on formulas and other rule-based approaches (TIS 2006). Structural Funds dramatically reversed this trend. In both the 2004 -2006 and the 2007-2013 period, a high percentage of the allocation of funding was vulnerable to corruption, with vulnerability defined on the basis of three criteria: political discretion in the allocation of projects, imbalances between supply and demand, and a high level of private sector involvement either through state aid or procurement (TIS 2006).
As a result, the Structural Funds gradually became a symbol of corruption in the distribution of government subsidies, with Slovakia having the dubious distinction of having a front-page article in the New York Times on the topic (Castle 2010). Table 3 shows that in the 1998-2004 period, only one minister was dismissed for corruption/abuse of power related to EU funding (Mr Hamžík, as mentioned above)-out of five ministers who had to leave for reasons related to corruption or abuse of power. During the term of the second Dzurinda government, two ministers had to resign between 2004 and 2006 because of allegations of impropriety and both of these were linked to EU funding. In the Fico government, the 'Structural Funds factor' (Sičáková-Beblavá et al. 2011) concerned three ministers out of eight who had to resign between 2006 and 2010 for impropriety (though it should be noted that an additional three ministers out of the same group of eight were dismissed because of an emissions trading scandal-also resulting from EU policies). This shows that the growth of importance of EU aid was reflected in its growing role in high-level political corruption scandals. Also, as  show in their analysis, the scandals leading to ministerial resignation were generally more serious during the Fico government than they had been during the Dzurinda governments, so they cannot be attributed to increasing sensitivity. Therefore, we see a growing negative impact of EU funding on the state of corruption in Slovakia, increasing both with decentralisation to domestic authorities and with the rapidly expanding volume of aid.
Direct EU intervention related to competition in the single market (state aid, procurement) In EU policy and legislation, issues of state aid, procurement and competition are closely intertwined and linked to one of the key objectives of the Union-to ensure smooth functioning of the single European market. 5 EU legislation subjects all state aid to scrutiny by the Commission unless it fits into one of the so-called block exemptions agreed to by the Commission and the member states. The Union also has extensive procedural legislation on major procurement actions, to ensure that there is an open competition without discrimination against suppliers from other member states. Additionally, both the Commission and the member states have an obligation to prosecute cases of anticompetitive behaviour by businesses and the power to approve or forbid mergers (Council of the European Union 2004) depending on their impact on competition in local, national and European markets. The European-level policy in this area is promulgated through direct legal instruments-regulations, which are directly enforceable and applicable in each member state. As we shall see, this transfer of executive power to Brussels was significant in a country such as Slovakia with weaker domestic institutions (unlike, for example, in Sweden, where Lennerfors (2007) also identifies impacts of the EU procurement legislation on corruption, but deals exclusively with domestic enforcement).
As Bedirhanoglu (2007) notes, the neoliberal discourse generally associates corruption with rent-seeking, and it links anti-corruption with higher levels of competition. Specifically in the EU context, the competition, procurement and state aid powers are important both in obvious and subtler ways. They introduce significant limits to the discretion of public authorities in a member state by obliging them to follow certain procedures (procurement) or even request approval by the Commission for certain actions (state aid). These powers have enabled the EU, and particularly the Commission, to shape domestic policies of the member states in far-reaching ways. 6 Examples of EU power include the following: if the Commission believes a member state did not respect the procurement legislation in a specific case, it can sue the member state in the European Court of Justice; increasing the transparency of relationships between public authorities and real estate developers by mandating the sale/ transfer of land and real estate at market prices (sales for lower prices are considered state aid); limits on the ability of governments to provide subsidies for public transport on a discretionary basis; and constraints on the ability of for-profit bodies including state-owned ones to be recipients of Structural Funds (or more precisely, constraints on conditions under which they can receive them). The importance of these rules has also increased in interaction with the Structural Funds. The primacy of public procurement in corruption in the new member states had been evident long before then (see Bohatá (1997) for the case of the Czech Republic), but it has grown in the 2000s (Grødeland 2007).
In the pre-negotiation stage, Slovakia established a competition authority (1990), modelled on what was perceived as European best practice, with EC involvement in an advisory capacity and largely based on bilateral links with national authorities in the existing member states. In 1992, Slovakia approved procurement legislation, but only with a limited scope and not related to EU requirements. Slovakia's Association Agreement with the EU in 1997 brought a more rigorous definition of anti-competitive practices based on EU legislation and brought state aid rules into the picture. However, the legislation continued to be implemented by domestic institutions and the Association Agreement was not really enforced due to political reasons.
In the pre-accession stage between 1999 and 2004, the competition, procurement and state aid acquis were transposed into domestic law. In 1999, the State Aid Authority (Ú rad pre štátnu pomoc) and the Public Procurement Authority (PPA-Ú rad pre verejné obstarávanie) were established, though enforcement was still weak and, in both instances, this can be seen as a learning period.
In the immediate post-accession stage, the fact of membership meant the demise of domestic state aid institutions, with all powers being transferred to the Commission (the State Aid Authority was abolished). On the other hand, the roles and powers of the Public Procurement Authority expanded even though the Commission also gained authority to monitor major procurement and, if necessary, initiate court cases against Slovakia. The national competition authority continued to deal with national and local issues, but the Commission could pre-empt cases of a multinational nature or even cases of a national nature where the national authority was unable to make headway.
The same trend has been apparent during the later post-accession stages, though the growing influx of Structural Funds made the powers of the Commission more relevant and intrusive. The importance of European intervention in the post-accession stage can be illustrated by three major cases that arose in Slovakia during the 2006-2010 period. These related to a tender to supply a toll system for commercial vehicles (value: e600-800 million) (Krajanová 2010); consulting services in the management of Structural Funds (value: e120 million funded from Structural Funds) (Ruttkayová 2013); and grants to so-called pilot social enterprises (eight grants worth e24 million funded from Structural Funds) (Ragáčová 2009). These cases have been dogged by allegations of clientelism and corruption and also by related allegations of infringements of EU and domestic law on public procurement and/or state aid. Domestic regulatory institutions (PPA, Ministry of Finance, Antimonopoly Office) proved to be ineffective-either refusing to become involved or unable to establish any illegality. 7 7 Grødeland (2007) found that informal networks have a strong influence on public procurement outcomes in the CEE countries.
The European Commission, however, declared that all three cases were in contravention of EU law-on public procurement in the first two instances, and on state aid in the third case. In cases where the costs were being reimbursed from the Structural Funds (the second and third case), this resulted in the cancellation of the projects by the domestic authorities at a stage where approximately 20 -30% of the expenditure was already pre-financed by the national budget. The first and largest case concerning the toll system for commercial vehicles remains under negotiation between the Slovak government and the European Commission.
We therefore conclude that EU intervention in this area has had a major impact since the accession of Slovakia in 2004. Its relevance increased during the late post-accession stage, since the burgeoning inflow of Structural Funds has increased the importance of the mechanisms concerning procurement, state aid and competition both for Slovakia and for the Commission. In the words of Ekiert et al. (2007), compliance was attractive and noncompliance visible and costly in this area.

Synthesis and conclusions
This article has examined the changing faces of Europeanisation in a post-communist state before and after its accession to the EU. To this end, we analysed the influence of five potential channels through which the European Union could have had an impact on corruption in Slovakia. In the final section, we bring together the results, but also draw some more general conclusions.
The results for Slovakia are aggregated in Table 4, which summarises the contribution of each channel to the overall state of corruption.
Before 1999, the influence of the Union on Slovak corruption was close to zero, since the Slovak government refused to change its policies under EU pressure and direct intervention therefore became possible. This changed significantly in the pre-accession period between 1999 and 2004, when the EU was able to deploy a wide range of instruments and strengthen  its influence. The political conditionality of accession meant both the exclusion of the parties most prone to corruption from the government and pressure to introduce and implement a formal anti-corruption policy. Reforms were also supported by technical assistance and the transfer of EU rules concerning competition, state aid and procurement into Slovak legislation, including the creation and strengthening of national regulators. At the same time, EU influence was limited in areas where structural reforms were most effective in reducing corruption, and thus the overall effect of policy conditionality and technical assistance was minor. The increasing flows of pre-accession aid and their gradual decentralisation to domestic authorities, on the other hand, meant an increased risk of corruption. Overall, we can conclude that prior to 2004 the EU had a strong, positive influence on corruption, with the most positive influence related to the exclusion of corruption-prone parties from the government. Accession, however, changed the situation significantly. Conditionality was lost both in terms of government composition and in policy areas, which was manifested in the increasing acceptability of Vladimír Mečiar and Ján Slota as government partners and in a reduced emphasis on anti-corruption policies amongst all mainstream parties. Anti-corruption technical assistance, which was part of the pre-accession funding, was also gradually wound down. The sharply increasing aid flows began to dominate the EU influence on corruption in Slovakia, with a major negative influence due to their discretionary nature and the decentralisation of management to national government authorities. On the other hand, accession meant that the European Commission and the European Court of Justice finally had the possibility to intervene directly in the areas of competition, state aid and procurement, thereby strengthening the impact of EU rules. Nonetheless, the overall assessment is that whereas before accession the EU undoubtedly had a positive influence in the field of corruption, its post-accession impact is much more ambiguous, primarily due to the effect of the Structural Funds.
EU membership has therefore been associated with noticeable changes in the scale of EU influence on corruption in Slovakia and with even more substantial modifications in the composition of the channels through which it was exercised.
Looking at the results from a broader perspective, the article shows that EU influence on such an important issue as corruption in a candidate or a member state can be significant, both in positive and negative terms. However, it also shows that in order to assess the influence correctly, we need a more holistic view of policy instruments and channels of influence rather than looking just at the acquis and its implementation. The conclusion in the case of Slovakia and corruption is that other channels dominated the impact of policy conditionality and technical assistance.
Consequently, the influence of the Union can benefit from an analysis that does not conceptualise influence as a two-player game concerned solely with legislation and its implementation, but recognises the multiplicity of actors on both sides, and particularly the multiplicity of channels. The general lessons are that research needs to look beyond direct instruments/policies to those with indirect influence. Research also needs to look beyond regulatory/institutional changes when examining the effects of Europeanisation.
The article also points out that the concept of post-accession backsliding might not do justice to the reality, since accession brought a strengthening of EU influence compared to the pre-accession period in terms of direct enforceability of some of the EU rules. While the role of Structural Funds in particular remains under-researched in this respect, it is clear that what we are seeing is a transformation of EU influence rather than a mere loss of influence.
The analytical framework developed here can be applied not only to other countries with regard to the issue of corruption, but also other areas-e.g. the civil service, where EU financial aid but also sectoral institutional requirements (Beblavý 2009) can have an influence that dwarfs formal policy conditionality. In other words, the specific channels of influence might not be directly transferable to other policy areas, but the underlying logic has the potential to be useful.