Transparency Watchdog : Guarding the Law and Independent from Politics ? The Relationship between the European Ombudsman and the European Parliament

This article investigates whether the European Ombudsman acts as an ‘independent’ institution vis-à-vis the European Parliament (EP). This is a relevant question because while the Ombudsman is appointed by and reports to the EP, it can also conduct inquiries into the work of the EP, in instances of alleged maladministration. Based on the empirical examination of all decisions following an inquiry by the Ombudsman in cases against the EP for an eleven-year period (2004–2015), plus the review of two recent landmark own-initiative inquiries, we inductively construct three roles played by the Ombudsman in relation to the EP, namely: ‘arbitrator’, ‘transparency watchdog’, and ‘vessel for civil society concerns’. These roles are used to operationalize the concept of independence. We conclude that the Ombudsman acts independently and is not a mere auxiliary organ of the European legislature. This is most apparent in the ‘transparency watchdog’ role, where the European Ombudsman has ensured the release of information empowering citizens to hold the Parliament accountable, or—failing that—has stimulated debate concerning such information (for instance, on theMEPs’ financial allowances) both within the Parliament itself and in the wider public domain.


Introduction
The academic debate on the European Union's (EU) democratic deficit has been described as being 'crowded waters' (Kohler-Koch & Rittberger, 2007).This article contributes to one strand of this 'crowded' debate, namely that of 'procedural' legitimacy (Lord & Magnette, 2004).According to this view, legitimacy may be enhanced as long as certain procedures-such as transparency, balance of interests, proportionality, legal certainty and the consultation of stakeholders-are adhered to, so as to increase public accountability (Meijer, Grimmelikhuijsen, & Brandsma, 2009).
The creation of the European Ombudsman (EO)1 in 1995 represents perhaps the strongest illustration of the growing importance given to procedural legitimization in the EU.The EO is expressly tasked with investigating maladministration within EU institutions and bodies-therefore, it constitutes a channel of scrutiny dedicated exclusively to how public officials carry out their activities.
Although academic research regarding the EO is rather scant, most existing contributions highlight the success of this institution.Magnette (2003) argues that, despite some initial scepticism, the EO managed to define and disseminate a set of defining principles of 'good administration', and more generally contributed to wideranging reforms in European governance.Along similar lines, Kostadinova's (2015) systematic analysis of cases spanning a period of over 15 years demonstrated that EU institutions have accepted a significant share of the EO's recommendations, which lead to the implementation of practices that boosted both their transparency and accountability.
Taking as a premise this rather positive account of the EO's performance, we focus on its 'hybrid' nature as a 'cross-over' between parliamentary control body and judicial organ (on this see Magnette, 2003).To be clear, on the one hand, the EO is appointed by and reports to the European Parliament (EP). 2 On the other hand, it independently investigates cases brought to its attention and solves them by defining and applying 'general principles' (much in the manner of a court of law).This ambiguity makes for a rather paradoxical relationship with the EP, which formally appoints the EO, while at the same time being potentially subject to its investigations in instances of alleged maladministration, as with any other EU body or agency.
The question thus poses itself whether the EO acts as an 'independent' institution as stipulated in the Treaty.As the concept of independence is difficult to grasp, it will be operationalized by means of certain roles that will be inductively drawn from the analysis of cases brought in front of the EO and against the EP.The conceptualization of roles is thus derived in a bottom-up (inductive) fashion, through the empirical analysis of all decisions following an inquiry by the EO in cases against the EP for a period of more than ten years (January 2004-May 2015).This period was chosen to cover two complete legislative terms of the EP (i.e., 2004EP (i.e., -2009EP (i.e., and 2009EP (i.e., -2014)), and the work of two out of the three people who have served as EOs so far, namely Nikiforos Dimandorous (2003Dimandorous ( -2013) ) and Emily O'Reilly (2013-present). 3urthermore, given the relative infrequency of inquiries against the EP, this rather lengthy time-frame also enables us to cover sufficient cases to be able to trace consistent patterns over time.In sum, these research design choices allow our findings to be generalizable.
Note that the thrust of the data used in this article is towards the 'internal' politics of the EU in general, and of the EP in particular (see Leino, 2017, in this issue).Not only has the EP a great role to play within this domain due to its co-legislative function (Hix & Hoyland, 2013), but also the cases brought to the EO mostly focus on internal EP issues (such as staff matters).Nevertheless, the role of the EP in the 'external' politics of the EU has been upgraded since the Lisbon Treaty, as the Parliament gained the right to veto international trade agreements, to just give one example (Rosén, 2016).To reflect these developments, but also to show how the EO engages the EP when it is acting in a more entrepreneurial manner, we complement the case analysis with a discussion of two landmark own-initiative inquiries of the EO.One of these concerns the internal/ institutional politics of the EU (case OI/8/2015/JAS, on the transparency of trilogues), while the other relates to the external dimension of the EU (case OI/11/2014/RA, on transparency and public participation in the Transatlantic Trade and Investment Partnership [TTIP] negotiations) (see Gheyle & De Ville, 2017, in this issue).
Due to the fact that there were no own-initiative inquiries which targeted the EP exclusively, we chose to focus on two cases that involved the EP together with other institutional players, namely the European Commission and the Council of the EU.These two particular inquiries are relevant for our analysis for two reasons.Firstly, they both deal with domains of activity and practices that have been traditionally marked by secrecy.Consequently, one of the main endeavours of the EO, the advancement of institutional transparency, is likely to come up against significant institutional resistance, including on behalf of the EP.These inquiries are thus significant tests for the EO's independence.Secondly, this choice of cases allows us to contrast a situation where the EO aligned with the EP to challenge the Commission and the Council (i.e., the TTIP inquiry) with one where the EO acted alone to challenge all three institutions equally (i.e., the trilogues inquiry) (see Abazi & Adriaensen, 2017, in this issue).
Observing this contrast is relevant for the EO's independence because it shows us whether or not an alliance with the EP has an impact on its activity.
The article proceeds as follows: the following section presents the attributions and the powers of the EO.Then we present an account of cases against the EP, which were brought forward to the EO.Based on this, we define two main roles assumed by the EO, and show how they are linked to the different subject matters of the cases under review.The third section deals with the aforementioned own-initiative inquiries and brings to light a third role played by the EP.Conclusions follow.

The Office of the EO and Its Relationship with the EP
The relationship between the EO and the EP can be meaningfully viewed through the lenses of agency theory.We build here on Majone's (2001) seminal distinction between two logics of delegation: on one hand, the logic of efficiency, where the principals' core aim is to reduce transaction costs and take advantage of the agent's expertise, and, on the other hand, the logic of credibility, where delegation serves to render the principals' (policy) commitments trustworthy by shielding them from ex-post legislative or administrative tinkering.This second type of delegation is prevalent where principals may face short-term interests to default on their initial commitments, and/or where there are reputational benefits to be reaped by having an agent whose decisions are not 'contaminated' by politics (Alter, 2008;Majone, 2001).In the EU, relevant examples of such 'fiduciary' agents or 'trustees'-as they are called in the literature-include the European Central Bank, the European Court of Justice, and the European Commission in some of its functions (Franchino, 2002;Majone, 2001;Pollack, 2007).
The different rationales for delegation presented above have far-reaching implications for the relationship between principals and agents.As Alter (2008) and Handke (2010) show, traditional agents are chosen because they have similar views and values to their principals, and are expected to execute their duties as if representing their principals.By contrast, trustees are selected primarily due to reputational and professional credentials, which may sometimes mean that their values are systematically different from those of their principals.Furthermore, trustees enjoy comparatively more discretion and are expected to carry out their mandates according to their own professional norms and best judgement.Importantly, while traditional agents act on behalf of their principals, trustees act on behalf of a third party-a beneficiary-towards whom both the trustees and their principals are bound.This beneficiary can be an artificial construction, for instance, the citizens in a democratic polity (Alter, 2008;Gehring & Plocher, 2009).Due to these factors, the principals' control over trustees is significantly looser compared to traditional agentsin particular, once appointed, a trustee will be less vulnerable to re-contracting sanctions (i.e., dismissal, budget cuts, re-writing of their mandate).These fundamental differences in both the rationale of delegation and the relationship between the relevant parties have led some authors (e.g., Alter, 2008;Handke, 2010) to argue that fiduciary relations cannot be adequately captured by the principal-agent model, while dissenting voices (e.g., Brandsma & Adriaensen, 2017;Pollack, 2007) point out that this is not the case as Majone's (2001) two logics of delegation represent opposite ends of the same continuum (as opposed to being dichotomous categories) and no trustee is ever fully independent from its principal(s).
It is beyond our scope to settle this theoretical dispute here.Rather, we draw on the distinction between traditional and trustee agents, presented above, to evaluate the independence of the EO vis-à-vis the EP, both from the standpoint of institutional design, as well as inquiry activity.In this section, we show that Majone's (2001) second logic of delegation applies in the case of the EO.While on paper the EO benefits from guarantees which are characteristic of trustees, we nonetheless identify two constraints on its independence, both stemming from its close relationship to the EP.The impact of these constraints will be assessed in the following section by looking at how the EO handles complaints against the EP.

The EO as a Trustee Agent
The EO is a trustee appointed by the EP: the EP alone elects the Ombudsman, with no role for the Member States or other European Institutions.To be accepted in the selection process via the EP, a candidate needs the support of at least 40 Members of the European Parliament (MEPs), from at least two EU member states.Those declared admissible are asked to present their priorities in a hearing in front the EP Committee on Petitions.The new Ombudsman is elected by secret ballot and by a majority of the votes cast.In line with what the literature suggests regarding the primacy of reputational and professional credentials in the appointment of trustees, the EP has so far chosen candidates with a history of being national Ombudsmen, and who were independent from the European Institutions and from their respective national governments (Former General Secretary of the EO, personal communication, May 11, 2015).
In line with Art.228 of the Treaty on the Functioning of the European Union (TFEU), the EO's main function is to inquire into and report on instances of maladministration arising from activities of the EU institutions and bodies (only the Court of Justice of the EU, when acting in its judicial capacity, falls outside the EO's mandate).This represents a rather broad mandate, which as suggested by Handke (2010) is typical for trustee delegation (traditional agents are authorised for a narrower remit, often for a single purpose).It also suggests that the EO acts not as a representative of the EP, but on behalf of a distinct beneficiary-namely, in defending 'good administration', the EO 'serves' the European citizens, who have a right to be treated appropriately by the EU institutions.
Importantly, neither the Treaty nor the Statute of the EO define 'maladministration'.This has allowed the EO to actively shape the limits of its own mandate.It has done so by consistently adhering to the position that 'maladministration' refers to unlawful behaviour and errors of legal interpretation, but it also goes beyond this by including failure to respect principles of good administration or fundamental rights (Harden, 2005).This interpretation results in a rather wide remit for the EO (Harden, 2008).
Particularly significant for the EO's status as a trustee is that it enjoys operational independence-meaning, it decides alone regarding the opening of inquiries, either in response to complaints from citizens or residents of the Union, or based on its own initiative.Art 228 TFEU states in no uncertain terms that the EO 'shall be completely independent in the performance of his duties' and furthermore 'shall neither seek nor take instructions from any government, institution, body, office or entity'.
Finally, the EO reports to the EP, insofar as it is required to present it with an annual activity report.However, it is significant that the EP cannot dismiss the EO on its own, but has to request that the European Court of Justice does so, and only on account of the EO's overall functioning or for 'serious misconduct'.Symbolically, the EO gives an oath to perform duties with 'complete independence and impartiality' (European Parliament, 1994) before the Court, not before the EP.In terms of budget, too, the EO is outside of the Parliament's direct control: being one of the seven formal institutions of the EU, its budget represents an independent section of the EU budget and is hence decided jointly by the EP and the Council, based on the Commission's proposal.
In conclusion, insofar as institutional design is concerned, the EO and the EP fit the characteristics of a typical trustee-principal relationship.Namely, while the EP has exclusive prerogatives in appointing the EO, it does not wield any other re-contracting tools.Thus, it cannot dismiss the EO alone (but can only ask the ECJ to consider doing so), it does not directly control its budget, and, although the Statute of the EO is formally an EP decision, it cannot fundamentally re-write the EO's mandate, as the crucial provisions (i.e., its mission to investigate instances of maladministration and the guarantees of operational independence) are inscribed in the Treaties, namely Art 228 TFEU.

Constraints on the Independence of the EO
Although the EO does enjoy a broad mandate, its powers are, as Peters (2005) observes, more modest compared to some of its national counterparts.Most significantly, the EO lacks the power to refer suspected illegalities to the courts, which-among other factors-leads it to cultivate a cooperative style of control.Concretely, in cases where maladministration is found, the EO has no way of obliging the institution concerned to take any redress measures.Instead, what it can do is attempt reconciliation by proposing a 'solution' to which both parties may submit observations.If the solution is accepted, this usually means that the offending institution has admitted wrongdoing, apologised to the complainant, and offered compensation for any damages (Cadeddu, 2004).In more serious cases, the EO may also choose to issue recommendations, where it proposes guidelines for good administrative practice to prevent similar instances of maladministration from occurring in the future.The institution concerned has three months to send a detailed opinion on the draft recommendations, in which it explains whether and how these would be implemented.
Additionally, the EO has at its disposal two other actions which are not explicitly mentioned in the Treaties or its Statute, but which have been shaped by practice over the years.Thus, the option of 'further remarks' allows the EO to make recommendations to the institution concerned even if no maladministration is found.On the other hand, 'critical remarks' are used when the institution cannot be persuaded to rectify the matter, or in situations where maladministration is of such nature that it cannot be rectified.The follow-up measures taken as a response to critical remarks are published annually.Cadeddu (2004) notes that 'critical remarks' are gener-ally used in cases where the maladministration has no general or serious implications.
The 'sharpest' tool in the EO's arsenal is the special report to the EP, which only applies in cases 'of significant public interest' (European Ombudsman, 2016a, p. 4), and where the Ombudsman has issued recommendations, but the offending institution has failed to satisfactorily accept them.The importance of special reports lies in the fact that they must be debated within the EP and as such they receive political attention.Therefore, although the institution under inquiry cannot be obliged to rectify maladministration, it can be directed towards compliance through public 'naming-and-shaming'. Special reports are used very sparsely (the EO has produced only 19 by the end of 2016), precisely because they are considered 'of inestimable value' for the EO's work and regarded as its 'ultimate weapon' (Former General Secretary of the EO, personal communication, May 11, 2015).
To sum up, even though the EO's mandate provides it with a broad remit, its inability to take binding decisions leads the Ombudsman to rely exclusively on the be the guardian of citizens' rights vis-à-vis other European Institutions, and the PC already provided a channel for collecting and addressing citizens' complaints (Magnette, 2003).The unclear demarcation of roles between the EP PC and the EO could potentially result in a situation where the former sees the latter as its adversary and tries to hinder its work by, for instance by cutting resources-or, conversely, it might try to 'subordinate' the Ombudsman to its needs, i.e., to have the EO do dedicated work for the Committee (Former General Secretary of the EO, personal communication, May 11, 2015).In other words, due to the overlap between their responsibilities, a risk exists that the EP would try to control the EO as if it were a traditional agent and not a trustee enjoying discretion over the execution of its own mandate.
These scenarios, however, have not come to pass, as both parties have made efforts to informally define lines of separation between their respective activities.Thus, in time, it became obvious that the petitions which represent the 'bread and butter' of the PC generally concern the Member States' alleged failure to comply with EU law, and hence matters that lie outside the EO's mandate (Former General Secretary of the EO, personal communication, May 11, 2015).For its part, the Ombudsman chose to navigate the ambiguous relationship with the EP by adhering-at least on paper-to the principle that the political work of the EP is outside its mandate, and that the concept of maladministration does not include the work of EP committees. 4 This is apparent in the very first annual report of the Ombudsman: All complaints against decisions of a political rather than an administrative nature are regarded as inadmissible; for example, complaints against the political work of the European Parliament or its organs, such as decisions of the Committee on Petitions.(European Ombudsman, 1996, p. 9, emphasis added) The EO's decision to steer clear of the EP's political work represents a self-imposed constraint.It does not automatically diminish its independence as a trustee, but it does create the risk that the EO might be less assertive when handling cases against the EP that touch on its political role.This is significant because the distinction between political and administrative matters is not always clear-cut, and many of the principles of good administration which the EO defends (e.g., transparency, absence of discrimination) can easily have political implications.
In conclusion, we have identified two constraints on the independence of the EO vis-à-vis the EP.The first constraint is of a general nature: the EO depends on the support of the EP in the framework of special reports to persuade the offending EU institutions to follow its recommendations.To maintain this instrument as a credible threat-and hence use it to its full potentialthe EO needs to cultivate an overall cooperative relationship with the EP.The second constraint is more specific and applies only when the EO deals with complaints against the EP: here, the EO needs to limit its inquiries to the administrative aspects of the case.Together, these two constraints create the risk that, when dealing with cases against the EP, the EO might be reluctant to decide against the Parliament in general, and especially so in cases of inquiries that have implications for or touch upon its political work.

Decisions by the Ombudsman Following Inquiries Against the EP
In this section, we follow up on the observations presented above by analysing the EO's performance in cases where it had to investigate alleged maladministration within the EP.By focusing on cases, we trace whether the two constraints identified in the previous section have had an impact on the EO's independence.
Before we shed light on the cases against the EP, these should be set into the political context.Note that the vast majority of cases handled by the EO concern the European Commission or EU agencies and not the EP itself.To give just one example: inquiries conducted by the EO in 2015 concerned the Commission in 145 cases (or 50.6% of the cases).30 cases (or 11.5%) fell within the realm of EU agencies and only 8% (or 21 cases) concerned the EP (European Ombudsman, 2016c).Concretely, for the period under examination here-January 1, 2004, to May 1, 2015-a total of 124 inquiries were carried out against the Parliament.5 It is noteworthy that in most inquiries against the EP no maladministration was found, or the institution settled the case: this applies to 83 out of the 124 cases reviewed here (in 13 of those instances the Ombudsman did choose to add further remarks).In the remaining cases, where the Ombudsman did find that maladministration had occurred, it issued critical remarks in all but 2 instances.
We used a coding scheme to classify cases according to subject matter.As shown in Table 1 below, they fall into two broad categories: first, cases that relate to the role of the EP as an employer (alleged violations of the Staff Regulations or selection procedures), and secondly, cases pertaining to the relationship of the EP with European citizens.The former category is considerably more numerous than the latter.
To offer an in-depth account of how the Ombudsman deals with cases against the EP, in what follows we discuss several illustrative examples for each of the two main categories above.

Cases Related to Staff Matters
To begin with, many (around 20) of the cases concerning alleged violations of staff rules pertain to promotion issues and specifically to the allocation of the so-called 'merit points'.These complaints have repeatedly given rise to findings of maladministration, and the EO has striven to use some of them as basis for defining best practices of general applicability.In one interesting case, which related to lack of impartiality in the award of merit points for an EP official, the Ombudsman was even requested to consider submitting a special report to the EP.He did not do so, justifying that the case was 'not important enough to merit Parliament's attention in its role as a political body designed to represent EU citizens' (see case 3289/2008/BEH).The EO, however, issued a critical remark urging the Parliament to avoid situations where the person or authority called to decide on staff matters could be perceived as partial (the case at hand concerned the Secretary-General of the Parliament).
The EO has also striven to establish best practices with staff cases that entailed financial implications, and with cases concerning competition and selection procedures, often advocating for the EP to enhance the transparency of the decisions it takes.For instance, in case 3732/2004/GG, further remarks were issued, urging the EP to consider measures whereby persons dealing with tenders would be asked not only to declare any potential conflicts of interests, but also to provide relevant information on any previous dealings with, or activities involving the tenderers.In case 2222/2004/TN, where a participant in a selection procedure was excluded due to lack of professional experience, the EP committed-at the EO's behest-to provide more information and clarify certain requirements regarding future recruitment procedures.
In other inquiries, however, the Ombudsman has limited its intervention to finding a way of reconciling the complainant and the Parliament.For instance, a recent decision concerned an EP official who had been granted derogation from the mobility policy that had been put in place, because of her daughter's severe and irreversible disease.The EP administration decided to disregard this derogation, which led the complainant to approach the Ombudsman.The EO concluded that the EP could 'not lawfully revoke its derogation' (case 118/2013/AN).Although the Parliament refused to acknowledge that its position was unwarranted, it did eventually accept the EO's recommendation to respect the complainant's derogation.

Cases Concerning the Relationship with EU Citizens
Cases concerning how the EP interacts with EU citizenswhen it comes access to documents and transparencyhighlight the tensions inherent in the relationship with the Parliament, as they tend to often touch on its 'political' role, also in the field of external relations.The controversial Anti-Counterfeiting Trade Agreement (ACTA) has brought the Ombudsman two interesting complaints that are illustrative of these dynamics.
Firstly, case 2393/2011/RA concerns the EP's refusal to grant the complainant public access to documents regarding the negotiations leading up to the finalisation of ACTA.While no maladministration was found (the EP had validly invoked exceptions provided for in Regulation 1049/2001), the EO issued further remarks where she highlighted the EP's position as the legislature representing all EU citizens.Accordingly, the Parliament-as a 'political body'-was called upon to intervene with the Commission and the Council to ensure that, in future, the 'very nature of Parliament, which is to openly deliberate on such issues', is not undermined (European Ombudsman, 2011).
Secondly, case 262/2012/OV stemmed from a complaint regarding the refusal of public access to the minutes of meetings of coordinators of several EP Committees relating to the ACTA negotiations (see Abazi & Adriaensen, 2017).After some debate regarding the status of these documents (the EP claimed that there were no separate minutes of meetings of committee coordinators, rather that these were included in the minutes of the committee meetings themselves) the Parliament agreed to the Ombudsman's recommendation that decisions or recommendations adopted by the coordinators would, after their endorsement by the respective EP committee, be included in the committee minutes, and be accessible via the public register as of July 2014 onwards.The Ombudsman expressed her hope that 'for the sake of consistency' the rule would also be applied retrospectively, for the 2009-2014 parliamentary term (case 262/2012/OV).
Another two significant cases-this time concerning the internal rather than the external dimension of EU politics-have raised questions regarding financial transparency in MEPs' activity, thus also touching on the political role of the Parliament.In one case, dating back to 2005, the EO dealt with a journalist's complaint regarding the refusal of Parliament to give public access to details regarding MEP's allowances, allegedly on grounds of data protection (case 3643/2005/(GK)WP).After consulting with the European Data Protection Supervisor-who advocated that the public has a right to be informed about the behaviour of MEPs-the Ombudsman called on the EP to disclose the requested information.The EP refused (again invoking data protection), but announced that it would publish general information on MEPs' allowances and alluded to the possibility of re-assessing the situation in 2009.The Ombudsman consequently issued a critical remark and through media coverage, the case gave rise to a more general debate on MEPs' allowances in the public domain.
In a somewhat related case, the Parliament refused to give access to the list of MEPs participating in the EP's supplementary pension scheme.The Ombudsman made a preliminary finding of maladministration and proposed a friendly solution, which the EP rejected.Significantly, the Parliament as a whole voted down a concrete proposal from its own Budgetary Control Committee to publish this list of names.The Ombudsman thus decided to close the case as the EP's action had made the issue one of 'political responsibility', on which as a legislature it would be accountable to the European electorate, and not the EO.

The Roles of the EO vis-à-vis the EP
The review of cases conducted above brings to the fore two main roles that the EO plays vis-à-vis the EP: 'arbitrator' and 'transparency watchdog'.In the first role, the EO focuses primarily on finding a solution that allows for rec-onciliation between the complainant and the EP.In most cases this leads to a settlement.When playing the 'arbitrator' role, the EO often tries to set 'best practices' concerning good administration within the EP, by giving guidance beyond the specific issue or case at stake.In its second role-'transparency watchdog'-the Ombudsman's focus is on requesting that the EP put certain documents into the public domain, or-failing that-on stimulating public debate and interest regarding EP documents.
Returning to the categories identified in Table 1, the distribution of these two roles is as follows.Firstly, with inquiries concerning violations of staff regulations (with or without financial implications), and the treatment of citizens by the EP as an institution (i.e., sub-categories A.1, A.2, and B.3), the EO performs the role of arbitrator.The cases concerning promotion decisions and the allocation of merit points, discussed briefly in the previous subsection, provide a relevant illustration.Secondly, with cases concerning requests for information and access to documents (sub-category B.1), the EO acts as transparency watchdog, as exemplified in the ACTA inquiries and the case on MEPs' allowances.Finally, in cases dealing with unfair treatment, either in competition and selection procedures (sub-category A.3), or in the award of tenders or grants (sub-category B.2), the EO acts mainly as an arbitrator, but it sometimes chooses to take on the role of transparency watchdog as well, by urging the EP to offer more information as to why certain decisions were taken.Therefore, in these few selected instances, the EO plays both roles simultaneously.This is exemplified in two cases presented above: case 3732/2004/GG and case 2222/2004/TN.
What does this tell us about the independence of the EO vis-à-vis the EP? Firstly, both roles identified above presume that the EO has positioned itself against the (short-term) interests of the Parliament.Thus, when acting as arbitrator, the EO has pushed the EP to acknowledge (and rectify) certain shortcomings in its handling of staff matters and other administrative issues.As a transparency watchdog, the EO has similarly advocated that the EP disclose information which it obviously preferred-at the time-to keep out of the public domain.This is indeed the sort of behaviour that one might expect of a trustee, which executes its mandate in line with relevant professional norms and their own best judgement, and can, therefore, take actions against its principal(s).The cases analysed here show that the first constraint identified in section 2-the EO's reliance on a cooperative relationship with the EP-does not undermine the EO's capacity to act independently.
However-and this is an important point-many of the cases that come to the attention of the EO do not have systemic value, and hence the two roles identified here do not carry equal weight.Generally speaking, inquiries related to staff matters concern the grievances of specific individuals, and although some of these may have a broader relevance, the interested audience is still relatively contained.Cases that pertain to the relation-ship between the EP and European citizens, however, are quite different, because they often touch on the EP's political role as a democratic representative institution.In this context, the EO acting as a 'transparency watchdog' was particularly challenging for the EP, because the information requested by some complainants was of such nature as to empower citizens to hold MEPs accountable.These less numerous but far more significant cases, therefore, represent a tougher test for the independence of the EO, and our analysis has shown that here the Ombudsman has spoken out against the EP (for instance by issuing critical remarks).Thus, we can conclude that the second constraint identified in section 2, regarding cases with a political substance, also does not seem to undermine the EO's capacity to act independently.

The Relationship with the EP in the Context of Own-Initiative Inquiries
The EO's performance in the context of own-initiative inquiries lends further insight into its relationship with the EP.This is because the EO enjoys full discretion in choosing the issues to be investigated in the framework of these inquiries-in other words, this instrument allows the EO to set its own agenda.Thus, own-initiative inquiries are very different from regular cases, which account for most of the EO's activity, but where the space for manoeuvre is confined to the complaints received.By contrast, here the EO is a pro-active actor, which makes own-initiative inquiries particularly significant for assessing its independence.
During the period considered here (2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015), the EO opened 43 own-initiative inquiries.It is telling that most of these concerned either the Commission or European agencies, and none were directed specifically at the EP.However, there have been four 'horizontal' inquiries (i.e., dealing with a specific subject and encompassing several European institutions) which have involved the EP, all of them carried out after 2010.Out of these four horizontal inquiries, case OI/8/2015/JAS concerning the transparency of the so-called trilogues is by far the most consequential and has the potential to stimulate a new level of openness in what has traditionally been a tightly closed-door decision-making process.This is thus a pertinent test case when it comes to role that the EO assumes as a trustee vis a vis the EP.
Trilogues are informal tripartite meetings attended by representatives of the EP, the Council and the Commission.The level of secrecy inherent in trilogues is defended by the institutions as necessary to find a common position within the Ordinary Legislative Procedure in first reading without the pressure and exposure of the regular legislative process (Reh, Héritier, Bressanelli, & Koop, 2013).The Ombudsman's inquiry indicated that trilogues are 'increasingly heralded as the place where the negotiated content of the final legislation text is decided upon', and asked for clarifications regarding: whether and how upcoming trilogues are publicly announced; the documents produced by each institution in the context of trilogues; the public accessibility of these documents (including any requests for public access received in relation to these); and, finally, the language regime of trilogues (case OI/8/2015/JAS).
The trilogues inquiry proved controversial, with all three institutions openly stating that it partly exceeded the Ombudsman's mandate.The EP's response was the least outspoken.It merely indicated that the EO's questions required careful consideration given 'the fact that trilogues are an expression of the more political role of the Parliament', and pointed out that while the handling of requests for information regarding trilogue documents represents an administrative exercise, the organization of the trilogues as such (including the regime of minutes, the languages used etc.) was rather within the legislators' prerogatives and hence outside the EO's mandate (Schulz, 2015).The Commission and Council essentially advanced the same arguments, but the tone of their letters was certainly more outspoken than the EP's.The Council in particular chose to remind the Ombudsman of its own established practice of distinguishing between questions that pertain to the political responsibility of the EU legislators, and those that involve possible maladministration.(Tranholm-Mikkelsen, 2015).
Despite the tense situation, all three institutions eventually responded to the EO's questions and allowed her team to conduct the desired document inspections.On this basis, the EO issued a series of recommendations to make trilogue documentation publicly available. 10The EO further noted that some of this information could be made available while trilogues were ongoing, whereas other aspects might have to wait until after their conclusion.In particular, the EO highlighted the value of transparency for building up citizens' trust in governing institutions: If citizens are to participate effectively in the democratic life of the European Union, by holding their representatives to account, and by voicing their opinion, then they need access to this information….This goes to the heart of EU law-making legitimacy.(European Ombudsman, 2016b) In conclusion, in the concrete case of the trilogues inquiry, the EO has successfully tested the limits of her mandate, by obtaining cooperation in a matter that was considered to bite into what may have previously been taboo territory, namely the political activity of the EP.From this perspective, the trilogues inquiry represents a gamechanger not only regarding transparency practices in the EU but also the relationship between the EO and the EP.
The own initiative inquiry on transparency and public participation in the TTIP negotiations (case OI/11/2014/RA) is of similar magnitude to the trilogues inquiry, in that it represents a bold push for transparency in a domain of activity traditionally characterised by secrecy-in this case, international trade negotiations.Although the EP is not directly targeted in this inquiry, we discuss it here given its strategic significance, and to contrast it with the trilogues inquiry where the Ombudsman did directly challenge the EP (alongside the Commission and the Council).
The background of this case is that the European Commission is currently negotiating a wide-ranging trade and investment partnership agreement with the United States (TTIP).The negotiations have attracted 'unprecedented public interest'.In July 2014, the Ombudsman opened an own-initiative inquiry, which was justified as responding to concerns raised by the EP together with civil society actors.The inquiry ended with the Council and the Commission agreeing to the pro-active publication of a range of relevant documents, including the EU's negotiating directives and opening positions.
What we see in this case is a prime example of how the EP and the EO 'join forces', which is clearly a different dynamic compared to the trilogues inquiry.However, in both inquiries, the EO was equally bold in her push for more transparency.This bodes well for the independence of the institution.Also significant is that in the TTIP case the EO has chosen to stress the role of the EP as the only directly elected EU institution.Thus, she explicitly pointed to the special 'democratic responsibility of elected representatives, at the European and national levels, in scrutinising the negotiations on behalf of their constituents'.While she acknowledged that there may always be circumstances in which elected representatives will have 'privileged access', the direct involvement of citizens is to be encouraged as much as possible (case OI/11/2014/RA).This shows that while in the TTIP inquiry the EO has responded to the concerns of the European legislature, and their interests were aligned, the Ombudsman did not act as an auxiliary organ of the EP, but rather used the opportunity to point out that the Parliament also has its own responsibility in ensuring transparency in international trade negotiations.Before concluding, it should be stressed that EO's decisions in both the trilogues and the TTIP inquiries were preceded by public consultations.In recent years this tool has been used regularly in connection with own-initiative inquiries: seven of the own-initiative inquiries closed since 2010 have incorporated public consultations.By voluntarily using this instrument (no mention of it exists in the EO's statute or in the implementing provisions) the EO makes her office a 'vessel' for the voices of NGOs and citizens, which in turn strengthens her stance vis-à-vis other European Institutions.Thus, when looking at own-initiative inquiries, we can distinguish a third role that the EO plays, mostly in relation to other institutions, but also-albeit exceptionally-vis-àvis the EP, as illustrated with the trilogues inquiry.This role of 'vessel' for civil society concerns is clearly linked to that of 'transparency watchdog', but it also goes be-yond it.The EO does not act on behalf of a case that is brought to it by one or several parties (as in the ACTA case, for instance), but actively asks for input by citizens and NGOs, and thus creates a larger democratic basis for her actions.To be clear, the consultations serve not only to provide the Ombudsman with more relevant information, but also to justify the societal relevance of the own-initiative inquiries, and to shore up the kind of popular support which might strengthen its stance against the institutions that are subject to these inquiries.Therefore, with the role of 'vessel', the EO as a trustee seeks to demonstrate that it is acting on behalf of (representing) civil society concerns, i.e. a third party (a beneficiary) that is distinct from EP.Hence, this role can be seen as signalling the Ombudsman's independence.

Discussion and Concluding Remarks
This article pursued the question of whether the EO acts as an 'independent' institution towards the EP.This is a pertinent question because the two institutions are in a principal-trustee type of relationship.Thus, the EO is appointed exclusively by the EP, with no role to play for other EU institutions or Member States.The EO can be removed by the Court of Justice of the EU only at the request of the EP, inter alia if he is guilty of serious misconduct.On the other hand, however, the EP is among the EU institutions and bodies covered by the EO's mandate.While the political activity of the Parliament, including the work of its committees, remains outside the EO's remit, it does have the power to probe into questions of maladministration within the Parliament and to examine issues such as the refusal of access to documents or decisions taken in competitions and selection procedures.In looking at how the EO acts vis-à-vis the EP, an institution it partly depends upon, this article deals with a 'hard case' for the EO's independence.Thus, the analysis conducted here is relevant for the general performance of the EO as a guardian of good administration in the EU.
The concept of 'independence' is (obviously) not only far-reaching but also difficult to operationalize.We have opted to discern patterns of interaction between the two institutions, by probing into all cases brought forward to the EO against the EP over a time-span of more than 10 years.We then inductively established reoccurring roles the Ombudsman adopts via the European legislature.
Findings show that in the majority of cases no maladministration was found, or the EP settled the case.Thus, the most common role played by the EO is that of 'arbitrator' between the complainant(s) and the EP.The cases where maladministration was found concerned, on one hand, (alleged) violations of staff regulations, and, on the other hand, the Parliament's relation to EU citizens.Here the Ombudsman played its 'arbitrator' role, but also that of 'transparency watchdog' (by ensuring that certain documents are made public, or by bringing certain topics into the public debate).Finally, the two own-initiative inquiries reviewed here-on trilogues and on the TTIP negotiations-reveal a third role assumed by the EO, namely that of 'vessel for civil society concerns'.This is closely allied with the role of 'transparency watchdog' and represents a more recent development, as public consultations have been a feature of (some) own-initiative inquiries only in the past few years.
The roles summarised above demonstrate that the EO acts as an 'independent' actor.From this perspective, the cases where transparency represents a core issue are the most significant, as they tend to concern the EP as an institution at the heart of the 'democratic life of the EU' and as such clearly touch on its political responsibility.
Here the Ombudsman faces clear limits to its mandate, and it has generally confined the inquiries to the legal elements of the case (i.e., the respect for rules governing public access to information).Some of these cases, however, had political repercussions for the Parliament, and thus their salience lies in their 'spin off' effects.On one hand, debates about sensitive issues such as the MEPs' allowances have shifted into the public domain, and pressure for more transparency increased.On the other hand, more political debate within the EP itself was generated, along with concrete action points.
In the context of own-initiative inquiries, the EO has more clearly entered 'political' territory, on issues pertaining to both the internal politics of the EU, as well as its external relations.Both the trilogues and the TTIP inquiries have pushed the boundaries of institutional transparency within the EU in significant ways.In these contexts, the Ombudsman has called upon the EP to exercise its role as a representative institution.
Overall, the role of the EO seems to have evolved over time and it has consistently not shied away from assuming its role as an independent institution also towards the EP.