Improving the Transparency and Accountability of EU Institutions: The Impact of the Office of the European Ombudsman

This research focuses on the inquiries into alleged maladministration by EU institutions completed by the Office of the European Ombudsman (EO) during the period 1996–2012, seeking to highlight the ways in which the Ombudsman advances democracy in the European Union (EU). The article analyses the content of an original dataset consisting of 515 critical remarks and 113 follow‐up inquiries issued by the EO. The study shows that concerns about transparency and accountability dominate citizens’ complaints to the Ombudsman. Further, the analysis concludes that, when EU institutions were subject to critical remarks, in the majority of instances they adopted the EO's recommendation to increase the transparency and/or accountability of their practices. The EO's impact is observable both during the inquiry process and after it issues critical remarks. The article links these conclusions to a broader discussion of the democratic dilemmas in the EU.


Introduction
The European Ombudsman (EO), created in 1995 and thereafter elected by the European Parliament for five-year renewable terms, acts on the intersection of a parliamentary review committee and a quasi-judicial institution and, by monitoring the activities of other EU institutions and modifying EU administrative practices, plays an important role in shaping the EU's supranational democracy. Yet the impact of the EO on other EU institutions, and on EU democracy more generally, is rarely assessed systematically, despite frequent criticisms and widespread scepticism about its place in the institutional framework of the EU. Using an original dataset and quantitative analysis of maladministration cases, this article contributes to filling this scholarly gap and addresses the following questions: Theoretically, what role does the EO play in alleviating some of the EU's democracy dilemmas, especially with respect to the transparency and accountability of its other institutions? Analytically, to what extent do EU citizens use the EO as a venue to exercise their rights to accountable and transparent governance? How likely are EU institutions to respond to pressure from the EO and modify their policies and enhance transparency and accountability? The article starts with an overview of some of the main challenges for democracy at the EU level, followed by a discussion of the ways in which the European Ombudsman can contribute to addressing them. This section gives special emphasis to how the EO enhances citizenship rights and how it can modify the activities of other institutions, especially with respect to their transparency and accountability. The research design section introduces an original dataset compiled for this project and discusses the ways in which this study captures whether or not the Ombudsman's recommendations help improve the transparency and accountability of EU institutions and their policies and practices.
The analysis that follows shows that the Ombudsman acts as an important venue for EU citizens to practise their rights of transparency and accountability, and that it also serves to improve those aspects of EU institutional behaviour. As the article demonstrates, in the majority of instances when the EO issued a reprimand to an EU institution, the latter accepted the Ombudsman's recommendation and modified its institutional practices to increase its transparency and accountability. The next section provides examples of the types of issues dealt by the EO in these cases, highlighting its substantively significant impact. The final section concludes by pointing out the limitations of this studynamely, that conclusions are drawn from broad quantitative data and thus it is not possible to address, for example, why EU institutions are willing to comply with the EO's recommendations. The final section also suggests directions for future research. 1

I. The EO in the Context of Democratic Governance of the European Union
The EU's Democratic Dilemmas How does a complex organization like the European Union address the challenges of ensuring and enhancing democracy across its institutions? While it is beyond the scope of this article to address whether the EU is a democratic organization, the rest of the section proceeds as follows. 2 It first highlights some of the main problems associated with practising democracy at the EU level, and then provides an account of how the Office of the European Ombudsman helps alleviate some of them. Throughout the section, special attention is given to the issues of democratic accountability and transparency, as both are linked to the work of the EO in addressing some of the gaps in EU-level democratic governance. The section concludes by drawing testable propositions, which are evaluated in the analysis section.
Neither a state nor just an international organization, the EU is often discussed from the perspective of multi-level governance (MLG), defined as 'a system of continuous negotiation among nested governments at several territorial tiers' (Marks, 1993, p. 392; see also Bache and Flinders, 2004;Benz, 2007;Harlow and Rawlings, 2007;Hooghe and Marks, 2000, among many others). The governance of the EU, described often as suffering from a 'democratic deficit', compounded by its MLG nature and the long 'chains of electoral control' (Eriksen and Fossum, 2012, p. 16), presents several challenges for the democratic process, including representation, participation and accountability. As many have argued, the EU as a government by and of the people is very difficult to achieve, 1 The author has no affiliation with the Office of the European Ombudsman and has not received any funding from it to conduct this study. 2 For a flavour of the discussion on the state of democracy in the European Union, see for example Hix (2006), Majone (1998), Moravcsik (2002) and Scharpf (2009), among many others. for at least two reasons. The lack of a European demos, the multitude of diverse populations speaking different languages, and more generally 'the absence of a unified public sphere ' (DeBardeleben and Hurrelmann, 2007, p. 6) all undermine the social preconditions for democracy at the EU level. 3 Further, despite recent gains in the powers of the European Parliament, compared to national-level democracies, the EU has fewer channels for political participation and democratic input in policy-making (e.g. Hurrelmann and Debardeleben, 2009).
A second dimension of EU-level democracy relates to practising it for the people. Arguably, among the best accomplishments of the European Union are its focus on and ability to produce policy outcomes that benefit the entire community: most importantly peace and stability, but also economic prosperity (e.g. Hurrelmann, 2014). However, the capacity of the EU to provide solutions to collective action problems, and thus enhance its output legitimacy, is counterbalanced by diminished accountability on the input side. This represents what scholars (e.g. Hurrelmann, 2014) call efficiency-accountability trade-off that is compounded by the inability of voters to directly punish most EU policy-makers for failed policies. The latter issue is intensified in part by the complex MLG structures, which make it difficult to identify who is to 'blame' for such policies (e.g. DeBardeleben and Hurrelmann, 2007;Papadopoulos, 2010). Peters and Pierre (2004) call such trade-offs a 'Faustian bargain', 'in which core values of democratic government are traded for accommodation, consensus, and the purported increased efficiency in governance' (p. 85). Such critiques reflect perhaps the most serious challenge to the legitimacy of the European Union, as not only far removed from citizens but also driven by 'managerial concerns' (Bache and Flinders, 2004, p. 202) rather than democratic values.
A third, often ignored, aspect of EU-level democracy is what Schmidt (2006) refers to as government with the people, which 'opens decision-making up to citizens qua organized interests ' (p. 25). In contrast to democratic participation by the people through voting, government with the people relies on civil society, business, consumer advocates and other interest groups taking part in the consultative phase of EU policy-making and implementation. Such participation increases both awareness and transparency of the EU, its decision-making processes and their complexity, and it also enhances the EU's legitimacy on the input side. As Schmidt points out, though, democracy with the people, as practised across the EU, is hindered by the low involvement of individual citizens (rather than by their organized groups) in the EU processes. Yet for Schmidt, who argues that comparing the EU to the standards of national-level democracies is not appropriate, this third aspectdirect individual citizen participation in the EU institutions and processes, outside of elections for the EPhas great potential for strengthening supranational democracy.
Thus, in summary, on the input side of democracy, the EU has limited channels for direct participation and representation of individual citizens, and even fewer opportunities for the latter to directly challenge the decisions of policy-makers. The lack of mechanisms for citizens to replace most EU-level policy-makersi.e. accountability through elections even if they can do so at the national level is seen as one of the most significant weaknesses of EU democracy. On the output side, the EU does better, and it also presents more opportunities for post hoc answerability of administrators during the policy implementation phase (Heritier, 1999). Thus, accountability spans both sides of the EU's democracy dilemma, and while direct electoral accountability is far from being a reality in the EU, the Office of the EO is among the main mechanisms for enhancing the EU's administrative transparency and accountability.
The next few paragraphs turn to discussing how the EO can help alleviate some of the democratic challenges facing the EU, namely by serving as a platform for citizens exercising some of their supranational rights and by improving institutional transparency and accountability, while the analysis section further below quantitatively evaluates the extent of the EO's impact on both accounts. 4 The EO's Role in Enhancing Transparency and Accountability in the EU The work of the Office of the European Ombudsman is focused on improving the practices of EU institutions and agencies, through investigating individual complaints or launching its own initiatives into matters of maladministration and, where necessary, submitting reports to the European Parliament (EP). As such it promotes transparency and accountability in the EU in two ways: by safeguarding individual citizens' rights to such, and by modifying the practices of EU institutions.
Transparency, understood as the openness of the EU institutions, implies on one hand the right of citizens to have access to information about the activities of EU institutions, and on the other these institutions' responsibility to provide such access and information. The first part of the concept, citizens' right to have access, is indisputable and well embedded in the legal framework of the European Union. For example, Articles 8b2 and 8b3 and Amendment 28d of the Treaty of Lisbon are concerned with the transparency of the European Union in general or the European Commission specifically. Additionally, other documents such as Regulation No. 1049/2001 and Article 253 of the Commission's Code of Good Administrative Behaviour, as well as the EO's own European Code of Good Administrative Behaviour, govern access to documents and information about activities of EU institutions. Further, the Treaty of the Functioning of the European Union articles 20d and 24 give EU citizens explicit rights to petition the European Ombudsman, whose functions are fully articulated in article 228.
While information about EU activities is not lacking, the EO is one of the few EU institutions through which citizens can solicit documents that are not publicly available, and further where they can directly question the decisions of EU officials. In this respect, the Office of the European Ombudsman, by serving as a platform for citizens to exercise their rights to transparency and accountability, also contributes to strengthening the democracy with the people aspect of the European Union. Further, the EO is one of the few institutions through which individual citizens, not only citizen groups, can get involved in the oversight of EU institutionsan aspect which, as Schmidt (2006) points out, is greatly lacking in EU-level democratic governance. The extent to which citizens do exercise their rights to transparency and accountability through the institution of the EO, though, is an empirical question that is addressed in the analysis section of this article.
The second part of the concept of transparencythe responsibility of EU institutions to provide access to informationis what is often investigated by the Office of the European Ombudsman, and is also analysed in this project. While the right of citizens to be informed is legally guaranteed and widely accepted, the extent to which EU institutions practise their responsibility for transparency varies, as does the meaning of what is acceptable access to information. Thus, in addition to serving as a platform for citizens' expressions of their rights to transparency and accountability, the EO's activities can change the policies and behaviour of EU institutionsand this role of the EO is in addition to, or even separate from, citizens' demands for greater transparency and accountability. In other words, the act of citizens filing complaints with the EO is separate from the impact (if any) that the subsequent activities of the EO might have on EU institutions. Both aspects are important: the EO cannot perform the majority of its functions unless citizens initiate the process of demanding greater transparency and/or accountability, although the Ombudsman does have the right to own initiative. However, it is through the process of investigating citizens' complaints that the EO has the potential to impact EU institutions.
The second aspect of the EO's functionits investigative roleis linked to the concept of post hoc administrative accountability. While accountability is most often associated with a normative evaluation of the behaviour of elected officials in a democratic polity, it does have applicability in the context of supranational EU institutions. Much of the recent literature on this topic, especially that related to the European Union, follows Bovens' (2007) definition of accountability as 'a relationship between an actor and a forum, in which the actor has an obligation to explain and to justify his or her conduct' (p. 107, italics in original). This definition is consistent with principal-agent theory, and specifically with Mulgan's (2000 and and Busuioc's (2013) understanding of accountability as including elements of not just information but also control and explanation, as well as the possibility for remedies and sanctions (see also Diamandouros, 2006;March and Olsen, 1995;Martinsen and Jorgensen, 2010;Sabel and Zeitlin, 2010;Skogstad, 2011). Following this definition, where 'the stipulation that the body to whom one has to do the explaining is (even if only indirectly) the electorate is dropped' (Page, 2010(Page, , p. 1011, one can apply the concept of accountability beyond electoral politics. In fact, as Lord and Pollak (2010) argue, limiting our understanding of accountability mechanisms to elections would ignore 'how far democracy, representation, and the electoral process themselves depend on accountability to the courts and independent administrative bodies such as Ombudsmen, especially for protections against the grave misconduct of representatives' (p. 973). Further, as Mulgan (2003) demonstrates, Ombudsmen and other public audit institutions are considered agents of government accountability that goes beyond electoral politics.
Understanding accountability as providing justifications for one's behaviour makes it possible to apply this concept to processes outside of elections, and it also allows scholars to study accountability outside of the scope of national politics. The European Union, with its institutional complexity and supranational aspects, is a good example of how accountability can be practised outside of the nation-state, which puts it at the 'cutting edge in terms of developing accountability mechanisms of various kinds at the supranational level of governance' (Curtin et al., 2010, p. 933). While there are multiple ways in which EU institutions could be held accountableincluding through electionsfor the purpose of this study, and adapting Bovens' definition to the context of multi-level governance (Papadopoulos, 2010), the office of the European Ombudsman is the forum and the EU institutions that are subject to the EO's investigations are the actors who need to provide information and explanation of their actions and behaviour.
The work of the EO can induce EU institutions to modify their behaviour in two ways: first through investigations into specific cases/complaints, which are concluded by issuing critical remarks with specific recommendations to the involved institutions, and second by following up the cases where critical remarks were issued to check if the institution has complied. 5 In all instances of alleged maladministration the Office opens an inquiry, engaging in private communications with the institution(s) against which the complaint has been filed, and tries to resolve it. If the Ombudsman concludes that maladministration has occurred, and when the case is not resolved through a friendly solution, the EO publishes a statement called a 'critical remark' which summarizes the complaint, the steps taken to address it and the recommended modifications that the involved institution needs to make. Finally, the cases in which critical remarks were issued are followed up by the EO to find out whether and how the institution has responded to the EO's recommendations for modifications. Thus, the EO's influence can be seen at both stages in this processthe investigation, which ends with a critical remark, and the follow-upand it is logical to assume that its impact would be different at each stage. Since the inquiry stage is private until a critical remark has been issued, one would expect the EO's influence over EU institutions would be weaker at that stage, especially given that the Ombudsman has no legal powers over other EU institutions. Once a critical remark has been issued, though, and the recommendations of the EO have been made public, one could assume a stronger impact. Even though an EU institution would still not be required to comply with the EO's recommendations, given the 'naming and shaming' public nature of the critical remark and the specificity of the issued suggestions, the impact of critical remarks and their subsequent follow-up should be stronger than that of the inquiries themselves.
The above discussion of the democratic dilemmas facing the European Union, and the ways in which the Office of the European Ombudsman can alleviate some of their aspects, can be summarized in the following propositions, which are evaluated in the analysis section.
H 1 : EO serves as a venue for citizens' concerns with transparency and/or accountability of the EU and its institutions.
H 2 : The EO enhances transparency through its investigations of allegations against EU institutions. 5 The EO receives a great number of complaints each year, and as a first step it screens them for admissibility. Once a case is deemed admissible and the EO opens an inquiry regarding the complaint, the process could lead to several outcomes. The institution against which a complaint is filed could settle the matter and agree to a friendly solution proposed by the EO. Additionally, in some cases, after an initial reading, the EO could decide that no further inquiries are justified or that there was no instance of maladministration. Finally, in a selection of cases, which are the focus of this article, the office of the EO finds that one or several instances of maladministration have occurred, and completes an investigation in the case, concluding it with a critical remark. These cases are later followed up to assess compliance with the recommendations.
H 3 : The EO enhances accountability through its investigations of allegations against EU institutions.
H 4 : The EO enhances transparency through its follow-up of cases featuring such documented maladministration.
H 5 : The EO enhances accountability through its follow-up of cases with such documented maladministration.
H 6 : EU institutions are more likely to respond to recommendations of the EO during the follow-up phase of the Ombudsman's investigations, after a critical remark has been issued, than they are during the private inquiry phase.

II. Research Design
The purpose of this analysis is to assess how the Office of the European Ombudsman acts to improve the functioning of democracy at the European Union level by impacting the policies and administrative behaviour of EU institutions. The rest of this section proceeds with a discussion of the measures used to capture the EO's impact on the transparency and accountability of EU administrative practices. 6 Based on the broader theoretical discussion in the previous section, transparency is defined as having access to documents, materials, etc., about the activities, policies and practices of EU institutions, while accountability is understood as EU officials providing explanations and justifications for decisions, actions and policies.
In an attempt to understand the impact of the EO on the functioning of EU institutions, this research focuses on the cases where the Ombudsman has found that maladministration has occurred, where no friendly solution was reached in the process of investigation and the EO has issued critical remarks specifying how the procedures, policies or behaviour of EU institutions should be modified to improve their transparency and/or accountability, as well as on the subsequent follow-up by the EO office on maladministration cases. Only cases with inquiries completed between 1996 and 2012, where the Office of EO concluded that maladministration had taken place, were coded. 7 Not all complaints featuring maladministration involved a case of lack of transparency and/or accountability. 8 To find out the share of such cases, each inquiry by the EO was coded in terms of whether or not the alleged maladministration related to transparency or accountability (equals one, otherwise, zero). To provide a more fine-tuned assessment, the coding scheme included two separate questions, one for transparency and one for 6 The source of all of the data used in the analysis is the website of the Office of the European Ombudsman, and in particular the EO's (1) detailed accounts of maladministration cases, (2) annual reports and (3) collection of follow-ups to investigations of critical remarks. Information is available at: http://www.ombudsman.europa.eu/en/cases/decisions.faces, http:// www.ombudsman.europa.eu/en/activities/annualreports.faces, and http://www.ombudsman.europa.eu/en/cases/followups. faces, respectively, and was accessed between January 2013 and March 2014. 7 These cases then represent the most stringent tests of the impact of the EO on EU institutionsthose where a friendly solution was not possible, and where resolution of all the issue(s) was not achieved in the process of investigation. In other words, these cases represent situations in which the EU institution(s) involved in the case already had an opportunity to learn about the allegations put forward by the complainant and, through their communication with the EO, to respond to these allegations and/or to try to resolve the issue. In the cases analysed here, no such resolution occurred. 8 Many of the critical remark cases are very complex (and lengthy), and involvein addition to transparency or accountabilityother issues of maladministration, such as undue delays in responding to inquiries or failure to adhere to the institutions' own policies (for example of non-discrimination). accountability. In this way the coding also accounts for cases in which the maladministration concerned both transparency and accountability. These measures, alleged lack of transparency and alleged lack of accountability, capture indirectly citizen demand for such, and are used as a proxy measure to ascertain the extent to which those who file complaints with the EO do so to exercise their citizenship rights for transparency and accountability. The two measures are used to assess hypothesis 1.
The remainder of the coding scheme captures the extent to which the activities of the Office of the European Ombudsman, through its recommendations, inquiries and follow-up, helps modify the practices of EU institutions towards greater transparency and accountability. One set of variables, suggested improvement in transparency and suggested improvement in accountability, are coded as equal to one (otherwise, zero) when the critical remark issued by the EO suggested such improvements. Again, the two issues were coded separately, acknowledging that in a number of cases the EO calls for improvements in both aspects. These measures are not directly used to assess any of the hypotheses spelled out in the previous section, but are instead used as an indicator of whether or not the EO agrees with the nature of the citizens' complaints.
A third set of variables, inquiry improved transparency and inquiry improved accountability (coded as equal to one or zero), reflect whether or not the EO investigation of the case led to modifications of administrative practices with respect to transparency and/or accountability. These measures are used to assess hypotheses 2 and 3, respectively. Finally, since 2006, the EO has published annual summaries of follow-ups of its investigations after critical remarks were issued. For each case, the EO office recounts the response received by the offending institution and whether or not the institution has accepted the Ombudsman's recommendation and modified its practices and policies to avoid future instances of maladministration of the type that was identified by the EO. For these cases, the coding reflects whether or not the recommendations of the Ombudsman have been adopted by the offending institution (variables follow-up improved transparency and follow-up improved accountability, respectively). These variables are used to evaluate hypotheses 4 and 5, and the difference between compliance during the inquiry phase and compliance during the follow-up is used to assess hypothesis 6.
Thus, the coding scheme looks at how the issues of transparency and accountability are reflected in each maladministration case: (1) from the perspective of the complainant; (2) as part of the inquiry initiated by the EO; (3) in terms of the critical remark issued by the EO; and (4) in the final response of the institution involved in the investigation. The first set of variables in the coding scheme reflects citizens' demands for greater transparency and accountability of EU institutions. The last three sets capture whether or not the Ombudsman's investigations, critical remarks and/or follow-up on cases are successful in modifying the practices of EU institutions with respect to transparency and accountability. 9 Appendix A shows the full text of the coding scheme, which also accounts for the following questions: who is the complainant (individual or legal person); what is the country of residence (if available); which institution(s) are involved; and in which year did the EO issue a decision on the case?
The next section presents a discussion about the quantitative characteristics, regarding transparency and accountability, of the maladministration cases, as coded using the above scheme. The section that follows briefly presents a qualitative assessment of the substantive nature of the complaints dealt with by the EO.

III. Quantifying the Impact of the Office of the European Ombudsman
The analysis focuses on, first, 515 complaints and their subsequent inquiries completed between 1996 and 2012 in which the EO found an instance of maladministration, as defined above, and issued critical remarks; and second, 113 follow-ups of these critical remarks issued between 2006 and 2012. 10 These cases were selected because they represent the most serious violations of good administrative practices, with the EO issuing the sternest of reprimands, where a friendly solution was not reached, and thus where the impact of the Ombudsman is least likely. Thus a focus only on cases of maladministration presents the most stringent test of the impact of the EO, as it concerns cases where options for resolution have been exhausted during the preliminary inquiry stage. Table 1 presents the share (with numbers in parentheses) of complaints, critical remarks, inquiries and follow-ups that concern (or suggest to improve) the transparency and/or accountability of EU institutions.
Starting with the nature of the complaints of maladministration filed with the EO, the second column of Table 1 shows the share of instances of alleged lack of transparency (39 per cent, or 203 cases) or of accountability (58 per cent, or 308 cases). Overall, three quarters of all complaints (385 cases) that resulted in a critical remark being issued by the EO concerned allegations against an EU institution or agency involving a situation of either compromised transparency or accountability. Further, close to one quarter of them (122 cases) were concerns about lack of both transparency and accountability.
The high share of inquiries involving a complaint about lack of either transparency or accountability in cases where an EU institution had committed maladministration is an indication that concerns about how EU institutions function dominate the investigations undertaken by the EO office. Such a high share also indicates that EU citizens (and legal persons) are engaged in demanding greater transparency and/or accountability of EU institutions, thus exercising their rights to such. This conclusion confirms H 1 , which states that EU citizens utilize the EO as a venue to express their dissatisfaction with the transparency and/or accountability of the EU institutions. While the Ombudsman is certainly not the only venue for such expressions, it is among the main tools that EU citizens have to directly challenge decisions by other institutions, and to practise democracy with the people.
Focusing on the substance of the critical remark issued by the EO after the completion of each maladministration inquiry reveals that in the majority of cases, around two thirds (354 cases), the EO office recommended that the EU institution involved in the case change its procedures and practices to improve the transparency and/or accountability of its practices (third column of Table 1). One third of critical remarks (159 cases) suggested the need to increase the transparency of the institution, while 50 per cent (or 256 cases) suggested increasing the accountability of its actions. In around 14 per cent of cases (70 inquiries), the EO office recommended improvements in both transparency and accountability practices. The critical remarks issued by the EO are about future changes that each institution or agency needs to undertake to address the concerns brought by the Ombudsman in the process of investigating each complaint.
While the EU institutions and agencies investigated by the Ombudsman are not obligated to follow the recommendationand, further, the EO has no formal mechanism to impose sanctions on offending institutionsrepeated occurrence of similar malpractice offences by the same institution could lead to the EO launching its own investigation into the matter, followed by special reports sent to the European Parliament. Further, two additional sets of data show that European Union institutions investigated by the Ombudsman do in fact follow through on recommendations for improved transparency and accountability of their administrative practices. As the fourth column of Table 1 demonstrates, in one quarter (121) of the cases, the actual process of EO inquiry led to the offending EU institution's compliance by providing access to information (improving transparency) or justification for its actions (improving accountability). More specifically, 14 per cent of the inquiries (69 cases) led to an improvement in transparency, 15 per cent (79 cases) in accountability, and 5 per cent (27 cases) in both. Such findings confirm hypotheses 2 and 3, stipulating that through its inquiries regarding maladministration practices, the EO can impact the policies and behaviour of EU institutions.
Additionally, an analysis of follow-up of the Ombudsman's inquiries shows that rates of compliance with its recommendations are even higher. The follow-ups summarized in the last column of Table 1 represent situations in which the EO's critical remarks suggested an improvement in the transparency and/or accountability of the EU institution's administrative practices, and where the inquiry itself did not lead to such improvements. Thus, after issuing a recommendation in the form of a critical remark, and after taking into account the formal reply issued by the involved institution(s), the EO makes a decision regarding whether or not the institution has taken measures addressing the EO's concerns, and this decision and the justification for it are published in the follow-up report. While follow-up decisions are available only for the past seven years of the EO's work, analysis of the EU institutions' responses indicates that they are very likely to adopt the recommendations of the Ombudsman with respect to instances of maladministration. The information in the last column of Table 1 should be interpreted as follows. Of the critical remarks concluding transparency-related maladministration where the inquiry itself did not result in improved transparency, in 68 per cent of the cases (43 instances) the institution adopted the EO's recommendation. In the same number of instances (43), representing 58 per cent of critical remarks regarding accountability-related maladministration where the inquiry itself also did not result in improved administrative practices, in its follow-up the EU institution(s) also adopted the EO's recommendation. A small number of cases (18) concern critical remarks about lack of both transparency and accountability, and in 67 per cent of these the institution(s) responded favourably to the Ombudsman's recommendation. Overall, in 60 per cent of maladministration cases involving issues of transparency or accountability, the EO's recommendation(s) were adopted by the offending EU institution, which confirms hypotheses 4 and 5.
The fourth column in Table 1, when compared to the numbers in the third column, also helps to evaluate hypotheses 6, which states that EU institutions are more likely to comply with EO recommendations after a critical remark has been issued. The higher rate in the follow-up cases (60 per cent) where transparency and/or accountability was documented to be improved, compared to such rates during the inquiry phase (24 per cent), indicates that the public nature of the critical remarks and subsequent follow-up provides incentives for EU institutions to adhere to EO recommendations. Overall, the EO helps improve two of the democratic aspects of EU institutions with both of its toolsinquiries and followupsyet the latter has a much greater impact.

IV. The Substantive Nature of the Maladministration Cases -A Few Examples
Using quantitative measures, the above section shows that the EO's inquiries and investigations can increase the transparency and accountability of the EU. The numerical assessment, however, does not reveal the substantive impact of the EO. Does the Ombudsman investigate cases about trivial issues, or are they only concerned with matters that are important for the EU and its citizens? While a systematic discussion of the essence of maladministration cases is beyond the scope of this article, discussing some of the issues that typical inquiries involve helps strengthen the conclusion that the Ombudsman plays an important role in the EU, addressing issues that have substantive meaning.
The following sets of examples represent the four most common categories of issues found in the maladministration cases investigated by the Ombudsman. One common concern is the way in which the European Commission (EC) performs its function as a 'guardian of the treaties'. For example, case 783/1998 involved an initial complaint filed with the EC regarding Spain violating the principle of mutual recognition of diplomas. After several years of failing to obtain information from the Commission regarding its investigation of this case, the complainant filed with the European Ombudsman. The EO agreed that by not responding to the information requests of the complainant, the EC failed to act in a transparent way. Similarly, case 39/2002 was about how the EC handled a complaint alleging that the United Kingdom (UK) failed to apply Council directive 85/337/EEC. After the Commission closed the case without an explanation, the complainant addressed the EO, which concluded that the EC must justify its closing of the initial case and that failing to do so was a breach of accountability. Case 1288/2002 is also an example of the EC closing an investigation of another Member State (Greece) violating Community directives, without informing the complainant and without providing an explanation (breach of both transparency and accountability).
Failure to obtain access to documents represents another set of common complaints about maladministration. In case 1463/2006, the complainant submitted to the European Commission, the European Parliament and the Council of the European Union a request for access to documents relating to national plans by the UK, France and the Slovak Republic for the allocation of greenhouse gas emission allowances. After the Commission failed to provide most of the documents, citing in its refusal the privacy of decisionmaking processes, the EO in its investigation stated that the refusal to grant access represents a violation of the principle of transparency and the EO did not accept the reasons for this refusal (breach of accountability). Cases 1443Cases /2007Cases , 1972Cases /2009Cases and 1131Cases /2007 are also about lack of access to documents on the part of the EC and the European Parliament, respectively.
Another set of cases that are representative of the substantive issues in maladministration cases involves funding, tenders and grants from the European Commission. For example, when the Commission failed to explain why it rejected a request to sponsor a community event (case 23/1998) or why it reduced the contracted amount that a German organization caring for sick children was receiving (case 1087/1998), the EO agreed with the complainants that these instances violated the principle of accountability (represented lack of reasoning). Cases 1043/2011 and 949/2003 represent examples where the Commission either failed to inform tender participants of the status of their applications or, in the judgement of the EO, did not apply transparent selection criteria.
Finally, a number of cases concern issues of hiring and human resource management. Cases 2472Cases /2005Cases , 14/2011Cases , 16/1998Cases and 905/1999 represent examples in which an applicant in open competition failed to obtain a position with, respectively, the Commission of the Regions, the European Personnel Selection Office, the Council of the European Union and the European Commission. In all these cases, the EO found that the institution failed to provide information and/or explanation to the applicant as to why they were not selected. Cases 1751Cases /2002Cases , 998/2002Cases and 2954Cases /2008 are examples of situations in which current EU employees (of the Commission and the European Court of Justice, respectively) did not receive full payment for their work, and the employing institution did not provide explanations for the missed payments.
That these four categories -'guardian of the treaties' complaints, access to information, administration of community funding and hiring and employment by EU institutionsare the most common issues in maladministration cases reveals that the EO investigates matters that are of substantive importance beyond each individual case. Thus the impact of the Ombudsman in increasing the transparency and accountability of EU institutions, when they deal with such important matters, goes beyond remedying administrative shortcomings and has broader implications for the legitimacy of the EU institutions.

Conclusions
Despite being in existence since 1995, the Office of the European Ombudsman remains significantly understudied. Yet the EO is an important venue for citizens to directly express their concerns with the ways in which EU institutions function. Between 1995 and 2011, the Ombudsman's office received over 36,000 complaints, an average of 2,127 per year. Over time, the number of EU institutions and agencies against which a complaint was filed also increased, peaking at 40 in 2011. By compiling an original dataset of the Ombudsman's critical remarks and follow-up decisions, this article discusses how and to what extent the EO improves democracy at the European Union level, specifically focusing on the transparency and accountability of EU institutions. A few conclusions are worth reiterating. The Office of the Ombudsman works to enhance the individual rights of EU citizens in two ways. Directly, the EO is a strong proponent of considering transparency and accountability to be citizenship rights, and the latter are encoded as such in EU legal documents, thanks in part to the work of the Ombudsman. The EO also serves as a platform for citizens to express these rights, as indicated by the high share of maladministration complaints pertaining to lack of transparency and/or accountability filed with the Ombudsman, thus promoting democracy with the people at the EU level.
Further, content analysis of maladministration cases investigated by the EO, as well as of the follow-ups provided on these cases since 2006, indicates the following. First, the majority of maladministration cases involve a lack of either transparency or accountability, and this holds true both for the allegations made by the complainants and for the nature of the critical remarks issued by the EO office. Of those, more cases involve violations of accountability than of transparency. This is perhaps due to the push in recent years by the European Union to increase the visibility of its activities through online and print publications and other media. Thus, it is difficult to argue that lack of information is an issue plaguing the EU and its institutions. Finally, the analysis concludes that the actual process of EO investigation of a maladministration case leads to improvements in both the accountability and transparency of the involved institution(s). This finding is strengthened by the significant share (at least 56 per cent) of cases in which the EU institution(s) responded positively to the EO's recommendations after an inquiry was completed and adopted practices to improve the transparency and accountability of its activities, as suggested by the Ombudsman. This is an especially positive conclusion given that these cases involve documented violations of administrative practices where friendly solutions or settlement were not possible before the EO investigation, and considering that the recommendations of the EO are non-binding. Thus, the EO Ombudsman can have an impact on the how EU institutions practice certain aspects of democratic governance in two waysthrough the inquiries that it opens into accusations of maladministration, and through the critical remarks and subsequent follow-ups that it issues. Conclusions of the quantitative assessment of the EO's impact are strengthened by the substantively important nature of complaints investigated by the Ombudsman.
While this project provides a systematic analysis of the extent to which the EO is able to induce European Union institutions to make their practices more transparent and accountable, much remains to be addressed in future research. As a piece utilizing quantitative data, the article is not able to address some of the underlying mechanisms that lead to the Ombudsman having an impact on EU-level democracy. Further research is needed in two directions: what motivates citizens to reach out to the European Ombudsman, and to what extent is filing complaints to the EO a form of participation in the EU? With respect to the direct impact of the EO on EU institutions, this project is not able to address the motivations behind the positive responses to Ombudsman recommendations. Further studies using such methods such as surveys and/or in-depth interviews are needed to ascertain why EU officials and bureaucrats comply with the EO's recommendations, and what are the implications of their compliance for EU-level democracy. (1) Yes (2) No 9. Did the EO find maladministration related to lack of (insufficient) accountability of the EU institution involved in the case?/Was the case about accountability?

Correspondence
(1) Yes (2) No 10. Did the critical remarks by the EO encourage/suggest an improvement in transparency?
(1) Yes (2) No 11. Did the critical remarks by the EO encourage/suggest an improvement in accountability?
(1) Yes (2) No 12. Did the EO's inquiry lead to increased transparency of the respective EU institution?
(1) Yes (2) No 13. Did the EO's inquiry lead to increased accountability in the respective EU institution?
(1) Yes (2) No 14. Did the EU institution's follow-up comply with the Ombudsman's recommendation on how to improve the transparency of the institution's practices?
(1) Yes (2) No 15. Did the EU institution's follow-up comply with the Ombudsman's recommendation on how to improve the accountability of the institution's practices?