UvA-DARE ( Digital Academic Repository ) Reasons of Control and Trust : Grounding the Public Need for Transparency in the European Union

Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: http://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.


Introduction
One of the first commitments of Jean-Claude Juncker, during his inauguration as President of the European Commission, was that the European Union would be renewed democratically and improve its relations with the citizens.Better rules on transparency are one of the promised means to attain these democratic aspirations. 1Yet, these promises do not seem to extend to a renewed effort in bringing the reform of the most central measure for transparency forward 2 -the reform of Regulation 1049/2001 3despite the fact that there is an established consensus in the EU that access to documents furthers the democratic nature of the institutions. 4The reform proposal has already been stuck in a legislative deadlock for seven years, 5 without any progress in sight, as the parties appear to be entrenched in their incommensurable positions. 6owever, even though some of the arguments have been repeated ad nauseam, this does not mean that everything has already been said.Instead, we submit that some relevant arguments have not been made and a relevant debate has not taken place.This is the debate about the justification and use of a transparency regime.We submit that there are two kinds of reasons for which transparency is relevant in a democracy.These are reasons of control and reasons of public trust.
Public access to documents is an important policy effort to foster citizens' understanding of EU decision-making and possibly to increase its acceptability.Indeed, access to documents is generally necessary in the relation between citizens and institutions conferred with public authority taking into consideration that any relation depends on information exchange for its establishment and continuation. 7owever, the view on the role of information and the expected results from public access policy vary depending on the different approaches to citizen-authority relations.A predominant approach, despite its intrinsic limitations, has been what we call a control-based approach.According to this approach, public authority is based on citizens' consent and thereupon functions by mechanisms of control.In this paper, we also develop on a somewhat neglected perspective: citizens' political entrustment.The need for citizen trust is already ever present in the political rhetoric in the EU, 8 and pronunciations of a lack of trust are widely used as a measure for the 'democratic deficit' of the EU. 9 It is thus time to take a better look at this aspect of the citizen-authority relation.
The trust approach leads to a new understanding of the role of information in the exercise of public authority.This approach is currently a subject of a number of scientific inquiries, yet its full potential for transparency is underexplored. 10Whereas some versions of the control-based approach incorporate quite a few of the ideas of what we label as trust, we believe that there are paradigmatic differences, which make it analytically useful to separate these two clearly.
The paper focuses on how these approaches manifest themselves in the EU access to documents policy with the underlining goal of broadening ideas and points of evaluation for openness in the EU.Towards this aim the paper offers an account of the two approaches and maps the role that information plays in each respectively.A focus on citizen control relies on different dynamics than an approach that builds on political entrustment.Hence, we additionally point out what evaluation of the access to documents regime either of these approaches dictates.
Our conceptual distinctions provide more innovative thinking on how access to documents could work, therefore possibly helping to erode the stalemate we currently find ourselves in and hopefully providing grounds for new proposals for reform.
The paper proceeds as follows: Section 2 provides an elaboration of the two approaches.Section 3 maps the reasoning of control and trust for information sharing (or not) between citizens and institutions endowed with public authority.Building on this, Section 4 focuses on access to documents in the EU.Section 5 brings together the conclusions.

A control-based approach
The notion of democracy as delegated public authority is typically understood as governing with the consent of the people.From a consent approach to public authority, the recognition of popular sovereignty implies that only the consent of the people themselves can 'morally legitimate the control that governments typically attempt to exercise' and that only their own consent can create an obligation for them to obey the law. 11A distinguishing characteristic of this approach is that public authority, formed by the consent of the people as the holders of the ultimate authority, is maintained through control in order to ensure that consent is not misused or abused.
Consent is thus predominantly carried through political and administrative processes of control.12This control is based on a general suspicion that the interest of citizens would not be upheld fully and truthfully by those exercising power, either due to the personality of officials and/or the circumstances in which they function. 13These mechanisms aim to ensure that citizens' interests are realised and consequently aim to limit discretion in the exercise of public authority as much as possible.The control approach accepts a narrow understanding of discretion as the application of judgment, which implies that the appropriate principal has set guiding standards and principles in advance. 14Control is exerted through electoral mechanisms and through the possible conflict of interests created by the separation of powers. 15While elections are viewed as the primary direct mechanism of democratic control, 16 the separation of powers is a more systemic means to prevent an abuse of power, manifested through appropriate checks and balances. 17The latter is meant to 'discipline' the different branches of power, through conflicting interests combined with the necessity to reach an agreement on a particular policy to the citizens' advantage. 18he main limitation of the approach is that genuine and effective control is often lacking due to the structure and mutual dependencies of control mechanisms.This limitation is not merely a result of practical issues such as a lack of resources or difficulties in organizing checks, regardless of their fundamental importance. 19Rather, the limitation rests on two principal reasons: firstly, due to the inherent reliance of the controller on the controlee, often mechanisms of control resemble what has been described as 'asking for evidence from the suspect' . 20Secondly, the controller is an 'imperfect' actor in the sense that it has organizational limitations, it is also subject to suspicions of (self-serving) action not covered by popular consent, and would in turn have to be controlled, and lastly adopts certain working methods that are similar to the controlee, which could be damaging to its control function. 21Hence, the mechanisms of the approach result in not sufficiently constraining the discretion of the delegated authority.Full consent ensured merely by control mechanisms is more an ideal than a complete enforceable approach.
The multiplicity of actors in the EU, its different layers of decision-making, and the complexity of the division of competences with the Member States, result in a situation where the control mechanisms are in a state of constant readjustment de jure and de facto. 22This is not to imply that there is a lack of control mechanisms (or accountability as a more often used term for the EU) to the effect that additional mechanisms would improve the situation. 23However, it does imply that control is a strenuous act, especially when seeking to preserve both the efficiency and expediency of public action. 24he difficulties with mechanisms of control directly relate to the discussion on Regulation 1049/2001.Such difficulties relate both to its application in the current form, and its revision process.Regarding the former, issues of control particularly arise with exceptions under Article 4(1) relating to security and international relations, issues for which the institutions retain a wider margin of discretion and where many requests for access to documents are denied. 25egarding the reform of Regulation 1049/2001, there seem to be insurmountable differences concerning the scope of possible control via public access.These differences relate to core issues on the implementation of the access regime.One such issue relates to the Commission's proposal to limit the scope to documents specifically relating to policies, activities, and decisions falling under that institution's responsibility as opposed to broadly maintaining that the regulation applies to all documents.Significantly and related to this is the Commission's revision of the wording of what constitutes a 'document' and hence affecting the scope of applicability of the access regime. 26This proposal met with strong opposition from civil society and the European Parliament.The debate and positions seem to be in a deadlock concerning these issues and despite the efforts of the Danish Presidency in 2012 and the European Parliament's motion in 2013 regarding a move forward, it seems that we are left with only the promises of prospects for better transparency rules made by the Commission's President. 27t seems that when control mechanisms reach their limitation, albeit that they include exceptions to access or for more inherent reasons, or when we disagree about their implementation, trust is invoked to compliment the functioning of the framework. 28Treated as a measure of last resort and as the ultimate goal at the same time, trust emerges as a relevant factor and already points at the significance it has for the conceptual construction of public authority.

A trust-based approach
From a historical perspective, the EU 29 may be seen as a trust-building system in that it aimed to establish cooperative relations between European states (and citizens). 30Towards this goal some level of independence from any single European polity was necessary, 31 and hence some space for EU action to ensure the credibility of commitment and to defeat easy capture by any singly polity.This, in turn, required the constituting polities to vest trust in the Union to enable independence.
Very early on, the EU demonstrated an additional intention to directly connect with European citizens through a sui generis relation.As was evident in the case of Costa v ENEL, 32 the EU claimed to create rights and obligations which are directly binding on citizens.However, the EU's connection with European citizens, perhaps due the Union's developing and specific character, is still a challenging one. 33To some extent the struggles arise from the ambivalence from the outset regarding how the EU is supposed to establish and maintain this relation.
Trust lies at the heart of the conceptual construction of public authority because it is a necessary condition for any kind of supportive relationship. 34Thus it is not surprising that trust is often incorporated into some forms of 'consent';35 after all, the choice for public office does not fall on people whom one actively distrusts.36'Consent' , in the political sense, and 'entrustment' cannot meaningfully be defined in a way in which one would exclude the other.Rather, the difference between the two approaches becomes clear when we consider how this 'consent' or 'entrustment' is maintained.
Before further engaging in the argument, it is necessary to first provide some clarification on and a definition of the term 'trust' as a scientific concept.The understanding of trust developed in this paper is a rather uncontroversial one, although it is not feasible to provide a definition which would be truly consensual in all the fields dealing with trust.Our elaboration is supported in some form or another in many contexts. 37rust is the belief in the goodwill of another. 38To trust means to be convinced of the good intentions of another, or at least to be convinced that these intentions can be defined as 'good' in the relevant circumstance, and to act on that conviction.This belief can be taken to also include the tacit acceptance of vulnerability to the possible bad faith of the institution. 39Yet, even though much of the literature about trust focuses on the vulnerability resulting from trust,40 this is certainly not its core feature.Risk and thus vulnerability might constitute the main philosophical puzzle and thus be of particular interest for researchers.Yet, from the perspectives of the parties involved, trust relations are not about vulnerability, instead they are about cooperation, community, and building strength. 41he term 'goodwill' as used above is somewhat peculiar, as it does not need to refer to anything which is specifically morally 'good' .One would thus trust, for example, not only a genuinely 'good person' with high moral qualities.Trust in one's parents is paradigmatic and based on the belief that they generally have one's best interest at heart.Even further, trust is also the belief that the interests of a Member of Parliament to best defend the interests of her constituency and to become re-elected works in favour of the electorate.Here, 'goodwill' is understood as the congruence of interests thereby reducing or eliminating the need to disregard one's own interest on the side of the trustee. 42In both cases, one will have certain indications that these assumptions, i.e. that of the general goodwill of one's parents and that of specialised 'good' interests of the Member of Parliament are correct.Yet, it is an important characteristic of such assumptions that they cannot ultimately be verified.In reality it is sadly possible that parents could actually exploit or abuse their children and that a Member of Parliament could serve a specific interest rather than her constituents' general best interest.
Trust, as belief in goodwill, creates the propensity to follow the directives of the trustee. 43The belief that the trustee has the best interest of the people in mind translates into the entrustment of (some of their) interest.Hence, following the directives of the trustee becomes the logical consequence for the very substantiation of one's own best interest.Through their trust people create authority for the public body, i.e. the normative force of its directives. 44However, trust is usually not universally granted.There is a delimitation to trust: A will believe in the goodwill of B not universally, but with respect to a defined area C. 45 In this way trust not only founds but also delimits public authority. 46Authority built on trust is thus inherently constrained by the limits of that trust; in as far as the exercise of public authority is a cooperative it simply cannot exist outside the bounds of citizens' trust.However, it should be noted that there is no simple description of 'C' in the case of trust in public authority.In such a multifaceted relationship as that between people and public institutions it might be useful to think of the subject of trust as a kind of multidimensional space, which is defined both by the goals and the limiting conditions put on institutions' actions.
Trust puts the trustee under an obligation while at the same time empowering the institution; therefore it creates a two-way relationship.On the one hand, it creates the authority of the institutions and thus their capacity to make the general public act according to a specific interest.On the other hand, trust works so as to place an obligation on these institutions through the normative expectation and moral belief that they will act in the interest of citizens. 47Trust thus creates normative expectations, 48 which bind and constrain institutions.
These limitations do not only describe the 'outer limits' of the space in which the trustee has the power to act, but also work in a 'third dimension' , namely in shaping conditions for the use of this power.A relationship involving institutionalised trust creates a presumptively legitimate space for action. 49here are certain areas of public policy where the need for trust is more pronounced.This is the case especially in areas of more long-term, more progressive (as least untested), and more potentially fatal policies, as for example the security policy tends to be.
Public authority created by trust is of course not an entirely empirical phenomenon, but also takes the form of a political and legal fiction which is needed to provide a narrative for public authority and the institutions' powers.Yet, it should be kept in mind that trust is no more of a fiction than citizen consent is.Due to the practical limitations of elections, consent as a basis for legitimacy could likewise be empirically contested.
The EU is an exemplary case of a regime that requires trust in order to function.Although in line with the stylized and typified view of the logics of public authority given here, we do not actually ask for, or ascertain, the actual trust of citizens in the EU.The EU requires trust to function because the distance and thus the independence of the EU from the citizens is particularly great, in terms of community, culture, and geography as well as in terms of organisational structure.This distance creates a large measure of disconnection between the EU and its citizens, as it necessarily makes direct influence unfeasible.Because the EU is in many ways distant, it needs to rely on citizen trust to validate the justification for its existence.Using the trust approach for the EU-citizens' relation thus enables a new understanding of the specific requirements of European democracy, facilitating the (further) development of a specific EU system of governance, as it is particularly suitable to take account of those factors that set the EU apart from national governments and thus also makes the relation between the EU and its citizens different from most national systems of governance.
Even though the need for trust is ever present in the political rhetoric of the EU, its actual dynamics and requirements do not seem to be reflected by policy makers.However, as will be visible below, control and trust have differing needs for information.

Two approaches, two logics: the role of information for control and trust
Democracy, through the lens of the control approach, relies on the accessibility of information in order to achieve representation and accountability as its essential elements. 50These elements of democracy are grounded in the presumption that the flow of relevant information and its proper handling are not obstructed. 51Keeping control implies having full and detailed information about: who is doing what?when? and how? 52 Information is important to enable citizens to reconstruct the exercise of public authority.Hence, citizens have a 'need to know' regarding all the actual decisions made in order to be able to control the delegated authority and to participate in its exercise by public bodies.
As was mentioned above, the primary mechanisms for citizen oversight are elections in order to ensure that the will of citizens is represented as well as inter-branch accountability arrangements.Access to internal and detailed substantive information and information flows are necessary for these primary mechanisms to deliver the promised results. 53The availability of information is a means of making visible whether the actions of public authorities are representative of the interest of the citizenry, or, to frame the issue in terms of the control approach, openness enables citizens to form an opinion and to formulate the conditions for consent, the adherence to which can then be checked. 54Citizens cannot participate if information on what has been decided is lacking or is only made available after a considerable passage of time. 55Moreover, substantive information on what decisions have been taken is indispensable to the obligations to explain and justify public conduct understood as part of control mechanisms as well as scrutiny over any possible abuse of power. 56n the trust approach, information is necessary to sustain the relation of trust on which deference and authority are premised.Trust exists within a relation, yet not all relations necessarily involve trust.For a relation to do so, the parties need to deem the opposite party trustworthy.Trustworthiness connotes the degree to which a person can be trusted.Where trust is A's belief in the goodwill of B regarding subject C, trustworthiness connotes in how far B invites and stimulates A's faith.Some (personal) characteristics are more conducive to trust than others: honesty, predictability and virtue will generally be deemed to construct trustworthiness, whereas evasiveness, variability, volatility or a deviation from socially accepted values will generally not be conducive to trustworthiness. 57Yet, not only certain characteristics construct trustworthiness.As explained in the examples of trust above, goodwill can also mean a conformance of goals and motives.Specific trustworthiness can then be constructed by establishing this conformity.In sum, to establish this trustworthiness, we need information about the characteristics of the prospective trustee.
The information which we need in order to establish trustworthiness is based not necessarily on the actual behaviour or actual facts of each case, as its function is not to reconstruct specific decisions which have been made.Empirical research has shown that trustworthiness is established on the basis of inferences about the motives of the prospective trustees. 58To retain trust, the trusting party does not need to be informed about the specific calculations and facts in every case, but rather about the motives informing these calculations.Instead of knowledge about the specific situation, knowledge about the actors and their characteristics and values is primarily relevant. 59Indeed, many trust relations involve a situation where the trusting party will have problems in assessing or even understanding the material substance of a trustee's decision. 60As an example of public policy, on average a citizen would not be able to assess the merits of equipping the military with one weapons system instead of another, but she would certainly be in a position to assert the personal and motivational level.
The need for information is not absent in the trust approach, and this need is not necessarily smaller either. 61Rather, we argue that it is different.A call for information in the context of trust is useful insofar as the information is used as a springboard for the 'final leap of faith' that is connoted by the term 'trust' . 62here it is sought to replace this 'leap' , it will very likely be detrimental to trust, 63 as facts cannot found a belief.Especially where material detailed information is not intelligible to citizens, in the best case it will have no impact, and in the worst case it will create confusion and disaffection.
Whereas control relies on full and detailed substantive information about specific decisions, it is a mark of trust for citizens to primarily focus on exercising oversight on the trustee, not strictly on her decisions.This implies that in a trust approach, it is not essential for the trust of citizens to reconstruct and reproduce decisions made by the institutions and consequently the information sought does not necessarily refer to them.Instead, there is a need to support the trust relationship.For this reason, trust requires information on what drives the decision maker.Information has a different function in this approach: to establish the conditions upon which the system supports the two-way relation between citizens and institutions.
Furthermore, both control and trust rely on 'relevant' information and not information per se.For example, a control approach would not be considered satisfactory if, despite the fact that detailed information is shared, such information would not necessarily be relevant for a specific aspect that citizens or oversight bodies aim to control.Similarly, trustworthiness concerning the motives of decision-makers would be established if there is disclosure of relevant information and not selectivity in the processes. 64his brings us to a key aspect of both approaches: the way in which they deal with limits on the provision of information.

Lack of information: challenged control, invincible trust?
Effectively exercising public authority also at times necessitates the withholding of information from the public.Some legitimate democratic policies and processes, typically security policies and negotiations, require some degree of limited access to information in order to be realized effectively or at all. 65Generally in democratic societies a limitation on the information provided is allowed in order to ensure the goals set, but this is nevertheless considered to be an exception to the rule of openness. 66imitations on access to information pose a challenge to the democratic mechanisms of control.The ability of the 'government' to manipulate information through limiting the disclosure of information distorts the key principles of representation, participation, and accountability as rudiments of democracy and threatens a certain sense of arbitrariness, which is against the idea of the rule of law. 67The extent of the limitation on openness that could be considered legitimate, and thus acceptable, in democracies depends on the extent to which democratic mechanisms of control over public authority and their reliance upon openness are not obstructed.Controlling the limits of legitimate non-disclosure is an approach towards reaching an acceptable framework in democratic processes and an instrument which aims to disentangle legitimate from illegitimate reasons, thereby preserving non-disclosure in the case of the former and exposing the information in the case of the latter. 68From a control perspective, the necessity to conceal information (even if only partially and for a limited period of time) gives rise to the challenge of distinguishing genuine reasons for keeping information undisclosed, as in the example of military plans or a specific negotiation position, from illegitimate and unnecessary bureaucratic or political reasons, such as avoiding embarrassment or concealing corruption. 69The objective of control over non-disclosure is to limit permissible non-disclosure to those areas in which confidentiality is particularly relevant.Further, there are procedural mechanisms seeking to prevent any overuse of non-disclosure exceptions.These mechanisms can be found internally within the executive and also in inter-branch mechanisms. 70n other words, information is relevant for the control approach primarily for the functioning of the control mechanisms as well as for an effective administration with the aim of best ensuring citizens' interests.For certain policies, public institutions are given a right to withhold certain sensitive information for a limited period of time.In such cases, citizen control is obstructed.Hence, in turn the prerogative for non-disclosure itself is subject to control.Nevertheless, the discretion to limit access to information poses a serious challenge to the control approach because it creates a structural dependency between the controller and the controlee.At this junction, the control approach's crucial reliance on information and its vulnerability to the limitation of information surfaces.
From a trust perspective, one of the relevant characteristics of a trusting relationship between citizens and public institutions is the possibility for trust to exist without having full and detailed access to the specific decisions. 71In the case of institutional secrecy this means that in a trust-based relation it will not immediately deteriorate simply because citizens might not know the information that a public institution keeps secret.For example, information about operations, market participants or technologies is not necessarily relevant to support trust.This does not imply that the trust approach automatically justifies secrecy on a higher level than the control approach, but it means that for the trust perspective the lack of this kind of information does not immediately threaten trust.
Yet, people will need to know why the institution is keeping information secret and what procedure it applies in distinguishing secret from open information in order to gauge the motivational context.The latter 'need to know' corresponds with the meta-information needed to establish trustworthiness.It should be noted that this requirement is not in itself less demanding than access to substantive and specific information.Instead, the different direction that this inquiry takes implies substantiating the democratic need for authority of public institutions by releasing the motivational information that supports trust.
What people will need to know is that the officials taking decisions about national security do so in their interest and towards a goal that they substantively agree with.Mechanisms dealing with secrecy in a trust approach thus need to focus on knowledge about the motives of officials and institutions and about the interests and ties that these institutions have.This information is not necessarily general or abstract, but should contain a significant level of detail.The difference with the control approach here lies not in trust being less burdensome to maintain, but in that it focuses on relational, rather than factual information.
While both approaches give information a significant role, they argue for different kinds of information for different purposes.As a logical consequence they also judge differently on which information institutions can legitimately withhold.These different treatments and uses of information play a role when evaluating (but also when designing or reforming) specific access to information policies.In order to demonstrate their implications, below we apply both views to the main instrument of the information regime of the EU.

Public access to information in the EU
In the European Union public access to documents is stipulated in Article 15(3) of the Treaty on the Functioning of the EU and further specified in Regulation 1049/2001. 72Furthermore, it is upheld as a fundamental right in the EU's legal order, as laid down in Article 42 of the Charter of Fundamental Rights of the EU.The scope of the right to access encompasses both legislative and administrative procedures and accessibility is required to be as wide as possible.In practical terms this means that any documents requested should in principle be provided in full and when limitations due to public or private interest occur they should be interpreted and implemented narrowly. 73he right to public access to documents, as is legally stipulated in the EU, must be judged differently depending on whether it is evaluated from a control or from a trust perspective.To recall, from a control perspective, the EU's regime should enable full access to any information which is necessary to reconstruct the decision taken.From a trust approach, the access to documents regime would be appropriate if the information which is necessary to establish the EU's trustworthiness and to make citizens understand how public authority is exercised can be found in documents produced or used by the European institutions.The salient question is thus whether the access to documents regime is able to fulfil these requirements.
Looking at the first requirement, when access to documents is granted under Regulation 1049/2001, this in principle enables the citizen to reconstruct the decision-making process.While such an assessment must be qualified by practical arguments about the capacity and resources of average citizens, it is a necessary conclusion from the control perspective.Significantly, access to documents translates into access to 'raw data' .This means that the information is 'real' and has not been tampered with, a characteristic which is important from a control perspective.Practical limitations are to a certain extent remedied by the role of intermediaries, which could be both NGOs or MPs, even though this situation gives rise to another set of questions, issues such as the interest of the actor in question, but most importantly it forgoes the main idea of access to documents as a mechanism of direct control.
From a trust perspective, the citizen does not necessarily gain trust by the mere fact that she has obtained the documents in question.These documents, which are often prepared for internal discussions and decision-making, do not necessarily contain information on the motivations of the decision-maker.Indeed, it is rather unlikely that an administrative organisation would debate its motivation internally.The citizen is left in a position of having to decipher the intentions of officials and their motives regarding a course of action taken since the documents per se do not include such statements.In this regard it is very relevant that Regulation 1049/2001 does not oblige the EU institutions to create new documents in response to questions from citizens; instead requests may only be made for exiting documents.
Besides being cumbersome, extracts from documents are prone to errors and thus misunderstandings.Information on facts seems more liable to misinterpretation, 74 and unpleasant surprises, 75 thereby creating backlashes.When facts are misinterpreted, they lead to incorrect inferences concerning the underlying motives.Additionally, the data provided in these documents might also simply not allow for any inferences about the information needed to establish trustworthiness.
Taking the above arguments into account, it is not surprising that it is not the citizen who seeks access to documents from the EU, but rather academics or special interest representatives. 76This serves as evidence of the problems involved in the system.It does not appear too speculative to claim that this can easily be due to the overly heavy burden which the system places on citizens who want to obtain information which is relevant to them.
Secondly, the most pressing characteristic of access to documents is the many limitations to which it is subject due to other public or private interests. 77From a control perspective, the limitations on access, due to for example the internal functioning of the administration or international negotiations, block the flow of information.For example, for issues of security and decisions as significant as whether there should be military intervention in a conflict in a neighbouring state or public security regarding the legitimate and acceptable scope of state surveillance, informed public opinion and debate are necessary in order to form consent, regardless of whether the consent will lead to negative or positive answers to these questions.These issues of security and personal liberty touch upon the most sensitive aspects of citizens' lives and hence require a high threshold of consent for their possible limitation. 78Yet, the examples given are also situations where institutions do not want the information to fall into in the wrong hands, and where the success of certain actions depends on confidentiality. 79The imperfection of the access mechanism is precisely a reflection of the limitation of the control approach more broadly.Namely, the mechanisms are structurally bound and limited by the very discretion they aim to control.Consequently, control in practice needs a complementary approach, one that requires a different understanding of information.
From the trust perspective, the exceptions can be divided, for the purpose of this inquiry, into three categories: those of having as their motive the protection of polity (for example, national security), the protection of the government as an organisation (for example, the so-called 'space to think' exception), and those protecting certain persons against other persons.This differentiation follows from the trust and discretion approach in that it puts the motives at the centre.To recall, trust is the belief in the goodwill of another.It is built on the expectation of the existence of this goodwill.Where an institution uses the powers conveyed in trust, it also needs to act with this goodwill.Goodwill can be defined as having the goal or motive of acting in the trusting persons' (best) interest.From a trust perspective, what motivates the exceptions to public access is thus relevant.
Generally, the protection of polity is a justifiable motive for such an exception, because it is commensurate with the purpose of the public institution.Thus, secrecy for the purpose of security is, from the trust perspective, in principle normatively justified.In the same light, secrecy as confidentiality, as protecting certain persons from the general public or from competitors is in line with the trust approach, as it protects the members of the polity as individuals.Where the trust relation involves a group of people, it is clear that staying true to the trust of one party might mean treating another party less favourably.So long as this is done in fairness, there is, in principle, nothing apprehensible. 80However, the above includes the qualification that the lack of objection is principled: whether the justification for the information withheld and the explanation of the motivations of the public authority are indeed conducive to its trustworthiness is thereby not yet given.Instead this depends on the substantiation of the information requirements of the trust approach.
The most problematic case is that of the organisation protecting itself, the ground for rejection which is most often invoked by most Commission and Council institutions. 81The involvement of an organisation's own interest is liable to taint an act and to prejudice the trust relationship through a suspicion of betrayal. 82While these latter grounds are most problematic for the trust perspective, they are not impossible to defend and implement.What would be necessary is a convincing justification of how the protection of organisational self-interest is nevertheless a realisation of goodwill towards the citizens.
To sum up, the form of Regulation 1049/2001 (providing access to documents as a mechanism of transparency), on the one hand is in line with the theoretical considerations of the control approach; more so than the trust approach.The criticism from a control view is here mainly of a practical nature, rather than principles.On the other hand, the exceptions and limitations are more problematic from a control approach than from a trust approach.While in either case practical problems remain or would be likely to emerge, these should be subject to a different discussion than the one we put forward which concerns the appropriateness of information-sharing instruments at a more fundamental level.What we would want to draw from the above discussion is that each perspective provides relevant views about the working and appropriateness of the access to information regime.In being under-used, the trust perspective in particular could serve to bring the discussion further.
Employing this perspective could bring movement into the debate, because the aspects in access to information which are relevant for citizen trust have so far not been widely discussed.It still appears to be likely that an agreement can be reached on how to make the transparency regime more sensitive to the needs of motivational information, even where it does not seem possible to widen the scope of access to existing documents.Thus, much of the above discussion of Regulation 1049/2001 from a trust perspective has concluded that the final evaluation depends on further information being given, information that is not yet part of the access to documents regime.
Indeed, motivational information can be provided independently of documents.It can accompany documents, but, more significantly, it can also be contained in the notice refusing access to a specific document.The latter possibility is more significant precisely because it opens up a potential route for transparency, even where access to a document has been denied.Thus, while it would be far from our task to claim that a denial of access to a document would in itself engender trustworthiness, it is also important for us that such a denial does not in itself endanger trustworthiness.Instead, from the trust perspective it is prima facie neutral.
Nevertheless, the notice denying access establishes a route of communication along which the information that is necessary to establish trustworthiness can be conveyed to an interested and engaged citizen.The fact that a notice refusing access takes a much more personal form, as well as the fact that it should contain reasons and explanations, 83 gives it the potential to establish the trustworthiness of the author.The specific reasoning and explanations contained in a denial give the institution the opportunity to explain how its actions constitute the actions of a confidential and reliable partner, and how it is consistent with, or even a proof of, the loyalties of the institution towards citizens.
The Court has affirmed that a denial of access to documents needs to include specific reasons for denying access for each individual request. 84From here it is only a small step further to also include more information about the motivations of the institutions in the process to which the document refers.
Instead of only treating a denial of access as a 'zero option' of non-access and no control , a reform of the transparency regime should recognise the potential of direct communication with citizens to provide grounds for citizen trust.

Conclusion: the potential of trust
In an era when most information is one click away, the notion and practice of public access to information seems commonplace.However, the bottomless pit of information is not of much use if we are not quite certain what it is that we require to know and most importantly, why do we want to know.In this article we have tried to answer the latter question both on a theoretical level and by focusing on the EU practice of Regulation 1049/2001 on public access to documents.On a theoretical level, we explained that there are different reasons for public access policy which are based on the (broader) idea of understanding the relation between citizens and public authority.We identified two main bases for such a relation -control or trust -which thereby establish different requirements for accessibility to information.We want to make clear that we do not think that the EU can conclusively be described by either of these approaches.Neither normatively nor descriptively is either a paradigm of control or a paradigm of entrustment able to explain the EU institutional and procedural framework.Indeed, we are convinced that despite their fundamental differences, a successful EU unites both.
On the theoretical level it seems as if trust and control are very different.So different in fact that one would have to choose for one perspective.Indeed, control proponents will likely be very suspicious of trust-based argumentation, and entrustment proponents might find that a control-based system is too rigid to provide for flexible and cooperative governance.However, it is also notable that at the practical level the two perspectives do not seem to be mutually exclusive -at least in the present case.In our view, it is possible for one and the same access to information policy to cater for both the needs of control and trust to a significant extent.Of course, it cannot be excluded that for some cases, one rationale would figure more prominently whereas the other would play a supplementary role, just as the control of and trust in institutions both have their place and time.The aim is not to set the goal of trust against the goal of control.Through their common need for information, up to a certain extent the two approaches should be seen as mutually reinforcing.Information needed as a basis for trust would facilitate some minimal control and substantive control would deliver some minimal trustworthiness information.Additionally, substantive information on decision-making is not necessarily harmful to citizens' trust and neither is information on the decision-makers a motivation for being harmful to citizens' control.Consequently, public information can cater for both attitudes of control and of trust.In this respect, sharing information could fulfil its function to bridge the gap between citizens and institutions by addressing both control and trust.
Especially in cases where suspicions cannot be quelled through open and transparent access to substantive information, a trust approach can point to what can be done to alleviate citizens' distrust and restore citizens' support in public policy.One could argue that where there is real control, there is no need for trust mechanisms; control can here serve as a sufficient basis for democratic authority.But where control is found wanting or is not possible, there is a need for mechanisms of trust to sustain deference to public authority.The trust approach thus reveals an opportunity to move the debate forward and to overcome the legislative stalemate by thinking along new lines and to argue with a different kind of reason.¶ Commission, [2010] ECR II-2245, Para.84.Limits apply so as not to undermine the protected interests under the exceptions provided in Article 4. See Case T-105/95, WWF UK v Commission, [1997] ECR II-313, Para.65;Case C-266/05, P Sisón, [2007] ECR I-01233, Paras.81-82.